Essays in Analytic Theology: Volume 2 (Oxford Studies in Analytic Theology) 9780198866817, 019886681X

Table of contents :
Cover
Essays in Analytic Theology
Copyright
Dedication
Acknowledgements
Contents
Introduction
1. Incarnation, Sin, and Atonement
2. Evil, Divine Hiddenness, and Worship
References
Part I: Incarnation, Sin, and Atonement
Chapter 1: The Metaphysics of Original Sin
1. Theories of Original Sin
1.1 The Nature of Our Corruption
1.2 For What Are We Guilty?
1.3 AG Theories
1.4 PG Theories
2. Jonathan Edwards and the Doctrine of Original Guilt
2.1 Edwards’s Theory of Imputation
2.2 Theories of Persistence
2.3 The Organic Whole Theory
2.4 The Fission Theory
2.5 Conclusion
3. Original Sin and Conditional Transworld Depravity
4. Conclusion
References
Chapter 2: Hylomorphism and the Incarnation
1. The Basic Framework and the Neo-Aristotelian Theory
2. Natures as Fundamental Powers
3. Natures as Uniting Other Powers
4. Natures, ‘Individuators’, and ‘Distinguishing Properties’
5. Further Terminology
6. The Trinity
7. The Incarnation
References
Chapter 3: The Ill-Made Knight and the Stain on the Soul
1.
2.
3.
4.
References
Part II: Evil, Divine Hiddenness, and Worship
Chapter 4: In Defence of Sceptical Theism: A Reply to Almeida and Oppy [with Michael Bergmann]
1.
2.
3.
4.
References
Chapter 5: Wright on Theodicy: Reflections on Evil and the Justice of God
1. Wright against Theodicy
2. Theodicy and Biblical Narrative
3. Conclusion
References
Chapter 6: Sceptical Theism and the ‘Too-Much-Scepticism’ Objection
1. What is Sceptical Theism?
2. The Too-Much-Scepticism Objection
3. Global Scepticism
4. Scepticism about Value
5. Scepticism about (Other) Knowledge of God
6. Moral Paralysis
References
Chapter 7: Narrative, Liturgy, and the Hiddenness of God
1. Divine Hiddenness and Divine Silence
2. Divine Silence and Divine Concern
3. Narrative, Liturgy, and the Presence of God
4. Postscript
References
Chapter 8: Hiddenness and Transcendence
1.
2.
3.
4. Postscript
References
Chapter 9: Protest, Worship, and the Deformation of Prayer
1. Divine Validation of Protest
2. The Duty to Worship and the Deformation of Prayer
3. Anger, Worship, and Prayer
4. Damaged Prayer?
References
Index

Citation preview

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OXFORD STUDIES IN ANALYTIC THEOLOGY Series Editors

MICHAEL C. REA and OLIVER D. CRISP

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OXFORD STUDIES IN ANALYTIC THEOLOGY Analytic Theology utilizes the tools and methods of contemporary analytic philosophy for the purposes of constructive Christian theology, paying attention to the Christian tradition and development of doctrine. This innovative series of studies showcases high quality, cutting edge research in this area, in monographs and symposia.   : Metaphysics and the Tri-Personal God William Hasker

The Theological Project of Modernism Faith and the Conditions of Mineness Kevin W. Hector

The End of the Timeless God R. T. Mullins

Ritualized Faith Essays on the Philosophy of Liturgy Terence Cuneo

In Defense of Conciliar Christology A Philosophical Essay Timothy Pawl

Atonement Eleonore Stump

Humility and Human Flourishing A Study in Analytic Moral Theology Michael W. Austin

Humility, Pride, and Christian Virtue Theory Kent Dunnington

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

Essays in Analytic Theology Volume 2 MICHAEL C. REA

1

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Michael C. Rea 2021 The moral rights of the author have been asserted First Edition published in 2021 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2020944362 ISBN 978–0–19–886681–7 DOI: 10.1093/oso/9780198866817.003.0001 Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

To Matthias Rea

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

Acknowledgements With the exception of Chapter 9 and the postscripts to Chapters 7 and 8, all of the essays in this volume have been previously published. I have made no changes to the previously published material except to correct a few minor errors, add an occasional editorial note, and make some formatting changes for the sake of uniformity. I am grateful to the following publishers for permission to reprint the material listed here. ‘The Metaphysics of Original Sin’, pp. 319–56 in Persons: Human and Divine, edited by Dean Zimmerman and Peter van Inwagen (Oxford: Oxford University Press, 2007). Reprinted by permission of Oxford University Press. ‘Hylomorphism and the Incarnation’, pp. 134–52 in The Metaphysics of the Incarnation, edited by Anna Marmodoro and Jonathan Hill (Oxford: Oxford University Press, 2011). Reprinted by permission of Oxford University Press. ‘The Ill-Made Knight and the Stain on the Soul’, European Journal for Philosophy of Religion 11 (2019): 1–18. ‘In Defence of Sceptical Theism: A Reply to Almeida and Oppy’ (with Michael Bergmann), Australasian Journal of Philosophy 83 (2005): 241–51. ‘Wright on Theodicy: Reflections on Evil and the Justice of God’, Philosophia Christi 10 (2008): 461–70. Philosophia Christi is the journal of the Evangelical Philosophical Society (http://epsociety.org). ‘Skeptical Theism and the “Too-Much-Skepticism” Objections’, pp. 482–506 in The Blackwell Companion to the Problem of Evil, edited by Justin McBrayer and Daniel Howard-Snyder (Malden, MA: Wiley-Blackwell, 2014). ‘Narrative, Liturgy, and the Hiddenness of God’, pp. 76–96 in Metaphysics and God: Essays in Honor of Eleonore Stump, edited by Kevin Timpe (New York: Routledge, 2009). ‘Hiddenness and Transcendence’, pp. 210–25 in A. Green and E. Stump (eds), Hidden Divinity and Religious Belief (Cambridge: Cambridge University Press, 2015). Copyright Cambridge University Press. Reprinted with permission. I am also grateful to Michael Bergmann for his permission to reprint our ‘In Defence of Sceptical Theism: A Reply to Almeida and Oppy’, to Oliver Crisp, Hud Hudson, and the anonymous referees for Oxford University Press for very helpful comments on the introductions and postscripts, and to Callie Phillips

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

viii



for preparing the index. Finally, I would like to thank Oliver Crisp for encouraging me to publish these essays here and in the companion volume, for our ongoing collaboration on all things analytic-theological, and, most of all, for our many years of friendship. It is in gratitude for all of this that I dedicated the first volume to him. This second volume I dedicate to my youngest son, Matthias.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

Contents Introduction

1

PART I. INCARNATION, SIN, AND ATONEMENT 1. The Metaphysics of Original Sin

13

2. Hylomorphism and the Incarnation

53

3. The Ill-Made Knight and the Stain on the Soul

73

PART II. EVIL, DIVINE HIDDENNESS, AND WORSHIP 4. In Defence of Sceptical Theism: A Reply to Almeida and Oppy [with Michael Bergmann]

91

5. Wright on Theodicy: Reflections on Evil and the Justice of God

103

6. Sceptical Theism and the ‘Too-Much-Scepticism’ Objection

113

7. Narrative, Liturgy, and the Hiddenness of God

143

8. Hiddenness and Transcendence

172

9. Protest, Worship, and the Deformation of Prayer

193

Index

211

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

Introduction This book is the second of two volumes collecting together the most substantial work in analytic theology that I have completed between 2003 and 2019. Volume I contains chapters focused, broadly speaking, on the nature of God; this second volume contains chapters focused more on doctrines about humanity, the human condition, and how human beings relate to God. The chapters in Part I deal with the doctrines of the incarnation, original sin, and atonement; those in Part II discuss the problem of evil, the problem of divine hiddenness, and a theological problem that arises in connection with the idea that God not only tolerates but validates a response of angry protest in the face of these problems. The section headings of this introduction match the part divisions of the book; but, as in Volume I’s introduction, the aim here is not to summarize the chapters included in each section’s corresponding part, but rather to supplement them with a more general discussion of some of my past and current thinking on the various loci covered by the chapters in the volume.¹

1. Incarnation, Sin, and Atonement The Westminster Shorter Catechism opens with the question, ‘What is the chief end of Man?’ and then offers the following answer: ‘To glorify God and enjoy him forever.’ Gender-exclusive language notwithstanding, this captures the heart of Christian teaching about the human telos. It implies that we cannot flourish outside of a relationship with God, that the purpose for which we are created is wholly oriented towards God, that we are capable of living forever, and that our purpose includes eternal enjoyment of God. But Christianity also teaches that human beings are not capable on their own of coming anywhere close to realizing their telos. They need divine help, owing to a further (contingent) fact about human nature. In short, human nature has been damaged, or become corrupted. (Different traditions within Christianity have different views both about whether it is better to say that human nature is ‘wounded’ or ‘corrupted’ and also about the precise nature and extent of the ¹ For some of the content of this introduction I have drawn on part of another essay not included here—one which I characterized at the time I wrote it as ‘a miniature sketch of a partial systematic theology’ (Rea 2017). I am grateful to the publisher for permission to reuse this material.

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0001

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

2

   :  

wound or corruption.² Henceforth I will simply speak of ‘damage’ in an effort to straddle the lines between these different views.) This damage is supposed to be something we are born with, a result somehow of the first human sin, and a condition that makes it very likely—many would say inevitable—that we fall into further sin. These claims constitute the main part of the doctrine of original sin.³ The other part, more controversial, is the doctrine of original guilt, which implies that the damage to our nature is sufficient, even in the absence of voluntary sin in our earthly lives, to preclude us from eternal life with God unless it is somehow remedied.⁴ Both parts of this doctrine are puzzling; both parts are theologically important. Why should the first human sin (assuming there was a single, definite event that constituted the first sin) result in universal damage? Why should such damage present in us from birth pose an obstacle to our relationship with God even in the absence of voluntary sin on our part? There are no easy answers to these questions, as the extended discussion of alternatives in Chapter 1 makes clear. But neither is it easy simply to abandon the doctrine. Original sin (taken to include original guilt) is supposed to explain two facts about the human condition. First, sin is universal. Everyone is disposed to sin, and everyone who lives long enough to become a full-blown moral agent does sin. Second, everyone needs salvation. The supposition that there was a first sin that damaged human nature explains the universality of sin without implying that God created us in a damaged condition, or that it is sheer coincidence that we are all damaged. The supposition that it is human nature that got damaged, and damaged in such a way as to separate us from God, explains why every human being needs salvation. I think that the two facts just mentioned can be accepted independently of the doctrine of original sin, simply on the strength of the scriptural evidence that supports them. I also think that the doctrine itself can be reasonably accepted as an article of faith, even in the absence of answers to the challenging questions mentioned above. Still, it would be nice to have at least some idea of how the first sin might have resulted in the consequences that the doctrine affirms.⁵ I do not have a full theory to offer; but I can take some initial steps in that direction.

² For example, the Catechism of the Catholic Church emphasizes the ‘wound’ to our nature, and explicitly denies that it is totally corrupted (see 1.7.3, esp. para. 405), whereas theologians in the Reformed tradition typically insist that original sin is a form of corruption. ³ Or ‘ancestral sin’ in Eastern Christianity; but my characterization more closely follows Western lines of thought. ⁴ In the confessions of the Reformed tradition, the doctrine of original guilt is normally taken to include the claim that we are guilty for the corruption of our nature, or that God blames us for it. It is also commonly said that God is angry with us for it. I do not reject these statements outright, but I think that they are apt to mislead; and I think that the ‘divine wrath’ claims are particularly unfortunate in this regard. Since I cannot possibly hope to do them justice in the short space allotted here, I simply set them aside. ⁵ We might also ask how belief in a ‘first sin’ or an ‘historical Adam’ could be reconciled with evolutionary theory. This is a matter of interesting controversy, and several proposals strike me as promising; but I shall not pursue this issue further here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi



3

Suppose that it is part of the human design plan for us to exist in a kind of emotional and psychological union with God (analogous to but deeper even than the sort of union that takes place between close friends or spouses). Under ‘normal’ circumstances, we would experience this union in rudimentary form from the first moment of our existence as psychological beings, and our experience of it would grow stronger and deeper throughout our lives as our cognitive capacities develop and mature. Furthermore, such union is absolutely necessary for proper moral and psychological development. Being apart from this relationship is like deep sea diving without proper equipment: we become damaged, distorted, and subject to further moral and psychological deterioration for as long as we are without it. Suppose that the first human person(s) came into the world already united with God in the requisite way, but that one consequence of the first sin was that God partially withdrew God’s presence from creation, so that the union for which we were designed was no longer readily available—it could be had only dimly in this life and only with special divine help and as a result of actively seeking God. This is a story according to which the first sin does indeed result in universal damage. But is it also a story on which the first sin results in damage to human nature? Despite the prevalence in the tradition of the idea that human nature itself is altered by the Fall, it is not immediately obvious how the withdrawal of God’s presence (apparently an extrinsic change) could be or result in such a change; and there is, unfortunately, a dearth of explicit commentary on this issue in the tradition. But perhaps we might (speculatively) flesh out the story as follows. Let us simply deny that the withdrawal of divine presence was an extrinsic change: it was, before the Fall, part of human nature itself to be, in a certain way, imbued with the divine presence. Pre-Fall and post-Fall humanity both count as ‘versions’ of humanity by virtue of their deep resemblance to one another. But, unlike preFall humans, post-Fall human beings are by nature bereft of an intrinsic feature that they desperately need in order to achieve their telos, and they are damaged and moving toward further ruin from the first moment of life.⁶ On the supposition that living in a world apart from the (relevant mode of) divine presence results in distinctively moral damage so utterly devastating as to pervade our entire psyche, it is even appropriate to say, with Calvin, that one result of the Fall is the total depravity of the human race. The story just given explains the universality of sin. Sin is universal because humans can avoid it only by being fully in the presence of God, and the first sin

⁶ How could being imbued in a certain way with the divine presence be intrinsic? Simple: Intrinsic properties are ones that cannot differ between duplicates or, following Langton and Lewis (1998), ones that are independent of accompaniment by contingent beings. It is not implausible that two human beings who differ with respect to the relevant mode of divine presence would not count as duplicates; for one, but not the other, would be bound for a life of increasing sinfulness absent divine rescue. Likewise, if (as I think) God is a necessary being, it is easy to see how being imbued with God’s presence in a certain way would be independent of accompaniment by other contingent beings.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

4

   :  

resulted in the partial withdrawal of God’s presence. It also provides the resources to explain why God’s plan of salvation is relevant to everyone. Standard Christian soteriology maintains that the work of Christ makes us fit for God’s presence and contributes to our sanctification. We might suppose, then, that even infants who die without voluntarily sinning require (as a result of their being conceived and born in the conditions just described) divine help to become fit for the presence of God, without which help they would remain damaged in their afterlife and would experience precisely the sort of moral deterioration and ruin that characterize natural human life. The human condition, then, fundamentally includes sin and misery. The idea is not that we are constantly committing sin and feeling miserable, never experiencing pleasure, never displaying virtues, always displaying vices, and so on. Rather, the idea is this: First, our lives are characterized by sin, in that we are unable without divine assistance to order our desires in the right way, and doing the right thing involves moral struggle against strong and pervasive self-oriented inclinations. Second, this situation is one in which we are ‘objectively miserable’, not happy in the Aristotelian sense, failing to flourish, and subject as a result to feeling miserable far more often than we should expect in a world created by a loving God. The Christian gospel, however—the good news—is that this tale of sin and misery is not the whole story about the human condition. The rest of the story is that, despite our sin and despite how things may look, God still loves us, desires union with us and wants us to flourish, and has therefore intervened dramatically in human history in order to save us from our condition. The essential details of this propitious intervention, sans explanatory comments, are as follows. The second person of the Trinity became human and lived among us as the man, Jesus of Nazareth. He lived a perfectly sinless life, and fulfilled the human telos, showing us in the process both what God the Father is like and what human beings were meant to be like. During his life on earth, he worked miracles— healing the sick, walking on water, feeding his followers, raising the dead, and much else besides. At the end he suffered unjust persecution, torture, and death at the hands of his contemporaries, after which he rose bodily from the dead and ascended into heaven. All of this, but perhaps especially his suffering, death, and resurrection, somehow deliver us from the power of sin and death and contribute to reconciling the whole world to God. Moreover, after Jesus’ ascension, the Holy Spirit came to dwell within individual believers and to help them realize the sort of union with God that they were intended to have. I believe this story, as I have told it, in its entirety; and I believe that the miracles reported therein literally occurred. I believe all of this in part because I take the New Testament authors to be reliable reporters of the events in Jesus’ life. But, of course, there is much in the story that merits extended discussion. First, how shall we understand the claim that the second person of the Trinity became human? As with the doctrine of the Trinity, the Christian tradition does not

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi



5

offer a full-blown theory of the incarnation but simply imposes boundaries on our theorizing. Whatever else we say about the incarnation, a fully orthodox theory (i.e. one that respects the pronouncements of the ecumenical creeds) must at least say this: In becoming human, the second person of the Trinity retained his divine nature, so that the incarnate Christ is one person with two natures rather than (say) one person with a single hybrid nature, or two persons in one body, each with his own nature; and, whatever else it involved, taking on human nature at least meant coming to have a rational soul, or mind, and a physical human body, and having two wills, human and divine. Chapter 2 offers a model of what this might involve, one that appeals to the same Aristotelian doctrines about matter and form that undergird my solution to the problem of the Trinity. Second, what shall we say about how the suffering, death, and resurrection of Jesus contribute to rectifying the human condition? Our condition, again, is one of sin and misery, brought on by a primordial change in the relationship between God and creation. Whereas God’s presence in the world and to human beings was once vivid and readily available, now it is hidden and available only with difficulty. But scripture tells us that the work of Christ has changed all of this for the better. As a result of Christ’s work, God’s presence and assistance is now more readily available. We who embrace Christ’s work on our behalf have been reconciled with God;⁷ we therefore have access to the divine help we need in order to avoid sin and reach our telos. Although we cannot fully achieve our telos in this life, we are assured that our lives will continue after our physical death and that we will in the afterlife be able to reach it. The New Testament employs a variety of terms (in addition to salvation) to describe what the work of Jesus accomplished on our behalf: e.g. justification, redemption or ransom, reconciliation with God, deliverance from sin, re-creation or rebirth, the offering of an atoning sacrifice, abundant life, and eternal life. But, I take it, the very simple message is that somehow, through Christ, the human condition has been rectified so that we are now able ultimately to glorify God and enjoy God forever. But how exactly does it all work? Which of the aforementioned terms are to be taken literally, and which are mere metaphors? Different decisions on these matters push one in radically different theoretical directions. Taking the justification and atoning sacrifice language quite literally and treating ransom language as more metaphorical, for example, tends to push theologians in the direction of a penal-substitutionary model: Jesus’ death on the cross was a sacrifice to God the Father, wherein Jesus bore in his body and soul exactly the penalty that we ourselves deserved in order to satisfy the wrath of God. Taking the redemption and ransom language more literally, on the other hand, pushes in the direction of ⁷ And what of those who, for whatever reason, have not embraced Christ’s work? My thoughts on this topic are still evolving; but I can say at least this much here: I have argued in Rea 2018 that there is good reason to think that everyone who at least tries to seek God will ultimately have a salvific relationship with God, and I have also argued in Rea (forthcoming) that it is irrational to have (as I do) the unconditional hope that soteriological universalism is true without believing that it is true.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

6

   :  

a Christus victor model, in which concerns about justification are (at least) deemphasized and Jesus’ death is seen as a literal transaction of some sort which delivers us from genuine bondage to the Devil, or to the power of sin and death, or to some other kind of evil other-worldly force. The view that the legal/penal imagery deserves pride of place, and that the justification of sinners is first and foremost what was accomplished by Christ’s atoning sacrifice on the cross, has sometimes been referred to as the ‘Protestant Orthodoxy’. (Cf. Aulén 1931.) I do not deny this view. But, at this stage in my thinking about the matter, neither can I defend it. For it is not clear to me that there is sufficient scriptural data for elevating any of these images over the others for theory-building purposes. Furthermore, it seems that one available theoretical option is to say simply this: The main soteriological message of the New Testament is that the work of Christ accomplished, in some sense, all of these things for us.⁸ It made us justified in the eyes of God; it delivered us from the power of sin, evil, and death and resulted in their utter defeat and humiliation; and it brought us new life, eternal and abundant, and made us into new creations. But as to how and why and in exactly what sense all of these things happened, perhaps we cannot say without offering a model that ultimately lapses into metaphor, leaves out important truths, or otherwise misleads. That said, however, I do think that it is possible to develop partial theories about how Christ’s work accomplishes its effects that do real justice to the various images used to characterize Christ’s work without necessarily giving any of them pride of place. Chapter 3 offers an initial foray into some of these issues in the context of a response to Eleonore Stump’s recent book, Atonement (Stump 2018).

2. Evil, Divine Hiddenness, and Worship Part II of this book deals with the two most long-standing and formidable challenges to Christian belief, the problem of evil and the problem of divine hiddenness, as well as with some puzzles that arise in connection with the idea that God both authorizes and validates a response of protest in the face of these problems. Chapters 4, 5, and 6 address the problem of evil; Chapters 7 and 8 focus on the problem of divine hiddenness. Chapter 9 addresses the concerns about protest. Both the problem of evil and the problem of divine hiddenness might fruitfully be characterized as problems arising out of violated expectations about the character and likely observable consequences of divine love, goodness, and power. Good people who have the power to prevent very bad things from ⁸ My view of the atonement resembles the so-called ‘kaleidoscopic theory’ of Mark Baker and Joel Green. See Green and Baker 2000 and Green 2006.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi



7

happening to those they love generally do so—except, of course, in certain cases where there is a special, good reason for allowing the bad thing to happen. Good people who have the power to communicate with those they love generally tend to do so often, openly, and especially in times of trouble—except, again, in certain cases where there is a special, good reason for not doing so. Hence the problems of evil and divine hiddenness. We are told that God is perfectly good, maximally powerful, and infinitely loving; we are told that God is our heavenly parent. Why, then, do bad things, even horrendous things, happen in the world with such soulcrushing frequency and, often enough, on a staggeringly grand scale? Why doesn’t God communicate openly and frequently with us, offering comfort in times of suffering and sorrow and assurance of God’s love and providential control over the world? These, in brief, are the two problems. It is fashionable in some quarters now simply to dismiss the problem of evil as a mere philosophers’ conundrum. N. T. Wright, for example, maintains that the real problem of evil is simply that evil is bad and needs to be dealt with (Wright 2006, ch. 1). Some go so far as to suggest that efforts to address the philosophical problem of evil are crass, or immature, trivializing evil by treating it as a mere puzzle for religious belief. The suggestion is that those spending their time trying to solve the philosophical problem of evil are not behaving with moral integrity; they should be reflecting instead on what God has done about evil and on what God expects us to do about evil. But I think that this critique of the literature is itself guilty of trivializing something important. The problem of evil is made worse by the phenomenon of divine hiddenness. Divine hiddenness makes the experience of evil all the more agonizing; and our most natural thoughts about what perfect goodness and perfect love would look like conspire, in the face of evil and hiddenness together, to steer us away from belief in God. As Robert Anderson wrote at the end of the nineteenth century, It is no novel experience with men that Heaven should be silent. But what is new and strange and startling is that the silence should be so absolute and prolonged; that through all the changing vicissitudes of the Church’s history for nearly two thousand years that silence should have remained unbroken. This it is which tries faith, and hardens unfaith into open infidelity. (Anderson 1897: 62–3)

And, towards the end of the twentieth century, G. Tom Milazzo expressed the same idea all the more pointedly as follows: The God that chooses to be silent in the presence of suffering and death is a God whose face is darkened by our agony and whose hands are covered with our blood. If the face of God is darkened by our suffering, a God whose reason and purpose is unknown to us, the faithful cannot but protest God’s injustice, hiddenness, and absence. Yet inasmuch as this lament is a protest against

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

8

   :   God’s silence, it is also a call to God to come out of the darkness that God might absolve itself of complicity in our suffering and tragedy. . . . Either God is not there, and God’s hiddenness is really absence, or the God who is there is a cruel, angry, brutal God that seems to relish human suffering. Either God is absent, or God is implicated in, if not responsible for our death. (Milazzo 1991: 44–5)

As I see it, then, integrity in the faithful requires us to grapple with these problems, not necessarily by trying to solve them (not everyone is a philosopher-theologian), but at least by attending to their seriousness and respecting the activity of trying to solve them. Moreover, both problems are rich mines for theological insight. Arising as they do out of natural expectations on divine love, goodness, and power, they invite us to reflect theologically on those attributes and offer an accounting of them. Indeed, I would go so far as to say that, just as believers ought not to glibly dismiss these two challenges to their faith, so too atheists and agnostics ought not to glibly treat them merely as reasons to deny the existence of God. As I have said elsewhere, it takes a remarkable abundance of faith in human theoretical capacities to give up belief in God, or to decline to investigate the matter further, simply on the basis of one’s alleged rational insight into the premises of the hiddenness argument. Given the importance and complexity of the question whether God exists, it does not seem right or reasonable to think that the matter could be decided by the hiddenness argument in the absence of detailed, historically and theologically informed exploration and defense of the assumptions about God that are embedded in it. (Rea 2018: 7–8)

In response to both problems, I am inclined to invoke a thesis I have elsewhere dubbed ‘  ’:⁹ H  E: Suppose F is an alleged intrinsic attribute of God; and suppose we have formed expectations about the manifestation of Fness on the basis of our grasp of a non-revealed concept of F-ness (i.e., a concept of F-ness whose content is not fully given in divine revelation).¹⁰ In that case, the violation of those expectations does not by itself support (i.e., imply, render probable, or justify belief in) the conclusion that sentences predicating F-ness of God are not true.

⁹ Rea 2018: 55–6. ¹⁰ The content of a concept is ‘fully given in divine revelation’ just if the complete content of that concept is part of or derivable from the content of divine revelation.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi



9

This thesis bears on the problem of evil in the following way. Suppose (as I think is correct) we have no good reason to think that the content of our concepts of divine goodness, love, or power are fully given in divine revelation. Then we are not entitled to infer that God is not good or loving or all-powerful, or that God does not exist, from the fact that our grasp of these concepts leads us to expect that God could have no justifying reason for permitting evils of whatever sort figure into our preferred formulation of the problem of evil. Accordingly, I endorse what is now commonly (even if somewhat misleadingly) called ‘sceptical theism’—the view (as I characterize it) that no human being is justified in thinking the following about any evil e that has ever occurred: there is (or is probably) no reason that could justify God in permitting e. As I argue in Chapter 6, this view does not imply that we know nothing about what God would or would not permit; but it does imply that the limits on our knowledge undercut our ability to draw inferences from actual instances of evil to the non-existence of God. My response to the problem of divine hiddenness is likewise grounded in H  E, although this fact does not come out as clearly as I would like it to in the chapters I have included in this volume. Chapter 7 was my first attempt at grappling philosophically with the problem of divine hiddenness. I formulated the problem in a non-standard way—I focused on the problem as I experienced it, rather than on the version that has dominated the literature— and I solved it by suggesting that the good for the sake of which divine hiddenness was justified had something to do with the expression of God’s unique and deeply alien personality. In effect, this solution maintains that our expectations on divine love should be tempered by humility because God is deeply other than we are. Given God’s ‘otherness’, it should not be terribly surprising that God violates our expectations in significant ways in manifesting love towards us. Furthermore, I argued, it might be that what justifies these violations of our expectations is not some good that comes to us as a result of what we suffer, but rather the goodness involved in God’s living out the maximally good and beautiful divine personality. In complementary fashion, Chapter 8 focuses more on a version of the problem that has dominated the literature and grounds the solution to the problem (again, an appeal to something like H  E) more explicitly in divine transcendence. As the postscripts make clear, neither chapter is a fully adequate representation of my current thinking about the problem of divine hiddenness; nor do they individually or together fully express what I now think of as the best way of addressing the problem.¹¹ The volume concludes with a chapter that explores concerns about whether impious protest against God—protest that does not arise in a context of steadfast faith in God’s love and goodness—can possibly count as ‘well-formed’ prayer that

¹¹ For the latter, see Rea 2018.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

10

   :  

is consistent with our duty to worship God. In The Hiddenness of God (2018), I argued that one way in which God shows love to people who are angry with God as a result of divine hiddenness, suffering, or related issues is by authorizing and validating a response of protest. I stand by this conclusion; but, as I note at the beginning of Chapter 9, the conclusion stands in tension with plausible and widespread assumptions about our duties to worship God and about the nature of well-formed prayer. The aim of this chapter is to address those tensions.

References Anderson, Robert. 1897. The Silence of God. New York: Dodd, Mead, and Company. Aulén, Gustaf. 1931. Christus Victor: An Historical Study of the Three Main Types of the Idea of the Atonement. London: SPCK. Green, Joel B. 2006. ‘Kaleidoscopic View’. In The Nature of the Atonement: Four Views, edited by James K. Beilby and Paul R. Eddy, 157–85. Downers Grove, IL: IVP Academic. Green, Joel B. and Mark D. Baker. 2000. Recovering the Scandal of the Cross: Atonement in New Testament and Contemporary Contexts. Downers Grove, IL: InterVarsity Press. Langton, Rae and David Lewis. 1998. ‘Defining “Intrinsic” ’. Philosophy and Phenomenological Research 58: 333–45. Milazzo, G. Tom. 1991. The Protest and the Silence: Suffering, Death, and Biblical Theology. Minneapolis, MN: Fortress Press. Rea, Michael. 2017. ‘(Reformed) Protestantism’. In Inter-Christian Philosophical Dialogues, edited by Graham Oppy and Nick Trakakis, vol. 4: 67–88. London: Routledge. Rea, Michael. 2018. The Hiddenness of God. Oxford: Oxford University Press. Rea, Michael. Forthcoming. “Hopeful Universalism.” Religious Studies. Stump, Eleonore. 2018. Atonement. Oxford: Oxford University Press. Wright, N. T. 2006. Evil and the Justice of God. Downers Grove, IL: InterVarsity Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

PART I

INCARNATION, SIN, AND ATONEMENT

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

1 The Metaphysics of Original Sin Various different doctrines in the history of Christian thought have gone under the label ‘the doctrine of original sin’. All of them affirm something like the following claim: (S0) All human beings (except, at most, four) suffer from a kind of corruption that makes it very likely that they will fall into sin. Many (perhaps most) go on to affirm the following two claims as well: (S1) All human beings (except, at most, four) suffer from a kind of corruption that makes it inevitable that they will fall into sin, and this corruption is a consequence of the first sin of the first man. (S2) All human beings (except, at most, four) are guilty from birth in the eyes of God, and this guilt is a consequence of the first sin of the first man. The ‘exceptions’ referred to in S0–S2 are the first human beings (Adam and Eve), Jesus of Nazareth, and (according to those who endorse the doctrine of the Immaculate Conception) Mary, the mother of Jesus. S2 is known as the doctrine of original guilt. It is now common for S2 to be treated as a doctrine separate from the doctrine of original sin, which many philosophers and theologians simply identify with S0 or S1. But it was not always so; and it will be convenient for present purposes just to stipulate that S2 is part of the doctrine of original sin. I will also stipulate that S1 is part of that doctrine. Thus, for purposes here, nothing counts as a theory of original sin or as an expression of the doctrine of original sin (hereafter, DOS) unless it includes commitment to both S1 and S2. For ease of exposition, I will talk as if the story of the Fall as recorded in Genesis 3 is literally true. I do not think that this story must be literally true in all of its details in order for S1 and S2 to be true. But I will not discuss here questions about which details are required by suitably developed versions of DOS, nor will I discuss questions about which details would have to be modified if, as many now believe, the Genesis account of creation were literally false. DOS has played an important role in the history of Christian thought. Among other things, it provides an explanation for the universality of sin, and it also

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0002

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

14

   :  

provides critical underpinning for the view that all human beings—even the youngest of infants—are in need of a saviour.¹ It was accepted by most of the medieval philosopher-theologians from Augustine through Duns Scotus, and it is affirmed by many of the most important post-Athanasian creeds of the Orthodox, Roman Catholic, and evangelical Protestant churches.² Prima facie, however, it conflicts with the following intuitively plausible ‘principle of possible prevention’: (MR) A person P is morally responsible for the obtaining of a state of affairs S only if S obtains (or obtained) and P could have prevented S from obtaining. The reason is simple. According to DOS, human beings are born guilty. But one cannot be guilty simpliciter. If one is guilty, then there must be something— presumably, the obtaining of some state of affairs—for which one is guilty. But, one might think, whatever states of affairs obtained at or before the time we were born were not states of affairs whose obtaining we had the power to prevent. So if MR is true, it would seem to follow that we can be guilty only for things that happen after we are born. But then we cannot be guilty from birth as DOS requires.³

¹ Hence, Augustine writes in one of his anti-Pelagian treatises: Now, seeing that [the Pelagians] admit the necessity of baptizing infants—finding themselves unable to contravene that authority of the universal Church, which has been unquestionably handed down by the Lord and His apostles—they cannot avoid the further concession, that infants require the same benefits of the Mediator, in order that . . . they may be reconciled to God. . . . But from what, if not from death, and the vices, and guilt, and thraldom, and darkness of sin? And, inasmuch as they do not commit any sin in the tender age of infancy by their actual transgression, original sin only is left. (On the Merits and Forgiveness of Sins, bk 1, ch. 39, in Augustine 1999: 30) ² See, for example, The Canons and Decrees of the Council of Trent, Fifth Session (in Schaff 1998a: 83–89); The Orthodox Confession of Faith, pt I, Q. 24 (in Mohila 1975); The Acts and Decrees of the Synod of Jerusalem, ch. VI, Decree XVI (Orthodox Eastern Church 1899: 139–43); The Augsburg Confession, pt I, Arts II–III (in Schaff 1998b: 8–9); The Heidelberg Catechism, Questions 4–11 (in Schaff 1998b: 308–11); The Thirty-Nine Articles of the Church of England, Art. IX (in Schaff 1998b: 492–3); and The Westminster Confession, ch. VI (in Schaff 1998b: 615–17). ³ It has been suggested to me that perhaps the alleged conflict with MR could be dismissed out of hand on the grounds that MR talks about ‘individual guilt’ whereas original sin concerns ‘collective responsibility’, the idea being that we humans are somehow collectively, though not individually, guilty or responsible for the behaviour of Adam. Peter Forrest (1994) develops a view of original sin roughly along these lines, a view according to which a society itself might be viewed as a moral person and the individuals who comprise it might, accordingly, be held collectively responsible for its acts. My own inclination, however, is to think that groups of persons are not themselves moral persons, and that whatever collective guilt or responsibility might be, it will, in any case, depend on facts about individual guilt or responsibility. For example: The mob is collectively guilty for the damage to the city; but what that means is just that various individuals who were parts of the mob are individually guilty for their contributions to the damage. The notion of collective debt is, to my mind, more promising. (As is the notion of collective liability. See, on this, Wainwright 1988: 45ff.) A group might owe $1,000 to someone even if there is no specific amount of money that any particular member of the group owes to that person. But as Richard Swinburne (1989) emphasizes, claiming that we collectively inherit only a debt from Adam is precisely to reject the doctrine of original guilt, which I am here taking as central to DOS.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

15

Whatever scriptural or systematic theological objections one might have against DOS, the apparent conflict with MR is almost certainly the primary source of purely philosophical resistance to it. On the other hand, some theologians, particularly in the Reformed tradition, treat the apparent conflict between DOS and MR as reason to reject MR.⁴ In the hands of these theologians, DOS plays an important role in paving the way for the view that moral responsibility is compatible with determinism—a conclusion which, in turn, constitutes an important premise in defence of the view that freedom is compatible with determinism. Thus, Christians who are interested in preserving their commitment to DOS while at the same time resisting compatibilism about freedom and moral responsibility would do well to examine carefully the question whether there is straightforward conflict between DOS and MR. In what follows, I will show that there is no straightforward conflict. My discussion will be divided into three sections. In Section 1, I will provide a brief survey of theories of original sin. With the exception of Edwards’s theory, which shall be deferred to Section 2, all of the theories that I will discuss there are in tension with MR. We will see, however, that none of these theories explicitly contradicts MR. Rather, the tension arises because none of the theories offers the resources for denying the following very plausible assumption which, in conjunction with DOS, does contradict MR: (A1) No human being who was born after Adam’s first sin could have done anything to prevent Adam’s first sin; and no human being who is born corrupt could have done anything to prevent her own corruption. The conflict to be resolved, then, is not, strictly speaking, between DOS and MR; rather, it is between DOS, MR, and A1. Of course, it is a hollow victory to show that DOS can be reconciled with MR if the price for reconciliation is denying what any sane person would be inclined to accept. A more substantive victory would be achieved if one could actually develop a theory of original sin that rests on metaphysical assumptions that are both deserving of serious consideration and inconsistent with A1. Thus, in Sections 2 and 3 I will describe two such theories. One is a development of a view defended by Jonathan Edwards.⁵ The other rests on assumptions that naturally accompany a Molinist account of divine

⁴ Jonathan Edwards most notably (Freedom of the Will, pt 3, sec. 4, in Edwards 1992: 47–51); but see also, for example, Hodge 2001, ch. 8, and Schreiner 1995. ⁵ Interestingly, the fact that Edwards’s theory of original sin can be reconciled with MR is bad news for Edwards, since Edwards wants to appeal to the alleged conflict between DOS and MR to support the claim that moral responsibility is compatible with determinism. Thus, an additional and important lesson to be drawn from the discussion in Section 2 is that, given Edwards’s own metaphysical commitments, it turns out that a crucial premise in his defence of compatibilism is undermined.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

16

   :  

providence. Section 2 describes the Edwardsian view; Section 3 describes the Molinist view.⁶ Both of the views described in Sections 2 and 3 come with substantial and controversial metaphysical commitments. But in each case, the commitments in question are ones that have been ably defended and taken very seriously in the contemporary literature. I do not, in the end, claim that any of these commitments ought to be accepted; nor do I claim that they must be accepted by anyone who wishes to endorse both DOS and MR. For all I am willing to commit myself to here, it might be that there are other ways of reconciling DOS and MR, and it might also be that none of the ways of reconciling those two doctrines is worth the intuitive price. My aim, again, is simply to show in some detail that there are ways of reconciling those doctrines, and that those represent ‘live options’ that cannot simply be dismissed out of hand.

1. Theories of Original Sin I will begin by providing a brief sketch of the various lines along which the central claims of DOS (i.e. S1 and S2) have been fleshed out.⁷ The purpose of doing so is to help make it clear where on the landscape of possible views the views developed in Sections 2 and 3 will fall. Doing so will also make it clear just how hard it is to generate a plausible theory of original sin that avoids conflict with MR. I will organize my discussion around three questions that might be raised about S1 and S2: (i) What is the nature of the corruption mentioned in S1? (ii) What is it that we are guilty of from birth? and (iii) Is what we are guilty of something that we have done, or not? It is perhaps tempting to think that once the answer to (ii) is settled, the answer to (iii) will be settled as well. But, as we shall see, that is not the case.

1.1 The Nature of Our Corruption S1 says that all human beings (except three or four) are corrupt. But there are at least two different ways of understanding the nature of this corruption. On one view, Adam’s first sin brought about a fundamental change in human nature. Whereas human beings prior to the Fall lacked the inclination to disobey God, all

⁶ In calling these theories ‘Edwardsian’ and ‘Molinist’, respectively, I do not at all mean to suggest that Edwards, Molina, or any of their contemporary followers would necessarily endorse these theories as I am developing them. Edwards, Molina, and their followers might be blamed for saying things that inspired and encouraged the development of these views, but that is all. ⁷ In addition to the sources cited throughout this section, the following works have influenced the discussion of different theories of original sin in this section: Adams 1999; Kelly 1978; Quinn 1984, 1997; Urban 1995; and Wiley 2002.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

17

human beings after the Fall possess such an inclination. Thus, for example, Augustine writes: Man’s nature, indeed, was created at first faultless and without any sin; but that nature of man in which every one is born from Adam, now wants the Physician, because it is not sound. All good qualities, no doubt, which it still possesses in its make, life, senses, intellect, it has of the most High God, its Creator and Maker. But the flaw, which darkens and weakens all those natural goods, so that it has need of illumination and healing, it has not contracted from its blameless Creator—but from that original sin, which it committed by free will. (On Nature and Grace, ch. 3 in Augustine 1999: 122)

And Calvin: Original sin, then, may be defined as the hereditary corruption and depravity of our nature. This reaches every part of the soul, makes us abhorrent to God’s wrath and produces in us what Scripture calls works of the flesh. . . . Our nature is not only completely empty of goodness, but so full of every kind of wrong that it is always active. Those who call it lust use an apt word, provided it is also stated . . . that everything which is in man, from the intellect to the will, from the soul to the body, is defiled and imbued with this lust. To put it briefly, the whole man is in himself nothing but lust. (Institutes of the Christian Religion, bk 2, ch. 1, sec. 8 in Calvin 1986: 90–1)

This sort of view was also endorsed by Luther, and it has been the typical view of theologians in the Reformed tradition.⁸ Another view, however, maintains that the change brought about by the Fall was not so much the positive addition of a new kind of wickedness to a once pristine human nature, but rather the withdrawal of a certain kind of grace that made perfect obedience to God possible. On Anselm’s view, for example, original sin is the loss of original justice, where original justice is the God-given rightness of will that Adam and Eve possessed but lost for themselves and their posterity when they sinned.⁹ Aquinas likewise identifies original sin with the loss of original ⁸ See, e.g. Luther 1976: 95; Edwards, The Great Christian Doctrine of Original Sin Defended (in Edwards 1992: 146ff.); Shedd 2003: 577ff.; Turretin 1992: 639–40. Cf. also The Formula of Concord, Art. I (in Schaff 1998b: 97–106); The Heidelberg Catechism, Questions 4–11 (in Schaff 1998b: 308–11); and The Westminster Confession, ch. VI (in Schaff 1998b: 615–17). ⁹ See his The Virgin Conception and Original Sin (1969). As Jeff Brower explains, Rightness of will, as Anselm conceives of it, is not something that rational creatures, at least in the first instance, are responsible for acquiring; rather it is something they are responsible for preserving once it has been given. In this respect, rightness of will, on Anselm’s view is more like what Aquinas and other medieval eudaimonists would call a theological virtue than it is like one of the cardinal virtues—that is to say, it is something supernaturally infused as opposed to

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

18

   :  

justice, but he characterizes original justice not as a sort of God-given rectitude of will possessed by our first ancestors, but rather as a supernatural gift that made it possible for Adam and Eve to appropriately order the various inclinations that (in us) give rise to sin. Insofar as they were, in Eden, capable of ordering their inclinations appropriately, Adam and Eve were able to refrain from sinning. The corruption brought about by the Fall was the disordering of our inclinations as a result of the withdrawal of the supernatural gift.¹⁰ This sort of view, according to which original sin consists in the loss of a supernatural gift rather than the acquisition of a new kind of corruption in our nature is sometimes characterized, by way of contrast with the Augustinian view, as one according to which human nature is wounded rather than totally corrupted.¹¹ The major confessions of the Roman Catholic and Eastern Orthodox Churches strongly suggest this sort of view, and it was also tentatively endorsed by James Arminius.¹² So far as I can tell, the Edwardsian development of DOS described in Section 2 is neutral between the Augustinian and Anselmian views. As we shall see in Section 3, the Molinist development of DOS can be made to accommodate both views as well.

1.2 For What Are We Guilty? According to S2, we are guilty from birth. But for what are we guilty? As far as I know, all of the existing theories of original sin give one of two answers: (i) we are guilty both for the corruption that makes it inevitable that we will fall into sin, as well as for the particular sin of Adam that caused that corruption, or (ii) we are guilty only for our corruption. The difference between these two answers is commonly characterized as a difference with respect to the question whether the imputation of Adam’s sin to his posterity is immediate (answer (i)) or mediate

acquired by repeated action. Indeed, according to Anselm, God created rational nature—both angels and the first human beings—with rightness of will precisely because they could not be happy without it. . . . According to traditional Christian doctrine, the first human beings and certain of the angels fell from grace by sinning. Anselm explains their sin in terms of their abandoning, or failing to preserve, rightness for its own sake. . . . [I]n the case of the bad angels (i.e., Satan and his cohorts), Anselm thinks their loss is permanent or irretrievable. In case of the first human beings, however, and their descendants (to whom the original loss was transmitted), Anselm thinks that, at least prior to death, their rightness of will can be recovered—though here again the recovery is primarily a matter of grace (co-operating with free will) rather than the result of any effort on the part of individual human beings. (Brower 2004: 249–50) ¹⁰ See, especially, Aquinas, Summa Theologica, pt 1 of pt 2, Q. 82, Art. 2 in Aquinas 1945: 674–5. ¹¹ Cf. The Catechism of the Catholic Church, 2nd edn, secs 400 and 405 (Catholic Church 1994: 112, 114–15). ¹² See The Canons and Dogmatic Decrees of the Council of Trent, Fifth Session (in Schaff 1998a: 83–89); The Catechism of the Catholic Church, 2nd edn, secs. 400 and 405 (Catholic Church 1994: 112, 114–15); The Orthodox Confession of the Eastern Church, Questions 23 and 24 (in Mohila 1975); The Acts and Decrees of the Synod of Jerusalem, ch. VI, Decrees VI, XIV, and XVI (Orthodox Eastern Church 1899: 118–119, 132–5, 139–43); and Arminius 1999: 150–7, 374–6, 717.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

19

(answer (ii)).¹³ On both views, our own corruption is a consequence of Adam’s sin and something for which we are guilty. Thus, either way we bear guilt as a result of something Adam has done. The difference is that answer (i), but not answer (ii), maintains that we are directly accountable for Adam’s first sin. Augustine, Aquinas, Luther, and Calvin all explicitly endorsed the doctrine of immediate imputation, and endorsement of that view is typical of theologians in the Reformed tradition. It is harder to find theologians who explicitly endorse the doctrine of mediate imputation. Anselm does.¹⁴ So, too, does the seventeenthcentury Reformed theologian, Joshua La Place, though his view was formally condemned at the National Synod of France in 1645, and condemned again shortly thereafter by other churches and theologians throughout Europe in the seventeenth century.¹⁵ The view is also sometimes, though I think mistakenly, taken to be the official position of the Roman Catholic Church.¹⁶ The Molinist view that I will develop in Section 3 is also committed to it. The main question that arises in connection with the doctrine of immediate imputation is the question of how we can be guilty of Adam’s sin given the apparent fact that none of us is identical to Adam and none of us existed when Adam sinned. Here there are only two possibilities. One is to deny the appearance, maintaining that we are guilty of Adam’s sin because there is some meaningful sense in which we ourselves committed or participated in the committing of that sin. The other is to claim that it is somehow just for God to impute to us guilt for a sin in whose commission we did not participate. Adapting some terminology from G. C. Berkouwer (1971, ch. 12), we may refer to views that embrace the first possibility as Personal Guilt (PG) theories and to views that embrace the second as Alien Guilt (AG) theories. I will discuss these views in reverse order in Sections 1.3 and 1.4. The doctrine of mediate imputation, by contrast, faces only the general problem of explaining how we could justly be held responsible for a state of affairs that we could not have prevented. In other words, it faces only the general problem of apparent conflict with MR. Notably, Anselm seems content to reject MR.¹⁷ ¹³ Cf. Crisp 2003 and Quinn 1997 for useful discussion. ¹⁴ See The Virgin Conception and Original Sin, ch. 22 (in Anselm 1969: 197–8). ¹⁵ See Hodge 2001, vol. 2: 205ff. for useful discussion. ¹⁶ See Murray 1955: 153–5. There, Murray claims (again, I think mistakenly) to find the position expressed in the Decree on Original Sin produced by the Council of Trent. ¹⁷ In defending the view that even infants deserve condemnation by God, he writes: If you think it over . . . this sentence of condemnation of infants is not very different from the verdict of human beings. Suppose, for example, some man and his wife were exalted to some great dignity and estate, by no merit of their own but by favor alone, then both together inexcusably commit a grave crime, and on account of it are justly dispossessed and reduced to slavery. Who will say that the children whom they generate after their condemnation should not be subjected to the same slavery, but rather should be gratuitously put in possession of the goods which their parents deservedly lost? Our first ancestors and their offspring are in such a condition: having been justly condemned to be cast from happiness to misery for their fault, they bring forth their offspring in the same banishment. (The Virgin Conception and Original Sin, ch. 28 in Anselm 1969: 209–10) See also chapter 22 of the same work.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

20

   :  

1.3 AG Theories AG theories of original sin maintain that we are guilty both for the corruption of our nature and for the sin of Adam, and that we are so guilty despite the fact that we in no way participated in the committing of Adam’s sin. The main challenge for such a theory is to explain how it could possibly be just for God to hold a person guilty for a sin she did not commit. The standard response to this challenge is to claim that we are guilty for Adam’s sin because Adam is the federal head, or representative of the human race. The basic idea is that Adam represented us before God in much the same way that a head of state might represent one nation before another. If a head of state commits a crime against another nation, the nation she represents may well be implicated in that crime and be held accountable for it. War might ensue, and it might turn out that peace can be restored only if the nation whose representative started the war manages to find another representative who can behave in such a way as to rectify the trouble. Thus, for example, Francis Turretin explains: [T]he bond between Adam and his posterity is twofold: (1) natural, as he is the father, and we are his children; (2) political and forensic, as he was the prince and representative head of the whole human race. Therefore the foundation of imputation is not only the natural connection which exists between us and Adam (since, in that case, all his sins might be imputed to us), but mainly the moral and federal (in virtue of which God entered into covenant with him as our head). Hence Adam stood in that sin not as a private person, but as a public and representative person—representing all his posterity in that action and whose demerit equally pertains to all. For Adam to be a public and representative person, it was not necessary that that office should be committed to him by us, so that he might act as much in our name as in his own. It is sufficient that there intervened the most just ordination of God according to which he willed Adam to be the root and head of the whole human race, who therefore not only for himself only but also for his (posterity) should receive or lose the goods. (Turretin 1992: 616)

This view is known as the federalist theory of original guilt. It is endorsed by many theologians in the Reformed tradition (including Turretin) and also tends to be endorsed by theologians in the Arminian tradition (e.g. John Wesley and Richard Watson).¹⁸

¹⁸ Wesley, The Doctrine of Original Sin, According to Scripture, Reason, and Experience, especially pt 3, sec. 6 (in Wesley 1978: 332–4); Watson 1834: 52ff.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

21

Not surprisingly, federalism is typically coupled with a doctrine of the atonement according to which Jesus counts as another representative of the human race—a ‘second Adam’—whose behaviour, unlike Adam’s, is sufficient to restore us to fellowship with God if only we embrace him as our representative. On this view of the atonement, just as the guilt of Adam is imputed to all of us from birth, so too the righteousness of Jesus is imputed to those who embrace him. It is perhaps worth noting that the imputed-righteousness theory of the atonement does not go hand in hand with the federalist or any other AG theory of original sin. That is, one can and many do accept the former without accepting the latter. But, obviously enough, it is hard to see why one should find the imputation of alien guilt objectionable if one is not inclined to object to the imputation of ‘alien righteousness’. But the AG theory, as it stands, is in obvious tension with MR. For nothing in the theory even so much as suggests that there was anything that any of us could have done that would have prevented Adam’s sin. In other words, nothing in the theory suggests any reason for thinking that A1 is false. But, again, A1 together with DOS flatly contradicts MR. Is it possible to produce a credible AG theory that is inconsistent with A1? I doubt it. One might be tempted to suggest that we have counterfactual power over Adam’s sin. To say that we all have counterfactual power over Adam’s sin is to say that, for each of us, there is something we could have done such that, had we done it, Adam would never have sinned. If we do have such power, then A1 is surely false. But, leaving aside worries about the very possibility of our having counterfactual power over the past,¹⁹ the problem with this proposal is that there is absolutely no reason—and certainly no reason arising out of the AG theory—for thinking that it is true. At best, then, it could only be an ad hoc addition to the AG theory. Alternatively, one might be tempted to resist A1 by arguing that there is some sense in which we were all present and able to act at the time of Adam’s sin. If that were true, then we would at least be moving in the direction of a reason to think that A1 is false; and, as we will see shortly, there are various stories one might tell that imply that we were present and able to act at the time of Adam’s sin. Unfortunately, however, all of the extant stories of this sort are either incredible or have the implication that we all actually participated in Adam’s sin (as some PG theories, but no AG theories, maintain). Thus, it is doubtful that any of these stories could be used to supplement the AG theory in such a way as to make even remotely plausible the denial of A1. It is worth noting that, on the federalist theory, since part of the explanation for our guilt from birth is the fact that Adam represents us, and since it is within our

¹⁹ In Section 3 we will consider another view that presupposes that we can have counterfactual power over the past, and there I will briefly explain why many philosophers think that worries about this presupposition ought to be left aside.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

22

   :  

power to do something—namely, embrace Jesus as our representative—that will make it the case that Adam no longer represents us, it is to some extent up to us whether we remain guilty for Adam’s sin. But, as the quotation from Turretin makes clear, the federalist theory still takes it for granted that the fact that Adam represented us from birth was not something we could have prevented. Thus, it looks as if those who wish to endorse both DOS and MR must endorse some sort of PG theory of original guilt.

1.4 PG Theories The main challenge faced by someone who wants to say that we bear personal guilt for the sin of Adam is to explain how we could possibly have participated in the committing of Adam’s sin. One way of meeting this challenge is to endorse a view according to which all of us existed as distinct individuals at the time of Adam and somehow participated in or concurred with Adam’s sin. One way to motivate this sort of view is to endorse a doctrine of pre-existing souls.²⁰ Another way is to urge a metaphysically loaded reading of the suggestion (in Heb. 7:9–10) that Levi was present as an agent in the loins of Abraham, and then to extend this idea to all members of the human race, claiming that everyone was present as an agent in the loins of Adam.²¹ But these sorts of views are neither plausible nor popular. More popular are views according to which we do not co-exist as distinct individuals with Adam, but we do somehow enjoy a certain kind of metaphysical unity with him. Here we have two main views, one sometimes, though perhaps mistakenly, credited to Augustine; the other associated with, among others, Aquinas and Edwards. The former view goes under the label ‘Realism’, and its chief and most explicit proponent is W. G. T. Shedd (2003). The latter view, the one associated with Aquinas and Edwards, is what is sometimes called the ‘Organic Whole’ theory. I will discuss each in turn. In three of his anti-Pelagian works, Augustine makes remarks that suggest the rather startling view that somehow we are Adam, and that not just Adam but human nature itself committed the sin that brought about our corruption. For example:

²⁰ Shedd attributes a view of this sort to Ashbel Green (Shedd 2003: 447). Origen also famously endorsed a doctrine of pre-existing souls, according to which human souls sinned ‘before their birth in the body’ and ‘contracted a certain amount of guilt’ which, in turn, is supposed to explain at least some of the distribution of pain and suffering in the world (see Origen, De Principiis, bk 3, ch. 3, in Origen 1999: 336–7). But it is not clear whether or to what extent Origen’s doctrine of pre-existence is supposed to be connected with the doctrine of original sin. ²¹ Berkouwer (1971: 440ff.) and Murray (1956: 26) both briefly discuss views of this sort without citing references. Anselm (1969: 199ff.) also seems to take this sort of view seriously.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

23

By the evil will of that one man all sinned in him, since all were that one man, from whom, therefore, they individually derived original sin. (On Marriage and Concupiscence, bk 2, ch. 15 in Augustine 1999: 288; emphasis mine) All good qualities, no doubt, which [human nature] still possesses in its make, life, senses, intellect, it has of the Most High God, its Creator and Maker. But the flaw, which darkens and weakens all those natural goods, so that it has need of illumination and healing, it has not contracted from its blameless Creator—but from that original sin, which it committed by free will. (On Nature and Grace, ch. 3 in Augustine 1999: 122; emphasis mine)

Anselm likewise makes remarks along these lines: Each and every descendant of Adam is at once a human being by creation and Adam by generation, and a person by the individuality which distinguishes him from others. . . . But there is no doubt from what source each and every individual is bound by that debt which we are discussing. It certainly does not arise from his being human or from his being a person . . . [for] then Adam, before he sinned, would have to have been bound by this debt, because he was a human being and a person. But this is most absurd. The only reason left, then, for the individual’s being under obligation is that he is Adam, yet not simply that he is Adam, but that he is Adam the sinner. (The Virgin Conception and Original Sin, ch. 10 in Anselm 1969: 183–4; emphasis mine)

According to Shedd, what Augustine and Anselm are both trying to express with these rather cryptic remarks is roughly this: Human beings have two modes of existence. We can exist as individuals, or we can exist en masse as a ‘single specific nature not yet individualized by propagation’ (Shedd 2003: 446). When Adam sinned, all of Adam’s posterity were literally present in Adam in the latter way, as the undifferentiated human nature. Moreover, as Augustine suggests, it was as much that nature as Adam who committed Adam’s sin. Human nature did not act consciously (not being the sort of thing that can be conscious); but, he thinks, the nature of its union with Adam and Eve is sufficient to make it blameworthy for their crime. And since all of humanity together is nothing other than human nature as ‘individualized by propagation’, we too are blameworthy.²² ²² As I have already indicated, however, Anselm explicitly (and repeatedly) denies that anyone other than Adam bears personal guilt for Adam’s sin. For example, in chapter 22 of The Virgin Conception and Original Sin, he says ‘I do not think the sin of Adam passes down to infants in such a way that they ought to be punished for it as if each one of them had personally committed it, as Adam did’ (Anselm 1969: 197). On Anselm’s view, as I have already said, what we are guilty of is simply the corruption of our nature. The passage quoted above, and cited by Shedd (2003: 445), is from a chapter wherein he

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

24

   :  

Shedd’s view is a heroic attempt to reconcile the claim that each of us is to blame for Adam’s sin with the principle that no one can justly be blamed for a sin in whose commission she did not participate. But even if his view could be made plausible (which seems unlikely), it still would not fare well with respect to MR. It might turn out on his view that human nature ‘not yet individualized by propagation’ could have prevented Adam’s sin; but it will not at all follow from this that any of us could have prevented Adam’s sin. (For example: The unruly mob could have prevented the riot; but it does not follow that Fred, who was part of the mob, could have prevented the riot.) Thus, though Shedd’s view might turn out to be of some help in reconciling DOS with some of our moral intuitions, it will not help us to save MR. The ‘Organic Whole’ theory faces similar problems. The idea, in short, is that humanity, human nature, or the human race is an organic whole with the following properties: (a) it is a moral agent; (b) every individual human being is a part or instance of it; and (c) it committed the sin of Adam by virtue of having a part or instance—namely, Adam—that committed that sin. On this view, it is by virtue of being parts, instances, or members of this whole that individual human beings other than Adam participated in Adam’s sin and share the guilt for it. But, as is clear even from this rough sketch, the obvious challenge for the view is to explain in what sense, if any, the non-Adamic parts or instances of the whole could have prevented the sin of Adam. Prima facie, they could not have. The problem is seen most clearly in Aquinas’s version of the view. Aquinas develops his version by way of analogy. Roughly, the analogy is as follows: If you move your hand in such a way as to commit a crime, we won’t blame your hand as such; but your hand will share in your guilt and will justly suffer the consequences of your sin. Your hand shares in your guilt because it is a part of the whole person who committed the sin, and it is a part that was involved in the sin.²³ Likewise, all human beings together comprise an organic whole, and human nature itself was involved in Adam’s sin. Indeed, says Aquinas, [A]ll men born of Adam may be considered as one man inasmuch as they have one common nature, which they receive from their first parents; even as in civil matters, all who are members of one community are reputed as one body, and the whole community as one man. Indeed, Porphyry says that by sharing the same

attempts to explain how we could be guilty of that, not how we could be guilty of Adam’s sin. (Of course, the claim that each of us is Adam is in superficial tension with the claim, clearly implied by the remark from ch. 22, that none of us committed the sin of Adam. But I will not attempt to sort that out here.) ²³ Cf. Summa Theologica, pt 2, sec. 1, Q. 81, Art. 1 (in Aquinas 1945: 664–7), and De Malo, Q. 4, Art. 1 (in Davies 2001: 327–41).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

25

species, many men are one man. Accordingly the multitude of men, born of Adam, are as so many members of one body. (Summa Theologica, pt II, sec. 1, Q. 81, Art. 1 in Aquinas 1945: 666; emphasis in original)

Thus, when Adam sinned, Humanity—the body of which all human beings are parts—sinned.²⁴ And just as all of your parts share in the guilt of whatever sins proceed from your will and involve your whole body, so too all of the parts or members of Humanity share in the guilt of this one sin that proceeded from Adam’s will and involved human nature; for it was by Adam’s will that Humanity committed that sin. Of course, one worry with this analogy is that it looks like it might imply that more than just the guilt for Adam’s first sin could be imputed to Humanity and thus, ultimately, to everyone. Why not Adam’s second sin, for example? Or, for that matter, why not my sins or yours? Aquinas is aware of this worry, and his response, in short, is that only the guilt for Adam’s first sin can be imputed to Humanity (and thus to everyone) because Adam’s first sin was the only sin that involved human nature as such. Aquinas’s view is more satisfying than Shedd’s if for no other reason than that it is somewhat easier to see how all human beings could be at least analogically treated as parts of a common whole than it is to see how we all could exist in an ‘unindividualized’ way in a single person. But it still leaves important questions unanswered. For example, it is hard to see why Adam’s first sin, and that sin alone, would involve all of human nature in the way required by the analogy. Even if we grant that there is a sense in which your hand, but not your foot, is to blame for sins you commit with your hand, still it is hard to see why Adam’s first sin was a sin committed with his whole nature, as it were, rather than a sin that simply involved him as an individual. Most important for our purposes, however, is the fact that, as indicated above, Aquinas’s view lacks the resources to explain how we could have prevented the sin of Adam. Indeed, if we take the analogy seriously, his view straightforwardly implies that we could not have prevented Adam’s sin. According to the analogy, individual human beings other than Adam are related to the impetus behind Adam’s sin as a hand is related to a particular movement of the will of the person of which it is a part. But then, just as your hand is powerless ²⁴ It is not clear to me how seriously Aquinas really wants to take the idea that there is a physical object composed of every human being who ever did or ever will live. Some of his remarks suggest that the idea might just be a metaphor—that it is not literally the case that Adam and the rest of humanity comprise a single body, but that things are only ‘as if ’ that were true. But if this is so, then it is hard to see how the hand analogy manages to illuminate the doctrine of imputation. For, after all, the main initial question about the claim that we bear guilt for Adam’s sin is how it can be just for God to treat us as if we had committed that sin when, to all appearances, we did not commit it. And it is hardly helpful to answer this question by saying simply that God is also treating us as if we were members with Adam of a single body, even though we are not. But for now I will simply ignore this concern.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

26

   :  

to prevent any particular exercise of your will, so too we must be powerless to prevent the exercise of Adam’s will that resulted in the Fall. Thus, Aquinas’s view, like Shedd’s, is of no help in preserving MR. We come now, at last, to Edwards’s theory (though, as I will note in Section 2, it is ultimately only on one of several possible interpretations that his view properly counts as a version of the Organic Whole theory).²⁵ Famously, Edwards appeals to a sort of divine command theory of persistence over time to account for the possibility of our bearing guilt for Adam’s sin. I will save the details of his view for the next section; but what will become clear in that section is that, on either of the two main ways of fleshing out Edwards’s view, conflict with MR can easily be avoided. Edwards, of course, has no interest in reconciling his views with MR. But the fact that his view of original sin is consistent with MR constitutes at least an ad hominem argument against his claim (in Freedom of the Will) that attention to the doctrine of original sin provides reason to think that MR is false.²⁶

2. Jonathan Edwards and the Doctrine of Original Guilt As I have already indicated, there are at least two different ways in which Edwards’s theory of imputation may be fleshed out. On one way of developing it, Edwards’s view counts as a version of the Organic Whole theory, is committed to a theory of persistence that I’ll refer to below as ‘worm theory’, and suffers from some of the same problems that Aquinas’s view suffers from. On the other way of developing it, there is no commitment to worm theory, and the main problems associated with the Organic Whole theory do not arise. I will begin in Section 2.1 by presenting, largely in his own words, the main lines of Edwards’s view about how it is that we bear guilt for Adam’s sin. In Section 2.2, I will digress briefly and describe several different theories of persistence. I will argue that, contrary to what seems widely to be taken for granted, there is no compelling reason to attribute to Edwards belief in a worm theoretic account of

²⁵ Interestingly, Edwards’s view is often, perhaps even typically, characterized as a federalist theory; but I think that this characterization is mistaken. (But see Crisp 2003 for a persuasively argued opposing view.) Part of the problem is that Edwards seems to appeal rather freely to various models for understanding Adam’s relation to the rest of the human race. (For example, Charles Hodge 2001: 207–8 finds not only an affirmation of federalism in Edwards, but also an outright endorsement of Shedd’s realist theory.) The theory I will present here, however, is the carefully worked-out view that he offers in direct response to the question of how it could be just for God to impute Adam’s sin to his posterity. And I think that that view is not properly understood as a federalist view, even though it is consistent with the claim that Adam is the federal head of the human race. The reason is that, whereas federalism takes it that it is Adam’s federal headship that explains the imputation of guilt to the rest of humanity, Edwards’s response to the question of how it could be just for God to impute Adam’s sin to his posterity takes it that something else—a kind of metaphysical unity with Adam—is the basis for the imputation. ²⁶ Pt 3, sec. 4, in Edwards 1992: 47–51.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

27

persistence. Then, in Sections 2.3 and 2.4, I will describe in more detail the two different ways of fleshing out Edwards’s theory of imputation. We will see that both ways provide theories of imputation that are consistent with MR, but I will argue that the one that carries no commitment to worm theory has distinct advantages over its rival.

2.1 Edwards’s Theory of Imputation Edwards’s theory of imputation is presented in its fullest detail in the last part of The Great Christian Doctrine of Original Sin Defended (in Edwards 1992). Whereas Aquinas uses the metaphor of a body in developing his version of the Organic Whole theory, Edwards relies more heavily on the metaphor of a tree. It is worth quoting him at length since, despite the fact that his theory of original sin is well-known and widely discussed, it is often mischaracterized. He begins thus: I think, it would go far towards directing us to the more clear conception and right statement of this affair, were we steadily to bear this in mind: that God, in every step of his proceeding with Adam, in relation to the covenant or constitution established with him, looked on his posterity as being one with him. And though he dealt more immediately with Adam, it yet was as the head of the whole body, and the root of the whole tree; and in his proceedings with him, he dealt with all the branches, as if they had been then existing in their root. From which it will follow, that both guilt, or exposedness to punishment, and also depravity of heart, came upon Adam’s posterity just as they came upon him, as much as if he and they had all co-existed, like a tree with many branches; allowing only for the difference necessarily resulting from the place Adam stood in, as head or root of the whole. Otherwise, it is as if, in every step of proceeding, every alteration in the root had been attended, at the same instant, with the same alterations throughout the whole tree, in each individual branch. I think, this will naturally follow on the supposition of there being a constituted oneness or identity of Adam and his posterity in this affair. (Edwards 1992: 220; emphasis in original)

Then, in a note, he goes on to develop the tree metaphor more fully as follows: My meaning, in the whole of what has been said, may be illustrated thus: Let us suppose that Adam and all his posterity had co-existed, and that his posterity had been, through a law of nature established by the Creator, united to him, something as the branches of a tree are united to the root, or the members of the body to the head, so as to constitute as it were one complex person, or one moral

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

28

   :   whole: so that by the law of union there should have been a communion and coexistence in acts and affections; all jointly participating, and all concurring, as one whole, in the disposition and action of the head: as we see in the body natural, the whole body is affected as the head is affected; and the whole body concurs when the head acts. Now, in this case, all the branches of mankind, by the constitution of nature and law of union, would have been affected just as Adam, their common root, was affected. When the heart of a root, by a full disposition, committed the first sin, the hearts of all the branches would have concurred; and when the root, in consequence of this, became guilty, so would all the branches; and when the root, as a punishment of the sin committed, was forsaken of God, in like manner would it have fared with all the branches; and when the root, in consequence of this, was confirmed in permanent depravity, the case would have been the same with all the branches; and as new guilt on the soul of Adam would have been consequent on this, so also would it have been with his moral branches. And thus all things, with relation to evil disposition, guilt, pollution, and depravity, would exist, in the same order and dependence, in each branch, as in the root. (Edwards 1992: 221, n.; emphasis in original)

Here we are just invited to imagine that ‘through a law of nature’ Adam and his posterity are unified as parts of a single moral agent. But later in the essay Edwards makes it clear (a) that he endorses a theory about laws of nature according to which laws are just divine decrees, (b) that he endorses a theory about persistence according to which facts about persistence depend solely on divine decrees, and (c) that, by divine decree, Adam and his posterity are ‘one’ in the same sense in which a sapling and the tree that it grows into are one. Thus: Some things are entirely distinct, and very diverse, which yet are so united by the established law of the Creator, that by virtue of that establishment, they are in a sense one. Thus a tree, grown great, and a hundred years old, is one plant with the little sprout, that first came out of the ground from whence it grew, and has been continued in constant succession; though it is now so exceeding diverse, many thousand times bigger, and of a very different form, and perhaps not one atom the very same: yet God, according to an established law of nature, has in a constant succession communicated to it many of the same qualities, and most important properties, as if it were one. It has been his pleasure, to constitute an union in these respects, and for these purposes, naturally leading us to look upon all as one. . . . And there is no identity or oneness [between the successive stages of a created substance] but what depends on the arbitrary constitution of the Creator; who by his wise sovereign establishment so unites these successive new effects, that he treats them as one, by communicating to them like properties, relations, and circumstances; and so, leads us to regard and treat them as one. When I call this an arbitrary constitution, I mean, that it is a constitution which depends on

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

29

nothing but the divine will; which divine will depends on nothing but the divine wisdom. In this sense, the whole course of nature, with all that belongs to it, all its laws and methods, constancy and regularity, continuance and proceeding, is an arbitrary constitution. In this sense, the continuance of the very being of the world and all its parts, as well as the manner of continued being, depends entirely on an arbitrary constitution. (Edwards 1992: 224; emphasis in original)

So, on Edwards’s view, the unity that obtains between Adam and his posterity is metaphysically on a par with the unity that obtains between the successive stages of any ordinary persisting thing. But here we encounter a fork in the road; for there are two different ways of unpacking the claim that the unity that obtains between Adam and his posterity is metaphysically on a par with the unity that obtains between successive stages of ordinary persisting things. I will refer to these two ways of characterizing Adam’s unity with his posterity as the Organic Whole theory and the Fission theory. According to the Organic Whole theory, Adam and his posterity are all together parts of a single, spatiotemporally extended object. On this view, Adam and his posterity comprise successive stages of a persisting individual which is (in some sense) a moral agent and which is such that all of its stages, or temporal parts, are personally accountable at least for the one salient crime committed by its Adamic parts. I said earlier that it is only under one interpretation of his view that Edwards’s theory counts as a version of the Organic Whole theory, and this is it. The Fission theory, on the other hand, says that Adam and his posterity are distinct individuals who share a common temporal stage or set of temporal stages (namely, whatever stages of Adam were involved in Adam’s sin, and perhaps all of the preceding ones as well). On this view, Adam undergoes fission at the time of his first sin, splitting into billions of different people, only one of whom gets kicked out of Eden, fathers Cain and Abel, and does the various other deeds traditionally attributed to Adam. As we will see more clearly in Sections 2.3 and 2.4, the Organic Whole theory presupposes the worm theoretic account of persistence, but the Fission theory may be developed independently of that assumption. But first I want briefly to distinguish several different theories of persistence and explain why there is no compelling reason to attribute to Edwards belief in the worm theoretic account.

2.2 Theories of Persistence Notably, though Edwards is commonly cited as a proponent of the view that familiar material objects are four-dimensionally extended ‘spacetime worms’,²⁷ ²⁷ Chisholm (1976: 138–9), Helm (1997, ch. 7), and Sider (2001: 75) are among those who characterize him as holding this view.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

30

   :  

the whole of his view as presented above is consistent with an alternative account of persistence. Let me explain. An object persists just in case it exists at multiple times. But what does it take for an object to exist at multiple times? A fairly commonsensical view about persistence says that existing at multiple times is just a matter of being wholly present at more than one time. In other words, an object persists just in case the whole thing exists at more than one time. Persisting in this way is typically referred to as ‘enduring’; and so the corresponding theory of persistence is typically called ‘endurantism’. According to endurantism, every moment of an object’s career is occupied by the object itself. The main rival to endurantism is ‘perdurantism’, which I will simply characterize as the thesis that objects persist without enduring. According to the most familiar version of perdurantism—I’ll call it ‘worm theory’, for reasons that will become clear shortly—objects persist by having distinct temporal parts at every moment at which they exist. On this view, material objects are extended in time just as they are extended in space; and just as objects have distinct spatial parts in every subregion of the total region of space that they fill at a time, so too they have numerically distinct temporal parts at every time or period of time in their careers. An object exists at a time, then, just in case it has a temporal part at that time; and an object exists at multiple times just in case it has proper temporal parts at multiple times. A temporal part T of an object X, according to the common intuitive definition, is just an object that exists for part of the total duration that constitutes X’s career, and that has X’s spatial boundaries at all of the times at which T exists. As it is usually fleshed out, worm theory says that whatever name we use for an ordinary material object will typically refer to the four-dimensionally extended ‘spacetime worm’ that fills the entire spacetime region that we would normally say is filled by the ‘career’ or ‘lifetime’ of that object.²⁸ Thus, for example, the name ‘David Letterman’ typically refers to the four-dimensionally extended object that fills the region occupied by the event that we would call Letterman’s lifetime; the expression ‘that table’ refers to the spacetime worm that fills the region occupied by the event that we would normally characterize as the career of the table in question; and so on. Attributions of temporary properties to things are to be analysed in terms of attributions of permanent properties to their temporal parts. So, for example, to say that Letterman was short but is now tall is just to say that Letterman has a temporal part that is (eternally) short and another temporal part that is (eternally) tall, and that the short part is earlier than the tall part.

²⁸ I say ‘typically’ because worm theorists also say that sometimes (perhaps often) familiar referring expressions refer to temporal parts of things rather than to the things themselves. Thus, for example, I might now say not only that I am human, but also that I am hungry. In the first case, ‘I’ can clearly refer to a spacetime worm; but in the second case ‘I’ plausibly refers only to my present temporal part.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

31

But there is another version of perdurantism, usually called ‘stage theory’. Assume that there are instants of time.²⁹ The stage theorist will agree with the worm theorist that every instant of an object’s career is occupied by a distinct thing—a stage of the object. She will probably (though not necessarily) also agree that the stages of an object compose a larger, temporally extended object, a spacetime worm of which those stages are temporal parts. But the stage theorist will not say that ordinary names typically refer to spacetime worms. Rather, according to stage theory, an ordinary name typically refers to a stage—to what a worm theorist would call the thing’s current temporal part. Thus, for example, Letterman is nothing other than whatever momentary Letterman-stage exists right at this very instant; and this table is just the present table-stage that stands before us. According to stage theory, attributions of presently possessed temporary properties are unproblematic. The claim that Letterman is tall, for example, is not given an analysis in terms of temporal parts as it is on the doctrine of temporal parts. Rather, it can simply be taken at face value as expressing the proposition that Letterman himself (the whole person) has the property of being tall. Past- and future-tense predications, however, are another story. Letterman was short (when he was a child); but if Letterman is identical to whatever Letterman-stage presently exists, then, strictly speaking, Letterman never existed before now and will not exist later than now. Stage theorists handle this problem by offering a counterparttheoretic analysis of temporal predications. In short, the claim is that predications of the form ‘x was φ’ or ‘x will be φ’ are equivalent, respectively, to claims like: ‘There is a y such that y is φ, y exists at an earlier time than x, and y is a counterpart of x’; and ‘There is a y such that y is φ, y exists at a later time than x, and y is a counterpart of x’.³⁰ The counterpart relation is then analysed in terms of relevant similarity, which, in turn, is normally taken to be a context-sensitive notion. In most contexts, stage theorists argue, the past stages that are relevantly similar to you are precisely those that the worm theorist would take to be your past temporal parts; and these, in turn, are just the stages that an endurantist would identify with you at various times. Thus, the stage theorist is able, by and large, to affirm

²⁹ There is some question about whether stage theory can be developed apart from the assumption that there are instants, but I won’t pursue that here. See Stuchlik 2003 for relevant discussion. ³⁰ This way of telling the stage theorist’s story about temporal predications presupposes that merely past and merely future objects are somehow available to have properties, stand in relations, and fall within the scope of the quantifier. Can this presupposition be done away with? I think that it can be. As I see it, stage theory will fare as well (or not) under the supposition that there are no merely past or future objects as a counterpart theoretic account of modal properties will fare under the supposition that there are no merely possible objects; and most counterpart theorists think that the supposition that there are no merely possible objects poses no problem whatsoever for their view. This view is controversial, but resolving the controversy would take us too far afield. For reasons to doubt that counterpart theory is viable if there are no merely possible objects (reasons which carry over as reasons to doubt that stage theory is viable if there are no merely past or future objects), see Merricks 2003.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

32

   :  

temporal predications (like ‘I was once a baby, but I was never a baby alligator’) that respect our commonsense intuitions. Both stage theory and worm theory are typically—and some would say, necessarily—developed under the supposition that presentism is false, where presentism is the view that it always has been the case and always will be the case that there are no non-present objects. Moreover, as I have indicated, stage theory is normally developed under the assumption that some things have temporal parts. Given this assumption, stage theory, like worm theory, is committed to the view that composition is not restricted in such a way that only objects existing at the same time can compose something. But suppose we drop these assumptions and yet retain stage theory’s counterpart theoretic analysis of temporal predication. We will then have a view according to which, strictly speaking, (a) nothing that now exists did, does, or will exist, at any time other than the present, (b) nothing has temporal parts, and yet (c) claims like ‘Fred was once a child’, ‘this table will probably be here ten minutes from now’, and so on still express truths. Insofar as stage theory counts as a theory of persistence (which is debatable, but generally accepted), this view, too, should qualify as a theory of persistence. It would be a version of perdurantism without any commitment to the existence of temporal parts. It is important to point out here that stage theory, unlike worm theory, belongs to a family of theories about persistence whose members maintain that familiar objects exist at multiple times in the ‘loose and popular’ sense while at the same time denying that they do so in the ‘strict and philosophical’ sense. In other words, stage theory is one among several views according to which it is appropriate and meaningful, but strictly and literally false, to say of familiar objects that they exist at more than one time. David Hume endorsed a view like this, as did Anthony Collins.³¹ Hume is often characterized as a believer in temporal parts. But, in fact, the view he describes—which seems basically the same as Collins’s view—sounds a lot more like a view that has, in recent times, been defended by Roderick Chisholm (1976), who is not a temporal parts theorist. According to Chisholm, only mereologically constant things (masses of matter, simples, etc.) persist in the strict sense. But other things (most familiar objects—tables, chairs, human bodies, etc.) persist in a ‘loose and popular’ sense by virtue of having ‘stand-ins’ at the various times that constitute what we take to be their careers. Chisholm’s view is not quite stage theory. For one thing, Chisholm believes that some things endure, whereas the paradigmatic stage theorist does not. But still, the two views are similar—and more similar to one another, I think, than either is to worm theory. I mention all of this because it is relevant to the question of how to interpret Edwards. Clearly enough, worm theory provides one way—and perhaps the most ³¹ Hume 1978; Collins 1709. See also Bishop Butler’s characterization of Collins’s view, in Butler 1849: 307–8.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

33

natural way—of fleshing out Edwards’s claim that Adam and his posterity are ‘one’ in the way that the root and branches of a tree are one. And this is the view that is commonly attributed to Edwards, particularly by contemporary philosophers interested in saying something about the history of the worm theoretic account of persistence. But to move from Edwards’s use of the tree metaphor to the conclusion that Edwards was definitely presupposing a metaphysic of temporal parts is to rest a lot of interpretive weight on the details of that metaphor; and it is not clear that this is warranted. For one thing, Edwards’s tree metaphor is substantially similar to Aquinas’s body metaphor. But no one wants on that basis to credit Aquinas with endorsing the existence of temporal parts. More importantly, we have to reckon with the fact that (a) in the eighteenth century, no doctrine of temporal parts had yet been clearly articulated (even by Edwards), (b) the explicit (and non-metaphorical) metaphysical claims that Edwards commits himself to are clearly consistent both with stage theory and with the views of Collins and Hume, and (c) the views of Collins and Hume were already in circulation at the time when Edwards wrote his treatise. I will not go so far as to say that it is a mistake to attribute to Edwards belief in a worm theoretic account of persistence rather than belief in the Collins/Hume view. But I do think that attributing to Edwards something like the latter view is at least as reasonable as attributing to him belief in the worm theoretic account. Indeed, superficially it seems more reasonable to do so, in light of his remarks to the effect that each successive stage is a ‘new creation’ that is ‘treated as one’ with its predecessors by ‘arbitrary divine constitution’. That said, let us now compare the virtues and vices of our two interpretations of Edwards. I will begin with the Organic Whole theory which, again, carries commitment to the worm theoretic account of persistence. After that, I will discuss the Fission theory, which can be developed independently of worm theory.

2.3 The Organic Whole Theory According to the Organic Whole theory, every human being is part of Humanity, a four-dimensionally extended object composed of every individual human being, including Adam. If the worm theory were false, there would be no such thing as Humanity (or, at any rate, it would not be the sort of thing that could include both Adam and us as parts). It is for this reason that the Organic Whole theory is committed to that view. And, on this view, we all bear guilt for Adam’s sin because we are all temporal parts of Humanity, which committed the sin of Adam by way of its Adamic temporal part. But now four questions immediately arise. First, is it really true that we hold the temporal parts of a person guilty for the sins committed by that person? That is, if I, by way of my current temporal part, commit a crime, do we really blame any of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

34

   :  

my temporal parts for that crime? Or do we simply blame me, the entire spacetime worm? Second, is this view consistent with MR? Third, is it really appropriate to view Humanity as a thing that acts and is thereby subject to praise and blame? Fourth, why, if this account is correct, do we bear guilt only for Adam’s first sin and not (say) for his second sin, or for the sins of people other than Adam? I will take each of these questions in order. Consider what a worm theorist will say about ordinary ascriptions of praise and blame. Initially, one might think that the temporal parts of a person are fitting objects of praise and blame because those temporal parts have all of the right equipment, so to speak, to think and act in the ways that ordinary persons do. Indeed, on worm theory, the only way an ordinary person can think and act is by having a temporal part that tokens particular thoughts and acts. But is tokening a thought or act sufficient for having the thought or doing the act? In my view, the worm theorist should say ‘no’. The reason is that if she says that tokening a thought or act is sufficient for having the thought or doing the act, then she will be committed to the view that, for every thought I have, there is at least one other thinker (namely, the temporal part of me in which it is tokened) that shares that thought; and for every act I perform, there is at least one other agent that performs that act.³² But that is absurd. If I am a spacetime worm, then the thoughts tokened in my temporal parts are my thoughts, not theirs; and the acts tokened by my temporal parts are my acts, not theirs. But then the responsibility for those acts is my responsibility, not theirs. And so I am the appropriate object of praise or blame for my acts, and they are not. To be sure, if I am punished for my acts, my temporal parts will receive the blows. But that no more implies that they are punished or blamed for my acts than the fact that my hand is slapped as punishment for a crime implies that my hand is blamed or punished. In the case of the hand-slap, I am punished by having damage inflicted upon my hand. Likewise, in the case of ordinary punishment, the agent is punished by having something inflicted upon her temporal parts. So far, then, the Organic Whole theory seems to be in trouble. Note, however, that the views just expressed depend crucially on the assumption that the temporal parts of thinkers are not themselves thinkers. But what if this assumption were false? What if each of our temporal parts were an agent and a thinker in its own right? Then it would seem that directing condemnation at or inflicting damage upon a later temporal part for the crime of an earlier one would be a way of blaming and punishing the later part for what the earlier one had done. Would this be unjust? Not obviously so. But if not, then it must be the case that those later

³² Some worm theorists are apparently content with this consequence. See, e.g. Lewis (1983: 74ff.). Donald Smith, (2004, ch. 2) presses this point as an objection against worm theory; but, as the present discussion makes clear, I doubt that worm theorists are, as such, committed to the view that temporal parts of thinkers are themselves thinkers.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

35

parts are in some sense guilty of the crime of the earlier parts. Of course, it would be misleading to say that the later parts bear guilt for the crime in precisely the same sense in which the earlier temporal parts, or the person as a whole, bear guilt for it. But perhaps we could do justice to our intuitions here by saying that the parts that commit the crime, and the person as a whole, bear guilt in the primary sense whereas the later parts bear guilt for it in a derivative sense (derivative upon their standing in the relation of genidentity to the criminal parts). Presumably this is the sort of thing that Edwards (taken as an Organic Whole theorist) would want to say about Humanity. On this view, Humanity is a moral person that committed the sin of Adam by way of its Adamic temporal part. Both Adam and Humanity are blamed for that sin, and both bear guilt in the primary sense for it. But the post-Adamic parts suffer the consequences of that sin, and they do so justly. Thus, they bear guilt in the derivative sense for that sin. But doesn’t this violate MR? Initially, one might think that it does. The later temporal parts of Humanity could not have prevented Adam’s sin, and yet they are held guilty. Note, however, that once we have the distinction between primary and derivative responsibility, MR is ambiguous. We can resolve the ambiguity by identifying three distinct readings: (MRa) One is morally responsible in the primary sense for the obtaining of a state of affairs only if one could have prevented that state of affairs from obtaining. (MRb) One is morally responsible in the derivative sense for the obtaining of a state of affairs only if one could have prevented that state of affairs from obtaining. (MRc) One is morally responsible in any sense for the obtaining of a state of affairs only if one could have prevented that state of affairs from obtaining. The Organic Whole theorist who believes that the temporal parts of persons can themselves be persons can insist that it is MRa rather than MRb or MRc that best expresses the intuitions that initially led us to endorse MR; and so she can claim that, once it has been suitably clarified, her view is consistent with MR. It is so consistent because, though later temporal parts of Humanity are held responsible for something they could not have prevented, they are not held responsible in the primary sense.³³ Whether this move will be plausible or not is, of course, debatable. But the point is just that once the distinction between primary and derivative responsibility is on the table, the conflict between the Organic Whole theory and MR is not at all straightforward.

³³ Note that this strategy enables the Organic Whole theorist to preserve MR without rejecting A1. Obviously enough, then, the claim (which I have made in several places throughout this chapter) that the conjunction of DOS and A1 contradicts MR presupposes that MR is not ambiguous in the way described here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

36

   :  

Still, this version of DOS faces some serious problems. For one thing, it seems wholly inappropriate to view something like Humanity as a moral agent.³⁴ This is for the same reason that it seems inappropriate to view the temporal parts of persons as moral agents. Many, if not all, of my post-natal temporal parts have the right equipment to be moral agents. They have brains (or, at any rate, temporal parts of brains), and their brains (or brain-parts) token thoughts, acts of will, and the like. But, so I would say, none of my temporal parts is the subject of its thoughts, and so the thoughts tokened in my temporal parts are not appropriately ascribed to them.³⁵ For the same reason, none of my temporal parts are appropriately regarded as the agents of the acts of will tokened in them. It is not their experiences, beliefs, and desires that give rise to those acts of will; and so there is no reason to regard them as the agents of those acts. And the same is true for Humanity. It has a brain—indeed, multiple brains. And its brains token thoughts, acts of will, and the like. But, like my temporal parts, Humanity is not the subject of the thoughts tokened in those brains, and so there is no reason to regard it as the agent of the acts of will that are tokened in them. Moreover, like Aquinas’s view, the Organic Whole theory lacks the resources to explain why it is only the guilt for Adam’s first sin that gets imputed to all of the temporal parts of Humanity. And whereas Aquinas could at least try to insist that only Adam’s first sin involved all of human nature, Edwards (on this interpretation) could not do so, for the metaphysical presuppositions that support the attribution of Adam’s guilt to all of us transparently imply otherwise.

2.4 The Fission Theory The Fission theory, on the other hand, is much more promising. For one thing, it is more exegetically plausible since it, unlike its competitor, is compatible with theories of persistence that were actually in circulation at the time that Edwards wrote his treatise on original sin. Moreover, it provides the resources either to answer or to obviate all four of the troublesome questions that arose in connection with the Organic Whole theory. To see this, let us begin by considering a straightforwardly stage theoretic development of the Fission theory. Recall the following remark: And there is no identity or oneness [between the successive stages of a created substance] but what depends on the arbitrary constitution of the Creator; who by ³⁴ Wainwright (1988) raises this objection against Edwards, though he does not develop it in the way that I do. ³⁵ As indicated above (note 32), there is room for disagreement on this point. But, as we have seen, saying that each of my temporal parts is the subject of its thoughts leads to an absurd multiplication of thinkers.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

37

his wise sovereign establishment so unites these successive new effects, that he treats them as one, by communicating to them like properties, relations, and circumstances; and so, leads us to regard and treat them as one. (Edwards 1992: 224; emphasis in original)

In the light of a counterpart theoretic account of persistence (together with a counterpart theoretic understanding of modal predications), we may flesh out remarks like this and others along the following lines. What temporal predications are (objectively) true of an individual depends entirely upon what stages God chooses to treat as counterparts of that individual. The counterpart relation may still be analysed in terms of relevant similarity; but, on this view, relevant similarity is an objective relation grounded in God’s judgments. For the most part, we may assume that God’s judgments coincide with our own intuitive judgments. In other words, for the most part, those stages that are objectively relevantly similar to us, or to other objects, are precisely the stages we would expect to be relevantly similar to us if our commonsense judgments about persistence were true. And so those judgments are true. I was once a baby; I was never a baby alligator. And so on. However, we learn from revelation (plus, perhaps, a bit of systematic theologizing) that a rather unexpected set of temporal predications is true of each of us. It turns out that, according to revelation, the stages of Adam that committed Adam’s first sin are relevantly similar to us in a way that suffices for their being our counterparts. In other words, for each of us, there is an x such that x is our counterpart and x committed Adam’s sin. Thus, given our counterpart theoretic account of persistence, it is true of each of us that we committed Adam’s sin. Notably, it is also true of each of us that we were Adam. But, as we have seen above, that hitherto cryptic remark has been affirmed by luminaries of the Church since the time of Augustine. Only now we have the resources to make sense of it. On the present view, it is literally true that we sinned in Adam, and that by Adam’s sin, the many were made sinners.³⁶ Consider now the four troublesome questions that arose in connection with the Organic Whole theory. Do we really blame later stages for the sins committed by earlier stages? On this view, yes. For, on this view, what it means to say that I (a momentary stage) committed some sin in the past is just that there is some earlier stage that committed the sin and is my counterpart. Is there conflict with MR? No. For, though I am blamed for Adam’s sin, it is also true that I could have prevented Adam’s sin. After all, I was Adam, and, by hypothesis, Adam could have prevented Adam’s sin. Thus, A1 is false, and MR is preserved. Is Humanity a moral agent?

³⁶ The claim that ‘we sinned in Adam’ is based on an inaccurate translation of Romans 5:12b, one which greatly influenced Augustine’s development of the doctrine of original sin as well as much subsequent thought on the topic (see Wiley 2002: 51 for discussion). The claim that ‘the many were made sinners’ is from Romans 5:19.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

38

   :  

On the present view, that question is obviated; for the present view makes no commitment even to the existence of such an object, much less to its moral agency. Is there an answer to the question of why only Adam’s first sin and not his later sins or the sins of other ancestors of ours are imputed to us? Yes: the answer is that, so far as revelation teaches us, only the stages of Adam that were involved in committing his first sin stand to each of us in the (objective) counterpart relation. Of course, one might well note that, at this point, a fifth difficult question arises: On this view, it is entirely up to God whether, for any person P, the parts of Adam that committed Adam’s sin are counterparts of P. Thus, except in the case of those very stages of Adam that actually committed Adam’s sin, it is entirely up to God whether claims of the form ‘p committed Adam’s sin’ are true. Likewise, then, it is entirely up to God whether claims of the form ‘p is to blame for Adam’s sin’ are true. Why, then, would God choose for everyone to have those sinning stages of Adam as counterparts? Wouldn’t we expect a loving, compassionate, and forgiving God to arrange things so that as few people as possible (rather than as many people as possible) are to blame for the sin of Adam?³⁷ Perhaps; but it is important to keep in mind here that, just as it is up to God whether to hold me guilty for Adam’s sin, so too it is up to God whether to hold later stages of Adam guilty for Adam’s sin. If it were really true that a good God would minimize overall guilt, then it should follow (if the Fission theory is correct) that a good God would not even hold later stages of Adam guilty for Adam’s sin. But that is a counterintuitive consequence. The Fission theorist therefore has reason to reject the claim that a good God would minimize overall guilt; and, if her theory is to have any hope of respecting ordinary moral intuitions, she will have to sign on to a view according to which there is something good, fitting, or wise about God’s choosing to ascribe guilt to a great many more stages than those that are actually involved in the commission of the various sins that have been committed throughout human history. Once she has accepted this sort of view, however, the way is open for her to argue that precisely what makes it good, fitting, or wise for God to ascribe guilt (say) to me for the sins of some of my yesterday-stages also makes it good, fitting, or wise for God to ascribe guilt to me for the sins of Adam’s stages. Notably, this is precisely the sort of approach that Jonathan Edwards himself takes in response to the question of why a good God might choose to ascribe guilt for Adam’s sin in the ways that the Fission theorist says that he does (Edwards 1992: 225). I have so far been fleshing out the Fission theory under stage theoretic assumptions, but it is important to note that the story could as easily be fleshed out under other assumptions. All we need is a theory of persistence that enables us to make coherent and plausible sense of the central claim that Adam underwent fission,

³⁷ I thank Michael Murray for raising this objection.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

39

splitting into billions of different people. That claim is singularly implausible under, say, endurantist assumptions; for there was simply no event in Adam’s life that looked even remotely like an enduring substance splitting into billions of different people. But, to my mind, the claim that Adam underwent fission will be equally plausible (or not) on any theory of persistence according to which, at least for the most part, the persistence of ordinary things is only persistence in the ‘loose and popular’ sense and temporal predications are to be analysed in terms of predications of ‘stand-ins’ or counterparts. For, on any of these theories, it will not be hard to tell a story according to which Adam, or some stage of Adam, counts as a suitable stand-in for all of us, thus grounding the attribution to all of us of the property having committed Adam’s sin. But what about worm theory? After all, worm theory does not fit into that family of theories whose members say that familiar things, for the most part, do not exist at multiple times. On worm theory, the central claim of the Fission theory amounts to the claim that all human beings overlap Adam, having some relevant temporal part of him as their first temporal part. There is nothing incoherent in this; but there is at least one worry to be raised. The worry is that this claim does not fit naturally with assumptions that typically accompany worm theory. Worm theorists typically want to say that the temporal parts of persons are unified by spatiotemporal and causal relations of a sort that seem not to hold between (say) Adam’s temporal parts and mine.³⁸ Thus, there is a real question of motivation here: Why, apart from the fact that it is required by a particular theory of original sin, should we believe that Adam has undergone fission and split into billions of different people? Here, worm theory has trouble accommodating the Fission theory for much the same reason that endurantism does: there is no event in Adam’s life that looks like his splitting into billions of different people. And so it is hard to see what would explain, or ground, the alleged fact that Adam’s temporal parts are among my temporal parts. After all, my temporal parts bear relations of biological and psychological continuity to one another that they do not bear to any part of Adam; and it is hard to see any other plausibly relevant spatiotemporal or causal relations that my parts bear to Adam’s that they do not bear to the parts of many other people. Thus, absent further argument, the claim that Adam and I share temporal parts in common is implausible. One might reply by saying that the temporal parts of Adam and me (and so of persons generally) are unified by brute, unanalysable genidentity relations. But saying this sheds no light on why Adam’s initial temporal parts and none others are shared by everybody. To claim that it is just a brute fact that this is so is perfectly coherent, but it is, to my mind, unacceptably ad hoc. But there is a more promising move that can be made. One might say that (a) sometimes, even if not ³⁸ See, for example, the discussion of identity criteria and persistence across temporal gaps in Hudson 2001 (chs 4 and 7).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

40

   :  

always, the temporal parts of persons are unified by nothing more than certain kinds of similarity relations, and (b) only the temporal parts of Adam up through the time of his first sin are similar enough to the temporal parts of everyone else to count as temporal parts of everyone else. So far as I can tell, adding this claim to the worm theoretic development of the Fission theory puts it on a par with the stage theoretic development of that theory. So long as one is prepared to analyse our genidentity with Adam in terms of relevant similarity, there seems to be no reason to prefer one to the other apart from whatever reasons there are in general to prefer worm theory over stage theory or vice versa.

2.5 Conclusion I have argued in this section that Edwards’s theory of original sin is consistent with MR regardless of whether it is interpreted as affirming a worm theoretic account of persistence. Moreover, I have identified two interpretations of Edwards’s theory (the Organic Whole theory and the Fission theory), and I have argued that the Fission theory is both more plausible exegetically (since it, unlike the Organic Whole theory, is compatible with theories of persistence that were actually in circulation at the time Edwards wrote his work on original sin) and also more philosophically satisfying than the Organic Whole theory. The fact that either way of fleshing out Edwards’s view is consistent with MR is actually bad news for Edwards, since Edwards wants to argue that attention to the doctrine of original sin provides reason to reject MR and related principles. Appeal to the alleged conflict between original sin and MR is an important premise in his argument for compatibilism about determinism and moral responsibility. But for those who wish to retain MR without giving up DOS, this fact is good news—at least if they are willing to reject endurantism and to analyse genidentity at least partly in terms of relevant objective similarity. For many of us, however, this will be too high a price. It would be nice, therefore, if an alternative were available. Happily, one is (though, as we shall see, it too comes with controversial metaphysical commitments). I will develop that alternative in the next section.

3. Original Sin and Conditional Transworld Depravity The version of DOS that I will develop in this section depends on two assumptions, the first of which is central to a Molinist account of divine providence and the second of which is a natural concomitant. Those assumptions are as follows: (M1) For every human person P, there are counterfactuals of freedom, including some with false antecedents, that are true of P.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

41

(M2) For any counterfactual of freedom C that is true of a human person P, P is or was able to prevent C from being true of P. For purposes here, a counterfactual is any conditional of the form ‘if P were the case then Q would have been the case’.³⁹ Counterfactuals of freedom, then, are conditionals of the form ‘if S were in circumstances C, S would freely do A’. Many philosophers are inclined to reject M1 on the grounds that, in the case of counterfactuals of freedom with false antecedents, it is hard to see what could possibly ground their truth. The idea, roughly, is that if a person S is free and would remain free if (non-actual) circumstances C were to obtain, then there is nothing about S that makes it the case that she would do one sort of action rather than another. Perhaps it is true that S would probably do one sort of action rather than another; but, according to those who are inclined to lodge the so-called ‘grounding objection’, that is the strongest that can be said. Many philosophers are also inclined to think that, even if there are true counterfactuals of freedom with false antecedents, the truth values of those counterfactuals are not in any meaningful sense up to us. It is tempting to say that such counterfactuals are grounded in our character and that, if we are free, our character is up to us. The trouble with this, however, is that our character seems to be entirely constituted by facts about our history plus a variety of ‘would probably’ facts; and it is hard to see how these facts alone could ground claims about what we would (definitely) do in various kinds of non-actual circumstances. Moreover, those who endorse a Molinist account of divine providence typically want to say that God’s knowledge of counterfactuals of freedom entered into his decision about what world to actualize. But this claim, together with the claim that it is up to us which counterfactuals are true of us, might seem to generate a kind of explanatory circle. Since God’s knowledge of counterfactuals of freedom plays a role in his decisions about what worlds (and so what individuals) to create, it looks as if the truth of any particular counterfactual of freedom C about an agent S must be explanatorily prior to the existence of S. But if the truth of C is supposed to be up to S (or at least preventable by S), then it looks as if S’s existence must be explanatorily prior to the truth of C. And, assuming that explanatory priority is a kind of dependence, it would appear that this little circle is vicious: the truth of C depends on the existence of S which, in turn, depends on the truth of C.⁴⁰ For these reasons and others, M1 and M2 are highly controversial. And so, unless the objections can be addressed, any theory of original sin that depends on ³⁹ The label applies most naturally when the relevant conditional has a false antecedent; and sometimes the label is used in such a way that a conditional counts as a counterfactual only if the antecedent is false. Often enough, though, the label is also used in the way that I am proposing to use it—to cover any sort of ‘if . . . would’ conditional. ⁴⁰ For discussion of these and related objections, see (for starters) Adams 1977, 1991; Craig 1991, 2001; Flint 1998; Hasker 1989; and van Inwagen 1997.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

42

   :  

them must be seen as a theory with substantial metaphysical baggage. My own view is that Molinists have gone a long way, though not the whole way, towards answering the objections that have been levelled against M1 and M2.⁴¹ But it is beyond the scope of this chapter to defend that claim. For the remainder of this section I will simply assume that M1 and M2 are true and attempt to build a theory of original sin around them. Let us begin by defining some terminology:⁴² • x is significantly free = significant for x.

df

x is free with respect to some action that is morally

• A is a morally significant action for x = df A is such that it would be morally wrong for x to perform it, but right for x to refrain, or vice versa. • x strongly actualizes S = df x causes S to be actual. • The state of affairs S includes the state of affairs S* = df necessarily, if S obtains then S* obtains. • S is the largest state of affairs strongly actualized in W by x = df x strongly actualizes S in W and, for every state of affairs S* that x strongly actualizes in W, S includes S*. • P suffers from transworld depravity = df for every world W such that P is significantly free in W and P does only what is right in W, there is a state of affairs T and an action A such that (i) T is the largest state of affairs strongly actualized in W by God, (ii) A is morally significant for P in W, and (iii) if God had strongly actualized T, P would have gone wrong with respect to A. Given these definitions, being free and suffering from transworld depravity (TWD) guarantees that one will fall into sin. This may be shown as follows. Let T* be the largest state of affairs strongly actualized in some world W by God; and let P be a person who is both free and suffers from TWD in W. Either W is a world in which P freely does something wrong, or not. By definition, if P suffers from TWD in W, then if God were to actualize T*, P would freely do something wrong. Thus, W cannot be a world in which P fails to freely do something wrong. Thus, necessarily, if P suffers from TWD in a world W, P will freely do something wrong. Obviously, however, it does not follow from this that it is necessary simpliciter that P fall into sin. The fact that suffering from TWD guarantees that P will fall into sin is perfectly consistent with the claim that it is possible that P not fall into sin (assuming, of course, that there is a possible world in which either P is not free or P does not suffer from TWD). ⁴¹ See especially Craig 1991, 2001, and Flint 1998. ⁴² All of the following definitions are either duplicated or adapted from Plantinga 1974: 166, 173, 186.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

43

Given the truth of M2, each of us has the power to prevent our suffering from TWD. The reason is that a necessary and sufficient condition for a person P’s suffering from TWD is that a certain range of counterfactuals of freedom be true of P; and, according to M2, for any counterfactual of freedom C that is true of P, P has the power to prevent C from being (or having been) true of her. Still, even though it is up to us whether we suffer from TWD, there is good reason to think that TWD is not an acquired property. To see why, suppose there is a person P who, up until time t, does not suffer from TWD and then, at t, comes to suffer from it. Let T* be the largest state of affairs that God strongly actualizes. Now, consider the following counterfactual: (CF) If God were to strongly actualize T*, P would freely do something wrong. Given the definition of TWD, if P suffers from TWD after t but not before, then CF is true after t but not before. Could CF change its truth value like that? Some counterfactuals, of course, can become true or false. Suppose you undergo a change of heart towards your enemy. In such a case, it may well be that after the change, but not before, if you were given the opportunity to become reconciled with your enemy, you would do so. Thus, a certain counterfactual would have been true of you at one time but not at another. But CF is not like that; for CF is equivalent to a claim whose consequent quantifies over all times, i.e.: (CF*) If God were to strongly actualize T*, it would be the case that there is (was, or will be) a time at which P freely does something wrong. But now consider the following premise: (P1) If P does A at t, then the proposition that P will do A at t was true at every time prior to t. P1 is very plausible. Moreover, though some philosophers (including some theists) reject it, traditional Christians have compelling reason to accept it. For, after all, traditional Christians believe, among other things, that God foreknows all of the future free acts of all of his creatures; but such foreknowledge is impossible unless, for every free act A, the proposition that A will occur was true prior to A’s occurrence. This, to my mind, constitutes good reason even apart from its intrinsic plausibility to endorse P1. But if P1 is true (as I shall henceforth assume), then, obviously enough, CF* cannot change its truth value. And if CF* cannot change its truth value, then TWD cannot be an acquired property. Thus, if it is ever true that P suffers from TWD, it is always true that P suffers from TWD. TWD, then, is a condition we have from birth.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

44

   :  

Some, no doubt, will find it hard to swallow the claim that each of us now has the power to prevent the obtaining of a state of affairs that obtained when we were born. But the claim might go down a bit more easily if we keep in mind that one of the most popular responses to fatalist arguments is to say that we have counterfactual power over a great many facts about the past.⁴³ As we saw in Section 1.3, an agent S has counterfactual power over the obtaining of a past fact F just in case there is some act A that S has the power to do such that, had S done A, F would not have obtained. There seems to be no in-principle obstacle to our having such power over at least some facts about the past (e.g. facts like its having been true one million years ago that I would mow my lawn today, or God’s having believed one million years ago that I would mow my lawn today); and Alvin Plantinga (1986) has argued persuasively that divine foreknowledge together with the possibility of divine ‘fore-cooperation’ imply that most facts about the past are such that we might have counterfactual power over them.⁴⁴ Thus, it is at least prima facie plausible that we might have such power over the fact that, from birth, we have suffered from TWD. One might object here that, if we are willing to invoke counterfactual power over the past in our theory of original sin, then preserving MR becomes too easy: one might simply say that, for each of us, there is something that we could have done such that, had we done it, Adam would never have sinned. There is, then, no need for controversial Molinist assumptions or a stage theoretic apparatus; MR can be saved by the simple expedient of postulating counterfactual power over the past. But, as I argued in Section 1.3, the trouble with this proposal is that there is absolutely no reason to think that it is true. Moreover, and perhaps more importantly, even if it were true, we would still face difficult questions about how we could be held accountable for Adam’s sin. After all, if there is something I can do (or could have done) such that, had I done it, Adam would never have sinned, I have no idea what it is. So it is hard to see how I could be held accountable for not having done it. One might be tempted to think that refraining from sin is the thing that I could have done that would have prevented Adam’s sin, and I can certainly be held accountable for not doing that. But keep in mind that, according to traditional Christian belief, Jesus of Nazareth refrained from sinning; and he did not thereby prevent Adam from sinning. Thus, it is hard to see why we should think that our refraining from sinning would have prevented Adam’s sin. On the other hand, by refraining from sinning, Jesus of Nazareth arguably did prevent himself from suffering from TWD; and, likewise, if we were to refrain ⁴³ This is the ‘Ockhamist’ response to fatalism. For discussion and development of this response, see the papers reprinted in Fischer 1989, especially Plantinga 1986. ⁴⁴ At any rate, there seems to be no obstacle to our having counterfactual power over the past if presentism is false. As Alicia Finch and I have argued elsewhere, however, if presentism is true, the Ockhamist response is untenable (Finch and Rea 2008).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

45

from sin, we would prevent ourselves from having suffered from TWD. And so here is one of the main advantages that the present Molinist proposal enjoys over the proposal that we have counterfactual power over Adam’s sin: Suppose that TWD is identified with the corruption of our nature that was produced by the Fall. Since refraining from sin would keep us from suffering from TWD, and since refraining from sin is clearly something that we can be blamed for not doing, it is easy to see how we might both have counterfactual power over the fact that we have been corrupt from birth and be held accountable for failing to act in a way that would have prevented our being corrupt from birth. One option, then, for those interested in developing a theory of original sin under Molinist assumptions is to identify TWD with the sort of corruption that DOS takes to be a consequence of the Fall. After all, it seems to be the right sort of property. We have it from birth, and we have it contingently. Moreover, there is no in-principle obstacle to supposing that our suffering from it is, in some sense, a consequence of Adam’s sin. We have already acknowledged that, though the counterfactuals that constitute us as TWD-sufferers have been true from the beginning of time, there are nevertheless things we can do (or could have done) such that, had we done them, we would not have suffered from TWD. But if it is coherent to say this, then surely it is also coherent to suppose that if there had been no Fall, we would not have suffered from TWD. This by itself doesn’t guarantee that our suffering from TWD is a consequence of the Fall. But my point here is just that there is no obvious reason to deny that our suffering from TWD could be a consequence of the Fall. Finally, since we have the power to prevent our ever having suffered from it, if TWD were identified with the corruption that is brought about by the Fall, the resulting theory of original sin would be consistent with MR. But there is a complication worth mentioning. Earlier, I said that the Molinist theory of original sin that I’d be developing would be consistent with both an Augustinian and an Anselmian view of the nature of the corruption that is original sin. But if we identify TWD with the corruption in question, it looks as if Anselmian views are ruled out. The reason is that, according to Anselmian views, our corruption consists mainly in the loss of a supernatural gift, possession of which would enable us to remain free of sin. Of course, any view according to which we are free and according to which one can be blamed only for things that one freely does will be a view according to which we are in some sense able to remain free of sin. But I take it that, on the Anselmian view, it is not the case that the supernatural gift merely makes it possible for us to remain free of sin. (That was possible already.) Rather, the supernatural gift is such that, had God given it to us, we might have remained free from sin. In other words, on the Anselmian view there is something that God can do for us (namely, restore to us the supernatural gift that Adam and Eve lost for the human race) such that, had he done it, we might have always freely done what is right. But to say that we suffer from TWD is precisely to deny this. To say that we suffer from TWD is, in effect, to say that even

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

46

   :  

if God had done whatever he does in worlds where we always freely do what is right, we still would have sinned. Thus, there is nothing God could have done (consistent with our being free) such that, had he done it, we might always have done what is right. The Anselmian view, then, is ruled out.⁴⁵ Perhaps it is not such a bad thing to rule out the Anselmian view.⁴⁶ But it would be nice to be able to accommodate it if possible. Thus, I offer the following, second option to the Molinist: build a theory of original sin around a notion of conditional transworld depravity (CTWD) rather than around TWD. Informally, to say that someone suffers from CTWD is just to say that there is some condition C such that, even if God had done whatever he does in worlds where both condition C obtains and we always freely do what is right, we still would have sinned. More formally, CTWD may be defined as follows: P suffers from conditional transworld depravity = df there is some condition C that does not include any of P’s free acts and is such that, for every world W such that P is significantly free in W, P does only what is right in W, and C obtains in W, there is a state of affairs T and an action A such that (i) T is the largest state of affairs strongly actualized in W by God, (ii) A is morally significant for P in W, and (iii) if God had strongly actualized T, P would have gone wrong with respect to A. Like TWD, CTWD will be a permanent property of the persons who suffer from it; it will be a contingent property; and whether we suffer from it will be preventable by us. But, unlike TWD, suffering from CTWD is consistent with there being something God might have done such that, had he done it, you might (or even would) always have freely done what is right. We may then flesh out our CTWD-based theory of original sin as follows. Consider again Aquinas’s theory about the nature of the corruption that is original ⁴⁵ Here I assume (what is standard in the literature on counterfactuals) that ‘if p were true, then q would have been true’ entails and is entailed by ‘it is not the case that, if p were true, then q might not have been true’. ⁴⁶ One reason for thinking that it would not be so bad to rule out the Anselmian view is that the Anselmian view might be thought to raise questions about the goodness of God. On the Anselmian view, we are subject to sin and death only partly, and not entirely through our own fault. For God could have chosen to withhold the sort of grace that was present in Eden only from those who sinned in the way that Adam and Eve did (and he could also have chosen to quarantine such people so that they could not interact with those who had not yet sinned). If he had so chosen, then at least some of us might have enjoyed the great benefit of a perfectly sinless life and a robust friendship with God. And so it seems that it would have been better for God to have so chosen. But if that is right, then it looks as if God’s choosing to withhold the grace that was present in Eden from all of Adam’s posterity is inconsistent with his perfect goodness. To my mind, however, this objection is far from decisive. For the problem here seems just to be an instance of the problem of evil generally; and so it seems that familiar strategies for responding to the latter problem will also apply to the former. Thus, perhaps there are great goods that God could obtain only by withdrawing his supernatural gift from the human race; or perhaps God’s withholding his grace from us is the permission of a gratuitous evil, but, contrary to our intuitions, it is not inconsistent with God’s perfect goodness to permit gratuitous evils; and so on.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

47

sin. On his view, the inclinations that lead us into sin were present in human nature from the beginning, but God had given Adam and Eve a supernatural gift, or a certain kind of grace, that enabled them to order their inclinations in such a way as to avoid falling into sin. On Aquinas’s view, absent that grace, it is inevitable that we fall. Thus, we might say, it has always been true that human beings (at any rate, all of those who will in fact be created) have suffered from a form of CTWD whose relevant condition is just the absence of whatever gift or grace was initially bestowed upon Adam and Eve. This form of CTWD is not itself original sin; but, as Aquinas might put it, it is the ‘matter’ of original sin whereas the absence of the supernatural gift is the ‘form’.⁴⁷ Moreover, we might add, the first sin of Adam brought it about that the relevant condition was satisfied. That is, it is partly because of Adam’s sin that God chose to withhold the supernatural gift thenceforth from Adam and his progeny. And so the first sin of Adam is among the salient causes of our being such that we will inevitably fall into sin. But, since (by refraining from sinning) we are able to prevent our having ever suffered from CTWD, and since, on this proposal, we have the power to refrain from sinning, A1 is false: there is something we could have done such that, had we done it, we would not have suffered from the corruption that makes it inevitable that we will fall into sin. And so our being held guilty for the fact that it is inevitable that we will fall into sin is not contrary to MR. We have seen, then, two ways in which our central Molinist assumptions M1 and M2 might contribute to the development of a theory of original sin—one Augustinian, the other Anselmian. Both views, however, must come to grips with at least one significant cost (besides commitment to M1 and M2) and one important objection (besides those that might be levelled against M1 and M2). I’ll close this section by discussing each of these in turn. The cost is that neither version of the Molinist view offers any real explanation for the universality of either TWD or CTWD. On the Augustinian version TWD is universal as a consequence of Adam’s sin; but it is hard to see why Adam’s sin should have universal TWD as a consequence. On the Anselmian version, CTWD is apparently universal simply by divine decree (even Adam and Eve suffer from it). The cost is important since, as indicated at the outset of this chapter, one of the main historical functions of the doctrine of original sin has been to explain (ostensibly in a deep, rather than merely superficial, way) the universality of sin. The two versions of the Molinist theory now under consideration purport to explain the universality of sin either by appeal to the universality of TWD or by appeal to the universality of CTWD and the absence of a certain kind of divine grace. But the depth of the explanation is threatened by the fact that it is hard to

⁴⁷ Cf. Aquinas, Summa Theologica, pt I of pt 2, Q. 82, Art. 3 in Aquinas 1945: 676–7.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

48

   :  

see what further explanation could be offered either for the connection between TWD and Adam’s sin or for the universality of CTWD. The cost, I think, is bearable. But reflection on the cost suggests an objection that, if sound, would be harder to bear. So far, I have simply taken it for granted that the universality of either TWD or CTWD would explain, at least in part, the universality of sin. But one might object that in fact this presupposition is false. For (one might argue) what counterfactuals are true of us depends in large part on what we do; thus, it appears that our behaviour explains our suffering from either TWD or CTWD rather than the other way around. If this is right, then if the doctrine of original sin were developed along either of the two Molinist lines I have here suggested, it would be unable to fulfil one of its main historical functions. Perhaps we could live with this; but it would be better if the objection could be shown to be unsound. And I think that it can be. Consider an analogy. The crystal vase is fragile. What this means, in part, is that, under ‘normal’ circumstances, if it were struck (by a suitably hard, suitably fast-moving object) it would break.⁴⁸ But its being struck, even in circumstances that count as ‘normal’, does not entail its breaking: there are worlds where it is struck and does not break. So what shall we say about such worlds? Are they worlds in which the vase is not fragile, or are they worlds in which it is fragile but (miraculously) fails to break? Plausibly, they are worlds in which the vase is not fragile. For, after all, a vase that does not break when struck under normal circumstances is clearly not such that it would break if struck in such circumstances; and so, ceteris paribus, it does not satisfy one of the defining conditions of fragility. Whether a vase counts as fragile, then, depends in part upon what it actually does if and when it is struck; but if it is struck and breaks, its breaking will nevertheless be partly explained by its fragility. Likewise, then, in the case of TWD and CTWD. Those two deficiencies are relevantly like (though perhaps not exactly like) dispositions to sin. To be sure, whether one has it depends in part on what one freely does; but (as in the case of other dispositions) that is consistent with the claim that what one freely does is partly explained by the fact that one suffers from it.

4. Conclusion I have shown in this chapter that there are at least two ways of reconciling the traditional doctrine of original sin with MR, the principle that one is morally ⁴⁸ The qualifier ‘under normal circumstances’ is, of course, hopelessly vague. But I include it simply to signal the fact that I am here ignoring complications that arise from the possibility of more unusual circumstances—e.g. circumstances in which the vase’s disposition to break is masked, or ‘finkish’, etc. Taking account of these issues would add greater complexity to the present discussion but, I think, would not substantially affect my basic point.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

49

responsible for the obtaining of a state of affairs only if that state of affairs obtains and there was something one could have done that would have prevented it from obtaining. The most significant metaphysical commitments associated with the strategies that I have developed are, on the one hand, a commitment to some sort of non-endurantist, probably similarity-based understanding of persistence over time, or, on the other hand, a commitment to the claim that there are true counterfactuals of freedom (including ones with false antecedents) and that it is up to us what counterfactuals of freedom are true of each of us. Neither of these commitments is wildly popular; but, if the arguments in this chapter are sound, embracing one or the other will provide one with metaphysical underpinnings for an MR-friendly development of a fully traditional doctrine of original sin.⁴⁹

References Adams, Robert. 1977. ‘Middle Knowledge and the Problem of Evil’. American Philosophical Quarterly 14: 109–17; reprinted with additional notes in Adams, The Virtue of Faith and Other Essays in Philosophical Theology. New York: Oxford University Press, 1987. Adams, Robert. 1991. ‘An Anti-Molinist Argument’. Philosophical Perspectives 5: 343–53. Adams, Robert. 1999. ‘Original Sin: A Study in the Interaction of Philosophy and Theology’. In F. J. Ambrosio (ed.), The Question of Christian Philosophy Today, 80–110. New York: Fordham University Press. Anselm. 1969. Why God Became Man and The Virgin Conception and Original Sin, translated by Joseph Colleran. Albany, NY: Magi Books. Aquinas, Thomas. 1945. The Basic Writings of Saint Thomas Aquinas, vol. 2, edited by Anton Pegis. New York: Random House. Arminius, James. 1999. The Works of James Arminius, vol. 2, translated by James Nichols. Grand Rapids, MI: Baker Book House. Augustine. 1999. On the Merits and Remission of Sins, and on the Baptism of Infants, in Philip Schaff (ed.), Peter Holmes and Robert Wallis (trans.), Nicene and Post-Nicene Fathers, First Series, vol. 5: Augustin: Anti-Pelagian Writings. Peabody, MA: Hendrickson Publishers. ⁴⁹ Work on this chapter was supported in part by an NEH Summer Stipend (2004). A version of this chapter was discussed at the Metaphysics of Human Persons Workshop (February 2004), sponsored by a grant from the Pew Charitable Trust, and also by the weekly reading group hosted by the University of Notre Dame’s Center for Philosophy of Religion. I am grateful to the participants in these groups for valuable advice and criticism—especially (from the Pew Workshop) Godehard Bruntrup, Peter Forrest, Hud Hudson, Trenton Merricks, Richard Swinburne, Peter van Inwagen, and Dean Zimmerman; and (from the Center for Philosophy of Religion group) E. J. Coffman, Jeff Green, Tom Flint, Carl Gillett, Todd Long, Michael Murray, James Rissler, and Kevin Timpe. I would also like to thank Michael Bergmann, Jeff Brower, and Tom Crisp for their very helpful comments on earlier drafts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

50

   :  

Berkouwer, G. C. 1971. Sin. Grand Rapids, MI: Eerdmans. Brower, Jeffrey. 2004. ‘Anselm on Ethics’, in Brian Davies and Brian Leftow (eds), The Cambridge Companion to Anselm, 222–56. Cambridge: Cambridge University Press. Butler, Joseph. 1849. The Works of Joseph Butler, vol. 1. Oxford: Oxford University Press. Calvin, John. 1986. The Institutes of the Christian Religion, edited by Tony Lane and Hilary Osborne. Grand Rapids, MI: Baker Book House. Catholic Church. 1994. The Catechism of the Catholic Church, 2nd edn. New York: Doubleday. Chisholm, Roderick. 1976. Person and Object. La Salle, IL: Open Court. Collins, Anthony. 1709. An Answer to Mr. Clarke’s Third Defense of His Letter to Mr. Dodwell, 2nd edn. London: J. Darby. Craig, William Lane. 1991. Divine Foreknowledge and Human Freedom. Leiden: E. J. Brill. Craig, William Lane. 2001. ‘Middle Knowledge, Truth-Makers, and the Grounding Objection’. Faith and Philosophy 28: 337–52. Crisp, Oliver. 2003. ‘On the Theological Pedigree of Jonathan Edwards’s Doctrine of Imputation’. Scottish Journal of Theology 56: 308–27. Davies, Brian (ed.). 2001. The De Malo of Thomas Aquinas, translated by Richard Regan. Oxford: Oxford University Press. Edwards, Jonathan. 1992. The Works of Jonathan Edwards, vol. 1. Edinburgh: Banner of Truth. Finch, Alicia and Michael Rea. ‘Presentism and Ockham’s Way Out.’ Oxford Studies in Philosophy of Religion 1: 1 – 17. Fischer, John Martin. 1989. God, Freedom, and Foreknowledge. Stanford, CA: Stanford University Press. Flint, Thomas (ed.). 1998. Divine Providence: The Molinist Account. Ithaca, NY: Cornell University Press. Forrest, Peter. 1994. ‘Inherited Responsibility, Karma and Original Sin’. Sophia 33: 1–13. Hasker, William. 1989. God, Time, and Knowledge. Ithaca, NY: Cornell University Press. Helm, Paul. 1997. Faith and Understanding. Grand Rapids, MI: Eerdmans. Hodge, Charles. 2001. Systematic Theology, vol. II: Anthropology. Peabody, MA: Hendrickson Publishers. Hudson, Hud. 2001. A Materialist Metaphysic of the Human Person. Ithaca, NY: Cornell University Press. Hume, David. 1978. A Treatise on Human Nature, 2nd edn., edited by P. H. Nidditch. Oxford: Oxford University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    

51

Kelly, J. N. D. 1978. Early Christian Doctrines. San Francisco, CA: Harper San Francisco. Lewis, David. 1983. ‘Postscripts to “Survival and Identity” ’. In Philosophical Papers, vol. 1, 73–7. Oxford: Oxford University Press. Luther, Martin. 1976. Commentary on Romans, translated by J. Theodore Mueller. Grand Rapids, MI: Kregel Publications. Merricks, Trenton. 2003. ‘The End of Counterpart Theory’. Journal of Philosophy 100: 521–49. Mohila, Peter. 1975. The Orthodox Confession of Faith, translated by Peter Popivchak, http://esoptron.umd.edu/ugc/ocfi.html. Murray, John. 1955. ‘The Imputation of Adam’s Sin’. Westminster Theological Journal 18: 146–62. Murray, John. 1956. ‘The Imputation of Adam’s Sin: Second Article’. Westminster Theological Journal 19: 25–44. Origen. 1999. De Principiis. In Philip Schaff (ed.), Peter Holmes and Robert Wallis (trans.), Ante-Nicene Fathers, vol. 4: Tertullian, Part Fourth; Minucius Felix; Commodian; Origin, Parts First and Second, 239–384. Peabody, MA: Hendrickson Publishers. Orthodox Eastern Church. 1899. The Acts and Decrees of the Synod of Jerusalem, translated by J. N. W. B. Robertson. London: Thomas Baker. Plantinga, Alvin. 1974. The Nature of Necessity. New York: Clarendon Press. Plantinga, Alvin. 1986. ‘On Ockham’s Way Out’. Faith and Philosophy 3: 235–69. Quinn, Philip. 1984. ‘Original Sin, Radical Evil, and Moral Identity’. Faith and Philosophy 1: 188–202. Quinn, Philip. 1997. ‘Sin and Original Sin’. In Philip L. Quinn and Charles Taliaferro (eds), A Companion to Philosophy of Religion, 541–55. Oxford: Blackwell. Schaff, Philip (ed.). 1998a. The Creeds of Christendom, vol. II, The Greek and Latin Creeds. Grand Rapids, MI: Baker Books. Schaff, Philip (ed.). 1998b. The Creeds of Christendom, vol. III, The Evangelical Protestant Creeds. Grand Rapids, MI: Baker Books. Schreiner, Thomas. 1995. ‘Does Scripture Teach Prevenient Grace in the Wesleyan Sense?’ In Thomas Schreiner and Bruce Ware (eds), Grace and the Bondage of the Will, 365–82. Grand Rapids, MI: Baker Book House. Shedd, W. G. T. 2003. Dogmatic Theology, 3rd edn, edited by Alan W. Gomes. Philipsburg, NJ: P&R Publishing. Sider, Theodore. 2001. Four-Dimensionalism. Oxford: Oxford University Press. Smith, Donald. 2004. Persistence, Persons, and Vagueness. PhD dissertation, University of Notre Dame. Stuchlik, Joshua. 2003. ‘Not all Worlds are Stages’. Philosophical Studies 116: 309–21.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

52

   :  

Swinburne, Richard. 1989. Responsibility and Atonement. Oxford: Clarendon Press. Turretin, Francis. 1992. Institutes of Elenctic Theology, vol. 1, First through Tenth Topics, translated by George Musgrave Giger, edited by James T. Dennison, Jr. Philipsburg, NJ: P&R Publishing. Urban, Linwood. 1995. A Short History of Christian Thought. Oxford: Oxford University Press. van Inwagen, Peter. 1997. ‘Against Middle Knowledge’. Midwest Studies in Philosophy 21: 225–36. Wainwright, William. 1988. ‘Original Sin’. In Thomas V. Morris (ed.), Philosophy and the Christian Faith, 31–60. Notre Dame, IN: University of Notre Dame Press. Watson, Richard. 1834. Theological Institutes; or A View of the Evidences, Doctrines, Morals and Institutions of Christianity, vol. 2. New York: B. Waugh and T. Mason. Wesley, Jonathan. 1978. The Works of Jonathan Wesley, 3rd edn, vol. 9: Letters and Essays. Grand Rapids, MI: Baker Book House. Wiley, Tatha. 2002. Original Sin: Origins, Developments, Contemporary Meanings. New York: Paulist Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

2 Hylomorphism and the Incarnation The Christian doctrine of the incarnation tells us that the Son of God, the second person of the Trinity, became incarnate as a first-century man, Jesus of Nazareth. According to the doctrine, the Son took on a human nature while at the same time retaining his divine nature. Thus, he became a two-natured individual. He also acquired distinctively human parts and characteristics—among them, a human body and soul. What the doctrine does not tell us, however, is what the relations are supposed to be between the various ‘elements’ involved in the incarnation: the divine nature, the human nature, the body of Jesus, the human soul of Jesus, the man Jesus, and the Son of God. The history of metaphysical speculation on the doctrine provides us with a bewildering variety of options. Some identify the human nature with the mereological sum of the body and soul; others insist that natures are abstracta and that the proposed identification therefore makes no sense. Some identify the Son with Jesus of Nazareth and Jesus, in turn, with the sum of his two natures. Others say that the Son merely ‘inhabited’ the human nature (construed as a concrete object) as if it were a sort of outer garment. Some say that the Son functioned as the soul of Jesus’ body and in so doing became a human soul. Others say that Jesus had two souls—one human and one divine. We could carry on with this list for quite a while.¹ Discussion of the relations among the elements involved in the incarnation usually arises in contexts where the main question in view is something like, ‘How can we coherently suppose that a fully divine being has some of the very specific and distinctively human limitations that Jesus is represented as having—e.g. ignorance of the time of the Second Coming, or the “ability” to grow in wisdom?’ Addressing this question usually requires at least a brief foray into controversies about the nature of natures and about what, exactly, would be involved in having multiple natures; but, in the contemporary literature at any rate, these issues tend not to take centre stage. It is also common for views about the metaphysics of the incarnation to be developed independently of views about the metaphysics of the Trinity—in some cases with the result that an author’s views about the incarnation

¹ For a valuable critical survey of the major positions in the contemporary and historical literature, including references to the sorts of views just mentioned, see Cross 2009. See also Adams 2006, ch. 5; Crisp 2007, ch. 2; and part I of Cross 2002 for further explanation and critical discussion of some of these views.

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0003

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

54

   :  

are in tension with his or her views about the Trinity.² Both tendencies in the literature seem methodologically problematic, the latter especially so. My goal in this chapter, then, is to provide a metaphysical account of the incarnation that starts from substantive assumptions about the nature of natures and about the metaphysics of the Trinity and that develops in light of these a story about the relations among the elements involved in the incarnation. Central to the view I will describe are two features of Aristotle’s metaphysics, though I do not claim that my own development of these ideas is anything of which Aristotle himself would have approved: (i) a hylomorphic understanding of material objects, (ii) a doctrine of numerical sameness without identity, and (iii) the view that the nature of a thing can appropriately be identified with its form. These ideas, along with other important aspects of the metaphysical framework with which I shall be working, are laid out in Sections 1–5, followed in Section 6 by a brief sketch of the account of the Trinity that Jeffrey Brower and I have presented in detail elsewhere. In Section 7, I present my account of the incarnation.

1. The Basic Framework and the Neo-Aristotelian Theory Central to Aristotle’s metaphysics is the idea that every material object is a structured entity with two constituents, matter and form. Explanations of these concepts commonly encourage the thought that matter is the stuff of which something is made whereas a form is a property, like humanity or felinity, instantiation of which accounts for an object’s being the kind of thing that it is. It is furthermore common to characterize Aristotelian forms as abstract, immanent universals. I won’t comment here on the extent to which these common characterizations are correct as interpretations of Aristotle’s views.³ I note them simply to acknowledge awareness of them, even as I depart from them in articulating my own story about matter and form.

² For example, advocating the view that human persons are material whereas divine persons are (normally) immaterial, Trenton Merricks (2007) argues that the incarnation is best understood as involving the Son of God becoming a material object—namely, Jesus of Nazareth. Yet, with regard to the Trinity, he argues that the relation between the divine persons is to be understood on analogy with the distinct spheres of consciousness of a split-brain patient (Merricks 2006). It is, at the very least, difficult to see how the incarnate second person of the Trinity (a material object) could be related to the (immaterial) other persons of the Trinity in the way that the spheres of consciousness of a split-brain patient are related. Or, to take another example: Thomas Flint, in the conference version of his 2009 paper, sets up the central problem by assuming that the Son of God, the second person of the Trinity, is identical to the divine nature of the incarnate Christ. But, of course, if the Son is identical to the divine nature then it is very hard to see how the Son could also share his divine nature with the Father and the Spirit. ³ But see, e.g. Witt 1989, wherein it is argued that forms are not properties and not universals but rather, individuals. Cf. also Koslicki 2008: 252ff.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

55

Natures, for Aristotle, are internal principles of change and rest; and talk of the nature of a material substance will primarily involve reference to its form. Although Aristotle acknowledges that both matter and form each have some claim to being called ‘the nature’ of a material substance, he indicates that form has the greater, or primary claim.⁴ What tips the scale in favour of form is, in a nutshell, the fact that forms, on his view, determine kind-membership for a substance and provide more fundamental explanations of its behaviour and development.⁵ For material substances, then, sharing a nature will primarily involve sharing a common form. Within Aristotle’s metaphysics, matter is shareable as well. The best-known examples used to illustrate this involve material substances (e.g. Socrates) and substance-accident compounds (e.g. seated-Socrates, which exists when and only when Socrates is seated) that are made of the very same stuff.⁶ Things that share matter in this way are, on Aristotle’s view, numerically the same, but not identical—of which more below. All that I have said in this section thus far (minus what I have called the ‘common characterizations’ of matter and from) is what I shall have in mind when I talk below of the ‘basic (Aristotelian) framework’. Like many philosophers, both contemporary and historical, I think that the basic framework offers fruitful resources not only for solving central problems in metaphysics, but also for explicating and solving problems with central doctrines of Christianity. My own use of it, however, depends on thinking of matter and form in such a way that even immaterial things might be said to have a matterform structure. As shall become clear below, I also find it useful to allow that one and the same thing can be the form of one object and the matter of another. Thus, as I have already indicated, I find it difficult to endorse, without a lot of qualifying assumptions, the idea that matter is stuff and forms are kind-properties. I also find it difficult, for other reasons, to endorse various other familiar claims about matter, forms, and natures—e.g. that matter is ‘potentiality’, or that forms (and therefore natures) are final causes or ‘principles of life’.⁷ One reason why I am not inclined to endorse these claims is that the central terms, like ‘potentiality’ and ‘final cause’, either are, or are explained in terms of, primitive concepts that I think many contemporary metaphysicians will find unintelligible. Though I am somewhat sceptical of claims to the effect that ‘my primitives are more intelligible than yours’, I would prefer to develop a metaphysic that retains some of the central

⁴ Physics, II.1, esp. 193b8–19. See also Metaphysics, IV.4 1014a35–b19 and Metaphysics, VII.3 1029a5–7, and Loux 1991: 81–2 and ch. 5. ⁵ Cf. Waterlow 1982: 58–66 and Witt 1989: 65–79. These passages in Waterlow and Witt focus primarily on the Physics. On the case for form in the Metaphysics, see Loux 1991, esp. chs 3 and 5. ⁶ See, e.g. Topics, I.7 103a23–31 and Metaphysics, IV.6 1015b16–27. Perhaps all of Aristotle’s examples are like these, but I cannot claim to have examined the entire corpus. ⁷ Cf. Witt 1989: 68, 126ff.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

56

   :  

aspects of the basic Aristotelian framework (in particular, a place for some of the central roles associated with the concepts of matter, form, and nature, as well as the relation of numerical sameness without identity) but whose explanations invoke primitives whose intelligibility is more widely acknowledged by contemporary metaphysicians. For the purposes of this chapter, then, I want to work with what is probably best thought of as a neo-Aristotelian theory of natures and substances. This theory can be roughly divided into two parts: the Aristotelian part and the neo part. The ‘neo’ part of the theory attempts to forge connections between contemporary ideas about powers and fundamental properties on the one hand and, on the other hand, certain Aristotelian ideas about natures and forms—namely, that whatever plays the role of form is also suited to play the role of nature, that natures are intimately connected with the distinctive powers and capacities of their corresponding natural kinds, and that natures are principles of unity. Let us start, then, with the Aristotelian part of the theory. (T1)

Every substance that is not a nature is a compound of matter and form.

(T2)

Forms are constituents of objects, not transcendent universals.

I have discussed these two claims to some extent already at the beginning of this section; but let me add here just two further comments. First, note that T1 allows that natures are substances. I believe that this is Aristotle’s view as well, but there are controversies in the neighbourhood that I have neither the space nor the expertise to discuss in appropriate detail.⁸ I do not here mean to take a position on those controversies. Second, hylomorphic compounding is not the same as composition, or mereological summation.⁹ In my terminology, matter and form are constituents of material objects, but not parts. I don’t suppose that there is any pre-theoretical distinction to be drawn between parts and constituents. But I do think that there is a substantive, and probably pre-theoretical, distinction to be drawn between composition (or summation) construed as an operation on concrete particulars and the sort of compounding that might take place between stuff and a form or between a concrete particular and a form. Even a child can grasp the idea of imposing a form upon a piece of clay; but it would be a substantially further step—and one that I am not inclined to take—to regard the form of a statue as one ⁸ But see, for starters, Loux 1991. ⁹ Whether composition and mereological summation are the same depends in part on whether one wants to reserve the term ‘mereology’ for, e.g. classical extensional mereology while at the same time denying that composition obeys the axioms of that system. My point here is just that compounding should not be assimilated to either. Alternatively, one might endorse compositional pluralism, the view that there are multiple fundamental composition relations; and one might then say (e.g.) that mereological summation and compounding are two such relations. (Cf. McDaniel 2009, as well as the remarks about Fine in note 10.) I am not, in principle, opposed to this way of talking; but it does represent a different terminological choice from what I have opted for here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

57

of its parts.¹⁰ The terminological distinctions between parts and constituents and between summation and compounding are meant to help us keep track of that difference. The neo part of the theory involves three further claims about the nature and function of natures: (T3)

Natures are fundamental powers.¹¹

(T4) The natures of composite objects unite other powers (in particular, the powers that are the natures of their parts). (T5) Natures can enter into compounds with individuators, and with distinguishing properties. In compounds with individuators, natures play the role of form; in compounds with distinguishing properties, they play the role of matter. Each of T3–T5 requires comment and, in the case of T4 and T5, explication of central terms. I’ll take each in turn.

2. Natures as Fundamental Powers In saying that natures are fundamental powers, I mean three things. First, they are perfectly natural properties—not in the sense that contrasts with ‘supernatural’, but rather in the sense of marking objective similarities and joints in nature (cf. Lewis 1983). Second, they are not reducible to other powers. The power to tell a lie, for example, is reducible (if it is a genuine power at all). It is nothing over and above the more basic powers involved in its exercise: the power to form beliefs, the power to speak, the power to entertain false propositions and to intend to report them as true, and so on. Negative charge, on the other hand, is plausibly nonreducible, and so fundamental. Third, they ground non-natural powers or, if there are no such things, they explain the truth of (putative) non-natural powerattributions. For example: Fundamental particles have the power to repel other fundamental particles. If there really is such a thing as the power to repel other fundamental particles, presumably it is a non-natural power that is grounded in one of two plausibly fundamental powers—negative or positive charge. (For purposes here, I’ll follow George Molnar in my understanding of grounding:

¹⁰ My hylomorphism differs in at least this respect, then, from that of Johnston (2006) and Koslicki (2008). Fine (2008) suggests that forms are parts, but not the sorts of parts that are joined by fusion (cf. also Fine 1999). Perhaps, then, our difference on this score is merely terminological. There are, of course, other differences, however. ¹¹ I have located ‘T3’ in the ‘neo’ section not because I think that Aristotle would disagree with it (I don’t). Rather, I have located it here simply to avoid having to defend the claim that it belongs in the ‘clearly Aristotelian’ part.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

58

   :  

‘The ground of a power, P, is the set of properties (all of which are conceptually distinct from P) by virtue of which a thing has P’ (2003: 147)). On the other hand, if there is no such thing as the power to repel other fundamental particles, still, the claim that fundamental particles have that power will be made true by facts about fundamental powers; and so the fundamental powers will explain the truth of that power-attribution. I take it that the difference between reducible powers and nonreducible non-natural powers roughly corresponds to the difference between conjunctive properties and irreducibly disjunctive properties.¹² I should also say what I don’t mean by the claim that natures are fundamental powers. Most importantly, I don’t mean to say that natures are ‘basic’ in the sense of being entities upon which all other things depend for their existence. For example, it might turn out that the natures of composite objects depend in some sense upon the natures of their parts. Also, I don’t mean to take any position on the question of whether there are fundamental empirical properties beyond those investigated by physics. If there are, then perhaps there are biological and chemical natures as well as physical natures; or perhaps there are mental natures. If there aren’t, then the only natures exemplified by material things are natures of physical objects. Of course, the doctrine of the incarnation is committed to the existence of human nature and the nature of God. But I don’t think anyone would say that the divine nature is a fundamental empirical property; and it is an open question both philosophically and from the point of view of traditional Christian doctrine whether humanity is. In saying that natures are powers and that natures can play the role of form, it might seem that, from the point of view of traditional hylomorphism, I have got my metaphysics upside down. For powers, one might argue, are dispositions, or potencies, not ‘acts’ or ‘manifestations’; but the role of ‘potency’ is commonly associated with matter, whereas the role of ‘act’ is commonly associated with form. Addressing this concern in detail would take us too far afield; but let me offer just two brief remarks. First, Aristotle himself distinguishes between active (causal) power and mere potentiality, and it is the latter, not the former, that is associated with the matter role.¹³ This makes intuitive sense, too. Being feline, for example, might just be a certain complex power or capacity to develop and behave in certain ways, even if, at the same time, felinity is the actualization of the potentiality on the part of some matter for being a cat. Second, the metaphysic of properties with which I am working in this chapter is a version of dispositional monism, according to which all properties (and hence all manifestations of dispositions) are powers,

¹² This is not to say, of course, that there is no overlap between the two; i.e. if there are indeed both conjunctive powers and irreducibly disjunctive powers, then surely there are irreducibly disjunctive conjunctions as well. ¹³ Witt 2003, esp. ch. 2.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

59

and therefore dispositions. There are, of course, well-known difficulties for dispositional monism, not least of which are the threat of infinite regress and the (related) problem of ungrounded dispositions. I cannot possibly hope to articulate and defend my own responses to these problems here; but I can at least refer interested readers to promising responses already available in the literature.¹⁴

3. Natures as Uniting Other Powers Thesis T4 is meant to express and do justice to that part of traditional hylomorphism that says that natures are principles of unity. The relation of uniting is to be understood roughly as follows: one power—a nature—unites some other powers just in the case that the nature is so connected to the other powers that its manifestation depends upon the cooperative manifestation of the united powers and, furthermore, the latter do not confer any powers on the object that has the nature that are both intrinsic to the object and independent of the nature. A bit more precisely: A power p₀ of an object x unites distinct powers p₁ – pn = df (i) p₀ is intrinsic to x,¹⁵ (ii) each of p₁ – pn is a nature of at least one of x’s parts, (iii) p₀ is grounded in or identical to a certain sort of cooperative manifestation (CM) of p₁ – pn,¹⁶ (iv) every power intrinsic to x that is at least partly grounded in CM is identical with, reducible to, or at least partly grounded in p0, and (v) there is no power intrinsic to x that is distinct from both p₀ and CM and that grounds p₀. Consider a human organism, for example, and suppose that humanity is indeed a biological nature. The manifestation of humanity in a region depends causally upon the cooperative manifestation of the natures of the simple parts of the human organism. Not just any sort of cooperative manifestation will do, however. Take all of the simple parts of a human and force-fit them into a one-quart cylindrical container and you will not have a human organism, even if, at that time, the natures of the erstwhile parts of the human being are engaged in some sort of cooperative manifestation. Thus, the presence of humanity in a region depends upon a particular sort of cooperative manifestation of the natures of the

¹⁴ See Bird 2007 and Molnar 2003. ¹⁵ This might be a problem if, as some think, dispositions—and therefore powers—turn out to be extrinsic. Jennifer McKitrick (2003), for example, argues for the thesis that dispositions are extrinsic. Molnar (2003), on the other hand argues that they are intrinsic, and Bird (2007) provides replies to McKitrick’s arguments. Obviously I’m taking sides with Molnar and Bird. ¹⁶ In other words: Let CM be a property such that, necessarily, CM is had by an object x iff p₁ – pn manifest in a particular sort of way. Then x’s having p₀ is grounded in or identical to x’s having CM.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

60

   :  

relevant parts. (Perhaps it just is that sort of cooperative activity. I don’t commit to this; but neither do I intend to rule it out.) Finally, every power intrinsic to a human being whose manifestation depends upon the relevant sort of cooperative activity—e.g. the capacity for rational thought, the power to grow and develop as a human organism, the power to run and dance—is plausibly dependent upon the power that is humanity; and humanity is not itself so dependent upon some further uniting power.¹⁷ Thus, humanity unites the natures of the parts of a human being in the sense described above. This is the sense in which I think that humanity is a ‘principle of unity’, and it is part of what is involved in its being a nature. At this juncture, it may be helpful briefly to contrast my own understanding of principles of unity with another one available in the literature. Mark Johnston, in developing his own version of hylomorphism, also regards forms as principles of unity; and he regards principles of unity as relations. Thus, for example, he writes: Consider HCl, a kind of molecule. The principle of unity for individual hydrogen chloride (HCl) molecules is the relation of bipolar bonding. . . . The principle of unity holds of the ions, and its holding is the essential condition for existence of the molecule. It is simply an essentialist elaboration of a proposition of chemistry that what it is for a given hydrogen chloride molecule to be is for there to be a hydrogen ion and a chlorine ion together in a bipolar bond. (2006: 653)

On the assumption that relations can be identified with polyadic properties and properties with causal powers (both controversial, of course) Johnston’s version of hylomorphism comes out in one respect very similar to my own: principles of unity, and so the forms of material objects, turn out to be powers. But it should also be clear that, on my view, relations like bipolar bonding will not at all be the right sorts of powers to function as natures. To see why, one need only attend to the fact that HCl and CO both have their parts united (in Johnston’s sense) by the bipolar bonding relation, and yet they don’t share a nature. As I see it, a nature should be something that unites the powers of the parts of the object in the sense described above. But it is hard to see bipolar bonding as playing that role. HCl and CO have very different causal powers, after all; thus, even if bipolar bonding satisfied conditions (i–iv) of the definition of power-uniting, it is hard to imagine that it would satisfy condition (v). Presumably condition (v) is satisfied in each case by a more ‘all-encompassing’ power—i.e. a power that we would normally name by reference to a kind, like being hydrogen chloride, or being carbon monoxide.

¹⁷ I don’t, of course, mean to suggest that these powers are unique to human beings; only that they are intrinsic.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

61

4. Natures, ‘Individuators’, and ‘Distinguishing Properties’ According to the version of hylomorphism that I am developing, the natures of material objects play the role of form, and they enter into compounds with things or stuffs that play the role of matter. On one common way of understanding the roles of form and matter, forms are constituents that are shared among objects of the same kind, whereas matter is what individuates objects of a kind. I can unqualifiedly endorse the claim about forms, but not the claim about matter. In accord with the spirit of that claim, I want to say that (for material objects, anyway) what natures enter into compounds with are individuators. But for those of us who endorse the view that distinct objects might nevertheless share the same matter in common, and for those of us who believe that immaterial things might have a hylomorphic structure, the claim that ‘matter is what individuates objects of a kind’ cannot be affirmed without qualification. Individuators can’t be seen as accounting for distinctness, since distinct items can have the same individuator as a constituent. Furthermore, matter can’t always be what individuates, since immaterial things have no matter, strictly speaking. (Another way of putting this: it is not always matter, literally speaking, that plays the matter-role.) Let me therefore say a few words about how I’m thinking about individuators, about matter-sharing, and about the hylomorphic structure of immaterial things. Let us begin by considering simple material objects. A simple material thing—a point-sized particle, for example—will have a nature which it shares with other particles of the same kind. But what accounts for the fact that there are many particles with the same nature rather than just one scattered, mereologically complex particle located wherever we find the nature in question? The answer, I take it, is just this: The nature itself doesn’t divide across disconnected regions of spacetime; it is a power that must be, in some sense wholly or fully concentrated at, or attached to, point-sized regions (or, better, line-sized ones for the typical case of a moving, spatially point-sized particle) rather than to scattered regions. In light of this, it is natural to suppose that, in the case of each particle of the relevant kind, the nature has compounded with some further constituent that accounts for its concentration at or attachment to the region in question. From here, the story might be fleshed out in a variety of different ways. My own inclination is to think that, for simple objects, the individuators are regions—presumably point-sized, but perhaps not—of spacetime. I think of the powers of simple material objects as physically locatable properties—qualities that exist at multiple regions of spacetime. Thus, it is natural to regard the objects themselves as compounds whose matter is a spacetime region and whose form is the quality located at that region. What about complex material things? Here, it seems, we have fundamentally the same situation, but on a larger scale. What is it that accounts for the fact that there are many distinct human beings rather than just one spatiotemporally scattered human being? Presumably it is that human nature doesn’t divide widely

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

62

   :  

across spacetime; it is the sort of power that is fully and completely located in what we think of as human-shaped regions.¹⁸ More precisely: it is a power that unites only the natures of certain kinds of objects standing in certain kinds of relations, and it is this fact that explains why human nature concentrates at human-shaped regions rather than scattered regions. Unlike the case of simple particles, however, it is not so natural to suppose that the individuating constituent of a human being is its region of spacetime. The reason is just that there is an intuitively better candidate available—namely, the collection of objects whose powers are united by the nature. (By using the terms ‘a better candidate’ and ‘the collection of objects’, I don’t mean to commit to the claim that ‘the collection’ is an individual thing distinct from the human being it constitutes.) Thus, in general, the individuators for mereologically complex things will just be collections of objects whose powers are united by the natures of those things. But what if we believe that distinct things share all of the same matter in common? Suppose, for example, you think that a clay statue is distinct from the lump of clay that constitutes it. (Perhaps you think that the lump can survive things—squashing and reshaping, for instance—that the statue cannot, and that by virtue of this difference, the lump cannot be identical to the statue.) In that case, it will be at least somewhat misleading to say that matter individuates. Matter would not be what accounts for the distinctness of the two objects. Nevertheless, if you also believe—as I do—in a relation of numerical sameness without identity, you might think that there is still a perfectly good sense in which matter does individuate. Proponents of numerical sameness without identity say that, in the statue/lump case, though the statue and the lump are distinct, they nevertheless count as one material object. They are two hylomorphic compounds, two things, two entities, etc.; but they are, nevertheless, one material object. Thus, on this view, material objects are individuated by their matter. That is, whether x and y count as the same material object just depends on whether they share the same matter. I have defended this view elsewhere and do not have the space to rehearse that defence here (1998b). But what I want to note is just this: If one has this sort of view, the right thing to say about individuators is not that they account for distinctness simpliciter, but rather that they account for absence of numerical sameness. In the case of material objects, matter plays that role. (What, then, accounts for distinctness simpliciter? Perhaps divergence of properties.) We have considered simple material things and complex material things. But there is one further sort of case we must consider given our present interest in the Trinity and the incarnation. What should we say about simple immaterial things? ¹⁸ But, one might wonder, how is it located at those regions? For some such region R, does the nature entend, pertend, or span the region? (Roughly, this question boils down to the following: ‘Is the nature wholly present at every subregion of the region in question, or does it have parts at every subregion, or neither?’ For definitions of entension, pretension, and spanning, and related notions, see Hudson 2005: 99–101. See also Parsons 2007.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

63

The nature of a simple immaterial thing will not be located anywhere in spacetime, so there is no literal sense in which matter could individuate such things. One option then is simply to deny that immaterial things have a hylomorphic structure. If we say this, then we can go on to say one of the following three things: that they have matter but no nature, that they have neither matter nor nature, or that they are identical to their natures. Affirming the first option would be extremely bizarre if what motivates us to deny that immaterial things have a hylomorphic structure is the fact that they have no literal matter. Taking the second option commits us to the view that some substances have natures, but not all of them do. It is hard to see what the motivation for such a view might be. Thus, it seems that the last alternative would be the clear choice. But suppose we believe in immaterial things that share a nature? We cannot appeal to matter or spatiotemporal separation or anything of the sort to individuate them—i.e. to account for an absence of numerical sameness among them. Thus, a natural thing to say, given that we already accept a relation of numerical sameness without identity, is to say that immaterial things that share a nature are, on that account, one in number, and they are distinguished from one another by their possession of some further property—let us call it, uncreatively, a distinguishing property. Note, however, that on this view there is a clear sense in which the nature of an immaterial thing does, or can, play the role of matter. Just as matter-sharing explains numerical sameness among distinct material objects, so too naturesharing explains numerical sameness among distinct immaterial objects. Likewise, the property (perhaps complex or conjunctive, or perhaps a mere thisness) that accounts for the distinctness of these objects plays at least part of the role of form. Differences in form account for major differences in character among material objects; and, in the case of two distinct things that count as the same material object, it will be differences in form that fundamentally account for the distinctness of those things. And so too in the case of the distinguishing properties of immaterial things that share a nature in common. Thus, it turns out that immaterial objects have a hylomorphic structure after all.

5. Further Terminology Having now finished commenting on theses T1–T5, I want to close my discussion of hylomorphism as such by explaining (in light of the foregoing) my use of terms like constituent, matter-form compound, matter, and form: x is a constituent of y = df x plays in y the role of matter or the role of form. x is a matter-form compound = df x has constituents that play the roles of matter and form.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

64

   :  

m plays the role of matter in x = df (i) m is an individuator that exactly overlaps x or, if there is no such individuator, m is a nature of x and (ii) it is not possible for a distinct thing to share m with x without being numerically the same as x. f plays the role of form in x iff f is a nature of x (or a sum of natures of x, in the case of multi-natured beings), or a distinguishing property instantiated by x, and f does not play the role of matter in x.¹⁹ As noted earlier, I allow that mere pluralities—collections—can function as individuators, and hence can play the role of matter. But, in light of the above definitions, this poses a small terminological problem. Consider the collection of particles, C, that plays the role of matter in some particular cat. Suppose the xs are the members of the collection. Given that C is a mere collection, ‘C’ is just a device for referring collectively to the xs. Thus, the xs play the role of matter. But now do we say that the xs are constituents of the cat? If we do, the definitions imply (falsely) that each of the xs plays the role of matter in the cat. If we don’t—if we say, instead, that the xs are a constituent of y—then we violate grammar. There is no substantive issue here. We just need to recognize that, just as ‘the collection’ is only apparently a singular referring expression, so too ‘being a constituent of the cat’ only apparently picks out a role that can be played just by an individual. (Analogously: We can’t grammatically say that the members of the Notre Dame football team are the winner of Saturday’s game. We have to say either that the team itself was the winner or that the members of the team won. The problem in the case of constituency is just that there is no corresponding verb analogous to ‘won’. We could invent one; but once the point here has been appreciated, there is no need to.) There is one additional notion that ought to be introduced here as well: the notion of consubstantiality. It is important to talk about consubstantiality because, according to the tradition, Jesus is supposed to be consubstantial with the Father with regard to his divinity and consubstantial with us as regards his humanity. It is generally taken for granted that the substance or essence of a thing is its nature, and that what it is for objects (material or immaterial) to be consubstantial is for them to share a nature. I affirm all of this. Given what I have said above, distinct but consubstantial immaterial things will stand in the relation of numerical sameness without identity—they will be counted as one F, where F specifies the nature they

¹⁹ In the case of multi-natured beings, I say that the two natures (or more) comprise one form because, as indicated above, different forms account for the distinctness of objects that share the same matter in common. Thus, if Christ’s two natures were two forms that inform the same matter, then there would be no way to account for the fact that Christ is a two-natured person who is distinct both from the compound of his matter plus the human nature and the compound of his matter plus the divine nature.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

65

share in common.²⁰ But consubstantial material things will not stand in that relation: matter-sharing (in general), not nature-sharing (in general) is what puts things into the relation of numerical sameness without identity. Since God is an immaterial thing, it will follow from all of this that x is the same God as y if and only if x and y share a divine nature (and there will be exactly one God if and only if there is exactly one divine nature, as traditional Christian doctrine maintains). Obviously we are now in a position to say something about the metaphysics of the Trinity.

6. The Trinity So let me provide a brief sketch of the ‘constitution model’ of the Trinity that I favour. In several other papers (Brower 2004a, 2004b; Brower and Rea 2005; Rea 2003, 2006, 2007) Jeff Brower and I have tried to motivate this model by highlighting its intrinsic virtues, defending it against objections, subjecting its main rivals to criticism, and arguing that the view is at least consistent with and in salient respects similar to views defended by some of the most important patristic and medieval figures writing on the Trinity. For present purposes, however, I’ll leave all of that aside and simply present the main lines of the model. According to the model I favour—‘Constitution Trinitarianism’—each divine person is an immaterial substance whose constituents are the divine nature (or substance) and a ‘person-making’ property (let’s refer to the relevant properties as Fatherhood, Sonship, and Procession). The divine nature plays the role of matter, and the person-making properties each play the role of form in the senses just described. Since each divine person is a substance, the persons cannot be regarded as mere aspects of a common substance, and so the heresy of modalism is avoided. Moreover, though the persons are distinct from one another, they nevertheless count as one God and as numerically the same divine substance, and so the heresy of tri-theism is avoided. This, again, is because immaterial things are individuated by their nature, and Father, Son, and Holy Spirit share the same divine nature. The relation between the persons of the Trinity is, therefore, analogous to the relation of material constitution—the relation between objects that share all of the same matter at the same time. Thus, on this view, there is just one divine substance, and so the view allows us to affirm, along with the creed, that there is ‘ . . . one God, the Father almighty . . . and . . . one Lord Jesus Christ . . . begotten, not made, being of one substance with the Father . . . ’. ²⁰ This, of course, implies that if (say) angels are immaterial and share a common nature, then all angels are the same angel, even if they are distinguishable under some other sortal. One way to avoid this consequence is to suppose that ‘angel’ isn’t a natural kind, and that each of the beings we characterize as angels in fact has its own distinct nature. Another possibility is to deny that angels are immaterial. But if these options are somehow untenable, I am content to accept the consequence.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

66

   :   Elsewhere I have said this about the model: What is it that plays the role of matter in the Trinity? And is it a substance itself? Here I want to offer only a partial view that might be developed in a variety of different ways. What plays the role of matter in the Trinity is the divine nature; and the divine nature is a substance. It is not a fourth substance, for reasons already discussed; nor is it a fourth person (since it is not a compound of ‘matter’ plus a person-defining-property). But it is a substance, since (again, taking cues from Aristotle) natures are substances. What I don’t want to take a position on here is the question of what, exactly, a nature is. Is it concrete or abstract? Is it particular or universal? Is it a property or something else? These questions I will not answer. I think that they must be answered in a way that allows the divine persons to be concrete particular non-properties; but I think that there are various ways of answering these questions that are compatible with that view. (Rea 2007: 420)

In the present chapter, obviously enough, I am not leaving all of these questions open. I have come down in favour of the view that natures are powers. I have not said whether natures are universals or particulars—I have said that they are sharable and multiply locatable, but nothing beyond that. This might suggest that natures are universals; but I don’t want to commit to this, partly because I am unsure whether I buy into the universal/particular distinction.²¹

7. The Incarnation I turn now, at last, to the metaphysics of the incarnation. Recall our central question: How are the elements involved in the incarnation—the divine nature, the human nature, the body and human soul of Jesus, the man Jesus, and the Son of God—related to one another? I’ll begin with some abbreviations that I think will make the presentation of the model go more smoothly; then I’ll lay out the model. Here are the abbreviations: ‘X++Y’ abbreviates ‘the matter-form compound whose matter is X and whose form is Y’ M = Jesus’ physical matter DN = the divine nature S = the Son ²¹ On this issue, see MacBride 2005 and Ramsey 1925.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

67

Sonship = the person-making property in the compound that is the Son HN = humanity—Christ’s (and our) human nature Now for the model. As should already be clear, I do not identify the human nature of Jesus (as some do) with either his body, or his soul, or the mereological sum of his body and soul. The human nature of Jesus is a power, and it is something that he is supposed to share with us. On the present view, then, humanity—HN—is one of two formal constituents of Jesus (the other being DN), and Jesus himself is a compound of those formal constituents and a material constituent—M. One might wish to identify M with Jesus’ body; or one might wish to identify Jesus’ body with the compound, M++HN. Which alternative one chooses depends in part upon what one says about Jesus’ soul. What should one say about Jesus’ human soul? At the very least, I want to say that, for all x, x has a human soul iff x is identical to a matter-form compound whose form is a human nature. But it is not clear to me what more to say. I am reluctant to take on commitments because I suspect that with minor modifications here and there, my model is indifferent between a variety of options (including an option according to which ‘having a soul’ doesn’t imply the existence of anything that would ordinarily count as a soul). That said, though, here is an option that I mostly like: A human soul is something with two modes of existence—material and immaterial. In the immaterial mode, the soul is a compound whose matter is the human nature and whose form is some distinguishing property—perhaps a thisness. In the material mode, the human nature plays the role of form in a compound whose constituents are humanity and some matter, and the (same) distinguishing property becomes simply a property of the whole compound. One likely consequence of this view is that Jesus’ human soul is identical to the compound that was the Son prior to the incarnation. The reason this is a likely consequence is that there is no candidate other than Sonship in the incarnate Jesus for being the relevant ‘distinguishing property’. Thus, on this view about human souls, what happens in the incarnation is that the compound that was the Son becomes the human soul of Jesus. Since this view is consistent with—indeed, entails—that the incarnate Christ has a human soul, there is nothing obviously heretical about it. And I find it intuitively rather satisfying as an account of what the incarnation fundamentally involves. A consequence of this view that I strongly dislike, however, is that all human souls that exist in the immaterial mode count as the same human being.²² On the other hand, views in the neighbourhood of this pervade the patristic and medieval literature on original sin; so perhaps it deserves to be taken seriously. At any rate, the consequence is mitigated by the fact that God can easily see to it that no

²² Or, at any rate, the same something, if being human requires having a body.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

68

   :  

human being ever exists in that immaterial state by seeing to it that the relevant distinguishing property is only ever exemplified by a genuine matter-form compound. (This might be part of why resurrection is so important.) Leaving aside concerns about the nature of the soul, what should we say about the relations between Jesus, the Son, and the two natures of Jesus? On the present model, Jesus is identical to the Son; and prior to the incarnation, the Son is identical to the matter-form compound whose constituents are the divine nature and the person-making property, Sonship. After the incarnation, the Son is identical to a matter-form compound with the following features: (i) his matter is M, (ii) his form comprises two natures, DN and HN, and (iii) he has the individuating property, Sonship. One might worry that this view implies the following claim, which appears at first glance to be incoherent: (C1)

S = DN++Sonship = M++(DN&HN)

One might also worry that Sonship seems simply to disappear in the incarnation. To the latter worry, I reply that Sonship doesn’t disappear; it simply ceases to be a constituent. Prior to the incarnation, Sonship is the formal constituent of S; after the incarnation, it is merely a property of S. To the former worry, I reply that the claim is implied, but it is not incoherent. If ‘++’ represented mereological summation (understood according to standard axioms of mereology), then the claim would be incoherent, since standard mereology includes an extensionality axiom. But there is no reason to think that hylomorphic compounding obeys the same rules as mereological summation. Compounds are not defined by their constituents—or, at any rate, to say that they are is to add a substantive and controversial thesis to one’s hylomorphic theory. One might also worry that C1 violates Leibniz’s Law. After all, DN++Sonship has the property having DN as matter whereas M++(DN&HN) lacks that property. I reply that the ‘problem’ here is to be solved in whatever way we solve ordinary problems with material change. Let MF be the eight-pound lump of matter that is Fred’s matter when he is newborn. Now, Newborn Fred has the property having MF as matter; Adult Fred lacks that property. But, as is well known, we can solve the problem by indexing either the property or the having of it.²³ One might furthermore worry that, post-incarnation, the Son is no longer the same God as the Father, since Father and Son no longer share the same matter: the ²³ For discussion, see (e.g.) Rea 1998a. There are other ways of solving the problem in Fred’s case, too: invoke temporal parts, appeal to presentism, or (following a recent suggestion of Jeffrey Brower 2010) account for the change by positing different compounds (i.e. Newborn Fred and Adult Fred) which have different properties but have Fred himself as a common constituent. Each of these strategies, however, seems problematic as applied to the incarnation.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

69

Son has M for matter whereas the Father has DN. I reply that, because the compound DN++Sonship is present in Jesus—because Jesus still has as a constituent the very thing that plays the role of matter in the Father—that is sufficient for Jesus still to stand in the immaterial analogue of the matter-sharing relation with the Father. (And likewise for the Spirit.) Thus, the doctrine of the Trinity is preserved. Moreover, because HN is present in Jesus, he shares a nature with us and so counts as human; and because Jesus exemplifies Sonship, he counts as the Son incarnate. Finally, one might object that the official story about the incarnate Christ is not consistent with what I am inclined to say about ordinary cases of material constitution. Consider, for example, a block of marble (‘B’) that constitutes both a statue (‘X’) and a pillar (‘Y’). Since the pillar could survive erosion that would destroy the statue, X and Y are distinct. Thus, it looks like we have two matterform compounds that share the same matter. So far so good. But now let F1 be the ‘statue form’ and let F2 be the ‘pillar form’. Given what I say about the incarnation, it seems that I ought to be willing to say that, in the region occupied by the statue, there is a two-natured object—a statue-pillar, perhaps—whose matter is B and whose two natures are F1 and F2. In other words, I ought to say that there is a thing S* such that S* = B++(F1&F2). But it looks as if I do not say this—I say, instead, simply that X and Y stand in the relation of numerical sameness without identity. So, one might think, I am not giving uniform treatment to similar cases. There are two ways to reply. One is to take two-natured beings to be metaphysically rare and special, and to insist that (unlike in the case of the incarnation) in the case of X and Y, there is simply no reason to believe that there is a third thing, S*, that has two natures and stands in the relation of numerical sameness without identity with them. The view, then, would be that sometimes when two natures inform the same matter, a two-natured being results, but most of the time that doesn’t happen. One would then appeal to divine revelation (i.e. the Christian scriptures and the arguments that move from them to the doctrine of the incarnation) as our reason for thinking that the incarnation is one of the special cases where a two-natured being results. Alternatively, one might say that two-natured beings aren’t all that special after all—that, e.g. when a statue and a pillar coincide, what we have is a statue, a pillar, and a statue-pillar, all of which are the same material object, and likewise for other cases of material constitution. One might then also concede that, in the incarnation too, there are three compounds—M++DN, M++HN, and M++(DN&HN)— all of which coincide, all of which count as the same material object (and perhaps the same F for a variety of other Fs) but only one of which counts as a person.²⁴

²⁴ Note that, on this view, DN and HN do not together compose a third, hybrid nature; rather, M+ +(DN&HN) is to be understood as a genuinely two-natured being. Note, too, that belief in multiple hylomorphic compounds—or even belief in many multi-natured beings—in the incarnate Christ poses

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

70

   :  

Overall, this second reply strikes me as more natural and more easily motivated for a believer in numerical sameness without identity, so I am inclined to favour it. But those who dislike the almost indiscriminate proliferation of multi-natured beings might prefer the first reply instead.²⁵

References Adams, Marilyn. 2006. Christ and Horrors. Oxford: Oxford University Press. Bird, Alexander. 2007. Nature’s Metaphysics: Laws and Properties. Oxford: Clarendon Press. Brower, Jeffrey. 2010. ‘Aristotelian Endurantism: A New Solution to the Problem of Temporary Intrinsics’. Mind 119: 883–905. Brower, Jeffrey. 2004a. ‘The Problem with Social Trinitarianism: A Reply to Wierenga’. Faith and Philosophy 21: 295–303. Brower, Jeffrey. 2004b. ‘Abelard on the Trinity’. In J. E. Brower and K. Guilfoy (eds), The Cambridge Companion to Abelard, 223–57. Cambridge: Cambridge University Press. Brower, Jeffrey and Michael Rea. 2005. ‘Material Constitution and the Trinity’. Faith and Philosophy 22: 487–505. Crisp, Oliver. 2007. Divinity and Humanity. Cambridge: Cambridge University Press. Cross, Richard. 2002. The Metaphysics of the Incarnation. Oxford: Oxford University Press. Cross, Richard. 2009. ‘The Incarnation’. In T. Flint and M. Rea (eds), The Oxford Handbook of Philosophical Theology, 452–75. Oxford: Oxford University Press. Fine, Kit. 1999. ‘Things and Their Parts’. Midwest Studies in Philosophy 23: 61–74.

no problem for orthodoxy. What matters for orthodoxy is that there is one person in the incarnate Christ, and that that person has exactly two natures. ²⁵ This chapter began life as a handout presented to my ‘Metaphysics and Christian Theology’ seminar in spring 2007. I am grateful to the students in that seminar—especially Andrew Bailey, Jennifer Martin, Luke Potter, and Luke van Horn—for helpful discussion. I have also been helped by comments from audiences at Wheaton College and Western Washington University, the ‘Incarnation: Perspectives from the Philosophy of Mind’ conference in Oxford, the Joseph Butler Society in Oriel College, Oxford, and the Baylor Philosophy of Religion Conference. On these occasions, Maria Rosa Antognazza, Trent Dougherty, Daniel Howard-Snyder, Hud Hudson, Jodeph Jedwab, Shieva Kleinschmidt, Brian Leftow, Anna Marmodoro, Alex Pruss, Christopher Shields, Christina van Dyke, and Ryan Wasserman were particularly helpful. Finally, I am especially grateful to Oliver Crisp for valuable conversation in the early stages of my work on this project, and to Jeff Brower and Alex Skiles who provided detailed and very helpful comments on earlier drafts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

   

71

Fine, Kit. 2008. ‘Coincidence and Form’. Proceedings of the Aristotelian Society, Supp. 82: 101–18. Flint, Thomas. 2009. ‘Should Concretists Part with Mereological Models of the Incarnation?’ Paper presented at the Logos Workshop in Philosophical Theology, University of Notre Dame. Hudson, Hud. 2005. The Metaphysics of Hyperspace. Oxford: Oxford University Press. Johnston, Mark. 2006. ‘Hylomorphism’. Journal of Philosophy 103: 652–98. Koslicki, Kathrin. 2008. The Structure of Objects. Oxford: Oxford University Press. Lewis, David. 1983. ‘New Work for a Theory of Universals’. Australasian Journal of Philosophy 61: 343–77. Loux, Michael. 1991. Primary Ousia. Ithaca, NY: Cornell University Press. MacBride, Fraser. 2005. ‘The Universal-Particular Distinction: A Dogma of Metaphysics?’ Mind 114: 565–614. McDaniel, Kris. 2009. ‘Structure-Making’. Australasian Journal of Philosophy 87: 251–74. McKitrick, Jennifer. 2003. ‘A Case for Extrinsic Dispositions’. Australasian Journal of Philosophy 81: 155–74. Merricks, Trenton. 2006. ‘Split Brains and the Godhead’. In T. Crisp et al. (eds), Knowledge and Reality: Essays in Honor of Alvin Plantinga, 299–326. Dordrecht: Springer. Merricks, Trenton. 2007. ‘The Word Made Flesh: Dualism, Physicalism, and the Incarnation’. In D. Zimmerman and P. van Inwagen (eds), Persons: Human and Divine, 281–300. Oxford: Oxford University Press. Molnar, George. 2003. Powers: A Study in Metaphysics, edited by Stephen Mumford. Oxford: Oxford University Press. Parsons, Josh. 2007. ‘Theories of Location’. Oxford Studies in Metaphysics 3: 201–32. Ramsey, F. P. 1925. ‘Universals’. Mind 34: 401–17. Reprinted in F. P. Ramsey, Philosophical Papers, edited by D. H. Mellor. Cambridge: Cambridge University Press, 1990, pp. 8–30. Rea, Michael. 1998a. ‘Temporal Parts Unmotivated’. The Philosophical Review 107: 225–60. Rea, Michael. 1998b. ‘Sameness without Identity: An Aristotelian Solution to the Problem of Material Constitution’. Ratio 11: 316–28. Rea, Michael. 2003. ‘Relative Identity and the Doctrine of the Trinity’. Philosophia Christi 5: 431–46. Rea, Michael. 2006. ‘Polytheism and Christian Belief ’. Journal of Theological Studies 57: 133–48. Rea, Michael. 2007. ‘The Trinity’. In T. Flint and M. Rea (eds), The Oxford Handbook of Philosophical Theology, 403–29. Oxford: Oxford University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

72

   :  

Waterlow, Sarah. 1982. Nature, Change, and Agency in Aristotle’s Physics. Oxford: Clarendon Press. Witt, Charlotte. 1989. Substance and Essence in Aristotle. Ithaca, NY: Cornell University Press. Witt, Charlotte. 2003. Ways of Being: Potentiality and Actuality in Aristotle’s Metaphysics. Ithaca, NY: Cornell University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

3 The Ill-Made Knight and the Stain on the Soul Eleonore Stump’s Atonement (2018) is a masterful and historic contribution to the project of Christian soteriology. Among its many virtues is the fact that it manages to be richly novel and innovative while at the same time hewing close and doing justice to what has been most widely and traditionally affirmed about the salvific work of Christ. One of the most interesting and important novelties in the book is her treatment of what she, following Aquinas, calls the problem of the stain on the soul. In this chapter, I will present that problem and Stump’s solution to it, explain why I think her solution falls short, and then suggest an alternative way of addressing it. I believe that the suggestion I will sketch is, in broad outline at least, compatible with Stump’s theory of the atonement; so I take myself to be recommending a supplement to what she says in her book, rather than rejecting any significant part of it.

1. A soteriology is a theory of the salvific significance of the work of Christ, which work comprises whatever aspects of his total earthly career that have salvific significance. More fully, it is a theory that identifies a salient problem (for human beings or God’s creatures in general) from which salvation is necessary, provides some specific content to the notion of salvation, and provides an explanation of how some salient work of Christ contributes to salvation from that problem for those affected by it. Christian philosophers and theologians have traditionally understood sin and its consequences to be the most salient problems afflicting creation. But the consequences of sin, according to traditional Christian doctrine, are legion. Some of the more important ones are the Fall of humanity, the breaking of creation, guilt before God, the cultivation of vice, enslavement to the flesh, suffering of various kinds and legacies of further sin and suffering in one’s own life and the lives of others, enmity towards God, separation from God, spiritual death, and eternity in hell. Salvation has, accordingly, been variously construed as Christ’s defeating or in some other way rectifying some combination of these consequences. Offering a theory of the atonement as one’s contribution to Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0004

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

74

   :  

soteriology represents a choice to focus on the consequences of sin for our relationship with God as the problem, and to understand the salvific significance of some aspect of Christ’s career—usually his suffering and death on the cross, but sometimes also some combination of his sinless life, resurrection, and ascension— mainly in terms of its contributions towards rectifying that relationship. For most contemporary philosophers working on the atonement, it seems as if the problem and its solution can be captured roughly as follows: Sin makes us guilty before God; guilt is what separates us from God (at least temporarily, if not permanently); therefore, we need forgiveness from God, and the work of Christ is a vital part (and maybe the whole of) what makes that possible, fitting, or both. Stump does not deny any of this; but, in contrast to most contemporary philosophers, she does offer a view according to which this is far from the whole story about the nature of the rift in our relationship with God and about what Christ does to help repair it. According to a fairly standard picture, we are separated from God mainly because (for some reason or other) God, being perfect, cannot or will not tolerate the presence of guilty creatures without some act of atonement (which act, of course, is done by Christ, thus removing or in some other way remedying our guilt and making it possible for us to re-enter the presence of God). In human relationships, however, remedying guilt by way of atoning acts and subsequent forgiveness may not be sufficient for full reconciliation and restoration of relationship. When human beings sin in serious ways against one another, there remains even after repentance and forgiveness what Stump characterizes as a kind of morally lamentable ‘residue’. She acknowledges that it is hard to say exactly what this residue consists in; but, as she characterizes it, it seems mainly to include the following components: memory of having committed the sins in question; experiential knowledge of what it is like to both desire and commit that kind of sin; and relational damage arising partly out of the damage wrought upon one’s psyche by one’s sin, but also out of the fact that this sin is now part of one’s own history and the history of one’s relationship with her victims, and will typically be remembered as such by all parties involved. So, for example, if you perpetrate serious abuse upon another person, however much you might repent and she might forgive you and absolve you of guilt, you or she or both will still remember you as the one who committed that abuse; you will remember what it is like to have both desired and then committed the abuse that you did; and all of this will morally diminish or taint both you and your present relationship, even if and after you have been forgiven by her. Moreover, even if you both manage to forget the abuse and all of its lingering effects upon your psyche, Stump points out that there will still be the history: you will still be one who did that thing; and this alone will morally diminish or taint both you and your

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

75

relationship with her, even if and after you have been forgiven by her. These effects of that sin together comprise a stain on your soul. The problem of the stain on the soul, then, is almost captured in Stump’s own summary of it as follows: Wrongdoing not only distorts the wrongdoer’s intellect and will, but it also has other morally lamentable effects; for example, it has deleterious effects on memory and on the cognitive capacities underlying mindreading and empathy. Furthermore, wrongdoing leaves relational characteristics altered for the worse. Something sad can remain for the wrongdoer in his relations with those hurt by him, even if he is repentant, even if he is forgiven by his victims. (2018: 340–1)

But I think that this statement does not fully capture the problem, because what it omits is the fact (apparent in Stump’s characterization of it, but never made fully explicit) that the stain on the soul—or some stains on the soul, anyway—linger also even after this-worldly therapies and other instruments and processes of psychological healing and restoration have run their course. If this were not so, then the problem of the stain on the soul would not be a problem from which we need divine salvation, and it would not properly figure very significantly into a theory of the atonement. The problem of the stain on the soul is, in my view, both genuine and serious; and it is one that has been almost entirely neglected in contemporary discussions of the atonement. But how can it possibly be solved if the stain lingers even after repentance and forgiveness, if some instances of it resist this-worldly processes of healing and restoration, and if it is partly constituted by history, the immutable past? What more could even an omnipotent God do about the effects of our sin after we have repented, been forgiven by God and others, and undergone the best, most effective processes of healing that this world has to offer? Are divine instruments of psychological healing so very different? Can they change the past? If not, then what can be done? The stain would seem to be indelible.

2. Stump’s answer to the question of how the stain can be removed, and how the work of Christ accomplishes its removal, comes in the final third of the penultimate chapter of her book. She begins by addressing what may well be the thorniest aspect of the problem: the fact that even after repentance and forgiveness, our sins against one another, and a fortiori our sins against an eternal and omniscient deity, live on in memory in a way that can still damage or otherwise diminish the quality of our relationship.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

76

   :  

Her solution, in short, is to point to the way in which the sting of a painful memory can sometimes be drawn. To illustrate, she imagines her familiar example-characters, Paula and Jerome, coping with some sin of Jerome’s against Paula in the following way. Jerome repents and Paula forgives Jerome; she then tells him that they can ‘forget this’ sin, and she expresses her willingness to be reconciled. In saying that they both can forget Jerome’s sin, she does not intend to convey that the sin will literally be wiped from their memories; rather, her point is simply that ‘this past event, now a subject of pain to both of them because of its evil, will remain in their memories but without its ability to cause either of them pain’ (373). But how can Paula’s declaration draw the sting of the painful memory? How can one make a memory no longer painful simply by fiat? Stump’s answer comes in the following paragraph: That a memory which was once painful to both Jerome and Paula can stay in their memories but lose its painful character stems from the fact that the harm which Jerome did Paula and now so regrets has become part of their ongoing joint story of mutual love and care. Their relationship is stronger because Jerome has come to Paula in repentance and Paula has accepted him as the repentant person he is. In his repentance and her acceptance of him, their relationship has not been restored to the same condition it had before he harmed her. His repentance cannot return either of them or their relationship to the relative innocence of the period before he hurt her. But his repentance and her willingness to be reconciled to him alter the relationship by making it more deeply rooted in each of them. Through his repentance and her reconciliation with him, the past hurt has been interwoven into a renewed commitment on the part of each of them to the other. This fact—that there is such a deepening of their relationship because of Jerome’s hurting Paula but repenting it and Paula’s accepting his repenting—does not mean that, retroactively, the very harm that Jerome did Paula is now not harm or that his harm should be welcome to her or that his harm is in some other way not the evil that it was. But this episode in their shared lives, in which Jerome did real and unwelcome harm to Paula, may nonetheless become precious to both of them because of what they have gone through together in it. (374)

In short, then: the sting is drawn by the combination of Jerome’s repentance, Paula’s expressed willingness to be reconciled, and the deepening of the relationship that comes in the wake of those things. It is perhaps easy to imagine things working as they do in Stump’s example when the sins in question are serious but not horrendous. Jerome spends the better part of a year drinking himself into oblivion, imposing heavy burdens upon Paula; finally there is a confrontation and Jerome is booted from the house and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

77

sent to rehab, after which Paula receives him back in something like the manner just described. Jerome badly and with much culpable negligence mismanages the finances for Paula’s business; heavy losses are sustained and Paula fires Jerome, whereupon he sues her, all to the near destruction of their friendship; but later he repents, and Paula receives him back in the manner described. And so on. But can things also work in this way when grievous, horrendous wrongs are in view? Stump says yes. Referring back to the sense, just described, in which Jerome’s sins against Paula, or the pain of a toddler’s frustratingly bad behaviour, can be forgotten, Stump says: This is the sense in which one can forget even great sins too if they are part of the history of human salvation. The acts are remembered, and so is their character as wrongful. But the remembered wrongful acts lose their power to produce pain in virtue of being wrongful, because they have become interwoven into a story of love that is worth prizing. And so there is a sense in which the sinfulness of those acts is forgotten. (374; emphasis in original)

Moreover, and crucially for Stump’s project, the ‘forgetting’ here described can come about with no actual loss of memory; and so it is a forgetting that even an omniscient God can experience. But even with such forgetting accomplished, more yet needs to be remedied to remove the stain on a person’s soul. For memory is not the only ingredient in the stain. There is also one’s history, and one’s experiential knowledge and its effects upon one’s personality and one’s ability to be close to others. At this juncture in her discussion, however, things start to move rather quickly. Shifting focus to Jerome’s reconciliation with God rather than Paula, and having effectively established that Christ’s atonement effects God’s forgetting of Jerome’s sins, Stump says that, if God forgets Jerome’s sins in the sense just described, those sins will also be forgotten in this sense by Jerome and a fortiori by all the redeemed in heaven for whom Jerome’s wrongdoing is also visible. . . . And so this last part of the stain on the soul from past sin, the shadows in memory and their connection to the empathic capacities, is healed also through the atonement of Christ . . . (375–6)¹

One might worry that the historical component of Jerome’s sin still lingers—he is, after all, still the person who did those things, whatever they were, to Paula. But I take it that on Stump’s view this part of the stain, too, is healed by being rendered irrelevant. Presumably the idea is that simply being the person who committed the sins one has committed is not by itself a stain; rather, it contributes to the staining ¹ The ‘connection to empathetic capacities’ that Stump mentions here are what I have been discussing under the description ‘experiential knowledge and its effects’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

78

   :  

of one’s soul only by way of its effects. Once those effects have been healed and only the history remains, however, its staining power has been thoroughly neutralized.

3. There is much that I like in Stump’s account of how the stains on our soul get removed. But it is not the way of contemporary philosophy to dwell at length on points of agreement, and so I turn now to objections. The upshot of my objections will not be that Stump has said the wrong thing about the healing of the stains on our soul. I think that everything I will say here is fully consistent with at least the broad contours of her account of the atonement. Rather, the upshot is that she has not said enough: there are stains on souls that her account thus far does not address. But, as I shall explain later, I think that there are ways of supplementing her account that will remedy this problem. To facilitate the discussion, I turn to an example from Arthurian legend. The tale of Sir Lancelot, as told in T. H. White’s (1940) The Ill-Made Knight, is the story of a man with a stain on his soul. Early in his career, before his ruinous and tragic affair with Guinevere, Lancelot is a model of purity, chivalric heroism, fidelity to his king and God—the greatest knight in all the land, a man destined for noble deeds and, indeed, one who fully expects, partly by way of remaining always a virgin, to be able to work miracles. The fall of Sir Lancelot comes while he is away from Camelot—partly in an effort to flee the temptation of his growing love for Guinevere. On his travels, he meets and rescues Elaine of Corbenic, who repays his good deed by getting him drunk and then deceiving him into having sex with her by posing as Guinevere. This rape costs him his virginity and so, in his mind, also his honour, all of his prospects for ultimate greatness, and his hope for one day being able to work miracles. Moreover, as he sees it, he has now not only betrayed Arthur, his king and best friend (never mind that the woman was not in fact Guinevere; for he thought that she was), and Guinevere, his one true love (never mind that he did not do it consensually). Overwhelmed by grief and guilt and shame, he returns to Camelot, whereupon he soon consummates his affair with the actual Guinevere, thus setting in motion the events that ultimately lead to the breaking of the round table and the tragic end of Arthur’s kingdom. The end of The Ill-Made Knight gives us Sir Lancelot utterly broken by shame and guilt over his many sins. He is in important ways a model case for the problem of the stain on the soul as Stump has set it up, for his sins—not just the loss of his virginity, with their attendant experiential knowledge and memories, but his actual betrayal of and lies to Arthur, as well as the consequences of his nonvoluntary ‘betrayal’ of Guinevere—have literally driven him to serious mental illness and, even after his return to some semblance mental health, have left him

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

79

deeply alienated from himself, his closest friend, his lover, and even his God. In the final pages we find Arthur badly in need of a miracle to save a fallen knight; and we see Lancelot, earlier a paragon of courage and honour, hiding away in a cell in abject fear of public failure, watching every other knight in the land try and fail to work the miracle. Eventually Lancelot is brought forth and persuaded to try. Fully expecting failure followed by public humiliation, Lancelot goes to the fallen knight and does work the miracle. God has blessed him; God has apparently forgiven him; God is apparently willing to be reconciled. Even so, we find Lancelot still plagued by the marks left on memory and other quarters of his psyche by his many sins. The case of Sir Lancelot is, as I have said, in some important ways a model illustration of a man whose soul is stained in the ways that Stump describes. But in some ways he is not. For it is not only his sins that have left the stain. The precipitating factor for the fall of Sir Lancelot—the event that moved him from being a faithful friend and servant of God actively resisting hard temptation to being a betrayer, an adulterer, and someone estranged from God and the people he loved— was an act to which he did not consent. Admittedly, as the story is told, Lancelot was not wholly without agency in the loss of his virginity; but he was mostly so, insofar as he was drunk, deceived, and partly under the influence of sorcery. We would not hesitate nowadays to say that Lancelot was victimized in this event; and yet it is among the most important events contributing to the stain on his soul. Here, then, is the first problem: As the idea of a stain on the soul is described— lingering, morally undesirable leftovers of great moral evil within the psyche, leftovers that come in the form of what is now unwelcome experiential knowledge and the memories thereof, together with a history whose consequences impair our relationships and even alienate us from ourselves—such stains are not caused by our sins alone. Look closely at the case of Lancelot. He remembers first intending, and acting on the intention, to betray Arthur with Guinevere because he was deceived by someone posing as Guinevere; the loss of his virginity that looms so large in his conception of himself and his vocation and that has left him at least partly estranged from God happened because he was deceived; his further sins— including his actual betrayal of Arthur—came not just in the wake of but partly as a result of this initial thing that happened to him; and so on. Things that happen to us can stain our souls no less than things that we do. The story of Sir Lancelot as described here is, of course, probably fictional. It is also dependent in significant ways upon ideas about men, women, God, human sexuality and its significance, and so on that many will find quaint at best, and in many ways positively problematic. But I hope it is evident that none of this matters for the main points I want to make. For it is similar enough in its details to plenty of true stories; and even quaint, false, and otherwise problematic beliefs can contribute in salient ways to the feelings, attitudes, and memories that enter into the stains on our souls. Trauma of all sorts, including traumas in which we have no agency whatsoever, can produce precisely the same kinds of effects that

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

80

   :  

are supposed to comprise the stain on the soul—including (importantly) unshakeable feelings of guilt and shame, together with the alienation from self and others that they cause. The problem, though, is that stains left by trauma rather than by sin do not at all seem to be the sorts of stains that will inevitably vanish and be forgotten by us simply in response to God’s letting us know that God has forgotten it, or is willing to forget it. Why think that a victim of horrific abuse whose soul has been stained thereby will suddenly ‘forget’ (in the relevant sense of having the sting drawn from her memories) the abusive events that have left her feeling guilty, ashamed, and alienated from self, other people, and God in response to God’s own willingness to forget about it? Moreover, and perhaps more importantly, why think that a divine ‘let’s forget about this’ response is an appropriate (or even morally acceptable) way of dealing with the stains left by victimization? A very natural reply, at this juncture, might be that these questions of mine have rhetorical force only because I am leaving out most of Stump’s story about how God gets, through Christ’s atoning work, to the stage of ‘forgetting’ human sins. The most important part of that story (so I would say, anyway) involves Christ’s empathetic engagement with every individual human being and their particular sins while dying on the cross. According to Stump, this is the way in which Christ ‘bears our sins’, and, on Stump’s account, this is a large part of what is behind the cry of dereliction from the cross. Christ’s psyche is somehow joined with the minds of everyone else so that he bears all of the psychological effects of our sins even without having actually committed any sin himself. He gets the experiential knowledge; he gets the feelings evoked by now unwanted memories; he knows the alienation; he knows what it is like to have the relevant history. And it is easy to imagine that knowing that we’ve been thus empathetically engaged would matter to those who have trauma-inflicted stains upon their souls. Easy to imagine, yes; but would it inevitably matter? That is harder to accept, and this for two reasons. First, there is no clear connection between Christ’s empathetic engagement with our experiential history and our no longer being pained (or guilty, or ashamed, or anything else) by it. Although it is certainly easy to imagine that one might feel a kind of comfort in knowing that Christ has felt all that we have felt, even in our worst moments as perpetrators or victims, it is just as easy to imagine that one might simply think ‘Well, I’m sorry for you too, then; but that doesn’t really help me at all.’ So that is one reason for doubting that this aspect of Stump’s story can do the work my imagined objector is suggesting it might do.² But there is a second reason, which is also the second problem that I want to raise for Stump’s account. Victims of serious evil do not always blame only their

² For this point I am indebted to Michelle Panchuk (although she raised it not specifically in connection with Stump’s theory of the atonement, but rather in connection with ideas I was developing in Rea 2018).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

81

human perpetrators. Some blame God simply for standing by and watching. Some victims of what Michelle Panchuk calls ‘religious trauma’ might blame God for putting into the hands of their perpetrators certain tools—for example, passages of scripture that seem to encourage submission to abuse—that contributed to their victimization.³ Some might even count God to be among the perpetrators. It makes no difference whether some or all of these ‘blaming God’ beliefs and feelings are false or perhaps even unwarranted (as might be the case, depending on their content, given the traditional view that God is perfectly good and loving and has perfectly good reasons for all that God does and allows). They are there nonetheless and contribute to people’s alienation from God. And here it seems that a ‘let’s forget about this’ response on the part of God is exactly the wrong approach to dealing with the stains left by their traumas, and pointing to Christ’s empathetic engagement with their trauma will (insofar as they partly blame Christ as one of the causes) be of no psychological help whatsoever. We might sum up the two objections I have raised against Stump’s solution to the problem of the stain on the soul as follows: Stump focuses on stains left by a person’s own sins; and her solution places us in the position of the person who, by rights, ought to be seeking and taking the first steps towards reconciliation. It is by virtue of the fact that we, rather than God, are in the latter position that the divine ‘we can forget about this’ comes as such sweet grace and comfort. But in fact the psychological components of what she is calling the stain on the soul can be caused by things other than one’s own sins; and, though I do not believe that God can sin or otherwise be in the wrong, and though I hesitate to speak of divine duties towards human beings, it nonetheless seems to be a mistake to locate us so firmly in the position of the one who, by rights, ought to be taking the first steps towards reconciliation. My smallest children sometimes get angry with and temporarily alienated from me for reasons that have much more to do with their own lack of understanding than any wrongdoing on my part. Maybe in some of those cases I have no duty to take the first steps towards putting things right between us. But, given their lack of understanding and its inevitability in light of their cognitive capacities, it does seem rather cold and unloving not to take those first steps, or to expect that my own willingness to forget about the whole thing would come anywhere close to doing the job.

4. What, then, can be done about these other stains on our souls? I said earlier that the project of soteriology is to develop a theory about the restorative significance (in relation to some salient affliction) for human beings ³ See Panchuk 2018.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

82

   :  

and the rest of creation of the salvific work of Christ, which work includes some or all of his life, suffering, death, resurrection, and ascension to heaven. Within this project, Christ’s death has been interpreted as a gambit, a vicarious punishment, a substitutionary punishment, a straightforward non-substitutionary punishment (of sin, or death, or human sinful nature), an expiatory sacrifice, a communicative sacrifice, a ransom payment, a divine apology, an act of identification with humanity, a great example of obedience and love, and much more; and this is to say nothing yet of other aspects of Christ’s earthly career. So, again, it is important to bear in mind that casting one’s soteriological theory specifically as a theory of (the) atonement is already to make a decision that might well be contested. As Stump herself points out, however, understanding Christ’s work as contributing to atonement does not preclude understanding it in other ways as well; and, to my mind, a promising way forward in light of the objections I have just raised is first to acknowledge that the atoning function of Christ’s work is not the whole story about how the stains on our souls are addressed in the context of the divine– human relationship, and then to supply some further story about what the work of Christ accomplishes for us beyond atonement. My own view (following Marilyn McCord Adams (1990, 2006)) is that at least part of this further story must be that Christ’s work somehow both defeats the badness of the evils—particularly the horrendous evils—in which we participate as victims and perpetrators, and redeems for us the parts of our lives that have been touched by those evils.⁴ Constraints of time and space permit only a sketch of what I have in mind. I leave the fuller details for another time. Theories of the atonement generally start from the idea that there is a rift caused by sin in God’s relationship with human beings, and they maintain that the atoning work of Christ is, fundamentally, God’s gracious way of repairing that rift as it appears from God’s side of the relationship. Christ’s atoning work is God’s way of addressing God’s grievances against us. But if what I have said in earlier sections of this chapter is correct, what we learn from reflection on the problem of the stain on the soul is that God’s grievances against us are not all there is to that rift. Instead—in some cases, at least—the rift is partly also caused and maintained by grievances that human beings have against God and also by, as it were, ‘grievances’ that human beings have against themselves. These latter grievances are often not considered in soteriological theorizing, largely (I suspect) because they are considered to be unjustified, irrational, or both. But I think that they deserve to be ⁴ Adams characterizes horrendous evils as ‘evils the participation in which (that is, the doing or suffering of which) constitutes prima facie reason to doubt whether the participant’s life could (given their inclusion in it) be a great good to him/her on the whole’ (Adams 1990: 26). On her view, defeating something bad is not the same as compensating someone for it, or seeing to it that the badness is outweighed. Instead, it is a matter of the bad thing’s being ‘included in some good enough whole to which it bears a relation of organic (rather than merely additive) unity’; and an instance of evil or suffering is defeated within the context of someone’s life if their life ‘is a good whole to which [that instance of evil or suffering] bears the relevant organic unity’ (Adams 1990: 28).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

83

taken seriously, I think that they are taken seriously by God, and I think that part of the function of Christ’s work is to address them. But they are not addressed by atonement, for the simple reason that they are neither grievances for which we need to atone nor, given divine sinlessness, are they grievances for which God needs to atone.⁵ As I have argued elsewhere, one way in which God deals with human grievances against God is via the scriptural authorization of lament and protest. But the work of Christ also has a role to play. Here is what I said there, drawing on ideas in Marilyn McCord Adams’s work,⁶ about how it might play such a role: In the early 1990s, in a manuscript that was never subsequently published, Jesse Hobbs argued that the atonement was, at least in part, a kind of divine apology for all of the evils in the world. [footnote omitted] This sort of view is untenable on the assumption that God is morally perfect; for, presumably, a morally perfect being would never do anything for which genuine apology is appropriate. But a morally perfect being might well sorrow over the pain inflicted on uncomprehending creatures by the pursuit of good ends that are ultimately beyond their ken; and such a being might take steps to validate the complaints that arise out of it, to take dramatic steps to identify not only with human victims of horrendous evil but also with the perpetrators, and to secure for people a blessed life at the end of all things—all with the aim of defeating, rather than merely compensating, the badness of the evils they have suffered. . . . Identifying through his own suffering with victims puts God in a kind of solidarity with victims; and identifying in his own suffering with perpetrators allows victims to see in the work of Christ both a divine acknowledgment that God has participated somehow as perpetrator of horrors and that the badness of these horrors, and of participation in them, merits some kind of condemnation. This kind of acknowledgment falls short of apology or penance, since it includes no actual admission of guilt; so it is an acknowledgment that can in principle be given by a morally perfect being. And it is the sort of acknowledgment, too, that can fit within a variety of different stories about what else, exactly, happens in and is accomplished by the work of Christ. (Rea 2018: 177–8)

I think that this is all true, as far as it goes; but what is omitted from what I said there is an explanation of the connection between the defeat of the evils in which we have participated and the removal of the stains on our souls. It is at just this point that I think talk of redemption needs to enter the picture. The idea of redemption can be woven into our discussion in the following way. Stains claim for themselves territory on the things that bear them. Removing a

⁵ Cf. Rea 2018, ch. 9.

⁶ See Adams 2006. See also Adams 2003, esp. note 4 on pp. 48–9.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

84

   :  

stain is a matter of reclaiming its territory—redeeming it from the presence that has taken it over. When we sin in serious ways, or when we are victimized, we might think of the moral leftovers that Stump takes to comprise the stain on the soul not via imagery of a blemish on an otherwise pristine surface but rather via imagery of a colonizing presence. The latter imagery is no less scriptural than the former; and I like it in the present context because I think it more readily facilitates the kind of supplemental story I would like to sketch about how the work of Christ might address the stains left by victimization, shame and self-blame, and the like. Over the past several decades, there has been growing appreciation in developmental and personality psychology of the vital role played by narrative in shaping our sense of self, our representation to ourselves of who we are.⁷ Facts about who we are (as contrasted, say, with facts about what we are, or about what person we are identical with) include, most importantly, facts about our core values, personality traits and dispositions, our salient social roles, and the trajectory of our lives and the significance (for us) of the events within it. Empirical research suggests, furthermore, that there is a bi-directional causal link between our narrative representations of who we are and who we in fact are, or become.⁸ In other words, our narrative representations of ourselves are not only shaped by the relevant real-world facts (most saliently by way of autobiographical memory) but they also contribute to shaping those facts over time. For example, there is evidence that recovery and growth in the wake of trauma are often facilitated by learning how to fit one’s traumatic experiences into a broader life-narrative that has a redemptive arc.⁹ Some, in fact, go so far as to say not only that narrative shapes our sense of self and impacts who we are at any given time, but that it constitutes us as persons and thus defines who we are.¹⁰ This is an interesting and suggestive idea; but I will not here recommend or develop it, in large part because I am not yet sure how best to understand it. Importantly, too, there is good reason to think that the narratives that shape— or, as some would have it, constitute—our sense of self are substantially coauthored by our families and peers, and that family narratives and the selfdefining memories from which they are built play a particularly important role in shaping who we take ourselves to be.¹¹ If this is true, and if everything else I have just said about self-shaping narratives is true, then the stories that others tell

⁷ Cf. McAdams 2001, 2008; McLean and McAdams 2013; Habermas and Köber 2015; and Singer 2004. ⁸ Cf. McLean, Pasupathi, and Pals 2007. ⁹ McAdams 2001, McLean 2015: 27–9. Note, however, that research also indicates that narrative meaning-making may not always be a healthy process in which to engage, and that in people with certain kinds of abuse-histories it is linked with greater incidence of PTSD (McLean 2015: 79. Cf. Greenhoot et al. 2013). To what extent these latter findings might force qualifications on what I say below about how Christocentric meaning-making might help to address the problem of the stain on the soul is not yet evident to me. ¹⁰ See, e.g. Schechtman 1996, 2014. ¹¹ Cf. McLean 2015.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

85

about us play an important role, too, in shaping not only who we take ourselves to be, but also who we in fact are. Reflecting on all of this in connection with familiar scripturally grounded claims to the effect that, in Christ, we have been reborn, adopted into a new family, given a new life and a new identity in Christ, and so on suggests a further way in which the work of Christ contributes to the defeat of the evils in which we have participated and the redemption of territory claimed within our souls by sin (both ours and other people’s) and its consequences. Serious sin and its consequences, as well as victimization, trauma, and their consequences, make a tremendous impact on our self-defining memories and threaten to consume the narratives that shape our sense of self. But, as the New Testament scriptures richly illustrate, the work of Christ provides tremendous resources for embedding our sins and traumas in new, redemptive narratives of our lives; and it constitutes a new source of ‘family stories’ and memories that are vividly called to mind by the Lord’s Supper and other liturgies of the Church, and are available for use in the coauthoring of our sense of self. To the extent that these new narratives of our lives can be overall beautiful and good, and to the extent that they can play a role in making us into (even if not fully constituting us as) who they say that we are, we would seem to have here a promising story about how the work of Christ and his solidarity with both victims and perpetrators might defeat, or contribute to defeating, the evils—even the horrors—in which we have participated. So likewise, to the extent that the work of Christ contributes (in the way just described) to making it the case that sin and its consequences for us (as either perpetrators or victims) no longer dominate the story of who we are, we would seem to have here a promising story about how the work of Christ redeems us from sin, reclaiming the territory within us—territory in our histories, our memories, our relationships and relational capacities, and much more—that has been colonized by sin and its consequences.¹² I acknowledge, of course, that I have come nowhere close to delivering on the promise I claim for the line of thinking I have just sketched. Among other things, I have yet to explain whether the mechanisms of defeat and redemption that I have just described depend on our consciously reflecting on and appropriating in this life the narrative resources that Christ’s work provides (I think it doesn’t), and what precise relationship the available Christocentric narrative of who someone is ¹² It is instructive to recall, at this juncture, the following part of Stump’s answer to the question of how the sting can be drawn from Paula and Jerome’s painful memory of his offence against her: ‘That a memory which was once painful to both Jerome and Paula can stay in their memories but lose its painful character stems from the fact that the harm which Jerome did Paula and now so regrets has become part of their ongoing joint story of mutual love and care’ (2018: 374). It is not clear just how much Stump wants to lean here on the relevance of the story of Paula and Jerome’s relationship, but I take the fact that she clearly sees the story as being of some relevance as a small bit of evidence that the supplement I am sketching here should be viewed as a welcome and friendly amendment to Stump’s overall project.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

86

   :  

might bear on the actual facts about who she is (I am, as of yet, uncertain). Filling in such details and delivering on the promise is a project that must be left for another time. But if the promise can be delivered on, it provides the kind of supplement to Stump’s theory that I think is needed to fully address the problem of the stain on the soul.¹³

References Adams, Marilyn McCord. 1990. ‘Horrendous Evils and the Goodness of God’. In The Problem of Evil, edited by Marilyn McCord Adams and Robert Merrihew Adams, 209–21. Oxford: Oxford University Press. Adams, Marilyn McCord. 2003. ‘In Praise of Blasphemy!’ Philosophia 30: 33–49. Adams, Marilyn McCord. 2006. Christ and Horrors: The Coherence of Christology. Current Issues in Theology. Cambridge: Cambridge University Press. Greenhoot, Andrea Follmer, Shengkai Sun, Sarah L. Bunnell, and Katherine Lindboe. 2013. ‘Making Sense of Traumatic Memories: Memory Qualities and Psychological Symptoms in Emerging Adults with and without Abuse Histories’. Memory 21: 125–42. Habermas, Tilmann and Christin Köber. 2015. ‘Autobiographical Reasoning Is Constitutive for Narrative Identity’. In Oxford Handbook of Identity Development, edited by Kate C. McLean and Moin Syed, 149–65. Oxford: Oxford University Press. McAdams, Dan P. 2001. ‘The Psychology of Life Stories’. Review of General Psychology 5: 100–22. McAdams, Dan P. 2008. ‘Personal Narrative and the Life Story’. In Handbook of Personality: Theory and Research, edited by Oliver John, Richard Robins, and Lawrence Pervin, 3rd edn, 242–62. New York: Guilford Press. McLean, Kate C. 2015. The Co-Authored Self: Family Stories and the Construction of Personal Identity. New York: Oxford University Press. McLean, Kate C. and Dan P. McAdams. 2013. ‘Narrative Identity’. Current Directions in Psychological Science 22: 233–8. McLean, Kate C., Monisha Pasupathi, and Jennifer Pals. 2007. ‘Selves Creating Stories Creating Selves: A Process Model of Self-Development’. Personality and Social Psychology Review 11: 262–78.

¹³ This chapter draws substantially on research done while funded by three different grants from the John Templeton Foundation: the Analytic Theology Project, the Experience Project, and a planning grant for a larger project on narrative conceptions of the self. I am grateful to the John Templeton Foundation for their generous support. Naturally, the views developed in this chapter are my own and are not necessarily endorsed by the Foundation. I am also grateful to Kate McLean, Sam Newlands, and Michelle Panchuk for helpful comments and conversation about the ideas herein.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 -       

87

Panchuk, Michelle. 2018. ‘The Shattered Spiritual Self and the Sacred: Philosophical Reflections on Religious Trauma, Worship, and Deconversion’. Res Philosophica 95: 505–30. Rea, Michael. 2018. The Hiddenness of God. Oxford: Oxford University Press. Schechtman, Marya. 1996. The Constitution of Selves. Ithaca, NY: Cornell University Press. Schechtman, Marya. 2014. Staying Alive: Personal Identity, Practical Concerns, and the Unity of a Life. Oxford: Oxford University Press. Singer, Jefferson. 2004. ‘Narrative Identity and Meaning Making Across the Adult Lifespan: An Introduction’. Journal of Personality 72: 437–60. Stump, Eleonore. 2018. Atonement. Oxford: Oxford University Press. White, T. H. 1940. The Ill-Made Knight. New York: G. P. Putnam’s Sons.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

PART II

E V I L , DI V I N E HI D D E N N E S S , AND WORSHIP

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

4 In Defence of Sceptical Theism A Reply to Almeida and Oppy [with Michael Bergmann]

Some evidential arguments from evil (e.g. those presented in Rowe 1979, 1988, 1991) rely on a ‘no-seeum’ inference of the following sort: NI If, after thinking hard, we can’t think of any God-justifying reason for permitting some horrific evil then it is likely that there is no such reason. (The reason NI is called a ‘no-seeum’ inference is that it says, more or less, that because we don’t see ʼum, they probably ain’t there.¹) Sceptical theists, us included, say that NI is not a good inference and that evidentialist arguments from evil that depend on it are, as a result, unsound. Michael Almeida and Graham Oppy (2003) argue that Michael Bergmann’s (2001) way of developing the sceptical theist response to such arguments fails because it commits those who endorse it to a sort of scepticism that undermines ordinary moral practice. Sceptical theism, they say, commits us to the claim that, for any terrible evil E we might consider preventing, we can assign no probability to the claim that great goods would be secured by our failure to prevent E. But, they write, if this is true then, contrary to what we all believe, ‘we cannot arrive at a reasoned view about whether or not to intervene to prevent E’ (2003: 516). In this chapter, we defend Bergmann’s sceptical theist response against this charge.

1. In developing his sceptical theist position, Bergmann relied on the following three claims (2001: 279): ST1 We have no good reason for thinking that the possible goods we know of are representative of the possible goods there are.

¹ The name for this sort of inference was introduced by Stephen Wykstra (1996).

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0005

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

92

   :  

ST2 We have no good reason for thinking that the possible evils we know of are representative of the possible evils there are. ST3 We have no good reason for thinking that the entailment relations we know of between possible goods and the permission of possible evils are representative of the entailment relations there are between possible goods and the permission of possible evils. Here, two clarificatory remarks are in order. First, note that a sample of xs can be representative of all xs relative to one property but not another. For example, a sample of humans can be representative of all humans relative to the property of having a lung while at the same time not being representative of all humans relative to the property of being a Russian. For ST1–ST3, what we are interested in is whether our sample of possible goods, possible evils, and entailment relations between them (i.e. the possible goods, evils, and relevant entailments we know of) are representative of all possible goods, possible evils, and entailment relations there are relative to the property of figuring in a (potentially) God-justifying reason for permitting the evils we see around us. Although that property is not explicitly mentioned in ST1–ST3, it is representativeness relative to that property that ST1–ST3 are speaking of. Second, it should be noted that, absent further explanation, it is hard to tell what sorts of things are referred to by the term ‘goods’. In some contexts, the term might refer to concrete objects or events that have some sort of positive value; in other contexts it might refer to abstract states of affairs or properties whose occurrence or exemplification would have or constitute some sort of positive value. For our purposes (in accord with what we take to be fairly standard practice in the literature on the problem of evil), we take ‘goods’ to be abstracta—properties or states of affairs. Actual goods, then, would be either states of affairs that actually obtain or properties that are actually exemplified; merely possible goods would be states of affairs that exist but do not obtain, or properties that exist but are not exemplified. Once this is clear, it is easy to clear up a certain confusion about how to read claims like ST1–ST3. Almeida and Oppy (2003: 505, n. 7) complain that ST1 admits of the following two readings: ST1a We have no good reason for thinking that the goods we know of are representative of goods that there are in the world. ST1b We have no good reason for thinking that the goods we know of are representative of the goods that there are in all possible worlds. On the plausible assumption that both properties and states of affairs are necessary beings, however, ST1a and ST1b are equivalent. It is only on the assumption that goods are contingent things—concrete objects or events, say—that ST1a and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    :      

93

ST1b come apart. Presumably it is this assumption that motivates Almeida and Oppy to say that ST1b is both ‘less controversial’ than ST1a and also ‘of no use to the sceptical theist, since—as Rowe insists—goods in other possible worlds cannot justify God’s actions in our world’ (2003: 505, n. 7). These remarks make sense on the assumption that goods are contingent beings (though, in that case, ST1a is hardly a plausible reading of ST1; it is better thought of as a replacement for ST1). But they make no sense on the assumption that we are making, that goods are necessary beings.² That said, we may now examine more closely Almeida and Oppy’s objection to Bergmann’s argument. Bergmann began his response to evidential arguments from evil that depend on NI by noting that theists and nontheists alike should agree with ST1–ST3.³ He then pointed out that the truth of ST1–ST3 should convince us that NI is a bad inference. Almeida and Oppy, on the other hand, claim that by endorsing ST1–ST3, Bergmann is committed to a sort of scepticism that undermines ordinary moral practice. The basic idea is as follows. Suppose you conclude on the basis of ST1–ST3 that there might, for all you know, be some great good that would result from God’s permitting some terrible evil E—a good that would provide God with reason to permit E. Then, Almeida and Oppy argue, you should likewise conclude that there might, for all you know, be some great good that would result from your permitting E—a good that would provide you with reason to permit E. If you conclude this, however, then it seems you are committed to the claim that, for all you know, it would be best, all things considered, if you did not prevent E. But then, they say, it is hard to see how you could reasonably conclude that you would not be justified in permitting E. But, of course, we can (in many such cases) reasonably conclude that we ought to intervene (and hence that we are not justified in failing to intervene). If, for example, you see a small child being abducted in a supermarket and you can easily and at no cost to yourself prevent the abduction from continuing, it would be morally abhorrent for you not to intervene. Indeed, it would be especially appalling for you to refrain from intervening on the grounds that the continuation of the abduction might, for all you know, result in some very great good. Thus, Almeida and Oppy conclude, insofar as endorsing ST1–ST3 commits us to the (false) conclusion that we cannot arrive at a reasoned decision about whether to prevent terrible evils, the sceptical theist strategy defended by Bergmann fails. The rest of this chapter will proceed as follows. In Section 2, we argue that the strongest conclusion that Almeida and Oppy’s argument could establish is not the general claim that sceptical theism as such undermines ordinary moral practice but

² Note, too, that our reading of ST1 in no way departs from the reading explicitly recommended by Bergmann in the paper to which Almeida and Oppy are replying (Bergmann 2001: 293, n. 4). ³ For a defence of the claim that theists and non-theists both should agree with ST1–ST3, see Bergmann 2001: 284–5 as well as Alston 1991: 44–5; 1996: 316–19.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

94

   :  

rather the significantly weaker claim that endorsement of ST1–ST3 in the absence of various background beliefs that theists are very likely to possess undermines ordinary moral practice.⁴ We then go on in Sections 3 and 4 to argue that Almeida and Oppy have not even successfully established this weaker conclusion.

2. Bergmann (2001) characterizes sceptical theism as traditional theism (the view that there exists an omniscient, omnipotent, omnibenevolent being) plus further ‘sceptical theses’ like, but not necessarily identical to, ST1–ST3. For present purposes, then, let us say that a sceptical theist is anyone who is both a theist and a proponent of ST1–ST3. And let us say that the ‘sceptical theist strategy’ is just the strategy of appealing to ST1–ST3 (or relevantly similar claims) in an effort to rebut the evidential argument from evil. According to Almeida and Oppy, the central idea behind the sceptical theist strategy is that ‘considerations of human cognitive limitations are alone [their emphasis] sufficient to undermine [evidential arguments from evil]’ (2003: 498). We don’t dispute this; but, obviously enough, the sceptical theist strategy will not be deployed in a vacuum. Sceptical theists, after all, are theists. Thus, when they consider the bearing of sceptical theism on their moral practice, they will inevitably and quite sensibly do so in a way that takes account of other things that they believe. But once this fact is appreciated, it is clear that most sceptical theists will find themselves completely untouched by Almeida and Oppy’s argument. The reason is simple: theists very typically believe that God has commanded his creatures to behave in certain ways; and they also very typically believe that God’s commands provide all-things-considered reasons to act. Thus, a sceptical theist will very likely not find it the least bit plausible to think that ST1–ST3 leave us without an all-things-considered reason to prevent harm to others in cases like the abduction scenario described in Section 1. For even if ST1–ST3 imply that we do not know much about the realm of value, they do not at all imply that we know nothing about that realm; and, in particular, they do not imply that we lack knowledge of God’s commands as God’s commands. Almeida and Oppy explicitly grant that one might reply to their argument by appealing to the fact that God has somehow revealed to us the fact that great goods cannot be secured by our failing to prevent terrible evils like the supermarket abduction. But, they say:

⁴ Almeida and Oppy indicate that what we call the ‘weaker’ claim is the claim that they aim to establish. But they take that ‘weaker’ claim to imply the claim that sceptical theism as such undermines ordinary moral practice (2003: 498, 508).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    :      

95

in making this sort of reply, one would be giving up the sceptical theist ambition: it is no longer true that it is merely considerations about our cognitive limitations that yield the desired conclusion. The whole point of the sceptical theist response is that it is supposed to avoid appeal to the other evidence which theists possess for the existence of a perfect being and the directives that that being makes in connection with our behaviour. (2003: 508)

But this reply just confuses the sceptical theist’s argument with her assessment of how the premises of that argument bear on other beliefs that she holds. To appeal to divine revelation in arguing that ST1–ST3 undermine the crucial no-seeum inference in some evidential argument from evil would be to give up the sceptical theist’s ambition. But to appeal to such revelation in showing that (contra Almeida and Oppy) ST1–ST3 do not undermine the sceptical theist’s own moral reasoning is not to give up that ambition. Almeida and Oppy are simply wrong, then, to claim that their argument ‘carries over to all versions of “sceptical theist” responses to evidential arguments from evil’ (2003: 498). That is, they have not established the strong claim they have set out to prove, namely: SC All sceptical theists are in the predicament of having their ordinary moral practice undermined. But it is important to note that, for all we have said so far, their argument might at least apply to anyone who endorses ST1–ST3 in the absence of standard theistic background assumptions. And surely there can be such people. Indeed, sceptical theists are typically recommending that their atheistic opponents accept claims like ST1–ST3 without, in the same breath, recommending that they abandon their atheism. For all we have said so far, then, Almeida and Oppy’s argument might at least show the truth of the weak claim: WC Nontheists who endorse the sceptical theses like ST1–ST3 are in the predicament of having their ordinary moral practice undermined. In the next two sections, however, we argue that Almeida and Oppy have not even successfully established this significantly weaker conclusion.

3. Consider again a case where we see someone, with obviously malevolent intentions, grab a child in a supermarket. Suppose we can easily and at no cost to ourselves prevent the abduction from continuing. Is there anything the non-

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

96

   :  

theistic proponent of ST1–ST3 can sensibly say on behalf of the claim that we have all-things-considered reason to prevent the abduction? Let us begin by stipulating, as Almeida and Oppy are prepared to do, that we have pro tanto reason to prevent the abduction. Since Almeida and Oppy do not want their argument to rest on the assumption that consequentialism is correct (2003: 510, n. 21), we are free to assume that our pro tanto reason in this case might include, say, awareness of a serious prima facie duty to prevent great harm to others together with awareness that the abductor’s success will likely result in great harm to others (whether or not that harm might ultimately be outweighed by countervailing goods). Now, according to Almeida and Oppy, apart from sceptical theistic assumptions, our natural line of reasoning in this sort of case would be as follows (cf. 2003: 507): (1) There is pro tanto reason for us to intervene to prevent the abduction. (2) We have found no pro tanto reason for us to permit the abduction. (3) Therefore: There is no pro tanto reason for us to permit the abduction. (From 2) (4) Therefore: We have all-things-considered reason to prevent the abduction. (From 1, 3) Importantly, the inference from (2) to (3) is a no-seeum inference of the sort that normally figures in evidential arguments from evil. But, they argue, just as the noseeum inferences that figure in evidential arguments are undermined by ST1–ST3, so too the inference from (2) to (3) is undermined by ST1–ST3. The reason is as follows. According to Almeida and Oppy, if we are convinced of ST1–ST3, then we cannot sensibly assign any likelihood to the truth of claims like P1 and P2: P1 Permitting the abduction will produce some outweighing good or prevent some worse evil. P2 Permitting the abduction (or something as bad) is required to produce some outweighing good or to prevent some worse evil. Plausibly, we also would be unable sensibly to assign any likelihood to the truth of P3: P3 If only we were smarter or better informed, we would be aware of goods the awareness of which would override whatever prima facie duties we might have to prevent the abduction. But, if any of P1–P3 is true, then (3) might well be false. Thus, on their view, if we endorse ST1–ST3, then we must remain agnostic about the likelihood of claims

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    :      

97

like P1–P3; but if we do that, then we cannot reasonably infer that there is no pro tanto reason for us to permit the abduction from the premise that we have found no pro tanto reason to permit the abduction. Are they right about this? Suppose that they are. That is, suppose that premise (2) does not in fact support premise (3); and, for good measure, suppose that we are not independently justified in believing (3). What follows for ordinary moral practice? Not much of interest, it seems. For even if the argument from (1)–(4) is bad, surely the following line of reasoning is sensible: (1a)

There is strong pro tanto reason to prevent the abduction.

(2a)

We have found no pro tanto reason to permit the abduction.

(3a) There is no reason to think (and, indeed, good reason to doubt) that any investigation that we could possibly conduct before having to make a decision about whether to prevent the abduction would turn up evidence pointing to even a weak pro tanto reason to permit the abduction. (4a)

Therefore: We ought to prevent the abduction.

Note that we leave open the question whether (1a)–(3a) support (4)—the claim that we have all-things-considered reason to prevent the abduction. Almeida and Oppy say virtually nothing about their conception of what it is for there to be a reason to do x, of what it is to have a reason to do x (can there be reasons one does not have?), or of what it is to have all-things-considered reason to do x; nor do they say much of anything about their conception of how these notions are connected with properties like being such that one ought to do x or being such that one has sufficient reason to do x. Absent such clarification, it is a bit hard to say whether (1a)–(3a) support (4). But it is hard to deny that (1a)–(3a) support (4a); and (4a) is all we need in order to save ordinary moral practice and to conclude that we are not justified in permitting the abduction. The point, in short, is that even if it is true that a proponent of ST1–ST3 is committed to thinking that, for all she knows, there is some all-things-considered reason (presently unknown or at least unappreciated by her) to permit some terrible evil E, it doesn’t at all follow that she is committed to doubting that she ought to prevent E. The following claim is (for all Almeida and Oppy have shown, anyway) perfectly consistent: (*) For all I know, there is some all-things-considered reason (presently unknown or at least unappreciated by me) for me to permit E; but still, I know that I ought to prevent E. If that sounds odd, just consider the fact that—as Almeida and Oppy themselves admit (2003: 512)—we often do not know the long-range consequences of our

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

98

   :  

actions. Perhaps the child is a nascent monster, and preventing his abduction and subsequent murder would result in the suffering of millions upon millions of people and the ultimate destruction of the world. For all you know, the world might be better, all things considered, if he were murdered. Still, apart from only the most naïve and implausible versions of consequentialism, it might well be that you ought to prevent his abduction. At this juncture, it is instructive to consider the above remarks in conjunction with an example that Almeida and Oppy offer in support of their contention that sceptical theism undermines ordinary moral practice. They write: When faced with a decision about whether or not to perform an action, rational people act on the basis of the relevant considerations that are available to them (whether or not those considerations count as knowledge or even, perhaps, rational belief). Second, and more importantly, we have to be able to factor self-confessed ignorance into our process of deliberation. It need not be ‘foolish to avoid action because of something we don’t know not to obtain’. Suppose that we don’t know that it is not the case that the floorboards are rotten (or . . . that we are not prepared to make any estimation of whether or not it is the case that the floorboards are rotten). How can we then step with complete confidence onto the floor? In practical considerations—no less than in theoretical considerations— probabilities are the stuff of deliberation. And whereof one is not prepared to assign probabilities, thereof one is simply not able to deliberate. (2003: 515)

The idea here, we take it, is that if we are aware of our inability to assign likelihood to the truth of the proposition that the floorboards are rotten, then we are unable to move from claim (5) to claim (6): (5) We have found no pro tanto reason to refrain from stepping on the floor. (6) Therefore: There is no pro tanto reason to refrain from stepping on the floor. But suppose they are right. Still, the following line of reasoning (if the premises were true) would be perfectly sensible: (7) We have strong pro tanto reason to step on the floor. (8) We have found no pro tanto reason to refrain from stepping on the floor. (9) We have no reason to believe (and, indeed, good reason to doubt) that any further investigation we could make before we would have to make a decision about whether to step on the floor would turn up evidence pointing to even a weak pro tanto reason to refrain from stepping on the floor. (10) Therefore, we ought to step on the floor.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    :      

99

One might think that our agnosticism about the probability that the floorboards are rotten constitutes reason to doubt (9). Perhaps whatever considerations counsel agnosticism about the relevant probability also constitute reason to think that further investigation would turn up at least a weak pro tanto reason to refrain from stepping on the floor. More plausibly, perhaps those same considerations themselves constitute at least weak pro tanto reason to refrain from stepping on the floor. But these points are neither here nor there as far as our argument is concerned. If anything, they only point to the fact that the floorboard analogy was bad to begin with. For, after all, by Almeida and Oppy’s own lights (2003: 515), neither the awareness of our agnosticism about claims like P1–P3 nor the considerations that lead us to such agnosticism are, by themselves, supposed to constitute either pro tanto reason for permitting the abduction or reason for thinking that such a reason would be turned up by further investigation. In sum, then, even if Almeida and Oppy are right in thinking that belief in ST1– ST3 undermines a no-seeum inference from the premise that we have found no pro tanto reason to prevent some terrible evil to the conclusion that there is no pro tanto reason to prevent that evil, it does not follow that proponents of ST1–ST3 cannot sensibly reach the conclusion that they ought to prevent that evil. Furthermore, it should be clear that the moves made to save ordinary moral practice do not have parallels that could be employed by someone aiming to save the evidential argument from evil. The basic move we have made is to concede (at least for the sake of argument) that the no-seeum inference from (2) to (3) is bad, and to argue that one does not need to believe that there is no pro tanto reason to refrain from doing x in order to believe (reasonably) that one ought to do x. The idea is that in cases like the abduction scenario, even if there are reasons that we are not aware of to permit the abduction, those reasons do not bear on what we ought to do. To concede that the crucial no-seeum inference in the evidential argument is bad, however, is to give up the game. For, after all, to say that God might, for all we know, have reason to permit some instance of terrible evil is just to say that there might, for all we know, be reasons of which we are not aware that do bear on what God ought to do.

4. One way of getting at the source of our disagreement with Almeida and Oppy is to notice an inconsistent triad of claims that emerges in light of the above discussion of the Almeida and Oppy’s argument. Consider a person S who is in a situation in which she can easily prevent—in a way that (as far as she can see) will cause no significant loss—some terrible evil E. Now consider these three propositions in connection with such a case:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

100

   :  

(i) S reasonably believes that S ought to prevent E. (ii) S is unable sensibly to assign any probability to any of the following three propositions: Q1: S’s permitting that terrible evil E will produce some outweighing good or prevent some worse evil. Q2: S’s permitting that terrible evil E (or something as bad) is required to produce some outweighing good or prevent some worse evil. Q3: If only S were smarter or better informed, S would be aware of goods the awareness of which would override whatever prima facie duties she might have to prevent E. (iii) S reasonably believes that S ought to prevent E only if S can sensibly assign a low probability to Q1, Q2, or Q3. Almeida and Oppy clearly think that (i) will be true of the abduction case. We agree. But we disagree with Almeida and Oppy about what to say about (ii) and (iii). We find ST1–ST3 extremely plausible. And we think that proponents of ST1– ST3 are committed to (ii). So we accept (ii) and reject (iii). Almeida and Oppy, on the other hand, seem to be committed to (iii) and to the denial of (ii). At any rate, if they were to abandon (iii), it is hard to see what reason they could give for thinking that the truth of (ii) undermines ordinary moral practice. By our lights, however, Almeida and Oppy have not adequately defended (iii); nor, in our view, is (iii) independently plausible. The conjunction of ST1–ST3 represents a moderate and sane scepticism concerning our abilities to discover facts about goods and evils. As Bergmann says in the paper to which Almeida and Oppy are objecting, ‘It just doesn’t seem unlikely that our understanding of the realm of value falls miserably short of capturing all that is true about that realm’ (Bergmann 2001: 279). (iii), on the other hand, seems clearly to be false, as evidenced by the cogency of the argument from (1a)–(4a) and the consistency of (*). At one point in their paper, Almeida and Oppy make remarks that suggest that they recognize the attractiveness of our approach, which favours (ii) over (iii): In sum, then, the sceptical theist response to evidential arguments from evil, if successful, really would pose a serious threat to ordinary moral practice. In any decision situation, we would be in the position of the person who is ‘out of her depth’ and who knows that she is ‘out of her depth’. What is objectionable about this is not the thought that we might always be ‘out of our depth’, in the sense that we are unable to fully evaluate the considerations which bear on our decision; for, of course, none of us can know all of the long term consequences of any action we perform—no doubt there were all kinds of good deeds which were causally necessary for Hitler to be born—and to this extent we are always ‘out of our depth’ in deciding what to do. Rather, the problem is that, if we are

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    :      

101

always ‘out of our depth’ and if we always aware that we are ‘out of our depth’, then we can never give first personal endorsement to any of our actions; moral deliberation can never end in anything more than the equivalent of tossing a coin. (Almeida and Oppy 2003: 512)

On the one hand, they say that it’s obvious that we are always out of our depth in our moral reasoning. On the other they say that our moral practice is undermined if we always realize we are always out of our moral depth. But, given that sensible people realize what is obvious, this is just to say that they think the moral practice of sensible people is always undermined. If they think that, why charge sceptical theists with having any special problem in this regard? More to the point, though: isn’t this just a reductio of the claim that the realization that we are out of our depth (in the sense described above) inevitably undermines ordinary moral practice? We do not need to believe that we have fully fathomed the realm of value in order to (reasonably) believe that we ought to save a child from being abducted from a supermarket. Nor do we have to have opinions about the likelihood that the abduction will produce some great outweighing good. All we need is our awareness of a strong prima facie duty to prevent harm for others, together with the rational conviction that the abduction will cause harm and that no evidence that could be turned up by further investigation (in the time available) would support the claim that we ought to permit that harm. And this we can have regardless of our ability to assign probabilities to future consequences, and regardless of whether we accept ST1–ST3.⁵

References Almeida, Michael and Graham Oppy. 2003. ‘Sceptical Theism and Evidential Arguments from Evil’. Australasian Journal of Philosophy 81: 496–516. Alston, William. 1991. ‘The Inductive Argument from Evil and the Human Cognitive Condition’. Philosophical Perspectives 5: 29–67. Alston, William. 1996. ‘Some (Temporarily) Final Thoughts on Evidential Arguments from Evil’. In The Evidential Argument from Evil, edited by Daniel Howard-Snyder, 311–32. Indianapolis: Indiana University Press. Bergmann, Michael. 2001. ‘Skeptical Theism and Rowe’s New Evidential Argument from Evil’. Noûs 35: 278–96. Rowe, William. 1979. ‘The Problem of Evil and Some Varieties of Atheism’. American Philosophical Quarterly 16: 335–41. Rowe, William. 1988. ‘Evil and Theodicy’. Philosophical Topics 16: 119–32. ⁵ Thanks to Jeffrey Brower, Daniel Howard-Snyder, Patrick Kain, Klaas Kraay, William Rowe, and two anonymous referees for Australasian Journal of Philosophy for comments on earlier drafts.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

102

   :  

Rowe, William. 1991. ‘Ruminations about Evil’. Philosophical Perspectives 5: 69–88. Wykstra, Stephen. 1996. ‘Rowe’s Noseeum Arguments from Evil’. In The Evidential Argument from Evil, edited by Daniel Howard-Snyder, 126–50. Indianapolis: Indiana University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

5 Wright on Theodicy Reflections on Evil and the Justice of God

For the past four decades, the problem of evil has been one of the most important and widely discussed topics in analytic philosophy of religion. For those participating in this discussion, what the problem of evil is, fundamentally, is just the problem of figuring out whether and to what extent the existence of evil counts as evidence against the existence of God. Nobody doubts that evil is problematic in other respects too. And, of course, it seems almost silly to mention that nobody doubts that one of the main problems with evil is just that . . . Well . . . evil is bad, and we’d like for it to be dealt with and to go away. But this hasn’t stopped any of us from thinking that the philosophical problem of evil is worth our time, and well worth the ink that has been spilled to talk about it. In Evil and the Justice of God, however, N. T. Wright (2006) presses the point that attempting to solve the philosophical problem of evil is an immature response to the existence of evil—a response that belittles the real problem of evil, which is just the fact that evil is bad and needs to be dealt with. Though Wright does not spend a great deal of time defending this point, it is reiterated several times in the text and seems to play a significant role not only in motivating Wright’s own conception of what the real problem of evil is, but also in motivating his approach to solving that problem. If the point is correct, then the vast majority of work on the problem of evil in the analytic philosophical tradition has been worthless at best, and possibly even pernicious (by virtue of trivializing a serious theological issue). As you might expect, I’m not inclined to endorse this sort of sweeping indictment of the entire field of research on the philosophical problem of evil. (I sort of doubt that Wright really meant to either.) But I do think that there is a kernel of truth in what I take to be Wright’s fundamental objection to attempts to solve the philosophical problem of evil. In the first section of what follows, I will try briefly to explain why. I will then go on to argue that, despite this fact, certain efforts at solving the problem of evil avoid Wright’s objection. Indeed, drawing on recent work by Eleonore Stump, I will argue that one perfectly legitimate way to try to solve the philosophical problem of evil is to follow precisely what seems to be the main piece of advice in Evil and the Justice of God: namely, to look more seriously than we have at the attitudes taken towards evil by the human authors of and characters in the Bible, and to attend more carefully to what the Bible says about how God deals with evil. Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0006

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

104

   :  

1. Wright against Theodicy In the literature on the philosophical problem of evil, there have been, broadly speaking, two kinds of solution offered: (a) solutions that attempt to show how God might be justified in permitting evil of the sort and amount we find in our world, and (b) solutions that attempt to explain why, though we have no idea how God might be justified in permitting evil, the existence of the sort and amount of evil that we find in the world still does not count as reason to disbelieve in God. I’ll use the term ‘theodicy’ to refer to any solution of the first type. (I thus ignore, for ease of exposition, a familiar distinction between theodicy and defence.) Moreover, when I talk of ‘the problem of evil’, I will have in mind just what I have been calling the philosophical problem of evil—again, the problem of figuring out whether and to what extent the existence of evil of the sort and amount we find in our world is compatible with the existence of the all-powerful, all-knowing, perfectly good God of traditional theism. Much of the literature on the problem of evil—both contemporary and historical—has focused on theodicy. According to Wright, however, this focus is misguided. Indeed, the project of theodicy, on his view, is immature and even pernicious. Thus, in the course of discussing ‘immature reactions’ to evil, he writes: Lashing out at those you perceive to be ‘evil’ in the hope of dealing with the problem [of evil]—say, dropping copious bombs on Iraq or Afghanistan because of September 11th 2001—is, in fact, the practical counterpart of those philosophical theories that purport to ‘solve’ the problem of evil. Various writers have suggested, for instance, that God allows evil because it creates the special conditions in which virtue can flourish. The thought that God decided to permit Auschwitz because some heroes would emerge is hardly a solution to the problem. In the same way, the thousands of innocent civilians who died in Iraq and Afghanistan bear mute testimony to the fact that often such ‘solutions’ simply make the problem worse . . . (2006: 28)¹

Later, he seems to characterize the effort at theodicy as something we ought to outgrow. Thus: The big question of our time . . . can be understood in terms of how we address and live with the fact of evil in our world. Growing out of the traditional philosophers’ and theologians’ puzzlement, the problem of evil as we face it today on our streets and in our world won’t wait for clever metaphysicians to solve it. (39)

¹ All Wright quotations in this chapter are from this edition of his book.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

105

A moment’s reflection reveals that these are serious charges. Attempting theodicy is somehow akin to carpet bombing our enemies; and it is an intellectual project that we must ‘grow out of ’—an opportunity, perhaps, to manifest a kind of cleverness; but not a serious project to engage in for people who are genuinely concerned about evil. But the question that receives no very explicit answer in Wright’s book is just this: Why should we think of the project of theodicy this way? In what way, exactly, is it the intellectual counterpart of carpet bombing? So far as I can tell, lurking behind the charges levelled in the passages just quoted are two different kinds of objection to the project of theodicy—one misguided, in my view; the other on target. The first objection is that theodicy minimizes either the badness of evil or the goodness of God’s creation. Thus, he writes: We cannot and must not soften the blow; we cannot and must not pretend that evil isn’t that bad after all. That is the way back to cheap modernism. As I said earlier, that is the intellectual counterpart to the immature political reaction of thinking that a few well-placed bombs can eliminate ‘evil’ from the world. No: for the Christian, the problem is how to understand and celebrate the goodness and God-givenness of creation and how, at the same time, to understand and face up to the reality and seriousness of evil. It is easy to ‘solve’ the problem by watering down one side or the other, saying either that the world isn’t really God’s good creation or that evil isn’t really that bad after all. (40–1)

The second objection is that there is some hubris in the attempt at theodicy. The relation between the two objections is just that both failings—the hubris involved in theodicy and the minimization of the badness of evil—are, Wright seems to think, a product of our failure fully to appreciate the depth and degree to which God’s beautiful creation has been dipped in and tainted by evil. It is easy enough to see why such a failure to appreciate the depths of the problem might lead one to minimize the badness of evil; but it will take a bit of explanation to show how it might also lead to hubris. Wright makes it clear that one of the reasons why carpet bombing the enemy is an immature reaction to evil is that it neglects the way in which we ourselves are tainted by evil. As he puts it in one context: The gospels tell the story of Jesus as a story in which the line between good and evil runs not between Jesus and his friends on the one hand and everyone else on the other . . . but down the middle of Jesus’ followers themselves. (82; emphasis in original)

This feature of the gospel is precisely what is missed by those who think that we can simply eliminate evil from the world by first identifying the evil people (them,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

106

   :  

not us) and then by bombing evil out of existence. Evil permeates the world, and it permeates us. Only by failing to see this can we even begin to think that the problem of the existence of evil could be solved in so simple a way as dividing the world into good guys and bad guys and then blasting the bad guys off the face of the earth. But just as it is a sign of immaturity unreflectively to group oneself with the ‘good guys’ who ought to be ridding the world of ‘bad guys’, so too, Wright seems to say, it is a sign of immaturity to think of oneself as possessing clear enough moral vision to determine what sorts of reasons might or might not justify God in permitting certain kinds of evil and suffering. Here, again, I quote him at some length: . . . ever since the garden, ever since God’s grief over Noah, ever since Babel and Abraham, the story has been about the messy way in which God has had to work to bring the world out of the mess. Somehow, in a way we are inclined to find offensive, God has to get his boots muddy and, it seems, to get his hands bloody, to put the world back to rights. If we declare, as many have done, that we would rather it were not so, we face a counter-question: Which bit of dry, clean ground are we standing on that we should pronounce on the matter with such certainty? Dietrich Bonhoeffer declared that the primal sin of humanity consisted in putting the knowledge of good and evil before the knowledge of God. That is one of the further dark mysteries of Genesis 3: there must be some substantial continuity between what we mean by good and evil and what God means, otherwise we are in moral darkness indeed. But it serves as a warning to us not to pontificate with too much certainty about what God should and shouldn’t have done. (58–9)

In short, just as a properly mature appreciation of the way in which we ourselves are tainted by evil ought to make us resist carving the population of the world up into white-hats and black-hats, so too it ought to make us doubt the need for theodicy (as if people as corrupt as we are have any business sitting in judgment over a pure and holy God), and it also ought to make us doubt our ability to find one. What shall we think of these two objections? I am inclined to doubt that there is anything essentially trivializing about the project of theodicy. To be sure, there is something deeply appalling about suggesting, for example, that God might permit the rape and murder of a small child so as to build character in her parents. More controversially, one might also think that there is something appalling about, say, Leibnizian theodicies, according to which our world is the best of all possible worlds, and the many and varied instances of evil fit into our world in roughly the way in which various instances of an ‘ugly colour’ might fit into an overall beautiful painting. But, at best, all this shows is that some attempts at theodicy

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

107

are bad ones. Why go on to conclude from this that just any effort at showing why God might permit certain kinds of evil and suffering must either minimize the badness of evil or minimize the goodness of the world? At the very least, a lot more argument would be needed to make such a claim plausible. The second objection, on the other hand, seems right on target. Indeed, in recent years, many philosophers (myself included) have turned away from attempts at theodicy, advocating instead a position known as ‘sceptical theism’.² And their reasons for doing so have been roughly in line with what I have taken to be Wright’s second objection to theodical projects. At the heart of sceptical theism is the idea that our cognitive faculties are so limited that there is no reason to believe that we would be able to see what goods actually or even possibly justify God in permitting the various kinds of suffering we find in the world. And it is not implausible to think that one of the reasons why our faculties are so limited—one reason why our moral vision is so darkened—is just the fact that, as Wright highlights, we ourselves are tainted by evil. But once the central thesis of sceptical theism is granted, the problem of evil as I have characterized it is solved, though not by theodicy. For the existence of some sort of evil counts as evidence against the existence of God only if we have reason to think that a good God, if he existed, would not be justified in permitting that sort of evil. But, says the sceptical theist, since we have no reason to think that we could tell what goods might justify God in permitting the sorts of evil we find in the world, we also have no reason to think, for any evil we might find, that God could not be justified in permitting it. Thus, the existence of evil does not count as evidence against the existence of God. Not surprisingly, many find this ‘solution’ to the problem of evil unsatisfying. But (philosophical objections to the solution aside) the dissatisfaction arises primarily from the fact that what we would like are definite answers to questions about why God permits this or that sort of evil; and these are precisely the sorts of questions that, if the sceptical theist is right, we are unequipped to answer. In sum, then, I reject the suggestion that theodicy essentially minimizes either the goodness of creation or the badness of evil; but I think that the kernel of truth in Wright’s rejection of theodicy lies (if I read him right) in his view that certain attempts at theodicy reflect an overestimation of our cognitive abilities and an underestimation of the depths to which they have been infected by evil. But now does it follow from this that all attempts at theodicy are hopeless? Not obviously. In fact, a case can be made for the conclusion that one way of providing a theodicy is to follow Wright’s own advice and take a close look at the attitudes towards evil that we find in the Bible, and, most importantly, at what God has done throughout history and in the person and work of Jesus Christ to deal with

² For a useful overview and references, see Bergmann 2009.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

108

   :  

evil. In the next section, I draw on the work of Eleonore Stump to argue briefly for this conclusion.

2. Theodicy and Biblical Narrative Suppose the sceptical theist is right: our cognitive faculties are so limited that we have no good reason to think that we could identify goods that might justify God in permitting many of the sorts of evils that we find in our world. Still, it does not follow from this that we cannot come to see clearly the truth of the following proposition: (T) There are goods that justify God in permitting all of the various kinds of evil that we find in the world. Just as we do not need to be able to identify particular electrons in order to be justified in believing that there are some, so too we do not need to be able to identify particular God-justifying goods in order to be justified in believing that there are some. How might we come to be justified in believing (T) even in the absence of identifying particular God-justifying goods? Here is an easy way: We come to believe in God on the basis of argument or religious experience; we then infer from the existence of God, together with the existence of the various kinds of evil that we find in the world, that (T) is true. But, of course, demonstrating the truth of (T) in this way does not count as offering a theodicy. For we haven’t come to see the truth of (T) directly, as it were. Though we have inferred the truth of (T), we haven’t come to understand the goodness of God in such a way that we can just see that that sort of goodness is compatible with the permitting of this or that evil. If we could identify particular goods that justify God, then we would have the right sort of direct appreciation of (T). But there may be another way of getting that. In Wandering in Darkness, the book based on her Gifford Lectures, Eleonore Stump (2010) aims to provide a theodicy via reflection on biblical narratives that portray God’s dealings with suffering human beings.³ Stump’s treatment of the biblical narratives is masterful, well worth extended discussion and reflection in its own right. For our purposes here, however, what is most relevant is just her methodology, and her justification for that methodology. For if what she says in defence of her particular way of approaching the problem of evil is correct, then it ³ Stump 2010. Editorial Note: In the originally published version of this chapter in the journal Philosophia Christi in 2008, I made reference to and quoted from the manuscripts of Stump’s Gifford Lectures. Here, I refer instead to the book since that is what is most readily accessible to readers, and I have updated the quotation below so that it matches what is in the book. The differences between the manuscript and the book make no difference for purposes of my discussion here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

109

is possible to arrive at theodicy via reflection on biblical narrative without ever supposing that it is possible for us to discover God-justifying reasons for permitting particular instances of suffering. The main novelty in Stump’s approach to the problem of evil (at least from the point of view of the analytic philosophical tradition within which she writes) is her attempt to arrive at theodicy by way of an extended literary-critical treatment of a variety of biblical narratives. Rather than treating, say, the story of Job or the story of Abraham and Isaac as brief toy examples to illustrate particular principles that then go on to get discussed in the analytic mode, she presents these and other stories in full detail, tries as much as possible to help us understand the motives, desires, and experiences of the characters involved, and aims in so doing to show us that we can learn philosophically useful things from the narratives that simply cannot be expressed propositionally. I cannot here present her full defence of this methodology; but I can at least offer a brief sketch. First some terminology. According to Stump, the biblical narratives that she discusses—the story of Job, the story of Abraham and Isaac, the story of Samson, and so on—are second person accounts of the events they relate. A second person account, in her terminology, is just a narrative that communicates the content of a second person experience. A second person experience is, roughly, a conscious experience of another conscious person as a person. A conversation with your child, an exchange of glances at a coffee shop, a hug—these are paradigm instances of second person experiences, to be contrasted, say, with a surgeon’s ‘objectifying’ experience of an unconscious patient on an operating table or with your own conscious awareness of yourself. This terminology in hand, we can now present the main elements of Stump’s defence of her methodology. (1) Second person experiences provide a particular kind of non-propositional knowledge—knowledge that simply cannot be conveyed as ‘knowledge that’. (2) Second person accounts are able to communicate roughly the same knowledge that one gets from a second person experience by making that experience available to us through the narrative. (3) Many biblical narratives—in particular, the ones Stump considers in the manuscript—are second person accounts of suffering human beings’ second person experiences of God. (4) Some of the knowledge gained via biblical narrative can be put to philosophical use in formulating a theodicy. If all of these claims are right, then there will be value for the project of theodicy in trying to address the problem of evil via extended consideration of biblical narratives. As I have said, I cannot here provide in detail Stump’s defence of these

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

110

   :  

claims. But I do want to provide a very brief indication of why she endorses the first claim; and I will also register a concern about the fourth claim. On behalf of the first claim, Stump adapts a familiar example from the literature on materialism in the philosophy of mind. In the familiar example, we are invited to consider a woman—Mary—who has all of the physical information about the colour red that one might ever want, but who has never actually experienced the colour red.⁴ We are then invited to ask whether Mary learns anything when she experiences red for the first time. Many think that she does; and those who think this typically say that what she learns—roughly, what it is like to experience redness—is something that goes beyond any of the physical information that she has. Various conclusions have been drawn from this example; but at least one conclusion one might draw (if one is convinced that Mary learns something) is that some of the things we might learn cannot be expressed propositionally. (For if what it is like to experience red could be expressed propositionally, it would have been among the information she had prior to experiencing it—in which case she wouldn’t have learned anything upon experiencing it.) And likewise, Stump says, if we suppose that Mary had been deprived not of the experience of red but of second person experiences of other persons, in this case too she would learn something upon first meeting another person (say, her mother)—and this would be so even if she had all of the propositional information she could possibly want about persons, personal interactions, and the particular person whom she met for the first time. This, Stump argues, lends support to the idea that there is information conveyed in second person experience that cannot be expressed propositionally. If we take this claim seriously, however, the fourth element of Stump’s defence of her methodology starts to look problematic. To see the problem clearly, it will help to quote a brief passage from her manuscript. She writes: The problem of suffering is, in a sense, a question about interpersonal relations, insofar as the problem has to do with possible morally sufficient reasons for God, an omnipotent, omniscient, perfectly good person, to allow human persons to suffer as they do. The narratives in the subsequent chapters present for us direct interaction between a loving God and suffering human beings. They are therefore, in effect, descriptions of (part of) a world in which God exists and has a morally sufficient reason for allowing human beings to suffer. If we can learn from the narratives the [non-propositional] knowledge they have to convey, then we can use that knowledge in the . . . philosophical project of formulating a defense and spelling out the nature of God’s morally sufficient reason. (Stump 2010: 61)

⁴ Some of the most important discussions of this argument can be found in Ludlow et al. 2004.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

111

If this is right, however, then we face the following puzzle: If what we glean from the second person accounts provided by the biblical narratives is (as Stump seems to suggest) a non-propositional understanding of what justifies God in permitting human beings to suffer, how could that knowledge possibly be put to use in spelling out (presumably propositionally) a story about what justifies God in permitting human beings to suffer? On the other hand, if the narratives give us the resources for an explicit theodicy, then it would seem that whatever knowledge we gain from them that is of relevance to the project of theodicy could, in principle, be gained apart from reflection on the biblical narratives—in which case such reflection is not nearly as indispensable as Stump seems to think that it is. I am not entirely sure how Stump would want to resolve this tension; but neither am I convinced that it is irresolvable. In any case, my goal here is not to press the objection but rather to point out that, regardless of whether what we learn from biblical narrative can be put to use in generating an explicit theodicy, if Stump is right about the other claims that support her methodology, reflection on biblical narrative might nevertheless provide us with all of the resources for a satisfying (even if non-explicit) theodicy. The reason is simple: As Stump notes in the passage just quoted, what we might glean from attention to the right sorts of biblical narratives is non-propositional direct insight into the fact that a perfectly good, loving God can be justified in permitting exactly the sorts and amounts of evil that we find in our world. In other words, we can come to see in a direct way the consistency of God’s goodness with the existence of evil; and so we can come to see in a direct way the truth of (T). But seeing the truth of (T) in a direct way suffices for theodicy. And, importantly, if Stump is right about claims (1)–(3), then seeing this is fully consistent with the suggestion that we might never be able to articulate, or even fully grasp, God’s specific reasons for permitting the suffering that he permits. Sceptical theism is left standing, hubris is avoided, but theodicy is nonetheless provided.

3. Conclusion In this chapter, I have considered some of what N. T. Wright has to say against the project of theodicy, and I have defended three conclusions: first, that one of his reasons for rejecting theodicy seems to be unsound; second, that his other, apparently main reason for rejecting theodicy is right on target; but, third, despite this fact, certain ways of doing theodicy escape the objection. Evil and the Justice of God is a rich and interesting book filled with excellent advice for how we ought to reflect on and address the evil in our world; and what I like most about the book is Wright’s insistence that we attend more fully to exactly what God himself has done to deal with evil, as well as his characteristically gripping and persuasive

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

112

   :  

story about what the biblical texts actually have to say on that score. My main goal in this chapter has simply been to argue that, far from having to set aside the philosophical problem of evil in order to follow Wright’s advice, following his advice might actually be a way of acquiring the resources for solving that problem.

References Bergmann, Michael. 2009. ‘Skeptical Theism and the Problem of Evil’. In The Oxford Handbook of Philosophical Theology, edited by Thomas P. Flint and Michael C. Rea, 374–99. Oxford: Oxford University Press. Ludlow, Peter, Yujin Nagasawa, and Daniel Stoljar (eds). 2004. There’s Something about Mary: Essays on Phenomenal Consciousness and Frank Jackson’s Knowledge Argument. Cambridge, MA: MIT Press. Stump, Eleonore. 2010. Wandering in Darkness: Narrative and the Problem of Suffering. Oxford: Oxford University Press. Wright, N. T. 2006. Evil and the Justice of God. Downers Grove, IL: InterVarsity Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

6 Sceptical Theism and the ‘Too-MuchScepticism’ Objection The recent literature on the problem of evil has focused on two kinds of ‘apparently gratuitous suffering’: that of non-human animals dying alone and in great pain, and that of children victimized by various forms of abuse.¹ Suffering like this does not seem to have any point, or purpose. It is hard to imagine how an omnipotent, omniscient, perfectly loving being could be justified in allowing such suffering. It is hard to imagine how the cost of preventing it might be the loss of some important greater good or the permission of something equally bad or worse. Consequently, it is widely held that reflection on suffering like this provides evidence that some of the suffering in our world is gratuitous. We do not have to turn to the extreme sufferings of children and animals to find fodder for discussion, however. Most people have experienced suffering, even as adults, that does not seem to have any justifying purpose. Even if we can see great goods that have come out of our sufferings, none of us, I submit, can see with clarity that there are goods both great enough and strongly connected enough to our sufferings to justify an omnipotent, omniscient, perfectly good and loving being in permitting every single moment of every painful experience that we have ever endured as adults. Suffering that does not seem to serve any suitably greater good is ubiquitous. Let us assume that all suffering is intrinsically evil, even if it contributes to great goods. Let us say that an evil is gratuitous if, and only if, there is no God-justifying reason for permitting it. Let us furthermore assume that there is a God-justifying reason for permitting some evil only if there is a good to which it contributes such that its contribution to that good would suffice to justify an all-powerful, allknowing, perfectly good and loving being in permitting the evil. Finally, let us assume that avoiding evils equally bad or worse might be among the goods to which an evil contributes.² Let E be the proposition that our world contains gratuitous evil. There are, then, two ways in which reflection on an instance of suffering might support E: (a) Perhaps we can just see directly that the suffering in

¹ I put ‘apparently gratuitous suffering’ in scare quotes because part of what is contested in this debate is precisely the claim that the suffering in question appears to be gratuitous. (Cf. Wykstra 1984.) ² These are standard assumptions in the literature. I am not sure that I wish to endorse all of them, but taking issue with them here would unnecessarily complicate the discussion.

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0007

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

114

   :  

question is gratuitous, without even bothering to speculate about what sorts of goods might justify it.³ (b) Perhaps, after speculating about what sorts of goods might justify it, we can justifiably infer from our failure to detect a God-justifying good that the suffering is, or is probably, gratuitous. There are several well-trodden routes from the claim that E is probably true or that we have evidence sufficient to justify belief that E is true to the conclusion that there is no God. Following standard convention, let us call these arguments, collectively, the ‘evidential problem of evil’. Sceptical theism is a response to the evidential problem of evil which opposes both (a) and (b). There are two components to sceptical theism: theism, and a sceptical thesis. Insofar as the sceptical thesis is separable from theism, the sceptical theist’s strategy for addressing the evidential problem of evil—namely, endorsement of the sceptical component— can be adopted by theists and non-theists alike (cf. Bergmann 2009: 375). As I shall characterize it, the central sceptical thesis of sceptical theism—the view that sceptical theism puts forth (at any given time) as a response to the evidential problem of evil—is this: ST No human being is justified (or warranted, or reasonable) in thinking the following about any evil e that has ever occurred: there is (or is probably) no reason that could justify God in permitting e.⁴ ST leaves open the possibility that an evil might someday occur about which we can justifiably think that it is gratuitous.⁵ Obviously enough, however, any decently principled defence of ST will imply that, if the world carries on pretty much as it has to date, with more or less the same sorts of evils continuing to occur, human beings will never be in a position to think justifiably about some evil that it is gratuitous. The most prominent objection against sceptical theism is that the sceptical theses typically adduced in support of ST have ramifications that range far more widely than sceptical theists hope or should tolerate: they lead to scepticism about various aspects of commonsense morality, about divine honesty and goodness, about the evidential value of religious experience, and much else besides. There are, in the literature, multiple ways of defending this objection. My view is that

³ Cf. Dougherty 2008, esp. sec. III; Gellman 1992. ⁴ I assume (for ease of exposition) that the variable ‘e’ ranges over aggregates of evils—e.g. the murder of twelve innocent victims, say, or even all the evil that has ever occurred—as well as individual evils. ST implies that sceptical theism is a view whose precise content changes over time (as new evils are added to the history of the world). Still, I think it is faithful to what sceptical theists actually say in response to the evidential problem. Note, too, that ST could, in principle, be accepted by someone who takes herself to be in possession of a theodicy. Whether a theodicy is available is one question; whether anyone is in fact justified in thinking of some actual evil that it is probably gratuitous is a wholly separate question. ⁵ Henceforth, for convenience, I use ‘justifiably’ to mean ‘justifiably, warrantedly, or reasonably’.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

115

none is successful. I do not, however, see any way of establishing that conclusion outright—that is, without considering and responding to each particular defence on its own terms. Since space limitations preclude me from attempting to take them all on at once, I shall restrict my focus to (multiple and various) defences of this objection that have appeared in papers published by Stephen Maitzen, David O’Connor, and Ian Wilks.⁶ Several of my replies will apply to similar arguments that have appeared elsewhere in the literature; but I do not pretend that anything I say here will lay all defences of the objection to rest. This chapter unfolds as follows. In Section 1, I characterize sceptical theism more fully. This is necessary in order to address some important misconceptions and mischaracterizations that appear in the essays by Maitzen, O’Connor, and Wilks. In Section 2, I describe the most important objections they raise and group them into four ‘families’ so as to facilitate an orderly series of responses. I respond to the objections in Sections 3–6.

1. What is Sceptical Theism? Sceptical theism has been characterized in various different ways. Let me begin with three examples. In the article that introduced the term ‘sceptical theist’, Paul Draper (1996) suggests (without offering an explicit definition) that a sceptical theist is someone who invokes a limited sceptical thesis in order to defend theism against the evidential problem of evil. The two theses under discussion in his article are these: ST1D Humans are in no position to judge directly that an omnipotent and omniscient being would be unlikely to have a morally sufficient reason to permit the evils we find in the world. ST2D Humans are in no position to compare theism’s ability to explain certain facts about good or evil to some other hypothesis’s ability to explain those facts. In his contribution on the topic to the Oxford Handbook of Philosophical Theology, Michael Bergmann offers a somewhat different characterization. He describes sceptical theism as a view with two components: theism and a sceptical component. The latter, he says, is best explained as an endorsement of some skeptical theses, among which these three are prominent:

⁶ Editorial Note: The papers under discussion here by Maitzen, O’Connor, and Wilks were published (along with the present chapter) in McBrayer and Howard-Snyder 2013.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

116

   :  

ST1B We have no good reason for thinking that the possible goods we know of are representative of the possible goods there are. ST2B We have no good reason for thinking that the possible evils we know of are representative of the possible evils there are. ST3B We have no good reason for thinking that the entailment relations we know of between possible goods and the permission of possible evils are representative of the entailment relations there are between possible goods and the permission of possible evils. (Bergman 2009: 376) ST1B–ST3B might plausibly be thought to lend support to ST1D and ST2D, but the former are neither individually nor jointly equivalent to either the latter, or to their conjunction. In his contribution on this topic to the Internet Encyclopedia of Philosophy, Justin McBrayer characterizes sceptical theism as theism plus the following thesis: ST1M We should be skeptical of our ability to discern God’s reasons for acting or refraining from acting in any particular instance. In particular . . . we should not grant that our inability to think of a good reason for doing or allowing something is indicative of whether or not God might have a good reason for doing or allowing something. (McBrayer 2010) Again, Bergmann’s sceptical theses and Draper’s might be thought to lend support to ST1M. But McBrayer’s thesis is not equivalent to any of those others, nor is it equivalent to a conjunction of any of them. These are by no means the only characterizations of sceptical theism in the literature either. So experts diverge to some extent on the question as to what, exactly, sceptical theism is. Moreover, neither Draper, Bergmann, nor McBrayer has identified a thesis to which all sceptical theists as such can be expected to agree and with which all opponents of sceptical theism can be expected to disagree. For example, I count myself a sceptical theist, but I reject ST1M because I think that we do have ways of discerning God’s reasons for acting on some particular occasions. (Scripture, for example, tells us that one of God’s reasons for becoming incarnate was love for the world.) Similarly, it is easy to imagine a sceptical theist accepting some of theses ST1D, ST2D, and ST1B–ST3B without accepting all of them. It is also easy to imagine someone accepting (say) ST2D or ST1B while at the same time maintaining (in opposition to sceptical theism) that we can see directly that certain evils are gratuitous. By contrast, my own characterization offers a thesis (namely, ST) to which all sceptical theists will agree and with which all opponents of sceptical theism will disagree. Maitzen, O’Connor, and Wilks all focus on ST1B–ST3B as somehow lying at the heart of sceptical theism. I will concede this for the sake of argument. But, in light

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

117

of the foregoing, I think that it is more accurate to think of them as comprising an important part of a typical defence of sceptical theism rather than to take their conjunction as part of sceptical theism. But let us set this quibble aside for now. Even having done so, it seems that all three authors are labouring under serious (albeit, in some cases, rather common) misconceptions as to the nature of sceptical theism. I will single out four for consideration. First: Maitzen says that ST1B–ST3B are ‘couched in broadly consequentialist terms, or at least they presuppose justifications couched in those terms’ (2013: 449). He talks on the next page about the ‘strongly consequentialist flavor’ of ST1B–ST3B Nor is he alone in making this association. Wilks also associates sceptical theism with consequentialism, though instead of finding it in ST1B– ST3B, he declares it to be a thesis that operates ‘in the background’ (2013: 458). But these claims are mistaken. The only hint of consequentialism in either sceptical theism itself or in ST1B–ST3B is the supposition that consequences are sometimes relevant to the moral status of an action. But, as Michael Bergmann argues (ironically, in a paper with which both Wilks and Maitzen are evidently familiar), ‘non-consequentialist ethical theories have no trouble allowing for considerations of consequences to play a role in moral decision-making’ (2009: 380). Second: According to Wilks, sceptical theism is a ‘stalemating technique’ whose ‘aim is to establish a confounding factor against the evidence of evil, and make that evidence inadmissible in debate over the existence of God’ (2013: 458). Note first that Wilks’s characterization of sceptical theism as a ‘stalemating technique’ uncharitably implies that the sceptical theist, as such, has abandoned the pursuit of truth in favour of simply stalling discussion. It also confuses replying to someone’s argument with trying to end the debate with neither party having ‘won’. More importantly, this characterization reflects confusion about the aims and content of sceptical theism. Sceptical theism does not deny of any actual evidence that it is admissible. Moreover, depending on one’s views about the nature of evidence, sceptical theists might even be happy to concede that reflection on certain kinds of suffering does, after all, provide evidence that our world contains gratuitous evil, albeit highly defeasible evidence. What the sceptical theist denies is that our awareness of (or reflection on) any actual or hypothetical instance of evil constitutes evidence sufficient to justify belief that our world contains gratuitous evil. Third: Wilks also claims that the ‘core of skeptical theism’ is ‘the thesis that we may not be aware of all the goods and evils there are, so we may not always be able to discern the reasons that justify God’s actions and permissions’ (2013: 458). He seems, furthermore, to think that this thesis implies that ‘[w]e should be skeptical about any claim to know what it would be evil for God to do or permit’ (458). Again, there are problems on multiple fronts. Sceptical theism is a controversial thesis; but the thesis that Wilks calls the core of sceptical theism is a platitude that should be affirmed by anyone. It is platitudinous that we human beings may not be aware of all of the various types (or tokens) of goods and evils there are; it is

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

118

   :  

likewise platitudinous that we may not always be able to discern the reasons that would justify a divine being’s actions and permissions. More importantly, it does not follow from these obvious truths—nor must sceptical theists affirm—that we should be sceptical about any claim to know what would be evil for God to do or permit. Most of us, sceptical theists included, think that it would be evil for God to do the following: permit horrendous suffering for absolutely no reason whatsoever. Indeed, if the sceptical theist had reason to doubt this claim, she would likely have a different reply to the evidential problem of evil. Nor is there any obstacle to a sceptical theist affirming (say) that it would be evil for God to permit a hundred people to be burned alive in a furnace just so that another person could enjoy the pleasure of a warm bath. Fourth: In O’Connor’s essay, we find the striking claim that it is a ‘basic tenet’ of sceptical theism ‘that we have no good reason to think that either our concepts or measures of goodness in human persons are representative of those applying to infinite, non-human persons’ (2013: 474). I assume that this is simply a failed paraphrase of ST1B–ST3B. If it is not, then I cannot see why he would attribute it to a sceptical theist. But it is important to see why it fails as a paraphrase. ST1B–ST3B claim that we have no good reason to think that our sample of possible goods, possible evils, and the relations among them is representative with respect to the property of being apt for justifying God’s permission of evil. They do not claim (or imply) that our concepts and measures of goodness fail to represent (or be a representative sample of) the concepts and measures of goodness that apply to infinite, non-human persons. Goods and evils are one thing; concepts and measures of good and evil are another. The claim that O’Connor puts in the mouth of a sceptical theist is far more radical than what the sceptical theist herself affirms.

2. The Too-Much-Scepticism Objection We are now ready to consider the too-much-scepticism objection. The core objection, again, is that the sceptical theist’s scepticism is infectious—it ramifies throughout her belief system, undermining a wide variety of beliefs about God, value, and other matters. Maitzen, O’Connor, and Wilks each defend this core general objection by arguing for some more specific version of it. Sceptical theism is the nominal target of the objection; but in the essays by Maitzen, O’Connor, and Wilks, it seems that the real target is something more in the neighbourhood of ST1B–ST3B. As noted earlier, I will concede for the sake of argument that an attack on ST1B–ST3B amounts to an attack on sceptical theism itself; but we should bear in mind that strictly speaking it is not sceptical theism that is the target of these essays but these other theses instead.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

119

The most important objections presented in the essays by Maitzen, O’Connor, and Wilks are sensibly grouped into four families, as follows: G S A. Sceptical theism ‘is implicated in skepticism of the extremist sort’, posing ‘a general problem about the possibility of knowing anything at all’ (Wilks 2013: 466, 459; cf. Maitzen 2013: 446–7). S  D C  V B. Sceptical theism implies that ‘we have no good reason to think that [terrible evils like the abduction, brutalization, and murder of a child] are sins’ (O’Connor 2013: 470). C. Sceptical theists have no good reason for believing that ‘what they regard as God’s “thou shalt not” commands are unrestrictedly forbidding certain actions as sins’ (O’Connor 2013: 470). D. Sceptical theists cannot take themselves to be able to ‘identify God’s commands, resolve questions about their relative importance, or apply them to [their] actual circumstances’ (Maitzen 2013: 451). E. Sceptical theism induces doubt about divine goodness and about our understanding of divine values (Maitzen 2013: 453; O’Connor 2013: 471–2). S  (O) K  G F. Sceptical theism undermines our ability to ‘[take] particular goods as signs of God’s presence’, and to ‘[see] things and events in the world around us as manifestations of God’ (Wilks 2013: 460). G. Sceptical theism undermines the belief that human suffering is not gratuitous (O’Connor 2013: 453–4). H. Sceptical theism precludes the warranted attribution of divine purposes to mundane events, thereby undermining belief in miracles and certain natural theological arguments (O’Connor 2013, 478–9). S  O O  P H I. Sceptical theism leads to scepticism about our obligation to prevent harm, and so it also leads to a kind of moral paralysis (Maitzen 2013: 451–3). In the next four sections, I will present the reasons given by Maitzen, O’Connor, and Wilks for thinking that theses A–I above are true. My ultimate conclusion will be that none of these claims has been shown to be true and so the different versions of the too-much-scepticism objection as defended by Maitzen, O’Connor, and Wilks, are failures.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

120

   :  

3. Global Scepticism Let us begin with the global scepticism objection. This is by far the most ambitious of the objections. But as Wilks has formulated it there is some question as to what exactly the objection is supposed to be. What does it mean to say that sceptical theism is ‘implicated’ in an extreme form of scepticism? What does it mean to say that sceptical theism poses ‘a general problem about the possibility of knowing anything at all?’ Here are some possibilities, none equivalent to the others: (i) Endorsing sceptical theism automatically provides one with an undefeated defeater for all of one’s beliefs. (ii) Endorsing sceptical theism automatically provides one with an undefeated defeater for most of one’s beliefs. (iii) Sceptical theism implies that every person (whether a proponent of sceptical theism or not) has an undefeated defeater for most of his or her beliefs. (iv) Proponents of sceptical theism who reflect rationally on its content and consequences have an undefeated defeater for most of their beliefs. There are others we might add to the list as well. Maitzen gestures at a similar objection, but he likewise refrains from stating it outright. Instead, what one finds in his paper are remarks like ‘skeptical theism implies radical skepticism’ (2013: 447)—a claim which he attributes to Bruce Russell (1996) and then goes on to defend without restating—and ‘theism threatens our knowledge’ (Maitzen 2013: 447). My guess is that both Maitzen and Wilks are getting at something like (iv), which I will henceforth refer to by the label ‘’. The objection I take them to be raising, then, is simply that  is true. Wilks provides the most detailed argument for the claim that sceptical theists are stuck with some kind of radical scepticism. Below is my best attempt at charitable and faithful reconstruction (treating  as the intended conclusion). Let ‘DH’ be the ‘Deception Hypothesis’—the hypothesis that we have come to be deceived (either by divine design or divine permission) in systematic and comprehensive ways. Let ‘reflective sceptical theists’ be just those sceptical theists who have rationally reflected on the content and consequences of sceptical theism. Then: 3.0

If sceptical theism is true, then God exists.

3.1

If God exists, then God has the power to bring it about that DH is true.

3.2 If 3.0 and 3.1 are true, then reflective sceptical theists cannot rationally deny that God exists and has the power to bring it about that DH is true.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

121

3.3 Therefore, reflective sceptical theists cannot rationally deny that God exists and has the power to bring it about that DH is true. (From 3.0–3.2) 3.4 Reflective sceptical theists cannot rationally affirm that it would be evil for God to bring it about that DH is true.⁷ 3.5 If one can neither rationally deny that God exists and has the power to bring it about that DH is true nor rationally affirm that it would be evil for God to bring it about that DH is true, then one cannot rationally dismiss DH as impossible. 3.6 Therefore, reflective sceptical theists cannot rationally dismiss DH as impossible. (From 3.3–3.5) 3.7 If sceptical theism is true, it is not absolutely unreasonable to believe that God has brought it about that DH is true for some justifying reason beyond our ken.⁸ 3.8 If 3.7 is true, then reflective sceptical theists cannot rationally dismiss DH as absolutely unreasonable to believe. 3.9 If sceptical theism is true, then one cannot rationally dismiss as false or unlikely the claim that God has a justifying reason for bringing it about that DH is true. 3.10 If 3.9 is true, then reflective sceptical theists cannot rationally dismiss as false or unlikely the claim that God has a justifying reason for bringing it about that DH is true.

⁷ According to Wilks, sceptical theism implies that we ‘should be skeptical about any claim to know what it would be evil for God to do’ (2013: 458, 461). It is only by supporting something like 3.4 that this remark is of relevance to Wilks’s ultimate conclusion. So that is why I take 3.4 to be part of Wilks’s argument. It is perhaps worth noting that this claim is also of relevance to 3.1. For if one could know that it is evil for God to bring about the truth of DH, then one would have reason to deny that God is able to bring about the truth of DH; and, on some ways of understanding the relation between divine ability and divine power, this would imply that God lacks the power to bring about the truth of DH. Obviously questions about the nature of omnipotence would then arise; but pursuing that issue would take us too far afield. ⁸ By ‘absolutely unreasonable’ I mean ‘unreasonable for any subject, regardless of what else she may believe’. So, to cast 3.7 in other terms: if sceptical theism is true, then it is at least possible for someone reasonably to believe DH. I attribute 3.7 to Wilks because he invokes the following premise from Bruce Russell en route to his conclusion: 3.7a. ‘If it is not reasonable to believe that God [has] deceived us, for some reason beyond our ken [with respect to the age of the universe], it is not reasonable to believe that there is some reason beyond our ken which, if God exists, would justify him in allowing all the suffering we see’. (Wilks 2013: 462; cf. Russell 1996: 197) As Wilks goes on to observe, there is nothing special about the age of the universe: one might replace ‘the age of the universe’ in the antecedent with just about any proposition about which it is possible for us to be deceived. Moreover, he seems to think that sceptical theism commits one to rejecting the consequent of 3.7a. (We might question this. But never mind that for now.) Given this, sceptical theism implies, for any proposition p about which human beings can be deceived: 3.7b. It is false that it is not reasonable to believe that God has deceived us with respect to p. This does not imply that it is reasonable for us to believe that God has deceived us with respect to p. What follows is just that it is not absolutely unreasonable to believe that we are so deceived. If that is true, then 3.7 is true.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

122

   :  

3.11 Therefore, reflective sceptical theists cannot rationally dismiss DH as absolutely unreasonable to believe, nor can they dismiss as false or unlikely the claim that God has a justifying reason for bringing it about that DH is true. (From 3.7–3.10) 3.12 Citing evidence against DH presupposes the falsity of DH (since any proposition that might be cited as evidence against DH will be a proposition about which we might be deceived). 3.13 If one cannot rationally dismiss DH as impossible or absolutely unreasonable to believe, and if one cannot rationally dismiss as false or unlikely the claim that God has a justifying reason for bringing it about that DH is true, and if citing evidence against DH presupposes the falsity of DH, then one has a defeater for most of one’s beliefs. 3.14 Therefore, reflective sceptical theists have an undefeated defeater for most of their beliefs. (From 3.6, 3.11–3.13) As I have reconstructed it, the argument is valid. Thus, I shall focus my critical remarks on the premises—specifically, premises 3.4, 3.7, 3.9, and 3.13. Before turning to criticism, however, I wish to make two preliminary observations. First, premises 3.7, 3.9, and 3.13 employ the notion of justifying reasons. I have employed this notion because Wilks does so in his own presentation of the argument (see note 7). But we should bear in mind that doing so represents a subtle shift in the topic of conversation. ST and ST1B–ST3B are all theses about what human beings can justifiably believe about the space of possible goods and evils. Goods and evils are interestingly connected with reasons, but it is controversial at best to identify them with reasons. According to some philosophers, for example, to call something a reason for action is to say something about its relationship to the relevant agent’s motivational structure; but calling something a good does not necessarily carry such implications. The significance of this fact in the present context is just this: There seems to be very good reason for thinking that sceptical theism implies that we cannot dismiss as false or unlikely the claim that there is some good capable of justifying God in bringing about DH. But, by virtue of the difference between justifying goods and justifying reasons 3.9 does not obviously follow from this claim. Further argument would be required, which argument Wilks has not provided. The second observation is that the premises that invoke the notion of justifying reasons omit any distinction between pro tanto reasons and decisive reasons. Pro tanto reasons for doing something are considerations that weigh in favour of doing it. Decisive reasons settle rational deliberation in favour of a particular course of action.⁹ So, since God is maximally rational, God has decisive reason ⁹ Cf. Murphy 2006: 413. Murphy, in turn, follows Raz 1979.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

123

to do x if, and only if, God actually does x. Given this, the term ‘justifying reason’ in premises 3.9 and 3.13 cannot sensibly be interpreted as meaning decisive justifying reason. For if we do interpret it that way, then 3.9 entails that it is a consequence of sceptical theism that we cannot dismiss as false or unlikely the claim that DH is true. But that claim is not at all obviously true, nor has Wilks argued that it is true. Thus, ‘justifying reason’ must be interpreted in this argument as meaning pro tanto justifying reason. Doing so renders premise 3.9 more plausible. (More exactly: it renders 3.9 more plausible on the assumption that 3.4 and 3.7 are true.) But it raises other problems that shall become apparent later in this section. I turn now to critical remarks. In connection with 3.4, 3.7, and 3.9 all I shall say is this: Wilks has not argued, nor is there any obvious reason to believe, that the following propositions are inconsistent: 3.15

Theism, ST, and ST1B–ST3B are all true.

3.16 It would be evil for God to bring it about that DH is true, and this fact can be known to be true on the basis of rational intuition. That is, there is no clear way of deriving the denial of 3.16 from the proposition that (a) God exists, (b) we cannot be justified in believing of any actual evil that it is gratuitous, and (c) we cannot tell whether the sample of possible goods, evils, and connections among them of which we are aware is representative. But this is precisely what would have to be derived in order to defend any of 3.4, 3.7, or 3.9. Bear in mind here that the denial of DH is, like the denial of any other radical sceptical hypothesis, a perfectly rational starting point. If it were not so, everyone would face insurmountable sceptical threats.¹⁰ But from the denial of DH we can deduce that God lacks decisive reason to bring it about that DH is true. This comports well with our independent intuition that it would be evil for God to bring it about that DH is true. Thus, there is no reason to think that our intuition on this score is challenged by the recognition that our sample of possible goods, evils, and connections among them might not be representative. That said, I think that some sceptical theists will not wish to affirm what 3.4 says they cannot rationally affirm, and some will also want to accept the claims that 3.7 and 3.9 say that sceptical theism implies. Moreover, though Wilks himself

¹⁰ Wilks might concede this much but insist that the sceptical theist’s scepticism precludes her from taking DH as a starting point. But, of course, that is a thesis that would have to be supported by argument, as it is hardly self-evident. A more interesting move would be to concede the point and then to ask why it does not apply to sceptical theses like ST1B–ST3B. That is, why are we not entitled to take the denials of those theses as perfectly rational starting points? Since this question is not raised by Maitzen, O’Connor, or Wilks, and since it has (in my opinion) been adequately addressed elsewhere (Howard-Snyder 2009: 25–8), I shall simply refer the reader to that discussion.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

124

   :  

has not defended 3.4, 3.7, and 3.9, I think that most philosophers will take at least one of those premises to be defensible. So, to my mind, the really interesting question here is whether 3.13 is true. I think that it is not. Suppose we concede the following (which is just a slightly clarified version of the antecedent of 3.13 specified to a particular subject, S): 3.13a (i) S cannot rationally dismiss DH as impossible or absolutely unreasonable to believe; (ii) S cannot rationally dismiss as false or unlikely the claim that God has a pro tanto justifying reason for bringing it about that DH is true; and (iii) citing evidence against DH presupposes the falsity of DH. Now, what follows? It seems to me that if anything here implies that S has a defeater for most of her beliefs, it is 3.13a(ii), and this by way of implying that she has a defeater for her (tacit or explicit) belief that DH is false. 3.13a(i) implies that the defeater cannot itself be defeated by S’s rationally coming to the belief that DH is impossible or unreasonable to believe; and 3.13a(iii) implies that the defeater cannot be defeated by evidence. So the next question to ask is whether 3.13a(ii) really implies that S has a defeater for most of her beliefs. Let me pause to note that Wilks and Maitzen both seem to think that the answer is ‘yes’. In discussing Bergmann’s ‘commonsensist’ view that we know via commonsense that we are not victims of radical sceptical scenarios, Wilks argues that there is a difference between the sceptical theist’s position vis-à-vis DH and her (and everyone else’s) position vis-à-vis more familiar sceptical scenarios. For most sceptical scenarios, we can dismiss as false or unlikely the hypothesis that there is a being willing and able to enact them; but, according to Wilks, the sceptical theist already admits the existence of a being who is able to bring about the truth of DH, and she cannot dismiss as false or unlikely the hypothesis that this being is also willing to bring it about. This difference, he thinks, is sufficient to plunge the sceptical theist into radical scepticism. But that consequence follows only if 3.13a(ii) implies that S has a defeater for most of her beliefs. Maitzen, in presenting an argument very similar to Wilks’s, gestures at the same point (which seems to be the very crux of his argument): [If] fraud and deception are . . . consistent with God’s perfection, [then] we can rule them out only by presuming that God can have no morally sufficient reasons for committing them. (Wilks 2013: 447)

This seems likewise to presuppose that 3.13a(ii) implies that S has a defeater for most of her beliefs. But I think that we can see clearly that 3.13a(ii) does not imply that S has a defeater for the belief that DH is false. If it does not imply this, then it also does not imply that she has a defeater for more ordinary beliefs. Consider the following case: You have just been diagnosed with cancer. A doctor whom you do not know very well (so not your doctor) tells you that if

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

125

you maintain a healthy lifestyle and are diligent about your treatments, your chances of beating the cancer are very good. You have no independent information on this topic (and no opportunity to get independent information any time soon), and you have no reason to suspect that this doctor is deceiving you. So you believe her and you go home feeling quite good. Still, upon reflection you realize that you cannot tell how likely it is that she would have a pro tanto justifying reason to deceive you about your chances of beating the cancer. You can imagine some reasons a doctor—even a highly virtuous doctor—might have for deceiving you. You have heard that there are correlations between optimism about survival and increased survival rates, for example. But you have no idea whether such correlations apply to your case, nor do you even have any idea whether your sources of information on this topic are reliable, nor do you have any clue whether or to what extent such considerations would weigh, in her estimation, in favour of deceiving you. Furthermore, you realize that pro tanto reasons come very cheaply: any consideration in favour of doing something, no matter how weak, constitutes a pro tanto reason for doing it. So, you reflect, obviously you can have no clear idea how likely it is that the doctor has a pro tanto reason for deceiving you. Do you now have a defeater for your belief that this doctor is telling you the truth (or for your belief that your chances of beating the cancer are good, so long as you follow the relevant advice)? No. Clearly you do not. It is perfectly rational for you to dismiss the doctor-deception hypothesis on the grounds that you have no good reason to think that it is true. Or consider this case: You go to the supermarket and discover that they are out of your favourite kind of beer. You ask the store clerk where else in town you might find some. She says that she has checked everywhere within a ten-mile radius and nobody else carries it. (It is a special import.) You find that plausible; it fits your experience. But, upon reflection, you realize that you have no idea what the likelihood is that she has a pro tanto justifying reason for deceiving you on this topic. You do not know her at all—you just met her today. So you have no idea how considerations of loyalty to her place of employment (say), or the desire to appear more thorough in her beer-searching than she really has been, would weigh against whatever propensities towards truth-telling she might have. And, again, you recognize that pro tanto reasons come cheaply. Do you now have a defeater for your newly acquired belief that no one else carries your favourite brand of beer? Again, you do not. You are rationally entitled simply to dismiss the clerk-deception hypothesis on the grounds that you have no good reason to think that it is true. As a general rule, we are justified in believing the testimony of others in the absence of defeaters that either rebut their testimony or cast doubt on their reliability. So, as a general rule, we are justified (absent defeaters) in dismissing both the hypothesis that the testifier is deceiving us and the hypothesis that she has a decisive reason for deceiving us. But, again, pro tanto reasons come cheaply. So our ignorance of other people’s values, desires, particular needs and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

126

   :  

circumstances, and so on means that we often will have no idea how likely it is whether they have pro tanto reasons for deceiving us or for doing all manner of other bad or crazy things. Knowing this, it is not rational simply to dismiss the hypothesis that testifiers whom we do not know have pro tanto reasons for deceiving us. Nor, however, should we think that our inability to dismiss this hypothesis constitutes a defeater for our presumption that they are in fact telling the truth. For if it did constitute such a defeater, we would almost never be justified in trusting testimony. What goes for the above cases goes likewise for the divine case. Grant that ST1B–ST3B imply that we cannot dismiss as false or unlikely the hypothesis that God has a pro tanto reason for deceiving us in significant ways. Still, for the sorts of reasons just given, this in no way implies that we cannot dismiss as false or unlikely the hypothesis that DH is true; nor does it supply us with a defeater for our presumption that DH is probably false; nor, therefore, does it supply us with a defeater for more ordinary beliefs grounded in our trust in sense perception, testimony, reason, and the like.

4. Scepticism about Value Both Maitzen and O’Connor give arguments for the conclusion that sceptical theism poses a problem for our beliefs about putative divine commands: we cannot know exactly what God means by them, we cannot know what God intends for us to do in response to them, and we cannot know what, if anything, they reflect about divine values. For O’Connor, this scepticism about divine commands also induces scepticism about what counts as sin. Furthermore, he thinks that sceptical theists face problems in believing that God is good. I’ll discuss each of these arguments in the order listed. First, divine commands. In sum, O’Connor’s argument runs as follows: () Philosophically informed, reflective, and otherwise rational theists will think that our understanding of the content and normative force of God’s commands comes only from one or more of the following sources: religious experience, scripture, church teaching, or reason. However, (ii) what they will know about religious experience provides them with an undefeated defeater for the belief that religious experience is a reliable source of information about God’s intentions, values, or commands.¹¹ Furthermore, (iii) sceptical theists have an undefeated defeater for ¹¹ O’Connor says simply that philosophically minded sceptical theists will have ‘good reason to hesitate’ before taking religious experiences at face value and so ‘may, plausibly, become less than quite confident about believing that [their] own purported revelation experiences are truly experiences of a divine being or of a supernatural surrogate’ (2013: 471). These remarks fall somewhat short of the stronger claim in (ii). However, replacing the stronger claim with these remarks leaves the sceptical theist with an easy reply. She might well have ‘good reason to hesitate’ and even to be less than ‘quite

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

127

the view that church teachings and scriptural texts, and human reason are reliable sources of information about God’s intentions, values, or commands. Reason is ruled out because (iv) sceptical theism implies that human moral judgments and moral values are not reliable indicators of divine values and commands.¹² Scripture and church teaching are ruled out because each is the product of human attempts to interpret putative divine communication and (v) sceptical theism implies that that we have no good reason to believe that such attempts will reliably succeed.¹³ Therefore, (vi) sceptical theists have an undefeated defeater for the belief that they understand the content and normative force of God’s commands, and so (vii) they have no good reason to think that ‘what they regard as God’s “thou shalt not” commands are unrestrictedly forbidding certain actions as sins’ (O’Connor 2013: 470–2). There are two general (and related) problems with this argument. The first is that, as we have already seen (in Section 1), O’Connor has confused sceptical theism with more radical claims about divine transcendence. Were he correct in his understanding of sceptical theism, premises (iv) and (v)—the ones that report what sceptical theism implies—might be quite plausible. As it stands, however, there is no obvious reason for thinking that those premises are true. That is, there is no obvious reason for thinking that ST or ST1B–ST3B imply what O’Connor says that sceptical theism implies. Which brings us to the second problem: O’Connor has offered no argument in support of premises (ii), (iv), or (v). This is startling in light of the absolutely crucial role these premises play in establishing his final conclusion. But there it is anyway.¹⁴ He seems to think that the truth of confident’ that her religious experiences are experiences of God. But all of this is consistent with the claim that philosophically minded sceptical theists can ultimately be rational in taking their religious experiences to be transparently intelligible communications from God. ¹² O’Connor writes: ‘it is a basic tenet of [the sceptical theist’s] skeptical defense that we have no good reason to suppose that actual or possible goods, evils, and the entailment relations between them, as we are aware or can conceive of them, are representative of such things to God, no matter how good and bad the goods and evils respectively are to us or how inconceivable to us a justification for certain actions and kinds of actions may be. So, on the basis of that, she has no good reason to think that, when it comes to the particular circumstances of a particular homicide, theft, abduction, or sexual abuse of a child, or even when it comes to considering such things in the abstract, how she or the rest of us may see them is representative of how they are in the eyes of God’ (2013: 472). ¹³ O’Connor says: ‘given . . . skepticism about supposing that possible goods, evils, and entailment relations between them, as we are aware of those things, are representative of such things to God, it is only a small step to justifiably thinking that, by virtue of the same gap between human and divine cognition that drives [the first sort of scepticism, one] also has no good reason to think that how she interprets supposed experiences of, or communications from, God is representative of their meaning in the mind of God or of what God . . . intends her to understand by them’ (2013: 471). ¹⁴ A casual glance at the remarks quoted in note 12 might lead one to think that premise (iv) is defended by argument. In fact, however, those remarks simply report (inaccurately) a basic tenet of sceptical theism and an alleged consequence of that tenet. There is no argument for the claim that sceptical theism (or the alleged basic tenet thereof) implies the consequence in question. Roughly the same is true of premise (v) and the remarks quoted in note 13. As for premise (ii), O’Connor (2013: 471) notes some facts awareness of which he takes to constitute a defeater for beliefs formed on the basis of religious experience: e.g. facts about the ‘prima facie appalling actions carried out by people,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

128

   :  

these claims is just obvious, and this despite the fact that (e.g.) philosophers like William P. Alston (1991) and Alvin Plantinga (2000) have devoted hundreds of pages to rebutting (ii), and despite the fact that most sceptical theists believe the denials of (iv) and (v). Highlighting these two problems constitutes sufficient rebuttal to O’Connor’s argument. But I think it will be instructive to pause to consider the profile of a particular (fictional) sceptical theist so that we can see more clearly the argumentative burden that O’Connor faces. Moreover, it is instructive not only in relation to the argument we are presently considering, but also in relation to other arguments by Maitzen, O’Connor, and Wilks that I shall consider below. Thus, we shall have cause to refer back to this profile when we consider those arguments later on. In presenting the profile, I shall fill in a variety of background beliefs and circumstances that I think make some difference in our overall assessment of the subject’s rationality. The profile presupposes that premises (ii), (iv), and (v), as well as some additional premises that show up in later arguments, are all false. My expectation is not that readers will take the profile to refute these premises; rather, it is that they will see that more detailed and careful argumentation is required to establish them. The premises whose falsity I shall presuppose are intuitively appealing, it seems to me, only when one is not sufficiently attentive to the variety of ways in which the moral, religious, and philosophical beliefs of sceptical theists can be acquired, combined, and prima facie justified. I expect that readers will find my description of the profiled subject to be psychologically plausible; I also expect that they will not regard her as manifestly irrational. But the fact is that, if the arguments considered in this section and the next are correct, the person I shall describe must be irrational. The challenge, then, for the proponents of those arguments is to show why she is irrational. I do not think that this challenge has been met. Along the way I shall explain why. Consider Lucy, a Christian who has come to hold a wide variety of beliefs about God and God’s commands in a way that is quite common for reflective Christians. She has listened to the testimony of respected authorities, read the Bible from cover to cover, and reflected on the nature, love, and goodness of God over the course of many years during Bible studies, church services, and her own personal times of prayer and meditation. In reflecting on these things, she has attended carefully to how all of her various sources portray God’s love and goodness, and she has also attended to how well (and not) these portrayals comport with her own intuitions about love and moral perfection. Religious experience has while supposedly inspired by personal experience of the voice or will of God’ or about (the dangers of?) ‘wishful thinking, emotional need or projection, personification, [etc.]’. But he offers no argument for the conclusion that awareness of such facts inevitably supplies one with a defeater for beliefs formed on the basis of religious experience, or for belief that one’s own religious experience is a reliable source of information about divine intentions, values, or commands.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

129

entered the picture too. Like many theists, she has had a number of what we might call ‘low-grade’ religious experiences: the feeling of blessed assurance that she is loved and forgiven by God, the strong sense of God’s presence in the world and to her in particular, and so on. She believes these experiences to be veridical; and, although they do not carry much content about the particular nature of divine love and goodness, they serve to reinforce in her the conviction that God is loving and good. In forming all of these beliefs, she takes it for granted that her moral intuitions are generally (though far from perfectly) reliable, that she has a fair bit of moral knowledge, and so on. But she also takes it for granted that God’s goodness far outstrips her own, that her moral intuitions are biased in many ways by her own self-interested desires, and that (as she has learned to her chagrin over and over throughout her life) she is subject to moral ‘blindspots’ and, in particular, to errors (sometimes serious ones) about what might justify what. In this her condition is no different from that of virtually any other morally decent—indeed, even morally excellent—person. She knows this, and she does not find in this knowledge any reason to think that fallible human beings like herself are doomed to the abyss of radical moral scepticism. So far so good. There is nothing obviously irrational in Lucy’s overall set of beliefs; nor is there any clear reason for thinking that she is wrong to reject moral scepticism. Now suppose she adds to this belief set ST, and ST1B–ST3B. That is, she becomes a sceptical theist. Suppose further that she adds these beliefs partly on the basis of her own keen awareness that no good she knows of seems even close to sufficient to justify a perfectly loving God in permitting the severe psychological and sexual abuse that she herself suffered at the hands of a neighbour when she was a child. In adding ST and ST1B–ST3B, she does not moderate any of her other moral beliefs, because she takes those to be independent of her assessments about how representative her sample of possible goods, evils, and connections among them might be. Indeed, she realizes that it takes a great deal of moral knowledge even to sort her sample of states of affairs into the categories of possible good and possible evil. It also takes a great deal of moral knowledge to reach the judgment (as she has) that no known good is sufficient to justify God in permitting all of the suffering we see in our world (or even all of her own). She has therefore not come to question her sorting procedure, nor has she come to question any of the other moral knowledge she has. She simply realizes that for all she knows, her sample might be comparatively small and strongly biased—perhaps in the direction of human goods and evils, perhaps in the direction of humanly graspable goods and evils, perhaps in the direction of relatively superficial goods and evils, etc. Now, let us ask: Does she have ‘good reason to hesitate’ before thinking that she has good reason to take her religious experiences at face value? Should she become ‘less than quite confident about believing that her own purported revelation experiences are truly experiences of a divine being’? Premise (ii) says that the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

130

   :  

answers to these questions are both affirmative. But O’Connor has said nothing that supports this verdict (see again notes 11 and 14), nor is it plain from the description of the case that the questions should be answered affirmatively. Does she have good reason to doubt her reflective interpretations of her own religious experiences? Premises (ii) and (v) imply that the answer is yes. But, again, neither anything that O’Connor has said nor anything manifest in the description of the case seems to support an affirmative answer (cf. notes 11, 13, and 14). So if premises (ii) and (v) are true, it is not at all obvious that they are. What about Lucy’s own moral judgments? Suppose she is thinking now about whether it is permissible to divorce her husband. She finds in scripture the claim that ‘God hates divorce’, which she takes to be evidence that God does not generally permit divorce—at least not for the reasons why she would be getting a divorce. But her moral intuitions tell her that divorce for those reasons is entirely permissible. After digging into the scriptures she finds that the waters in the neighbourhood of this question are quite murky. So she finds herself with reason to doubt that ‘God hates divorce’ implies what it seems to imply. On the other hand, she realizes all too well that she has a strong personal stake in having the ‘verdict of scripture’ come down in favour of the permissibility of her divorce. She will feel much less conflicted about the course of action she is considering if she can convince herself that God does not object. So, in other words, she realizes that her interpretation of scripture might be tainted by self-interest. She also realizes that her intuitions might be so tainted as well. Does she now have reason to doubt that her own moral judgment about the moral status of divorce is ‘representative of its moral status in the eyes of God’? Yes, of course. This is not because she is a sceptical theist, however. It is because of a prior belief that partly grounds her sceptical theism—namely, the belief that she is morally fallible and subject to bias, and that God is not morally fallible. But she still does not seem to have any reason to question beliefs like ‘God detests murder’ or ‘God disapproves of adultery’. Again, as a sceptical theist she thinks that her sample of possible goods, evils, and connections between them is biased. She recognizes that she is fallible in other ways, and that self-interest and other factors sometimes cloud her moral judgment. But she has not given up thinking that she is nonetheless pretty good at distinguishing goods from evils; and moral intuition screams out to her—wholly in accord with the voice of scripture— that murder and adultery are wrong, and that God therefore disapproves of them. Do ST, or ST1B–ST3B even so much as suggest that her beliefs here are irrational? They do not. At any rate, O’Connor has given no argument for the conclusion that they do (cf. notes 12 and 14). So premise (iv) stands in need of substantial defence as well. Maitzen also offers an argument from sceptical theism to scepticism about divine commands. The first part of the argument, with which I shall not take issue,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

131

proceeds as follows (note that I am paraphrasing, not quoting, from Maitzen 2013: 451–3): We find apparently conflicting commands in the scriptures of the different theistic traditions. None of these putative revelations is self-authenticating; so theists must decide which to take as genuine. Moreover, within the scriptures of each tradition, we find commands (e.g. to avoid mixed-fibre clothing or to stone disobedient children) that adherents of the relevant tradition regard as having been superseded by other commands or as being for other reasons inapplicable or not reflective of genuine divine preferences about the behaviour of contemporary human beings. In order to decide which to take as genuine, we must determine which to take as most likely expressing God’s will and intentions. Therefore, identifying God’s genuine commands requires human insight into God’s reasons and intentions. Let us grant the sub-conclusion. Let us further grant (as Maitzen seems to affirm) that just as human insight into God’s reasons and intentions is needed in order to identify God’s commands, so too it is needed to ‘resolve questions about their relative importance, or apply them to our actual circumstances’ (Maitzen 2013: 451). Then: 4.0 One can identify God’s genuine commands, resolve questions about their relative importance, or apply them to one’s actual circumstances only if one has human insight into God’s reasons and intentions. At this point, Maitzen introduces a further premise, followed by a rhetorical question: 4.1 One has human insight into God’s reasons and intentions only if one ‘independently understand[s] the realm of value [i.e. understands it in a way not dependent upon our prior beliefs about the content of God’s commands] well enough to tell which acts and omissions a perfect being would be likely to command’ (Maitzen 2013: 452). 4.2 ‘How then can we understand the realm of value well enough to tell which actions and omissions a perfect being would be likely to command and yet, as skeptical theists insist, not understand that realm well enough to tell which cases of horrific suffering a perfect being would be at all likely to permit?’ (Maitzen 2013: 452). I assume (but am not certain) that the question is supposed to function as if it were equivalent to the following conjunction:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

132

   :  

4.3 If sceptical theism is true, then (a) we do not understand the realm of value well enough to tell which cases of horrific suffering a perfect being would be at all likely to permit; and if (a) is true, then (b) we cannot understand the realm of value well enough to tell which actions and omissions a perfect being would be likely to command. Replacing 4.2 with 4.3 introduces no additional problems into Maitzen’s argument, and it has the virtue of rendering valid his inference to the following conclusion: 4.4 Therefore, ‘[if sceptical theism is true, then] we can’t identify God’s commands, resolve questions about their relative importance, or apply them to our actual circumstances . . . ’ (Maitzen 2013: 451). Still, the argument is unconvincing, for there is no reason to think that either 4.1 or 4.3 is true. Consider Lucy. Is her understanding of the realm of value independent of her views about what God has commanded? Apparently not. Her moral beliefs are intimately connected to her religious upbringing. Indeed, we may speculate that (like many Christians) she grew up thinking that moral value is grounded somehow in divine preferences and commands. Does it follow from this that she ‘lacks the requisite insight to identify God’s genuine commands’? Hardly, and this for two reasons. First, Maitzen has not ruled out the possibility of reliably arriving at beliefs about divine commands via routes other than reflection on facts about value. Suppose Christianity is true, and suppose the Christian scriptures are—when properly understood—reliable sources of information about divine commands. Suppose further that careful, prayerful attention to the scriptural texts themselves and to the vast body of commentary on those texts within the Christian tradition will resolve many, even if not all, questions about which putative commands in scripture are genuine and which are merely apparent, and about which are still applicable and which are not. Many Christians, including some who are sceptical theists, will affirm these suppositions; and Maitzen has given no argument for the conclusion that they are false. But if they are true, then (contra Maitzen) someone like Lucy may very well not ‘lack the requisite insight to identify God’s genuine commands’. Second, it seems clear that all one really needs in order to discern some facts about what a perfect being would be likely to command is knowledge that there is at least one type of act that is morally wrong. But so far as I can tell—and so far as Maitzen has argued—there is no reason to think that one can have such knowledge only if one’s grasp of the realm of moral value is independent of one’s beliefs about God’s reasons and intentions. Likewise, premise 4.3 has little to recommend it either. Suppose Lucy considers her belief that the commands in the Decalogue are indeed God’s commands, and

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

133

suppose she asks herself whether it is likely that a perfect being would issue a command prohibiting (say) murder. Moral intuition says that murder (the killing of someone for no justifying reason) is wrong, and she cannot imagine a perfect being approving of murder. So she reasons that it is indeed likely that a perfect being would issue such a command. Neither ST nor ST1B–ST3B raise any obstacle to these thoughts on her part; nor does her acceptance of those theses supply her with a defeater for the belief that it is likely that God would prohibit murder. Indeed, it does not even supply a defeater for her more ‘speculative’ beliefs, e.g. that it is likely that God would not approve of one person’s severely abusing another. She acknowledges that her sample of possible goods, evils, and connections among them may not be representative; she is likewise firmly of the conviction that we cannot be justified in believing (for example) that the abuse she suffered as a child is gratuitous. But these beliefs of hers obviously do not constitute evidence that God might approve of murder or abuse. So they do not defeat her justification for believing that it is likely that God would prohibit murder and abuse. Maitzen acknowledges a further possible reply to his argument: namely, that a sceptical theist might insist that, even if we do not have representative knowledge of the realm of value, we nevertheless know enough (independently of our beliefs about divine commands, I take it) to discern at least a few divine commands. This is an attack on premise 4.1 Obviously I have chosen to address that premise in a different way; but it is worth noting that even this objection is stronger than Maitzen gives it credit for being. In reply to the objection, Maitzen says: [a]ccording to [sceptical theism], we lack what it takes even to estimate the likelihood that some compensating good justifies a perfect being’s permitting [the instance of suffering under consideration in his paper]. . . . By the same token, we can’t estimate the likelihood that some reason lying beyond our ken turns what seems to us a diabolical command into just the thing a perfect being would tell someone to do in the particular circumstances. (2013: 452–3)

Strictly speaking, the first quoted sentence here is false: sceptical theists do not deny (for example) that one who believes that God exists and would not permit gratuitous evil can reasonably infer from these beliefs that there is no gratuitous evil. But set this aside for now. Still, it is hard to see why one should think that the truth of the second sentence follows. Again, suppose that Lucy considers the question of whether God might command the sort of abuse that she suffered as a child. It is part of her sceptical theism to say that, for all she knows, some good beyond her ken justifies God in permitting that suffering. So, she might reason, perhaps there is some good beyond her ken that would justify God in commanding the abuse that led to her suffering as well. Still, this is not evidence that God did

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

134

   :  

command the abuse she underwent; it is not evidence that God would ever command that sort of abuse; it is not evidence that God even might issue a general command that people engage in that sort of abuse. Is it at least evidence that God might, at some point, once command someone to engage in that sort of abuse? No; not even that. For even if there is some good that might justify God in commanding that sort of abuse, Lucy has no evidence whatsoever that the good in question is one that God might desire to bring about. (Cf. Section 3 above.) At best, then, Lucy has evidence only for what she already takes herself to know—namely, that God would, under some circumstances, permit the abuse that she suffered. So it seems that even if we grant the first part of Maitzen’s reply, the second (and crucial) part simply does not follow. The final two objections to be considered in this section are O’Connor’s claims that (a) sceptical theism implies that we cannot identify any actions as sins, and (b) sceptical theism implies scepticism about divine goodness and divine values. The first argument can be dismissed immediately, since its central premise is just the conclusion (4.4) which I have already shown to be unsupported. The second argument runs as follows: 4.5 ‘My believing that somebody is a morally good person requires my believing him to be good according to the standards of moral goodness that I myself accept’ (O’Connor 2013: 473). 4.6 ‘The basic tenet of [sceptical theism], namely, that we have no good reason to think that either our concepts or measures of goodness in human persons are representative of those applying to infinite, non-human persons’ implies that the sceptical theist ‘has no good reason to suppose she understands what she believes about divine goodness’ (O’Connor 2013: 474) and ‘has no good reason to think that she would recognize divine goodness as goodness’ (2013: 475). 4.7

Therefore, a sceptical theist cannot reasonably believe that God is good.

Let premise 4.5 pass. The main problem with the argument is that 4.6 is false (as we have already seen in Section 1). We might ask, however, what happens if we replace premise 4.6 with a premise that really does state a basic tenet of sceptical theism. Suppose, e.g. we replace 4.6 with 4.8: 4.8

Sceptical theism affirms both ST and ST1B–ST3B.

Now does the conclusion follow? It does only if the following premise is also true: 4.9 The conjunction of ST and ST1B–ST3B implies that we have no reason to believe that God counts as good according to standards of goodness that we ourselves accept.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

135

But 4.9 is false. Again, consider Lucy: Her standards of goodness are standards that she has drawn from scripture, moral intuition, and testimony from people she admires and respects. Her belief that God is good (according to standards that she accepts) is grounded in scripture, religious experience, testimony, and intuitions about the nature of perfection. Does her belief that her sample of possible goods, evils, and connections among them may not be representative somehow count against any of this? I cannot see that it does, and O’Connor has not argued that it does. What about her belief that human beings cannot be justified in saying of actual horrendous evils (like her own childhood suffering) that they are gratuitous? This belief actually seems to presuppose that God is good according to standards of goodness that we ourselves accept. Of course, she does believe that there may be something wrong with her standards—she is morally fallible. But in the course of defending premise 4.5, O’Connor concedes that we might reasonably believe of someone that she is morally good even if we do not think that she is good according to all of our moral standards. All that is required is that there be some significant overlap between our standards of goodness and the standards to which she conforms.

5. Scepticism about (Other) Knowledge of God O’Connor and Wilks both claim that sceptical theism generates problems for certain other beliefs we might have about God. O’Connor, for example, says that it undermines our reasons for thinking that our sufferings are not gratuitous, it undermines our ability to have a relationship with God, it deprives us of any basis for attributing purposes or reasons for action to God, and it undermines belief in miracles, as well as belief in crucial premises of certain natural theological arguments. The problem, however, is that all of these objections are predicated on the success of his argument for the conclusion that sceptical theists have defeaters for their beliefs about God’s goodness and about the contents of God’s commands. As we have already seen, however, those arguments are not successful; and so these conclusions likewise are left unsupported. In light of this, I shall focus attention in this section on Wilks’s argument. According to Wilks, sceptical theism undermines our ability to ‘[take] particular goods as signs of God’s presence’, and to ‘[see] things and events in the world around us as manifestations of God’ (2013: 460). Why should this be so? As I understand it, his argument runs as follows: 5.1 If sceptical theism is true, then ‘we should be skeptical about any claim to know what it would be evil for God to do’ (Wilks 2013: 458).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

136

   :  

5.2 ‘If we should be skeptical about any claim to know what it would be evil for God to do, then it seems we should also be skeptical about any claim to know what it would be good for God to do’ (2013: 459). 5.3 Therefore, if sceptical theism is true, we should be sceptical about any claim to know what it would be good for God to do. 5.4 Consequences of both goods and evils can run contrary to appearances: prima facie good events can have very bad consequences, and prima facie evil events can have very good consequences. 5.5 Therefore ‘we have equally little reason to suppose that we can discern their actual standing as good or evil on the basis of properties internal to them’ (2013: 460). 5.6 If we should be sceptical of any claim to know what it would be good for God to do, and if we have equally little reason to suppose that we can discern whether an event is good or evil on the basis of properties internal to it, then we have ‘no reason to think we see in a particular good a sign of God’s presence any more than we have reason to think we see in a particular evil a sign of his absence’¹⁵ (2013: 460). 5.7 Therefore, if sceptical theism is true, ‘we have no reason to think we see in a particular good a sign of God’s presence any more than we have reason to think we see in a particular evil a sign of his absence’ (2013: 460) The conclusion here is not quite the same as O’Connor’s conclusion that sceptical theism implies that we have no reason to think that we can see God’s purposes in particular events or claim any knowledge of God’s values. But obviously there are similarities. Like O’Connor, Wilks takes his conclusion to have untoward implications for sceptical theists’ acceptance of the argument from design. We saw in Section 1 of this chapter that premise 5.1 is false. So the argument to 5.3 is unsound. There is no independent reason to accept 5.3 either. So I reject that claim. Those, like me, who reject consequentialism will also reject the inference from 5.4 to 5.5. Wilks says that consequentialism ‘operates in the background’ of sceptical theism; so (presumably) he thinks that sceptical theists are bound to accept that inference. But, as we have already seen, that claim rests on a misconception. Finally, premise 5.6 is also false. Suppose you endorse for good reasons the view that it is part of the human design plan to have experiences as of the presence of God when we experience certain kinds of phenomena—a sublime vista, communion with fellow believers, etc. Suppose, furthermore, that you have some such experiences. You will then have good reason to think that you see in some particular phenomenon a sign of God’s presence; and if the phenomenon in ¹⁵ Wilks does not explicitly affirm 5.6; rather, the quoted material in this premise is part of his overall conclusion. I have supplied premise 5.6 simply in order to render the argument valid.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

137

question is, in fact, a good then (whether you recognize it as good or not) you will have good reason to think that you see in some particular good a sign of God’s presence. Moreover, you will have this reason even if you accept all of the premises of Wilks’s argument.

6. Moral Paralysis I turn finally to Maitzen’s moral paralysis objection. In defending this objection, Maitzen takes an argument by Michael Almeida and Graham Oppy (2003) as his springboard. In short, Almeida and Oppy argue that if you think (as the sceptical theist does) that there might be some great good that can justify God in permitting some evil, then you should also think that there might be some great good that can justify you in permitting the same evil. But if you think this, then you cannot sensibly think that you ought to intervene to prevent the evil in question. I have already replied to this argument in an article co-authored with Michael Bergmann (Bergmann and Rea 2005). Maitzen’s goal here is to rebut that reply, arguing that even if sceptical theists can somehow justify intervention to prevent some evil, they cannot sensibly take themselves to be obligated to prevent it. Before presenting Maitzen’s argument, I should note that even if the argument were sound, I am not sure that the conclusion would be all that damaging. It is well known that belief in a general obligation to prevent suffering yields highly counterintuitive conclusions. Peter Unger (1996), for example, reasons quite persuasively from the principle that U1 ‘If we can prevent something bad without sacrificing anything of comparable significance, it is wrong for us not to do so’ (Unger 1996: 8. Cf. Singer 1972). to the conclusion that U2 ‘On pain of living a life that’s seriously immoral, a typical well-off person, like you and me, must give away most of her financially valuable assets, and much of her income, directing the funds to lessen efficiently the serious suffering of others’ (Unger 1996: 134). But U2 seems to me (and most other people, I think) to be false. Thus, by rebutting Maitzen’s argument here, I do not intend to sign on to the presupposition that the denial of his conclusion is part of ‘commonsense’ morality. The core of Maitzen’s argument is the following analogy, which I quote at length: Imagine that a well-armed tribesman walks into a jungle field hospital and sees someone in strange garb (known to us as the surgeon) about to cut open the

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

138

   :  

abdomen of the tribesman’s wife, who lies motionless on a table. It certainly looks to the tribesman like a deadly assault, and thus he sees good reason to attack the surgeon and no particular reason not to. But suppose that the tribesman also believes that strangely garbed magicians (known to us as surgeons) travel around who miraculously save dying people by cutting them open. As a result, he occurrently believes “If this is one of those life-saving miracles, I shouldn’t expect to know it.” The incision is about to happen, and clearly there’s no time to investigate before acting. Given his beliefs, does it follow that he ought to attack the surgeon, i.e., that he would be wrong to refrain? I think not. At most what follows is that he may attack the surgeon, even at the cost of preventing his wife’s life-saving appendectomy: we get at most permission rather than obligation. Skeptical theism asks us to admit that we occupy the same position with respect to the realm of value that the tribesman occupies with respect to modern medicine: we shouldn’t expect to see how it works. Yet skeptical theists such as Bergmann and Rea claim to preserve, despite their skepticism, our ordinary moral obligation to intervene in such cases. . . . Their claim is correct only if the tribesman is obligated to attack the surgeon. (2013: 456)

So, in short, it is the tribesman’s ‘legitimate self-doubt’ (2013: 456) that mitigates any obligation he might have had to intervene; and Maitzen’s idea is that sceptical theists are likewise subject to such legitimate self-doubt when confronted with serious suffering or evil that they are in a position to prevent. Thus, like the tribesman, they have no obligation to intervene either. Moreover, Maitzen goes on to argue that the (typical theistic) belief that ‘someone exists who can make this suffering turn out for the best even if I don’t intervene’ mitigates whatever obligations we might have to intervene to prevent serious harm. For suppose we have such a belief. Then, says Maitzen, ‘[w]e ought . . . to feel less obligated (or less clearly obligated, if obligation doesn’t come in degrees) to prevent and relieve suffering than we would feel if we didn’t believe in such a potential guarantor of a good outcome’ (2013: 451). Here, then, it is not sceptical theism that causes the problem, but simply theism (in conjunction with an otherwise ordinary set of beliefs about morality). Maitzen notes that one reply that can be made by sceptical theists—and that has been made to Almeida and Oppy’s version of the moral paralysis argument—is that divine commands can ground our obligations to intervene to prevent serious suffering. But, says Maitzen, this reply is unworkable, partly because sceptical theism induces scepticism about the content and application conditions of putative divine commands (as seen earlier) but also because (a) there is no explicit command in scripture to intervene to prevent serious harm, and (b) if we try to derive the obligation from God’s other commands, we presuppose insight into God’s assumptions in issuing God’s commands—insight ‘skeptical theism says we have no right to think we possess’ (2013: 454).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

139

There are two points I wish to make by way of reply. First, suppose we pursue the line of reply that appeals to our knowledge of divine commands (or, more broadly, divine values). Maitzen is correct that we find in scripture no explicit command to intervene with preventive measures when we encounter someone who is about to cause grievous harm to another. But that does not mean that scripture is wholly silent on the matter. We are told to love our neighbours as ourselves; we are enjoined to help the poor, show compassion to those who are sick and suffering, visit people in prison, care for orphans and widows, and so on. The prophet Micah says, He has told you, O mortal, what is good; and what does the L require of you but to do justice, and to love kindness, and to walk humbly with your God? (Mic. 6:8, NRSV)

None of this can very well be done without intervening somehow in the lives of others; and it seems clear that those who refuse to intervene in the lives of others are doing something wrong. None of this entails (say) that we are obligated to prevent a kidnapping when we see that it is about to occur and are able to prevent it at little risk or cost to ourselves. But it seems clearly to constitute evidence in favour of the view that we have such an obligation. The question, then, is whether believing in such an obligation on the basis of evidence like this presupposes insight into God’s assumptions in issuing such commands that ‘skeptical theism says we have no right to think we possess’. If it does, Maitzen has not shown that it does. As we saw in Section 4, it simply does not follow from ST and ST1B–ST3B that we cannot sensibly assume that we understand at least some of the putative divine commands we find in scripture; nor does it follow that we cannot get some reasonably clear and action-guiding understanding of God’s values from scripture. Second, consider Maitzen’s claim, towards the end of the quoted passage above, that ‘[s]keptical theism asks us to admit that we occupy the same position with respect to the realm of value that the tribesman occupies with respect to modern medicine: we shouldn’t expect to see how it works.’ This claim is, of course, absolutely crucial to his argument-by-analogy. If the sceptical theist is not in the same position with respect to the realm of value that the tribesman occupies with respect to modern medicine, then the analogy is bad and the argument fails. But it should by now be clear that it is simply not true that sceptical theism asks us to admit that our position in relation to the realm of value resembles Maitzen’s tribesman’s position with respect to modern medicine. The tribesman has no understanding whatsoever of modern medicine. Indeed, as Maitzen presents him, he is worse off even than most elementary school children in the United States.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

140

   :  

For, after all, children do not usually regard successful surgeries as ‘life-saving miracles’, nor are they typically so ignorant and obtuse as to worry that those undergoing surgery are being assaulted by their physicians. So to say that the sceptical theist asks us to admit that our position in relation to the realm of value resembles the tribesman’s position with respect to modern medicine is to say that the sceptical theist asks us to admit that we are mostly clueless about the realm of value—we understand very little of it at all and, indeed, are inclined to reason rather stupidly about it. But as we have seen multiple times now in this chapter, that is not what the sceptical theist says. So, it seems to me, Maitzen’s argument by analogy fails. One might object, however, that I am still failing to address the basic point of Maitzen’s argument. The core worry is something like this: Sceptical theists maintain that we are in a position of some significant ignorance with respect to the realm of value. Moreover, sceptical theists typically also think that unknown goods justify God in permitting all manner of horrendous things to happen to his creatures.¹⁶ Given these two commitments, it seems (says the objector) that sceptical theists should also acknowledge that unknown goods, evils, and relations among goods and evils might justify us in permitting kidnappings and other serious evils. And if we do acknowledge this, then (the objection continues) it seems that we should be significantly less confident about our obligations to intervene than commonsense morality says that we should be. My own answer to this objection is that just as God’s obligations depend to some extent upon what God knows, so too our obligations depend to some extent upon what we know. True enough, there might be goods, evils, and relations among them such that if we knew about them, we would revise some of our beliefs about how or whether we ought to intervene in various kinds of circumstances. But there is no reason to think that awareness of this fact should make us any less confident in our views about what we ought (or are obligated) to do in light of what we actually know (cf. Bergmann and Rea 2005; Bergmann 2009, 2012). More exactly: I can see (and Maitzen has given) no reason for thinking this that does not somehow depend on controversial moral or epistemological principles that sceptical theists as such are free to reject (cf. Howard-Snyder 2009). Thus, absent further argument laying out the operative moral and epistemological principles and then somehow tying them to sceptical theism, the objection that sceptical theism undermines confidence in commonsense beliefs about our obligations to intervene is a failure.¹⁷ ¹⁶ Typically, but not necessarily, since one can be a sceptical theist while thinking either that some evils are gratuitous (and that this is not inconsistent with the existence of God) or that all evils are in fact justified by facts about known goods, known evils, and known relations among goods and evils. ¹⁷ For helpful comments on and conversations about an earlier draft, I am grateful to Michael Bergmann, Jeff Brower, Trent Dougherty, Daniel Howard-Snyder, Hud Hudson, Justin McBrayer, Jeff Snapper, and Meghan Sullivan. I am also grateful to the participants in the (Fall 2012) weekly

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

    ‘--’ 

141

References Almeida, Michael J. and Graham Oppy. 2003. ‘Sceptical Theism and Evidential Arguments from Evil’. Australasian Journal of Philosophy 81 (4): 496–516. Alston, William P. 1991. Perceiving God: The Epistemology of Religious Experience. Ithaca, NY: Cornell University Press. Bergmann, Michael. 2009. ‘Skeptical Theism and the Problem of Evil’. In Oxford Handbook of Philosophical Theology, edited by Thomas P. Flint and Michael C. Rea, 375–99. Oxford: Oxford University Press. Bergmann, Michael. 2012. ‘Commonsense Skeptical Theism’. In Reason, Metaphysics, and Mind : New Essays on the Philosophy of Alvin Plantinga, edited by Kelly James Clark and Michael Rea, 9–30. New York: Oxford University Press. Bergmann, Michael and Michael Rea. 2005. ‘In Defence of Sceptical Theism: A Reply to Almeida and Oppy’. Australasian Journal of Philosophy 83 (2): 241–51. Dougherty, Trent. 2008. ‘Epistemological Considerations Concerning Skeptical Theism’. Faith and Philosophy 25 (2): 172–6. Draper, Paul. 1996. ‘The Skeptical Theist’. In The Evidential Argument from Evil, edited by Daniel Howard-Snyder, 175–92. Bloomington: Indiana University Press. Gellman, Jerome. 1992. ‘A New Look at the Problem of Evil’. Faith and Philosophy 9: 210–16. Howard-Snyder, Daniel. 2009. ‘Epistemic Humility, Arguments from Evil, and Moral Skepticism’. Oxford Studies in Metaphysics 2: 16–57. McBrayer, Justin P. 2010. ‘Skeptical Theism’. Internet Encyclopedia of Philosophy. http://www.iep.utm.edu/skept-th/. McBrayer, Justin and Daniel Howard-Snyder (eds). 2013. The Blackwell Companion to the Problem of Evil. Malden, MA: Wiley-Blackwell. Maitzen, Stephen. 2013. ‘The Moral Skepticism Objection to Skeptical Theism’. In The Blackwell Companion to the Problem of Evil, edited by Justin McBrayer and Daniel Howard-Snyder, 444–57. Malden, MA: Wiley-Blackwell. Murphy, Mark. 2006. ‘Authority’. In Encyclopedia of Philosophy, edited by Donald M. Borchert, 412–18. 2nd edn. Detroit, MI: Macmillan Reference USA. O’Connor, David. 2013. ‘Theistic Objections to Skeptical Theism’. In The Blackwell Companion to the Problem of Evil, edited by Justin McBrayer and Daniel HowardSnyder, 468–81. Malden, MA: Wiley-Blackwell. Plantinga, Alvin. 2000. Warranted Christian Belief. New York: Oxford University Press.

discussion group of the Center for Philosophy of Religion at the University of Notre Dame—especially Dustin Crummett, Kate Finley, Samuel Newlands, Faith Pawl, Tim Pawl, Bradley Rettler, Lindsay Rettler, Joshua Rasmussen, Amy Seymour, and James Sterba. Finally, I am grateful to the John Templeton Foundation for funding that supported a research leave during which portions of this paper were written.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

142

   :  

Raz, Joseph. 1979. The Authority of Law: Essays on Law and Morality. Oxford: Clarendon Press. Russell, Bruce. 1996. ‘Defenseless’. In The Evidential Argument from Evil, edited by Daniel Howard-Snyder, 193–205. Bloomington: Indiana University Press. Singer, Peter. 1972. ‘Famine, Affluence, and Morality’. Philosophy & Public Affairs 1 (3): 229–43. Unger, Peter K. 1996. Living High and Letting Die: Our Illusion of Innocence. New York: Oxford University Press. Wilks, Ian. 2013. ‘The Global Skepticism Objection to Skeptical Theism’. In The Blackwell Companion to the Problem of Evil, edited by Justin McBrayer and Daniel Howard-Snyder, 458–67. Malden, MA: Wiley-Blackwell. Wykstra, Stephen J. 1984. ‘The Humean Obstacle to Evidential Arguments from Suffering: On Avoiding the Evils of “Appearance” ’. International Journal for Philosophy of Religion 16 (2): 73–93.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

7 Narrative, Liturgy, and the Hiddenness of God Next to the problem of evil, the most important objection to belief in God is the problem of divine hiddenness.¹ This latter problem starts from the supposition that God exists—after all, God is hidden only if God exists—and then reduces that supposition to absurdity by pointing out that if God exists, then the following mutually inconsistent claims are true: P1

God has allowed himself to remain hidden from many people.

P2 It would be bad for an omnipotent, omniscient God to remain hidden from anyone. P3

God, being perfectly good, cannot do anything that is bad.

In defence of the first premise, something like the following two (alleged) facts are standardly cited: I E For many people, the available a priori and empirical evidence in support of God’s existence is inconclusive: one can be fully aware of it and at the same time rationally believe that God does not exist. A  R E Many people—believers and unbelievers alike—have never had an experience that seems to them to be a direct experience or awareness of the love or presence of God; and those who do have such experiences have them rarely. Obviously not everyone will agree that these two claims are true. At the same time, many atheists and agnostics, and some theists, would insist that they are not only true, but understated. For purposes here I’ll grant that they are true and I won’t explore the question of whether stronger versions might be defensible as well.

¹ There is dispute in the literature over whether there is a problem of divine hiddenness that is genuinely separate from the problem of evil. Peter van Inwagen, for example, argues that there is, whereas Jon Kvanvig argues that there is not (cf. Kvanvig 2002; van Inwagen 2002). See also the last chapter of van Inwagen 2006.

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0008

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

144

   :  

Plausibly strengthened versions of these claims—e.g. that the evidence for most people or even for all people is inconclusive at best—wouldn’t affect the main argument of this chapter. But using them here would make it more difficult to motivate the problem, it would make it harder to get away with simply granting that they are true, and it would open me up to the charge of attacking an unsophisticated target. One might wonder whether the obtaining of   and     really implies that God is hidden. I’ll take up this question in the next section. But, obviously enough, nothing of great substance hangs on this matter either. The proponent of the hiddenness argument might simply replace talk about divine hiddenness with talk about the obtaining of   and    . The new argument would then proceed with much the same force as the original. Let us for the moment concede that   and     together imply that God is hidden. What now of the second premise? The basic problem with divine hiding is that it seems inconsistent with the following thesis: D C God strongly desires to promote the well-being of all of his rational creatures, both now and in the afterlife. The reason is straightforward. All of the major theistic religions agree that belief in God is vital for our present and future well-being. But a world in which God is hidden is one in which God is doing far less than he could (if he is omnipotent and omniscient) to promote rational theistic belief. Hence, it is one in which God is doing far less than he could to promote our well-being. Moreover, divine hiddenness is a source of suffering in believers, who often feel abandoned, neglected, unloved, or rejected by the being to whom they have devoted their lives and whom they have been taught to regard as their loving heavenly Father. Of course, it is conceivable that God have a good reason for remaining hidden. Suppose, for example, that divine hiddenness promotes some greater good for humanity that God rightly desires more than he desires the well-being of the individuals from whom he is hidden. Or suppose that, contrary to initial appearances, for every individual S from whom God is hidden, God’s hiddenness from S promotes S’s well-being. Then the apparent conflict with   disappears. So the problem of divine hiddenness remains only if we suppose that divine hiddenness does not promote any good the promotion of which would justify God in permitting whatever bad things come from divine hiddenness. In simpler terms, the problem remains only if we suppose that divine hiddenness does not promote any ‘God-justifying good’. Not surprisingly, once we set aside attempts to deny P1, the bulk of the remaining literature on divine hiddenness is aimed at identifying possible

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

145

God-justifying goods.² The presumption in this literature seems to be that the only viable candidates are human goods (e.g. freedom, the cultivation of certain kinds of virtue or prevention of certain kinds of vice, etc.).³ Though various other considerations might just as easily motivate the presumption, at least one powerful motivation is the fact that it is a logical consequence of the following general principle endorsed by various prominent writers on the problem of evil: B   S God is justified in allowing undeserved suffering to come to an individual X for the sake of greater goods only if among those greater goods are goods that benefit X.⁴ If     is true, then it looks like human goods have to be included among the goods (if any) that justify God in allowing   and     to obtain. In the present chapter, however, I want to defend a rather different line. In particular, I want to defend a response to the problem of divine hiddenness that is consistent with the following claim: N H G It is not the case that God permits   and     in order to secure human goods. N   is consistent with the claim that God permits   and     in order to secure greater goods. It is also consistent with the claim that the obtaining of   and     in fact promote human goods (that may or may not count as ‘greater’, God-justifying goods). What it rules out is

² For a variety of such responses, see the essays in Howard-Snyder and Moser 2002—in particular Garcia 2002, Moser 2002, Murray 2002, and Wainwright 2002. See also Moser 2008. Note, too, that one might posit God-justifying goods without attempting to identify them. In the works of his just cited, Paul Moseridentifies various God-justifying goods that might result from divine hiddenness, but he also denies that the goods he identifies explain every case in which God is hidden from someone. Nevertheless, his view seems to be that every case in which God is hidden someone is justified by some good that God aims to promote ³ I assume that the prevention of evils equally bad or worse is a good. So, in other words, if it turns out that God permits     and   to obtain simply because preventing them would result in human evils as bad as or worse than the suffering they cause, then, by my lights, the point of divine hiddenness is still to promote human goods. This is a terminological point, not a substantive one. ⁴ See, e.g. Stump 1985, 1990. For critical discussion of this principle, along with references to other philosophers who endorse it, see Jordan 2004. I think that a suitably qualified version of     is plausible when the suggestion in view is that God might be allowing someone to suffer for the sake of human (or other creaturely) goods; but the unqualified principle here seems to me to have counterexamples (along the lines discussed in Section 2 below).

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

146

   :  

the idea that whatever human goods may be promoted by divine hiddenness are the goods for the sake of which God remains hidden. Drawing in part on recent work by Eleonore Stump and Sarah Coakley, I shall argue that even if    is true, divine hiddenness does not cast doubt on  . My argument will turn on three central claims: (a) that     and   are better thought of as constituting divine silence rather than divine hiddenness; (b) that even if    is true, divine silence is compatible with   so long as God has provided a way for rational creatures to find him and to experience his presence despite the silence; and (c) that there is some reason to think that biblical narratives and liturgical acts are vehicles by which we might find and experience the presence of God. Each of these claims will be defended, in turn, in the three sections that follow.

1. Divine Hiddenness and Divine Silence Consider the following biconditional: 1 God is hidden $ God permits   and     to obtain. I don’t think that H1 has been explicitly affirmed by anyone writing on the problem of divine hiddenness. But it seems to be largely taken for granted in the literature.⁵ My goal in this section is to provide reasons for rejecting H1. In doing so, I don’t take myself necessarily to be arguing against widely held views about the nature of hiddenness. Rather, I take myself simply to be providing reasons to reject a widely accepted terminological convention. Still, I think that the argument of this section matters; for our terminology colours the way in which we think about the problem at hand. If God hides from his creatures, then it seems he ought to have a good reason for doing so—a reason that somehow involves their good and not just his own personal preferences. But if God is merely silent, then it is not at all clear that he needs to have any human-oriented reason for doing so. At any rate, so I will argue later on. The problem with H1 is just this: It is equivalent to the conjunction of two conditionals, one true and the other false. The true conditional is the left-toright one: ⁵ At any rate, this is so on the assumption that God’s providing strong evidence of his existence would imply the falsity of either     or  . See, especially, Schellenberg 1993: 4–6, 83–4. Many writers follow Schellenberg’s lead in formulating the problem of divine hiddenness, and so are plausibly taken as endorsing roughly the same views about what divine hiddenness consists in.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

147

H1a God is hidden ! God permits   and     to obtain. If   were false, then there would be empirical or a priori evidence for the existence of God in light of which atheism would be irrational. In that event, we would want to say that God is not hidden, but manifest in the empirical or a priori evidence. Likewise, if     were false, then God would be manifest to most people by way of private, subjective experience. Perhaps God would be hidden from some; but it wouldn’t make sense to say that God is hidden simpliciter. Thus, God is hidden only if God permits   and     to obtain. But now consider the other conditional involved in H1: H1b God permits   and     to obtain ! God is hidden. This one is open to a rather obvious objection. Suppose there’s an object—a car, perhaps—that is in plain sight in Wilma’s driveway but which Wilma can’t see because her eyes are closed. The car isn’t hidden from her; she’s just not looking. Indeed, even if someone had put the car in her driveway knowing that she wouldn’t be looking, we wouldn’t want to say that the person had hidden the car from her. Now suppose there’s something analogous to ‘opening our eyes’ that we all can do that would allow us to receive experiences or other evidence of the presence of God. And suppose that, as it happens, most of us haven’t opened our eyes in this analogous sense.⁶ In that event,   and     might well be true, and true by divine permission; but there seems to be no reason (yet) to say that God is hidden. Admittedly, more information might change the verdict. Suppose Fred knows that the one place in the world that Wilma can’t bear to look is her own driveway—she always closes her eyes when she approaches it. Suppose further that Fred has put the car there in a deliberate attempt to conceal it from her. In that case, the car is hidden—hidden in plain sight, as it were. What matters here is the intention to conceal. Likewise, if we think that     is due to some deliberate intention on God’s part to conceal himself, then (given  ) we have reason to say that God is hidden. Or maybe there is some other bit of information that we could acquire that would lead us to say that God is hidden. My point here is simply that the obtaining of

⁶ The idea that something like ‘failure to open our eyes’ explains why God’s existence is not more obvious is present in many writers. (See, for example, Wainwright 2002 and Moser 2008, esp. ch. 2.) But I think that if this sort of story is really correct, then it is incorrect to say that God is hidden unless the story is fleshed out along the lines suggested in the next paragraph.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

148

   :  

  and     by divine permission is not enough.⁷ The term I would prefer to use in characterizing what we seem to know about God’s self-disclosure to the bulk of humanity is, therefore, not hiddenness but rather silence. To say that something is hidden implies either that it has been deliberately concealed or that it has been concealed (deliberately or not) to such a degree that those from whom it is hidden can’t reasonably be expected to find it.⁸ This is why divine hiddenness would seem to require justification. If God cares about our well-being, one would think that, absent special reasons for doing otherwise, he would put us in circumstances such that we could reasonably be expected eventually to find him. But   and     don’t imply that God is deliberately concealing his existence from us; nor do they imply, on their own, that we can’t reasonably be expected eventually to find him. What they do imply is that God hasn’t made a special effort to ensure that most of his rational creatures detect (as such) whatever signs of his existence there might be or whatever messages he might be sending us. I don’t mean to suggest that talk of divine hiding is categorically inappropriate. Indeed, various biblical writers talk explicitly of God hiding his face and concealing his presence, and Isaiah exclaims outright, ‘Surely you are a God who hides himself!’⁹ But in all of these cases the suggestion is not that God has so obscured his presence that, for the most part, people cannot reasonably be expected eventually to find him. Rather, the suggestion is simply that God has made his presence less obvious, so that seeking is required in order to find him. In other words, the suggestion is that God is, at most, partially hidden, not that he is hidden simpliciter. I also don’t mean to suggest that   and     imply that God is totally silent. Both are compatible with God’s having delivered special revelations to select people—for example, the authors of scripture, the prophets, witnesses to the miracles of Jesus, and so on. ⁷ I assume here that divine omniscience and omnipotence do not obliterate the distinction between what God deliberately brings about and what God merely permits. The assumption isn’t problem-free; but it is standard nonetheless. Dispensing with it would push us towards the view that every event is a divine act, and so it would require substantial reframing of the present discussion in terms of whatever distinctions we would then use to separate purely divine acts from those divine acts that also count as natural events and acts of non-divine creatures. ⁸ Note, too, that hiddenness comes in degrees. Word search puzzles and Easter egg hunts, for example, are typically constructed so that the hidden items can reasonably be expected to be found after a certain amount and kind of effort. In cases like this, I think that we’d want to say not that these items are entirely hidden, but that they are merely partially hidden. Items that could be found only by sheer luck or extraordinary (unexpected) skill are the ones that count as completely hidden. ⁹ Isa. 45:15. See also, e.g. Ps. 10:1, where the psalmist asks, ‘Why, O LORD do you stand far off? Why do you hide yourself in times of trouble?’ and Job 13:24, where Job asks, ‘Why do you hide your face, and count me as your enemy?’. Translations are from the New Revised Standard Version, Copyright © 1989, Division of Christian Education of the National Council of the Churches of Christ in the United States of America.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

149

Nor do I mean to suggest that God isn’t, even now, sending ‘messages’ to all of his rational creatures that could be accessed if only we would do something analogous to ‘opening our eyes’. The point, again, is just that God is evidently not making any special effort to ensure that most of us receive communicative content from him. A man who chooses to whisper rather than shout instructions to his children, knowing all the while that they cannot (yet) hear him over the racket they are making, is being silent towards his children in the sense that I have in mind. Given   and    , it seems that we can safely conclude that, in just the same sense, God (if he exists) is being silent towards most of us. Henceforth, when I speak of divine silence I will be speaking simply of the fact that   and     both obtain. As I understand it, then, divine silence is compatible with God’s having provided some widely and readily accessible way for his creatures to find him and to experience his presence, albeit indirectly, despite his silence.

2. Divine Silence and Divine Concern The question, now, is whether divine silence casts any doubt upon  , given   . In other words, assuming divine silence doesn’t contribute to our well-being or to any greater human good, does the fact of divine silence give us any reason to doubt that God cares about us? It is easy to see why one might think that it does. Consider, for example, this excerpt from the private writings of Mother Teresa: Lord, my God, who am I that You should forsake me? The child of your love— and now become as the most hated one—the one You have thrown away as unwanted—unloved. I call, I cling, I want—and there is no One to answer—no One on Whom I can cling—no, No One.—Alone. The darkness is so dark . . . The loneliness of the heart that wants love is unbearable.—Where is my faith?—even deep down, right in, there is nothing but emptiness & darkness.—My God—how painful is this unknown pain. It pains without ceasing . . . . I am told God loves me—and yet the reality of darkness & coldness & emptiness is so great that nothing touches my soul . . . . The whole time smiling—Sisters & people pass such remarks.—They think my faith, trust & love are filling my very being & that the intimacy with God and union to His will must be absorbing my heart.—Could they but know—and how my cheerfulness is the cloak by which I cover the emptiness & misery.—What are You doing My God to one so small?¹⁰

¹⁰ Teresa 2007: 187.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

150

   :  

Mother Teresa devoted the bulk of her life to the service of God and, if we can believe her own assessments of her own inner life, loved and longed for union with God throughout her life in a way that was deeper, more fervent, and more authentic than most of the rest of us even aspire towards, much less attain. Yet here, towards the end of her life, she is driven to absolute despair and misery by persistent silence on the part of God. And this experience of hers is hardly unique: many religious believers experience, to varying degrees, precisely the same sort of misery and despair, the same feelings of abandonment and isolation, and the same heart-rending cosmic loneliness in the face of what they take to be unrelenting silence on the part of the God they serve and worship. What loving father would treat his children so? How could a compassionate God refrain from answering the cries of Mother Teresa and others like her? Wouldn’t any human parent—flawed and selfish as we are—find it irresistible to draw near to his child and whisper words of comfort and affection? Shouldn’t we expect at least as much from a being whose love is supposed to be perfect? The objection implicit in the rhetorical questions is altogether natural; but it is flawed. It is flawed in just the same way in which complaints about the behaviour of human persons are often flawed: it depends on a particular interpretation of behaviour that can in fact be interpreted in any of a number of different ways, depending upon what assumptions we make about the person’s beliefs, desires, motives, dispositions, and overall personality. A senior member of your department doesn’t greet you in the hallway. Is he offended by you? Does he think you’re beneath him? Is he depressed and having a bad day? Or is that just him, a little pre-occupied and not really noticing his surroundings? You’re on a day trip with a colleague from another country. You try a few times to strike up conversation, but it never takes off and shortly you find that over an hour has passed in almost total silence. Is your colleague disrespecting you? Is she playing a power game, trying to force you to carry the conversation or some such thing? Does she find you boring, or intimidating? Or is the silence an indication of nothing more or less than the fact that she is somewhat introverted and doesn’t happen to have a whole lot to say (to anyone) at the moment? Answering questions like this with any reliability requires substantial information about what sort of person one is dealing with—about the person’s cultural background, about what sorts of social norms he or she is likely to recognize and respect, about his or her views about what various kinds of behaviour (both verbal and not) communicate to others, about his or her general ‘style’ of interacting with other people, and so on. But if this is what it takes to interpret the behaviour of an ordinary human person, imagine how difficult it must be to interpret the behaviour of an invisible and transcendent divine person. Seen in this light, the suggestion that divine silence in and of itself somehow indicates disinterest or lack of love and concern on God’s part is absurd. Even granting the complete reliability and transparency of biblical testimony about

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

151

God, we have precious little by way of clear and reliable information about God’s personality and about his general ‘style’ of interacting with others; and to ask about God’s ‘culture’ or about what sorts of social norms God would likely recognize and respect seems to border on the overly anthropomorphic. God is as alien and ‘wholly other’ from us as it is possible for another person to be. Thus, it is hard to see how we could say with any confidence at all what his silence indicates. Indeed, even to suppose that divine silence is unlikely given   seems to me to involve quite a lot of unwarranted assumptions about the degree to which divine modes of interaction would likely resemble twentyfirst-century human modes of interaction. Granted, divine silence would indicate a lack of concern for rational creatures if we had good reason to think that God had provided no way for us to find him or to experience his presence in the midst of his silence. This would indicate a lack of concern because it would indicate that God is trying to prevent us from finding him, or at least doing nothing to help, and thus bringing about something that is both intrinsically very bad for us and totally beyond our control. In the next section, however, I’ll argue that, in fact, there is reason to think that God has provided ways for us to find him and to experience his presence in the midst of his silence. The point of this section thus far, then, might be summed up this way: Silence is an interpretable kind of behaviour; and, as with any other person, God’s behaviour doesn’t wear its interpretation on its sleeve—it can be understood only in the light of substantial background information. To be sure, divine silence could be an indication of divine rejection or lack of concern. But that interpretation is entirely optional, given our evidence. Divine silence might instead simply be a reflection of the fact that God prefers to communicate with us and to draw us into his presence in ways other than ones that would render either   or     false. It might just be a reflection of God’s personality, so to speak. The pressing question, of course, is what to do with the fact that God’s silence is painful for us. Many believers experience crippling doubt, overwhelming sadness, and ultimate loss of faith as a result of ongoing silence from their heavenly Father. On the assumption that God exists and that a loving relationship with God is a great good, it would appear that many people have been positively damaged by divine silence. Isn’t it just this that leads us to take divine silence as evidence of God’s lack of concern? Perhaps silence is just an outgrowth of God’s personality; but then, the objector might say, God’s personality is just that of a distant, unconcerned ruler rather than that of a loving and attentive parent. The problem with this objection is that it completely ignores the fact that sometimes our being pained by another person’s behaviour is our problem rather than theirs—due to our own dysfunctional attitudes and ways of relating to others, our own epistemic or moral vices, our own immaturity, and the like. In such cases,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

152

   :  

it is our responsibility to find a way out of our suffering rather than the other person’s responsibility to stop behaving in the ways that cause us pain.¹¹ As it happens, in cases like this we can benefit if the other person exercises her right to persist in her behaviour: we can grow in maturity or in our ability to relate to others, for example. But if we are indeed unreasonably pained by her behaviour, then her right to persist in her behaviour is independent of this potential benefit: she can persist regardless of whether she believes that the benefit is at all likely to be realized (indeed, even if she somehow knows that it will not be). To suppose otherwise—to suppose that she would have to stop if there weren’t some benefit to persisting that somehow outweighed or defeated our suffering—is, I think, just to deny that it is possible for one person to be unreasonably pained by the behaviour of another. Those who do deny this cannot make use of the response I am offering to the problem of divine hiddenness; but, to my mind, denying it is wholly implausible. Let me be clear here about terminology. As I am using the term, unreasonable suffering refers to suffering that is the result of vice, immaturity, or some other disposition on the part of the sufferer for which the sufferer is morally responsible, and which consequently places no general demand upon the person causing the suffering to stop her problematic behaviour. So, in cases where X is suffering unreasonably in response to the behaviour of Y, it will be wholly appropriate for Y to persist even if no greater goods come to X or to other people as a result of her doing so. Compassion for X might require that Y do something to help X get past her suffering. But it will not require that X desist from her behaviour. The suffering of others should always move us, of course; but it is simply not true that compassion, love, or any other virtue requires us to submit unboundedly to manipulation by the unreasonable responses of others. Moreover, if the problematic behaviour is silence, compassion will not even require that Y explain her behaviour or act in other ways that make it clear to X that she is trying to show compassion; for doing either of these things is incompatible with maintaining silence. Thus, if it is possible for one person to suffer unreasonably in response to the silence of another, then it is possible that there be persons X and Y such that X is pained by Y’s silence, but it is wholly appropriate for Y to remain silent even though her silence promotes no greater good for X or for anyone else, and even if it appears (falsely) to X and to others as if she is making no effort at showing ¹¹ Note that the view being articulated here is not quite the same as the view that ‘divine hiddenness is due to human blindness’. According to the latter view,   is false, and the relevant evidence for God’s existence could be seen if only we weren’t blinded to it by our own sin. (For a full presentation of this view, see Wainwright 2002.) On the view that I am presenting, however,   is taken for granted, our own faulty interpretation of divine silence is what is to blame for our finding it objectionable, and sin is only one among several possible explanations for why we interpret divine silence in the way that we do. (You might not be culpable for your own present immaturity, for example, even if you have a responsibility to overcome it.)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

153

compassion towards X. Indeed, not only might it be wholly appropriate for Y to remain silent under these conditions, but her doing so might be consistent with the supposition that she has perfect love for X. For to suppose otherwise is to suppose that her remaining silent is a failure of love on her part, which implies that X’s suffering, no matter what the ultimate explanation for it might be, places an automatic demand upon Y to desist from her silence, which, in turn, implies that X’s suffering is not unreasonable after all. Obviously the suggestion here is that believers who are pained by divine silence might be in just this sort of situation with respect to God. That is, it might be that our suffering in the face of divine silence is unreasonable, due more to our own immaturity or dysfunction than to any lack of kindness on God’s part. Perhaps it is largely just a result of our own untrusting, uncharitable interpretations of divine silence, or from an inappropriate refusal to accept God for who God is and to accept God’s preferences about when and in what ways to communicate with us. Perhaps there are ways of experiencing the world that are fully available to us (either as we are now or as we would be if only we would strive for maturity in the ways that we ought to) and that would allow us to be content with or even to appreciate the silence of God in the midst of our joys and sufferings. It helps, in this vein, to be reminded of a fact about God and a fact about ordinary human relationships. The fact about God is that the most enigmatic, eccentric, and complicated people we might ever encounter in literature or in real life are, by comparison with God, utterly familiar and mundane. The fact about human relationships is that experiencing the silence of another person can, in the right context and seen in the right way, be an incredibly rich way of experiencing the person—all the more so with a person who is sufficiently beyond you in intellect, wisdom, and virtue. A wise and virtuous person who is utterly beyond you intellectually and silently leads you on a journey might teach you a lot more about herself and about other things on your journey than she would if she tried to tell you all of the things that she wants to teach you. In such a case, objecting to the silence, interpreting it as an offence, or wishing that the person would just talk to you rather than make you figure things out for yourself might just be childish—an immature refusal to tolerate legitimate differences among persons and to be charitable in the way that you interpret another’s behaviour. And there is no reason to think that the person would owe it to you to cater to these objections— even if her decision to be silent was arrived at not for the sake of your greater good, but simply because that’s who she is, and that’s how she prefers to communicate with people like you.¹² ¹² Admittedly, leading on a journey is a communicative act; so the person in this example is not totally silent. Is this a relevant disanalogy? I don’t think so. Recall, for one thing, that, though I have granted     and  , I have not conceded that God is totally silent. But, more importantly, the illustration could be modified to get rid of the disanalogy. Drop the supposition that the person is leading you on a journey, and suppose instead

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

154

   :  

I imagine that some might be tempted to caricature the view I have been developing as follows: S F God, on this view, is like a man who neglects his children, leaving them bereft and unloved while he sits in stony silence thinking ‘I just gotta be me’. The suffering that results from this sort of behaviour is intense, and (if it truly promotes no human good) entirely gratuitous; for, just as it would cost a man nothing to break his silence momentarily to whisper words of love and encouragement to his children, likewise it would cost God nothing to deviate from his own preferred mode of interaction to communicate divine love more widely and fully to all of creation. But there are at least two problems with the caricature that, I think, undermine its force as an objection against my view. First, the Silent Father analogy is apt only if God’s behaviour towards the world is more like the God of deism than the God of classical theism. The deistic God is highly non-interactive; the God of classical theism, on the other hand, is intimately involved with creation, active in revealing himself (albeit not as widely or clearly as many of us would prefer), and—according to the Christian story— concerned enough about creation to become incarnate, suffer, and die on a cross for the sake of humankind. Of course, the objector will complain that we do not have enough evidence even about God’s existence, much less his behaviour, to be sure that these claims on the part of theists are true. But the relevant point here is that neither do we have enough evidence to be sure that they are false. There are, in other words, no grounds for saying with any confidence that if God exists, he is neglectful of creation in the way that the Silent Father is neglectful of his children. Second, the caricature fails to take seriously the possibility that God might have a genuine, robust personality, and that it might be deeply good for God to live out his own personality. One odd feature of much contemporary philosophy of religion is that it seems to portray God as having a ‘personality’ that is almost entirely empty, allowing his behaviour to be almost exhaustively determined by facts about how it would be best for others for an omnipotent being to behave. But why should we grant this portrayal, or anything like it? God is supposed to be a person not only of unsurpassable love and goodness, but of unsurpassable

that you are touring a museum with the following background beliefs: (a) the curator has arranged things in the museum with you in mind, and wants very much to show you things about yourself, about life, and about his personality via the arrangement; and (b) the curator, though in fact silent and invisible, is nevertheless watching as you tour, can be talked to, can talk with you at any time, and can alter things in the museum in response to your communications. (I have heard roughly this sort of analogy in a variety of sermons over the years. If it has an original printed source, I am unaware of it.)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

155

beauty.¹³ And it is not at all clear that God could be that sort of person if the portrayal of God as (effectively) a cosmic, others-oriented utility-maximizing machine were correct. For it is hard to see how a person could manage to be unsurpassably beautiful, or even very beautiful at all, without having a highly complex personality and motivational structure. But if we grant that God has a highly complex personality and motivational structure, and if we also grant that it would be deeply good for God to live out God’s personality, then even straightforward utilitarian calculations might rule in favour of God’s persisting in silence despite the suffering that it causes (provided God is behaving compassionately towards his creatures in other ways, and so on). Or the goods and evils here might be incommensurable: it might be good for God to be who God is, and bad for us to suffer; and there might be no metric for comparing the good with the bad. Either way, if it is deeply good for God to live out the divine personality, and if our suffering is furthermore unreasonable, the result of immaturity or other dysfunctions that we can and should overcome anyway, then I see no reason why even perfect love would require God to desist from his preferred mode of interaction in order to alleviate our suffering. On the view that I am developing, then, it is not true that the suffering produced by divine silence is gratuitous; for it is not true that divine silence serves no good whatsoever. If, as I am suggesting, divine silence is an outgrowth of the divine personality or of God’s preferences about how to interact with creatures like us, then divine silence is plausibly thought of as good in and of itself, or good as a means to the expression of the perfectly good and beautiful divine personality. Moreover, if, as I have suggested, there are ways of experiencing divine silence that we would find non-burdensome or even beautiful, and if God’s persisting in his silence provides opportunities for us to grow in maturity or in our ability to relate to others, then divine silence might even be good for us. My point is simply that it need not be good for us in order for God to be justified in persisting in silence; for divine silence might promote goods without promoting human goods. The question now is whether it is really plausible to suppose that our suffering in response to divine silence is the result of something like immaturity or dysfunction that we ought to overcome on our own. Can we really say such a thing about Mother Teresa, for example? Initially we might baulk at saying such a thing. Mother Teresa was widely revered in her lifetime as a woman of great wisdom and virtue.¹⁴ Isn’t it implausible at best (perhaps even scandalously offensive or worse) to suggest that her relationship with God was marred by dysfunction and immaturity? Moreover, even if one can say such things about Mother Teresa, isn’t it simply incredible to suppose that everyone who suffers in

¹³ Or a tri-person, depending on what brand of theism (or even what brand of trinitarian theism) you subscribe to. But let us ignore this complication for now. ¹⁴ Though not universally so. For a radically different perspective, see Hitchens 1995.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

156

   :  

response to divine silence does so as a result of some sort of dysfunction and immaturity? Here we must remember our dialectical context. We are assuming that if God exists, then God is perfectly good; we are also assuming that if God exists, divine silence is not permitted for the sake of greater goods or for the prevention of comparable or worse evils. So our only alternatives here are to say (a) that something like the story I have just offered is true, (b) that God does not exist, or (c) that some alternative story consistent with the assumptions laid out in the introduction of this chapter is true. Option (a) posits widespread cognitive or emotional dysfunction, even in the likes of Mother Teresa; but, notably, so does option (b). If God does not exist, then the religious experiences that led Mother Teresa to found the Missionaries of Charity and to devote herself to work in the slums of Calcutta were non-veridical, probably hallucinatory;¹⁵ the cognitive faculties responsible for the maintenance of her religious belief are (probably) unreliable;¹⁶ and her ongoing cries to God to break his silence are ultimately just one-sided conversations that express a longing for her childhood hallucinatory experiences to return and both validate and reinforce her ongoing devotion to a being that does not (and probably cannot) exist.¹⁷ More or less the same would hold true for the rest of us who suffer in response to ‘divine silence’ as well. In short, religious believers who suffer in response to divine silence suffer from serious cognitive dysfunction on both of the main options that this chapter has in view; so the fact that (a) posits serious cognitive dysfunction in the likes of Mother Teresa can’t be a reason for preferring (b) to (a). It could, I suppose, be a reason for preferring (c) to (a). But that reason is defeated, I think, by the fact that no viable alternative (consistent with all of the assumptions laid out in the introductory section) is yet available. And, in any case, option (c) isn’t something that an atheist can accept, so it’s not something that helps those who want to use the problem of divine hiddenness as an argument against the existence of God. Note, however, that everything I’ve said so far about why divine silence need not conflict with   has been said under the supposition that the following claim is true: D S-D God has provided some widely and readily accessible way of finding him and experiencing his presence despite his silence. ¹⁵ Teresa 2007, esp. p. 3, and chs 3–6. ¹⁶ Different religious epistemologies will yield different results on this score. Someone who thinks that Christian belief is produced and maintained by ordinary evidential reasoning, and that we might (though probably wouldn’t) have the same evidence we have even if Christian belief is false, wouldn’t necessarily be committed to thinking that false Christian belief would have to be produced by unreliable mechanisms. Someone like Alvin Plantinga, however, who thinks that Christian belief is likely produced and maintained by the operation of something like John Calvin’s sensus divinitatis—a faculty whose proper function is to detect the presence of God—would be so committed. (Cf. Plantinga 2000: 170–86.) ¹⁷ If, as many theists think, God exists necessarily if at all, then if God does not exist, God cannot exist.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

157

If that supposition is false—if there is no divine self-disclosure whatsoever—then most of what I have said in this section falls flat. For in that case, most of us have been entirely cut off from God’s presence; and if we have been entirely cut off from God’s presence, then I can’t very well argue that divine silence might possibly be interpreted as a result of God’s desire to communicate with or be present to us in other ways. If we have been entirely cut off from God’s presence, then God has done or permitted something that is both devastatingly harmful to us and totally out of our control, and it is much harder to make plausible the suggestion that God has taken reasonable steps to be compassionate towards us in the midst of our suffering. Thus, there is a more serious, perhaps even intractable problem reconciling God’s behaviour with    and  . Thus, in the next section I will argue that there is good reason to think that   obtains despite the fact that God is silent. This will complete my defence of the claim that divine silence is compatible with the conjunction of   and   .

3. Narrative, Liturgy, and the Presence of God My goal in this section is to defend  -. The literature on divine hiddenness seems largely to take it for granted that   is true only if either   or     is false. Presumably the idea is that the presence of God is widely and readily accessible only if either there is conclusive empirical or a priori evidence of the existence of God, or many people are having subjective experiences that at least seem to be direct experiences of the presence of God. What this supposition ignores, however, is the possibility of mediated experiences of the presence of God through media that are themselves widely and readily accessible. In order to make a case for the conclusion that this possibility ought to be taken seriously, I’ll first have to explain what I mean by ‘mediated experiences of the presence of God’; then I’ll have to provide plausible candidates for media through which those experiences might be had. The difference between mediated experiences of an object or event and direct experiences of the same thing is like the difference between Parfit-style quasimemory of a thing and direct perception of it. As Parfit characterizes them, quasimemories, or ‘q-memories’ are apparent memories that are genuinely about someone’s experiences, though not necessarily about experiences of the person having the q-memory.¹⁸ So, for example, if I step into a duplication machine, my duplicate will have q-memories of my life, but not genuine memories (since they will ‘feel like memory’ for him, without being genuine memories of his

¹⁸ Parfit 1971: 15.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

158

   :  

experiences). The crucial thing to note here is that q-memory provides a person with roughly the same sort of information about a thing as direct experience of it would provide, and it can do so even if the person has never in fact had a direct experience of the thing. In q-memory, you experience to an attenuated degree what it is like to be in the presence of the event or object that is q-remembered without ever having to experience its presence directly. There is a well-known and controversial thought experiment in the philosophy of mind, the upshot of which is supposed to be that first-hand experiences of sensible properties (like redness) convey information that could not possibly be conveyed by even the most complete physical descriptions of the property, things having the property, or what is involved in experiencing the property. In short, the thought experiment invites us to imagine a woman (Mary) who has never experienced redness but who has come to be in possession of a maximally complete physical description of everything pertaining to the phenomenon of redness and the experience thereof. We are then invited to consider the question whether she would learn anything upon coming to experience redness for the first time. At least initially, it is hard to resist the intuition that she would, and that the information she acquires is just information about what it is like to experience redness.¹⁹ Suppose this is right. (For present purposes I’ll ignore the fact that this conclusion is controversial.²⁰) The thing to notice is that even if it is true that the information Mary acquires couldn’t have been gotten by way of physical descriptions, it seems clear that it could have been acquired by means other than being in the presence of something red. Implanted q-memories of redness would do it, as would Matrix-style virtual experiences of redness. The case of qmemory is the one that interests me most, however, because there is no question that q-remembered experiences of redness would be and feel different from direct experiences of redness. (This isn’t so clear in the case of Matrix-style virtual experiences.) Importantly, in q-memory of redness we get some even if not all of the very same non-propositional information about redness that we get through direct experience thereof. It is in precisely this sense that q-memory is a way of having mediated experiences of redness. I turn now to the question of what sorts of things might plausibly be thought to mediate experiences of the presence of God. In her recent work on the problem evil, Eleonore Stump has pursued the project of theodicy by way of extended literary-critical treatment of a variety of biblical narratives.²¹ Rather than treating, ¹⁹ The example is due to Frank Jackson 1982. For discussion, see the essays in Ludlow et al. 2004. ²⁰ The controversy doesn’t matter much for present purposes. What ultimately matters is the much less controversial idea that there is a sui generis quality to experiencing a thing that is lost in mere descriptions of it but that can be mediated by something like q-memory. ²¹ See, especially, Stump 2009, 2012 [Editorial note: Stump 2012 was in draft form in 2010 when the present chapter was first published].

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

159

say, the story of Job or the story of Abraham and Isaac as brief toy examples that illustrate particular principles that then go on to get discussed in the analytic mode, she presents these and other stories in detail, tries as much as possible to help us understand the motives, desires, and experiences of the characters involved, and aims in so doing to show us that we can learn useful things from the narratives that simply cannot be expressed propositionally. What matters for theodicy is the philosophical usefulness of what we learn from the biblical narratives; but what matters for present purposes is more their subject matter. I’ll begin with some terminology. According to Stump, the biblical narratives that she discusses—the story of Job, the story of Abraham and Isaac, the story of Samson, and so on—are second person accounts of the events they relate. A second person account, in her terminology, is just a narrative that communicates the content of a second person experience. A second person experience is, roughly, a conscious experience of another conscious person as a person. A conversation with your child, an exchange of glances at a coffee shop, a hug—these are paradigm instances of second person experiences, to be contrasted, say, with a surgeon’s ‘objectifying’ experience of an unconscious patient on an operating table or with your own conscious awareness of yourself.²² According to Stump, second person experiences provide a particular kind of non-propositional knowledge, very much like whatever sort of knowledge Mary acquires upon coming to experience redness for the first time. Indeed, the example that she uses to illustrate and defend this point is just an adaptation of the Blackand-White Mary thought experiment. She considers a woman who has never experienced the presence of her own mother, but who has access to as much propositional information as you please about her. On Stump’s view, when the woman meets her mother for the first time, she acquires new knowledge: namely, the non-propositional awareness of what it is like to experience her mother. Stump then goes on to argue that second person accounts are able to communicate roughly the same knowledge that one gets from a second person experience by making that experience available to us through the narrative. Note too that the claim that second person accounts make second person experiences ‘available’ to us is to be taken quite seriously and robustly. Her view, I take it, is that the accounts make the experiences available not by giving us mere propositional information about the experiences (although they do in fact do this too), but by somehow putting us in touch with the experience in a way analogous to that in which q-memory might put us in touch with an experience.²³ Thus, putting (what I take to be) her view into my own terminology, I’d say that second person accounts mediate second person experiences, just as q-memories mediate sensory

²² On the concept of ‘objectifying’, cf. Bultmann 1984 and van Fraassen 2002, ch. 4. ²³ So I interpret her, anyway; she doesn’t herself invoke the concept of q-memory.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

160

   :  

experiences. This is why attention to biblical narrative figures so importantly in Stump’s theodical project. Many biblical narratives—in particular, the ones Stump considers in her manuscript—are second person accounts of suffering human beings’ second person experiences of God. On Stump’s view, attention to the narratives provides us with (what I would call) mediated experiences of God’s goodness; and the knowledge we acquire by way of these mediated experiences can be put to philosophical use in formulating a theodicy. How exactly the theodicy might go is, of course, tangential to my present concerns, so I’ll pass over that for now. What matters is just the fact that, if Stump is right, attention to biblical narrative is one way of acquiring mediated experiences of the divine presence. Note, too, that, a second person experience is, by definition, a conscious awareness of another person as a person. Second person experience is thus to be distinguished from what one might be tempted to label ‘third person’ or ‘objectifying’ experience of another person: e.g. the sort of experience one might have of another if one regarded her as a mere object. The import of this for present purposes is that a certain kind of ‘seeing as’ is a necessary condition for experiencing the presence of another person as such: one has to consciously regard the other as a person. Experiencing the person while seeing her as nothing more than a cleverly contrived automaton, for example, would be a wholly different kind of experience. Likewise, then, one would expect that a similar sort of ‘seeing as’ would be involved in having mediated experiences of the presence of another person. Thus, for example, if one were to read a story about Fred’s second person experiences of Wilma while failing to see Wilma as a (real) person—perhaps regarding her as a fictional character, or a figment of Fred’s imagination—the experiences conveyed by the narrative would be different and, in that event, there would be no reason to think that the narrative would in any sense be mediating Wilma’s presence. If this is right, then whether biblical narratives mediate the presence of God will depend importantly upon whether one takes those narratives to be reporting real experiences of God. Something similar might be said about liturgical actions.²⁴ Gesturing roughly in this direction, Nicholas Wolterstorff argues that the function of liturgical acts is primarily commemorative acts, and that commemorative acts are importantly linked to memory. Thus, he writes: Commemorations are meant to produce the memory of something in someone, or intensify the memory, or keep the memory alive; or to bring the remembered

²⁴ This idea was inspired by a paper (quoted here later on) given by Sarah Coakley at the Philosophy and Liturgy Conference at Calvin College in May 2008. The paper is ‘Beyond “Belief”: Liturgy and the Cognitive Apprehension of God’ [now published as Coakley 2013]. However, the paper doesn’t argue for or even assert the claim that liturgy mediates the presence of God, and the remarks therein that are suggestive of that view are not really fleshed out enough for me to attribute to her the sorts of views that I’m developing here.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

161

entity actively before the mind for a while, etc . . . . [T]o do something in commemoration of a certain event is to bring it about that the action signifies, or stands in for, that event; and . . . to produce something as a commemoration of some entity is to bring it about that the object signifies, or stands in for, that entity. (Wolterstorff 1990: 136, 139)

Liturgical acts, like the Eucharist, he argues, are commemorative of past events, like the Last Supper. Moreover, many (though not Wolterstorff, exactly) are inclined to see commemoration as a way of actualizing, or making present the things commemorated. Describing this view, Wolterstorff writes: Over and over in our century it has been said that the saving events of God which are commemorated in the liturgy are made present, or actualized, by way of the performance of the liturgy. The acts of God commemorated are not just acknowledged as having present significance; in some way the commemoration makes them actually present. One finds this position espoused, for example, in the comments of [Sigmund] Mowinckel and [Brevard] Childs on the liturgy of biblical Israel. Here is what Childs says in one place: “When Israel observes the Sabbath in order to remember the events of her redemption, she is participating again in the Exodus event. Memory functions as an actualization (Vergegenwärtigung) of the decisive event in her tradition.” (Wolterstorff 1990: 153; the quotation from Childs 1962: 53)

In the end, Wolsterstoff rejects the position described here as obviously false if taken literally. ‘God does not bring it about that the resurrection of Christ occurs when the liturgy is performed’, he writes. (155) But my own inclination is to think that what Childs and other adherents of the commemoration-as-actualization view are gesturing at is just the idea (which I have been articulating here) that, like q-memory, and like certain kinds of narrative, commemorative events, in the right context and undertaken in the right ways, mediate the presence of the events which they commemorate. If this is right, then liturgical acts, too, can be ways of experiencing the mediated presence of God. As with biblical narratives, a certain amount of ‘seeing as’ will be required for the divine presence to be mediated at all. For example, if the function of (some) liturgical acts is indeed to commemorate (or something like that), then, at the very least, one will have to see the liturgy as a vehicle for putting us in touch with historical events in which God himself was an actor. Moreover, it might be that one’s ability to experience the presence of God in the liturgy requires practice and training. Thus, for example, in the course of arguing for the conclusion that there might be some sense in which a liturgy might aptly be characterized as true, Sarah Coakley writes:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

162

   :  

. . . it could . . . be that the deeper ‘truth’ at stake in the liturgy is not propositional at all, but ‘truth’ in the particular sense intended by Christ when he said, according to John’s Gospel, that he was himself ‘the way, the truth, and the life’ (John 14:6). The intersection of liturgy and ‘truth’ would then consist in the liturgy’s capacity to train one’s sensibility to the presence of Christ in the same liturgy, and to knit one more deeply into his ‘true body’ through sacramental ingestion, attention to his Word, and the sharing of his communal love in the Spirit. (Coakley 2013: 134)

She goes on to argue that ‘what is distinctive to liturgical “knowing” . . . is the way that bodily movement, sensual acuity, affective longing, and noetic or intellectual response, are intricately entwined and mutually implicated in what is occurring, and indeed are being trained over time to intensify and deepen their capacity for response to the risen Christ’ (2013: 143–4). I must admit that I find it very difficult to see what the concepts of liturgical truth and liturgical knowing really amount to. But I have reproduced Coakley’s remarks here because I think that they are at least suggestive of the sort of view that I have been developing: namely, that (a) the right sort of participation in liturgical acts mediates the presence of God, and (b) the extent to which the divine presence is mediated depends importantly upon the (trainable) sensibilities that one brings to the experience. If the foregoing is correct, then (given that biblical narrative and the right sorts of liturgical forms—whatever those might be—are readily and widely accessible)  - is true. And if  - is true, then (so I have argued) divine silence is unproblematic. But we can go a step further, I think. I suggested earlier that there might be ways of experiencing divine silence contentedly, and perhaps even appreciatively—even while enduring suffering. I think that this suggestion is more plausible in light of the arguments just given than it is on its own. For if liturgy and biblical narrative mediate the presence of God, it is easier to see what the source of contentment and appreciation might be. Suffering human beings longing for the presence of God can go to the scriptures and the liturgy and find it—in small and mediated ways to be sure, but nevertheless in ways that provide them with the resources to see themselves not as lost and abandoned by God but rather as living daily in the presence of a loving but silent God.²⁵

²⁵ Work on this chapter was supported in part by a grant from the Institute for Scholarship in the Liberal Arts, College of Arts and Letters, University of Notre Dame. An earlier version of this chapter was read at the ‘Religious Experience and Knowledge of God’ conference, Dubrovnik, Croatia, 2008. I am grateful to the participants in that conference—especially Mylan Engel, Alvin Plantinga, and Bruce Russell—for helpful comments and conversation, and to Michael Bergmann, Jeff Brower, and Kevin Timpe for valuable advice and criticism on the penultimate version of the chapter.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

163

4. Postscript In this postscript, I want to offer some brief reflections on a few of the more interesting criticisms that this chapter has received—several raised by Ross Parker (2014), and one raised by Nick Trakakis (5). As shall emerge, I think that most of these objections do not quite hit their mark (though one of Parker’s does); but, at the same time, neither do I think that they are wholly without merit. I will begin with a brief discussion of Trakakis, followed by a more extended discussion of Parker. Trakakis’s concern is that my solution is overly anthropomorphic, ascribing ‘the kind of dependent finitude to God that the classical tradition of East and West have consistently sought to avoid’ (2015: 207). His basis for saying this is, first and foremost, my characterization of God as a person with a personality. On his view, one does better to respond to divine hiddenness by embracing (with much of the rest of the Christian tradition) a strong conception of divine transcendence. As he puts it, rather than respond to the hiddenness problem in ways that concede ground to the various anthropomorphic assumptions and analogies that tend to motivate the hiddenness problem, one might instead insist that ‘the appearance of hiddenness and silence may bespeak a failure to come to grips with the nature and otherness of the reality of God’ (2015: 208). As Trakakis notes in his brief and mostly sympathetic discussion of this chapter, I do gesture in the direction of this latter claim. His concern is simply that I do not go far enough. To an extent, I think he is right. The next chapter in this volume—‘Hiddenness and Transcendence’—was originally published in the same volume in which Trakakis’s paper appeared, and argues for a view very much in the neighbourhood of what he seems to prefer. That said, however, I still think that there is value in developing a solution more along the lines I have developed in the present chapter. The reason is simply that the nature of divine transcendence is deeply controversial; so if one’s solution to the hiddenness problem rests (as it must, if divine transcendence is the basis for the solution) on a particular conception of that attribute, one’s solution will be rather limited in its reach. Much better if one can also have something to say about divine hiddenness that will be of use to those who do not endorse one’s assumptions about divine transcendence.²⁶ Unlike Trakakis, Ross Parker objects to just about every aspect of the present chapter. He begins by arguing that the formulation of the hiddenness problem that I offer here ‘fail[s] to embody a plausible problem of hiddenness for theism’

²⁶ Accordingly, in The Hiddenness of God (Rea 2018), I develop a two-pronged solution, one prong of which appeals to divine transcendence and the other prong of which sets transcendence aside and reasons in accord with what I expect Trakakis would characterize as more anthropomorphic assumptions.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

164

   :  

(2014: 124). Moreover, he thinks my strategy for solving the problem is misguided: I aim at defence where instead I should be aiming at theodicy. On his view, contrary to what I say in the chapter, we do understand enough about God and God’s intentions and desires to be prima facie justified in taking divine silence as evidence against  . So, as he sees it, simply showing that divine silence is compatible with   (the strategy of ‘defence’) fails to address the problem; instead, what we need is a believable explanation for divine silence (something like a theodicy) to undercut the evidential force of divine silence against  . Finally, he thinks that even if defence rather than theodicy were sufficient to address the hiddenness problem, my defence is inadequate because it seems to rest on the mistaken idea that God’s provision of scripture and liturgy, together with the fact that scripture and liturgy can mediate divine presence, suffices for God’s having made divine presence widely and readily accessible. I will take each of these objections in turn. Parker’s main justification for the claim that I have failed to offer a plausible problem of divine hiddenness is that hiddenness as I characterize it here can obtain—i.e.     and   can both be true—without casting any doubt on  . The reason, he thinks, is that hiddenness as I characterize it can obtain and yet everyone still have warrant sufficient for knowledge that God exists (Parker 2014: 125). In support of this claim, he invites us to consider the following possibility: The many people who lack a priori and empirical evidence to justify belief that God exists (i.e. who make I E true) have frequent experiences of the loving presence of God. These people have direct evidence for God’s existence via their direct experiences of God, and because of their frequent religious experiences they could not reasonably feel neglected or abandoned by God. Additionally, the many people who fail to have direct experiences of God (i.e. who make A  R E true) have knowledge-level justification via their available a priori and empirical evidence. Further, their a priori and empirical evidence provides knowledgelevel justification for beliefs about God’s love for them and continual involvement in their life. (2014: 126)

Commenting on this scenario, Parker writes, ‘If it is possible that divine silence is true while every person has knowledge-level justification for belief that God exists and that God loves them and is active in their life, then the truth of   and     in itself is not in any way in tension with  .’ The problem with this objection, however, is that, contrary to what Parker supposes, the envisioned scenario is not one in which both     and   are true. If everyone has warrant sufficient for knowledge that God exists, then   is false.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

165

Parker seems to be supposing that one can have evidence—‘direct evidence’—that is neither empirical nor a priori. Even if it is not a conceptual truth that all evidence is either empirical or a priori, I was not assuming that there is such a middle category; and so, at worst, Parker has identified an infelicity in the formulation of  . Nevertheless, I do think that there are infelicities in my formulation of  , and I also think that my formulation of     is problematic. One problem common to both theses is that they suggest that a problematic form of divine hiddenness obtains only if many people possess inconclusive evidence, or suffer from absent (or, as I’d now put it, limited) religious experience. But that seems wrong: the aspects of divine hiddenness that the two theses were meant to capture are problematic even if only some, but not many, people experience them; and so the theses need to be reformulated to make that clear. But, as I now see it, there are substantive problems even beyond this. A    is meant to capture what I refer to in The Hiddenness of God as the experiential aspect of divine hiddenness (Rea 2018:18–20). As I formulate it here, however, it fails to do this because it fails to recognize that absence of the sorts of experiences described therein is problematic only for people who strongly desire such experiences. Accordingly, I am now inclined to formulate the idea I tried to capture with     as follows (Rea 2018: 20): L R E Some people have strong but persistently unfulfilled desires to have experiences that seem clearly to them to be experiences of the love or presence of God as such. I , as I formulate it here, is meant to capture the doxastic aspect of divine hiddenness. But it, too, fails at its task. The reason has partly to do with an ambiguity in the term ‘the available evidence’. Is the available evidence for a person the evidence she possesses or the evidence that she (in some sense) could possess? If the latter, then we really don’t know whether the evidence is inconclusive for anyone; for all we have access to is the evidence that we (collectively) in fact possess, plus maybe some reasonable speculation about what further types of evidence we could possess after diligent investigation of some sort. If the former, however, then   does not yet capture a problematic aspect of divine hiddenness; for the evidence a person possesses might be inconclusive simply because they are a sloppy investigator, or have self-deceptively blinded themselves to some important evidence. Moreover, a further problem with   is that, contrary to what the formulation here suggests, the problem is not just that some people’s evidence allows them to rationally believe that God does not exist. Rather, the problem is that some people’s evidence allows for other failures of belief in God—nonbelief in God’s existence that, for various reasons, may not rise to the level of disbelief. Accordingly, I am now

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

166

   :  

inclined to formulate the doxastic aspect of divine hiddenness by way of three separate theses (Rea 2018: 16): I E For some people, whatever evidence they possess in support of the existence of God is inconclusive in the following sense: even if they happen to see that this evidence at least weakly supports belief in God, it is not strong enough in relation to the rest of their evidence and background beliefs (however those might have been acquired) to produce in them rational belief in God. R N N N

Some people inculpably fail to believe in God. Some people nonresistantly fail to believe in God.

I do not, of course, say that (the doxastic aspect of) divine hiddenness obtains just if all three theses are true. For the sake of argument I grant that all three are true; but I think that one could generate a version of the hiddenness problem by appeal to any one of them, and I also think that there are other theses besides these that might reasonably be thought to be part of the doxastic aspect of divine hiddenness.²⁷ Returning to Parker’s paper, let me now examine his main objections to my solution. The first, again, is that my approach is wrong: I have offered defence when I should have offered theodicy. This objection rests substantially on his rejection of my claim that we don’t know nearly enough about God’s personality, motives, and so on to place an interpretation on divine silence. On his view, we do know enough; and the following quotation seems to capture his main reasons for saying so: Consider first that Rea acknowledges that we do know an important truth about God’s personality; namely,  :  : God strongly desires to promote the well-being of all his rational creatures, both now and in the afterlife. If   is true, then we have a solid understanding of one of God’s fundamental desires and motives. So when we attempt to interpret divine silence (or other of God’s actions) it’s not the case that we know little of importance concerning God’s beliefs and desires. Rather we interpret divine silence in light of the truth that if there is a God, he has perfect love for all humans and desires their ultimate well-being. Even if there is much we don’t know about the beliefs, desires, motives, and dispositions of a perfect Divine being, if divine silence

²⁷ Cf. Rea 2018: 17, n. 8.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

167

seems to clearly not promote human well-being, then we have some reason to interpret it as incompatible with  . (Parker 2014: 130)

A bit later, he goes on to say: It doesn’t follow from the truth that the behaviour of persons is interpretable and our lack of robust knowledge of the divine personality that there is no good reason to interpret divine silence as incompatible with   . . . . [I]t seems clear that we know enough about what a perfect being would do such that divine silence is a pro tanto bad state of affairs, such that if one wants to deny [that it would be bad for God to permit divine hiddenness to obtain], then one needs to provide a reason to think that divine silence promotes a good ‘the promotion of which would justify God in permitting whatever bad things come from divine hiddenness’ . . . . (Parker 2014: 131)

If this is right, he thinks, then defence—merely showing that divine silence is consistent with  —is not enough to address the hiddenness problem. What we need instead is something like theodicy. As he puts it, I need to ‘give a reason to think that divine silence, though pro tanto bad, is not allthings-considered bad by developing a plausible account of why God allows divine silence’ (2014: 132). Suppose Parker is correct about all of this. The implications reach far beyond my chapter. Parker’s idea seems to be that if we know that something like   is true, then we automatically know enough about God’s interests and desires to see divine silence as evidence against  . But the view that something like   is true, at least as a kind of a priori constraint on the concept of God, is not peculiar to my view. It is affirmed, more or less, by all parties to the hiddenness debate. (If something in the neighbourhood of   were not a point of general agreement, the hiddenness problem could not arise. Those arguing from hiddenness to the nonexistence of God would first have to motivate the claim that, if God exists, od strongly desires to promote our well-being.) So if Parker is right, then anyone trying to solve the hiddenness problem would have to aim for theodicy rather than mere defence. A tall order indeed! But I see no reason to think that Parker is correct. Grant that we know   is true, that it is ‘an important truth about God’s personality’, and that it functions as a partial constraint on our interpretation of divine silence. Grant, furthermore, that ‘if divine silence seems to clearly not promote human wellbeing, then we have some reason to interpret it as incompatible with  ’. It does not follow from this that one can resist the atheistic conclusion of the hiddenness argument only if one can ‘give a reason to think that divine silence . . . is not all-things-considered bad by developing a plausible account of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

168

   :  

why God allows divine silence’. To see why, just consider the example that Parker offers to motivate his reasoning: N-I C You meet a colleague in the hall, look him in the eye, and say ‘I have an important question I need to ask you’: He ignores you and looks the other way. Just then, you feel severe pain in your chest, and collapse to the floor. You look at your colleague again and ask ‘please call for help!’ but again he ignores you and looks the other way. (Parker 2014: 129)

Parker thinks that, given ordinary assumptions about the colleague’s ability to understand conversational English, the likelihood that he sees and hears you, and so on, you are ‘justified in believing that his behaviour is incompatible with his concern for [you]’ (129). But this is incorrect, at least if we add to the case the further features that are salient in the case of divine hiddenness. Suppose your understanding of your colleague includes something analogous to  : you think that if you know anything at all about your colleague, you know he strongly desires to promote your well-being. Suppose further that you also know that your colleague is in other ways socially unusual (but it is nonetheless a further constraint on your interpretation of your colleague that he is generally behaving in a very smart and resourceful way, with clear understanding of his surroundings). With these further features added to the case, it seems clear that you have no real justification for the positive affirmation that your colleague’s behaviour is incompatible with his concern for you. It is, of course, reasonable for you to be puzzled; and, absent at least a hypothetical answer to the question, ‘Why would he act this way if he really cares about me?’, it may also be reasonable to suspend judgment about your colleague’s goodwill towards you until you get further information. But it seems clear that nothing stronger than defence—a merely possible reason why your colleague might be acting this way, despite his great concern for you—is needed for it to be rational for you to maintain belief in his goodwill.²⁸ So likewise, I would say, in the case of divine hiddenness. Even if defence is enough, however, Parker thinks my own defence fails. My defence seems to rest on the idea that scripture and liturgy are what God has provided to make divine presence ‘widely accessible’. But, according to Parker, if this is the whole story about what God has provided, then the provision is inadequate for (at least) two reasons. First, by my own account, scripture cannot mediate the presence of God unless one takes it to be reporting real experiences of God. Thus, it looks as if scripture is not a way in which any non-theist can gain access to the presence of God; and something similar will be true of liturgy as well. Even if scripture and liturgy are in some sense widely and readily accessible, they

²⁸ Nor would I even insist that defence is required. If your antecedent belief in his goodwill is sufficiently grounded, this together with your own awareness of the gaps in your understanding of your colleague might be enough for you to rationally continue believing in his goodwill.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

169

are not sufficiently accessible to provide a fully general (even hypothetical) explanation of what God is doing for people experiencing divine hiddenness and suffering as a result of that experience. For example, the appeal to scripture and liturgy as loci of mediated divine presence tells us nothing about what God might be doing for people who lived and died prior to the development of the relevant scriptures and liturgies, or who live and die in parts of the world not yet reached by people in possession of those scriptures and liturgies. This is a valid objection, and points to a genuine defect in the chapter. As I would put it, the chapter seems to defend  - by way of the following thesis: L D God’s presence is only mediated through the liturgies and scriptures of particular religious traditions, none of which has been universally accessible to all human beings at all times and places. But  , all by itself, is insufficient to render divine silence unproblematic. I say that the chapter seems to defend  - by way of   because I do not in fact explicitly affirm anything like   (nor do I believe it to be true). But that should have been made clearer; and, more importantly, I should not have moved so quickly from the thesis that scripture and liturgy are loci of mediated divine presence to the claim that divine silence is unproblematic. From the vantage point of nearly a decade of further thinking about divine hiddenness, I am tempted to defend that move by saying that once we see that divine presence can be mediated through the (reasonably) widely accessible scriptures and liturgies of the Church, we can also see that there will be many other avenues to mediated experiences of God; and once we see that divine presence is, as I say in the chapter, mediated by way of a kind of ‘seeing as’, the door is open then to seeing how even those who are merely receptive to the existence of God without yet fully believing in God can have mediated experiences of God’s presence. But even this defence would need to be unpacked further; and so I will simply leave the matter here and refer interested readers to the more upto-date treatment of this issue in chapters 6 and 7 of The Hiddenness of God (Rea 2018). But, before concluding, there is one other defect in the present chapter that needs to be acknowledged. In appealing as I did to scripture and liturgy as loci of mediated divine presence, I was ignoring the significant obstacles to such mediation posed by people’s experiences of scripture, liturgy, and other religious tools and symbols as instruments of or contributing factors to various kinds of abuse. We are all familiar with ‘troubling’ passages in scripture—the binding of Isaac, the rape of Dinah, the story of Judah and Tamar, the handing over of Lot’s daughters to the crowd at his door, the sacrifice of Jephthah’s daughter, the conquest of Canaan, and many others—and it should be no surprise that these very passages,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

170

   :  

especially when construed as parts of God’s authoritative revelation to human beings, are for many people more a source of pain and bewilderment than a locus of mediated divine presence. The growing literature on what Michelle Panchuk (2018) calls ‘religious trauma’ provides ample testimony to the ways in which scripture, liturgy, and other aspects of organized religion can accomplish precisely the opposite of what I am claiming for them here. The question of what God is doing to show love to those who have experienced this sort of trauma is difficult and pressing, and this fact is one further reason why the move from the truth of  - to the claim that divine silence is unproblematic was far too hasty. For more on this topic, I refer interested readers to chapters 8 and 9 in particular of The Hiddenness of God (Rea 2018).

References Bultmann, Rudolf. 1984. ‘New Testament and Mythology’. In New Testament and Mythology and Other Basic Writings, edited by Schubert Ogden, 1–44. Philadelphia, PA: Fortress Press. Childs, Brevard. 1962. Memory and Tradition in Israel. London: SCM. Coakley, Sarah. 2013. ‘Beyond Belief: Liturgy and the Cognitive Apprehension of God’. In The Vocation of Theology Today: A Festschrift for David Ford, edited by Tom Greggs, Rachel Muers, and Simeon Zahl, 131–45. Eugene, OR: Cascade Books. Garcia, Laura L. 2002. ‘St. John of the Cross and the Necessity of Divine Hiddenness’. In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul Moser, 83–97. Cambridge: Cambridge University Press. Hitchens, Christopher. 1995. The Missionary Position: Mother Teresa in Theory and Practice. New York: Verso. Howard-Snyder, Daniel and Paul Moser (eds). 2002. Divine Hiddenness: New Essays. Cambridge: Cambridge University Press. Jackson, Frank. 1982. ‘Epiphenomenal Qualia’. Philosophical Quarterly 32: 127–36. Jordan, Jeff. 2004. ‘Divine Love and Human Suffering’. International Journal for Philosophy of Religion 56: 169–78. Kvanvig, Jon. 2002. ‘Divine Hiddenness: What is the Problem?’ In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul K. Moser, 149–63. Cambridge: Cambridge University Press. Ludlow, Peter, Yujin Nagasawa, and Daniel Stoljar (eds). 2004. There’s Something about Mary: Essays on Phenomenal Consciousness and Frank Jackson’s Knowledge Argument. Cambridge, MA: MIT Press. Moser, Paul. 2002. ‘Cognitive Idolatry and Divine Hiding’. In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul Moser, 120–48. Cambridge: Cambridge University Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

171

Moser, Paul. 2008. The Elusive God: Reorienting Religious Epistemology. Cambridge: Cambridge University Press. Murray, Michael. 2002. ‘Deus Absconditus’. In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul Moser, 62–82. Cambridge: Cambridge University Press. Panchuk, Michelle. 2018. ‘The Shattered Spiritual Self and the Sacred: Philosophical Reflections on Religious Trauma, Worship, and Deconversion’. Res Philosophica 95: 505–30. Parfit, Derek. 1971. ‘Personal Identity’. The Philosophical Review 80: 3–27. Parker, Ross. 2014. ‘A Critical Evaluation of Rea’s Response to the Problem of Divine Hiddenness’. European Journal for Philosophy of Religion 6: 117–38. Plantinga, Alvin. 2000. Warranted Christian Belief. New York: Oxford University Press. Rea, Michael. 2018. The Hiddenness of God. Oxford: Oxford University Press. Schellenberg, J. L. 1993. Divine Hiddenness and Human Reason. Ithaca, NY: Cornell University Press. Stump, Eleonore. 1985. ‘The Problem of Evil’. Faith and Philosophy 2: 393–423. Stump, Eleonore. 1990. ‘Providence and the Problem of Evil’. In Christian Philosophy, edited by Thomas P. Flint, 51–91. Notre Dame, IN: University of Notre Dame Press. Stump, Eleonore. 2009. ‘The Problem of Evil: Analytic Philosophy and Narrative’. In Analytic Theology: New Essays in the Philosophy of Theology, edited by Oliver D. Crisp and Michael C. Rea, 251–64. Oxford; New York: Oxford University Press. Stump, Eleonore. 2012. Wandering in Darkness: Narrative and the Problem of Suffering. Oxford: Oxford University Press. Teresa, Mother. 2007. Come Be My Light: The Private Writings of the Saint of Calcutta, edited by Brian Kolodiejchuk. New York: Doubleday. Trakakis, Nick. 2015. ‘The Hidden Divinity and What It Reveals’. In Hidden Divinity and Religious Belief, edited by Adam Green and Eleonore Stump, 192–209. Cambridge: Cambridge University Press. van Fraassen, Bas. 2002. The Empirical Stance. New Haven, CT: Yale University Press. van Inwagen, Peter. 2002. ‘What is the Problem of the Hiddenness of God?’ In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul Moser, 24–32. Cambridge: Cambridge University Press. van Inwagen, Peter. 2006. The Problem of Evil. Oxford: Oxford University Press. Wainwright, William. 2002. ‘Jonathan Edwards and the Hiddenness of God’. In Divine Hiddenness: New Essays, edited by Daniel Howard-Snyder and Paul Moser, 98–119. Cambridge: Cambridge University Press. Wolterstorff, Nicholas. 1990. ‘Remembrance of Things (Not) Past’. In Christian Philosophy, edited by Thomas P. Flint, 118–61. Notre Dame, IN: University of Notre Dame Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

8 Hiddenness and Transcendence For over two decades, the philosophical literature on divine hiddenness has been concerned with just one problem about divine hiddenness that arises out of one very particular concept of God. The problem—I’ll call it the Schellenberg problem—has J. L. Schellenberg as both its inventor and its most ardent defender. The concept of God in question construes God as a perfect heavenly parent, and seems to be the product of perfect being theology deployed within constraints imposed by modern ideals of parenthood. The idea that God is our heavenly Father is traditional within Christian theology (which shall be my focus in this chapter, as it is the tradition that I know best), and the method of perfect being theology has enjoyed an important place in that tradition as well. Nevertheless, one might reasonably wonder to what extent it makes sense to allow modern ideals about parenthood to drive our theological reflections in the ways that it has done in the contemporary hiddenness literature. Within the Christian tradition, theologians have typically allowed their views about the fatherhood of God to be shaped in light of their views about divine holiness and transcendence rather than the other way around. The same is true for the theology of divine motherhood that developed in monastic circles in the high Middle Ages. This is not to say that ideals about parenthood have been irrelevant to the theology of divine motherhood and fatherhood; far from it. But I think that it is fairly safe to say that, for the most part throughout the tradition, such ideals have rarely, if ever, played the sort of primary, driving role in theological reflection that they have done in the literature on the Schellenberg problem. Upon attending to this fact, one might conclude that Schellenberg and those who have embraced his method of theological reflection are simply taking Christian theology in a new and better direction. Alternatively, one might start to wonder whether the Schellenberg problem is in fact not a problem for traditional Christian theism at all, but rather an attack inadvertently mounted against a straw deity. In this chapter, I argue that the Schellenberg problem is an attack on a straw deity. More specifically, I argue that Schellenberg’s argument against the existence of God depends on certain theological claims that are not commitments of traditional Christian theology and that would, furthermore, be repudiated by many of the most important and influential theologians in the Christian tradition. I close with some very brief remarks about the implications of this conclusion for what I take to be the real import of the Schellenberg problem.

Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0009

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

173

1. Let me begin by stating the Schellenberg problem, with an eye to highlighting its most important underlying theological assumptions. The problem takes the form of an argument for the conclusion that God does not exist. In Divine Hiddenness and Human Reason, Schellenberg (1993) formulates the argument as follows:¹ S1

If there is a God, God is perfectly loving.

S2

If a perfectly loving God exists, reasonable nonbelief does not occur.²

S3

Reasonable nonbelief occurs.

S4

Therefore: no perfectly loving God exists.

S5

Therefore: there is no God.

Theists will not dispute S1; and, although S3 has been the subject of much dispute, it presupposes no substantive theological claims, nor are any such claims required for its defence. Accordingly, for the remainder of this chapter I shall focus on S2. Let me begin with some brief remarks about which God, exactly, Schellenberg has in view. Even Schellenberg acknowledges that not every conception of God is one on which S2 is plausible. For example, as we shall see in more detail at the beginning of Section 2, he seems happy to concede that those who regard God as absolutely incomprehensible might find S2 unacceptable. But if so, then whose God, exactly, is in view with this argument? Schellenberg’s answer to this question is quite explicit: the argument targets belief in ‘the personal God of traditional theism’ (2005a: 209). But what God is that? Theism itself is not a religious tradition in its own right; and the various religions that are paradigmatically theistic—Judaism, Christianity, Islam, and a few of their offshoots—embrace very different conceptions of God, very different views about how God is to be worshipped, and very different views on a wide range of other theological topics. Granted, all three of these religious traditions share some common views about God. Indeed, they have traditionally overlapped on a small family of theological claims that together comprise the philosophical-theological position known as classical theism. But there is no such thing as theistic orthodoxy to which one could appeal for support for S2; and, given the wide diversity of views about divine ¹ Schellenberg has expressed the argument in several slightly different ways over the years (including Schellenberg 2015a, his contribution to the volume in which the present chapter was originally published). But the differences among these formulations do not make a difference to the arguments that follow; for, as readers can easily verify, they are all predicated on the same basic theological assumptions (T1–T3 below). ² In discussing S2, Schellenberg has made it clear that by ‘reasonable nonbelief ’ he means ‘inculpable nonbelief ’, and that when he says that such belief ‘does not occur’ he means that it never occurs. So S2 should be understood as equivalent to the thesis that, if a perfectly loving God exists, inculpable nonbelief never occurs. (Cf. Schellenberg 1993: 25–9 and Schellenberg 2005a: 201, 203.)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

174

   :  

love and personality that have been developed within and across these various religious traditions, it is singularly implausible to suppose that there is any conception of either divine love or divine personality that could be considered a commitment of theism as such and that would be robust enough to lend support to S2. In light of this, it is perhaps unsurprising that Schellenberg’s own defences of S2 tend to appeal to very general considerations (e.g. analogies with human love, or alleged conceptual truths) rather than to particular theological doctrines. Moreover, when he does appeal to particular theological doctrines, he seems to draw only on small portions of the Christian tradition rather than either considering that tradition writ large or examining views about divine love and personality that are common to all three of the theistic traditions.³ The result of this methodology has been a rather remarkable detachment of his defence of S2 from virtually all of the theological work on divine love and personality that has been done in any of the major theistic traditions. This is noteworthy; for one would expect that if each of the theistic religions is committed to S2, then the easiest and most straightforward way to defend that claim would be to cite a variety of theologians in each tradition who more or less explicitly endorse it. Likewise, showing that many of the most influential theologians in any one of these traditions are not committed to S2 would suffice to undermine the claim that S2 is a commitment of theism in general, or of that tradition in particular. I am in no position to comment on the contours of Jewish or Islamic theology; but I think that one would be hard pressed to draw much unqualified support for S2 from the work of Christian theologians writing before the twentieth century (or even during the twentieth century, apart from the work of American evangelical Protestants). Be that as it may, in Section 2 I shall argue that many of the most influential theologians in the Christian tradition are, at any rate, not committed to S2. If that is correct, then the Schellenberg problem fails as an argument against the existence of the God of traditional Christianity, and a fortiori, it also fails as an argument against the God of theism in general. Before turning to that argument, however, let us examine in a bit more detail Schellenberg’s reasons for thinking that theists are committed to S2. Although Schellenberg has had a lot to say in support of S2 over the years, so far as I can tell, there is no final, master argument for that premise to be found in any one article or book chapter. Instead, what we find is an extended defence of S2 in Divine Hiddenness and Human Reason, followed by a variety of clarifications and supplementary remarks in subsequent articles written mostly in response to critics. Nevertheless, thanks to the steadfastness and internal coherence of Schellenberg’s views about the nature of divine love over the years, it is not difficult to reconstruct

³ See pp. 184–85 of the present volume.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

175

a master argument on his behalf.⁴ In doing so, we can also highlight some important underlying theological assumptions. The crucial premise in Schellenberg’s defence of S2 is the following claim: S6 Perfect love towards another person includes a strong disposition to seek personal relationship with him or her. According to Schellenberg, a personal relationship is to be understood as an explicit, reciprocally interactive relationship. Given this, S6 implies the following two theological claims: T1 One has a personal relationship with God only if one is involved in an explicit, reciprocally interactive relationship with God. T2 Divine love manifests a bias towards explicit, reciprocally interactive relationship with human beings. Schellenberg does not define what he means by ‘explicit, reciprocally interactive relationship’, but he does give some illustrative examples. On God’s side, such a relationship with a human person would involve such things as giving guidance, support, forgiveness, and consolation; on the human side, it would involve such things as worship and obedience; and the relationship would count as reciprocally interactive at least in part because what God gives in the relationship is relevantly connected to what the human being gives, and vice versa (Schellenberg 1993: 18–21). This suggests that what he has in mind is (at least) a relationship in which each participant is aware of the other as a person, and there is some kind of communicative interaction between the two such that each party to the relationship is evidentially in a position both to believe reasonably that the other person is intentionally communicating something to him or her and to understand the specific content of what is being communicated. The route from S6 to S2 is fairly simple. Given S6 and T2, it follows that if God is perfectly loving towards everyone, God will be strongly disposed to seek explicit, reciprocally interactive relationship with everyone. God will, in other words, manifest a strong bias towards explicit, reciprocally interactive relationship with human beings. But manifesting such a bias, he thinks, will involve at a minimum supplying every nonresistant person with enough evidence to form rational belief in God. As he puts it, a perfectly loving God ‘would, as it were, have to be convinced that there was reason to deprive us of the evidence for belief which an opportunity to enter into personal relationship with God requires’ (Schellenberg 2005b: 288; emphasis in original). So, if God is perfectly loving, reasonable ⁴ In doing this, I draw on what I take to be his most important extended discussions of S2— specifically, Schellenberg 1993, 2003, 2005a, 2005b.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

176

   :  

nonbelief does not occur—which is just to say that S2 is true. I note in passing that, on Schellenberg’s view, divine hiddenness—a term which surprisingly appears nowhere in the summary formulation of his argument from divine hiddenness—is just the fact that God has not provided evidence sufficient to form belief in God to every human being capable of a personal relationship with God. Given this terminology, Schellenberg’s argument from S6 to S2 boils down to this: a perfectly loving being will be biased towards explicit, reciprocally interactive relationship with everyone who is not resisting such relationship, and therefore such a being will be hidden from nobody except those who are resisting. I will not here contest this argument (except to challenge its starting point, S6). I present it simply in order to highlight the importance of S6 to the case for S2. Rejecting S6 undermines the argument, and it is hard to see how S2 could be defended without appeal to something like S6. So S6 is important. Why think it is true? Schellenberg has offered several different reasons. In Divine Hiddenness and Human Reason, he defends it by appeal to the following two claims: (i) divine love, insofar as it is analogous to the best human love, would seek to maximize the well-being of God’s beloved, and (ii) participating in a personal relationship with God would greatly enhance the wellbeing of any human person. Elsewhere, he takes a more direct route, defending S6 by saying that divine love would be analogous to parental love. Ideal parental love has been variously conceived across times and cultures; but in our time and culture, at any rate, ideal parental love is widely understood to include an overwhelming disposition to seek ongoing explicit and reciprocally interactive relationship with one’s child. Thus, taking modern ideals of parenthood for granted, the parent analogy seems to lend a great deal of support to S6 (Schellenberg 2003: 32–5). In a later article, he claims that S2 is a conceptual truth about divine love; and part of his basis for saying this seems to be the thought that S6 is a conceptual truth (Schellenberg 2005a: 212–13). By invoking the parent analogy, Schellenberg presupposes a further theological claim: T3 The fact that normal human parental love manifests a strong bias towards explicit reciprocally interactive relationship with one’s child is weighty evidence in support of the truth of T2. Moreover, each of the other two lines of defence seems to depend on it as well. If T3 were false, it would be implausible to suggest that S6 is a conceptual truth about divine love (cf. again Schellenberg 2005a: 212–13); and it would likewise be untenable to rest a case for S6 on an analogy with the best human love, given that parental love has strong claim to being among the best forms of human love. So Schellenberg’s case for S2 depends importantly on all three of the theological assumptions just highlighted. However, in the next section I will argue that each of

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

177

these claims is an uneasy fit with the broad theological framework endorsed by many of the most important and influential theologians in the Christian tradition and that, as a result, neither they nor S6 nor S2 can sensibly be thought to be commitments of traditional Christian theology.

2. In his own discussions of the implications of divine transcendence for the Schellenberg problem, Schellenberg has focused exclusively on the question whether and to what extent a transcendent God could act so as to render theistic belief reasonable. He grants that a strong theology of divine transcendence—one according to which God is absolutely incomprehensible, such that human concepts do not even analogically apply to God—makes it hard to see either how theistic belief could be evidentially supported or how one could meaningfully say that God has acted so as to provide evidence of God’s existence (1993: 46). But he says that only a doctrine of absolute incomprehensibility would have this result. So long as familiar predicates like ‘is just’ or ‘is loving’ at least analogically apply to God, God can supply us with evidence for theistic belief and, furthermore, divine love would require that God do so. Moreover, he insists that, in the context of the Schellenberg problem, . . . reference is being made not to an incomprehensible God, but to the personal God of traditional theism, whose love and justice, and so on, are conceived as sharing properties with their human counterparts, though of course they are thought of as perfected in various ways, and the manner of their instantiation or exercise might well be incomprehensible to us. (2005a: 209; emphasis in original)

On the strength of these considerations, and particularly in light of the alleged contrast between an incomprehensible God and the God of ‘traditional theism’, Schellenberg seems to think that the doctrine of divine transcendence has no significant bearing on the premises of the Schellenberg problem. I am with Schellenberg in thinking that many human predicates apply to God at least analogically. Indeed, I think that plenty of human predicates (e.g. is transcendent, is uncreated, and is either simple or not) apply univocally to God. I am also with Schellenberg in thinking that the method of perfect being theology, properly construed and implemented, is a route to genuine, even if only partial, understanding of the divine nature. Despite these points of agreement, however, I think that Schellenberg is mistaken both in his views about the importance of divine transcendence generally in the Christian tradition and in his understanding of the potential bearing of even a modest doctrine of transcendence for the premises of his argument.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

178

   :  

Note again the unqualified contrast between an incomprehensible deity on the one hand and ‘the personal God of traditional theism’ on the other. Divine incomprehensibility and divine transcendence go hand in hand as divine attributes. Often enough the terms ‘transcendence’, ‘incomprehensibility’, and ‘hiddenness’ are used interchangeably. But the claim that God is both transcendent and personal enjoys overwhelming support from the Christian tradition, and is a crucial part of the theological framework endorsed by theologians as diverse as Gregory of Nyssa, Pseudo-Dionysius, Thomas Aquinas, Jonathan Edwards, Karl Barth, and many others.⁵ In short, the personal God of traditional Christianity is a transcendent, incomprehensible deity. Indeed, many theologians would say that the personal God of traditional Christianity is absolutely transcendent. Witness, for example, the opening remarks of Elizabeth Johnson’s ‘The Incomprehensibility of God and the Image of God as Male and Female’: The holiness and utter transcendence of God over all of creation has always been an absolutely central affirmation of the Judeo-Christian tradition. God as God— source, redeemer, and goal of all—is illimitable mystery who, while immanently present, cannot be measured or controlled. The doctrine of divine incomprehensibility is a corollary of this divine transcendence. In essence, God’s unlikeness to the corporal and spiritual finite world is total; hence we simply cannot understand God. No human concept, word, or image, all of which originate in experience of created reality, can circumscribe the divine reality, nor can any human construct express with any measure of adequacy the mystery of God, who is ineffable. (1984: 441)

Johnson’s gloss on (absolute) divine transcendence is controversial. But that is neither here nor there as far as the present point is concerned. Her understanding of the tradition is entirely typical,⁶ whereas Schellenberg is committed to the first three sentences of the quoted paragraph being fundamentally mistaken. His unqualified contrast between the personal God of the tradition and an incomprehensible deity presupposes the following broad theological claim: T4 The Christian tradition as such is committed to an understanding of God’s personal attributes (love, justice, etc.) that straightforwardly conflicts with the claim that God is absolutely transcendent. T4 might indeed be true, but, at best, it will be extremely controversial. This is due partly to the fact that there are diverse understandings of ‘absolutely transcendent’ ⁵ Cf. Trakakis 2015 and references therein. ⁶ For a detailed, and extremely useful, survey of the doctrine of divine incomprehensibility from the Patristic period on through the Reformation, see Bavinck 2004: 36–41.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

179

in the tradition, not all of which are identical to Schellenberg’s (or Johnson’s, for that matter). But it is also due to the fact that there are diverse understandings within the tradition of the nature of divine love, justice, and other personal attributes. Thus, T4 cannot simply be taken for granted in an argument that relies heavily on a particular understanding of God’s personal attributes en route to the conclusion that the God of traditional Christianity does not exist. Schellenberg might insist that even if T4 is false on some precisifications, at least the following claim (which replaces the term ‘absolutely transcendent’ with Schellenberg’s own particular understanding of absolute transcendence) is true: T5 The Christian tradition as such is committed to an understanding of God’s personal attributes that straightforwardly conflicts with the claim that familiar predicates like ‘is just’ and ‘is loving’ neither univocally nor analogically apply to God. But T5 will also be extremely controversial. For many theologians will want to distinguish between analogy and metaphor, and will want to say that claims like ‘God is loving’ and ‘God is just’ are true or apt metaphors rather than univocal or analogical truths. Thus, again, insofar as Schellenberg’s target is the God of traditional Christianity, T4 and T5 cannot simply be taken for granted; they stand in need of substantial defence. But suppose we grant the truth of T5. Suppose we furthermore allow, as I think we must in order for Elizabeth Johnson’s understanding of the tradition to be genuinely uncontroversial, that the claim that God is utterly or absolutely transcendent admits of interpretations that are fully consistent with the claim that familiar predicates like ‘is loving’ apply analogically or even univocally to God. Let us also insist, as Schellenberg must, given the way in which he defends his premises, on the legitimacy of the method of perfect being theology as way of discovering truths about God. Should we then agree that, despite the centrality of divine transcendence to the Christian tradition, the doctrine has no bearing on the premises of the Schellenberg problem? I think that we should not, and this largely for methodological reasons. In the remainder of this section I explain why. Commitment to the method of perfect being theology is, first and foremost, commitment both to the thesis that God is a perfect being and to the viability of relying on at least some of our intuitions about perfection—e.g. about what it would take for a being to be perfectly loving, or perfectly knowledgeable—as a means for arriving at further true claims about God. Importantly, it is no part of perfect being theology to suppose that this method is perfectly reliable, or that intuitions about perfection are evidentially superior to or even on a par with the claims of scripture as evidence about what God is like. Nor is it any part of perfect being theology to suppose that the thesis that God is perfect (or any other claim about God) can be known independently of divine revelation. One can, I believe,

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

180

   :  

deploy the method of perfect being theology even within Barthian constraints, affirming that God is known, independently of revelation, only by God, that unaided human cognitive activity is inadequate to the task of arriving at substantive theological truths, and that ‘[n]o one has ever said, or can say, of himself . . . what God is; God is inexpressible . . . . He is, therefore, visible only to faith and can be attested only by faith’ (Barth 1957: 190. Cf. the distinction between ontotheology and ‘theo-ontology’ in Vanhoozer 2010: 104). The import of all of this is as follows. One can, as a perfect being theologian, start with the thesis that God is perfectly loving or that God is a perfect parent and rely on one’s intuitions about love in general or about parental love in particular to arrive at conclusions like T1–T3. In doing so, one would then very naturally attenuate one’s understanding of divine transcendence (or other divine attributes) in light of one’s understanding of divine love or divine parenthood. But that is just one way of deploying the method. Alternatively, one might start with the thesis that God is transcendent (or with other theses that lead via the method of perfect being theology to the conclusion that God is transcendent) and allow one’s understanding of divine transcendence to shape one’s understanding of divine love and divine parenthood. Moreover, in taking this alternative approach, one need not abandon the idea that divine love is analogous to human love in general or to parental love in particular; but one might well endorse very different views from Schellenberg about the extent to which these loves are analogous, or about which features of human love or human parenthood are most salient for understanding the nature and attributes of God. The alternative approach just described is not merely hypothetical; it, or something very much like it, has dominated the Christian tradition. The personal God of traditional Christianity has, for many of the most important and influential theologians throughout history, been the transcendent, simple, immutable, and a se God of classical theism. The idea that these are non-negotiable divine attributes has traditionally been seen to be one of the results of the method of perfect being theology, and it has exerted enormous influence both on the conceptions of divine love that are to be found in the tradition and also on the conceptions of what it might look like to enter into unitive, loving relationship with God. In the work of theologians who lay emphasis on transcendence as a divine attribute, divine love towards creatures is commonly understood not as the homey yearning of a human parent for an explicitly communicative and mutually reassuring relationship with her child, but rather simply as God’s goodness towards creation, God’s willing the good for particular creatures, God’s use of creation for good purposes, God’s grounding and illuminating creaturely goodness, or some combination of these.⁷ Some of these theologians seem to eschew ⁷ See, for example, Pseudo-Dionysius c.1987: 79–84; Augustine, On Christian Doctrine, 1.31.34; Aquinas, Summa Theologica, 1.20; and the discussion of divine love in the work of John of the Cross in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

181

talk of divine desire as a component of divine love altogether.⁸ Others identify the desire(s) involved in divine love either as a desire for the good of human creatures (or creation generally), or a desire for union with human creatures, or both. But even those who identify the desire for union as a component of divine love typically do not envision such union as something that God longs to have with every human being under just any conditions whatsoever, nor do they envision it as something that requires rationally supported belief in God. Instead, union with God is typically seen as something that God brings about in a person only after she has directed her will and desire towards God, apparently wholly independently of the degree or epistemic status of her belief in God.⁹ In the writings of the apophatic mystics—among the most important works in the Christian tradition on the subject of just how human beings might achieve union with God—the concept of contemplative union seems to stand in for the notion of a personal relationship with God. But the idea that this might regularly, reliably, or essentially involve anything like explicit, reciprocal interaction is largely foreign to that tradition. Instead, achieving contemplative union is commonly construed as mainly a matter of bringing one’s will into conformity with the will of God. To be sure, most of these authors envision the process of attaining contemplative union as one in which God sometimes causes within a person intense and vivid religious experiences and provides various ‘consolations’ in prayer. But the experiences in question are generally not seen as the ultimate goal of our quest for union with God, nor are they even typically seen as favours that ought to be explicitly sought for their own sake.¹⁰ Broadly speaking, then, in the work of those theologians who lay heavy emphasis on God’s transcendence, aseity, simplicity, and immutability, we find no substantial support for the idea that God might have anything like a bias towards mutual reciprocal interaction, or that a relationship with God would have to be what Schellenberg would think of as mutually reciprocally interactive, or that empirical and a priori evidence might be a necessary condition for such a Williams 2014. See also Bavinck 2004: 215–16, and Peckham 2015, esp. ch. 2. Cf. also Anselm’s treatment of divine mercy (Proslogion 8), according to which God counts as merciful simply by virtue of the fact that God acts as merciful people do rather than by virtue of possessing any of the characteristic emotions or desires of mercy. Similar things have been said, mutatis mutandis, about divine love throughout the history of Christianity. (Thanks to Jordan Wessling for this last point. Thanks also to Jordan Wessling and Peter Martens for the references to Augustine and Aquinas.) ⁸ Cf. Augustine, On Christian Doctrine, 1.31.34; Peckham 2015: 66, 74. ⁹ Cf. Aquinas, Summa Theologica, 1.20; Pseudo-Dionysius c.1987: 81–2; Anonymous c. 1981, sec. 34; Teresa of Avila, The Interior Castle, IV.1.7 (Rodriguez and Kavanaugh 1980: 319) and John of the Cross, The Ascent of Mount Carmel, bk 2, ch. 5.3 (Kavanaugh 1988: 89) See also Muller 2003: 561–69, esp. 564–65, 567. ¹⁰ See, e.g., Teresa of Avila, The Interior Castle, IV.2.9 (Rodriguez and Kavanaugh 1980: 326), and John of the Cross, The Dark Night of the Soul, 1.5 (Kavanaugh 1988: 173–5). For John of the Cross, such consolations seem to be little more than an impediment and distraction that one must hope to overcome in pursuing union with God.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

182

   :  

relationship.¹¹ Not only this, but when we look to the details of what these theologians have had to say about divine love and union with God, Schellenberg’s ideas about the nature of divine love and about the nature of ‘personal relationship’ with God are, at best, an uneasy fit.¹² This is not to say that such authors never speak of God as relating to human beings in a mode of explicit, mutually reciprocal interaction. Aquinas, for example, seems to think that the indwelling of the Holy Spirit involves (or can involve) something like relationship in that mode.¹³ Similarly, John of the Cross seems to think that the highest mode of divine presence to believers involves at least what Schellenberg would call explicit relationship.¹⁴ Rather, my point is that, in contrast to Schellenberg, none of these authors seems to think that God counts as perfectly loving only if God has a bias towards this sort of relationship and brings it about that everyone can have this sort of relationship with God just by willing it. The indwelling of the Holy Spirit is available only to believers; and it was surely as obvious to Aquinas as it is to us that not even every believer has what Schellenberg would call an explicit, mutually interactive relationship with God. Similarly, the highest mode of divine presence is, according to John of the Cross, available only to those who have devoted a lot of effort towards progressing along the route to contemplative union; but there is no indication in his works that God counts as less than perfectly loving to those who do not—and, in their present state, cannot, even if they desire it—enjoy that mode of presence with God. But what about the force of the parent analogy in its own right? The idea that God is our heavenly Father is a scriptural and creedal mainstay of the Christian tradition; and if God is perfect and a Father, then God is a perfect Father. Thus, if we are prepared to grant (as I think that many of us would) that perfect parental love is much like what Schellenberg takes it to be, one might very well wonder what would justify leaving that analogy and our intuitions about parental love in the dust as we theologize about the nature of divine love.¹⁵ The answer, in short, is ¹¹ See Peckham 2015, esp. chs 2–3, for a brief survey of different conceptions of divine love in the history of Christian theology and for detailed discussion of what he calls the ‘transcendent-voluntarist model’, which develops a conception of divine love within constraints imposed by an emphasis on divine transcendence. ¹² In discussing Augustine’s conception of divine love, for example, John Peckham writes: ‘Augustine’s ontology . . . prohibits a dynamic, reciprocal relationship between God and creatures’ (2015: 66). ¹³ On this, see Stump 2011: 36–9. ¹⁴ Cf. Payne 1990: 53–4. ¹⁵ One interesting answer, which I will not pursue in detail here, is that the force of the parent analogy needs to be understood, and perhaps somewhat mitigated, in light of other scriptural imagery apparently aimed at illuminating the nature of divine love for creatures. For example, God is portrayed (e.g. in the divine speeches of the book of Job) as showing loving concern for non-human animals; but presumably God’s love towards those creatures would not have to involve a quest for personal relationship as Schellenberg perceives it. Likewise, God’s concern for human beings is communicated in the New Testament not only via parent analogies but also via shepherd analogies, which themselves lend no support whatsoever to the idea that divine love for humans would involve a bias towards explicit, reciprocally interactive relationship. (Thanks to Helen De Cruz and Rebecca Chan for suggestion along these lines.)

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

183

that, although the claim that God is our heavenly Father is entrenched in the tradition, and although the claim that God is our Mother also has a surprisingly important place in the tradition, these claims have not nearly always been seen as telling us anything important about divine biases towards relationship with human beings in general. According to Peter Widdicombe (1994: 255), for example, in the writings of Athansius, Origen, and other Patristic authors, the fatherhood of the first person of the Trinity is understood in terms of his being ‘the unoriginate first principle’, the creator of all things, an ‘inherently generative’ deity, and ‘the fount of the Godhead’. Moreover, he writes: . . . It is notable that Origen and Athanasius, and the other Fathers discussed in this study, did not support their picture of God as Father either by drawing on the biological or on the psychological and sociological dimensions of human fatherhood. Contemporary ideas about the family and about adoption play no role in their discussions of the divine being or of the Father’s relation to us. (255)

Widdicombe makes this remark at the beginning of a discussion of how the views of Origen, Athanasius, and other church fathers about the fatherhood of God might be brought to bear on contemporary controversies about the patriarchal assumptions that might be involved in calling God ‘Father’ rather than ‘Mother’ or ‘Parent’ or something else entirely. But what he says here bears just as much on the question whether biblical, creedal, or historical theological affirmations of the fatherhood of God presuppose anything about the distinctively parental nature of divine love. In a word, they do not, at least not in the writings of the Patristics.¹⁶ It is also worth noting that even when scriptural affirmations of divine fatherhood do seem to tell us something about the nature of God’s love for human beings, it is almost always God’s love for believers, or for Israel, or for others who already believe in and in some sense worship God (e.g. ‘those who fear him’ in Psalm 103:13) that is in view in these verses. There is little, if any, scriptural support for the claim that God is a father to everyone; thus, even if perfect parental love were as Schellenberg envisions it, one would be hard pressed to find support from scriptural affirmations of divine fatherhood for anything as strong as T1 or S6. In light of all this, I am inclined to think that anyone looking to support general claims about traditional Christianity’s understanding of divine love by appeal to parental imagery found in the tradition would do better to draw on the theology of

¹⁶ But not just the Patristics. Bavinck, for example, seems to reduce God’s fatherhood to his unbegottenness (2004: 306) and says explicitly that the name of ‘Father’ is ‘not a metaphor derived from the earth and attributed to God. Exactly the opposite is true: fatherhood on earth is but a distant and vague reflection of the fatherhood of God’ (307). He cites Eph. 3:14–15 in support of this claim.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

184

   :  

divine motherhood that developed and flourished in the high Middle Ages. Unlike paternal imagery for God, maternal imagery in theological writings has often been used in a way that is clearly designed to exploit for theological use human understandings of parental love. Still, such imagery seems not to have been employed to convey anything like the idea that divine love manifests a bias towards explicit, reciprocal interaction. Rather, its most central uses were to convey by way of maternal metaphors facts about how God nourishes us through scripture and the teachings of the Church, facts about divine compassion, and facts about human dependence upon God (Bynum 1982: 146–69). My point is not that there is no precedent at all in the Christian tradition for thinking that the fact that God is our Father (or Mother) tells us something about the nature of God’s love for human beings. Rather, my point is simply that certain very prominent and influential understandings of the significance of paternal and maternal imagery as applied to God lend no support whatsoever to T1–T3. Likewise, the main point of my earlier discussion about divine love and unitive knowledge of God was not to say that the Christian tradition speaks unequivocally against T1–T3, but rather that the tradition neither unequivocally endorses them nor even relegates their denials to the outer fringes. Let me be clear, then, about how the response to the Schellenberg problem offered in this chapter differs from other responses in the literature. Most responses to the Schellenberg problem—at least those that focus on S2 rather than other premises of the argument—have operated with a specific conception of God (usually Schellenberg’s) and have tried to find some reason that might justify God thus conceived in permitting the occurrence of reasonable nonbelief. What I have done instead is to argue that the whole problem is predicated on a theology that is not part of traditional Christianity and is, furthermore, an uneasy fit with commitment to one of the historically major tenets of traditional Christianity— namely, the view that God is transcendent. In short, the theological credentials of T1–T3 are shaky at best. Although they are not outright denied by the Christian tradition, they can hardly be regarded as commitments of traditional Christian theology. Insofar as S2 and S6 depend on them, the same is true for those theses as well. As I mentioned earlier, Schellenberg himself shows some concern for the theological credentials of S2; but he treats the topic only briefly and, interestingly, the theological case he offers rests not on insights about divine love that are common to the theistic religions, but rather more on claims about the nature of Christian salvation. He says that his case for the conclusion that theologians as such are committed to S2 is summed up ‘nicely’ by the following quotation from Grace Jantzen: Salvation is not (or at least not primarily) about our future destiny but about our relationship to God and the gradual transforming effect of that relationship in

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

185

our lives . . . . If religious experience is centrally the sense of the loving presence of God, gradually helping people to reorient and integrate their lives in accordance with their love for him, is this not precisely what salvation is? Salvation must, surely, be religious experience if anything ever is: not in the sense of being a single climactic experience . . . but in the sense of a gradual opening of all life, all of experience to the wholemaking love of God. (Jantzen 1987: 128–9, quoted in Schellenberg 1993: 29)

Schellenberg follows this quotation immediately with the remark, ‘Hence theologians, too, seem committed to the affirmation of [S2]’ (1993: 29). But, of course, that is true (of the Christian tradition, anyway) only if the Christian tradition generally is on board with the claim that salvation is religious experience, and that the religious experience that salvation consists in is something that God’s love would lead God to provide for everyone at every moment of his or her life. Not even Jantzen seems to affirm the latter of these two claims; and, as is readily seen from my earlier discussion of what the apophatic mystics have to say about unitive knowledge of God, the conjunction of these two claims is neither an unequivocal affirmation of the Christian tradition nor even a clear majority view.

3. I noted earlier that Schellenberg regards S2 as a conceptual truth about divine love. Although he does not say it explicitly, he pretty clearly also takes S6 and T2 to be conceptual truths as well. If he is right, then all that I have said in Section 2 might seem irrelevant to the Schellenberg problem, for the Christian tradition would then be committed to S2 and S6 regardless of what anyone has had to say about the nature of divine love or about what it might be like to enter into a unitive relationship with God. But here again the centrality of divine transcendence to the Christian tradition is vitally important. For one very plausible consequence of even a very modest doctrine of divine transcendence is that we have no revelation-independent concept of divine love. There may well be purely conceptual truths about creaturely love, and about perfected creaturely love; but insofar as God is transcendent, there is good reason to think that it will always be at least partly an exegetical or systematictheological question (rather than a matter of mere conceptual analysis) to what extent divine love would resemble a hypothetically perfected creaturely love. Karl Barth, towards the beginning of his discussion of the perfections of divine freedom, makes roughly the same point in a very general way. Having already asserted that divine hiddenness (which, for him, is equivalent to divine incomprehensibility) is ‘the first word of the knowledge of God instituted by God himself ’ (1957: 183), Barth writes:

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

186

   :  

The recognition of divine attributes cannot be taken to mean that for us God is subsumed under general notions, under the loftiest ideas of our knowledge of creaturely reality, and that He participates in its perfections. It is not that we recognize and acknowledge the infinity, justice, wisdom, etc. of God because we already know from other sources what all this means and we apply it to God in an eminent sense, thus fashioning for ourselves an image of God after the pattern of our image of the world, i.e., in the last analysis after our own image . . . . God is subordinate to no idea in which He can be conceived as rooted or by which He can be properly measured. There are not first of all power, goodness, knowledge, will, etc. in general, and then in particular God also as one of the subjects to whom all these things accrue as a predicate. (Barth 1957: 333–4)

The idea here and in the surrounding context is that our knowledge of attributes like wisdom, power, goodness, and love is subordinate to what we learn by way of revelation, rather than the other way around. This has not been a minority view in the tradition; nor is it a minority view among contemporary theologians. But, if it is correct, it has as a straightforward consequence the claim that we have no revelation-independent concept of divine love; and if that is correct, then S2 and S6 are conceptual truths about divine love only if the concept of divine love that they presuppose is one that is somehow grounded in divine revelation. I have no argument for the conclusion that this is not the case; but neither has Schellenberg done the exegetical or systematic-theological work that would be required to show that it is the case. Suppose I am right about all of this. What conclusions should we draw about the significance of the Schellenberg problem? Even if the Schellenberg problem fails as an argument against the existence of the God of theism in general, or as an argument against the existence of the God of traditional Christian theism in particular, it still poses a threat to belief in a God about whom Schellenberg’s theological assumptions are true. Many theists do accept those assumptions; so, although it is an attack on a straw deity if the God whose existence it targets is supposed to be the God of theism or of traditional Christianity, the Schellenberg problem can easily be reframed as an argument with a real, definite target. I suspect that Schellenberg’s God has some claim to being the God of certain strands of contemporary American evangelicalism. Thus, in light of the arguments of the present chapter, one might reasonably see the Schellenberg problem as a referendum on that concept of God, and as a general challenge to rethink the biblical and systematic-theological warrants for thinking about divine love and about personal relationship with God in a way that privileges parent analogies understood in light of contemporary (probably also predominantly American) ideals of parenthood. For the most part, respondents to the Schellenberg problem have tacitly agreed with Schellenberg in thinking that the salient questions about the nature of divine love and personal relationship can mostly be settled a priori

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

187

rather than by taking a more systematic- or historical- theological approach. The arguments of the present chapter are meant primarily to pose a challenge to that way of thinking.¹⁷

4. Postscript Soon after this chapter was originally published in 2015, John Schellenberg wrote to me about it; and our correspondence convinced me that, despite my efforts to represent his views accurately, I had still in some ways missed the mark. I have also learned, partly from that correspondence and partly from more careful attention to some of his more recent work, that the difference between us in our views about the importance of paying attention to theological work on divine love arises out what seem to me to be deeper differences about the appropriate methodology for handling philosophical-theological topics. In this postscript, I want to call attention to what I take to be the two most important misrepresentations of Schellenberg’s views, and also say a few words about our methodological disagreements. In the opening paragraph of this chapter, I say that the Schellenberg problem arises out of a concept of God that ‘construes God as a perfect heavenly parent’. Later, in developing the contrast between the views about divine love that seem to drive Schellenberg’s argument and the views about divine love that seem to have greater currency among theologians, I suggest that the conception of divine love in play in the Schellenberg problem includes a sort of ‘homey yearning of a human parent for an explicitly communicative and mutually reassuring relationship with her child’. Neither of these two claims, however, is quite faithful to what Schellenberg has actually said in print. It would have been a lot more accurate to say that, in Schellenberg’s work, parental love is one important model among several for divine love (spousal love, love among friends, and so on being salient others); and it would have been a lot more accurate simply to focus on Schellenberg’s idea that divine love centrally includes bias towards relationship, without going on to suggest that the bias is motivated or attended by some sort of affective longing. I do not think that these corrections force substantive revisions in, or retractions of, the main critical points that I made in the chapter; but it is true that, had I represented Schellenberg’s views more accurately, the contrast that I drew between the conception of divine love that seems to motivate his argument ¹⁷ For helpful comments on earlier drafts, I am grateful to Rebecca Chan, Helen De Cruz, Adam Green, Carl Mosser, Sam Newlands, Eleonore Stump, and the members of audiences at the weekly discussion group of the University of Notre Dame Center for Philosophy of Religion and at the 2014 annual meetings of the American Academy of Religion and the Evangelical Theological Society. Work on this chapter was supported by The Experience Project, funded by The John Templeton Foundation in partnership with the University of Notre Dame and the University of North Carolina, Chapel Hill.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

188

   :  

and views that have been more prevalent throughout the history of theology would have been less stark. A more important misrepresentation comes in my characterization of the argument itself. The version of the Schellenberg problem on which I focused is the one that appeared in his 1993 book, Divine Hiddenness and Human Reason; and, though I gave a nod to the rather different formulation that appeared in the same volume in which this chapter was first published,¹⁸ I did not say much about the differences, nor did I allow those differences (or, really, any other work of Schellenberg’s on the hiddenness problem published between about 2005 and 2015) to play much role in shaping my own arguments in this chapter. Although I do not believe that I misrepresented the version of the hiddenness argument (and subsequent development) on which I focused, the discussion would certainly have been more faithful to Schellenberg’s current views had I expanded my focus to encompass more recent work of his. I did acknowledge in note 1 that I was ignoring more recent formulations (neither of which was published at the time I wrote this chapter); and I further commented that I did not think the differences would make a difference to the main argument of the paper. But I should have made good on that latter claim—as I will try to do now—instead of assuming readers would arrive at the same conclusion on their own.¹⁹ In my opinion, the best of the two most recent formulations of the Schellenberg problem comes in his book, The Hiddenness Argument.²⁰ The argument runs as follows: S1* If a perfectly loving God exists, then there exists a God who is always open to a personal relationship with any finite person. S2* If there exists a God who is always open to a personal relationship with any finite person, then no finite person is ever nonresistantly in a state of nonbelief in relation to the proposition that God exists. S3* If a perfectly loving God exists, then no finite person is ever nonresistantly in a state of nonbelief in relation to the proposition that God exists (from S1 and S2). S4* Some finite persons are or have been nonresistantly in a state of nonbelief in relation to the proposition that God exists. S5*

No perfectly loving God exists (from S3 and S4).

S6*

If no perfectly loving God exists, then God does not exist.

S7*

God does not exist (from S5 and S6).

¹⁸ Green and Stump 2015. ¹⁹ I hope to have rectified this mistake in the first two chapters of Rea 2018. ²⁰ The version in Schellenberg 2015a is, so far as I can tell, basically the same argument expressed in a way that trades away some of the simplicity and clarity of the book’s version of the argument for somewhat greater precision; but, to my mind, the simplicity and clarity of the book’s version is of significantly greater value.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

189

The most important differences between this version and the one I focused on in the present chapter are, first, the focus specifically on nonresistant nonbelief rather than on reasonable or inculpable nonbelief and, second, the explicitness and nature of the rationale for thinking that if there is a perfectly loving God, then the problematic form of nonbelief does not occur. Whereas in the previous argument the underlying idea was that a perfectly loving God would not permit reasonable nonbelief because perfect love includes bias towards relationship that would naturally lead to the removal of obstacles to relationship (like lack of sufficient evidence for belief in God), the present version emphasizes that perfect love includes openness to relationship, which Schellenberg seems to construe as partly including a disposition to remove such obstacles to relationship as one can without violating the autonomy of the beloved.²¹ I am inclined to think that this new version of the argument is stronger than the old: the revisions reflected in the new formulation smooth over some of the rough patches in the previous argument that became clear in the literature over the years, and they also stave off various misunderstandings that Schellenberg has been at pains to correct in the literature.²² But, despite the greater strength of the new version, I think it is equally vulnerable to the criticisms raised in the present chapter. As I initially construed it, the Schellenberg problem took S2 as a premise and relied for the defence of that premise on S6: S2

If a perfectly loving God exists, reasonable nonbelief does not occur.

S6 Perfect love towards another person includes a strong disposition to seek personal relationship [viz. an explicit, reciprocally interactive relationship] with him or her. I argued in the chapter that—owing in large part to its commitment to divine transcendence, neither of these two claims are commitments of the Christian tradition and, indeed, both would be rejected by many of the most important theologians within that tradition. But I think the same can be said for the first premise of the 2015b formulation: S1* If a perfectly loving God exists, then there exists a God who is always open to a personal relationship with any finite person.

²¹ This is, of course, just my rough summary of what Schellenberg means by ‘openness’. For his own characterization, see especially Schellenberg 2015b: 41. ²² A variety of misunderstandings are helpfully noted and corrected in Schellenberg 2005a and 2005b. Readers can judge for themselves how many of those are blocked by the 2015b formulation of the argument.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

190

   :  

The Christian tradition is, of course, committed to claims in the neighbourhood— most importantly, that God in some sense desires the salvation (which, we may safely assume, will eventually include something like personal relationship); but one would be hard pressed to show that the conceptions of divine love that one finds in the tradition entail that God is at all times open to relationship in the sense of ‘not making it impossible by [God’s] own acts or omissions for the [beloved] to participate in personal relationship with [God] at that time should the other wish to do so’ (Schellenberg 2015b: 41; emphasis mine). To repeat just one of the relevant considerations from the present chapter: many in the tradition have been inclined to identify divine love with God’s goodness towards creation, God’s willing the good for particular creatures, God’s use of creation for good purposes, God’s grounding and illuminating creaturely goodness, or some combination of these. But it seems clear that God can love creatures in any of these different senses without being at all times open in Schellenberg’s sense to personal relationship with creatures. But why appeal to theological reflection on divine love at all? Here we come to the main methodological disagreement between Schellenberg and myself. Schellenberg’s view seems to be that the most reliable way of theorizing about God is to mute the voices of the theologians and to reason matters out for ourselves. As he puts it in The Wisdom to Doubt: We have—all of us—been influenced by the many attempts of theology to make God fit the actual world. Theology starts off by accepting that God exists and so has to make God fit the world: in a way, that is its job. But our job as philosophers, faced with the question of God’s existence, is to fight free from the distractions of local and historical contingency, to let the voice of authority grow dim in our ears, and to think for ourselves about what a God and a Godcreated world would be like. [footnote omitted] When we think at the most fundamental level about the idea of God, we cannot assume that probably God’s nature is in accord with the actual world, and so we cannot take as our guide a picture of God fashioned by theology over the centuries on that assumption. (2007: 197–8; emphasis in original)

If this expresses the right way in general to theorize about God, then presumably it also tells us the right way to theorize about divine love: do so in a way that is selfconsciously unconstrained by theological doctrine or deference to theological tradition, and follow your own arguments wherever they lead. In evaluating this advice, however, we have to raise again the question of just what ‘God’ we are theorizing about. If our theoretical question is something like, ‘What sort of being would most conform to my own (or commonly shared) intuitions about perfection?’, then it seems to me that Schellenberg’s methodological advice is roughly on target. But in the context of the hiddenness argument

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

  

191

construed as an argument against the existence of the Christian God, the salient question is something more like ‘What is the nature of the God who is allegedly revealed in the Christian scriptures?’ Theorizing responsibly about this question is neither an individual task nor a purely philosophical task. It is an interdisciplinary matter, requiring attention not only to the scriptures themselves and what experts have said about them, but also to the history of systematic theological reflection on their contents.

References Anonymous. c. 1981. The Cloud of Unknowing. Translated by James Walsh. Text is Free of Markings edition. New York: Paulist Press. Barth, Karl. 1957. Church Dogmatics, Vol. 2, Part 1: The Doctrine of God. Edited by G. W. Bromiley and T. F. Torrance. Translated by T. H. L. Parker and J. L. M. Haire. London: T&T Clark. Bavinck, Hermann. 2004. Reformed Dogmatics, Vol. 2: God and Creation. Grand Rapids, MI: Baker Academic. Bynum, Caroline Walker. 1982. ‘Jesus as Mother and Abbot as Mother: Some Themes in Twelfth-Century Cistercian Writing’. In Jesus as Mother: Studies in the Spirituality of the High Middle Ages, 110–69. Berkeley: University of California Press. Green, Adam and Eleonore Stump (eds). 2015. Hidden Divinity and Religious Belief. Cambridge: Cambridge University Press. Jantzen, Grace. 1987. ‘Conspicuous Sanctity and Religious Belief ’. In The Rationality of Religious Belief : Essays in Honour of Basil Mitchell, edited by William J. Abraham and Steven W Holtzer. New York: Clarendon Press. Johnson, Elizabeth A. 1984. ‘The Incomprehensibility of God and the Image of God as Male and Female’. Theological Studies 45 (3): 441–65. Kavanaugh, Kieran (ed.). 1988. John of the Cross: Selected Writings. New edition. New York: Paulist Press. Muller, Richard A. 2003. Post-Reformation Reformed Dogmatics: The Rise and Development of Reformed Orthodoxy, ca. 1520 to ca. 1725. 2nd edn. Vol. 3. Grand Rapids, MI: Baker Academic. Payne, S. 1990. John of the Cross and the Cognitive Value of Mysticism: An Analysis of Sanjuanist Teaching and Its Philosophical Implications for Contemporary Discussions of Mystical Experience. Dordrecht: Kluwer. Peckham, John. 2015. The Concept of Divine Love in the Context of the God-World Relationship. New York: Peter Lang. Pseudo-Dionysius. c.1987. Pseudo-Dionysius: The Complete Works. Translated by Colm Luibheid. New York: Paulist Press.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

192

   :  

Rea, Michael. 2018. The Hiddenness of God. Oxford: Oxford University Press. Rodriguez, Otilio and Kieran Kavanaugh (trans). 1980. Collected Works of St. Teresa of Avila (Collected Works of St. Teresa of Avila) Vol.2. First edition. Washington, DC: ICS Publications. Schellenberg, J. L. 1993. Divine Hiddenness and Human Reason. Ithaca, NY: Cornell University Press. Schellenberg, J. L. 2003. ‘Divine Hiddenness Justifies Atheism’. In Contemporary Debates in Philosophy of Religion, edited by Michael L. Peterson and Raymond Vanarragon, 30–41. Malden, MA: Wiley-Blackwell. Schellenberg, J. L. 2005a. ‘The Hiddenness Argument Revisited (I)’. Religious Studies 41: 201–15. Schellenberg, J. L. 2005b. ‘The Hiddenness Argument Revisited (II)’. Religious Studies 41: 287–303. Schellenberg, J. L. 2007. The Wisdom to Doubt: A Justification of Religious Skepticism. Ithaca, NY: Cornell University Press. Schellenberg, J. L. 2015a. ‘Divine Hiddenness and Human Philosophy’. In Hidden Divinity and Religious Belief, edited by Adam Green and Eleonore Stump, 13–32. Cambridge: Cambridge University Press. Schellenberg, J. L. 2015b. The Hiddenness Argument: Philosophy’s New Challenge to Belief in God. New York: Oxford University Press. Stump, Eleonore. 2011. ‘The Non-Aristotelian Character of Aquinas’s Ethics: Aquinas on the Passions’. Faith and Philosophy 28: 29–43. Trakakis, Nick. 2015. ‘The Hidden Divinity and What It Reveals’. In Hidden Divinity and Religious Belief, edited by Adam Green and Eleonore Stump, 192–209. Cambridge: Cambridge University Press. Vanhoozer, Kevin. 2010. Remythologizing Theology: Divine Action, Passion, and Authorship. Cambridge: Cambridge University Press. Widdicombe, Peter. 1994. The Fatherhood of God from Origen to Athanasius. New York: Clarendon Press. Williams, A. N. 2014. ‘The Doctrine of God in San Juan de La Cruz’. Modern Theology 30: 500–24.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

9 Protest, Worship, and the Deformation of Prayer The idea that lament and protest might have a valuable place in Christian liturgy and practice has lately become a topic of increasing theological and philosophical interest. To take just a few examples: Kathleen O’Connor (2002), Kevin Timpe (2020), and Todd Billings (2015) have, in different ways, pointed to the role lament can play in both public worship and private prayer in expressing and encouraging hope in God in the face of suffering. O’Connor also argues, following Walter Brueggemann (1986), that lament teaches us to resist inappropriate uses of power; and she furthermore maintains that it teaches us to face, rather than deny, suffering in our own lives and in the world around us. Along similar lines, Christina Fetherolf (2016) and Michelle Panchuk (2019) have argued that lament and protest can both empower and help to heal survivors of various kinds of abuse and other traumatic experiences. The inclusion in holy scripture of numerous Psalms of lament, together with books like Job and Lamentations that are chock full of lament and protest, provides powerful evidence that God is willing to tolerate lament and protest from human beings, at least in what we might call their pious forms, where the lament and protest are directed to God in faith that God is good and in hope that God will be motivated to respond. Moreover, the fact that so many benefits seemingly accrue to us as a result of engaging in such prayer helps to explain why God might tolerate it. In The Hiddenness of God, however, I went a step further than many writing on lament and protest have been willing to go, arguing largely on the strength of my reading of Job and Lamentations that God not only tolerates pious lament and protest, but both authorizes and validates even some instances of impious protest—protest whose primary and most salient motivation is outright anger, despair, or similar affective states in response to the apparent injustice, wrongness, or unlovingness of God’s behaviour, and which is neither expressive of nor significantly motivated by faith or hope in God’s love or goodness. In fact, I even went so far as to suggest that God might sometimes prefer impious protest over more pious modes of religious engagement. I stand by these conclusions; but, at least on the surface, they are in tension with important and widespread assumptions about worship and prayer that I do not want to give up. In particular, it is hard to see how God can authorize and validate Essays in Analytic Theology: Volume II. Michael C. Rea, Oxford University Press (2021). © Michael C. Rea. DOI: 10.1093/oso/9780198866817.003.0010

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

194

   :  

impious protest against God if, as seems correct, it is always true that everyone ought to worship God. Furthermore, it appears that impious protest is an instance of what Lauren Winner calls the ‘characteristic deformation of prayer’ (2018: 83). In this chapter, after briefly explaining and defending the idea that God validates impious protest (Section 1), I will explain these tensions more fully (Section 2) and then explain how to resolve them (Sections 3 and 4).

1. Divine Validation of Protest The books of Job and Lamentations are both known for their powerful and piercing complaints against God. Job gives us the cries of a man who is presented by the book’s own prologue as innocent, righteous, and afflicted by God ‘for no reason’ (Job 2:3). Lamentations offers a prophet’s complaint against God for inflicting suffering upon Israel beyond anything she could sensibly be thought to deserve. Both books undeniably contain expressions of hope and trust in God— indeed, heroic expressions of hope in light of the broader perspectives of the main speakers in each book. But I see no reason to think that all of the complaints against God in each book proceed from a hopeful, faithful frame of mind. Kathleen O’Connor argues persuasively, against a ‘long tradition of interpreting . . . ’, that one finds in Lamentations at best a wavering, uncertain, and substantially confused expression of hope—one that quickly fades as the book moves into its final, despairing chapters. And the speeches of Job draw a very clear picture of a character whose mind wavers, in the face of persistent divine silence, between hope and trust in God and angry despair. For example, although Job admonishes his wife with the faithful remark, ‘Shall we receive the good at the hand of God and not receive the bad?’ when she urges him to ‘Curse God and die’ (Job 2:9–10), later speeches read more like angry protest. Witness, for example, Job 7:11–20:¹ ¹¹ “Therefore I will not restrain my mouth; I will speak in the anguish of my spirit; I will complain in the bitterness of my soul. ¹² Am I the Sea, or the Dragon, that you set a guard over me? ¹³ When I say, ‘My bed will comfort me, my couch will ease my complaint,’ ¹⁴ then you scare me with dreams and terrify me with visions,

¹ All scripture quotations in this chapter are from the New Revised Standard Version Bible. copyright © 1989 the Division of Christian Education of the National Council of the Churches of Christ in the United States of America.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,      ¹⁵ ¹⁶ ¹⁷ ¹⁸ ¹⁹ ²⁰

195

so that I would choose strangling and death rather than this body. I loathe my life; I would not live forever. Let me alone, for my days are a breath. What are human beings, that you make so much of them, that you set your mind on them, visit them every morning, test them every moment? Will you not look away from me for a while, let me alone until I swallow my spittle? If I sin, what do I do to you, you watcher of humanity? Why have you made me your target? Why have I become a burden to you?”

Some of the speeches, furthermore, include outright accusation. Thus, Job 30:16–23: “And now my soul is poured out within me; days of affliction have taken hold of me. ¹⁷ The night racks my bones, and the pain that gnaws me takes no rest. ¹⁸ With violence he seizes my garment; he grasps me by the collar of my tunic. ¹⁹ He has cast me into the mire, and I have become like dust and ashes. ²⁰ I cry to you and you do not answer me; I stand, and you merely look at me. ²¹ You have turned cruel to me; with the might of your hand you persecute me. ²² You lift me up on the wind, you make me ride on it, and you toss me about in the roar of the storm. ²³ I know that you will bring me to death, and to the house appointed for all living.” ¹⁶

It is hard to imagine these words being uttered with an attitude remotely resembling steadfast hope or faithful trust in divine goodness. These speeches I have just quoted, then, are examples what I am calling ‘impious protest’. On the supposition that all scripture is divinely inspired and valuable for spiritual teaching, their inclusion in scripture is deeply puzzling except on the supposition that God somehow authorizes such protest. How could a congregation take Job 30:16–23 as the Sunday sermon text, read it aloud from the pulpit, and then say—as we do in my church—‘The word of the Lord . . .

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

196

   :  

thanks be to God’ apart from the supposition that God in some sense values and accepts those very words of Job? Passages like these do not read like passages wherein biblical characters are portrayed as behaving in manifestly sinful ways; nor do they even read like historical sketches, offered without comment, of morally ambiguous behaviour. Again, Job is presented as blameless; and a reading on which he is blameless only until he starts to complain against God is strained at best. Lamentations, similarly, is given as the words of a prophet; and, given that the whole point of the book seems to be to bring the prophet’s (sometimes impious) complaints to God, a reading on which the book as a whole is both divinely inspired and subject to God’s disapproval seems bizarre. Accordingly, it seems plausible to treat both books as evidence that even impious protest against God is sometimes acceptable to God. As I read the book of Job, however, God does more than merely accept Job’s protest; God validates it. This is not to say that God endorses or agrees with Job’s protest. Rather, the idea is simply that God accepts it and recognizes it as a reasonable response to Job’s circumstances on the part of someone who loves goodness and justice but whose understanding of goodness, justice, and the relations between particular goods and evils is occluded by familiar human limitations. I read this validation partly in God’s explicit remark at the end of the book to the effect that Job alone among the speakers in the book has spoken rightly of God, but also in God’s treatment of Job when God finally appears in response to Job’s summons. The divine speeches begin with God appearing as a whirlwind before Job, who has previously described himself as a pile of ‘dust and ashes’ and accused God of making him ‘ride on the storm’ and tossing him about. (Job 30:19, 22) God invites Job to ‘brace [himself] like a man’ and receive God’s questions, which questions seem calculated to impress upon Job an overwhelming portrait of divine power and majesty. Many commentators have read these speeches as a divine smackdown—an extended ‘How dare you question the almighty?’, designed to silence Job and to rebuke him for the preceding thirty-two chapters of complaint and protest. But the speeches climax with God asking who can stand before the mighty Leviathan and be safe, the implied answer being ‘only God’; and, of course, the obvious and irresistible continuation of that thought is, ‘a fortiori, who can stand before God and be safe?’ Yet Job, the pile of dust and ashes, does stand, safely intact, before the divine whirlwind. Job spends thirty-two chapters demanding an audience with God and accusing God of treating him with cruelty and undeserved violence; and, as I have already noted, at least some of Job’s protests are of the impious sort—motivated by and expressive of emotions more like anger and despair than hope and faith. Yet when God finally appears, God lifts Job up, invites him to stand before God like a man, demonstrates that even in the face of God’s overwhelming power Job can stand and remain safe, and then concludes by restoring Job’s material blessings to the extent that they can be restored and saying

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

197

that Job has spoken rightly of God. To my mind, it is hard not to read this as God validating Job’s protest in the sense just described. One might wonder how it is that Job can be said to have spoken rightly of God if (as I have also indicated) Job’s protest reflects confusion, conceptions of God, love, justice, and so on that are distorted by human limitations. But I think reflection on fairly mundane misunderstandings makes it quite easy to see how this might be said of Job. Imagine that X and Y both acquire misleading but persuasive evidence that X’s romantic partner has taken a new lover. X is angry, insisting that they do not deserve such treatment, that the relationship is too valuable to be thrown away so casually, that the love between X and their partner was deep and genuine and a betrayal like this is absolutely unacceptable, and so on. Y, on the other hand, draws the understandable conclusion that things must have been pretty poor between X and Y, that either X did deserve the treatment in question or the relationship was simply not what they thought it was, that in any case the love between X and their partner must have been relatively shallow, and the like. It is easy to imagine that when X’s romantic partner finally steps into the conversation to clear things up, they might well say that X alone spoke rightly of them, even though X was in the grip of serious misunderstanding. I believe that, on the assumption that scripture is divinely inspired and the formation of the canon was providentially orchestrated, the inclusion of Job in scripture provides powerful evidence that God not only accepts but validates human protest, both pious and impious. But it is hardly the only such evidence. The inclusion of Lamentations in the canon provides similar evidence; as do the various Old Testament stories wherein people argue with God, wrestle with God, and, time after time, receive loving accommodation from God rather than rebuke.² And yet, theologically speaking, the view that protest is validated by God faces at least two important challenges. These I take up in the next section.

2. The Duty to Worship and the Deformation of Prayer The first challenge is that impious protest looks to be inconsistent with our absolute duty to worship God, and so it is hard to see how it can possibly be authorized or validated by God. That we do have a duty to worship God is both plausible and widely endorsed. Aquinas, for example, maintains that human beings owe God worship (ST II, 2, Q. 81). Nicholas Wolterstorff, identifies an explicit affirmation of the duty to worship in the ‘Rite One’ version of the Episcopal Church’s ‘Great Thanksgiving’, which speaks of thanksgiving to God ² This is not to deny, of course, the stories wherein people are rebuked (to put it mildly) for various kinds of rebellion. But my point here is not that God accepts all forms of protest, but only that God accepts some, and that some of what God accepts counts as impious.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

198

   :  

as our ‘bounden duty’; and he finds implicit affirmations of the duty to worship in passages of scripture (like Psalm 96) that enjoin us, in the imperative mood, to worship God and to do things like ‘ascribe to the Lord glory and strength’ that are at least partly constitutive of worship. Commenting on Psalm 96 in particular, Wolterstorff says, it is too weak to say that it is a good and joyful thing for the church to assemble to enact the liturgy [in worship of God]. Assembling to enact the liturgy is something the church ought to do; it is its bounden duty. Should it fail to do so, it would be guilty of wrongdoing. (2015: 42)

A bit later, he goes on to add that failure to worship wrongs God: It’s not just a good thing for the church to enact its liturgy for the worship of God; it’s obligatory that it do so. But if it is obligatory that the church do so, then it wrongs someone if it fails to do that. Who could that be other than God? (2015: 44)

These claims, furthermore, are made against the backdrop of his book’s first chapter, the conclusions of which he partially sums up as follows: Worship . . . is a particular mode of Godward acknowledgment of God’s unsurpassable excellence; specifically, a person is worshipping God if her attitudinal stance, when engaging in Godward acknowledgment of God’s unsurpassable excellence, is that of adoration. . . . Christian adoration of God is awed, reverential, and grateful adoration of God. . . . [T]he understanding of God implicit in our worship of God, and often explicit, is that of God as unsurpassable in glory, holiness, and love. (2015: 38)

Accordingly, it looks as if, on Wolterstorff ’s view, the proper aim of Christian liturgy is worship, and the Church’s duty to enact the liturgy for the worship of God comprises a duty to maintain and express a certain kind of positive attitudinal stance towards God. This is not to suggest that the Church, or any individual person, has a duty to maintain conscious positive attitudes towards God at every moment, so that sleeping or focusing intently on some task would put one in violation of the duty. Rather, I take it that the idea is that the Church has a duty to provide regular opportunities for the cultivation and expression of worshipful attitudes, and individuals have a duty to integrate the maintenance and expression of such attitudes into one’s life in increasingly deep and meaningful ways; and all parties have a duty to refrain from the cultivation or expression of attitudes towards God that are incompatible with the sorts of positive attitudes that Wolterstorff describes. Conceived of in this way (or in relevantly similar, even if

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

199

somewhat more limited ways), fulfilling the duty to worship God is straightforwardly inconsistent with what I am calling impious protest. Wolterstorff acknowledges that some may be inclined to resist (albeit for misguided reasons) the idea that worship of God has the status of duty or obligation; and, indeed, one way to resist it might be to reject the entire deontological framework that such a claim presupposes.³ But I think that making this move simply relocates the problem rather than solving it. For even those who resist (either for technical philosophical reasons, or for the sorts of reasons Wolterstorff envisions) the idea that we have a duty to worship God will nonetheless surely agree that God is someone who, by nature, deserves worship and whom we therefore ought (rationally, and perhaps also morally) to worship. It is natural to capture this ‘ought’ by reference to a kind of rational or moral duty; but we need not do so. We might say instead that the concept of God is the concept of a being for whom worship is always fitting and failure to worship is always unfitting, or a being whom it is always virtuous to worship and unvirtuous not to worship. However we capture the idea that God absolutely ought to be worshipped, the divine authorization and validation of impious protest seems to makes sense only if it is not the case that God absolutely ought to be worshipped. So the problem, in short, takes the form of a trilemma: our concept of God needs to be adjusted, or our concept of worship needs to be adjusted, or we need to reject the idea that God authorizes and validates impious protest. The second problem—that impious protest is a ‘characteristic deformation’ of the practice of prayer—is rather more complicated to explain. In The Dangers of Christian Practice, Lauren Winner (2018) argues that all Christian practices (like all human practices, perhaps) are susceptible to characteristic deformations— kinds of damage that are proper, rather than merely incidental, to the practices. She explains this idea primarily by way of examples. One way for a novel to be damaged is for it to be overly sentimental. Another way for a novel to be damaged is for some of its sentences to manifest the poor writing style of an officious copyeditor. The first form of damage is, according to Winner, proper to the art of novel writing; the second is not.⁴ Similarly, one way for a family meal to go awry is for a fight about politics to break out. Another way for a family meal to go awry is for it to be interrupted by an intruder. The first is a form of damage proper to the practice of family meals; the second is not. And so on. Can we say in a more general and precise way what it is for a form of damage to be proper to a practice? Winner herself does not try to do this; nor, importantly, ³ The sorts of reasons Wolterstorff envisions are not technical philosophical reasons, but rather reasons arising out of natural associations with commonsense understandings of the terms ‘duty’ and ‘obligation’—e.g. that doing something because it is obligatory implies doing it grudgingly rather than joyfully or spontaneously. ⁴ The examples here are somewhat modified versions of examples Winner herself gives to illustrate the concept.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

200

   :  

does she specify what exactly she means by ‘practice’. Is drinking coffee a practice? Drinking coffee with friends? Being part of a coffee klatch?⁵ It is hard to know where to draw the lines. That said, however, I think that reflection on her illustrative examples points to the following rough account of what it is for a form of damage to be proper to a practice. Let us say that a practice is any type of activity (i) about which it makes sense to say that the activity has gone intrinsically well or badly, or that it has been done well by those who are intentionally participating in it, (ii) that is not inherently bad when it goes well, and (iii) in relation to which we can identify behavioural dispositions—let us call them ‘virtues associated with the practice’—whose manifestation by practitioners of the practice tend to contribute to its going well or being done well. The first two conditions are needed in order for the notion of damage to make sense. It is hard to see what sense it would make to say that (e.g.) someone’s aimless stroll through the forest was damaged if (as I believe) the very activity of aimlessly wandering cannot be evaluated as having been gone intrinsically well or poorly, or as having been done well or poorly by the person participating in it. (Having a heart attack while aimlessly wandering, or finding a pot of gold, are, to be sure, ways for the wandering to go poorly or well, respectively; but those happenings are not intrinsic to the stroll; and it makes little sense, except as a joke. to evaluate someone as having done a good or bad job at strolling aimlessly.) Similarly, it is hard to see what one would be saying if one applied the notion of damage to (e.g.) an instance of armed robbery or murder. Something has gone undeniably well if a murder is foiled; and this, of course, is because the practice of murder is inherently bad.⁶ With practices thus characterized, it seems to me that Winner’s notion of damage proper to a practice can be adequately captured as follows: A form of damage is proper to a practice just if it tends to arise only out of practitioners’ failure to manifest some virtue or collection of virtues associated with the practice and furthermore tends to contribute to the practice not going or being done well. Sentimentality in a novel tends to arise out of the novelist’s failure to manifest virtues associated with novel-writing; bad style introduced by a copyeditor does not so arise. A family dinner erupting in a political squabble tends to arise out of one or more family members failing to manifest virtues associated with the practice of family dinners; an intruder’s disrupting the family dinner typically ⁵ Winner’s conception of a practice seems to be in the neighbourhood of MacIntyre’s (1984: 187ff.). But she does not cite MacIntyre as a salient influence; and, towards the end of her book, she cites a variety of theologians who, as she puts it, have been ‘drawn to the category’ of practice, without specifying which, if any, exerted significant influence on her own conception (see 2018: 168; and see note 4 for further discussion and citations). In any case, as critics have pointed out, MacIntyre is also rather vague about what exactly it takes for a kind of activity to count as a practice. ⁶ Cf. MacIntyre 1984: 187ff., and the corresponding characterization of virtue on p. 191. I am, of course, not strictly following his characterization here; but I am trying to capture a conception in the neighbourhood of what both he and Winner (again, perhaps independently) seem to intend.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

201

does not. There are no intrinsic virtues associated with the activity of wandering aimlessly, and this is at least part of why Winner’s notion of ‘damage’ cannot be applied to it. And so on. With this characterization in hand, it is now easy to see how impious protest against God can look like a form of damage proper to the practice of prayer. As it is typically conceived, when prayer goes well, it is worshipful, it draws one closer to God, and it includes at least one of the following four components (perhaps, but not necessarily, expressed verbally): adoration, thanksgiving, confession, supplication. It is not immediately obvious exactly what virtues contribute to prayer going well; but a natural thought is that among them would be both a disposition to submit to God and a disposition to interpret God and God’s actions in a positive light—i.e. in a way consistent with continued adoration of and thanksgiving toward God. But, of course, impious protest against God is a manifestation of neither disposition; and, on the typical conception of what it is for prayer to go well, it is natural to think that a prayer that includes impious protest is one that, precisely because of the failure of virtue that such a prayer would manifest, has not gone entirely well (even if it has gone well in at least some respects, such as its honesty, or its being directed towards God rather than an idol, etc.). What, then, can we say about these two problems? Consider again the trilemma that constitutes the first problem: our concept of God needs to be adjusted, our concept of worship needs to be adjusted, or we need to reject the idea that God authorizes and validates impious protest. I have no interest in developing a revisionary concept of God; and I have already made clear my commitment to the idea that God authorizes and validates impious protest. Accordingly, I think the problem lies ultimately in our concept of worship. In contrast to what Wolterstorff argues, I think that worship of God is best understood primarily in terms of concepts like love, devotion, allegiance, and honouring rather than in terms of concepts that entail positive attitudes. In the next section, I will unpack this idea in some detail; and what I have to say on that score will, in turn, point the way towards addressing the problem of protest as a kind of defective prayer.

3. Anger, Worship, and Prayer I want to begin by raising the question whether anger might ever be an apt—i.e. fitting and morally permissible—response to God, or God’s actions. Obviously lament and protest against God do not always involve anger; but impious protest in particular does often involve that emotion, and I think that it is anger’s common involvement in impious protest that makes it hardest to see how God could authorize or validate it. So I think that if it can be shown that, and why, anger towards God can be apt, validated by God, and consistent with worship of God, then the first of the two concerns in focus in this paper is effectively defused.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

202

   :  

So, again: Can anger ever be an apt response to God? Initially one might think not. Many philosophers doubt that anger is ever an apt response to anything; and even if we concede that anger is sometimes at least fitting, it is commonly thought to have generally harmful effects. Moreover, many philosophers, both historical and contemporary, think that there is an intimate, perhaps even partly constitutive, connection between anger and the desire to harm or to exact revenge.⁷ Moreover, anger is often thought to impair one’s capacity to reason and, in both public and private discourse, to polarize its targets against one’s own position (Bell 2009: 171–3; Srinivasan 2018: 124–5). Against these claims, some philosophers have argued that anger is not inevitably counterproductive, and often has positive effects. For example, summarizing her own brief survey of feminist defences of the value of anger prior to 2009, MacAlester Bell writes: [F]eminist philosophers have argued that anger is a mode of protest that can help maintain agents’ self-respect, that anger has . . . important epistemological roles to play in correct moral perception, that it allows us to bear witness to injustice, and that it can directly motivate social change. As should be clear, these defences of anger stress its instrumental value. Anger is defended as a valuable tool to develop in response to oppression because it can help bring about certain ends—either full recognition of one’s oppression, self-respect, or social change. (2009: 169)

But even if anger is sometimes fitting, non-harmful, and even productive, this does not guarantee that anger towards God can ever be apt. To the extent that human anger plays a role in correct moral perception and can motivate us towards positive social behaviour, a case can be made that God might have good reason to tolerate it, even when it is directed towards God. But, all by itself, such a case falls short of establishing the aptness of anger towards God; nor does it provide reason to think that God might validate such anger. For even if we grant that anger towards God is sometimes permissible, it may still not be a fitting response to God. Amia Srinivasan (2018), for example, takes anger to be an evaluative affective response to moral wrongness, injustice, or perhaps other norm-violations (e.g. she grants that one might respond aptly with anger to violations of epistemic norms).⁸ She thinks that anger is often apt, but only in the presence of actual norm-violations, rather than merely perceived—even justifiably perceived—violations. If she is right, and if God is morally perfect,

⁷ See the discussion in Nussbaum 2016: 17–27. Nussbaum herself thinks that desire for retribution is partly constitutive of anger. ⁸ Cf. also Nussbaum 2016, ch. 2.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

203

then anger towards God would never be fitting, even if it might sometimes be understandable, and so it could never be validated by God. So what can be said in defence of the fittingness of anger towards God, and in support of the claim that God might validate such anger? My answer comes in three steps. The first is to follow Srinivasan in rejecting the idea that anger constitutively involves the desire to harm or to exact revenge. As she notes, even if ancient cultures had trouble separating anger from the desire to harm or exact revenge, anger in contemporary Western culture often occurs unaccompanied by such desires. Granted, X’s anger towards Y might well include a desire on X’s part to make Y suffer at least through the appreciation of how her actions have caused X to suffer; but this is different from a desire to harm or to exact revenge.⁹ The next step is to highlight some of what is intrinsically good about certain kinds of anger—in particular anger that arises in response to injustice and other kinds of wrongdoing. MacAlester Bell, for example, argues that ‘appropriate anger’ is a ‘mode of hating or being against evil’. Thus, she writes: If humiliation and pain are evil, then on the view I’ve sketched, the virtuous person will be disposed to hate humiliation and pain. . . . If malice and cruelty are themselves evil, the virtuous person will respond to these vices with some form of hatred. If we understand anger as a kind of hatred or con-attitude, then anger would be a prima facie apt response to these vices. On this picture, what makes the character trait of appropriate anger a virtue is not its instrumental value in bringing about a state of eventual flourishing. Rather, the idea is that loving the good and hating the evil is itself non-instrumentally valuable. (2009: 177)

⁹ Srinivasan 2018: 129–30. Martha Nussbaum (2016: 21–7) emphatically rejects this position, insisting that anger always, with but one exception, involves a desire for some kind of retribution. The exception is what she calls Transition Anger, an allegedly ‘rare’ emotion whose ‘entire content’ [emphasis in original] is ‘ “How outrageous! Something must be done about this!”” (35–6) But Nussbaum’s argument that anger always involves a desire for retribution is premised on intuitionbased assessment of just one example; and even if her intuitions about that case are entirely correct, it does not seem to me that the example is adequately representative of the wide diversity of ways in which anger arises and manifests itself. More importantly, the view that anger constitutively involves some kind of desire for retribution implies that people who believe themselves to be angry without any desire for harm or other forms of payback to come to the target of their anger are simply mistaken— either they have failed (and maybe persistently tend to fail) to properly grasp their own conscious mental states, or else the desire for payback is a sort of unconscious desire of theirs. Although I do not deny that people have unconscious desires and even conscious desires that they somehow mistake for other mental states, it is not credible—especially just on the strength of intuition-based assessment of a small number of example cases—to insist that this is always what must be going on whenever someone thinks that they are angry in the absence of any desire for payback. Of course, it is possible to handle this worry in the way in which she handles a similar worry that one might raise about anger against oneself—namely, say that every such case is either one that involves a desire for payback or is an instance of Transition Anger (128–9). But I do not see how one can keep making this move without turning the category of Transition Anger into a mere catchall category for a variety of (not obviously rare) cases of anger that don’t fit her characterization.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

204

   :  

She goes on to argue that, as a response to wrongdoing, anger is not merely understandable, tolerable, or permissible; rather, it is an excellent response. Thus, appropriate anger is an excellent response in the domain of slights in that it is an especially fitting and appropriately expressive response in this domain. While other negative emotions such as disappointment or sadness, might also have a role to play in our emotional attachment to our moral standards and principles, in the domain of slights and disrespect, anger does the best job of: (1) fitting the failure; and (2) expressing the victim’s integrity, respect for the object of her anger, and commitment to the moral standards in question. (2009: 177)

Amia Srinivasan similarly defends the intrinsic value of anger as a response to certain circumstances; and she is particularly interested in showing that even counterproductive anger might nonetheless be apt. On her view, the value of anger lies in the fact that it ‘is a means of affectively registering or appreciating the injustice of the world’ (2018: 132). Comparing our capacity for anger with our capacity for aesthetic appreciation, she goes on: Just as appreciating the beautiful or the sublime has a value distinct from the value of knowing that something is beautiful or sublime, there might well be a value to appreciating the injustice of the world through one’s apt anger—a value that is distinct from that of simply knowing that the world is unjust. Imagine a person who does everything, as it were, by the ethical book—forming all the correct moral beliefs and acting in accordance with all her moral duties—but who is left entirely cold by injustice, feeling nothing in response to those moral wrongs of which she is perfectly aware. I don’t want to say that such a person has done anything wrong. But I do think it is natural to say that there is something missing in her; indeed, that it would be better, ceteris paribus, if she were capable of feeling anger towards the injustice she knows to exist. (2018: 132)

These remarks comprise the bulk of Srinivasan’s positive defence of the aptness of anger, and I think that, conjoined with Bell’s remarks on behalf of the value of anger as a mode of standing for what is good and hating what is evil, the defence is compelling. Obviously not every expression of anger will manifest these intrinsic goods. But some do and, intuitively, the goods that Bell and Srinivasan point to are genuinely significant.¹⁰ There is something deeply important about having an affective response, as opposed to a dispassionate response, to injustice. Although it seems doubtful that anger is ever a morally mandatory response to injustice,¹¹ it does seem to be a permissible and fitting response on many occasions. ¹⁰ For further defence, cf. Kauppinen 2018 and Shoemaker 2018. ¹¹ See Bell 2009: 177–8 for discussion.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

205

Again, though, we must ask whether anger is apt only in response to actual wrongdoing or whether it might also be apt in response to merely apparent wrongdoing. Here is where I part ways from Srinivasan, and the support I offer for this departure constitutes the third step in my defence of the claim that anger towards God might sometimes be both fitting and validated by God. Srinivasan’s defence of the claim that anger is apt only in response to actual norm-violations consists in the following three sentences: What if I mistakenly but justifiably believed that your not coming to the party was a moral violation? I’m inclined to say that my anger would be excused but inapt. If I learned that you in fact had made no promise to come to the party, I would hardly insist that my previous anger about your non-attendance was fitting. (2018: 129, n. 23)

But I think that the example she offers does not clearly support her point. The reason is that present judgments about the fittingness of one’s past anger are easily influenced by one’s present grasp of the past circumstances. Knowing now that one’s past anger was based on a mistake, there is considerable pressure to judge it as having been unfitting. (I think that present evaluations of the reasonableness of past beliefs that were based on misleading evidence are susceptible to the same distorting influence.) Change the example to remove this distorting pressure, however, and it is harder to side with Srinivasan. Suppose it now seems to you, on the basis of evidence you cannot explain by other means, that your business partner has embezzled a large sum of money from your business. You are, naturally, angry. Is your anger apt? I think it is; but the more important point here is that, whatever the correct answer might be, it seems not to depend on whether, in addition to your having otherwise inexplicable evidence that she has betrayed you, you believe correctly that you have been betrayed. Notice, too, that in constructing this example I have not said that you actually believe you have been betrayed, or that you have justifying evidence that you have been betrayed. Seemings are different from beliefs. The premises that enter into familiar logical paradoxes all might seem true even if you know that one is false. Something seems true just when you experience a kind of cognitive inclination or pressure towards believing it; and I think anger can be an apt response to seemings even when the evidence that prompts them is insufficient to justify belief. Change the example again: If it powerfully seems to you that your child is in mortal danger, grave concern is apt even if, upon reflection, you would realize that the evidence prompting the seeming doesn’t justify the belief that she is in mortal danger. In fact, lack of concern would be strange, at best. Why then wouldn’t it also be apt to be angry if it also seems to you that your child is in mortal danger as a result of the negligent activity of one of her caretakers?

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

206

   :  

In light of the foregoing, I think it is clear that anger towards a morally perfect God might sometimes be fitting. Much of what God permits in the world, and some of what God is said to have done in the world, powerfully seems to fallible human beings to be inconsistent with perfect love and goodness; and, in light of the foregoing, it seems fitting for those who believe in God and love goodness and justice and their fellow human beings to respond with anger. Job, and the prophet of Lamentations, are two biblical figures who seem to have been fittingly angry with God—even if, as both I and the tradition would affirm, God did not in fact act wrongly on the occasions that provoked their anger. If this is right, then anger towards God is sometimes apt. Insofar as such anger is indeed a mode of loving what is good and hating what is evil, and insofar as it aligns a person with God’s values even if not with God’s own perception of the circumstances, it seems furthermore that such anger will be validated by God—again, not endorsed as a correct response to the circumstance, but rather one recognized as a reasonable response on the part of someone who loves what is good and is taking a stand for it insofar as one has a grasp on it. The question now is whether expressing apt anger towards God is compatible with worship. I take it that, even if it is psychologically possible to maintain an attitude of awed admiration for someone while in the grip of (apt) anger against them, and even if it is semantically possible to express anger against someone while at the very same time expressing one’s awed admiration for them, these things are (at best) extremely difficult and uncommon. So it seems clear that if God does authorize and validate impious protest motivated by and expressive of apt anger towards God, and if worship partly consists in maintaining or expressing an attitude of awed admiration towards God (in the ways I described earlier), then there is in fact no absolute duty to worship God. But why characterize worship in terms of awed admiration? True, awed admiration often, maybe even typically, at least accompanies paradigmatic acts and attitudes of worship. But so too do a cluster of other acts, attitudes, and dispositions, and several of these strike me as candidates equal to or better than awed admiration for capturing the essence of worship. It is, for example, natural to identify worship with a kind of love, or devotion; it is also natural to identify worship with a kind of honouring. The latter is at least in the same conceptual family as ‘awed admiration’; but, importantly, it seems entirely possible to honour someone in thought, word, and deed while at the same time being angry with them and expressing that anger.¹² It would be surprising, for example, if the command to honour one’s parents entailed that one could neither be aptly angry towards one’s parents nor communicate one’s anger to them. And, of

¹² In fact, one might think it is not only possible to honour someone in the midst of anger, but anger is sometimes an expression of one’s respect for the person or valuing of the relationship. Cf. Kauppinen 2018.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

207

course, it is obviously possible to be angry with and express anger toward those whom we love or bear other forms of intense and stable devotion. It is not my task here to provide a definitive analysis of worship; but what I am recommending as a way of defusing the first of the two problems that I have identified is to construe our absolute duty to worship God not as requiring us to cultivate, maintain, and express positive attitudinal stance towards God, but rather as a duty to cultivate, maintain, and express a complex, positive dispositional stance. What I have in mind is something more like love, devotion, allegiance, honouring, or some combination of these, rather than mere positive attitudes like awed admiration. Construing worship this way makes room for the possibility of negative evaluative-affective responses to God without violating the duty to worship. A further advantage of the construal I propose is that, unlike admiration and other affective-evaluative responses (like anger), stances or dispositions like love, devotion, and the like are plausibly within our voluntary control. Accordingly, these latter, unlike the former, are more appropriately thought to be among our obligations or duties. This is not to say, of course, that our affective-evaluative responses are not at all subject to moral evaluation. After all, the question of anger’s aptness is in part a question of its permissibility. But when we speak of the permissibility of an affective-evaluative response, what we are primarily focused on is the permissibility of acting so as to cultivate, maintain, or express that response. Whether we have the response at any particular time, however, will not be directly up to us; and so we can hardly be faulted for having it. What we would be faulted for, if anything, is, as in the realm of problematic beliefs, failing to take steps to correct it. If all of this is correct, then there is no very compelling reason after all to think that impious protest is inconsistent with our duty to worship God. Accordingly, there is nothing problematic about the supposition that God authorizes it. Neither is there anything problematic about the supposition that God validates it. The apparently impious protest against God that we find in Job seems to be at least partly motivated by and expressive of respect for God’s law and love of justice and the good; and the protest we find in Lamentations seems to be at least partly motivated by and expressive of a concern for justice and compassion towards others. To be sure, if God is morally perfect, these protests reflect confusion on the part of the protestor; but insofar as they arise out of the good attitudes just noted, these protests, though impious, still locate the protestors among God’s staunch allies. They are confusedly expressive of allegiance to God and devotion to things that God loves; they may, in their own way, even constitute confused expressions of love for and devotion to God. It is easy, I think, to imagine God validating and, indeed, valuing such protest. It is even easy to imagine God preferring such protest over other modes of engagement, given the protestor’s inaccurate grasp of circumstances.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

208

   :  

4. Damaged Prayer? I have defused the first challenge to the idea that God authorizes and validates impious protest by arguing that such protest is not inconsistent with our duty to worship God and may in fact both arise out of and express deep allegiance to God and devotion to things that God loves. With this response in hand, we can now defuse the second challenge—the concern that impious protest appears to be a damaged form of prayer. Recall that the second challenge boiled down to this: As it is typically conceived, when prayer goes well, it is worshipful, it draws one closer to God, and it includes at least one of adoration, thanksgiving, confession, supplication. Plausibly, too, well-formed prayer is accompanied by a disposition to submit to God and to interpret God’s actions in a positive light. But impious protest seems to be prayer that manifests none of these characteristics; hence, it is broken prayer. The first thing to note by way of response is that, if the arguments of the previous section are sound, impious protest is consistent with a worshipful stance towards God (even if it is not itself expressive of worship). Moreover, on the traditional assumption that God is good, loving, compassionate, merciful, a lover of justice particularly for the oppressed, and that God is all of these things to a far greater degree than any human being might be, it seems that much of what was highlighted as beneficial about the manifestation and expression of apt anger would ultimately draw a person closer to God, even when one’s anger is directed and expressed towards God. Indeed, for some people the path towards recovering their conception of God as an ally in their struggle against injustice, and as a protector of their own best interests and the best interests of others might well involve a great deal of prayer wherein they bear affective witness to injustice before God, angrily reminding God of what human beings need, what mercy, compassion, and justice require, and what love (as it is humanly conceived, anyway) necessarily involves. With regard to just those first two components, then, impious protest may well be highly functional prayer, at least for a person in certain kinds of circumstances. Once we see this, however, then we must ask whether the remaining components of paradigmatically well-functioning prayer are in fact necessary for prayer to go well, or just typical of well-functioning prayer in ordinary circumstances. I would submit the latter. Just as love, devotion, honour, allegiance, and, accordingly, worship in ordinary circumstances are paradigmatically positive and full of praise and admiration, so too prayer would be; and so likewise prayer would naturally be accompanied by a disposition to interpret God positively and to submit to God’s desires. But impious protest on the part of a still worshipful believer arises out of extreme and unusual circumstances; and I think it is ultimately better to say that impious protest is one form of prayer that is still

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

, ,     

209

functioning well in such circumstances rather than to say that it is prayer that has been broken in part by the circumstances in which it occurs. The fear of the Lord is the beginning of wisdom, and we are called to work out our salvation with fear and trembling. But, at the same time, perfect love casts out fear. Like so many ideas in scripture, these ideas need to be held and understood in tension with one another. Neither trumps the other. It is, in part, the first cluster of ideas here that makes impious protest seem so far out of bounds. But I think one way, perhaps the only way, in which the love of a fearsome person who surpasses their beloved along every conceivable positive dimension can cast out fear is by that person showing themselves to be receptive and meekly responsive to the worst that the beloved has to offer. This is how it goes on the cross; and I think this is part of what’s going on in the divine authorization and validation of protest.¹³

References Bell, Macalester. 2009. ‘Anger, Virtue, and Oppression’. In Feminist Ethics and Social Philosophy: Theorizing the Non-Ideal, edited by Lisa Tessman, 165–83. London: Springer. Billings, J. Todd. 2015. Rejoicing in Lament: Wrestling with Incurable Cancer and Life in Christ. Grand Rapids, MI: Brazos Press. Brueggemann, Walter. 1986. ‘The Costly Loss of Lament’. Journal for the Study of the Old Testament 36: 57–71. Fetherolf, Christina M. 2016. ‘Lamenting Abuse: Reading Psalm 22 as a Response to Intimate Partner Abuse’. PhD thesis, Berkeley, CA: Graduate Theological Union. Kauppinen, Antti. 2018. ‘Valuing Anger’. In The Moral Psychology of Anger, edited by Myisha Cherry and Owen Flanagan, 31–48. Lanham, MD: Rowman & Littlefield. MacIntyre, Alasdair. 1984. After Virtue: A Study in Moral Theory. 2nd edn. Notre Dame, IN: University of Notre Dame Press. Nussbaum, Martha C. 2016. Anger and Forgiveness: Resentment, Generosity, Justice. New York: Oxford University Press. O’Connor, Kathleen H. 2002. Lamentations and the Tears of the World. Maryknoll, NY: Orbis Books. Panchuk, Michelle. 2019. ‘Toward a Liturgy of Protest’. Unpublished ms. Shoemaker, David. 2018. ‘You Oughta Know: Defending Angry Blame’. In The Moral Psychology of Anger, edited by Myisha Cherry and Owen Flanagan, 67–87. Lanham, MD: Rowman & Littlefield.

¹³ This chapter was presented at the Logos Seminar in the Logos Institute for Analytic and Exegetical Theology at the University of St Andrews, and as part of the Sophia Lectures series at Azusa Pacific University. I am grateful to the audiences on those occasions for helpful discussion. I am also grateful to Joshua Cockayne and Scott Davison for helpful comments on an earlier draft.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

210

   :  

Srinivasan, Amia. 2018. ‘The Aptness of Anger’. Journal of Political Philosophy 26: 123–44. Timpe, Kevin. 2020. ‘Toward an Account of Lamenting Well’. In Reaching for God: New Theological Essays on Prayer, edited by Oliver Crisp, James Arcadi, and Jordan Wessling. Oxford: Oxford University Press. Winner, Lauren F. 2018. The Dangers of Christian Practice: On Wayward Gifts, Characteristic Damage, and Sin. New Haven, CT: Yale University Press. Wolterstorff, Nicholas. 2015. The God We Worship: An Exploration of Liturgical Theology. Kantzer Lectures in Revealed Theology. Grand Rapids, MI: Eerdmans.

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages. Adam and Eve, see Fall Adam, as federal head or representative of the human race 20–2 Adams, Marilyn McCord 53n.1, 82–3, 84n.9 Adams, Robert 17n.8, 42n.42 Alien Guilt theories 19–22 Almeida, Michael 91–101, 137–8 Alston, William P. 93n.3, 127–8 Anderson, Robert 7 anger as a morally permissible response to God 201 as harmful 202 compatibility with worship 206–7 intrinsic goods of 203–4 role in correct moral perception and motivation 202–3 Transition Anger 202n.8 Anselm 17–19, 23, 45–8 apophatic mystics 181 Aquinas, Thomas 19, 22, 24–6, 32–3, 36, 46–7, 178, 181–2, 197–8 argument against the existence of God, see problem of evil argument from design 136 Aristotle 4–5 metaphysics 54–5 on natures 55 Arminius, James 17–18 Athansius 183 atonement as defeat of evil 83 as redemption 83–4 theory of 73–4, 81–2 see also problem of the stain on the soul Augustine 13–14, 16–17, 19, 22–3, 39n.38 Barth, Karl 178–80, 185–6 basic Aristotelian framework, see Aristotle, metaphysics Bavinck, Hermann 178n.6, 180n.7, 183n.16 Bell, MacAlester 202, 204 Bergmann, Michael 91–4, 100, 108n.3, 115–17, 124, 137

Berkouwer, G. C. 19 Bible 103, 148; see also biblical narratives biblical narratives 109–11, 158–60, 168–70 Billings, Todd 193 Bird, Alexander 59n.15 Brower, Jeff 18n.10, 65 Brueggemann, Walter 193 Bultman, Rudolf 159n.22 Bynum, Caroline Walker 183–4 Calvin 17, 19, 156n.16 Chan, Rebecca 182n.15 Childs, Brevard 161 Chisholm, Roderick 30n.28, 32 Christ as providing us identity and new narratives 84–6 empathetic engagement of 80–1 works beyond atonement 82 Christian theology 172 Christian tradition 177–84, 189–90 Christus victor model 5–6 classical theism 173–4 Coakley, Sarah 146, 160n.24, 161–2 collective responsibility 14n.3 Collins, Anthony 32 commemorative acts 160 composition 56–7 Conditional transword depravity 16–48 constituents 56–7, 61–4 constitution trinitarianism 65–6 consubstantiality 64–5 contemplative union with God 181 Council of Trent 19n.16 Craig, William Lane 42n.42 Crisp, Oliver 26n.26, 53n.1 Cross, Richard 53n.1 deception hypothesis 120–6 Decree on Original Sin 19n.16 DeCruz, Helen 182n.15 deformation of prayer 193–4, 208–9; see also lament and protest

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

212



deontology and worship 199 developmental and personality psychology 84 distinguishing properties 63, 67–8 divine hiddenness and absence of religious experience 143 and benefit to the sufferer 145, 155 and divine concern 144–6, 149–57, 163–8 and human blindness 147–8, 152n.11 and human good 144–6, 154–5 and inconclusive evidence for God 143–4, 146–9, 164–6 and limited religious experience 165; see also divine hiddenness, absence of religious experience; and see divine hiddenness, and divine concern and nonresistant nonbelief in God 165–6 and reasonable nonbelief in God 165–6 and the silent father objection 154–5 and unreasonable suffering 151–3 and using the term hiddenness 146–9 as a feature of God’s personality 149–57 problem of 6–9, 143–6, 172, 175–6; see also Schellenberg problem vs divine silence 146–9 divine incomprehensibility 177–9 divine love and non-parental imagery 182n.15 and the nature of Christian salvation 184–5 as analogous to parental love 176, 182–4 as analogous to the best human love 176 conceptual truth about 176, 185–6 divine providence, Molinist account of 40–8 divine self-disclosure 148, 156–63; see also divine hiddenness divine silence and God’s concern for us 148; see also divine hiddenness, and divine concern experienced with contentment and appreciation 162 divine transcendence 9, 172, 177–87 Draper, Paul 115–16 Duns Scotus 13–14 duty to worship God 9–10 scriptural support for 197–9 Edwards, Jonathan 15–16, 15n.4, 22, 178 theory of imputation 26–9, 32–3, 35–6 evangelicalism 186–7 evidence for the existence of God, empirical and a priori 147, 157; see also divine hiddenness, and inconclusive evidence for God evidential arguments from evil 91–101 evidential problem of evil 114 evolutionary theory 2n.5

Fall, the 1–2, 13, 16–18, 44–5 fatalism 44 federalist theory of original guilt, see original guilt, federalist theory of Fetherolf, Christina 193 Fine, Kit 57n.10 finkish dispositions 48 Fission theory 29, 36–40 Flint, Thomas 42n.42, 54n.2 form defined 64 as part vs constituent 56–7 see also Aristotle, metaphysics Forrest, Peter 14n.3 free will 40–8 fundamental properties 56 Garcia, Laura L. 145n.2 Genesis 13 God-justifying goods 144–5 God as heavenly parent 172–6, 180–4 as immaterial 65 duties to worship, see duty to worship God otherness of 9 personality of 9; see also divine hiddenness, as a feature of God’s personality protest against, see lament and protest gospel 4; see also Christ gratuitous suffering 113–18 Green, Ashbel 22n.21 Gregory of Nyssa 178 grievances against God 82–3; see also problem of the stain on the soul, and blaming God Hasker, William 42n.42 Helm, Paul 30n.28 Hitchens, Christopher 155n.14 Hodge, Charles 15n.4, 19n.16, 26n.26 Holy Spirit 181–2 Howard-Snyder 123n.10 Hudson, Hud 41n.40 human condition, as characterized by sin and misery 4 human corruption, nature of, see original sin, doctrine of, the nature of the corruption human nature damage to 1–4 nature of 53, 61–2, 66–70 wounded vs. corrupted 17–18 human telos 1–2 Hume, David 32 humility about expectations thesis 8–9 hylomorphism 54, 56–7, 59–61

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

 Immaculate Conception, doctrine of 13 immaterial things 55–6; see also God, as immaterial; see also natures, of immaterial things immediate imputation, doctrine of 18–19 incarnation, metaphysics of 53–4, 66–70 Jackson, Frank 158 Jantzen, Grace 184–5 Jesus of Nazareth 4, 44–5, 53–4, 66–70 as representative of the human race 21 resurrection of 5 see also Christ Job, book of 194–7 John of the Cross 181–2 Johnson, Elizabeth 178–9 Johnston, Mark 57n.10, 60 Jordan, Jeff 145n.4 Kauppinen, Antti 203n.9, 204n.11 Kavanaugh, Kieran 181n.10 Kelly, J. N. D. 17n.8 knowledge argument 158 Koslicki, Kathrin 54n.3, 57n.10 Kvanvig, Jon 143n.1 La Place, Joshua 19 lament and protest 9–10 impious forms validated by God in scripture 194–7 impious protest as deformation of prayer 199–201, 208–9 impious protest consistent with duty to worship God 197–207 pious forms 193 scriptural authorization of 83 Lamentations, book of 194–7 Lewis, David 36n.34 liturgical actions 160, 168–70 liturgical truth and liturgical knowing 162 liturgy, worship as aim of 198–9 Loux, Michael 56n.8 Luther 17, 19 MacBride, Fraser 66n.21 MacIntyre, Alasdair 199n.4, 200n.5 Maitzen, Stephen 114–20, 124, 126, 130–4, 137–41 material objects 56–7, 61–3 matter-form compound 56–7, 63 matter, defined 64 McBrayer, Justin 116 McDaniel, Kris 56n.9 McKitrick, Jennifer 59n.15 Mclean, Kate C. 84n.9

213

mediate imputation, doctrine of 18–19 mediated experiences of God 157–62, 168–70 mereological summation 56–7 Merricks, Trenton 32n.31, 54n.2 Milazzo, G. Tom 7–8 modern ideals of parenthood 172, 176 Mohila, Peter 14n.2 Molinism, see divine providence, Molinist account of Molnar, George 57–8, 59n.15 moral paralysis 137–41; see also sceptical theism, and scepticism about our obligation to prevent harm Moser, Paul 145n.2, 147n.6 Mother Teresa 149–50, 155–6 Murray, John 19n.16 Murray, Michael 41n.39, 145n.2 narratives, as defining who we are 84–6; see also biblical narratives National Synod of France 19 natural properties 57–8 natures and individuators 61 as fundamental powers 57–9 as principles of unity 59–60 as uniting other powers 59–61 nature and function of 57 of immaterial things 62–3 of material objects 61–2 universals vs particulars 66 New Testament 5–6; see also gospel non-propositional knowledge 109–11 noseeum inference 91–3 numerical sameness without identity 54, 62–5 Nussbaum, Martha 202nn.7,8 O’Connor, David 114–19, 126–30, 134–5 O’Connor, Kathleen H. 193 Oppy, Graham 91–101, 137–8 Organic Whole theory 22, 24–6, 29, 33–6 Origen 22n.21, 183 original guilt doctrine of 1–3, 13 federalist theory of 20–2 original justice 17–18 original sin, doctrine of 1–3 definitions of 13 objections to 13–16 and the nature of the corruption 16–18 Orthodox Eastern Church 14n.2, 17–18 Panchuk, Michelle 169–70, 193 Parfit, Derek 157–8

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi

214



Parker, Ross 163–9 Patristic authors 183 Peckham, John 180n.7, 182n.11 penal-substitutionary model 5–6 perfect being theology 172, 177, 179–80 methodology 179–80 perfect love, and seeking personal relationship 175; see also divine love persistence 29 endurantist theories of 30–3, 38–40 perdurantist theories of 30–3 stage theories of 31–3, 36–40 worm theoretic account of 29–30, 33–6, 39–40 see also Edwards, Jonathan, theory of imputation Personal Guilt theories 19, 22–6 personal relationship, nature of 175 philosophy of religion, analytic 103 Plantinga, Alvin 44n.44, 127–8, 156n.16 possible goods, nature of 91–3 potentiality 55–6 powers 56–60 practice, characterized 199–200 pre-existing souls, doctrine of 22 presentism 32 principle of possible prevention 13–16, 19, 21, 24–6, 32–8, 40, 44–9 principles of life 55–6 pro tanto reasons 96–9 pro tanto vs decisive reasons 122–3 problem of divine hiddenness, see divine hiddenness problem of evil 6–9, 91, 103 attempts to solve it as immature 103–7 methodology 108–9 theodicy solutions to 104–11 problem of the stain on the soul 74–5, 78–81 and blaming God 80–1 and removing the stain 75–8 and stains caused by trauma 79–80 protest, see lament and protest Protestant Orthodoxy 6 Pseudo-Dionysius 178, 180n.7 quasi-memory 157–60 Quinn, Philip 17n.8 radical sceptical scenarios 120–6; see also sceptical theism, and global scepticism Ramsey, F. P. 66n.21 Rea, Michael 8, 65, 83, 163, 165, 169–70 Realism about metaphysical unity with Adam 22 reasonable non-belief 173, 184, 189; see also Schellenberg problem

redemption 83–6 redemptive narratives 85 reflective sceptical theists 120 Reformed tradition 15, 17, 19 religious experiences 126n.11, 128–30, 181, 184–5 religious trauma 169–70 Roman Catholic Church 17–18 Rowe, William 91–3 Russell, Bruce 120, 121n.8 salvation 73–5, 81–2 necessity of 2–4, 73 sanctification 3–4 sceptical theism 9, 107, 111, 114–15 and commonsense morality 114–15, 140; see also sceptical theism, and scepticism about our obligation to prevent harm and consequentialism 117, 136 and global skepticism 119–26 and scepticism about knowledge of God 119, 135–7 and scepticism about our obligation to prevent harm 119, 137–41 and scepticism about what counts as sin 126 and scepticism and God’s goodness 134–5 and to scepticism about divine commands and values 119, 126–35 characterized 91, 94, 115–18 as undermining ordinary moral practice 91, 93–101; see also sceptical theism, and skepticism about our obligation to prevent harm Schaff, Philip 14n.2 Schellenberg problem 172–91 and misrepresentations of the argument 187–91 as attack on a straw deity 172 see also divine hiddenness Schellenberg, J. L. 146, 172–91 methodology 174, 190–1 Schreiner, Thomas 15n.4 scripture, and intervening to prevent serious harm 138–9 second person accounts 109–11, 159–60 second person experience 109–11, 159–60 sense of self 84; see also narratives sensus divinitatis 156n.16 Shedd, W. G. T. 22–6, 22n.21, 26n.26 Shoemaker, David 203n.9 Sider, Theodore 30n.28 sin universality of 2–4, 13–14, 47–8 forgetting of by God 76–8, 80–1 psychological effects of 74–5, 80–1

OUP CORRECTED AUTOPAGE PROOFS – FINAL, 23/10/2020, SPi



215

Smith, Donald 36n.34 Sonship 66–9 soteriology 3–4, 73, 81–2 souls blemish on 83–4 see pre-existing souls, doctrine of see problem of the stain on the soul Srinivasan, Amia 202–5 staining of souls, see problem of the stain on the soul Stuchlik, Joshua 31n.30 Stump, Eleonore 6, 73–81, 103, 108–11, 145n.4, 146, 158–60 substances 55–7, 65–6 Swinburne, Richard 14n.3

Timpe, Kevin 193 Trakakis, Nick 163 Transition Anger, see anger, Transition Anger Transworld depravity 42–8 Trinity, doctrine of 4–5 metaphysics of 65–6 Turretin, Francis 20

temporal parts, doctrine of 31–2; see also persistence, perdurantist theories of temporal predications, counterpart-theoretic analysis 31–2, 37–8 theodicy 114n.4, 158–9 and hubris 105–7, 111 as minimizing evil and minimizing goodness of creation 105–7 theology Christian, see Christian theology; see also Christian tradition of divine motherhood 172, 183–4 perfect being theology, see perfect being theology

Wainwright, William 14n.3, 37n.36, 145n.2, 147n.6, 152n.11 well-formed prayer 9–10 Westminster Shorter Catechism 1 White, T. H. 78 Widdicombe, Peter 183 Wiley, Tatha 17n.8 Wilks, Ian 114–26, 135–7 Williams, A. N. 180n.7 Winner, Lauren 193–4, 199–201 Witt, Charlotte 54n.3 Wolterstorff, Nicholas 160–1, 197–8 Wright, N. T. 7, 103–8 Wykstra, Stephen 91n.1

Unger, Peter 137 universality of sin, see sin, universality of Urban, Linwood 17n.8 van Frassen, Bas 159n.22 van Inwagen, Peter 42n.42, 143n.1 Vanhoozer, Kevin 179–80