Ramified Natural Theology in Science and Religion: Moving Forward from Natural Theology 9780367373191, 9780429356131

Citation preview

‘Probabilistic natural theology is the production of arguments from the most general features of the universe to the probable existence of God. Ramified natural theology is the production of probabilistic arguments from historical evidence of a kind recognisable by atheists as well as religious believers, for the truth of doctrines about how God has acted in history. The detailed application to ramified natural theology of the probability calculus, and in particular of Bayes’s theorem, is a very recent development in philosophy of religion, to the details and power of which Rodney Holder introduces readers in this book. It helps us to see just how similar to those of any other historical argument, are some arguments from the New Testament (considered merely as an ordinary historical document) and other historical evidence to the conclusion that Jesus Christ rose from the dead on the first Easter day. Holder explains well the background of the past two centuries of theological scepticism towards rational discussion of this topic, a rational discussion which is crucially needed both by Christians and their opponents in today’s world.’ Professor Richard Swinburne, Emeritus Nolloth Professor of the Philosophy of the Christian Religion, University of Oxford ‘At a time when religious faith is often dismissed as private and irrational, Dr  Holder powerfully and incisively argues for the complete rationality of Christian belief. This book demands our attention.’ Professor Roger Trigg, Senior Research Fellow, Ian Ramsey Centre, University of Oxford ‘Distilling decades of deep reflection on the consonance between the natural sciences and religious commitment, Rodney Holder develops with remarkable clarity and care a potent cumulative case for theism in general and for Christian theism in particular. Few thinkers combine scientific expertise and philosophical acumen as effectively and effortlessly as the author of this book.’ Dr James Orr, University Lecturer in Philosophy of Religion, Faculty of Divinity, University of Cambridge

Ramified Natural Theology in Science and Religion

This book offers a rationale for a new ‘ramified natural theology’ that is in dialogue with both science and historical-critical study of the Bible. Traditionally, knowledge of God has been seen to come from two sources, nature and revelation. However, a rigid separation between these sources cannot be maintained, since what purports to be revelation cannot be accepted without qualification: rational argument is needed to infer both the existence of God from nature and the particular truth claims of the Christian faith from the Bible. Hence the distinction between ‘bare natural theology’ and ‘ramified natural theology.’ The book begins with bare natural theology as background to its main focus on ramified natural theology. Bayesian confirmation theory is utilized to evaluate competing hypotheses in both cases, in a similar manner to that by which competing hypotheses in science can be evaluated on the basis of empirical data. In this way a case is built up for the rationality of a Christian theist worldview. Addressing issues of science, theology, and revelation in a new framework, this book will be of keen interest to scholars working in Religion and Science, Natural Theology, Philosophy of Religion, Biblical Studies, Systematic Theology, and Science and Culture. Rodney Holder was Course Director of The Faraday Institute for Science and Religion, Cambridge, UK, before retiring in 2013, and is a Fellow Commoner of St Edmund’s College, Cambridge. He has published widely in the fields of science and religion and natural theology including the book Big Bang, Big God: A Universe Designed for Life? (2013) and articles in peer reviewed journals such as The British Journal for the Philosophy of Science, Theology and Science, and Philosophia Christi.

Routledge Science and Religion Series Series editors: Michael S. Burdett University of Nottingham, UK

Mark Harris

University of Edinburgh, UK

Science and religion have often been thought to be at loggerheads but much contemporary work in this flourishing interdisciplinary field suggests this is far from the case. The Science and Religion Series presents exciting new work to advance interdisciplinary study, research and debate across key themes in science and religion. Contemporary issues in philosophy and theology are debated, as are prevailing cultural assumptions. The series enables leading international authors from a range of different disciplinary perspectives to apply the insights of the various sciences, theology,  philosophy and history in order to  look at the relations between the different disciplines and the connections that can be made between them. These accessible, stimulating new contributions to key topics across science and religion will appeal particularly to individual academics and researchers, graduates, postgraduates and upper-undergraduate students. Science and the Truthfulness of Beauty How the Personal Perspective Discovers Creation Robert Gilbert Against Methodology in Science and Religion Recent Debates on Rationality and Theology Josh Reeves Interreligious Perspectives on Mind, Genes and the Self Emerging Technologies and Human Identity Edited by Joseph Tham, Chris Durante and Alberto García Gómez God, Evolution, and Animal Suffering Theodicy without a Fall Bethany N. Sollereder Ramified Natural Theology in Science and Religion Moving Forward from Natural Theology Rodney Holder For more information and a full list of titles in the series, please visit: www. routledge.com/religion/series/ASCIREL

Ramified Natural Theology in Science and Religion Moving Forward from Natural Theology Rodney Holder

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 Rodney Holder The right of Rodney Holder to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Scripture quotations are taken from Common Bible: Revised Standard Version of the Bible, copyright 1965, 1966 by the Division of Christian Education of the National Council of the Churches of Christ in the United States of America. Used by permission. All rights reserved. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-0-367-37319-1 (hbk) ISBN: 978-0-429-35613-1 (ebk) Typeset in Sabon by Apex CoVantage, LLC

Contents

Prefaceix Acknowledgementsxi   1 Theology as a scientific discipline: the place of natural theology and ramified natural theology

1

  2 Reason and religion: the role of natural theology

9

  3 Natural theology and modern cosmology: the cosmological and design arguments

36

  4 Moving on from natural theology: why we need ramified natural theology

65

 5 Pascal’s Pensées and Butler’s Analogy: foreshadowing ramified natural theology

82

  6 The rationality of belief in miracles

106

  7 Ramified natural theology and evidence for Christian claims about Christ: Jesus’ miracles

128

  8 Jesus and prophecy

148

  9 Ramified natural theology in action: outline of the argument for the historicity of the resurrection of Jesus

170

10 On the third day he rose again: Bayesian methodologies applied to the resurrection of Jesus

189

viii  Contents

11 Towards a fuller picture: the fruits of ramified natural theology

210

Bibliography219 Index228

Preface

Natural theology, as traditionally conceived, is concerned with what we can know about God purely by being human, and so without recourse to any special revelation. The subject has had a chequered history but is alive and well today, and discussed widely by philosophers, scientists, and theologians. As such, it has been the main focus of my academic work over many years now. In particular, I  have examined the fine-tuning argument which came out of modern cosmology in some detail in articles and books. I have also looked at the challenge posed to natural theology by the great Protestant theologian Karl Barth, including reactions to Barth from scholars sympathetic to his approach and those more critical. In short, I believe that natural theology provides good arguments for the existence of God. What natural theology can tell us about God is, however, very limited. Traditionally it has been held that we need revelation to tell us more, indeed to tell us what it is most important to know. In this book I argue that the distinction between natural and revealed knowledge of God cannot be strictly maintained. This is because what is purported to be revelation needs to be evaluated using reason and evidence in like manner to that in which what nature may tell us about God is evaluated. One reason for this is that there are competing revelations which say contradictory things about God, so critical evaluation is needed to choose between them. There has thus arisen a distinction between ‘bare’ or traditional natural theology and ‘ramified natural theology’ which argues for the particularities of religious belief, and indeed the focus of such work has been on the particularities of Christianity. My first foray into this territory was to write an article for a special issue of the journal Philosophia Christi dedicated to the topic of ramified natural theology. This book constitutes my further, more detailed work on the subject. It will be apparent that I  owe a great debt of gratitude in producing this book to many distinguished scholars. Richard Swinburne is clearly a major influence. I had the privilege of attending his lectures and seminars at Oxford while reading theology and training for ministry, and he has been a support and inspiration to me since then. I was also privileged to be taught by Alister McGrath whom I admire greatly, although my views on natural

x  Preface theology have diverged from his. John Polkinghorne has been an inspiration and mentor over many years. Although I was not taught by him, it will be clear that my take on New Testament scholarship has been very greatly influenced by Tom Wright. Similarly the late Wolfhart Pannenberg has been a considerable influence. Interaction with my colleagues at The Faraday Institute for Science and Religion has been invaluable since the Institute was founded in 2006, and I am grateful that my work was supported in earlier years by the Templeton Foundation and Templeton World Charity Foundation. I am grateful to Joshua Wells and Yuga Harini of Routledge for their help and guidance since the project’s inception. Finally, as ever, my wife Shirley has been the most wonderful support and encourager throughout.

Acknowledgements

I am grateful to the following for permission to republish articles or portions of chapters, with some editing and supplementary material, from the following: • Oxford University Press for Rodney D. Holder, ‘Hume on Miracles: Bayesian Interpretation, Multiple Testimony, and the Existence of God,’ British Journal for the Philosophy of Science 49 (1998), 49–65. • Templeton Press for extracts from Rodney D. Holder, The Heavens Declare: Natural Theology and the Legacy of Karl Barth (West Conshohocken: Templeton Press, 2012). • Oxford Publishing Limited for Rodney D. Holder, ‘Natural Theology in the Twentieth Century,’ in Russell Re Manning (ed.), The Oxford Handbook of Natural Theology (Oxford: Oxford University Press, 2013), 118–134. • The editor of Philosophia Christi for Rodney Holder, ‘Why We Need Ramified Natural Theology.’ This essay first appeared in Philosophia Christi 15(2) (2013). More information can be found about Philosophia Christi at www.epsociety.org. • Taylor and Francis for Rodney D. Holder, ‘Explaining and Explaining Away in Cosmology and Theology,’ Theology and Science 14(3) (2016), 234–255.

1 Theology as a scientific discipline The place of natural theology and ramified natural theology

Introduction: theological rationality This book aims to show that religious belief, and in particular Christian belief, is rational in a way similar to that in which science is rational. Traditionally knowledge of God has been seen to come from two sources, nature and revelation. Arguments for the existence of God have come under the domain of ‘natural theology’ and have often utilized evidence from scientific findings about nature. My position is that inferential reasoning from the data towards a divine mind behind nature can be construed in a similar way to that by which scientific hypotheses are evaluated on the basis of empirical data. Natural theology, however, only gets one so far—to a creator God behind the universe and its laws. To get to the specifically Christian God one needs to go further and to engage in ‘ramified natural theology,’ a term introduced by Richard Swinburne.1 Ramified natural theology is the providing of arguments for the specific claims of Christianity. For this it is not enough to accept what is claimed as revelation without further ado; one needs to provide reasons for acceptance. In particular one needs to take account of Biblical scholarship. Thus the division between knowledge of God from nature and knowledge of God from revelation is an artificial one, since what is claimed to be revealed also needs evaluation using similar modes of rational argument. The German theologian Wolfhart Pannenberg, whom we shall meet again in these pages, makes this very point. Pannenberg is adamant that theology is dealing with truth, but there is a job of work to be done: ‘dogmatics may not presuppose the divine truth which the Christian doctrinal tradition claims. Theology has to present, test, and if possible confirm the claim.’2 We shall follow through this programme, utilizing Bayesian confirmation theory to clarify the meaning of the term ‘confirmation.’ The idea that theology is indeed a science is the major theme of Pannenberg’s Theology and the Philosophy of Science.3 Pannenberg is concerned that theology is only eligible for a place in the university if it is a science. This approach contrasts with Friedrich Schleiermacher, who thought it

2  Theology as a scientific discipline enough for theology to be essential for the education of clergy. This is inadequate; theology must be there for what it is, for its pursuit of truth, and it cannot be immune to criticism.4 Importantly Pannenberg emphasizes that truth is universal: ‘My truth cannot be mine alone. If I  cannot in principle declare it to be truth for all—though perhaps hardly anyone else sees this—then it pitilessly ceases to be truth for me also.’5 Christian thinkers have been concerned from the beginning to demonstrate the truth of Christianity, says Pannenberg. He writes: ‘This same concern, to defend the truth of Christianity by generally accepted criteria, has been present since the thirteenth century in the argument about the scientific status of theology and its right to be included among the sciences taught in a university.’6 If theology were to lose its place in the university that would be a problem not just for theology itself, even if the church took over the training of its own ministers. It would be a problem for truth in general, since the collaboration of theology and philosophy is essential for integrating the various scientific disciplines into a united body of truth, as indeed systematic theology unites the sub-disciplines of theology itself. Thus, for Pannenberg, Christian theology must argue with atheism on the grounds of a shared rationality. Theological truth claims are made on the basis of universal and publicly accessible evidence. For Pannenberg, as in the present work, this evidence is primarily historical. He argues that history, which resembles the natural sciences in some, though not all, respects, can only be understood through Jesus Christ who has revealed the end of history proleptically through his resurrection. The resurrection of Jesus tells us who he is, confirming his divine nature. Importantly this is an objective ‘Christology from below,’ in contrast to that of nineteenth century theologians such as Schleiermacher for whom we merely reflect subjective human experience or feelings of redemption back onto Christ. Whereas older views preceded historical criticism of the Bible, it also needs to be recognized that with the Enlightenment and historical criticism naturalistic assumptions were imported into theology. Thus we need to move forward and reconsider the evidence for Christian claims—and supremely that for the resurrection which in turn tells us who Jesus is—in the light of these factors, and this is what we set out to do in this book. In the English-speaking world, John Henry Newman had made similar points to Pannenberg in the nineteenth century. Newman espoused the benefits of a multi-disciplinary university, teaching all branches of knowledge. Although the university should not be under church control, Newman saw it necessary to include theology as a subject precisely because it is a branch of knowledge, indeed a branch which ‘to say the least, is as important and as large as any of them.’7 He writes: Is not the being of a God reported to us by testimony, handed down by history, inferred by an inductive process, brought home to us by

Theology as a scientific discipline 3 metaphysical necessity, urged on us by the suggestions of our conscience? It is a truth in the natural order, as well as in the supernatural.8 Moreover, theology undergirds every other branch of knowledge: Admit a God, and you introduce among the subjects of your knowledge, a fact encompassing, closing in upon, absorbing, every other fact conceivable. How can we investigate any part of any order of Knowledge, and stop short of that which enters into every order? All true principles run over with it, all phenomena converge to it; it is truly the First and the Last.9 Newman concludes his chapter on theology as a branch of knowledge with these words: Religious doctrine is knowledge, in as full a sense as Newton’s doctrine is knowledge. University Teaching without Theology is simply unphilosophical. Theology has at least as good a right to claim a place there as Astronomy.10 Whether this assessment of theology as the pursuit of truth, in like manner to the sciences, is valid will I trust become clearer in what follows in the rest of this book. We return to Pannenberg in particular, for whom history is of supreme importance and resembles natural science, although it can only be understood through Jesus Christ, briefly in Chapter 2 on natural theology, and then more particularly when we consider the concept of ramified natural theology in Chapter 4.

Outline of the book In the book I begin (Chapter 2) with a survey of natural theology, looking at the historical development of this discipline, from St  Thomas Aquinas up to the present day. I look at how natural theology changed from general arguments about the existence and ordered structure of the universe to arguments from the particular in the so-called ‘physico-theology.’ I discuss some of the challenges made to natural theology, notably from David Hume and Immanuel Kant, yet how it continued in the physico-theological form at least up until William Paley. I note how Paley was more subtle than is often realized since he also used more general arguments. Whilst Darwin’s theory of evolution by natural selection is often portrayed as defeating natural theology, I show how figures such as Charles Kingsley and Frederick Temple welcomed evolution, which, to them, enhanced our view of God’s majesty in creation. Natural selection did not negate design but merely shifted the mode in which the design was executed. I bring us up to date with a look at three central figures who have contributed to natural theology today: Alister

4  Theology as a scientific discipline McGrath, John Polkinghorne, and Richard Swinburne. The last of these has pioneered the use of Bayesian confirmation theory in making arguments both of natural theology and ramified natural theology and I  introduce the theory at this point. It should be emphasized that this mathematical framework does no more than codify the judgments we make in our normal modes of rational argument (e.g. certain evidence is more likely to pertain if one hypothesis is true rather than another). I devote a chapter to modern cosmology (Chapter  3), which has provided a new impetus to natural theology, reviving traditional arguments but expressing them in new ways. An important question is whether we need natural theology to get us off the ground by providing arguments for the existence of God before we proceed to arguments for the particulars of the Christian faith. This is because, if we can show with reasonable probability that God exists, such claims as that Jesus performed miracles or rose from the dead become less improbable a priori. Modern cosmology raises two significant questions which, at least potentially, lead to arguments of natural theology: does the universe have a beginning, and why is the universe so apparently fine-tuned for life? In a significant paper Mark McCartney and David Glass of the University of Ulster ask whether science can explain away, or at least explain away in part, such features of the universe in cosmology and other sciences which may alternatively invite a theological explanation.11 Theology would then be at best redundant in this area of explanation and appeal to cosmology for justification of belief in God would be nullified. In this chapter I argue that two proposals made by cosmologists fail to explain away the universe’s beginning, and that science is powerless to explain away the more fundamental question as to why there is a universe at all. I argue similarly that scientific, or quasi-scientific, proposals such as the multiverse fail to explain away the fine-tuning. Indeed, I argue that theism is the most rational position to adopt to explain such features. The arguments here are powerful ones of natural theology, which, as I have indicated, is preliminary to ramified natural theology and lends credence to the latter. The chapter is based on my contribution to a project I was involved in with McCartney and Glass, the results of which were published in a special issue of Theology and Science.12 Importantly, however, the chapter is supplemented by a Bayesian analysis of each argument at the end. Of course, if the existence of the universe enhances the probability that God exists then the fact that the universe is fine-tuned will enhance it further, so we have the beginnings of a cumulative argument. If we opt for a multiverse then that is even less probable a priori than a single universe, besides which it does not solve the fine-tuning problem. Later chapters on ramified natural theology will also argue that several pieces of evidence can be put together to make an overall strong case for specifically Christian theism. The important point is that this example of natural theology gives powerful evidence for the existence of God and the same methodology can be used in ramified natural theology.

Theology as a scientific discipline 5 I devote a chapter to answer the specific question of why we need ramified natural theology (Chapter 4). Not only can we not take for granted what revelation tells us, but there are also competing revelations, so the context of religious pluralism becomes important. Thus what is purported to be revelation must be examined utilizing commonly accepted modes of rational discourse. Since there are many competing claims to truth, and they contradict each other, not all of them can be true. I thus now come to the point of directly challenging the notion that natural knowledge and revealed knowledge can be kept rigidly apart. On the one hand, what religions agree about may be seen as strengthening the case for theism: arguments from natural theology may have commonality with other religions. On the other hand, ramified natural theology is necessary if we are to resolve competing claims when it comes to the particularities. In this enterprise, because Christianity is nothing if not historical, historical science will be of vital importance, so we are very much echoing Pannenberg’s concerns in this chapter. In this chapter I briefly compare the problem of competing religious claims with that of competing theories in science which are difficult to resolve. Thus the steady-state theory and the Big Bang theory were direct competitors until empirical evidence was discovered which favoured the latter. Another example concerns differing interpretations of quantum theory, such as the Copenhagen interpretation which is indeterministic and the Bohmian pilot wave interpretation which is deterministic. Interestingly these interpretations are indistinguishable empirically. I  go on to discuss how much one needs to know about other religions in order to justify Christian claims. Thus one might not need to know too much if either of two criteria is met: the claims of other religions for an incarnation of God are not there and the arguments for Christian claims are in any case very strong. Nevertheless, I shall also include some brief discussion of Islam here and, especially with reference to miracles, in a subsequent chapter. Whilst I argue that one should not assume the truth of Christian claims at the outset, as some theologians such as Thomas Torrance and Alister McGrath do, I also strongly argue that such claims should not be excluded by fiat as some liberal Protestants such as Ernst Troeltsch in the nineteenth century did. I next discuss (Chapter 5) two important historical precursors of ramified natural theology, namely Blaise Pascal and Bishop Joseph Butler. A significant difference between these two eminent thinkers is that Pascal rejects natural theology but offers vigorous arguments, notably from miracles and prophecy, for the truths of Christianity. Butler accepts the arguments of natural theology as a starting point for an equally stout defence of specifically Christian claims, also based on purported miracles and prophecies. Probability is important for both thinkers, famously with Pascal’s wager but also for Butler for whom ‘probability is the very guide of life.’ Whilst both these thinkers are more sophisticated than sometimes they are given credit for, they are nevertheless of their time, and we need to fast

6  Theology as a scientific discipline forward to the present day. For Butler writing in the first half of the eighteenth century the deliverances of natural theology would be commonly agreed with his opponents the deists. That would be in line with our concern to begin with commonly accepted criteria of rationality. However, those results are not commonly accepted today so we need to go back and make these arguments again in a convincing way for our contemporaries. Moreover, both Pascal and Butler precede the rise of historical criticism of the Bible so what they say also needs to be revised with this in mind. Scottish Enlightenment philosopher David Hume is often taken today to have refuted the idea that we can believe in miracles, understood as ‘violations of the laws of nature.’ Whilst Hume’s definition of miracle may not be that of Scripture or of theologians, in Chapter 6 I nevertheless begin from this definition and argue that Hume has not shown that it is irrational to believe in miracles. I do this by showing that however initially implausible an account of a miracle may be, the accumulation of sufficient evidence of testimony can make its occurrence probable. Furthermore, the accumulation of evidence for many miracles can make it probable that at least some accounts were veridical. The occurrence of suitable miracles could be used as evidence for the existence of God and I  examine the circumstances in which it does so. On the other hand, as above, other arguments for the existence of God could enhance the prior probability of miracles before we take account of the evidence of testimony to them. In Chapter 7 I begin serious engagement with the evidence for Christian claims about Christ, and so I embark on modern ramified natural theology in earnest. This chapter deals especially with the Biblical accounts of Jesus’ purported miracles. I begin with a brief survey of historical criticism of the Bible as it arose in Germany in the late eighteenth century. Often this has been correlated with naturalistic assumptions coming out of the Enlightenment, with the result that one finds a rather bland Jesus who, as one commentator put it, merely reflects the liberal Protestant ambience from which this figure arose. There has always been a more orthodox criticism of the more outlandish claims and, as I show, a much more optimistic view, of the historicity of the gospels in particular, in the last fifty years or so. Engaging with modern scholars I show that the historical evidence is overwhelmingly in favour of Jesus performing miracles. Brief consideration of what Islam and Judaism say only goes to reinforce that point rather than undermine it, and indeed there is a total lack of counter-evidence. Chapter  8 builds on the foregoing through an examination of Biblical prophecy ostensibly fulfilled in Jesus. This, perhaps more so than the case of miracles, needs nuancing in the light of modern understanding of what prophecy actually is in the Old Testament. Thus it is not in the first instance prediction of an event in the far future by an all-seeing prognosticator. Rather it is very often speaking truth to power and with a message for the immediate context. However, as a number of scholars I engage with point out, that does not mean that there may not be a future reference, even if

Theology as a scientific discipline 7 not understood, or only dimly understood, by the prophet. If the prophet has some real insight into the mind of God, and understanding of the state of the world, it may be that an immediate and partial fulfilment will be followed many years later by a more complete consummation. C. S. Lewis is helpful here in drawing a secular parallel with Plato’s insight into the way a perfectly righteous man would be treated. Having made this point I go on to refer to the many ways in which Jesus is seen to fulfil prophecy in the gospels, not just in very particular ways through the circumstances of his death and purported resurrection but much more broadly in his life and teaching. The most important miracle of all for Christians, indeed the sine qua non of the Christian faith, is the resurrection of Jesus Christ from the dead. Chapter 9 looks at the evidence for this in detail, both in the gospels and, even more importantly, in St Paul’s first letter to the Christians in Corinth. Indeed 1 Corinthians 15 enables us to track back to within just a very few years of the purported event. However, I also look briefly at literature outside the New Testament. Again, there is no contradictory evidence but even some hints from non-Christian sources that the New Testament accounts may be veridical. Within the Christian fold the evidence of early second century writer Papias is particularly significant. I note that the gospels themselves give substantial evidence of the empty tomb and of Jesus’ appearances alive after his death, and that, while there are discrepancies between them, this is only to be expected of accounts which are independent and not the result of a conspiracy. Particularly telling is the evidence of women, but there are numerous strands of evidence which, when combined, make a compelling case for the veracity of what is claimed. Chapter 10 again sees ramified natural theology in action, now framing the argument in Bayesian terms. The resurrection of Jesus Christ is a priori improbable, yet can be rendered credible by an examination of the historical evidence, as will have emerged from the preceding chapters. Bayesian confirmation theory, which has been introduced and deployed in earlier chapters, is an invaluable tool for evaluating competing hypotheses, which each explain the evidence in question, in a rigorous way. However, authors utilizing Bayesian confirmation theory to evaluate the evidence for the resurrection differ in approach, in particular by requiring or not requiring the input of a high probability of theism into the analysis. Thus Timothy and Lydia McGrew only compare ‘Bayes factors,’ otherwise known as likelihood factors, which measure how much more likely the evidence is to obtain on one hypothesis compared with another; they disregard prior probabilities.13 Richard Swinburne goes all the way to derive a posterior probability for the truth of God incarnate, for which it is necessary also to take account of prior probabilities.14 Crucially, for him, the prior probability he uses for God becoming incarnate takes account of the deliverances of natural theology, so bare theism is already at least moderately likely. The McGrews and Swinburne also vary widely in the results they derive. This chapter reexamines the issues involved and suggests a resolution to the conflicting

8  Theology as a scientific discipline methodologies and results. It shows how feeding natural theology into ramified theology can make the end result highly probable. A lot of evidence will by now have been amassed. Hence in Chapter 11 I bring all the evidence and arguments of the preceding chapters together to build a cumulative case for the fundamental Christian claim that God became incarnate in Jesus Christ. I  follow Swinburne in including input from natural theology to enhance the prior probability of this claim, i.e. the probability before the historical evidence is taken into account. I then show that the combined evidence for Jesus having performed miracles, his fulfilling Messianic prophecies and supremely his resurrection, makes it overwhelmingly probable that he was indeed God incarnate.

Notes 1 Richard Swinburne, ‘Natural Theology, its “Dwindling Probabilities” and “Lack of Rapport,” ’ Faith and Philosophy 21(4) (2004), 533–546, 533. 2 Wolfhart Pannenberg, Systematic Theology, vol. 1, trans. Geoffrey W. Bromiley (Edinburgh: T & T Clark, 1991), 50. 3 Wolfhart Pannenberg, Theology and the Philosophy of Science, trans. Francis McDonagh (London: Darton, Longman & Todd, 1976). (First German edition, Wissenschaftstheorie und Theologie, Frankfurt am Main, Suhrkamp, 1973). A helpful review is given by Wentzel van Huyssteen, ‘Theology as the Science of God: Wolfhart Pannenberg,’ in his Theology and the Justification of Faith: Constructing Theories in Systematic Theology (Grand Rapids: Wm. B. Eerdmans, 1989), 71–100. 4 Pannenberg, Theology and the Philosophy of Science, 255; also Systematic Theology, vol. 1, 7. 5 Pannenberg, Systematic Theology, vol. 1, 51. 6 Pannenberg, Theology and the Philosophy of Science, 13. 7 John Henry Newman, The Idea of a University Defined and Illustrated: I. In Nine Discourses Delivered to the Catholics of Dublin; II. In Occasional Lectures and Essays Addressed to the Members of the Catholic University (Oxford: Oxford University Press, [1852, 1889] 1976), 33. 8 Newman, Idea of a University, 38. 9 Newman, Idea of a University, 38. 10 Newman, Idea of a University, 50. 11 David H. Glass and Mark McCartney, ‘Explaining and Explaining Away in Science and Religion,’ Theology and Science 12(4) (2014), 338–361. 12 Rodney D. Holder, ‘Explaining and Explaining Away in Cosmology and Theology,’ Theology and Science 14(3) (2016). 13 Timothy McGrew and Lydia McGrew, ‘The Argument from Miracles: A Cumulative Case for the Resurrection of Jesus of Nazareth,’ in William Lane Craig and J. P. Moreland (eds.), The Blackwell Companion to Natural Theology (Chichester: Wiley-Blackwell, [2009] 2012), 593–662. 14 Richard Swinburne, The Resurrection of God Incarnate (Oxford: Oxford University Press, 2003). Swinburne’s argument is summarized in Richard Swinburne, ‘The Probability of the Resurrection of Jesus,’ Philosophia Christi 15(2) (2013), 239–252; and, less technically in Richard Swinburne, Was Jesus God? (Oxford: Oxford University Press, 2008).

2 Reason and religion The role of natural theology

Introduction In this book I  argue that theology is a rational pursuit, just as science is. Moreover, I  am convinced that theology needs to justify its truth claims, in like manner to the way in which science does. The discipline of natural theology aims to do precisely that, by providing arguments for the existence of God which depend only on our common humanity and not on any putative revelation. In this chapter I give an overview of natural theology from Thomas Aquinas to the present day. There will be some discussion of the objections to the enterprise from Hume and Kant. I shall argue that William Paley is more subtle than often portrayed, implicitly answering some of Hume’s criticisms. Thus he sees a need for explanation of the existence of watches even if watches come from other watches, an argument reminiscent of Leibniz’s argument concerning an infinite sequence of geometry books with each copied from the preceding one ad infinitum (a point that will recur rephrased in Chapter 3—an infinite chain of causes needs an explanation). Paley also produced general as well as particular arguments for design, and I  shall argue that natural theology did not die with Darwin. Rather, as Archbishop Frederick Temple averred, what changed with Darwin was not the fact of design, but merely the mode by which the design was executed. I shall point out how the position of such figures as Temple and Charles Kingsley are consonant with the classical position on creation of St Augustine. Today arguments from such general features of the universe as its orderly nature open to human enquiry and the fine-tuned character of its laws offer much better arguments than appeal to individual organisms or structures within nature, for which we should seek scientific explanation. In biology, the phenomenon of convergent evolution can be invoked, whereas so-called ‘Intelligent Design’ fails. As Richard Swinburne says, scientific explanations are not competitors to theological ones; rather, theism explains why science explains. Swinburne is one of three present-day proponents of natural theology whom I discuss in the latter half of this chapter, the others being theologian Alister McGrath and scientist-theologian John Polkinghorne.

10  Reason and religion

The knowledge of God and natural theology1 When it comes to knowledge of God, it has traditionally been thought that there are two kinds of such knowledge: natural knowledge which is what we know about God purely by being human and the knowledge of God which is given through God’s special revelation. The former comes under the domain of ‘natural theology,’ and a typical way of understanding this term is given by James Barr: Traditionally “natural theology” has commonly meant something like this: that “by nature”, that is, just by being human beings, men and women have a certain degree of knowledge of God and awareness of him, or at least a capacity for such an awareness; and this knowledge or awareness exists anterior to the special revelation of God made through Jesus Christ, through the Church, through the Bible.2 Natural theology is an area of intellectual enquiry with a long, if chequered, history dating back at least to the era of classical Greek thought. Within Christian theology the expression theologia naturalis seems to have been first used by St Augustine, in critiquing the insights of classical philosophers, especially Plato who came closest to Christianity.3 St Thomas Aquinas, for example, thought we could know that God exists from human reason alone: The truths about God which St Paul says we can know by our natural powers of reasoning—that God exists, for example—are not numbered among the articles of faith, but are presupposed to them.  .  .  . God’s effects, therefore, can serve to demonstrate that God exists, even though they cannot help us to know him comprehensively for what he is.4 Hence, for Aquinas, we can know that God exists but we cannot know God in himself unless he reveals himself to us. And the Christian revelation informs us of what we could not know otherwise, that God is Trinity: Father, Son, and Holy Spirit. At the time of the European Reformation John Calvin made a similar distinction. Each human being possesses what he calls a sensus divinitatis, a sense of the divine. And the heavens are the theatre of God’s glory. As Psalm 19 says, ‘The heavens declare the glory of God.’ However, for Calvin as for Aquinas the far more important knowledge of God, simply than knowing that there is a Creator, is that which is specially revealed in Scripture, for it is the knowledge of God as Redeemer, in Christ, which secures our salvation.5 At the time of the scientific revolution, the philosopher Francis Bacon spoke about these two kinds of knowledge given in two kinds of book, the book of God’s works in nature and the book of God’s word in Holy Scripture. Science, known at the time as natural philosophy, is the study of the

Reason and religion 11 former and theology the study of the latter, and, as he says here, one cannot go too far in the study of either: To conclude, therefore, let no man out of a weak conceit of sobriety, or an ill-applied moderation, think or maintain, that a man can search too far or be too well studied in the book of God’s word, or in the book of God’s works; divinity or philosophy; but rather let men endeavour an endless progress or proficience in both.6 Natural knowledge of God has been the subject of natural theology, and could be construed as an immediate impression of God’s existence, power and majesty, coming from simply gazing in awe at the heavens. However, in practice natural theology has been about providing reasons and arguments as to why anybody might believe in God. In that sense it fulfils the requirement of John Locke (1632–1704) for faith to justify itself: ‘Faith is nothing but a firm Assent of the Mind: which if it be regulated, as is our Duty, cannot be afforded to anything, but upon good Reason.’7 In the century following Locke’s Essay David Hume (1711–1776) made a similar demand: ‘A wise man, therefore, proportions his belief to the evidence.’8 Twentieth century theologian Wolfhart Pannenberg’s understanding of faith chimes in nicely with these philosophers’ demands for evidence. Pannenberg writes: ‘a person does not come to faith blindly, but by means of an event that can be appropriated as something that can be considered reliable. True faith is not a state of blissful gullibility.’9 For Pannenberg, the main point is that God has acted in history through the life, death, and resurrection of Jesus Christ and we have publicly accessible evidence, in the form of witness testimony to those events, which can therefore be judged reliable. I agree with Pannenberg about this, and I shall examine the evidence for miracles, and for the resurrection in particular, in later chapters, since this takes us into the realm of ramified natural theology. Suffice it to say here that my own way of thinking about faith is in terms of a relationship of trust, like marriage, rather than in terms of belief without evidence. I have faith in my wife. When we married I had a certain amount of evidence and that has grown over the years, though it would be a strange kind of marriage in which I deliberately kept doing experiments to find out if she was still trustworthy. Similarly there is evidence for Christian belief, and that is important when it comes to sharing the faith with others—for having a reason for the hope that is in us, as Peter says (1 Pet 3:15), but we do not continue putting God to the test all the time. However, it is true that some Christians open themselves up to the charge of being immune to argument and not being interested in evidence, whereas in fact there is plenty of good evidence on which to base one’s faith, as will transpire. Examples of natural theology in practice include the classical cosmological and design arguments. The cosmological argument says that everything

12  Reason and religion which exists has a cause for its existence. Therefore there is a cause for the universe’s existence. The design argument appeals to the ordered structure of the universe as requiring explanation. The merits of different forms of these arguments have been the subject of philosophical debate for centuries and we shall return to them.

Natural theology: from Aquinas to the present day From Aquinas to William Paley in the nineteenth century there was a subtle shift in natural theology. Aquinas gave general arguments, whereas the scientific revolution brought in arguments based on the particular. Thus it was that particular structures within the natural order, rather than order per se, were seen as exhibiting the signs of design. This was the ‘physico-theology’ of such figures as Boyle, Hooke and John Ray, indeed many founders of the Royal Society. For example, the intricate detail of the eye of a fly shown under a microscope, illustrated in Hooke’s Micrographia, far outshone that of a human artifact such as a nail, which showed rough edges. However, arguments of this kind were often combined with an appeal to more general features, too, such as the law-like behavior of the natural world that the sciences were uncovering. The early Boyle Lectures gave expression to much natural theology at the beginning of the eighteenth century. In the eighteenth century natural theology came under fierce attack by Enlightenment philosophers Immanuel Kant and David Hume. Kant noted three forms of argument: the ontological, cosmological, and physicotheological (design). However, as Richard Swinburne notes, he failed to distinguish different forms of these. Kant said that we can only experience phenomena; we cannot experience ‘noumena,’ things ‘as they are in themselves.’ And we impose order on the phenomena via the structures of our minds. It is human mind-imposed order rather than real order that we can know nothing about. It is thus not ‘God-imposed’ order, so the design argument of natural theology is apparently undermined. However, scientists in particular find, contrary to Kant, that the empirical reality of the world does seem to force itself on us. Prime examples would be the totally counterintuitive phenomena described by quantum theory, and the four-dimensional curved space-time of general relativity. Indeed, Kant has been shown to be in error in claiming the necessity of describing the world through Euclidean geometry because this is embedded in the structure of our minds.10 He is also in error in arguing that the cause of an observable event must itself be observable, and therefore God must be observable if he is the cause of what we observe: ‘If the empirical law of causality is to conduct us to a Supreme Being, this Being must belong to the chain of empirical objects.’11 As Richard Swinburne points out,12 perhaps unlike their forebears in the eighteenth century, modern day scientists are quite used to inferring unobservable causes from their effects, e.g. the Higgs boson as just the latest unobservable sub-atomic particle.

Reason and religion 13 Hume made a number of criticisms of natural theology, particularly the argument from design.13 For example, he pointed to the lack of analogy between mechanical artifacts pointing to a human designer and natural phenomena supposedly pointing to God. A house points to an architect or builder because we have many instances of existing houses. Natural phenomena are not like houses in that respect, and one needs a very close analogy to make the inference to design. To transfer an inference about parts of the universe exhibiting design to the whole doing so would be disproportionate: there is no similarity between a house and a universe.14 Moreover, since a cause ought to be proportioned to its effect, one cannot infer an infinite deity from anything finite. And then, why infer one god as the creator and not several deities, if the analogy is with a house or ship which was built by many men?15 Perhaps a better analogy for the universe, says Hume, is an animal. God would then be its soul (curiously an idea that some modern theologians have actually favoured, though it certainly departs from classical orthodoxy’s maintenance of the distinction between the creator and the universe).16 The animal analogy also leads him to a different idea for the cause of the universe, namely generation; an ostrich lays an egg in the sand to produce another ostrich, and so on. Intelligence and design do not at any rate constitute the only option.17 These arguments do not apply to modern natural theology, as formulated for example by Richard Swinburne, since present day arguments are not really arguments from analogy. Rather, they ask, what would we expect to see on the basis of theism? Is a universe more likely to exist if there is a God then if there is no God? Is the universe more likely to exhibit the character our universe does if there is a God rather than if there is no God? God is introduced as a hypothesis to be compared with alternative hypotheses, such as atheism, with regard to how probable it makes the evidence. Moreover, to introduce an arbitrary multiplicity of gods violates the Occam’s razor principle of not multiplying entities needlessly, and would be more likely to lead to different laws of nature in different parts of the universe, contrary to observation. Indeed, the fact that the same laws apply across the whole space-time history of our 13.8 billion year old universe with its present complement of 1022 galaxies is surely more to be squared with a single Creator God than a pantheon of gods beavering away at different times and in different places. We refer again to these important points when considering cosmology in Chapter 3. It is also false to say that we need many examples of a universe, like many examples of a house, to make inferences about the universe’s design. We only have one instance of an intelligent life-form able to speculate about origins and design, yet we make inferences about how human beings came into existence, having evolved from more primitive creatures. And we can certainly look at the space of possible universes and make estimates of their probable occurrence under different hypotheses. Indeed, in Chapter 3 I shall

14  Reason and religion do just that. And against another of Hume’s points, if one universe is generated from another, one still needs an explanation for the whole sequence of universes—indeed, again, the question is whether the whole sequence with its eventual production of our universe is more likely on a theistic or atheistic hypothesis. It is worth noting that Leibniz, writing in 1697, anticipated just this point, using the example of geometry books copied from other geometry books: Let us suppose that a book of the elements of geometry existed from all eternity and that in succession one copy of it was made from another, it is evident that although we can account for the present book by the book from which it was copied, nevertheless, going back through as many books as we like, we could never reach a complete reason for it, because we can always ask why such books have at all times existed, that is to say, why books at all, and why written in this way. What is true of books is also true of the different states of the world; for, in spite of certain laws of change, the succeeding state is, in some sort, a copy of that which precedes it. Therefore, to whatever earlier state you go back, you will never find in it the complete reason of things, that is to say, why there exists any world and why this world rather than some other.18 Historically, Hume’s arguments seem not to have precipitated the downfall of natural theology, since, even in its physico-theological form, it continued flourishing well into the first half of the nineteenth century. Indeed the high point came with William Paley’s 1802 volume entitled Natural Theology, or Evidence of the Existence and Attributes of the Deity, Collected from the Appearances of Nature.19 In this book, Paley gave the famous example of a watch found on a heath. The watch, which possessed great intricacy, was obviously designed. How much more so the eye observing it? Indeed, the works of nature possess an incalculable degree of contrivance compared with that of the watch. This form of the argument from design is generally considered to be undermined by Darwin, who explained how eyes, and complex biological organisms in general, came about through the long process of evolution. However, I was surprised recently to find that Siegfried Sassoon, one hundred years on, would still find Paley ‘an essential preliminary’ to his career at Cambridge.20 In any case Paley was more subtle than he is often given credit for nowadays. Thus he clearly considered various objections to the argument, some of which remind one of Hume. For example, he asked whether the imperfection or malfunctioning of the watch should affect our conclusion that it was designed. The answer is no, because the original intention could still be discerned. A particularly interesting question is whether the existence of the watch would be accounted for if we were told that it was ‘one out of possible combinations of material forms’ and whatever we found would have

Reason and religion 15 had to take some configuration or other. That kind of argument is sometimes advanced today in order to deny that the fine-tuning of the universe requires explanation. Paley dismisses it as not something ‘any man in his senses’ would think.21 I have argued that it fails in the case of the argument from fine-tuning of the universe both because there is special significance in the configuration of the universe we observe and because, in fact, we have a good explanation for this particular configuration.22 Another objection Paley considers is whether it would make any difference to the argument if one found that the watch could generate other watches (directly facing Hume’s objection, and echoing Leibniz). In the first place, that would enhance the impression of design, not lessen it, he says, because it is evidence of even greater contrivance. Secondly, while a watch might generate another watch, it is not the maker in the same sense as a carpenter is the maker of a chair. It is not the source of order and purpose in the new watch, which must still be sought from an overall designer. Paley writes, ‘There cannot be design without a designer; contrivance without a contriver; order without choice; arrangement, without anything capable of arranging; subserviency and relation to a purpose, without that which could intend a purpose. . . .’23 Even if our watch were produced by a preceding watch, that would not explain the contrivance of watches and their fitness for purpose. A backward sequence of causes does not explain the existence of the sequence itself nor why it is productive of design and purpose. The remark of Charles Kingsley about evolution (see later), that it is even wiser of God to ‘make things make themselves’ than to create them all separately, seems to be in line with what Paley says here. Paley’s thought also ranged wider than the design or ‘contrivance’ of particulars within the natural world, such as the individual organs of animals, which are indeed explained by Darwinian evolution even if evolution itself requires explanation. Thus he also considered the design of the laws of nature. In particular he realized that the stability of planetary orbits is owed to the inverse square law of gravitation. Other laws would have resulted in the planets either spiralling into the sun or flying away from it. When the very narrow range in which the law must fall for preservation of the system is compared with the infinite range of possibilities, the only reasonable conclusion to draw, says Paley, is that a ‘designing mind’ is responsible for the choice.24 To my mind this is the best of Paley’s arguments, and it is one that has lasted. It has been resurrected in the modern discussion of anthropic fine-tuning, with the added nuance that only if there are three (out of infinitely many possible) spatial dimensions is there an inverse square law and hence a stable solar system.25 But the focus on the design of physical law, rather than the design of individual features separately, is most significant here, since individual features might come about through the operation of laws. That evolution by natural selection, as discovered by Darwin, can be readily accommodated to Christian faith will be evident from a brief look

16  Reason and religion at how the Christian doctrine of creation has been understood by the early Church Fathers. Thus Gregory of Nyssa sees God implanting causal principles in the creation from the beginning: The sources, causes, and potencies of all things were collectively sent forth in an instant, and in this first impulse of the Divine Will, the essences of all things assembled together: heaven, aether, star, fire, air, sea, earth, animal, plant—all beheld by the eye of God . . . There followed a certain necessary series according to a certain order . . . as nature the maker (technike phusis) required . . . appearing not by chance . . . but because the necessary arrangement of nature required succession in the things coming into being.26 St  Augustine is better known for saying the same thing: these embedded causalities, ‘rationes seminales’ or seed-like principles, are present from the cosmic beginning, and each of them contains the potential for the later development of a specific living kind. Indeed Augustine uses a whole host of expressions to convey this idea: causales rationes, quasi semina futurorum, rationes primordiales, primordia causarum, rationes seminales, quasi seminales rationes, and sometimes simply rationes.27 As Ernan McMullin tells us, Gregory and Augustine saw this notion as implicit in the language of ‘bringing forth’ in Genesis 1. It is not that God creates the plants as a new, special creation but that ‘God said, “Let the earth put forth vegetation, plants yielding seed, and fruit trees bearing fruit in which is their seed, each according to its kind, upon the earth.” And it was so. The earth brought forth vegetation, plants yielding seed according to their own kinds, and trees bearing fruit in which is their seed, each according to its kind.’ (Gen. 1:11–12). That Darwin did not deal the death blow to natural theology can be seen from the reaction of significant Anglican clerics to The Origin of Species. Charles Kingsley wrote a letter to Darwin on 18 November 1859 on receiving a complimentary copy of Origin prior to its publication a few days later on 24 November. In it he wrote: All that I have seen of it awes me; both with the heap of facts, & the prestige of your name, & also with the clear intuition, that if you be right, I must give up much that I have believed and written . . . I have gradually learned to see that it is just as noble a conception of Deity, to believe that he created primal forms capable of self development into all forms needful in tempore & pro loco, as to believe that He required a fresh act of inter-vention to supply the lacunas which he himself had made. I question whether the former be not the loftier thought.28 It may be that if Kingsley had been more familiar with Augustine, he would not have had to give up so much. Be that as it may, even clearer and more

Reason and religion 17 positive is Kingsley’s statement in his lecture of 1871 entitled ‘The Natural Theology of the Future’: ‘We knew of old that God was so wise that He could make all things; but behold, He is so much wiser than even that, that He can make all things make themselves.’29 Indeed Alister McGrath sees the ‘neat slogan’ about God choosing ‘to make all things make themselves’ as ‘nothing more than a creative reworking of Augustine’s notion of rationes seminales.’30 Frederick Temple, who was consecrated Archbishop of Canterbury in 1896, could, like Kingsley, see that the design argument of natural theology was still valid, though revised from the form it took in William Paley. Echoing Kingsley he wrote: What is touched by this doctrine [of Evolution] is not the evidence of design but the mode in which the design was executed . . . In the one case the Creator made the animals at once such as they now are; in the other case He impressed on certain particles of matter . . . such inherent powers that in the ordinary course of time living creatures such as the present were developed . . . He did not make the things, we may say; no, but He made them make themselves.31 Kingsley too realized that this means that natural theology is alive and well, just as it was in Paley; it simply needs to be expressed differently. And here is what he says about this, with very strong echoes of Augustine: All, it seems to me, that the new doctrines of Evolution demand is this. We all agree, for the fact is patent, that our own bodies, and indeed the body of every living creature, are evolved from a seemingly simple germ by natural laws, without visible action of any designing will or mind, into the full organisation of a human or other creature. Yet we do not say, on that account: God did not create me; I only grew. We hold in this case to our old idea, and say: If there be evolution, there must be an evolver. Now the new physical theories only ask us, it seems to me, to extend this conception to the whole universe: to believe that not individuals merely, but whole varieties and races, the total organised life on this planet, and it may be the total organisation of the universe, have been evolved just as our bodies are, by natural laws acting through circumstance.32 Here Kingsley is prescient. Modern cosmology has discovered how indeed the universe from its very beginning at the Big Bang has in a real sense been ‘pregnant with life.’ Both the initial conditions at the earliest time we can sensibly speak of and the fundamental constants of physics seem as it were ‘pre-set’ so that the universe could bring forth life at some stage in its evolution. This is a large topic, and we reserve fuller discussion to Chapter 3, yet it is worth noting here that this so-called ‘fine-tuning’ of the universe is also

18  Reason and religion entirely consistent with Augustine’s notion of seed-like principles.33 Indeed Augustine’s concept provides the theological lens through which, in Part 2 of A Fine-Tuned Universe, McGrath explores how ‘the cosmos appears to have come into existence with the potentiality for human existence.’34 The fine-tuning is fundamental to this, but McGrath also cites other evidence such as the life-conducive chemical properties of water, and the phenomena of convergence (signifying apparent directionality) and emergence (of new irreducible laws and properties) in nature. Further evidence that Darwin did not deal the death blow to natural theology is provided by the inauguration in 1888 of the Gifford Lectures in natural theology in Scottish universities, thanks to a handsome bequest from Lord Gifford.35 However, these lectures brought mixed blessings, especially for those who were concerned with theological orthodoxy and the centrality of God’s revelation in Jesus Christ, most notably perhaps for the Swiss Protestant theologian Karl Barth. Lord Gifford had stated in his will that the lectures established in his name were for ‘Promoting, Advancing, Teaching, and Diffusing the study of Natural Theology.’36 The lecturers could be of any religion or denomination or none and they should ‘treat their subject strictly as a natural science, the greatest of all possible sciences, indeed, in one sense the only science, that of Infinite Being, without reference to or reliance upon any supposed special exceptional or so-called miraculous revelation.’37 The lecturers should, furthermore, be under no restraint and should be able to discuss ‘all questions about man’s conceptions of God or the Infinite’ including ‘whether he can have any such conceptions.’38 In light of the above it is no surprise that the lectures and lecturers have been very diverse. Thus William James, while attacking materialism, argued for ‘pluralism,’ i.e. the fragmentation of the universe, and acknowledged that his study of ‘the varieties of religious experience’ could lead merely to polytheism. Many other early lecturers had scant regard for the orthodoxy of the Church, being sceptical about miracles, or espousing a less than personal God (say as G. W. F. Hegel’s ‘Absolute Spirit’), or seeing Jesus as just a special human being. Anthropological accounts of the origin of religion, most famously expressed by J. G. Frazer, could also be troubling to traditional Christian belief. Ultimately religion could be explained as evolved from magic, with science now superseding both.39 Perhaps, in view of all this, it is no wonder that Barth stormed in to give a lecture series denouncing natural theology altogether! He did that by choosing to speak, conveniently for his purpose, on the Scottish Confession of 1560, rather than the later Westminster Confession of Faith which is more positive for natural theology. Of course, while natural theology may not say much about the particulars of Christian doctrine, it was Barth’s concern that so many saw reconciliation with science effectively as the capitulation of theology to science and necessitating abandonment of basic doctrines. On the other hand some Gifford lecturers have espoused a traditional faith and re-presented traditional arguments. This has happened mainly in the

Reason and religion 19 post-Second World War period, though there is still no discernible unity or agreement. On the positive and orthodox side would be Austin Farrer, Stanley Jaki, E. L. Mascall, and Richard Swinburne, whose work has become central to the modern discussion. Yet the atheist logical positivist Freddie Ayer is also a contributor, maintaining the extreme diversity of the earlier period. For many proponents, from Aquinas to the present day, natural theology has been a preliminary for revealed theology. In my own work, I  see the task of natural theology as removing barriers to belief in God and providing good reasons for belief. It provides the groundwork for the more specific and important belief in Jesus Christ as Lord and Saviour. As I shall explain, however, I believe that the division between natural and revealed theology is a somewhat artificial one, especially because I  think we need to justify accepting what is purported to be revelation, ‘to give a reason for the hope that is within us’ as St Peter puts it (1 Pet 3:15).

Three modern advocates of natural theology40 In the following pages I briefly outline the approach to natural theology of three significant contributors to the modern discussion. Alister McGrath, heavily influenced by Scottish theologian Thomas Torrance, who in turn was heavily influenced by Karl Barth, sees natural theology as part of dogmatic theology, only telling us about the God who is already known. John Polkinghorne, a major contributor to the science-religion dialogue, agrees with McGrath that natural theology belongs within dogmatic theology, but sees it as saying something about what we know of God apart from revelation. He advocates a revivified but limited form of natural theology utilizing his concept of ‘motivated belief.’ Richard Swinburne adopts a cumulative case approach and deploys the analytic framework of Bayesian confirmation theory. Moreover, for Swinburne, the probabilities which go into Bayes’s theorem are objective, implying that any rational person ought normatively to accept the results of the argument. The similarities and differences between these thinkers will be explored in the context of specific arguments regarding the openness of the universe to scientific enquiry and its fine-tuned character.

Alister McGrath: natural theology redefined41 Natural theology is generally defined to be something like ‘the study of what can be known about God purely from being human, i.e. apart from any special revelation.’ We gave one definition along these lines, due to James Barr, at the beginning of this chapter. William Alston’s definition is similar: ‘Natural theology is the enterprise of providing support for religious beliefs by starting from premises that neither are nor presuppose any religious beliefs.’42

20  Reason and religion Alston’s definition is quoted by Alister McGrath, who is Andreos Idreos Professor of Science and Religion at Oxford, in several places,43 but it is directly opposed to his own definition. This is because, for McGrath, ‘natural theology cannot be conceived as an autonomous theological discipline, precisely because its foundational and legitimating insight—namely, that nature is to be viewed and recognised as God’s creation—is derived from divine revelation.’44 Natural theology is therefore to be pursued ‘within the scope of a revealed knowledge of God, rather than as an autonomous discipline outside its bounds.’45 This bears a very strong resemblance to the way Thomas Torrance sees natural theology. For Torrance, just as Einstein brought geometry into physics in his general theory of relativity, so natural theology must be brought within dogmatic theology, in its full NiceneChalcedonian formulation.46 McGrath’s stance is perhaps not surprising given that he is Torrance’s biographer.47 McGrath puts his basic point thus: ‘A Christian natural theology is about seeing nature in a specific manner, which allows the observer to discern in what is seen the truth, beauty, and goodness of a trinitarian God who is already known.’48 One might therefore be led to think that natural theology is a pursuit carried out by Christians for Christians to tell them more about this God whom they know already as Trinity. Theological belief is thus the starting point, rather than that which needs to be justified, in a way not so dissimilar from Alvin Plantinga’s invocation of theistic belief as properly basic. And that squares with much of what McGrath says. But it seems to me that there is a tension in McGrath who also wants to use natural theology apologetically. Thus he immediately adds to the last statement this: ‘and which allows nature to function as a pathway towards this same God for secular culture as a whole.’49 The implication would seem to be that a secular observer can learn something of the unknown God from natural theology after all. The detection of a tension, or ambiguity, here seems to be reinforced by McGrath’s appropriation of several insights from secular philosophy. Let us briefly examine three of these. Critical realism ‘Critical realism’ is a philosophical stance widely employed in the sciencereligion dialogue; both John Polkinghorne and Arthur Peacocke are exemplars. Alister McGrath also appeals to it, although the version to which he appeals is not that of these scientist-theologians but of the philosopher Roy Bhaskar. Bhaskar’s version takes account of ‘cultural, historical and social factors shaping human perceptions of reality.’50 This relates to McGrath’s view that nature is a socially constructed notion and therefore open, apparently equally, to both theistic and atheistic interpretations.51 I believe that this view is too negative and mutes McGrath’s criticism of atheist polemicist Richard Dawkins.

Reason and religion 21 McGrath also appeals to critical realism to assert, like Barth and Torrance before him, that theology is scientific because it treats its object in the way appropriate to it. He expresses this succinctly in the maxim ‘ontology determines epistemology,’52 which closely resembles Polkinghorne’s inverse expression ‘epistemology models ontology.’53 And then McGrath also uses Bhaskar to point to the multi-layered nature of reality, as did Torrance (though in his case, getting the idea from Michael Polanyi). This multi-layered texture of reality implies that different methods of enquiry are appropriate to different levels. So, while being scientific, theology will employ methods different from the other sciences which deal with other aspects of reality. Again, Polkinghorne would agree, though Swinburne would argue that in the realm of hypothesis comparison, the same Bayesian framework applies to both scientific and theological hypotheses. Tradition transcendent rationality McGrath’s second philosophical insight comes from Alasdair MacIntyre. According to MacIntyre, and McGrath concurs, there is no universal rationality, only different tradition-specific rationalities. This means that Christianity is free to advance its own form of rationality. However, that does not lead to relativism, McGrath says, because Christian faith has the capacity to account for the other rationalities, including those of other religions. That is to say, Christian rationality is not only tradition-dependent but tradition-transcendent. A  key reason why it can explain other religions comes from natural theology, which, after all, does indeed say that something can be known of God from outside the Christian framework!54 Moreover, to claim that Christian rationality is tradition-transcendent does not seem to me very far from saying that it ought to command universal assent, despite McGrath’s assertion to the contrary that there is no universal rationality. Inference to the best explanation McGrath’s third tool from philosophy is ‘inference to the best explanation,’ the expression being due to Gilbert Harman, but the idea itself closely resembling Charles Sanders Peirce’s notion of ‘abduction.’55 By this means, several competing hypotheses are evaluated to explain the evidence in question. The criteria for inferring which hypothesis is best are such factors as elegance, parsimony, and explanatory power (see later for the technical, Bayesian definition of the last term). The trouble for McGrath’s position is that these seem to be universal rational criteria, and McGrath has wanted to distance himself from universal rationality (albeit mainly from foundationalism). Anyhow, they are criteria used in the evaluation of scientific hypotheses and are not obviously dependent on McGrath’s starting point of Christian Trinitarian orthodoxy.

22  Reason and religion McGrath invites us to infer that the Christian faith offers the best explanation for such features of the universe as the fine-tuning of physical constants, the life-conducive chemical properties of water, and a number of other scientific facts.56 The Christian faith also offers the best explanation for the existence of its own competitors. But if this is so, on the basis of the criteria listed, then it is presumably so for all people and, despite their failure to recognize this, ought normatively to be given assent. Thus it seems to me that McGrath’s version of natural theology ends up being very close to the traditional kind of natural theology which he rejects.

John Polkinghorne: natural theology and motivated belief Scientist-theologian John Polkinghorne was Professor of Mathematical Physics at Cambridge University before training for ordination in the Church of England and is a pioneer of the modern science-religion dialogue. Polkinghorne shares with McGrath the view that natural theology is an undertaking to be carried out within dogmatic theology and integrated with it.57 Polkinghorne describes Thomas Torrance’s comparison of this to the way Einstein brought geometry into physics as a ‘brilliant use of a metaphorical resource.’ However, it seems to me that Polkinghorne’s understanding of natural theology is significantly different from that of Torrance and McGrath. Thus Polkinghorne writes: One may define natural theology as the attempt to learn something of God from the exercise of reason and the inspection of the world—in other words, from reflection on general experience rather than from specific revelatory events.58 That is a classic definition, which implies, contra Torrance and McGrath, the setting aside or ignoring of revelation while pursuing natural theology. And, while Polkingorne does not use the expression ‘providing support for religious beliefs,’ that is implicit in the rest of what he says, though with suitable caution and qualification. But in going on to say that natural theology has been revived in recent years by scientists including such as Paul Davies,59 who has no religious faith, Polkinghorne also distances himself from Torrance’s geometric metaphor and the pursuit of natural theology entirely from within dogmatic theology. Polkinghorne is in agreement with both McGrath and Swinburne (see later) that natural theology does not offer proof, as it had traditionally been considered to do. For Polkinghorne it offers ‘insight,’ indeed the ‘most satisfying insight into the way the world is.’60 Significantly, theism explains more about the universe than atheism does. That view edges Polkinghorne more towards Swinburne’s cumulative case-type argument. So, theism explains the fine-tuning of the universe, but that might also be explained by a multiverse. Yes, says Polkinghorne, but theism explains more than a multiverse

Reason and religion 23 does. It solves Einstein’s mystery of the comprehensibility of the universe and it explains why people profess to have experiences of the divine. An important concept for Polkinghorne is ‘consonance.’61 Again, Alister McGrath has adopted something similar but uses the word ‘resonance.’ What Polkinghorne appears to be saying is that science and theology speak about the world in different ways but what each says is variously ‘compatible,’ ‘congruent’ or ‘consonant’ with what the other says. Thus science tells us about the processes of the world; theology tells us that the universe is God’s creation and that God is providentially active within it. As an example of consonance, Polkinghorne notes, referencing John Zizioulas, that the triune God of Christian theology is ‘Being in Communion.’62 Modern science has provided a much more relational picture of the universe than classical science, for example through the interplay of chance and necessity, non-locality in quantum physics, and the self-organisation of dissipative systems held far from equilibrium. So what modern science has revealed is consonant with what Christian theology would expect on the basis of the doctrine of the Trinity, though of course it does not prove that doctrine. A further key concept for Polkinghorne is ‘motivated belief,’63 which seems to go beyond mere ‘consonance.’ Motivated belief arises through having evidence for what one believes, though this evidence will be different in theology from that involved in the sciences. For Polkinghorne, the evidence falls into two broad categories. The first is natural theology which gives rise to, at best, an etiolated view of God; indeed it is consistent with deism. The second, which, although he does not use the term, marks Polkinghorne out as a practitioner of ramified natural theology, is the evidence for the specific claims of the Christian revelation, and involves an assessment of the reliability of the documentary evidence of the New Testament. Neither of these presents ‘knock-down’ proof, but, like McGrath, Polkinghorne uses the language of ‘best explanation.’ For example, he says: The realistic aspiration is that of attaining the best explanation of complex phenomena, a goal to be achieved by searching for an understanding sufficiently comprehensive and well-motivated as to afford the basis for rational commitment.64 That aspiration is the best that can be achieved in any area of human enquiry, and therefore applies to both science and theology. In the case of the New Testament, belief in the resurrection represents abduction from the data provided about the empty tomb and post-death appearances of Jesus, and this in turn leads to Christological formulations which transcend traditional categories such as ‘prophet.’ In arguing that there is no absolute certainty Polkinghorne appeals to Michael Polanyi’s notion of personal knowledge. Even in science there is the inevitable involvement of a community of persons in assessing the truth claims of any discipline. Science cannot be wrapped up in some simple,

24  Reason and religion unique infallible method, but involves judgments made by persons. That means that complete objectivity is unattainable. A  favourite quotation of Polkinghorne’s from Polanyi sets out the aims of Personal Knowledge: ‘The principal purpose of this book is to achieve a frame of mind in which I may hold firmly to what I  believe to be true, even though I  know that it may conceivably be false.’65 Ultimately for Polkinghorne judgments of truth can only be made from inside the community—be that the scientific fellowship or the church. This is Anselm’s ‘faith seeking understanding’ and, says Polkinghorne, ‘there is no Archimedean point of detachment from which judgment can be made; insight is only gained through participation.’66 This is all part of how Polkinghorne understands ‘critical realism.’ Indeed Polkinghorne’s adoption of this, which pre-dates McGrath, also includes such aspects as the multilayered texture of reality, the subtle interplay between theory and experiment, and the recognition that there are no facts which are uninterpreted facts. He does not, however, endorse McGrath’s view that nature, as understood by science, is a socially constructed notion.67 This is because science has a universal aspect—the same experiment will give the same result in Cambridge or Canberra. The same applies to theology, even though this is more evidently influenced by culture, and Polkinghorne is perplexed by the phenomenon of religious pluralism. Thus, the fallibility of our judgements means that non-acceptance of the Christian faith does not render a person irrational.

Richard Swinburne and confirmation theory Richard Swinburne is Emeritus Nolloth Professor of the Philosophy of the Christian Religion at Oxford University and is one of the pre-eminent philosophers of religion of recent times. His pioneering approach to natural theology is entirely different from McGrath’s, and goes beyond Polkinghorne’s. Swinburne builds a cumulative case for the existence of God using successive pieces of evidence.68 Very importantly his case is a probabilistic one. Thus he uses traditional arguments for the existence of God, such as the cosmological and design arguments (including the fine-tuning version of the latter), but says that these do not constitute proofs of God’s existence but make God’s existence more likely than it would otherwise be. In other words, the existence of a physical universe is more likely if there is a God than if there is no God, and the universe is more likely to be ordered if there is a God. And if God’s existence makes the evidence likely then the evidence makes God’s existence likely. One way of looking at this is to see it as a mathematical formalisation of abduction or ‘inference to the best explanation,’ the methodology we met earlier. The criteria for a hypothesis being the best of competing hypotheses are that it makes the data more probably true than its rivals do, and it is simpler, more economical and more elegant than its rivals.

Reason and religion 25 Included in making up the case are also arguments from consciousness, morality, providence, history, miracles, and (for Swinburne, the decisive argument) religious experience. These all enhance the probability of God’s existence. Swinburne also recognizes the problem of theodicy. The existence of natural and moral evil diminishes the probability of the God hypothesis but this is outweighed by all the positive arguments, and in the end the hypothesis is overall more probably true than not. Swinburne also uses the term ‘ramified natural theology’ when it comes to assessing the probability that the particularities of the Christian religion are true.69 My point in this chapter is not to examine or critique the precise arguments adduced by Swinburne in any detail, but rather to comment on his method. The tool he uses to build up the case is Bayesian confirmation theory. It should be emphasized at the outset, especially for readers who are not mathematically trained, that Bayesian confirmation theory is simply a way of formalizing our judgments. Indeed, it captures the way in which scientists weigh up competing theories against each other using the same kinds of criteria—how well theories predict the data, and how simple, comprehensive, and elegant the theories are. It is worth stating that scientific theories cannot be proved. Rather, the evidence makes them more or less probably true. An example will illustrate this. Newton’s law of gravity could only be proved if we could measure the force between every pair of bodies in the universe. Even if we could do that, there would be the further problem that we cannot isolate any pair of bodies from the rest of matter in the universe and so there would be distorting effects from other gravitating bodies. Theological explanations are the same in this regard: the evidence we have renders our conclusions true only to some degree of probability. The starting point is a theorem in probability theory enunciated by the eighteenth century non-conformist cleric the Reverend Thomas Bayes. Bayes’s theorem is expressed in the following equation: P[H|E] =

P[E|H ].P[H ] (2.1) P[E]

Here H represents some hypothesis being evaluated against evidence E. The symbol ‘|’ is read ‘given’ so that probabilities such as P[E|H] are known as conditional probabilities, in this case the probability that one sees evidence E if hypothesis H is true. P[E|H] is also known as the ‘likelihood’ of the hypothesis. P[H] is the prior probability of the hypothesis, i.e. the probability that H is true before the evidence is taken into account, and so based simply on background knowledge such as the laws of logic. The symbol ‘.’ means multiplication (it is often omitted). The ratio P[E|H]/P[E] is called the ‘explanatory power’ of H with respect to evidence E. P[H|E] is the posterior probability that H is true given E. P[H|E] = 1 means that given E, H is certainly true, and P[H|E] = 0 means that given E, H is certainly false. Thus hypotheses are more probable the nearer their probability is to 1.

26  Reason and religion It should be noted that the probabilities in the equation are ‘epistemic’ probabilities and not physical probabilities. They are ‘degrees of belief’ and it can be shown that, for a rational agent, the degrees of belief which should be accorded to a proposition obey the probability calculus, and so can be interpreted as a form of probability. There are several important consequences of Bayes’s theorem which can be simply derived from equation (2.1). Thus the denominator in equation (2.1) can be expanded using the ‘total probability rule’ of probability theory: P  E  P  E|H .P  H   P  E| ~ H .P ~ H  where the expression ~H is read ‘not H’ and is the hypothesis that H is false. From this it follows that P  E|H .P  H 

P  H|E 

P  E|H .P  H   P  E| ~ H .P ~ H 

(2.2)

What the theorem tells us is how to revise the probability of some hypothesis in the light of evidence. It shows how the probability of the hypothesis given the evidence, P[H|E], depends on the other probabilities in the equation: the prior probability, P[H]; the likelihood, P[E|H]; and the probability of the evidence supposing the hypothesis is false, P[E|~H], which is the likelihood of ~H. P[~H] is simply equal to 1 ₋ P[H]. Another useful consequence of the theorem can be obtained by writing equation (2.1) for competing hypotheses H1 and H2 substituted for H. We can then divide these two equations to obtain P  H1 |E

P  H 2 |E



P  E|H1  P  H1  . (2.3) P  E|H 2  P  H 2 

If one wants simply to compare the probability of H with the probability of ~H, then this gives P  H|E

P ~ H|E



P  E|H 

.

P H 

P  E| ~ H  P ~ H 

(2.4)

In equations (2.3) and (2.4), the first ratio on the right hand side, P[E|H1]/P[E|H2] or P[E|H]/P[E|~H], is termed the ‘likelihood ratio’ or sometimes just the ‘Bayes factor.’ The second ratio is the ratio of the priors. Swinburne introduces further helpful terminology. Thus if P[H|E] is greater than P[H], i.e. the evidence has enhanced the probability of the hypothesis over its a priori value, we have a ‘C-inductive argument’ and E is said to confirm H. If we can go better than this and show that P[H|E] is greater than ½, in which case H is said to be ‘probable,’ we have

Reason and religion 27 a ‘P-inductive argument.’70 These are in contrast to deductive arguments, which make a hypothesis certain, but work in neither science nor theology. Let me give an example of Bayes’s theorem in practice. One dark and rainy night Sarah witnesses a hit-and-run accident. The car which leaves its victim behind is a taxi and Sarah tells the police that it was a blue taxi. There are two taxi firms in the town, one using green taxis and the other blue taxis. The green taxi firm is the dominant one with 85% of the taxis. Sarah is tested to see if she can correctly identify the colours of taxis under similar conditions to the night in question, and she is right 80% of the time. The jury is asked to judge whether it was indeed the driver of a blue taxi, as reported, who committed the crime.71 In this example the prior probability that the culprit was driving a blue taxi is 15%, i.e. 0.15; the likelihood that it was blue, i.e. the probability that it was reported blue if it was blue, is 0.8; the likelihood that it was green is 0.2. The correct answer to this problem contradicts what most psychological tests reveal people’s judgments to be. Many people confuse the likelihood, the probability that it was reported blue given that it was blue, with the posterior probability, the probability that it was indeed blue given that it was reported as blue. Most people think the criminal in this example was more likely to have been driving a blue car. However, Bayes’s theorem shows that the posterior probability that the car was blue, given the report that it was blue, is only about 0.41. Thus, despite a fairly reliable witness testifying to it being blue, it is in reality more likely to have been a green car. The fact that we are inclined to misjudge probabilities is rather worrying for the jury system in Britain, where, in a notable case, the Appeal Court ended up deciding that experts should not explain Bayes’s theorem to juries or guide them through the process of using it. Peter Donnelly of Oxford University has described how this could be done without baffling juries with mathematics.72 When it comes to assessing the probabilities relevant to the God-hypothesis Swinburne recognizes that one cannot be very precise, certainly much less precise than in the case of the taxi driver. But one can make fairly confident statements of the form ‘It is much more likely that humans have religious experiences if God exists than if he does not,’ or ‘It is much more likely that a physical universe exists and that the physical universe which exists is finetuned for life if there is a God than if there is not.’ One reason would be that a universe selected randomly from the set of possible universes is highly unlikely to be fine-tuned. The main problem with Swinburne’s approach is ascertaining the prior probability that God exists, i.e. the probability of God’s existence given only background knowledge such as the truths of logic. But Swinburne argues that God is much more likely to exist uncaused than a physical universe is. His main ground for saying this is that God is simple and a physical universe is complex. A suite of physical universes, as in the multiverse hypothesis, is vastly more complex still. Where there are competing alternative hypotheses

28  Reason and religion to explain any phenomenon the simplest will have the greatest prior probability, says Swinburne. This seems to be the case for science and there is no reason for it to be otherwise when we compare metaphysical hypotheses. The upshot of all this is that given the fine-tuned physical universe we observe, the probability that God exists as its creator is much higher than the probability that it exists uncaused. The main point is that for Swinburne, even though we cannot give precise figures, the probabilities used in Bayes’s theorem are objective. This is his most controversial step. If they were subjective, so that any individual might make subjective judgments about the relative probabilities, there would clearly be no problem. And that is how some philosophers, such as Colin Howson, Peter Urbach, and John Earman, see matters.73 On that basis Bayes’s theorem would essentially ensure that any individual’s belief system was self-consistent. For Swinburne, though, these are not merely epistemic probabilities in the sense of personal degrees of belief; they are logical probabilities. They relate to the way things are and relate to each other in the mathematical space of possibilities, and the objective measure of support one proposition gives to another. Logical probability is ‘that measure of inductive support that would be reached by a logically omniscient being.’74 Swinburne repudiates the subjective probability view, since he believes that would mean the end of objective truth. And that would be not just objective truth about the probability of God’s existence but about the probability of any scientific theory being true. Things may not be quite as bad as Swinburne paints them with the subjective view. Thus, as we shall see, it may turn out that an atheist is forced into a corner and challenged in having to assign an absurdly low prior probability to God’s existence, simply because the evidence is overwhelmingly more likely to pertain given the theistic hypothesis than given the atheistic hypothesis. Be that as it may, most importantly from the point of view of this book, Swinburne applies the Bayesian framework both to the arguments of ‘bare’ natural theology and to those of ramified natural theology and we follow in his footsteps.

Summary: McGrath, Polkinghorne, and Swinburne McGrath utilizes the language of natural theology, but redefines it. He exhibits a certain circularity in arguing on the one hand that natural theology tells us about the God who is already known and ambiguity on the other by utilizing such notions as ‘inference to the best explanation.’ Polkinghorne uses similar language to McGrath but in a more traditional and less ambiguous way. He offers a commendably cautious, rational approach to the subject of natural theology, and in some ways his position is a mediating one between those of McGrath and Swinburne. Polkinghorne is similar to Swinburne in offering arguments, such as the universe’s finetuning, for a Creator behind the universe, and historical evidence for the

Reason and religion 29 particulars of the faith. A  major difference is that Swinburne utilizes the analytical framework of Bayes’s theorem. However, even though Swinburne argues that the probabilities in the theorem are objective, he admits that we may make mistakes in our judgments of them from our limited perspective. There might be an ‘Archimedean point’ but it is not a point which any human being has perfect access to. In my own work I have utilized Bayes’s theorem in a manner similar to Swinburne, while attempting to judge criteria of simplicity, etc. in a similar way to experts in the scientific community, a procedure which Polkinghorne justifies on the grounds of historical success.75 An area of major difference between Swinburne and Polkinghorne is that of theodicy, which is a separate subject in its own right. In the areas I have considered the differences may to a considerable extent be matters of style rather than substance, but they give the impression of a more cautious critical realist approach to Polkinghorne and a more rationalist, possibly ultrarationalist approach to Swinburne.

What kind of explanation does God provide? All three authors discussed in the latter part of this chapter consider God to be explanatory in some sense or other. Traditionally, it has certainly been thought that God is an explanation, for the universe and all sorts of things in it. That may not be the primary role of religion, but it is a role. The question we now have to ask is what kind of explanation God provides. Explanations are of different kinds. And they do not necessarily exclude each other. God provides a personal explanation, e.g. for the existence of the universe.76 Science provides an explanation in terms of scientific laws and initial conditions, e.g. for how the universe evolved from the Big Bang. To give an example often used by John Polkinghorne, if I go into the kitchen and put the kettle on, there will be a scientific explanation for why the water boils in terms of bubbles forming and growing as the temperature rises and percolating upwards. But there is also the personal explanation that I want a cup of tea! Now let us think about an explanation for why we are here in the first place. Why am I here? Scientifically there is a story about a particular sperm hitting a particular egg and making me. There is also a personal story about my parents loving each other and wanting to share that love with a third person who reflects something of each of them. But we can take it back much further. Scientifically we are here because we are descended from our ancient ancestors and their predecessors. We are descended from simpler animals and ultimately from bacteria floating around in the primordial soup. The earth is here because matter collapsed due to gravity to form the sun and solar system. They are here because the universe started with a Big Bang and, as the universe expanded, galaxies and stars began to form through gravitational collapse. The chemical elements manufactured by

30  Reason and religion nuclear fusion in the interiors of the first generation of stars provided the material out of which our earth and ourselves were subsequently made. As we look back towards the Big Bang we can see that scientific explanations begin to run out. Indeed the whole story is open to a personal explanation as well as a scientific one. In fact the scientific explanation is not really answering the ‘why’ question at all. It is only answering the question ‘how?’ e.g. ‘How did we come to be here?’ So God and the Big Bang are not rival explanations. Rival scientific explanations would be the Big Bang theory and the steady-state theory for the origin and evolution of the universe, a live issue in cosmology until the clinching evidence for the Big Bang—the cosmic background radiation—was discovered in 1965. But what might be called rival metaphysical explanations for why there is a universe at all would be ‘God made it’ or ‘It just is and was not created by anybody’ or ‘It is some kind of gigantic accident.’ The first of these is a personal explanation; the other two are not personal explanations, but neither are they scientific ones. They could be described as ‘atheistic’ as opposed to ‘theistic’ explanations. So the rivals now are theism (‘God exists’) and atheism (‘there is no God’) and not ‘God or the Big Bang.’ The explanation that God provides to these questions goes like this. I am thinking of the Christian God, but most of what I say for the time being applies equally to the God of Judaism and of Islam. Indeed most members of these religions would believe that they are worshipping the same God, even if they have some different things to say about him (which I shall come to in a moment). Having said that, let us proceed. God as understood by Christians created the universe with intentions and purposes. He is eternal, all-powerful, all-good, all-knowing. Out of his overflowing love and creativity he made the universe. He intended it to produce creatures like us within it who would be able to contemplate his handiwork and who would be able to have a personal relationship with him. His purpose was that those creatures would love him and their neighbours and would take care of the world he had made. They might not have done that very well, but that was God’s intention. This idea of God explains why the universe is here and it explains why the universe produces human beings after its 13.8 billion year evolution from the Big Bang. Indeed the more we go into particulars about what is required in the way the universe is set up to produce humans, the more it seems that God explains and the harder a struggle atheism seems to have. We shall come back this issue of fine-tuning, which I have mentioned before, in more detail in Chapter 3. One could list many more features of the universe we inhabit which God explains and atheism either has to say are ‘just there’ (or perhaps are not, we have invented them) or are the result of some cosmic accident. Examples include: why we are in a universe in which consciousness arises, why there are moral values (why good and evil, right and wrong mean something), and why people have religious experiences of God or some divine presence.

Reason and religion 31 God even explains why we can do science in the first place! As noted earlier, Richard Swinburne uses these in building his cumulative case argument for the existence of God. From what I  have said so far, it looks as though theism—the existence of God—can provide explanations for why things exist and why they are as they are. Explanation in terms of God is clearly different from scientific explanation. And we can certainly argue that God provides a better explanation than atheism. But can this idea of God be tested in any more tangible way? That would make God more like a scientific hypothesis. One problem is that if one reads the Bible, it tells us that we should not put God to the test. And of course, God is not amenable to testing in the way a rat in a laboratory is. Even humans are not testable in that sense of course. I suggested earlier that faith is like marriage, a relationship of trust. If that is so, then faith is something that is harmed by testing, rather than helped. At least, that is the case if we construe testing in the rather narrow laboratory sense of the word. However, I also quoted John Locke’s statement that faith should only be accorded to anything on the basis of good reason, and Wolfhart Pannenberg to the effect that faith is in fact based on reliable evidence. One of the questions one needs to ask about God is whether he actually does anything in the world as opposed to just setting it up. If he does, perhaps one can know about it. This is where one of the major differences between the religions comes in, namely their historical origins. The Christian faith is centred on the life, death, and resurrection of Jesus Christ. When people ask a Christian what God is like, he or she could give the kind of abstract answer that we have been talking about: the eternal, omnipotent, and omniscient Creator of the universe. That would be true, but a much more distinctively Christian answer would be, ‘Look at Jesus Christ. That is what God is like.’ Karl Barth is absolutely right about that, though wrong to neglect the intimations of God in creation. So this is where it would be good to look for evidence. However, that evidence is not scientific but historical. It needs the methods of historical research to investigate it. Science cannot exclude an event like the resurrection of Jesus from the dead because science just describes the normal regularities of the world. But it still needs good evidence to back up the claim that someone rose from the dead. As it happens, that evidence is very powerful: the empty tomb, the many appearances of Jesus alive after his death, and the spread of Christianity started by a handful of disciples who fled from the crucifixion scared to death themselves and completely disillusioned. This just begins what needs to be the much fuller analysis which we shall go into in a later chapter, but it shows that in principle historical claims can be ‘confirmed’ in a similar way to scientific ones—or falsified if, for example, someone were to come up with compelling evidence that they have discovered Jesus’ bones or that the disciples were involved in some kind of deception.

32  Reason and religion To sum up, I have said that scientific explanation and theistic explanation are different kinds of explanation, but there are some similarities. Theistic explanation is personal. I have said that evidence cannot prove or verify a scientific theory but it can confirm it in the sense of making it more probably true. The same is true of evidence for theistic belief. Some of the evidence for the latter comes from the very existence of the universe and the way it is set up—as described by science. Evidence for God can also be drawn, inter alia, from the existence of morality, from religious experience, and, most of all perhaps, from history. At the very least science and religion can live quite happily together because they are not competitors. And finally, the most important point of all. We can know that God exists, from nature and the exercise of our natural reason. But we can only know God in himself, we can only know God as our Redeemer and Saviour, through his gracious revelation of himself in Jesus Christ. We can evaluate the evidence for that revelation being authentic, again using our natural reason, and we proceed to do this in later chapters. However, in coming to faith, putting our trust in him, we are appropriating for ourselves the reliable event of Christ’s life, death, and resurrection.

Notes 1 This chapter utilises material, edited and supplemented, from Rodney Holder, The Heavens Declare: Natural Theology and the Legacy of Karl Barth (West Conshohocken: Templeton Press, 2012), 3–14. 2 James Barr, Biblical Faith and Natural Theology: The Gifford Lectures for 1991 Delivered in the University of Edinburgh (Oxford: Oxford University Press, 1993), 1. 3 St  Augustine, The City of God (De Civitate Dei), Book VIII, in Philip Schaff (ed.), Nicene and Post-Nicene Fathers, First Series, vol. 2 (Peabody, MA: Hendrickson, 1994), 144–165. 4 St Thomas Aquinas, Summa Theologiae, Blackfriars edition (London: Eyre & Spottiswoode; New York: McGraw Hill, 1963), 1a. 2, 2. 5 John Calvin, Institutes of the Christian Religion, trans. Henry Beveridge (Grand Rapids, MI: Wm. B. Eerdmans, [1559] 1989), Book II. 6 Francis Bacon, Advancement of Learning, quoted facing the title page of the first edition of Charles Darwin, The Origin of Species. 7 John Locke, An Essay Concerning Human Understanding, edited with an introduction by P. H. Nidditch (Oxford: Oxford University Press, [1689] 1979), Bk IV, ch. XVII, § 24, 687. 8 David Hume, An Enquiry Concerning Human Understanding, Section X, ‘Of Miracles,’ in the first part of L. A. Selby-Bigge (ed.), Enquiries Concerning Human Understanding and Concerning the Principles of Morals by David Hume, third edition, revised by P. H. Nidditch (Oxford: Oxford University Press, [1748] 1975), 104–148, 110. 9 Wolfhart Pannenberg, ed., with Rolf Rendtorff, Trotz Rendtorff, and Ulrich Wilckens, Revelation as History, trans. D. Granskou (New York: Macmillan, 1968), 138. (First German edition, Offenbarung als Geschichte, Göttingen: Vandenhoeck and Ruprecht, 1961). 10 Immanuel Kant, Critique of Pure Reason, trans. J. M. D. Meiklejohn (London: J. M. Dent, [1781] 1934), 43–47. 11 Kant, Critique of Pure Reason, 370.

Reason and religion 33 12 See Richard Swinburne, The Existence of God, second edition (Oxford: Oxford University Press, 2004), 58, n. 4. 13 David Hume, Dialogues Concerning Natural Religion, ed. J. M. Bell (London: Penguin, [1779] 1990). 14 Hume, Dialogues, part 2. 15 Hume, Dialogues, part 5. 16 Hume, Dialogues, part 6. 17 Hume, Dialogues, part 7. 18 Gottfried Wilhelm Leibniz, ‘On the Ultimate Origination of Things,’ in Robert Latta (trans.), The Monadology and other Philosophical Writings (Oxford: Oxford University Press, [1697] 1971), 337–351, 338. 19 William Paley, Natural Theology, or Evidence of the Existence and Attributes of the Deity, Collected from the Appearances of Nature (Oxford: Oxford University Press, [1802] 2006). 20 Sassoon was writing of his preparation for Cambridge entrance in the year 1905: Siegfried Sassoon, ‘Seven More Years,’ in The Old Century and Seven More Years (London: Faber and Faber, [1905] 1968), 250. 21 Paley, Natural Theology, 9. 22 Rodney D. Holder, God, the Multiverse, and Everything: Modern Cosmology and the Argument from Design (Aldershot: Ashgate, 2004), 43–47. 23 Paley, Natural Theology, 12. 24 Paley, Natural Theology, 207. 25 Holder, God, the Multiverse, and Everything, 36. 26 Gregory of Nyssa, Apologetic Treatise on the Hexemeron, quoted in McMullin, ‘Introduction: Evolution and Creation,’ in Ernan McMullin (ed.), Evolution and Creation (Notre Dame, IN: University of Notre Dame Press, 1983), 12. 27 See St  Augustine, The Literal Meaning of Genesis (De Genesi ad Litteram), Ancient Christian Writers, vol. 1, translated and annotated by John Hammond Taylor SJ (New York and Mahwah, NJ: Paulist Press, 1982), translator’s note 67 to Book IV, 252–254. 28 Darwin Correspondence Project Database, letter no. 2534, accessed February 24, 2020, www.darwinproject.ac.uk/entry-2534. 29 Charles Kingsley, ‘The Natural Theology of the Future,’ in Westminster Sermons (London: McMillan, 1874), v–xxiii. 30 Alister E. McGrath, A Fine-Tuned Universe: The Quest for God in Science and Theology, The 2009 Gifford Lectures (Louisville, KY: Westminster John Knox Press, 2009), 169. 31 W. F. Temple, The Relations Between Religion and Science, The Bampton Lectures for 1884 (London: Macmillan, 1885), 114ff. 32 Kingsley, ‘The Natural Theology of the Future.’ 33 Whilst chapter 3 is devoted to this topic, book length treatments can be found in Holder, God, the Multiverse, and Everything; and, less technically, in Rodney D. Holder, Big Bang, Big God: A  Universe Designed for Life? (Oxford: Lion Hudson, 2013). 34 McGrath, Fine-Tuned Universe, 106. 35 See Rodney D. Holder, ‘Natural Theology in the Twentieth Century,’ in Russell Re Manning (ed.), The Oxford Handbook of Natural Theology (Oxford: Oxford University Press, 2013), 118–134. 36 S. L. Jaki, Lord Gifford and his Lectures: A Centenary Retrospect (Edinburgh: Scottish Academic Press; Macon, GA: Mercer University Press, 1986), 2. 37 Jaki, Lord Gifford and his Lectures, 73–74. 38 Jaki, Lord Gifford and his Lectures, 74. 39 Larry Witham, The Measure of God: Our Century-Long Struggle to Reconcile Science and Religion, The Story of the Gifford Lectures (San Francisco: HarperSanFrancisco, 2005), 71.

34  Reason and religion 40 The author is grateful to the Ian Ramsey Centre and the International Society for Science and Religion, joint organizers of the ‘God and Physics’ conference held in Oxford in July 2010 to celebrate John Polkinghorne’s 80th birthday, for the opportunity to present a short paper. The latter half of this chapter is a modified and extended version of the paper given at the conference. 41 For more on McGrath, see Holder, The Heavens Declare, 169–231. 42 William Alston, Perceiving God: The Epistemology of Religious Experience (Ithaca, NY: Cornell University Press, 1991), 289. 43 See Alister E. McGrath, The Open Secret: A New Vision for Natural Theology (Oxford: Blackwell, 2008), 7; Alister E. McGrath, A Scientific Theology Volume 1: Nature (Edinburgh and New York: T & T Clark, 2001), 241; Alister E. McGrath, The Order of Things (Oxford: Blackwell, 2006), 67. 44 McGrath, A Scientific Theology: Nature, 295. 45 McGrath, A Scientific Theology: Nature, 295. 46 E.g. Thomas F. Torrance, Space, Time and Incarnation (Edinburgh: T & T Clark, 1969; paperback edition, 1997), 69–70; Thomas F. Torrance, The Ground and Grammar of Theology (Belfast: Christian Journals, 1980), 91–92. 47 Alister E. McGrath, T. F. Torrance: An Intellectual Biography (Edinburgh: T & T Clark, 1999). 48 McGrath, The Open Secret, 147. 49 McGrath, The Open Secret, 147–148. 50 Alister E. McGrath, A Scientific Theology Volume 2: Reality (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2002), 208. 51 E.g. Alister E. McGrath, Dawkins’ God: Genes, Memes, and the Meaning of Life (Oxford: Blackwell, 2005), 53–57; Alister E. McGrath with Joanna Collicutt McGrath, The Dawkins Delusion: Atheist Fundamentalism and the Denial of the Divine (London: SPCK, 2007), 13. 52 E.g. McGrath, A Fine-Tuned Universe, 214. 53 John Polkinghorne, Science and Christian Belief: Theological Reflections of a Bottom-up Thinker (London: SPCK, 1994), 156. 54 McGrath, Order of Things, 64. 55 E.g. Alister E. McGrath, A Scientific Theology Volume 3: Theory (Grand Rapids, MI: Wm. B. Eerdmans, 2003), 153–165, 234. 56 McGrath, A Fine-Tuned Universe, Part 2. 57 John Polkinghorne, Faith, Science and Understanding (London: SPCK, 2000), 176–177. 58 John Polkinghorne, ‘Where is Natural Theology Today?’ Science and Christian Belief 18(2) (2006), 169–179, 169. 59 Polkinghorne, ‘Where is Natural Theology Today?’ 171. 60 Polkinghorne, ‘Where is Natural Theology Today?’ 171. 61 John Polkinghorne, Theology in the Context of Science (London: SPCK, 2008), 66ff. 62 Polkinghorne, Theology in the Context of Science, 82; John D. Zizioulas, Being in Communion: Studies in Personhood and the Church (New York: St Vladimir’s Seminary Press, [1985] 1997). 63 Polkinghorne, Theology in the Context of Science, 84ff. 64 Polkinghorne, Theology in the Context of Science, 85–86. 65 Michael Polanyi, Personal Knowledge: Towards a Post-Critical Philosophy (Chicago: University of Chicago Press, 1974), 214; cited for example in Polkinghorne, Faith, Science and Understanding, 33–34. 66 John Polkinghorne, Belief in God in an Age of Science (New Haven and London: Yale University Press, 1998), 115. 67 Polkinghorne, Belief in God in an Age of Science, 121.

Reason and religion 35 8 Most notably in Swinburne, Existence of God. 6 69 See Richard Swinburne, ‘Natural Theology, its “Dwindling Probabilities” and “Lack of Rapport”,’ Faith and Philosophy 21(4) (2004), 533–546. 70 Swinburne, Existence of God, 4–7. 71 This example is adapted from one by Amos Tversky and Daniel Kahneman, and discussed in Ian Hacking, An Introduction to Probability and Inductive Logic (Cambridge: Cambridge University Press, 2001), 72–73. I also quote it in Holder, Big Bang, Big God, 162. 72 Peter Donnelly, ‘Appealing Statistics,’ Significance 2(1) (2005), 46–48. 73 Colin Howson and Peter Urbach, Scientific Reasoning: The Bayesian Approach, second edition (Chicago and La Salle, IL: Open Court, 1993); John Earman, Bayes or Bust? A Critical Examination of Bayesian Confirmation Theory (Cambridge, MA: MIT Press, 1992). 74 Richard Swinburne, Epistemic Justification (Oxford: Oxford University Press, 2001), 64. 75 E.g., Polkinghorne, Belief in God in an Age of Science, 106. 76 Swinburne, Existence of God, 35–45.

3 Natural theology and modern cosmology The cosmological and design arguments

Introduction1 Modern cosmology tells us that the universe began with a gigantic explosion we now call the Big Bang, out of which all the galaxies, stars, and planets we observe today evolved over a period of 13.8 billion years. It seems to me that modern cosmology raises two fundamental issues which take us beyond the science into the realm of metaphysics and, I shall argue, theology. The first relates to the question of the beginning and to the traditional cosmological argument for the existence of God, in one form or another. The second relates to the so-called fine-tuning, the way in which the universe, and the laws of nature describing its structure and evolution, seem to be set up in a very special way in order for intelligent life to evolve at some stage in its history. This gives rise to a relatively new form of the traditional design, or teleological, argument for God’s existence. For each of these major issues, I want to ask whether, and to what extent, purely naturalistic explanations can explain the phenomena—that is, of the universe’s existence in the first place and of its fine-tuned character in the second—as alternatives to the explanations of divine creation and divine design. I shall present the arguments for each issue and then subject them to the questions posed by David Glass and Mark McCartney in their seminal paper, ‘Explaining and Explaining Away in Science and Religion,’ published in the November 2014 issue of Theology and Science.2 In that paper Glass and McCartney make a useful beginning to the fine-tuning issue, as Glass has also done in slightly more detail in his chapter ‘Can Evidence for Design be Explained Away?’ in a volume edited by Jake Chandler and Victoria Harrison.3 In this chapter I argue that two proposals made by cosmologists fail to explain away the universe’s beginning, and that science is powerless to explain away the more fundamental question as to why there is a universe at all. I argue similarly that scientific or quasi-scientific proposals such as the multiverse fail to explain away the fine-tuning. Indeed, I argue that theism is the most rational position to adopt to explain such features. It seems to me that the arguments here are powerful ones of natural theology, which is preliminary to ramified natural theology and lends credence to the latter.

Natural theology and modern cosmology 37 To aid our discussion and set the scene, let me first set out Glass and McCartney’s questions, and their implications, for the general case. Suppose we have evidence E, an initial hypothesis HI, and a proffered alternative hypothesis HA: Question 1. Is the alternative hypothesis HA incompatible with the initial hypothesis HI? If the answer to this question is ‘yes,’ then learning that HA is true will render HI false and, in Glass and McCartney’s terminology, ‘absolute explaining away’ results. Question 2. How likely is it that the alternative explanatory hypothesis HA would result in evidence E without the help of HI? Suppose HA is true and HI is false. Then, as Glass and McCartney say, other things being equal, the greater the probability of E in this case the greater the extent to which explaining away occurs. Question 3. Is the alternative explanatory hypothesis HA known to be true? Or, how strong is the independent evidence for it? Of course if HA is false no explaining away can occur, so independent evidence is required for HA to have any traction. The stronger the evidence there is for HA the greater the extent to which explaining away occurs. Question 4. Does the explanatory hypothesis HA depend on HI? If HA is more likely to occur if HI is true than if HI is false, then this decreases the extent to which explaining away occurs and if the dependence of HA on HI is strong enough, it may imply that HA provides additional confirmation of HI. On the other hand, if HA is less likely to occur if HI is true, this makes explaining away more likely to occur. Question 5. Is E taken as providing evidence for HI in the first place (i.e., before we find out that HA is true)? Unless the answer to this question is ‘yes,’ there is clearly no explaining away to be done. With this framework in mind, let us proceed to examine the issue of the ostensible beginning of the universe.

What do we make of the beginning? The cosmological argument proceeds from the existence of the universe to God. In recent times, evangelical Christian philosopher William Lane Craig has re-presented, with considerable force and vigour, a form of cosmological

38  Natural theology and modern cosmology argument that arose in mediaeval Islam,4 and, with equal force and vigour, the Roman Catholic apologist Robert J. Spitzer has done the same.5 This form of argument is known as the kalām cosmological argument. It could equally well be called the temporal cosmological argument—as opposed to the argument of Thomas Aquinas, which does not rely on there being a beginning to the universe and to which we shall come in due course. Likewise, it should be distinguished from Leibniz’s argument based on the principle of sufficient reason. The kalām argument goes as follows: 1 Everything that begins to exist has a cause of its existence. 2 The universe began to exist. 3 Therefore, the universe has a cause of its existence. Presented like that, the argument is a logical proof. If one accepts the premises, 1 and 2, then the conclusion, 3, follows by inexorable logic. Of course, there is another step needed to identify the cause in the conclusion with God. However, surely at least part of what is meant by the term ‘God’ is the cause of all contingent things, so that we may say with St Thomas, as he does at the end of his five proofs, ‘And this we call God.’ Now on the face of it, premise 1 seems very reasonable. Craig says it is intuitively obvious. Things do not just appear from nothing. Something has to exist to cause them. Premise 2 seems to be precisely what the Big Bang theory is saying. The universe began 13.8 billion years ago. However, I think there are reasons, both scientific and theological, to be cautious about drawing too definite a conclusion about this. I would prefer to see this as an argument which makes creation by God more probably true than it would otherwise be, rather than as a logical proof, which is how it seems to be portrayed by Craig. What are the arguments for either rejecting or accepting the kalām cosmological argument? Are there alternative naturalistic explanations for the origin of the universe that explain the universe’s existence or explain away the need for divine creation? Scientifically, there are real problems about the beginning. As one goes back further in time towards the beginning, the physics that applies becomes much less secure. Classically, if one takes the standard time-evolution equations according to general relativity and ignores quantum theory, the universe ends up infinitely dense and at an infinite temperature. This is a singularity, a point at which physics breaks down; and it marks the beginning of time. There are a number of theories which try to get around that problem, all of which are highly speculative, since we have no agreed theory which combines quantum theory and general relativity, which is what is required to handle the régimes in question.

Natural theology and modern cosmology 39

The Hartle-Hawking ‘no boundary’ proposal One such theory is due to the late Stephen Hawking and his colleague James Hartle. According to the Hartle-Hawking theory, as one goes back towards the beginning space-time gets ‘smoothed out’ and time becomes like a space dimension. Time itself is then imaginary in the technical sense of complex numbers. This four-dimensional space has no boundary or edge. This is what Hawking called the ‘no boundary proposal.’ It appears in his first book, A Brief History of Time;6 and it reappears in his later book, co-authored with Leonard Mlodinow, The Grand Design.7 Hawking seemed to agree with Craig that if the universe had a beginning then it would need God to create it. He thought that his no boundary proposal did away with a beginning and therefore with any need for God. Here is what he said, now quoting from A Brief History of Time: There would be no singularities at which the laws of science broke down and no edge of space-time at which one would have to appeal to God or some new law to set the boundary conditions for space-time. One could say: “The boundary condition of the universe is that it has no boundary.” The universe would be completely self-contained and not affected by anything outside itself. It would neither be created nor destroyed. It would just BE.8 And again: So long as the universe had a beginning we could suppose that it had a creator. But if the universe is really completely self-contained, having no boundary or edge, it would have neither beginning nor end; it would simply be. What place, then, for a creator?9 In The Grand Design Hawking and Mlodinow say the same thing: a universe with no beginning in time has no need for God to ‘light the blue touch paper’ to set it going.10 Thus for Hawking the no boundary proposal explained away the need for God. There are serious scientific and philosophical problems with Hawking’s proposal. Not least is the idea of imaginary time. For example, real time measures change from one state of a physical system to another. If time has become imaginary, in other words a fourth space dimension, the universe might ‘just be’ but how can it ever be otherwise than it is? In A Brief History of Time, Hawking discussed philosophical options for the meaning of imaginary time and the Euclidean 4-space which embraced it ‘prior,’ in some doubtful sense, to the emergence of real time. Either these were convenient mathematical devices, or maybe imaginary time was real and real time our own invention. In The Grand Design he embraces a rather

40  Natural theology and modern cosmology odd philosophical position he calls ‘model-dependent realism’ (despite saying on the opening page that philosophy is dead), according to which, he says, it is meaningless to ask which is real since they both exist only in our minds and it is only a matter of which is the more useful description. With his muddled philosophical reasoning, Hawking was essentially saying that we can believe what we like about imaginary time: we can perfectly well accept only real time in the mathematical sense as ontologically real, and the universe beginning, though not from a singularity, but at the surface where (real) 3-space and real time intersect the Euclidean four-space where time has become imaginary. Imaginary time is then just a useful calculating device, much as imaginary numbers are elsewhere in physics, serving to give us the radius of the universe at its beginning. Surprisingly, this is precisely the interpretation Hawking himself gave in his technical papers. Thus, the Euclidean (4-space) and Lorentzian (real space-time) regions in the model are demarcated by the wave function being exponential in the former and oscillatory in the latter. Hawking wrote: ‘We live in a lorentzian geometry and therefore we are interested really only in the oscillatory part of the wave function.’11 Moreover, ‘The Lorentzian geometries began at a non-singular minimum radius . . . The Lorentzian solutions will be the analytic continuation of the Euclidean solutions. They will start in a smooth non-singular state at a minimum radius equal to the radius of the 4-sphere and will expand and become more irregular.’12 As atheist philosopher Quentin Smith notes, the only physical reality in the model, as described by Hawking in his technical papers, is a classical universe that begins at a minimum radius, inflates, and then goes over to a normal Friedmann expansion, reaching a maximum radius and recollapsing to a singularity.13 In fact, the reality is that Hawking failed to avoid a temporal beginning to the universe. This conclusion is reinforced by a theorem of Alexander Vilenkin and others, according to which, under quite general conditions, the space-times of expanding universes are incomplete in null and time-like past directions, i.e. the paths of photons and test particles are finite in the past. The universe must therefore have a beginning. Vilenkin’s theorem applies to all the variant cosmological models presently on offer, including multiverse models which I shall describe later.14 Ironically, Vilenkin reaffirmed the results of his theorem at a meeting in Cambridge in early 2012 to celebrate Stephen Hawking’s 70th birthday, as reported by Lisa Grossman in New Scientist.15 Vilenkin is reported as saying, ‘All the evidence we have says that the universe had a beginning.’16 In reviewing the kalām argument in the light of Vilenkin’s theorems, Christian physicist Peter Bussey concurs: ‘If the summary of the current position by Vilenkin and others is correct, cosmologies with an infinite past history are not easily viable at present, and so the universe “probably” had a beginning. Therefore the Kalam argument would seem to hold.’17 What the kalām argument then demonstrates, says Bussey, is that a non-temporal and non-physical First Cause for the universe is ultimately unavoidable.

Natural theology and modern cosmology 41

Glass and McCartney’s questions to the no boundary proposal In Glass and McCartney’s notation, suppose the evidence E is the existence of the universe, our initial hypothesis HI is divine creation, and our alternative hypothesis HA is the no boundary proposal. Question 1. Is the no boundary proposal and/or lack of a beginning to the universe incompatible with divine creation? The answer is clearly ‘no.’ As long as imaginary time is a coherent concept (which is far from clear), God can make a universe in which time becomes imaginary. God can equally well create a universe with a history that goes back infinitely far in time, again if that is a coherent concept. Question 2. How likely is it that the no boundary proposal or a universe without a beginning would exist apart from divine creation? The question here is, ‘Why is there a universe at all? That is what is addressed by the Christian doctrine of creation—much more than the question ‘who lit the blue touch paper to set it off?’ which is how Hawking saw it and how others still mistakenly see it. I shall address that question in more detail shortly. Here, it is sufficient to note that the no boundary proposal starts with the existence of the universe, even if it does so as a 4-space rather than the universe of three spatial dimensions and one of time that we are familiar with but into which it supposedly transmutes. Thus, in principle it cannot explain the existence of the universe. What it might conceivably do is explain away the apparent beginning to the universe, as derived from the standard FLRW (Friedmann-Lemaître-RobertsonWalker) Big Bang models, if E were taken to be the beginning of the universe rather than the universe’s existence. However, even this is highly problematic, to say the least, from the above discussion. Question 3. Is the no boundary proposal/absence of a beginning known to be true? Or how strong is the independent evidence for it? The no boundary proposal is ingenious, but highly speculative and controversial. It is not known to be true and there is no independent evidence for it. Question 4. Does the no boundary proposal/absence of a beginning to the universe depend on divine creation? The answer here is similar to that to question 2. Essentially, we are faced with two alternatives: either the universe exists as a brute fact, whether it has a beginning or not, or the universe exists as a divine creation. As we shall see, God as necessary being explains why there is a universe, which is contingent. Richard Swinburne also argues that God is simple

42  Natural theology and modern cosmology compared with a complex physical universe and therefore a priori more probable.18 And, as we shall see when discussing the fine-tuning, the universe could have been different in infinitely many different ways, so the question arises, ‘Why this particular universe?’ Question 5. Is the existence of the universe taken as evidence for divine creation? Yes, but if the kalām argument is undermined, which is not the case at present (see Bussey’s comments above), the broader argument for creation still stands. In biology, a form of design argument is undermined by the theory of evolution, namely separate design of individual organisms. However, design of the laws of nature that underlie evolution is not undermined. Here, if there were good reason to deny that the universe had a beginning, then one form of cosmological argument would be undermined—namely, the argument from a beginning to the universe. However, the argument of Thomas Aquinas, and the Leibnizian cosmological argument from sufficient reason, are not undermined.

Quantum vacuum creation Another way, which atheist Lawrence Krauss, as well as Hawking before him, sees as explaining away the need for God is creation out of the socalled ‘quantum vacuum.’ In quantum theory, the vacuum is not just empty space but a seething hive of activity with particles spontaneously coming into existence and then annihilating. In his book A Universe from Nothing,19 Krauss redefines the notion of nothing so that the quantum vacuum can be identified with nothing. This really is sleight of hand. To philosophers, ‘nothing’ is the absence of anything at all. The quantum vacuum is a hive of activity with space acted on by quantum fields to produce particles and their anti-particles, and acted on by gravity. ‘Nothing’ seems to be a very complicated something with all kinds of properties, including, Krauss says at one point, the property of being ‘unstable.’20 And even if gravity were to do the trick because it has negative energy to cancel out positive matter energy, as is also claimed, we really would be entitled to ask where gravity came from in the first place. Krauss et al. are as confused as the King in this incident from Through the Looking Glass: “I see nobody on the road,” said Alice. “I only wish I had such eyes,” the King remarked in a fretful tone. “To be able to see Nobody! And at that distance too.” . . . “Who did you pass on the road?” the King went on, holding out his hand to the Messenger. . . . “Nobody,” said the Messenger. “Quite right,” said the King: “this young lady saw him too. So of course Nobody walks slower than you.” “I do my best,” the Messenger said in a sullen tone. “I’m sure nobody walks much faster

Natural theology and modern cosmology 43 than I do!” “He can’t do that,” said the King, “or else he’d have been here first.” (Lewis Carroll, Through the Looking Glass, ch. vii)21 Just as the King mistakes the absence of any person for a person called Nobody, so Krauss et al. ontologize the concept of ‘nothing,’ turning it from the absence of anything at all to a very sophisticated something. Well, this nothing of Krauss, Hawking, and others is certainly not nothing in the sense that it would trouble a theist. These scientists simply have not proposed a model in which the universe arises spontaneously from literally nothing without the need for God. Glass and McCartney’s questions to quantum vacuum creation Now suppose the evidence E is the existence of the universe, our initial hypothesis HI is divine creation, and our alternative hypothesis HA is quantum vacuum creation. Question 1. Is quantum vacuum creation incompatible with divine creation? The answer is clearly ‘no.’ God can create through a quantum vacuum. If the idea is correct, then that would simply be God’s way of creating. Question 2. How likely is it that quantum vacuum creation would result in a universe in the absence of divine creation? Creation via a quantum vacuum actually assumes the existence of a universe in the first place. Therefore, in fact, it provides no explanation for why there is a universe! Again, the real choice is between a universe as a brute fact, with no explanation, and a universe created by God. As with the no boundary proposal, what HA might explain is how the universe we are familiar with arose from a previous state. Question 3. Is creation through a quantum vacuum known to be true? Or how strong is the independent evidence for it? This too is a speculative idea. We know that particles are created and annihilated in the vacuum but this requires a runaway effect to create a whole universe. It is not clear how it can be verified, since quantum vacuum creation precedes the so-called ‘inflationary’ era. This is the idea that the universe expanded at an exponential rate in the first 10−32 seconds of its existence, and the problem is that inflation ‘forgets’ its initial conditions.22 Question 4. Does quantum vacuum creation depend on divine creation? As in question 2, we are really faced with a brute fact universe or a created universe. Also, again as we shall see when discussing fine-tuning, inflation needs to occur for just that tiny fraction of a second and then to be

44  Natural theology and modern cosmology turned off to give a universe like ours, so there are many possible ways in which quantum vacuum creation could have gone. Question 5. Is the existence of the universe taken as evidence for divine creation? Yes; and in this case quantum vacuum creation has no effect on the cosmological argument, because, despite what its proponents say, it does not result in a universe which creates itself out of nothing.

The Christian doctrine of creation Clearly the notion that the universe had a beginning is troublesome for atheists. However, even if any of these theories were true, Hawking, Krauss, and Fred Hoyle with his steady-state theory before them would all be mistaken in seeing God as just the cause of a temporal beginning to the universe. The main lesson to draw from the doctrine of creation out of nothing is that the universe is totally dependent on God for its existence moment by moment, continuously. Furthermore, creation is not confined to, or even dependent on, a first moment. As emphasized by, for example, Janet Soskice, the doctrine of creation embraces the universe’s dependence on God yesterday, today, and every day, not just some moment 13.8 billion years ago.23 Thus God does not light the blue touch paper at the Big Bang and then absent himself ever after. In Scripture Christ the Son of God is described as upholding the universe and sustaining it. Were he to cease doing that, the universe would collapse into nothingness. No; God sustains the universe in being, through the Son according to Colossians 1:17 and Hebrews 1:3, and God interacts with his creation bringing about his purposes within it. We might call this ‘continuous creation,’ creatio continua, not in the sense of Hoyle’s scientific theory of continuous creation, though that would be compatible with it. St Thomas Aquinas in the thirteenth century recognized that God would be the cause of the universe’s existence even if it had no beginning in time. He thinks that it can neither be proved nor disproved that it had a beginning, but he himself believed it does from Genesis (Summa Theologiae 1a.46.2), although, in fact, Genesis is not clear cut on this. The Christian doctrine is much more correctly considered as an answer to the question ‘Why is there anything at all?’ or ‘Why is there something rather than nothing?’ which I shall discuss in the next section than about the how of a temporal beginning. For Aquinas, as for Leibniz later, there can be an infinite chain of causes going back in time, but that infinite chain needs a cause for its existence. And God provides the first cause because he himself exists by necessity as I am about to explain. Aquinas did believe in a beginning of the universe,

Natural theology and modern cosmology 45 because that is how he read Scripture, but his argument is framed in logical, not temporal, terms.

Ultimate explanations—Why is there something rather than nothing? By an ultimate explanation, I mean an answer to the question ‘Why is there something rather than nothing?’ or ‘Why is there any universe at all?’ We have seen that, despite what they say to the contrary, both Lawrence Krauss and Stephen Hawking failed to answer this. The fact is that no scientific theory can answer it. Only God can provide the ultimate explanation. There is a universe because God freely created it. He wanted to bring about an environment in which free, rational creatures could flourish and have a relationship with him. Aquinas and many other theologians since have argued that it is the idea of God as ‘necessary being’ which provides a stopping point for explanation. To say that God is necessary means that he cannot but exist. He must exist. He cannot not exist. This is at least part of what the concept ‘God’ means. Another way of saying it is that there is no possible universe in which God does not exist. It follows from this that God was not himself created since, if God had been created, there would have been a time when God did not exist but something else did, namely whatever or whoever created God. Anything created is not God. Someone could, of course, doubt that such a being exists—many do doubt it—but it follows that if he does exist, then he has always existed and will always exist and everything else that exists depends on him. That is because everything else is ‘contingent.’ The word ‘contingent’ means the opposite of necessary. Something which may or may not exist is contingent. It did not have to exist. It might not have existed. And even if it exists, it could be different from what it is. God and the universe are thus completely different. Unlike God, the universe might or might not have existed. Hawking put this very eloquently himself back in A Brief History of Time, when he wrote: ‘What is it that breathes fire into the equations, and makes a universe for them to describe?’24 That is the fundamental question. One can have the most wonderful theory—it might be Hawking’s favoured M-theory (of which more will be said later), or whatever—but the question is, ‘Why is there a universe in which that theory is instantiated?’ Science is powerless to explain why the universe exists. The universe cannot explain its own existence. It cannot create itself, by lifting itself up by its own bootstraps, as it were, into existence. However, theism can explain why there is a universe and it can explain the particular character of the universe—it was created by God, who freely chose to make this particular universe with all the properties needed for it to produce life. And that brings me on to the second area where cosmology interacts with theology.

46  Natural theology and modern cosmology

Specialness of the Big Bang: cosmic fine-tuning Modern cosmology tells us that the universe is set up in a very special way indeed, seemingly in order for us to be here to observe it. This specialness relates to two areas: 1 First, the conditions right back at the beginning, shortly after the Big Bang, need to be just right to high degrees of accuracy for the universe to give rise to life. 2 Secondly, the constants which go into the laws of physics need to take the values they do, in order for the universe to give rise to life. These constants determine the relative strengths of the four fundamental forces of nature, namely gravity, the electromagnetic force which holds atoms together, the weak nuclear force responsible for radioactive decay, and the strong nuclear force which binds atomic nuclei together. They also include such quantities as the masses of the fundamental particles. They determine how key physical processes go at different stages of the universe’s evolution. There are many, many examples of this so-called fine-tuning, and I will give just one of each kind now: 1 First, at the very beginning a mere one second after the Big Bang the mean density of the universe has to be just right to 1 part in 1015. If it is smaller than it is by this amount then the universe will expand far too quickly for galaxies and stars to be able to form. If it is greater, then the whole universe will recollapse under gravity long before there has been time for stars to evolve. Either way, one has a boring universe with no possibility of life. If one naïvely extrapolates back to the earliest time we can speak of, 10−43 seconds, then an accuracy of 1 in 1060 is required. 2 Secondly, the strong nuclear force has to be of just the right strength for carbon and oxygen to be made inside stars. One of the great discoveries in astrophysics is how the chemical elements are manufactured inside stars, where the temperatures reach hundreds of millions of degrees, through nuclear reactions. Sir Fred Hoyle, the atheist Cambridge astrophysicist I mentioned before, was foremost in this discovery; it was he who discovered the particular ‘coincidences’ required for carbon to be made in the first place, and then for the carbon not to be destroyed in making oxygen. When he made this discovery he was moved to remark that ‘a superintellect had monkeyed with physics, as well as with chemistry and biology, and that there are no blind forces worth speaking about in nature.’25 This is a man who earlier in his life described religious belief as illusory.26 As I say, there are a host of these examples of fine-tuning which I cannot discuss in detail, although some more examples will crop up later. The cosmologist Paul Davies puts it like this: ‘Like the porridge in the tale of Goldilocks

Natural theology and modern cosmology 47 and the three bears, the universe seems to be “just right” for life, in so many intriguing ways.’27

Explanations for the fine-tuning This specialness of the universe, which is essential if there is to be life, just cries out for explanation. The most obvious explanation is the theistic one, that it was made that way; it was designed so that life would appear. Christians would say that God intended there to be living creatures with the capacity for reason and with free will, who would be able to have a relationship with him. Many scientists, however, regard any kind of design hypothesis, even this one, with loathing. They want to restrict their explanations, even for why the laws of physics are as they are, to within science itself. So what alternatives have scientists come up with? I’m going to contrast two strategies which scientists have pursued in order to explain away the need for design by God. 1 The first is to seek an explanation from within science for the values taken by the various constants of physics—to derive them from some more fundamental theory, a so-called ‘theory of everything’ (TOE). Interestingly, Einstein spent his later years in a fruitless search for such a theory: ‘What I  am really interested in is whether God could have made the world in a different way,’ he said—although this quote obviously indicates that he still saw no contradiction with God being behind it all. Connected with this search for a TOE, though different from it, is the aim to show that the initial conditions are not special: to argue that whatever they were, the universe would turn out much the same. 2 The second strategy is diametrically opposed to this. It is to postulate a multiverse. A multiverse is a vast, usually infinite, ensemble of existent universes, embracing the whole range of values of the constants and initial conditions. The idea is that if a multiverse exists you can then say: ‘Hey presto! Given the vast ensemble, our universe with its suite of parameters is bound to exist, and we should not be surprised to find ourselves in it, because we simply could not exist in the overwhelming majority of universes which differ from ours in their parameter values to the slightest degree.’ Strategy 1, then, explains why the universe is like it is, given that it exists. It could not have been different; so, with the big proviso that it exists, then it is necessarily the way it is. If that is so, then there is still a massive puzzle because we can now ask, ‘Why does the only self-consistent set of physical laws give rise to life?’ It could have given rise to an isolated amorphous lump of rock, or a few isolated particles floating about in otherwise empty space, and nothing else. Is it not utterly astonishing that it should give rise to a universe with all the

48  Natural theology and modern cosmology rich complexity including living creatures that we see? Given the infinite variety of outcomes we can imagine, it is desperately puzzling why the only possible set of laws gives a universe with human beings in it. So, given this massive puzzle, this seems a not very convincing explaining away of the need for divine design. There is a less all-embracing version of strategy 1, namely that the fundamental constants relating to the laws of nature may be derivable from a more fundamental theory, but there exist alternative fundamental theories. For God to have no choice, as Einstein speculated, there would have to be just one unique fundamental theory. But if there are alternative theories, then, of course, we have to ask the question why the particular fundamental theory which applies to our universe does so, and not one of the other theories; so this is very far from explaining away the need for divine design, which would amount to God choosing which among the possible theories to instantiate. Coming to strategy 2, the multiverse hypothesis says that the universe certainly can be different (the physical constants could indeed take other values) and indeed different universes actually exist. And it could be the case that the more universes you have the more chance there is of getting one with life. But there is a pretty big puzzle here too; namely, ‘Why does this particular multiverse exist as opposed to another?’ We now have a choice of equations into which fire somehow gets breathed, and we have a choice about how many sets of equations give rise to universes and how many universes they give rise to. What determines these choices? In answer to this, one cosmologist, Max Tegmark, has proposed that all possible mathematical structures have physical existence (his ‘Level IV multiverse’).28 This would certainly guarantee our universe’s existence, though totally to avoid the ‘why this universe (or multiverse)’ question one really needs to hypothesize that all universes, mathematically structured or not, exist. Either way, this takes us way beyond what physics can tell us, and some mathematicians and physicists have questioned whether Tegmark’s idea even makes sense. One soon runs into problems and paradoxes when one actually starts to try and write down ‘all possible mathematical structures.’ Certainly there seem to be conflicts in what actually exists, as opposed to what can possibly exist. For example, I cannot simultaneously deliver a talk on ‘Explaining and Explaining Away in Cosmology and Theology’ in Belfast, on which this chapter is based, and remain sitting at home watching TV. Some copy of me in another universe could conceivably have taken a different course, but I could not simultaneously do both.

Recent developments in cosmology A widely accepted way to solve some of the problems with the standard Big Bang, and the fine-tuning, is the theory of inflation, first proposed by Alan Guth at a seminar in 1980.29 This postulates that the universe underwent an

Natural theology and modern cosmology 49 incredibly rapid period of accelerating expansion—called inflation—from 10−35 to 10−32 seconds after the origin. In that time the universe expanded from, give or take a few orders of magnitude, 10−25 cm to 10 metres across. At that point, the much slower deceleration of the classical Big Bang took over. Now it is the case that such a rapid period of accelerating expansion, even if that short, drives the density of the universe to the critical value and smooths out the differences between different parts of the universe, ostensibly solving what was also a problem with the standard Big Bang. That sounds wonderful, but there were some serious problems with inflation. One serious problem from our point of view is that inflation itself needed fine-tuning, that is, parameters to be chosen specially! That is not very satisfactory for a theory which was meant to solve the problem of the need for fine-tuning. The next step was to propose that inflation occurs in some places and not others and at different rates in the different parts of the universe where it does occur. The parts where inflation does occur would ultimately swamp the small non-inflating regions. This represents a turn to strategy 2, namely a multiverse with different regions having different parameters and these regions being regarded as distinct universes. This picture was proposed by Andrei Linde of the University of Stanford, California, and is known as chaotic inflation. Another variant is eternal inflation, in which infinitely many different bubble universes are formed by inflation with bubbles forming within bubbles ad infinitum.30 Now we are still not quite at the TOE. That is the theory which is said to apply to the very first 10−43 seconds from the origin. During that time, one needs a theory which combines all the forces of nature. That is to say, it combines Einstein’s general theory of relativity, which is the theory of gravity, with quantum mechanics, which applies to the other forces and describes the very small. We do not know what that theory is, but the leading contender is string theory, one of whose main pioneers has been Leonard Susskind. M-theory is a generalization of string theory. String theory postulates that the ultimate building blocks of matter are not point-like particles but tiny, onedimensional objects called strings. By tiny I mean really tiny, some 10−33 cm across. String theory aims to solve some of the problems with the standard model of particle physics, especially the existence of infinite quantities like mass and charge. The elementary particles we observe are actually different modes of vibration of the strings. An important complication is that these vibrations occur in more than the three dimensions of space that we are used to. The reason we only see three extended dimensions is that these other dimensions get curled up very small. Quite why this is so remains something of a mystery. The original aim of string theory was to calculate particle masses, that is, strategy 1 was pursued. The theory has always been dogged by its lack of connection with observation and experiment, so the main motivation

50  Natural theology and modern cosmology has been that it is beautifully mathematically elegant and it solves some theoretical problems. It is still the aim of some string theorists to calculate everything and some believe that is possible in principle, although some parameters (like the cosmological constant) still seem to need strategy 2. Nevertheless, nothing has been calculated in practice, so some string theorists, notably Leonard Susskind, have turned to strategy 2. Susskind and his colleagues talk about the ‘landscape of string theory.’31 They find that there is not just one but many solutions of the theory, anything from 10500 to 101000 solutions. The further claim is that a universe can ‘tunnel’ between solutions. The solutions are stable for billions of years, then another universe pops up as a region moves to another solution of the equations. This feeds in very neatly to the eternal inflation idea. If it works, and it is a big if, eternal inflation would be the means whereby the string theory landscape is populated. It is also true that if there is a theory which in some sense naturally gives rise to many universes, then that gives plausibility to the idea of a multiverse. Susskind clearly sees the string landscape as explaining away the need for God to explain the fine-tuning: ‘If there is a God, she [sic] has taken great pains to make herself irrelevant.’ Echoing Laplace he declares, ‘I have no need of that hypothesis.’ Susskind does, however, see that string theory makes no difference to the cosmological argument, since just before the above statement he writes: The ultimate existential question, “Why is there Something rather than Nothing?” has no more or less of an answer than before anyone had ever heard of String Theory. If there was a moment of creation, it is obscured from our eyes and our telescopes by the veil of explosive Inflation that took place during the prehistory of the Big Bang.32 There are a number of other multiverse models out there, which puts me in mind of a remark once made by the great Russian physicist Lev Landau that ‘cosmologists are often in error, but seldom in doubt.’ With the proliferation of models and lack of observational constraint, it seems there is a considerable degree of truth in this. Of course, the idea of there being a multiverse is in any case perfectly compatible with divine creation. Indeed, some Christian scientists, theologians, and philosophers embrace the idea, as we shall now see.

Why might Christians welcome multiverses? Let me briefly look at three Christians who believe in both God and a multiverse. One is a scientist-theologian, the second a philosopher, and the third a cosmologist. Arthur Peacocke was a scientist-theologian who favoured a multiverse. What counts is the potentiality of the whole ensemble of universes to

Natural theology and modern cosmology 51 produce cognizing subjects, not that of one universe in particular.33 The finetuning argument still applies given that the multiverse of which our universe is a member has the right parameters for persons to evolve somewhere in it, in fact in our particular universe at least. Robin Collins is a philosopher who certainly thinks that the fine-tuning is real and in need of explanation, and in an earlier book seemed to treat the design and multiverse hypotheses as alternatives.34 But in the later volume on multiverses, Universe or Multiverse?,35 edited by Bernard Carr, he takes a similar view to Peacocke. On the theistic hypothesis, the creation would reflect God’s infinitely creative capacities, so physical reality might be much larger than a single universe. The idea would be that God expresses his infinite creativity, rather than simply and purposively creates a single universe with life in it. Collins thinks that creating a multiverse through a single physical universe-generating mechanism, as seems to be on offer with inflation, perhaps as combined with the landscape of string theory idea, would be a befittingly elegant way for God to do this. Moreover, if there is a universe-generator, then it needs design, just as a single universe needed to be designed if it were to give rise to life. On this view, the fine-tuning problem is merely shifted from the universe to the multiverse. However, more recently, Collins has argued that a multiverse does not in fact solve the fine-tuning problem, even if the multiverse turns up fine-tuned universes. As I shall explain shortly, this is essentially because our universe is not just fine-tuned but extra specially fine-tuned, in an important and well-defined sense. A cosmologist who is both an evangelical Christian and a multiverse proponent is Don Page, who was one of Stephen Hawking’s collaborators over many years. Page favours the Everett version of the multiverse, in which a new universe arises corresponding to every possible outcome of a quantum measurement.36 But he also argues that God might prefer an elegant theory such as string/M-theory without free parameters that would lead to a multiverse but with the deliberate intention that there be life somewhere within it. Both Page and Collins argue that the beauty and elegance of that theory might be an argument for its design by God.

Problems for multiverses The whole idea of multiverses, including the latest string landscape idea, is fraught with problems; here I shall just list a few of them.37 1 It is important to recognize that the physics is speculative, to say the least. The trouble with many universes is that they cannot even in principle be observed. They cause no effect whatever in our own universe, because no signal from them can ever reach us.   Martin Rees is one of Britain’s most distinguished cosmologists. In one of his books he describes himself as a ‘cautious empiricist’ who

52  Natural theology and modern cosmology starts to feel at home when familiar physics can be applied to the universe, which he says is the first thousandth of a second from the origin and later.38 However, in another book he expresses his preference for a multiverse over design, even though he describes the multiverse idea as ‘highly speculative’ and his preference ‘no more than a hunch.’39 The physics that would yield multiverses applies not to one thousandth of a second after the origin, but the first 10−32 seconds or even the first 10−43 seconds. It is a quite interesting example of an ideologically driven rather than evidence-based preference. Like Susskind, Rees recognizes that science cannot answer the question why there is anything at all; but, also like Susskind, he opts for a multiverse as apparently explaining away the need for God as designer. 2 There is a problem about the existence of actual infinities in nature. Mathematicians happily talk about and manipulate different degrees of infinity, but paradoxes arise when one starts to think about infinite numbers of things existing in the real world. One well-known paradox was formulated by the mathematician David Hilbert. ‘Hilbert’s Hotel’ has infinitely many rooms, all of which are full. Even so, the receptionist can very easily make room for infinitely many more guests! All she has to do is tell the person in Room 1 to move to Room 2, the one in Room 2 to go to Room 4, the one in Room 3 to go to Room 6, and so on. Then all the even-numbered rooms are full, but the odd-numbered ones are all free. Thus infinity is never complete but can always be added to.   One problem is that, if there are infinitely many regions with varying parameters there will be infinitely many identical copies of me. There will also be copies who differ very slightly. As noted earlier, some of the “Is” will fly over and talk about multiverses in Belfast, whereas others will decide to stay at home and watch TV instead. It is quite bizarre even to begin to think about this. Some philosophers and mathematicians think infinitely many universes are ruled out because of the paradoxes. I do not quite see the paradoxes as logically precluding them, but a theory without paradoxes is surely to be preferred. 3 The multiverse hypothesis is not a simple hypothesis. Scientists normally opt for the simplest of competing hypotheses, and this does not seem to be that. The principle of Occam’s razor tells us that we should not multiply entities needlessly. As I said earlier, another question one needs to ask is, ‘Why this multiverse?’ That applies to the string landscape idea as much as any of the others; even to produce the landscape some choices within string theory have been made. 4 In any case, the turn from strategy 1 to strategy 2 implies a move away from predictability, which had been a cornerstone of the scientific method. This is not just predictability of physical parameters, but predictability in general based on the existence of order in the universe. Suppose some unexplained feature arises in the universe. Instead of trying to explain it rationally using ‘normal’ science, the temptation is now to say,

Natural theology and modern cosmology 53 ‘We just happen to be in a universe which exhibits that feature.’ Such theories are not falsifiable. They also encourage a ‘multiverse of the gaps’ attitude akin to that of some Christian apologists who invoke a ‘God of the gaps’ to explain phenomena presently unexplained by science.   Matters are even worse than this. As philosopher Tim Mawson points out, for intelligent life to operate we need the principle of induction to work, whereby the future resembles the past.40 Mawson draws on Nelson Goodman’s famous ‘new riddle of induction’ whereby observations of emeralds which turn out to be green support both the hypothesis that ‘All emeralds are green’ and the hypothesis that ‘All emeralds are grue,’ where ‘grue’ means ‘green up until 31 December  2050 (or any other arbitrary time in the future) and blue thereafter.’41 Mawson notes that the ‘maximal’ multiverse in which all possible universes exist is the only version of the multiverse to avoid the ‘which multiverse’ question and thereby guarantee that our fine-tuned universe exists. This is akin to Tegmark’s Level IV multiverse (see above), in which all mathematical structures are instantiated in physically existing universes, though even more general and more like David Lewis’s modal realism with regard to universes.42 The problem is that in the maximal multiverse there will be infinitely many universes in which the inductive procedures of intelligent creatures operated up till now, but collapsed at some later point, completely overwhelming the number of universes in which induction continues to work indefinitely. 5 Possibly the most outstanding problem in cosmology is the fine-tuning of the cosmological constant, Λ. This is the term originally introduced into his equations by Einstein to make the universe static. If he had put it to zero, he would have predicted the expansion and arrived at the Big Bang theory.   Until relatively recently it had been thought that Λ was zero. The latest observations indicate that Λ probably takes a very small, but positive, value.   Now, physicists think they know where Λ comes from; namely, the energy of the quantum vacuum. The unfortunate thing is that when Λ is calculated, it gives a value 10120 times that which is compatible with observations. If Λ really took the calculated value, our bodies would be pulled apart in an instant, with body parts flying away to the ends of the universe.   The answer cosmologists have come up with to this one? By now the reader will not be surprised to hear that it is a multiverse. And in the string theory landscape, the different universes represent different values of Λ. If a universe starts with a very high value of Λ, it will spawn billions upon billions of universes until a universe eventually arises with the small value of Λ that our universe has.   This looks like a great success. But now there is another question we need to ask. According to the multiverse theory, the universe should

54  Natural theology and modern cosmology be regarded as typical of those with Λ values which permit life. It is a random member of the subset of universes which give rise to life. The question then is, ‘Does it look like it is that, or is it more special than that?’   Now, calculations show that the average value of Λ which would be compatible with life is quite a bit more than the value we observe. The first calculations showed that it could be a hundred times more; that figure came down with further calculations, but it still looks a bit too high. On the multiverse hypothesis, one would expect to observe a value nearer to that upper limit, not much lower. That the value we observe is lower than expected has been acknowledged by Steven Weinberg, who did the calculations.43 Paul Davies also thinks the value is rather too low to be ‘minimally biophilic,’ as would be expected on the basis of a multiverse. He says that it may even be ‘optimally biophilic,’ that is, take the value most compatible with life.44 Thus we seem to be observing a value of Λ that is a bit too special, though not enormously so by astronomical standards.   Of course, there could be many other parameters of our universe besides Λ that are more highly tuned than is strictly required for our own existence. It looks as though there are; and I shall return to one of them in a moment. 6 Some multiverse models require an element of fine-tuning for there to be a multiverse in the first place. An example is that the overall mean density must be less than or equal to the critical value, so that the universe as a whole is infinite and expands for ever. And that may not be likely, given that in principle the density can take any value from an enormously large range. It might well be greater than the critical value, in which case the universe is not infinite, but finite.   In fact, we could never be sure that we really inhabited an infinite universe. John Barrow makes just this point.45 Either of two options is possible. We may think we are in an infinite universe when we just inhabit an underdense part of a finite universe; or we think we are in a finite universe when we inhabit an overdense part of an infinite universe. 7 Roger Penrose poses a massive problem to inflation, and indeed all attempts to explain the specialness of the Big Bang on the basis of a multiverse.   Penrose is concerned with the amount of order there was at the beginning. Order is measured by a quantity called entropy; entropy increases over time and order correspondingly decreases. Penrose puts it like this concerning the entropy of the universe. He says that the Creator 123 had something like 1010 possible universe configurations to choose from, only one of which would have the order that ours does.46 That is the order necessary to produce a cosmos with all the galaxies, stars, and planets that our universe possesses. Note that 10123 is 1 with 123 123 noughts after it; 1010 is 1 with 10123 noughts after it.

Natural theology and modern cosmology 55   Now Penrose points to the fact that, for a universe to have life, you actually need a great deal of order but much less than this vast amount.47 You could create the entire solar system with all its planets and all its inhabitants by the random collisions of particles and radiation with a 60 probability of 1 in 1010 . This is a tiny probability, but much greater 123 than 1 in 1010 . The implication is that our universe is vastly more special than required merely in order for us to be here. It is much, much more special than a universe randomly selected from the subset of universes which are conducive to life. This is a very serious challenge for the multiverse idea, but totally consistent with design.   To summarize this point, if the multiverse explanation is correct then we ought to be in a universe with parameters just right for us but not vastly too special. The cosmological constant looks close to meeting this criterion, but the initial entropy of the universe fails catastrophically. There are other parameters, such as the constancy of the charge on the electron and the lifetime of the proton, which also look much too finetuned, again posing a problem for the multiverse hypothesis. A similar argument to this one is made by Robin Collins, but I leave that to the next section. Mawson’s point about our universe being extra-special in that induction works, noted in point 4, is also highly relevant here.

Glass and McCartney’s questions to multiverse theories In this case we take the evidence E to be the fine-tuning, our initial hypothesis HI divine design, and our alternative hypothesis HA the existence of a multiverse. Question 1. Is a multiverse incompatible with design? The answer is clearly ‘no.’ Many Christians favour a multiverse. For someone like Peacocke, this would just be God’s way of ensuring that some subset of universes produced ‘cognizing subjects.’ In a similar way, if there were only one universe with our physical laws and parameters, God could have created many planets to ensure that intelligent life arose on at least one of them. Even on this one planet God could have created (did create, if one is a theist) a process which gave rise to many different creatures around the world, but ensured that eventually an intelligent, rational, self-conscious, and morally aware species would arise. The multiverse would just be an extension of that idea. Question 2. How likely is it that a multiverse would result in fine-tuning in the absence of design? There are many possible multiverses, just as there are many possible universes. If a multiverse has a finite number of member universes, one has to ask: ‘If at least one is fine-tuned, then why are there sufficiently many

56  Natural theology and modern cosmology universes to ensure that?’ Even a multiverse with an infinite number of members far from guarantees that any member universes are fine-tuned. One always has to ask, ‘Why does that version of the multiverse, with just those physical laws, giving rise to just the right parameters for life in one or more of the universes, exist, as opposed to the myriad others?’ The only way to get around this problem of choosing between multiverses is to say that all possible multiverses exist—supposing, doubtfully, but for the sake of argument, that this idea is coherent. This would be something like Max Tegmark’s Level IV multiverse of all mathematical structures existing as real universes, or, a fortiori, Mawson’s maximal multiverse.   A Tegmark Level IV multiverse—better still, the Mawson maximal multiverse—would ostensibly explain away the need for design, though I now argue that even the Tegmark Level IV multiverse has a fatal flaw in that regard. Indeed this is a problem for every multiverse model on offer. And this leads me to give a different answer to question 2 from that which Glass and McCartney provided in their Theology and Science paper.   As Robin Collins points out, for the multiverse to solve the fine-tuning problem, it needs to be combined with an ‘Observer Selection Principle.’48 If there were no observers, there would be no one to ask the question, ‘How come the universe is observable?’ So that is the fundamental question: ‘Why is the universe observable? How is it that the universe gives rise to observers?’ We should accordingly regard ourselves as typical or generic observers in the multiverse.   The observer selection principle is required so that the multiverse hypothesis will not undercut normal claims of improbability.49 Thus, in a suitably large multiverse some observer somewhere (maybe even some copy of me) will toss a fair coin one hundred times and see it land on heads every time. But a generic observer (such as I should regard myself) will do so with a tiny probability.   Collins goes on to argue that, in fact, the universe is not fine-tuned for observers per se but for the very special kinds of observers that we are— namely embodied conscious agents (ECAs) who can, in addition, ‘significantly interact with each other’ and ‘develop scientific technology and discover the universe.’ For example, the gravitational constant could take any value between 0 and 1038 times its actual value, but above its actual value one starts to hit problems. For ‘mere’ observers, gravity could be a bit more than it is; but when you increase it tenfold, it becomes difficult to lift objects and science becomes difficult. Increase it a bit more, and brains the size of ours get crushed. For a while one can compensate for this by reducing the size of the planet (say up to one hundred-fold), but then it is more difficult for the ECAs that do evolve to develop technology. At a billion-fold, ECAs could not evolve at all. Thus one has maximal fine-tuning for ECAs of about 1 part in 1029, but the probability that gravity is fine-tuned for ECAs with the kinds of capacities we possess to

Natural theology and modern cosmology 57 interact and do scientific experiments, given fine-tuning for ECAs in the first place, is 1 in 107.   Roger Penrose’s argument, which is also made by Collins, is much more devastating here. Our universe is ordered and structured out as far as we can see—with one hundred billion galaxies’ worth of order. But 123 as generic observers, we are 1010 times more likely to be in an isolated solar system surrounded by utter chaos than we are to be in this totally ordered universe.   Of course, a universe like ours has the potential for there being more observers than a single solar system would have. For the sake of argument, one could estimate that there are maybe 1022 times as many observers given that there are 1022 stars in the universe and one assumes each one has planets. But that makes no difference to the argument, since 1022 123 is utterly swamped by 1010 .   The upshot of this line of argument is that a multiverse might be able to explain the existence of ‘Boltzmann brains,’ isolated observers which arise simply from thermal fluctuations of matter-energy. However, such isolated generic observers, although themselves rare, would nevertheless overwhelmingly outnumber ECAs, with the kinds of capacities we have, inhabiting the kind of vast ordered universe we do; and thus the latter would be completely unexplained. Add in Mawson’s criticism of the maximal multiverse, that universes in which induction fails to work overwhelmingly predominate, and the whole multiverse edifice looks utterly incapable of explaining the fine-tuning that is required. Question 3. Is the multiverse hypothesis known to be true? Or how strong is the independent evidence for it? The multiverse hypothesis is certainly not known to be true. It is not even clear whether it can ever be known to be true—and certainly not Tegmark’s Level IV multiverse or Mawson’s maximal multiverse. All multiverse models are plagued by the fact that they can never be directly observed. If one had a well-established theory which in some sense naturally gave rise to multiple universes, that would make the hypothesis more persuasive. However, most multiverse models depend on contentious and speculative physical theories supposedly applying in régimes far beyond experimental and observational capability (e.g. applying to the first 10–43 seconds from the origin at the Big Bang). The best hope is that some form of inflation is true, but at present inflation looks more like a generic set of theories (with at least 111 different versions according to Paul Shellard),50 rather than a unique theory. Originally, it was meant to apply to the Grand Unified Theory (GUT) era when the strong force was united to the electroweak force. The failure of GUTs has led instead to the postulation of ad hoc ‘scalar fields’ which do not sit easily with fundamental theories.51

58  Natural theology and modern cosmology Question 4. Does the multiverse hypothesis depend on design? Unless one adopts the wildly anti-Occamite maximal multiverse, a particular multiverse has to be chosen with the right properties for at least one of its members to produce intelligent life; hence the need for design is not negated. As Collins says, one needs a designed ‘universe-generating mechanism.’ Even then, it fails to explain the extra-special fine-tuning of our universe for us, as we saw in answer to question 2. Question 5. Is the fine-tuning taken as evidence for design? Yes; most cosmologists recognize the fine-tuning as requiring explanation, and that design is an option. Some physicists, philosophers, and theologians—including Richard Swinburne, Robin Collins, John Polkinghorne, and myself—opt for design as the best explanation. Some, as we have seen, opt for the multiverse as an alternative.

Comparing the explanations So how do we choose between the multiverse explanation and design? I have just listed a host of problems with the multiverse explanation in addition to the fact that no purely physical explanation will ever be ultimate. In contrast, creation and design by God does provide an ultimate explanation because God, if he exists, exists necessarily—that is at least part of what we mean by ‘God.’ In addition, design by God is a simple explanation, and much more economical than the multiverse. One is not invoking a whole multitude of complex entities with which one can have no possible interaction, but one intelligent being—like ourselves in some ways but so much greater: a being of unlimited power and knowledge (omnipotent and omniscient in classical terms), and perfectly good. Most importantly, God does explain the fine-tuning that is required for the existence of intelligent creatures able to interact with each other and comprehend, in some measure, his handiwork. Out of all possible universes, God freely chose to create this particular universe with the deliberate intention of its bringing forth creatures for a relationship with himself, and so is likely to have created an ultra-fine-tuned universe for this purpose.

Appendix: Bayesian evaluation of the arguments

In Chapter  2 we saw how Bayesian confirmation theory can be used to evaluate scientific hypotheses and, as particularly in the work of Richard Swinburne, metaphysical hypotheses. Let us see how this might work out for the hypotheses we have discussed in this chapter. We begin with the cosmological argument. Recall Bayes’s theorem as given in equation (2.1) in Chapter 2: P  H|E 

P  E|H .P  H 

P  E|H .P  H   P  E| ~ H .P ~ H 

(3.1)

Here let E = the evidence that there is a complex physical universe, by which I  mean (echoing Swinburne) a physical object containing many physical objects which are related to each other both in space and time and characterized by physical laws and perhaps initial conditions.52 Then let H = the hypothesis that there is a God, understood as something like the Christian God who exists necessarily and possesses such qualities as omnipotence, omniscience, and perfect goodness. Then ~H = the hypothesis that there is no such God. ~H is read ‘not H.’ Now ~H does not explain E, why there is a universe, at all. On hypothesis ~H the universe is just a brute fact which we have to accept and it is uncaused. As we explained earlier, the universe is contingent. It therefore does not explain itself. It thus seems reasonable to suppose that P[E|~H] is very low, especially when we add in the consideration that, even if a universe of some kind were, improbably, to exist, then a complex physical universe would be much less likely to exist than a simple one. On the other hand, God as necessary being does explain why there is a universe, i.e. hypothesis H does explain E. We may not be able to say with any precision how likely it is that God would create a universe, though we can think of reasons why God might choose to do so, for example God wanting to exercise his overflowing power and love to create something of great beauty and wanting to create finite creatures who reflect something of his own being and are able

60  Natural theology and modern cosmology to relate to him. Thus it would seem that a universe is much more likely to exist given God as prior cause than if there is no cause. It follows that P[E|H] is much larger than P[E|~H]. Expressing this mathematically, P[E|H] >> P[E|~H]. Remembering that by the laws of probability P[~H]  =  1 – P[H], with some simple algebra we obtain the following result: P[H|E] >> P[H ] (3.2) In other words, the evidence of a complex physical universe renders the existence of God much more probable than if this evidence were not taken into account. This is what Swinburne calls a ‘good C-inductive argument,’ meaning one which enhances the prior probability of God without necessarily bringing the probability of God above a half (the latter would be a ‘good P-inductive argument’).53 The above applies whether the universe has a temporal beginning or not. If, as is probably the case, it does have a beginning in time, it seems to me that this can only strengthen the argument. Indeed, the idea of the universe spontaneously erupting into existence from no prior state without a cause seems hardly coherent. And indeed, this anxiety is quite possibly behind Stephen Hawking’s (mistakenly) thinking that he got around the need for God by doing away with a temporal beginning. At least in the case of a universe with an infinitely long temporal past history each event is scientifically explicable in terms of preceding causes, even if the existence of the whole sequence is not explained. In the case of a universe with a temporal beginning there is also an event which has no scientific explanation in terms of a prior cause. When it comes to the fine-tuning argument, I  believe the Bayesian formulation shows this to be much more powerful than the cosmological argument given the discussion in previous sections. It may even be a good P-inductive argument in Swinburne’s terms. Given that a single universe does not explain the fine-tuning, and the popularity of the multiverse idea, we confine the discussion to a comparison of the latter (without God) and God. Thus we now let E′ = the evidence of the fine-tuning, H = the hypothesis of theism as before, and M  =  the hypothesis that there is a multiverse (it could be any of the multiverse types discussed above) but no God. We utilize the consequence of Bayes’s theorem which we encountered in Chapter  2, equation (2.3): P[H | E ] P[E  | H ] P  H  (3.3)  . P[M | E ] P[E  | M] P  M  Again we may not be able to be precise in estimating the probabilities going into the above equation. However, one thing that the ‘Observer Selection

Natural theology and modern cosmology 61 Principle’ referred to above teaches us is that P[E′|M] is very low. Indeed, on Penrose’s calculation, the probability that we as observers are in a member universe (out of the ensemble of universes comprising the multiverse) as 123 highly ordered and structured as our universe is of the order of 1 / 1010 . 123 If P[E′|M]  =  1 / 1010 then, under any reasonable values for the other probabilities, this term will completely dominate equation (3.3). It is vastly more likely that God would create a completely ordered and structured universe like ours than that this is what we would observe given a multiverse. Given the complexity and contingency of a multiverse, including the number of different ways a multiverse could be realized, there is little reason to think that a multiverse is a priori more likely to exist than God. Indeed, Richard Swinburne argues that God is much simpler and therefore much more likely to exist a priori than a complex physical universe, or a fortiori a complex multiverse.54 However, even if we do think that P[H] is less than P[M], there is no reason at all to think that P[H] is absurdly low in comparison with P[M]. Suppose for the sake of argument that the probability that we observe what we do given God is rather low, say only 1/10, i.e. P[E′|H] = 1/10 in equation (3.3). And even suppose (with Paul Davies55 as opposed to Richard Swinburne) that we are indifferent between theism and the hypothesis of a multiverse, so that P[H] = P[M]. Then equation (3.3) implies that 123 P[H | E ]  1010 (3.4)  P[M | E ]

i.e., given the evidence of fine-tuning, it is overwhelmingly better explained by God than by a multiverse. Any ‘moderate’ probabilities are washed out 123 by the monstrously huge number 1010 . Rewriting equation (3.1) with E′ substituted for E we obtain P  H|E   

P  E |H .P  H  P  E|H .P  H   P  E | ~ H .P ~ H 

(3.5)

If we want to make further progress and derive an absolute value for P[H|E′], using equation (3.5), then we need an estimate for P[E′|~H] rather than P[E′|M]. Here ~H includes all non-theistic alternatives to M, not just M itself. Now the main alternative is that there is just one universe U with no God, but of course the multiverse was introduced precisely because P[E′ |U] 123 is very low, indeed itself of the order of 1 / 1010 . The hypothesis M was introduced in order to increase the probability of fine-tuning, but, as we have seen, although ultra-fine-tuned universes will exist under M, the probability that we as typical observers see ultra-fine-tuning remains extremely low. In fact, to complete the analysis one should also consider other ‘theistic’ hypotheses such as there being multiple gods rather than the one God we

62  Natural theology and modern cosmology have been assuming, the one equating to the God of classical Christian theism. But, as Swinburne argues, the hypothesis of many gods is not simple like the single, unique God hypothesis.56 In particular, it violates the principle of Occam’s razor, that ‘entities are not to be multiplied beyond necessity.’ Moreover, again as argued by Swinburne, the fact that the universe is described by the same laws across the vastness of space and time betokens creation by one God rather than many. The upshot is that it is very difficult to see why, under any reasonable values for P[E′|H] and P[H], P[E′|H].P[H] is not very much larger than P[E′|~H] and hence than P[E′|~H].P[~H]. If this is the case, then, by equation (3.5), the argument from fine-tuning is a good P-inductive argument and makes P[H|E] very close to 1 indeed. To avoid this conclusion the atheist is virtually driven to saying that the concept of God is incoherent, since the prior probability P[H] has to be 123 made so close to zero, i.e. less than 1 / 1010 . This is something which I believe even the staunchest atheist would be hard put to justify.

Notes 1 This chapter, here very slightly revised and with a new appendix, was first published as Rodney D. Holder, ‘Explaining and Explaining Away in Cosmology and Theology,’ Theology and Science 14(3) (2016), 234–255. 2 David H. Glass and Mark McCartney, ‘Explaining and Explaining Away in Science and Religion,’ Theology and Science 12(4) (2014), 338–361. 3 David H. Glass, ‘Can Evidence for Design be Explained Away?’ in Jake Chandler and Victoria S. Harrison (eds.), Probability in the Philosophy of Religion (Oxford: Oxford University Press, 2012), 79–102. 4 William Lane Craig, The Kalām Cosmological Argument (London: Macmillan Press, 1979); see also the more recent Paul Copan and William Lane Craig, Creation out of Nothing: A Biblical, Philosophical, and Scientific Exploration (Grand Rapids, MI: Baker Academic, and Leicester: Apollos, 2004). 5 Robert J. Spitzer, New Proofs of the Existence of God: Contributions of Contemporary Physics and Philosophy (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2010). 6 Stephen Hawking, A Brief History of Time (London: Bantam, 1988). 7 Stephen Hawking and Leonard Mlodinow, The Grand Design: New Answers to the Ultimate Questions of Life (London: Bantam, 2010). 8 Hawking, Brief History, 136. 9 Hawking, Brief History, 140–141. 10 Hawking and Mlodinow, Grand Design, 134, 180. 11 S. W. Hawking, ‘The Quantum State of the Universe,’ Nuclear Physics B239 (1984), 257–276, 273. 12 S. W. Hawking, ‘Quantum Cosmology,’ in S. W. Hawking and W. Israel (eds.), Three Hundred Years of Gravitation (Cambridge: Cambridge University Press, 1987), 631–651, 650. 13 Quentin Smith, ‘The Wave Function of a Godless Universe,’ in William Lane Craig and Quentin Smith (eds.), Theism, Atheism, and Big Bang Cosmology (Oxford: Oxford University Press, 1993), 301–337, 316. 14 Arvind Borde and Alexander Vilenkin, ‘Eternal Inflation and the Initial Singularity,’ Physical Review Letters 72(21) (1994), 3305–3308. The theorem was extended from applying to inflationary models to include ‘brane’ cosmologies

Natural theology and modern cosmology 63 in Arvind Borde, Alan H. Guth, and Alexander Vilenkin, ‘Inflationary Spacetimes Are Incomplete in Past Directions,’ Physical Review Letters 90(15) (2003), 151301-1-151301-4. 15 Lisa Grossman, ‘Death of the Eternal Cosmos,’ New Scientist 213(2847) (January 2012), 6–7. 16 Grossman, ‘Death of the Eternal Cosmos,’ 7. 17 Peter J. Bussey, ‘God as First Cause—a Review of the Kalam Argument,’ Science and Christian Belief 25(1) (2013), 17–35. 18 Richard Swinburne, The Existence of God, second edition (Oxford: Oxford University Press, 2004), 96–109. 19 Lawrence M. Krauss, A Universe from Nothing: Why There is Something Rather than Nothing (New York: Simon and Shuster, 2012). 20 Krauss, Universe from Nothing, 159. 21 Lewis Carroll, Alice’s Adventures in Wonderland and Through the Looking Glass (Harmondsworth, Middlesex: Penguin, 1962), 286–289. 22 Alexander Vilenkin, Many Worlds in One: The Search for Other Universes (New York: Hill and Wang, 2006), 178. 23 Janet Soskice, ‘Creatio ex nihilo: Its Jewish and Christian Foundations,’ in David B. Burrell, Carlo Cogliati, Janet M. Soskice, and William R. Stoeger (eds.), Creation and the God of Abraham (Cambridge: Cambridge University Press, 2010), 24–39. 24 Hawking, Brief History, 174. 25 Fred Hoyle, ‘The Universe: Some Past and Present Reflections,’ Engineering & Science 45(2) (November 1981), 8–12, 12. 26 Fred Hoyle, The Nature of the Universe (Oxford: Blackwell, 1950), 115–116. 27 Paul Davies, The Goldilocks Enigma: Why is the Universe Just Right for Life? (London: Allen Lane, 2006), 3. 28 Max Tegmark, ‘Parallel Universes,’ Scientific American 288(5) (May  2003), 30–41. 29 Alan H. Guth, ‘Inflationary Universe: A Possible Solution to the Horizon and Flatness Problems,’ Physical Review D 23(2) (1981), 347–356. See also Alan H. Guth, The Inflationary Universe: The Quest for a New Theory of Cosmic Origins (London: Jonathan Cape, 1997). 30 A. D. Linde, ‘Particle Physics and Inflationary Cosmology,’ Physics Today 40(9) (1987), 61–68. 31 Leonard Susskind, The Cosmic Landscape: String Theory and the Illusion of Intelligent Design (New York: Little Brown and Company, 2006). 32 Susskind, Cosmic Landscape, 380. 33 See, for example, A. R. Peacocke, Theology for a Scientific Age, second edition (London: SCM Press (1993), 107–109. 34 Robin Collins, ‘Evidence for Fine-Tuning,’ in Neil Manson (ed.), God and Design: The Teleological Argument and Modern Science (London: Routledge, 2003), 178–199. 35 Robin Collins, ‘The Multiverse: A Theistic Perspective,’ in Bernard Carr (ed.), Universe or Multiverse? (Cambridge: Cambridge University Press, 2007), 459–480. 36 See for example Don N. Page, ‘Predictions and Tests of Multiverse Theories,’ in Carr (ed.), Universe or Multiverse? 411–430; Don N. Page, ‘Multiple Reasons for a Multiverse,’ in Rodney D. Holder and Simon Mitton (eds.), Georges Lemaître: Life, Science and Legacy (Heidelberg: Springer, 2012), 113–123. 37 More detail is given in Rodney D. Holder, God, the Multiverse, and Everything: Modern Cosmology and the Argument from Design (Aldershot and Burlington, VT: Ashgate, 2004); and in Rodney D. Holder, Big Bang, Big God: A Universe Designed for Life? (Oxford: Lion Hudson, 2013).

64  Natural theology and modern cosmology 38 Martin Rees, New Perspectives in Astrophysical Cosmology, second edition (Cambridge: Cambridge University Press, 2000), 138. 39 Martin Rees, Our Cosmic Habitat (London: Weidenfeld & Nicolson, 2001), 164. 40 T. J. Mawson, ‘Explaining the Fine Tuning of the Universe to Us and the Fine Tuning of Us to the Universe,’ Royal Institute of Philosophy Supplement 68 (2011), 25–50. 41 Nelson Goodman, Fact, Fiction, and Forecast (New York: Bobs-Merrill, 1946). 42 D. K. Lewis, On the Plurality of Worlds (Oxford: Blackwell, 1986). 43 Steven Weinberg, ‘Living in the Multiverse,’ in Carr (ed.), Universe or Multiverse? 29–42, 32. 44 Paul Davies, ‘Universes Galore: Where will it all End?’ in Carr (ed.), Universe or Multiverse? 487–505, 492. 45 John D. Barrow, The Infinite Book (London: Jonathan Cape, 2005), 144. 46 Roger Penrose, The Emperor’s New Mind: Concerning Computers, Minds, and the Laws of Physics (Oxford: Oxford University Press, 1989), 343. 47 Penrose, Emperor’s New Mind, 354. 48 Robin Collins, ‘Modern Cosmology and Anthropic Fine-tuning: Three Approaches,’ in Holder and Mitton (eds.), Georges Lemaître, 173–191. 49 Collins, ‘Modern Cosmology and Anthropic Fine-Tuning,’ 174. 50 E. P. S. Shellard, ‘The Future of Cosmology: Observational and Computational Prospects,’ in G. W. Gibbons, E. P. S. Shellard, and S. J. Rankin (eds.), The Future of Theoretical Physics and Cosmology: Celebrating Stephen Hawking’s 60th Birthday (Cambridge: Cambridge University Press, 2003), 755–780, 764. 51 Shellard, ‘Future of Cosmology,’ 764–765. 52 Swinburne, Existence of God, 133–135. 53 Swinburne, Existence of God, 6, 133–152. 54 Swinburne, Existence of God, 96–109, 165, 185–188. 55 Davies, Goldilocks Enigma, 249. 56 Swinburne, Existence of God, 145–147.

4 Moving on from natural theology Why we need ramified natural theology

Reason and revelation: a false dichotomy? In Chapter 2 we saw how it has been traditional in theology to recognize two sources of the knowledge of God, namely that which we can acquire simply by being human and exercising the natural human capacity for reason and that which is given to us by special revelation from God. I  gave James Barr’s definition of natural theology, and later that of William Alston. Here is a fuller version of the latter: Natural theology is the enterprise of providing support for religious beliefs by starting from premises that neither are nor presuppose any religious beliefs. We begin from the mere existence of the world, or the teleological order of the world, or the concept of God, and we try to show that when we think through the implications of our starting point we are led to recognize the existence of a being that possesses attributes sufficient to identify Him as God.1 We saw in Chapter 2 how St Thomas Aquinas makes precisely this point, namely that the existence of God can be known by reason alone: The truths about God which St Paul says we can know by our natural powers of reasoning—that God exists, for example—are not numbered among the articles of faith, but are presupposed to them  .  .  . God’s effects, therefore, can serve to demonstrate that God exists, even though they cannot help us to know him comprehensively for what he is.2 We also saw how, at the time of the reformation John Calvin made a similar distinction between what we can know simply by being human and what we can only know by revelation. For Calvin, each human being possesses a ‘sensus divinitatis,’ a sense of the divine. Historically, knowledge of God obtainable from the exercise of reason, and in principle available to all people, has been the subject matter of ‘natural theology’ as defined above. It is usually acknowledged, with Aquinas and Calvin, that this knowledge is

66  Moving on from natural theology limited, perhaps simply to inferring that some supreme being exists and is the creator of the universe. This knowledge needs to be supplemented by the far more important knowledge of God available only through special revelation, which is the subject matter of systematic theology. In this chapter I challenge the idea that these two sources of knowledge of God can be kept rigidly apart, and I argue for the vital necessity of ‘ramified natural theology,’ which concerns the evaluation of evidence for the particularities of the Christian faith and which was the subject of a special issue of the journal Philosophia Christi.3

Religious pluralism I have argued in previous chapters that theology is a rational pursuit just as science is rational. It thus needs to justify its tenets in a similar way to that in which science justifies its tenets, that is, by bringing reason and evidence to bear on its truth claims. We have seen how this can be done in a parallel way to science, through inference to the best explanation and through the application of Bayes’s theorem. This is precisely the domain of natural theology, as outlined in the preceding section. Given, as outlined, that the results of natural theology are limited, one major reason for going further and applying the same kind of reasoning to the particularities of the Christian faith is the phenomenon of religious pluralism. Not only are there competing claims about whether there is a God, but there are competing claims among religions about who this God is and how he has revealed himself. Distinguishing between these latter claims also requires a rational approach and this is the territory of ‘ramified natural theology.’ It is often said, with some truth, that there are many religions, and they say different and contradictory things. Sometimes atheists who make this observation move rather swiftly from it to the brusque conclusion that one cannot therefore choose between religions. There are many religions and they contradict each other; therefore all should be rejected. As a logical argument this is manifestly fallacious. Simply rephrase the argument. For a start, atheism is not monochrome. There are many belief systems, some atheistic, some religious. They contradict each other. Therefore all are false—including the atheistic ones. A different example would be to say that there are many political ideologies, all of which are competing for our votes. They contradict each other, so there is no way of choosing between them. One might conclude by arguing for the abolition of politics. It is also true that there are and have been rival, contradictory theories in science. Up until the early 1960s both the steady-state and Big Bang theories were viable cosmological models of the evolution of the universe. Today there are rival theories to unite quantum theory (the theory of the very small) and general relativity (Einstein’s theory of gravity), such as string theory and loop quantum gravity, which are currently empirically indistinguishable. It

Moving on from natural theology 67 is also true that rival interpretations of quantum theory itself, the standard Copenhagen interpretation of Niels Bohr and the Bohmian pilot wave theory, are empirically indistinguishable. Of course it would be ridiculous to say that all the theories are false because there are rivals which contradict them. The first case was settled when the discovery of the cosmic microwave background radiation verified the Big Bang theory which had predicted it and which could not be explained by the steady-state. The second case may (or may not) be settled by observation at some future date. The third is more of a metaphysical choice since both interpretations are consistent with the empirical evidence and are likely to remain so. Thus a convinced determinist will opt for the Bohmian theory whereas application of the criteria of simplicity (or economy) and elegance (or lack of ad hoc contrivance) would lead to a preference for the Copenhagen interpretation.4 But the argument is fallacious even without rephrasing it in the first way or appealing to the political or scientific analogies. It is either true or false that Jesus died on the cross. If he did, then Christianity is right about this particular point and a religion which denies it is in error on this particular point. It does not follow either that Christianity is true in all its affirmations or that the religion which denies that Jesus died on the cross is false in every respect. There may be things which these religions hold in common, for example that there is a God who created the universe. But this is where we begin to perceive a difficulty. If there is some general and minimal knowledge of God available to all human beings simply through being human, then different religions may agree about that. Perhaps this is an immediate sense, from the observation of nature, that there just has to be a supreme intellect behind the universe. Or perhaps there are arguments from reason and evidence which point towards there being such an intellect—the classical ontological, cosmological, and teleological arguments of natural theology for example, possibly reframed in probabilistic terms as in the work of Richard Swinburne. It is where the appeal is to special revelation, for the particular doctrines of a religion, that things are not so straightforward. If no arguments or evidence are produced for the validity of a putative revelation, how are we to decide whether it is true or not? Like Alvin Plantinga (see later), I may have some personal experience or encounter which convinces me that Jesus is Lord and that is fine for me. But then, how am I to convince my neighbour who has not had such an experience or encounter? It is at this point that certain trends in modern Protestant theology come unstuck. As we saw in Chapter 2, Karl Barth famously rejected natural theology (most especially, we may add, in his ‘Nein!’ to Emil Brunner).5 Barth offered instead what Dietrich Bonhoeffer described as a ‘positivism of revelation.’ According to Bonhoeffer this meant accepting the whole Christian package—Trinity, virgin birth, etc.—lock, stock, and barrel, at face value and without question. This was fine for the church, the already believing community, but not for the world at large. The church needs to ‘move out

68  Moving on from natural theology again into the open air of intellectual discussion with the world,’ says Bonhoeffer, rather than to isolate itself from such discussion, as it appears to do in the theology of Barth.6 Brian Hebblethwaite is a theologian in modern times who has advocated the extension of natural theology to the evaluation of revealed theology: [T]he old distinction between natural and revealed theology, between what reason can achieve and what required revelation, breaks down . . . reason and revelation cannot be treated as different sources of knowledge. On the contrary, revelation-claims, despite being channelled through particular historical traditions, are part of the data upon which reason has to operate.7 A major theologian who takes the evaluation of revelation claims seriously is Wolfhart Pannenberg.8 Pannenberg agrees with Bonhoeffer that Barth’s ‘positivism of revelation’ will not do: ‘Whereas other attempts to give theology a foundation in human terms sought support from common arguments, Barth’s apparently so lofty objectivity about God and God’s word turns out to rest on no more than the irrational subjectivity of a venture of faith with no justification outside itself.’9 Furthermore, ‘this reduction of the positive nature of the tradition to the subjective leads to a plurality of positions with no rational means of comparing them.’10 In light of the fact that, for Pannenberg, God reveals himself supremely by acting in history, it is to history that we must turn to evaluate the particular claims of Christianity. For Pannenberg, evaluating the historical evidence is essential since it constitutes ‘publicly checkable evidence.’ Although he does not use this expression, Pannenberg is thus engaged in what is now termed ramified natural theology. The task, he says, is to ‘make clear the primacy of God and God’s revelation in Jesus Christ and vindicate its claim to truth.’11 To carry out this task theology must not simply assert as Barth did, but argue with both atheism and religious pluralism on the grounds of a commonly shared rationality and on the basis of universal and publicly checkable evidence. As we saw in Chapter 1, Pannenberg argues that truth cannot be purely subjective: ‘My truth cannot be mine alone. If I cannot in principle declare it to be truth for all—though perhaps hardly anyone else sees this—then it pitilessly ceases to be truth for me also.’12 Pannenberg’s point here could equally well be applied to the Reformed epistemology of Alvin Plantinga. Accepting Calvin’s ‘sensus divinitatis,’ Plantinga argues that belief in God is ‘properly basic,’ just like the laws of logic and the deliverances of memory, and is not therefore inferred from any other propositions or empirical data. But then the question is, as in the case of Barth, how, if I believe in this sense, do I commend my faith to others? As Pannenberg puts it: If the really decisive matter, the revelatory and redemptive significance of the fate [Geschick] of Jesus of Nazareth, can be seen only by faith

Moving on from natural theology 69 and is in principle closed to rational investigation of this event, then it is impossible to see how the historicity of the pure facts should be able to protect faith against the reproach that it rests upon illusion and caprice.13 Pannenberg’s theology of other religions does not mean that he simply writes them off as false (in the simple way some atheists write off all religions). No, God is working within the non-Christian religions too. However, as Carl Braaten puts it: ‘Pannenberg makes the bold assertion that the special place of Jesus Christ in world history is a phenomenon that can be examined on generally applicable methodological grounds, without recourse to specifically Christian dogmatic principles.’14 Braaten notes a fundamental difference between Hendrik Kraemer, for whom a Christian theology of the religions derives from Christian presuppositions, and Pannenberg: ‘The difference between Kraemer and Pannenberg is that what for Kraemer is a premise of faith (the Christological starting point) is for Pannenberg a conclusion of reason.’15 And Braaten quotes this early programmatic statement of Pannenberg’s theology: ‘Jesus of Nazareth is the final revelation of God because the End of history appeared in him. It did so both in his eschatological message and in his resurrection from the dead. However, he can be understood to be God’s final revelation only in connection with the whole of history as mediated by the history of Israel. He is God’s revelation in the fact that all history receives its light from him.’16 Pannenberg’s view of other religions thus resembles that of a very different Christian thinker, C. S. Lewis, for whom other religions contain truths and hints, for example in their stories of dying and rising gods, of the one true revelation of God which comes through the incarnation, death, and resurrection of Jesus Christ. It is important to consider the extent to which one needs to study and evaluate the claims of other religions in drawing the kind of conclusion Pannenberg does about the place of Christ in history and, most notably, his resurrection. Pannenberg himself writes this: It is my thesis that the question of the divine reality cannot be settled independently from dealing with the particular and antagonistic claims of the different religious traditions.17 Brian Hebblethwaite makes a similar point: On any view, however, theology itself must reckon with the plurality of theistic and non-theistic religions in the world. Even if God does exist and can be thought about rationally in the discipline known as theology, each religious tradition which claims to provide knowledge of God must have something to say about the different claims of the other religions. Each tradition’s theology, therefore, must include from its own standpoint the theology of religion and the religions, the attempt, that is, to explain the plurality of religions.18

70  Moving on from natural theology Hebblethwaite argues that ‘Biblical interpreters must keep their eyes open to the other scriptures and rival revelation-claims in world religion.’ He moreover recognizes the problem here that a single scholar cannot do justice to all these revelation claims. As a simple illustration of this he says that he can learn Hebrew and Greek for Biblical study, but really ought to learn Sanskrit, Pali, and Arabic for in-depth study of other traditions.19 Richard Swinburne argues that it is not necessary to go down the route of detailed familiarization with other religions. He was accused by J. L. Schellenberg of failing to make the necessary comparison of Christian theism with other religious belief systems, and of showing ‘only the most superficial acquaintance with the beliefs and practices of other religious traditions.’20 Swinburne’s response to this is to argue that he does not need to make a detailed investigation of other religious traditions if he can show ‘that none of those religions even claim for themselves characteristics to be expected a priori of a true religion and claimed by Christianity, and that Christianity does have these characteristics.’21 Swinburne claims, first, to have established with significant probability that there is a God. So far, so good. He then goes on to argue that this God, given that he exists, would have good reason to do such things as become incarnate, reveal things to us, atone for human sin, and verify all this with a super-miracle. Given that no other religion makes such claims, that there is good evidence that Jesus led the kind of life attributed to him and did indeed rise from the dead, and that no corresponding evidence exists for any other prophet of any other religion, Swinburne regards his position as vindicated. Interestingly, Swinburne says he takes other religions at their face value. The fact that Islam makes no claim that Mohammed was God means that he does not need to investigate the claims Islam does make. Where I would demur somewhat with Swinburne is in the confidence with which he claims to know a priori what God would do—even what God is morally obliged to do. In any case, when he does so, it looks suspiciously as though he is working backwards from what he believes God has actually done. In fact, Swinburne recognizes this very point.22 I am inclined to agree with him that it does not really matter that it is the Christian tradition which gave us reasons for what God might do. If they are good reasons we can work with them and see how the evidence fits them. As we shall see in chapters 10 and 11, if we accept a moderately low probability a priori that God would become incarnate, we can derive an overwhelming probability from the historical evidence that he did so.

Excursus: the death of Jesus Where one might need to investigate the claims of other religions is where they flatly contradict the central claims of Christianity. To take one important example, the resurrection depends for its truth on the earlier historical

Moving on from natural theology 71 event of the death of Jesus on the cross. If Jesus’ death on the cross is denied then Christianity is mistaken in its central claim. It seems to me that the evidence for the crucifixion is very robust indeed, being reported in secular sources as well as being multiply attested to in Scripture. It is agreed to be indisputable across the spectrum of New Testament scholarship, from John Dominic Crossan to James Dunn and Tom Wright. John P. Meier suggests five main criteria for assessing what is authentic in what the New Testament tells us about Jesus.23 These are: 1 The criterion of embarrassment—this is not something Jesus’ followers would have invented but rather would have caused difficulty for the early Church. 2 The criterion of discontinuity (or dissimilarity)—this focuses on words or deeds of Jesus which could not be derived from Judaism at the time of Jesus or from the early Church after him. 3 The criterion of multiple attestation which we have already referred to—there is more than one independent literary source. 4 The criterion of coherence—other sayings or deeds fit with what we can learn from the first three criteria. 5 The criterion of rejection and execution—this directs us to Jesus’ violent end and asks what Jesus said or did to explain this. Regarding the last criterion Meier makes this ironic comment, relevant to what some of the liberal reductionist critical scholarship we shall meet in Chapter 7 has to say about Jesus: A tweedy poetaster who spent his time spinning out parables and Japanese koans, a literary aesthete who toyed with 1st-century deconstructionism, or a bland Jesus who simply told people to look at the lilies of the field—such a Jesus would threaten no one, just as the university professors who create him threaten no one. The historical Jesus did threaten, disturb, and infuriate people—from interpreters of the Law through the Jerusalem priestly aristocracy to the Roman prefect who finally tried and crucified him. This emphasis on Jesus’ violent end is not simply a focus imposed on the data by Christian theology. To outsiders like Josephus, Tacitus, and Lucian of Samosata, one of the most striking things about Jesus was his crucifixion or execution by Rome. A Jesus whose words and deeds would not alienate people, especially powerful people, is not the historical Jesus.24 For James Dunn the evidence, Biblical and extra-Biblical, puts Jesus’ death by crucifixion high on the ‘almost impossible to doubt or deny’ scale of historical ‘facts.’25 He makes a similar point to Meier: ‘One of the flaws of the most characteristic Liberal portrayal of Jesus was the unlikelihood that anyone would have wanted to crucify such an attractive moral teacher.’26 Like

72  Moving on from natural theology many modern scholars now (e.g. E. P. Sanders, Anthony Harvey, and Tom Wright), Dunn sees the ‘cleansing of the Temple,’ which we shall discuss in Chapter 8, as the primary cause for Jesus’ arrest by the Temple authorities and execution by Pontius Pilate.27 One would therefore need very strong grounds for denying the crucifixion, and the questions one would need to ask are ‘Has the historical-critical work been done on the text which is the source of the denial?’ and/or ‘Is there another way of interpreting that text so that the denial is only apparent?’ But some work in this direction might well be required, and as in the case of the Bible, it cannot be assumed a priori that any other holy text is infallible in its assertions. In fact the Qur’an in Sûrah IV (verses 157–158) does indeed appear to deny the crucifixion: And because of their [the Jews] saying: We slew the Messiah Jesus son of Mary, Allah’s messenger—They slew him not nor crucified, but it appeared so unto them; and lo! Those who disagree concerning it are in doubt thereof; they have no knowledge thereof save pursuit of a conjecture; they slew him not for certain. But Allah took him up unto Himself. Allah was ever mighty, wise. The traditional Islamic interpretation is given by Ghulam Sarwar: According to the Qur’ān, Prophet ‘Īsā [Jesus] was not crucified to death rather he was taken up to Allah, the Almighty and the Most Wise’ (4.157–158). Everything is possible for Allah. It was He who saved Ibrāhīm [Abraham] from the fire and Mūsā [Moses] from Fir’awn [Pharoah].28 Likewise Badru Kateregga: According to the true belief of Islam, it would seem most inappropriate for the Messiah to die through a shameful crucifixion. God, who is just, would not permit the righteous Messiah to suffer in that manner. Muslims believe that Allah saved the Messiah from the ignomy (sic) of crucifixion as Allah also saved the Seal of the Prophets from ignomy following the Hijra.29 Kateregga goes on to affirm Islam’s denial of humanity’s need for redemption and thus Christianity’s belief in the redemptive sacrificial death of Christ. In contrast to Christians, Muslims believe ‘that man has always been fundamentally good, and that God loves and forgives those who do His will.’30 However, it is worth noting that Kateregga and indeed Sûrah IV.157 are affirming that Islam shares with Christianity the belief that Jesus was the Jewish Messiah.

Moving on from natural theology 73 All I can say here is that, as we shall see in the case of the miracles of Jesus, from the perspective of historical analysis it would appear that the Qur’an is importing ideas from certain (heretical) Gnostic gospels of the second century. Now, while in fact it was the Romans who crucified Jesus, Mohammed may have acquired the idea of the Jewish claim to have killed him from a tradition like that in the Talmudic tractate Sanhedrin (43a) which describes Jesus as hanged on Passover Eve because he practised sorcery (a charge we return to in Chapter 7) and enticed Israel into apostasy.31 The explanation in Muslim tradition for the above assertion that they did not kill him is that a substitute with the appearance of Jesus was crucified while Jesus himself was caught up to heaven. Indeed, in his modern translation of the Qur’an N. J. Dawood renders the relevant part of Sûrah IV.157 as ‘They did not kill him, nor did they crucify him, but they thought they did,’ but adding in a footnote, ‘Or, literally, ‘he was made to resemble another for them.’32 This idea of a substitute was propagated by the Gnostic heresy of Docetism in the early Church, which heresy maintained that Jesus’ humanity and sufferings were only apparent and not real. Some Gnostics thus maintained that a substitute looking like Jesus, such as Simon of Cyrene, was crucified instead. Irenaeus, second century Bishop of Lyons, describes this heresy, ascribing it to Basilides: Wherefore he did not himself suffer death, but Simon, a certain man of Cyrene, being compelled, bore the cross in his stead; so that this latter being transfigured by him, that he might be thought to be Jesus, was crucified, through ignorance and error, while Jesus himself received the form of Simon, and, standing by, laughed at them. For since he was an incorporeal power, and the Nous (mind) of the unborn father, he transfigured himself as he pleased, and thus ascended to him who had sent him, deriding them, inasmuch as he could not be laid hold of, and was invisible to all.33 As it happens, Irenaeus may be mistaken here in his attribution of this denial of Jesus’ death to Basilides, since Clement of Alexandria quotes Basilides as saying that Jesus suffered like any other martyr. Nevertheless, as F. F. Bruce observes, it is ironic that Islam, which insists on the humanity of Jesus, should reflect this heretical view.34 To close this brief excursus on Jesus’ death, let me quote Charlie Moule’s masterly summary: Among the statements made about Jesus, one at least falls within the strictly historical field and, at any rate in principle, could have been objectively verified; and it is almost universally believed—the statement that he died. There have always been those, of course, who do not believe that he ever lived. But these are few and eccentric. There

74  Moving on from natural theology have been more who distinguish between the mortal being who died and a transcendental being who was unscathed: these are the docetists, alike in Christianity and in Islam. There have also been those who hold that Jesus did not die but only swooned: the resuscitation theorists. But the great majority, if they agree on nothing else, will allow that Jesus of Nazareth lived and truly died. It was a death, and took place when Pontius Pilate was prefect of Judaea. So much may be established as a historical event, as objectively as the death of Julius Caesar.35 A key feature of Pannenberg’s theology, which, although using different language, places him near the analytic philosophy (Bayesian confirmation theory) approach of Swinburne, is that the final vindication of the truth of Christian claims will only come at the eschaton. For Pannenberg the eschatological consummation of history is proleptically revealed in the resurrection of Jesus (which again makes Jesus the final revelation of God). So, although the resurrection of Jesus can be affirmed on the basis of publicly checkable evidence, that affirmation has a provisional—Swinburne would say, probable—claim to truth, rather than a certain one, until that final vindication.

Presuppositions Hugh Gauch says that the presuppositions made as the starting point for investigation of the truth claims of Christianity must command universal assent.36 The evidence admitted ‘must be rendered admissible by commonsense, worldview-independent presuppositions.’37 That is true. On the one hand one must not start by already assuming the Christian revelation to be true. This was the approach of Kraemer noted above and it seems to be the method of Alister McGrath, who himself follows in the footsteps of Thomas Torrance. For these theologians natural theology is not about producing arguments towards God from nature but about what nature can tell us, once we accept the full Trinitarian-Chalcedonian Christian belief system. This seems to be starting from what one is trying to show, and is inevitably circular. McGrath is, however, ambivalent, as we saw in Chapter 2, since he utilizes concepts from secular philosophy to claim, for example, that the Christian faith offers an inference to the best explanation for certain kinds of data, such as anthropic fine-tuning. In the context of religious pluralism, he also claims, building on the work of Alasdair McIntyre, that the Christian faith offers not just a tradition-specific rationality, but a tradition-transcendent rationality which has the power to explain the existence of other religious perspectives. I have argued that if these latter claims are correct, then they provide support for the Christian faith, and McGrath is really doing natural theology with its traditional meaning after all.38 Pannenberg identifies presuppositions of the opposite kind which should also not be made. Obviously, to begin with, if the existence of God is ruled out by fiat before one starts, one will never be able to infer to the existence

Moving on from natural theology 75 of God. As Pannenberg puts it, ‘the divine reality which is the subject of religious traditions must not be excluded by a definition of reality before it even comes to specific historical investigation and judgment.’39 In the nineteenth century, Ludwig Feuerbach took the approach criticized by Pannenberg. If there is no objective basis in reality for the concept of God, as Feuerbach assumed, it is not surprising that he would come up with the idea that God is a projection of human longing for significance. Materialistic premises are bound to lead to materialistic conclusions. This seems to be the way some researchers in the cognitive science of religion think.40 In the latter discipline, atheism is often assumed and a naturalistic explanation of the origin of religion put forward. That is supposed to render religion false, whereas in reality an explanation of its origin says nothing about its truth or falsity. Perhaps, on the contrary, we have a religious impulse because the Creator intended us to relate to him, and used ‘natural processes’ to bring it about that we could. This would be in line with Justin Barrett’s work on childhood theism,41 Calvin’s notion of the ‘sensus divinitatis,’42 and Plantinga’s view of religious belief as properly basic.43 After all, God’s normal way of working is through ‘natural processes,’ which are simply causal powers implanted in the created order by God, and which Aquinas called ‘secondary causes.’ A particular target for Pannenberg is Ernst Troeltsch. Troeltsch maintained three principles of historical enquiry. The first, the principle of criticism (Kritik), is the idea that historical events can only be established with a degree of probability rather than with certainty. Although both then and now some theologians would want to rest faith on certainty rather than probability, Pannenberg agrees with Troeltsch on this, as do analytic philosophers such as Swinburne today. Indeed faith is in the same position as science here, as evidenced by this quotation, which I also gave in Chapter 2, from Michael Polanyi in his book Personal Knowledge, on the philosophy of science: ‘The principal purpose of this book is to achieve a frame of mind in which I may hold firmly to what I believe to be true, even though I know that it may conceivably be false.’44 Even scientific theories have a provisional character, this also being a major theme of Karl Popper whose main criterion for a theory to be scientific was that it should be falsifiable. Troeltsch’s second principle is more controversial. This is the principle of analogy (Analogie), according to which past events must resemble present events. And his third principle, that every historical event is embedded in a web of interconnected events, is known as the principle of correlation (Korrelation, but also called Wechselwirkung, interconnectivity).45 By means of the principle of analogy, that is, by virtue of the historicalcritical method he adopted, Troeltsch was able effectively to deny the resurrection of Jesus by fiat. Indeed, he claimed that many had learned to be content with this approach by the time he was writing.46 In the Englishspeaking world, Troeltsch’s position resembles that of David Hume, especially in the latter’s famous essay ‘Of Miracles.’47 Among others, in recent times I in an article and John Earman in a book have argued against Hume,

76  Moving on from natural theology utilizing Bayesian probability theory, of which Hume was apparently ignorant.48 I return to this topic in Chapter 6. Pannenberg too refutes this a priori reasoning, while remaining fully committed to a historical-critical method. Pannenberg does not simply abandon the principle of analogy, but says that it must be qualified: Historicity does not necessarily mean that what is said to have taken place historically must be like other known events. The claim to historicity that is inseparable from the assertion of the facticity of an event simply involves the fact that it happened at a specific time. The question whether it is like other events may play a role in critical evaluation of the truth of the claim but is not itself a condition of the actual truth claim the assertion makes.49 This point is crucial when it comes to evaluating the central Christian claim that Jesus rose from the dead, which we discuss briefly in the next section but shall examine in more detail in later chapters. However, it is also worth emphasizing that it applies to other unique events as well. Thus, Richard Bauckham makes an interesting parallel with the Holocaust, an event which resembles the resurrection of Jesus only in the fact of its uniqueness and prima facie incredibility. Indeed both the resurrection of Jesus and the Holocaust fall into what Paul Ricoeur has termed ‘uniquely unique events.’50 As Bauckham says, ‘Holocaust testimonies are not easily appropriated by the historian, since they are prima facie scarcely credible and since they defy the usual categories of historical explanation.’ In brackets he quotes Charlotte Delbo as saying of new arrivals in Auschwitz ‘what is also true of any who read Holocaust testimonies: “They expect the worst—they do not expect the unthinkable.” ’ Bauckham goes on: ‘That is why the testimonies of survivors of the Holocaust are in the highest degree necessary to any attempt to understand what happened. The Holocaust is an event whose reality we could scarcely begin to imagine if we had not the testimonies of survivors.’51 The testimony of witnesses is likewise vital in establishing the historicity of Jesus, and of the resurrection in particular. As Bauckham says, ‘The history of Jesus discloses God’s definite action for human salvation, but only to those who attend to the testimony of the witnesses.’52

The resurrection of Jesus As above, our judgment of the historicity of the resurrection will be contested until the eschatological consummation of all things. It must for now simply be evaluated on the basis of the evidence without prejudging the issue. In his view on this Pannenberg appeals to no less an authority than St Paul, who makes the same point: Our judgment regarding the historicity of the resurrection of Jesus depends not only on examining the individual data (and the related

Moving on from natural theology 77 reconstruction of the event) but also on our understanding of reality, of what we regard as possible or impossible prior to any evaluation of the details. In this regard Paul is right that if we do not think the dead can rise in any circumstances, then we cannot regard the resurrection of Jesus as a fact (1 Cor 15:13), no matter how strong the evidence may be that supports it [a footnote refers to Hume’s classic essay]. We must concede, however, that a judgment of this type rests on a prior dogmatic decision and does not deserve to be called critical (in the sense of the evaluating of the transmitted texts).53 Judicious application of the principle of analogy can, ironically, work in favour of the historicity of the resurrection. For if Jesus’ tomb were empty, as even the Jewish authorities agreed, then it would violate the principle of analogy to assert both that the disciples had stolen the body and that they preached the resurrection of Jesus in Jerusalem as the fundamental component of the kerygma. This is because of the immense improbability of the disciples risking their lives for what they knew to be false—people do not do that today, so how can we conclude that they did it then? (This point is well brought out by Daniel Fuller54 and by Lydia and Timothy McGrew,55 and we return to it in our fuller discussion of Bayesian probability applied to the resurrection in Chapter 10.) Moreover, nothing in the background of the disciples would lead one to argue on historical grounds that the disciples were suffering from hallucinations or imagining that Jesus had risen from the dead. That would quite contradict the Jewish expectation, which was not of the resurrection of the Messiah alone ahead of time, but only of a general resurrection of the dead at the end of the age. They would comfort themselves with the belief of seeing Jesus then, not in seeing him somehow in the present. N. T. Wright makes a rather different, but equally valuable, point here, based on the same historical data. A general resurrection at the end of the age, as in Daniel 12, was indeed the predominant Jewish view in the time of Jesus,56 and what was not expected was the resurrection in the midst of the present age of a single human figure, Messiah or not. Wright argues that, whereas post-Enlightenment scholarship has often attributed belief in Jesus’ resurrection to later development and reflection by the early church, much the more plausible historical explanation reverses the order entirely. The actual bodily resurrection of Jesus, witnessed by the early disciples, is the source of the later reflection, rather than later reflection leading to belief in the resurrection: The accounts, in fact, make sense not as the final product of a development of theological and exegetical reflection within the early church, but as something like the source from which that development emerged. They are not the leaves on the branches of early Christianity. They look very like the trunk from which the branches themselves sprang, even though the writings in which we meet them are to be dated towards the end of the first generation, or even later.57

78  Moving on from natural theology In fact we can see something like the principle of correlation or interconnectivity at work here; it is just that the more rational interpretation reverses the order of cause and effect. In a nutshell, though, there is no bypassing the serious historical work, in which Pannenberg, Wright and others have been engaged, and which is the foundation for evaluating the fundamental truth claim of Christianity that Jesus of Nazareth rose bodily from the dead, left behind an empty tomb, and was seen alive again by many witnesses. We shall have more to say about this critical issue in chapters 9 and 10.

Conclusion To summarize, the traditional division between natural theology and revealed theology breaks down as soon as we ask why we should believe in a putative revelation and how we can commend our own perceived revelation to others. The question is particularly acute in the light of the multiplicity of apparent revelations in the world’s religions, and in light of atheist rhetoric about conflicting religious claims. While ‘bare’ natural theology can get us towards belief in a supreme being as creator of the universe, a view shared by many religions, it must be the task of Christian theology to present to the world reasons and evidence for the particular tenets of the Christian religion. William Alston, whom we quoted at the beginning, neatly outlines the task: The Christian may have recourse to natural theology to provide metaphysical reasons for the truth of a theism as a general world-view; and then, within the field of theistic religions, he may argue that historical evidence gives much stronger support to the claims of Christianity than to those of its theistic rivals—Judaism and Islam. To proceed in this way would be to argue from starting points that do not depend on prior acceptance of the Christian faith. . . . I believe that the attempt to argue from neutral starting points for the truth of Christian beliefs deserves much more serious consideration than is commonly accorded it today in philosophical and (liberal) theological circles.58 I agree with Alston and Pannenberg that we must seek to vindicate the primacy of God’s revelation in Jesus Christ in this pluralistic context. That will mostly mean examining the evidence for Christian claims, but it may well also mean dealing with at least some of the claims made by other religions which contradict those of Christianity. Of course it may also mean, as in the case of Muslim affirmation of Jesus’ Messiahship, pointing to areas where other religions support specifically Christian claims. We must also make the fundamental claims of Christianity credible on the basis of a commonly shared rationality and on the basis of the historical evidence. That means that we should not make any prior assumption of the correctness or accuracy of Christian claims, but nor should we exclude

Moving on from natural theology 79 Christian claims, such as the existence of God or the occurrence of miracles, a priori. These are the claims which are at stake and cannot be either assumed or excluded before we look at the evidence. As with science, widely believed to be the most objective and rational of pursuits, any results will have a provisional character and will only establish the facts with a degree of probability, as per Ernst Troeltsch’s principle of criticism. There is, however, only a qualified role for Troeltsch’s principles of analogy and correlation, since they must not be used to exclude what is at stake by fiat. Thus the question of whether human beings are prepared to die for what they know to be a lie (one might add, especially given the character and moral teaching exemplified by the witnesses to Jesus’ resurrection and, indeed, by the Master they claimed to serve) can be brought to bear in favour of the resurrection accounts being reliable. Later reflection by the church in the gospels can be correlated with the early Christians’ experience of having met the risen Christ. No doubt many more examples can be given as one examines the evidence in detail. In fact, anticipating the results of later chapters, as soon as we allow historical evidence its proper place, a very substantial case can be made to render it true with a high degree of probability that ‘the third day he rose again according to the Scriptures,’ thus enabling the Christian believer to recite the creed with confidence. This chapter has been chiefly about questions of principle, but it is to serious historical study that one must turn for the detailed material to support the programme of ramified natural theology, which is so vital to Christian apologetics and mission in the modern world. Chapter 6 is also methodological in nature, applying Bayesian confirmation theory to the notion of miracles in general. Following that, in chapters 7 to 10 we turn to the historical data, to make the case for Jesus’ own miracles, his fulfilment of prophecy, and his resurrection. Having compared Lydia and Timothy McGrew’s use of Bayesianism with that of Swinburne in Chapter 10, in the final chapter we bring together all the evidence we have considered in our own formalizing of the argument using the Bayesian approach.

Notes 1 William P. Alston, Perceiving God: The Epistemology of Religious Experience (Ithaca, NY: Cornell University Press, 1991), 289. 2 St Thomas Aquinas, Summa Theologiae, Blackfriars edition (London: Eyre & Spottiswoode; New York: McGraw Hill, 1963), 1a. 2, 2. 3 This chapter is based on my article in that issue: Rodney Holder, ‘Why We Need Ramified Natural Theology,’ Philosophia Christi 13(2) (2013), 271–282. 4 John Polkinghorne, Quantum Theory: A  Very Short Introduction (Oxford: Oxford University Press, 2002), 88–89. 5 Karl Barth, ‘No!,’ Response to Emil Brunner, ‘Nature and Grace,’ in Emil Brunner and Karl Barth, Natural Theology, trans. Peter Fraenkel, with an introduction by John Baillie (London: Geoffrey Bles/Centenary Press, 1946), 65–128. 6 Dietrich Bonhoeffer, Letters and Papers from Prison, enlarged edition, ed. Eberhard Bethge (London: SCM Press, 1971), 378. See my discussion in Rodney

80  Moving on from natural theology Holder, The Heavens Declare: Natural Theology and the Legacy of Karl Barth (West Conshohocken: Templeton Press, 2012), 90–96. 7 Brian Hebblethwaite, The Problems of Theology (Cambridge: Cambridge University Press, 1980), 79. 8 For more on Pannenberg, see Holder, The Heavens Declare, 99–138. 9 Wolfhart Pannenberg, Theology and the Philosophy of Science, trans. Francis McDonagh (London: Darton, Longman & Todd, 1976), 273. (First German edition, Wissenschaftstheorie und Theologie, Frankfurt am Main: Suhrkamp, 1973). 10 Pannenberg, Theology and the Philosophy of Science, 273. 11 Wolfhart Pannenberg, Systematic Theology, vol. 1, trans. Geoffrey W. Bromiley (Edinburgh: T & T Clark, 1991), 128. 12 Pannenberg, Systematic Theology, vol. 1, 51. 13 Wolfhart Pannenberg, Basic Questions in Theology, vol. 1, trans. George H. Kehm (London: SCM Press, 1970), 60. Originally published as Grundfragen Systematischer Theologie: Gesammelte Aufsätze (Göttingen: Vandenhoeck and Ruprecht, 1967). 14 Carl E. Braaten, ‘The Place of Christianity Among the World Religions: Wolfhart Pannenberg’s Theology of Religion and the History of Religions,’ in Carl E. Braaten and Philip Clayton (eds.), The Theology of Wolfhart Pannenberg: Twelve American Critiques, with an Autobiographical Essay and Response (Minneapolis: Augsburg, 1988), 287–312, 305. 15 Braaten, ‘The Place of Christianity Among the World Religions,’ 306. 16 Wolfhart Pannenberg, ‘The Revelation of God in Jesus of Nazareth,’ in J. M. Robinson and John B. Cobb, Jr. (eds.), Theology as History (New York: Harper & Row, 1967), 101–133, 125; quoted in Braaten, ‘The Place of Christianity Among the World Religions,’ 306. 17 Wolfhart Pannenberg, ‘A Response to My American Friends,’ in Braaten and Clayton (eds.), The Theology of Wolfhart Pannenberg, 314. 18 Hebblethwaite, Problems of Theology, 3. 19 Hebblethwaite, Problems of Theology, 91. 20 J. L. Schellenberg, ‘Christianity Saved? Comments on Swinburne’s Apologetic Strategies in the Tetralogy,’ Religious Studies 38 (2002), 283–300, 288. 21 Richard Swinburne, ‘Response to My Commentators,’ Religious Studies 38 (2002), 301–315, 310. Swinburne’s arguments are spelled out in detail in Richard Swinburne, The Resurrection of God Incarnate (Oxford: Oxford University Press, 2003), especially chapters 2 and 3. 22 Swinburne, Resurrection of God Incarnate, 35. 23 John P. Meier, ‘Criteria: How Do We Decide What Comes from Jesus?,’ in James D. G. Dunn and Scot McKnight (eds.), The Historical Jesus in Recent Research (Winona Lake, IN: Eisenbrauns, 2005), 126–136. 24 Meier, ‘Criteria,’ 136. 25 James D. G. Dunn, Jesus Remembered: Volume 1 of Christianity in the Making (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2003), 339. 26 Dunn, Jesus Remembered, 784. 27 Dunn, Jesus Remembered, 785–786. 28 Ghulam Sarwar, Islam: Beliefs and Teachings (London: The Muslim Educational Trust, third edition, 1984), 158. 29 Badru D. Kateregga in Badru D. Kateregga and David W. Shenk, Islam and Christianity: A Muslim and a Christian in Dialogue (Grand Rapids, MI: Wm. B. Eerdmans, revised edition 1981), 140. 30 Kateregga, in Islam and Christianity, 141. 31 F. F. Bruce, Jesus and Christian Origins outside the New Testament (London: Hodder and Stoughton, 1974), 55–56. 32 The Koran with a Parallel Arabic Text, translated with notes by N. J. Dawood (London: Penguin Books, fifth revised edition, 1990), 102.

Moving on from natural theology 81 33 Irenaeus, Adversus Haereses, I.XXIV.4, in Alexander Roberts and James Donaldson (eds.), Ante-Nicene Fathers, vol. 1, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson), 349. 34 Bruce, Jesus and Christian Origins, 176, and 176, fn. 22. 35 C. F. D. Moule, The Origin of Christology (Cambridge: Cambridge University Press, 1977), 107–108. 36 Hugh Gauch, ‘The Methodology of Ramified Natural Theology,’ Philosophia Christi 13(2) (2013), 283–298. 37 Gauch, ‘Methodology of Ramified Natural Theology,’ 288. 38 Holder, The Heavens Declare, 169–231. 39 Pannenberg, ‘Response to My American Friends,’ 319. 40 E.g. Pascal Boyer, Religion Explained: The Evolutionary Origins of Religious Thought (New York: Basic Books, 2001). 41 Justin L. Barrett, Born Believers: The Science of Children’s Religious Belief (London: Simon and Shuster, 2012). 42 John Calvin, Institutes of the Christian Religion, trans. Henry Beveridge (Grand Rapids, MI: Wm. B. Eerdmans, [1559] 1989), I.III.1, 43. 43 Plantinga makes this point well in Alvin Plantinga, Where the Conflict Really Lies: Science, Religion, and Naturalism (New York: Oxford University Press, 2011), 129–152. 44 Michael Polanyi, Personal Knowledge: Towards a Post-Critical Philosophy (Chicago: University of Chicago Press,1974), 214; cited in John Polkinghorne, Faith, Science and Understanding (London: SPCK, 2000), 33–34. 45 Ernst Troeltsch, ‘Über historische und dogmatische Methode in der Theologie,’ in Gerhard Sauter (ed.), Theologie als Wissenschaft (München: Chr. Kaiser Verlag, 1971), 105–127. Originally published in Studien des rheinischen Predigervereins, 1898. 46 Troeltsch, ‘Über historische und dogmatische Methode,’ 108. 47 David Hume, An Enquiry Concerning Human Understanding, Section X, ‘Of Miracles,’ in the first part of L. A. Selby-Bigge (ed.), Enquiries Concerning Human Understanding and Concerning the Principles of Morals by David Hume, third edition, revised by P. H. Nidditch (Oxford: Oxford University Press, [1748] 1975). 48 Rodney D. Holder, ‘Hume on Miracles: Bayesian Interpretation, Multiple Testimony, and the Existence of God,’ British Journal for the Philosophy of Science 49 (1998), 49–65; John Earman, Hume’s Abject Failure: The Argument Against Miracles (New York and Oxford: Oxford University Press, 2000). 49 Wolfhart Pannenberg, Systematic Theology, vol. 2, trans. Geoffrey W. Bromiley (Edinburgh: T & T Clark, 1994), 360–361. 50 Richard Bauckham, Jesus and the Eyewitnesses: The Gospels as Eyewitness Testimony (Grand Rapids, MI: Wm. B. Eerdmans, 2006), 499. 51 Bauckham, Jesus and the Eyewitnesses, 493. 52 Bauckham, Jesus and the Eyewitnesses, 500. 53 Pannenberg, Systematic Theology, vol. 2, 362. 54 Daniel P. Fuller, ‘The Resurrection of Jesus and the Historical Method,’ Journal of Bible and Religion 34(1) (1966), 18–24, 22. 55 Timothy McGrew and Lydia McGrew, ‘The Argument from Miracles: A Cumulative Case for the Resurrection of Jesus of Nazareth,’ in William Lane Craig and J. P. Moreland (eds.), The Blackwell Companion to Natural Theology (Chichester: Wiley-Blackwell, [2009] 2012), 593–662, 624. 56 N. T. Wright, The Resurrection of the Son of God: Volume 3 of Christian Origins and the Question of God (London: SPCK, 2003), 130. 57 Wright, Resurrection, 211. 58 Alston, Perceiving God, 270.

5 Pascal’s Pensées and Butler’s Analogy Foreshadowing ramified natural theology

Introduction This chapter will consider two great thinkers whose work prefigures ramified natural theology. Blaise Pascal, interestingly, rejected natural theology, but argued for the deity of Christ from miracles and fulfilled prophecy. Joseph Butler, in his Analogy of Religion (published 1736), discusses both natural theology and revelation and does so in terms of evidence and probability. Thus he prefigures the use of Bayesian probability theory as applied to ramified natural theology today. Both Pascal’s and Butler’s arguments will be picked up later when I discuss modern ramified natural theology. Butler was arguing against the deists. The latter accepted natural theology as pointing to the existence of a designer but rejected the particulars of Christianity because of the supposed incredibility of miracles. Butler argued for the probable truth of the Christian verities. He is thus an important forerunner for the kind of arguments I advance for ramified natural theology which, like him, I  see as pointing to the probable truth of Christian claims. Pascal and Butler, while they have important things to say, are, however, dated in view of the fact that they preceded the rise of historicalcritical study of the Bible, which occurred alongside Darwinism in the nineteenth century; also they precede the development of Bayes’s theorem in probability theory. Historical criticism of the Bible is possibly more important than Darwinism since it impacts the particulars of faith. However, it can also be seen positively as its more extreme outputs are challenged and critiqued, and it can feed into ramified natural theology. I shall argue in later chapters that, in drawing inferences from Scripture in an academically informed way, we cannot simply assume that it was directly delivered by God to the prophets and apostles. The Bible contains many different genres of literature and there are often complex processes of transmission and redaction at play. Nevertheless, a sober assessment can see the overall integrity of the Bible maintained. Indeed, the discoveries made can help towards a greater understanding of the text. But these arguments will be seen later in the light of the earlier apologetic of Pascal and Butler.

Pascal’s Pensées and Butler’s Analogy 83

Blaise Pascal (1623–1662) Blaise Pascal is remembered as a great French mathematician, pioneer of probability theory, and physicist. He invented a calculating device (the Pascaline), which probably explains the naming of a modern computer programming language after him. The SI unit of pressure is also named after him, remembering Pascal’s Principle in fluid mechanics. In religion, Pascal was a Roman Catholic who in 1646 became associated with the convent of Port-Royal, which under the influence of the abbé de Saint-Cyran had adopted the principles of Jansenism, an austere theological movement heavily influenced by St Augustine. After some years working intensely on mathematics and physics, Pascal underwent a dramatic conversion experience, as described by Alban Krailsheimer in the introduction to his translation of the Pensées: On 23 November 1654 he saw the light that guided him for the rest of his life, . . . The “hidden God” . . . was manifested that night to Pascal in the person of Jesus Christ. Whatever his doubts may have been, they were dispelled by this direct contact with God in the only form in which he henceforth sought him, and sought to communicate him to others. The witness of the two Testaments, the Passion of Christ, and Pascal’s personal obligations as a Christian are the three elements of a definitive realization of the truth, which brought him at last to certainty and joy.1 When his niece was suddenly cured of a fistula in her eye apparently by the application of a ‘relic of the Holy Thorn’ (believed to be from Christ’s crown of thorns), Pascal applied himself to collecting material for a work on miracles.2 The Pensées represent notes for a major work of Christian apologetics. Sadly they are all we have; we do not have the final product. In what follows individual fragments from the Pensées are preceded by the letter ‘L’ and are from the arrangement by L. Lafuma in his edition of the complete works of Pascal.3 This is the arrangement adopted by Krailsheimer in his translation of the Pensées.4 Pascal repudiates arguments from nature near the beginning of the Pensées: ‘ “Why, do you not see yourself that the sky and the birds prove God?” . . . “No. For though it is true in a sense for some souls whom God has enlightened in this way, yet it is untrue for the majority.” ’ (L. 3). John Cruickshank sees this rejection of natural theology as in keeping with the teaching of Augustine (who, as noted in Chapter 2, critiqued the insights of classical philosophers), and points to a number of passages in the Pensées where Pascal rejects such an approach.5 Pascal states: ‘Anyone who chose to follow reason alone would have proved himself a fool.’ (L. 44). The problem is original sin, which makes man the wretchedest of creatures: Man is nothing but a subject full of natural error that cannot be eradicated except through grace. Nothing shows him the truth, everything

84  Pascal’s Pensées and Butler’s Analogy deceives him. The two principles of truth, reason and senses, are not only both not genuine, but are engaged in mutual deception. (L. 45) As noted by John Cruickshank, the prison writings of the Jansenist SaintCyran find many echoes in the Pensées. ‘First among these,’ writes Cruickshank, ‘is his Augustinian emphasis on man’s total wickedness because of Adam’s sin and his utter dependence on God’s grace.’6 The following passage from the Pensées illustrates this point well. Noting that there are in faith ‘two equally constant truths,’ Pascal expounds these truths thus: One is that man in the state of his creation, or in the state of grace, is exalted above the whole of nature, made like unto God and sharing in his divinity. The other is that in the state of corruption and sin he has fallen from that first state and has become like the beasts. . . . Whence it is clearly evident that man through grace is made like unto God and shares his divinity, and without grace he is treated like the beasts of the field. (L. 131) Reason is important to Pascal but suffers from a number of inadequacies. Thus one must not exclude reason but neither should one admit nothing but reason (L. 253). One problem is that first principles cannot be derived from reason but come from the heart and from instinct, and reason works on what these deliver (L. 110). Moreover, there are an infinite number of things which are beyond reason and ‘If natural things are beyond it, what are we to say about supernatural things?’ (L. 188). In the light of this, it comes as no surprise that Pascal rejects traditional metaphysical arguments for the existence of God: The metaphysical proofs for the existence of God are so remote from human reasoning and so involved that they make little impact, and, even if they did help some people, it would only be for the moment during which they watched the demonstration, because an hour later they would be afraid they had made a mistake. (L. 190) In particular, arguments from design in nature are rejected since, while believers will immediately recognize what is described as the creation of God, unbelievers will think the arguments feeble: In addressing their arguments to unbelievers, their first chapter is the proof of the existence of God from the works of nature. Their enterprise would cause me no surprise if they were addressing their arguments to the faithful, for those with living faith in their hearts can certainly see at

Pascal’s Pensées and Butler’s Analogy 85 once that everything which exists is entirely the work of the God they worship. . . . to give them [unbelievers] no other proof of this great and weighty matter than the course of the moon and planets; to claim to have completed the proof with such an argument; this is giving them cause to think that the proofs of our religion are indeed feeble, and reason and experience tell me that nothing is more likely to bring it into contempt in their eyes. This is not how Scripture speaks, with its better knowledge of the things of God . . . . (L. 781) Furthermore, such arguments point at best to deism, and thus do not do what is required, namely to put a person into a right relationship with the living God: They imagine that it simply consists in worshipping a God considered to be great and mighty and eternal, which is properly speaking deism, almost as remote from the Christian religion as atheism, its complete opposite . . . God does not manifest himself to men as obviously as he might. . . . The Christian’s God does not consist merely of a God who is the author of mathematical truths and the order of the elements. That is the portion of the heathen and Epicureans. But the God of Abraham, the God of Isaac, the God of Jacob, the God of the Christians is a God of love and consolation: he is a God who fills the soul and heart of those he possesses: he is a God who makes them inwardly aware of their wretchedness and his infinite mercy . . . (L. 449) For Pascal, ‘It is the heart which perceives God and not the reason.’ (L. 424). One of his most famous aphorisms is this: ‘The heart has its reasons of which reason knows nothing.’ (L. 423). Nevertheless reason has a role, as is revealed in the famous wager (L. 418). Pascal was a pioneer of decision theory, which aids the making of rational decisions in the light of uncertainty. Thus given two possible outcomes of differing value or utility to me, I can rationally wager on one of the outcomes by calculating the expected utility. Suppose you are compelled to make a bet on the basis of a toss of a coin. Suppose that if the coin comes down heads, your return is £300 and if it comes down tails you forfeit £100. Then, since the probability of each outcome is ½, your expected gain is ½ x 300 + ½ x (- 100) = £100 if you call heads. Now we really are compelled to wager in the game of life. Supposing that there is a probability of ½ that God exists. If we wager that he does and we live our lives accordingly, our reward will not be finite but infinite, an eternal life of bliss in the Kingdom of God. If we are wrong our loss is one finite life. This means that now our expected gain from wagering on God’s existence is infinite, so it is the rational choice to make purely on the grounds of expected utility.

86  Pascal’s Pensées and Butler’s Analogy A more detailed analysis of Pascal’s Wager, including some objections, is to be found in Jon Elster’s chapter in The Cambridge Companion to Pascal.7 Besides some technical issues to do with infinities, Elster notes the ‘many gods’ argument. As we see below, Pascal is aware of the problem of religious pluralism and argues for the superiority of Christianity, but these arguments would need to be made to, and accepted by, the person challenged to make the Wager, so that his choice is a binary one between believing in the true God and rejecting him. A further difficulty with Pascal is squaring any kind of argument with his doctrine of predestination, which despite Pascal’s protests is not unlike that of Calvin.8 Pascal thought that science was the object of rational enquiry and not the subject of authority, but in theology the opposite is the case, with the authority being the Bible.9 In contrast to what natural theology delivers, Pascal asserts, in a manner not dissimilar to Karl Barth in the twentieth century, that we only know God through Jesus Christ, as testified in Scripture: We know God only through Jesus Christ . . . All those who have claimed to know God and prove his existence without Jesus Christ have only had futile proofs to offer. But to prove Christ we have the prophecies which are solid and palpable proofs . . . . (L. 189) Pascal’s view here no doubt owes a lot to his dramatic conversion experience, referred to above. He describes in his ‘Memorial’ how he discovered the ‘God of Abraham, God of Isaac, God of Jacob, not of philosophers and scholars. Certainty, certainty, heartfelt, joy, peace. God of Jesus Christ. God of Jesus Christ.’ (‘The Memorial,’ L. 913). Pascal indicates in L. 189 above the most important way in which he sees Christ as ‘proved’ by Scripture, namely through the prophecies. Miracles are important too, and we noted Pascal’s intention of producing a treatise on them above. In the Pensées Pascal states that ‘Miracles prove God’s power over our hearts by that which he exercises over our bodies’ (L. 903), and refers to miracles a number of times. He notes that ‘  .  .  . the scribes and Pharisees made much of his miracles, trying to prove them to be either false or the work of the devil, for they could not help being convinced if they once recognised them as coming from God.’ (In Part III, L. 854). That the scribes and Pharisees attributed Christ’s miracles to the work of the devil is a point to which we shall return in Chapter 7, since it still has purchase in the light of Biblical criticism. This is something the gospel writers would be most unlikely to include were it not true. Miracles are thus important. However, it is prophecies which are at the heart of Pascal’s apologetic: But to prove Christ we have the prophecies which are solid and palpable proofs. By being fulfilled and proved true by the event, these prophecies show that that these truths are certain and thus prove that Jesus is

Pascal’s Pensées and Butler’s Analogy 87 divine. In him and through him, therefore, we know God. Apart from that, without Scripture, without original sin, without the necessary mediator, who was promised and came, it is impossible to prove absolutely that God exists . . . But through and in Christ we can prove God’s existence . . . Therefore Jesus is the true God of men. (L. 189) The Christian faith is thus established as true. Pascal was aware, however, of the problem of religious pluralism: I see a number of religions in conflict, and therefore all false, except one. Each of them wishes to be believed on its own authority and threatens unbelievers. I do not believe them on that account. Anyone can say that. Anyone can call himself a prophet, but I see Christianity, and find its prophecies, which are not something that anyone can do. (L. 198) The issue of religious pluralism is indeed important, as we noted in Chapter 4, and we shall need to consider it in later chapters, albeit we shall be unable to provide a detailed treatment. The greatest rival to Christianity for Pascal, as many would see it today, is Islam. Pascal lists a number of fundamental differences between the two great faiths: Mahomet not foretold, Jesus foretold. Mahomet slew, Jesus caused his followers to be slain. Mahomet forbade reading, the Apostles commanded it. In a word, the difference is so great that, if Mahomet followed the path of success, humanly speaking, Jesus followed that of death, humanly speaking . . . (L. 209) And again: The Muslim religion has the Koran and Mahomet for foundation. But was this prophet, supposedly the world’s last hope, foretold? And what signs does he show that are not shown by anyone else who wants to call himself a prophet? What miracles does he himself claim to have performed? (L. 243) One can again see the importance of prophecy in the sense of ‘being foretold’ here, and indeed of miracles. Interestingly the Qur’an affirms that Jesus performed miracles, and it is not claimed that Mohammed performed miracles (the great miracle for Muslims is the Qur’an itself). The above may not be entirely fair, however. Pascal may have got the dubious forbidding of reading from fellow French philosopher Montaigne. In fairness, Pascal also admits that there are obscurities in Scripture as well as in Mahomet. However,

88  Pascal’s Pensées and Butler’s Analogy regarding Scripture, ‘I admit that there are obscurities as odd as those of Mahomet, but some things are admirably clear, with prophecies manifestly fulfilled.’ (L. 218). He cites this obscurity, again making an important point about non-collaboration between the gospel writers which we shall also return to in later chapters: ‘Thus all the most obvious weaknesses are really strengths. Example: the two genealogies of St Matthew and St Luke. What could be clearer than that there was no collaboration?’ (L. 236). Pascal is adamant that the prophecies foretell the coming of Christ and are demonstrative of Christian truth. For example, he writes: If a single man had written a book foretelling the time and manner of Jesus’s coming and Jesus had come in conformity with these prophecies, this would carry infinite weight. But there is much more here. There is a succession of men over a period of 4,000 years, coming consistently and invariably one after the other, to foretell the same coming: there is an entire people proclaiming it, existing for 4,000 years to testify in a body to the certainty they feel about it . . . (L. 332) The problem with Pascal’s approach to prophecy is that, in contrast to later critical scholars, Pascal takes the Bible literalistically where this would now be deemed inappropriate. Thus, in contrast to the theological academy today, the stories of the creation and flood are taken as literal and Adam and Eve are literally existing human beings. On the other hand, he does allow metaphorical interpretation if the literal would be plainly false: When the word of God, which is true, is false in the letter it is true in the spirit. Sit thou at my right hand is false literally, so it is true spiritually. . . “The Lord smells a sweet savour” and will reward you with a rich land . . . means that he has the same intention as a man who smells your sweet savour and rewards you with a rich land. (L. 272) We would readily recognize metaphors in the above. However, Pascal’s endorsement of spiritual or metaphorical interpretations is sometimes more sophisticated than we would expect. Thus prophecies about ‘defeat of enemies’ may mean defeat of iniquities since these are the real enemies, and sometimes, as in Isaiah 43:25 and elsewhere, the ambiguity of the term ‘enemies’ is removed and words such as iniquities, transgressions, sins, used explicitly (L. 278). This kind of interpretation of some aspects of prophecy will be important in Chapter 8. An argument that is as valid today as when Pascal formulated it is the following, and again we shall return to it in later chapters: The hypothesis that the Apostles were knaves is quite absurd. Follow it out to the end and imagine these twelve men meeting after Jesus’s

Pascal’s Pensées and Butler’s Analogy 89 death and conspiring to say that he had risen from the dead. This means attacking all the powers that be. The human heart is singularly susceptible to fickleness, to change, to promises, to bribery. One of them had only to deny his story under these inducements, or still more because of possible imprisonment, tortures and death, and they would all have been lost. Follow that out. (L. 310) Arguably in the seventeenth century Pascal had no reason to think the creation and flood stories were metaphorical or in any way non-literal (though we noted St Augustine’s more nuanced interpretation of creation in Chapter 2). Cruickshank makes a rather different point about Pascal’s handling of prophecy: Pascal paid little or no attention to the question of how far the texts he cites were always intended as prophecy or to what extent the New Testament writers described Christ’s life in terms consciously related to the prophetic phraseology of the Old Testament.10 Cruickshank’s point here is an important one, since it may well be that at least some ostensible ‘fulfilments’ of prophecy, related as such by the gospel writers, may not have had that intention in the original writers. Moreover, Jesus could have been consciously acting with Old Testament passages in mind. We return to these points in Chapter 8. They do not necessarily negate prophecy but require a more nuanced approach. Prophecies which could still today be regarded as fulfilled without much question would include the continuing existence (‘perpétuité’) of the Jewish people and the Christian Church. As Cruickshank says, ‘This perpetuity means that the Old Testament promises made to the Jews and the New Testament conception of the Church have continued to produce a myriad witnesses—prophets, martyrs, saints, teachers—up to the present time.’11 Pascal writes: He [the Messiah] came at last in the fullness of time, and since then we have seen so many schisms and heresies arise, so many states overthrown, so many changes of every kind, while the Church which worships him who has always been has continued without a break. (L. 281) In L. 189 quoted above Pascal sees prophecy fulfilment as providing proof of the Christian verities. This is in tension with another important passage: The prophecies, even the miracles and proofs of our religion, are not of such a kind that they can be said to be absolutely convincing, but they are at the same time such that it cannot be said to be unreasonable to believe in them. There is thus evidence and obscurity, to enlighten some

90  Pascal’s Pensées and Butler’s Analogy and obfuscate others. But the evidence is such as to exceed, or at least equal, the evidence to the contrary, so that it cannot be reason that decides us against following it, and can therefore only be concupiscence and wickedness of heart. Thus, there is enough evidence to condemn and not enough to convince, so that it should be apparent that those who follow it are prompted to do so by grace and not by reason, and those who evade it are prompted by concupiscence and not by reason. (In Part III, L. 835) Here it is not a question of absolute proof but rather of the weight of evidence, which could be translated into probability. Even if the evidence ought normatively to be persuasive, in that it outweighs that on the other side, it is still possible to decide against it, says Pascal. I would say on this point that deciding for the truth of Christian claims is not simply a matter of accepting certain propositions, but does indeed involve the heart as well as reason. To decide to follow Christ, and not simply to believe that he existed, wrought miracles, and was prophesied, is to commit one’s life to his service, and it is indeed grace which enables that decision. On Pascal’s pre-critical approach to prophecy, Krailsheimer remarks: ‘His quite uncritical reading of the prophetic books, especially Daniel, led him into open absurdity.  .  .  .’12 Pascal did indeed, as one might expect from someone writing in the seventeenth century, treat Daniel at face value and uncritically. Since we return to the book of Daniel in Chapter 8, we shall take up this particular point in some detail, with David Wetsel providing a helpful guiding hand to understand Pascal’s method.13 In series XIV of the Pensées (L. 485) Pascal lists numerous passages from chapters 2, 8, 9, and 11 of the book of Daniel. Pascal believes that these passages clearly predict the timing of the birth of the Messiah. Pascal does not explain how he arrives at this conclusion, though there are hints in the marginal comments included in Krailsheimer’s setting forth of series XIV. Thankfully we are helped by the commentaries on chapters 8 and 9 of Daniel by Le Maistre de Sacy, whose working sessions on a new translation of the New Testament were attended by Pascal. These commentaries are explained for us by Wetsel. Thus the four kingdoms in Daniel 8:20–25 are the Chaldeans, the Medes and the Persians, the Greeks, and the Romans. The seventy weeks (Daniel 9:24–27) are weeks of years, i.e. 490 years, and refer to the year of Christ’s death. Sacy also takes Daniel 9:27 to imply that in the middle of the last week Christ will be put to death. Thus the 490 years is reduced to 486 years. But 486  years from when? The answer is, from the time when Artaxerxes orders the rebuilding of Jerusalem (Dan 9:25), in the twentieth year of Artaxerxes according to Nehemiah 2:1–8. By consulting his colleague Lancelot’s Abrégé de la chronologie sainte Sacy dates Artaxerxes’ order to 3550 in the ‘Year of the World,’ i.e. 3550 years after the creation. The commonly accepted view in Sacy’s (and Pascal’s day) was that Jesus was born in

Pascal’s Pensées and Butler’s Analogy 91 the 4000th Year of the World, meaning that the creation occurred in 4000 BC. A simple sum now gives the date of Christ’s crucifixion: Date of crucifixion = – 4000 + 3550 + 486 = AD 36 Wetsel summarizes Pascal’s argument: With the help of Le Maistre de Sacy, we may reconstruct the substance of Pascal’s argument as follows. The prophets predicted the arrival of Christ in the fourth of four great monarchies, after the seat of power had been removed from Jerusalem, and in the fifth century following return from exile. Christ arrived in the fourth of four successive monarchies, during the Roman occupation of Jerusalem, and in the fifth century following Daniel’s prophecy.14 Of course this is not quite how modern scholarship views things. For example, we now know that the universe began, not 6000 years ago but 13.8 billion years ago. Nevertheless, the dating of Artaxerxes’ order at 450 BC (= – 4000 + 3550) is not far out. However, the modern scholarly take on the prophecies of Daniel is quite different from Pascal’s pre-critical perspective. In Daniel 9:24–27 the angel Gabriel says this to Daniel: Seventy weeks of years are decreed concerning your people and your holy city, to finish the transgression, to put an end to sin, and to atone for iniquity, to bring in everlasting righteousness, to seal both vision and prophet, and to anoint a most holy place. Know therefore and understand that from the going forth of the word to restore and build Jerusalem to the coming of an anointed one, a prince, there shall be seven weeks. Then for sixty-two weeks it shall be built again with squares and moat, but in a troubled time. And after the sixty-two weeks, an anointed one shall be cut off, and shall have nothing; and the people of the prince who is to come shall destroy the city and the sanctuary. Its end shall come with a flood, and to the end there shall be war; desolations are decreed. And he shall make a strong covenant with many for one week; and for half of the week he shall cause sacrifice and offering to cease; and upon the wing of abominations shall come one who makes desolate, until the decreed end is poured out on the desolator. The first thing to note, with André Lacocque,15 is that the seventy weeks are related to Jeremiah’s prophecy (Jer 25:11–14 and Jer 29:10, cf. Dan 9:2), and what Daniel comes to understand is that, indeed, the ‘seventy years’ of Jeremiah are to be interpreted as ‘weeks of years.’ The first seven weeks are already past and run from the beginning of the captivity in 587 BC to the enthronement of the High Priest Joshua, who is to rebuild the Temple, in 538 BC. The remaining 62 weeks or 434  years arun, from Jeremiah’s

92  Pascal’s Pensées and Butler’s Analogy oracle in Jeremiah 25:1, from 605 BC to 171 BC, which is the year of the murder of the second ‘anointed one,’ namely the High Priest Onias III. This occurred during the reign of Antiochus IV ‘Epiphanes,’ and is described in 2 Maccabees 4. Given the scholarly consensus that Daniel was written in the Antiochene period, what looks like a prophecy may be considered to be a vaticinium ex eventu, as the early pagan critic Porphyry saw it. Does all this mean that we cannot relate the prophecies of Daniel to the time of Christ? As we shall see in Chapter 8 this is by no means the case. John Goldingay offers a way to do this. Referring to Daniel 9:24–27, he says: The detail of vv 24–27 fits the second-century B.C. crisis and agrees with allusions to this crisis elsewhere in Daniel. The verses do not indicate that they are looking centuries or millennia beyond the period to which chaps. 8 and 10–12 refer.  .  .  . The passage refers to the Antiochene crisis. Yet its allusiveness justifies reapplication of the passage, as is the case with previous chapters, in the following sense. It does not refer specifically to concrete persons and events in the way of historical narrative such as 1 Maccabees, but refers in terms of symbols to what those persons and events embodied, symbols such as sin, justice, an anointed prince, a flood, an abomination. Concrete events and persons are understood in the light of such symbols, but the symbols transcend them . . . . This means that the symbols may be manifested in ‘other embodiments.’ Goldingay continues: What these other embodiments are is a matter of theological, not exegetical, judgment—a matter of faith, not of science. But if I am justified in believing that Jesus is God’s anointed, and that his birth, ministry, death, resurrection, and appearing are God’s ultimate means of revealing himself and achieving his purpose in the world, they are also his means of ultimately achieving what the symbols in vv 24–27 speak of. It is this point that is made in traditional categories by speaking of a typological relationship between the events and people of the Antiochene crisis and deliverance and those of the Christ event and the End we still await.16 I shall look at how this ‘typological relationship’ might be established in Chapter 8. But let us end this brief analysis of Pascal’s apologetic by citing Krailsheimer’s verdict: His quite uncritical reading of the prophetic books, especially Daniel, led him into open absurdity, but in the last analysis these errors, gross as they are, should increase our respect for him. If Christianity is more

Pascal’s Pensées and Butler’s Analogy 93 than a moral code or a pious legend, it must be related to historical fact and tradition, and, if Christians are to follow Christ in respecting the Old Testament, it is more meritorious to examine the credentials of the Jewish books, even getting them wrong, than to ignore the problem. For his day, and for an amateur, Pascal did very well.17

Joseph Butler (1692–1752) Joseph Butler was brought up a Presbyterian but decided that his home was in the Church of England. After graduating from Oriel College, Oxford, he was appointed Preacher at the Rolls Chapel (1718–1726), which stood on the site of the Public Record Office. The Chapel was annexed to the Keeper of the Rolls, and the Rolls themselves (records of Chancery) were kept there. It was in the Rolls Chapel that Butler preached the famous Sermons, which made his reputation. It was while in ministry in the country parish of Stanhope, Co. Durham, that Butler produced the text I shall discuss shortly, his seminal Analogy of Religion (1736). His next appointment was as Chaplain to the Lord Chancellor, when he came in contact with Queen Caroline who appointed him her Clerk of the Closet. On her death-bed the Queen commended Butler to King George II, which resulted in his preferment, first to the see of Bristol and finally to Durham.18 There are two main editions of the Analogy, that edited by W. E. Gladstone and that edited by J. H. Bernard.19 References to quotations from the Analogy are given in brackets in the paragraphing of Gladstone. Terence Penelhum draws attention to this statement in the Advertisement at the beginning of the Analogy as being the key to the whole work: It is come, I know not how, to be taken for granted, by many persons, that Christianity is not so much a subject of enquiry; but that it is, now at length, discovered to be fictitious. And accordingly they treat it, as if, in the present age, this were an agreed point among all people of discernment; and nothing remained, but to set it up as a principal subject of mirth and ridicule, as it were by way of reprisals, for its having so long interrupted the pleasures of the world. On the contrary, thus much, at least, will be here found, not taken for granted, but proved, that any reasonable man, who will thoroughly consider the matter, may be as much assured, as he is of his own being, that it is not, however, as clear a case, that there is nothing in it. There is, I think, strong evidence of its truth; but it is certain, no one can, upon principles of reason, be satisfied of the contrary. And the practical consequence to be drawn from this is not attended to by every one who is concerned in it. (Advertisement, 2)20 Penelhum likens Butler’s stance here to Pascal’s Wager. If there is no disproof of Christianity, the sheer importance of it implies that it should be taken

94  Pascal’s Pensées and Butler’s Analogy with the utmost seriousness. If Christianity is true or merely credible it is both irreverent and rash to treat it with levity. (II.I.25). On the other hand, Butler does oscillate within the Analogy between showing the counterarguments to be weak and showing that Christianity is effectively ‘proved.’21 Butler frames his arguments in the Analogy in terms of probability. Indeed, he begins the Analogy by distinguishing probability from demonstrative proof: Probable evidence is essentially distinguished from demonstrative by this, that it admits of degrees; and of all variety of them, from the highest moral certainty, to the very lowest presumption. We cannot indeed say a thing is probably true upon one very slight presumption for it; because, as there may be probabilities on both sides of a question, there may be some against it: and though there be not, yet a slight presumption does not beget that degree of conviction, which is implied in saying a thing is probably true. (Introduction, 1) Although he often uses the word ‘proof’ of some assertion, Butler clearly means that the matter in question is ‘of high probability.’ Indeed, his meaning for probability here could be read as resembling the modern, Bayesian concept of ‘epistemic probability,’ the rational degree of belief to be accorded a proposition, based on the evidence for it. Elsewhere, however, Butler also uses probability in the relative frequency sense. Butler also argues that a cumulative case can be made for a proposition, by mounting up individual pieces of evidence which in isolation may each be of low probability: ‘But that the slightest possible presumption is of the nature of a probability, appears from hence; that such low presumption, often repeated, will amount even to moral certainty’ (Introduction, 2). And again, he says that one piece of evidence need not be considered as a proof by itself but all the evidence together may be ‘one of the strongest’ (II.VII.9). But here comes the frequency interpretation: Thus a man’s having observed the ebb and flow of the tide today, affords some sort of presumption, though the lowest imaginable, that it may happen again to-morrow: but the observation of this event for so many days, and months, and ages together, as it has been observed by mankind, gives us a full assurance that it will. (Introduction, 2) Butler cites an important example originally due to John Locke, namely the prince living in a warm climate who would naturally conclude ‘that there was no such thing as water’s becoming hard’ (Introduction, 3). In Britain of course we are in a different position, since we have seen ice many times. The example is significant because it is also cited by David Hume.22 To my mind

Pascal’s Pensées and Butler’s Analogy 95 it represents a real problem for Hume, since there is an analogy between not believing in ice and not believing in miracles. Locke himself makes the point that belief for the prince will be wholly dependent on testimony, and then even ‘the most untainted Credit of a Witness will scarce be able to find belief.’23 I argue in Chapter 6, in line with Butler’s cumulative style of argument, that the testimony of multiple witnesses can make even the a priori most unlikely phenomenon probable. Reinforcing the epistemic nature of probability, Butler says that for an infinite Intelligence nothing at all can be merely probable. Such a being will discern anything which is the possible object of knowledge ‘absolutely as it is in itself, certainly true, or certainly false.’ Famously, he goes on, ‘But to us, probability is the very guide of life’ (Introduction, 4). Butler is of course ignorant of the apparent ontological uncertainty in nature revealed in quantum theory, and debates this gives rise to in modern times. Butler notes that prudence may dictate action even on a presumption of low probability. (Introduction, 5, 6). That would be the case in common matters. One would not detonate explosives to demolish a building if one thought there was even a small probability of anyone being inside. Moreover, Butler’s argument has an analogy with Pascal’s wager—it would be prudent to bet on the Christian faith being true and to act in accordance with it, given the potential payoff. Butler is arguing in the Analogy against the deists, who were prominent opponents of Christian belief at the time. Of course, they already believed that there is a God, the intelligent author of nature, and thus Butler can take this as his starting point: . . . taking for proved that there is an intelligent Author of nature, and natural Governor of the world. For as there is no presumption against this prior to the proof of it: so it has been often proved with accumulated evidence; from this argument of analogy and final causes; from abstract reasonings; from the most ancient tradition and testimony; and from the general consent of mankind. Nor does it appear, so far as I can find, to be denied by the generality of those who profess themselves dissatisfied with the evidence of religion. (Introduction, 10) From this quotation, we see that Butler supports natural theology and various arguments that fall under that head. As Penelhum notes, Butler does not make these arguments himself, though he seems to endorse them, and takes the deist position as his starting point.24 We shall need to consider in due course how necessary this is and how it affects the arguments. Controversially, Butler includes within natural religion God’s moral governance of the universe with associated future rewards and punishments. His arguments include drawing an analogy from nature (always important to Butler) between dramatic changes known to us in nature, such as that

96  Pascal’s Pensées and Butler’s Analogy from a baby in the womb to an adult human, or a worm metamorphosing into a fly, with that from life to a future state after death (I.I.2, 3). Such arguments may not be as compelling to us today as Butler found them. He also sees revelation as confirming natural religion with regard to the moral governance of the world (e.g. II.VII.32). Butler sees miracles and prophecies, first, as giving natural religion a firmer base, since various miracles he lists which Jesus purportedly worked would support his teaching (II.I.7). But then, in like manner to Pascal, secondly, of course, miracles and prophecies support the specific claims of Christianity, and thus Butler can be seen as another forerunner of ramified natural theology. He begins his discussion of miracles by arguing that analogy with nature implies no presupposition against their occurrence. He argues against the idea that ‘stronger evidence is necessary to prove the truth of them, than would be sufficient to convince us of other events, or matters of fact’ (II.II.2). Butler is arguing here that there is no presumption against a revelation which in itself is a miracle, but also contains miracles within it (II.II.4, 5). That a revelation may not be discoverable by reason does not imply that there should be a presumption from analogy against it. This is because there are many features of nature of which we are ignorant and many which are beyond the reach of our faculties. Thus we cannot tell whether the universe is infinite or finite, even though we know it to be vast. Butler writes: And doubtless that part of it, which is opened to our view, is but as a point, in comparison of the whole plan of Providence reaching through eternity past and future; in comparison of what is even now going on in the remote parts of the boundless universe; nay in the whole scheme of this world. (II.II.4) Butler argues that there is also no presumption against revelation because it contains things which are unlike the known course of nature. This is because there is no presumption against the whole course of nature, which is unknown to us, being unlike anything which is known to us. Moreover, within nature there are many things unlike each other: For there is no presumption at all from analogy, that the whole course of things, or divine government, naturally unknown to us, and every thing in it, is like to any thing in that which is known; and therefore no peculiar presumption against anything in the former, upon account of its being unlike to any thing in the latter. And in the constitution and natural government of the world, as well as in the moral government of it, we see things, in a great degree, unlike one another; and

Pascal’s Pensées and Butler’s Analogy 97 therefore ought not to wonder at such unlikeness between things visible and invisible. (II.II.5) Interestingly, some deists, and most of those in Butler’s audience, accepted the ‘moral government’ of the world by God, as Samuel Clarke explains in his 1706 work Discourse Concerning the Unchangeable Obligations of Natural Religion, with which Butler was no doubt familiar.25 A miracle for Butler is a departure from the normal course of nature: ‘a miracle, in its very notion, is relative to a course of nature; and implies somewhat different from it, considered as being so’ (II.II.7). Penelhum sees Butler and Hume as therefore in agreement that miracles are violations of natural law.26 Of course revelation includes God’s miraculous interpositions, which certainly seem to be against nature. However, again, he appeals to our ignorance of nature. Thus we know ‘several of the general laws of matter,’ but we know nothing about numerous other laws, such as ‘by what laws, storms and tempests, earthquakes, famine, pestilence, become the instruments of destruction to mankind’ (II.IV.4). It is thus only by analogy that we can see that such things are also governed by general laws. Thus he states: And if that be a just ground for such a conclusion, it is a just ground also, if not to conclude, yet to apprehend, to render it supposable and credible, which is sufficient for answering objections, that God’s miraculous interpositions may have been, all along in like manner, by general laws of wisdom. (II.IV.5) These ‘miraculous interpositions’ might, then, be conformable to laws, but these would be general ‘laws of wisdom.’ Later Butler avers, surprisingly but now perhaps understandably in the light of what he has said hitherto, that ‘Christianity is vindicated, not from its analogy to natural religion, but chiefly from its analogy to the experienced constitution of nature’ (II.VIII.6). A further argument that there should be no presumption against miracles is that common facts are also extremely improbable before one examines the evidence, and he cites the story of Caesar (II.II.11). Butler says: ‘For, if there be the presumption of millions to one, against the most common facts; what can a small presumption, additional to this, amount to, though it be peculiar? It cannot be estimated, and is as nothing’ (II.II.12). This is an intriguing argument, but it does seem to be fallacious.27 Thus, it does seem to me that a miracle is a priori improbable in a different way to an event which conforms to known laws, and thus there should be an initial ‘presumption’ in Butler’s terms against its occurrence. However, I agree with

98  Pascal’s Pensées and Butler’s Analogy Butler that sufficient evidence can make the miracle probable (see Chapter 6). He is on better ground comparing miracles with new and surprising discoveries in nature, which should not be excluded a priori (II.II.15). A further point with which I agree is that if we assume the existence of God, miracles are more likely a priori than if we do not. Here Butler’s argument is linked to the natural theology he accepts, and which his opponents the deists accept (II.II.14). This would obviate a difficulty with Butler’s concept of analogy, pointed out by Penelhum.28 Thus Butler defines analogy in terms of inferring a general law from many instances (Introduction, 3, 7). An example would be seeing the sun rise every morning for many years and inferring that it will rise tomorrow and go on rising. However, his use in the religious context seems to be different from this, as when he denies a presumption against the occurrence of miracles by arguing that nature throws up new and surprising phenomena. It is here that Hume’s argument comes into play, but we shall defer our main consideration of Hume until Chapter 6. As Penelhum says, and as I have suggested above, Butler’s assumption of the deist position, while not arguing for it, obviates much of Hume’s criticism. I also believe that the modern Bayesian way of doing natural theology, whereby hypotheses are compared in terms of the degree to which they predict the evidence in question, obviates Hume’s arguments against natural theology based on analogical reasoning in the Dialogues Concerning Natural Religion.29 Having removed presumptions against the Christian revelation, Butler moves on to the positive evidence in its favour (ch. VII). He writes: Now in the evidence of Christianity there seem to be several things of great weight, not reducible to the head, either of miracles, or the completion of prophecy, in the common acceptation of the words. But these two are its direct and fundamental proofs: and those other things, however considerable they are, yet ought never to be urged apart from its direct proofs, but always to be joined with them. (II.VII.2) Many of Butler’s arguments strike me as essentially valid, even if they need revising in the light of historical criticism of the Bible. Thus he argues that the historical evidence for the miracles of Moses and the prophets are the same as that for ‘the common civil history of Moses and the kings of Israel’ (II.VII.3). Likewise the Gospels and Acts ‘afford us the same historical evidence of the miracles of Christ and the apostles, as of the common matters related in them’ (II.VII.3). He seems to be saying that there is something natural in the way miracles are interwoven into the Biblical narratives. This seems to me to be true, though the situation with the Old Testament is much more complex than with the New Testament, and can hardly be generalized in such a sweeping way. The New Testament very much fits with the historical background as we know it, including

Pascal’s Pensées and Butler’s Analogy 99 the existence of wonderworkers and other claimants to Messiahship than Jesus, and it is the New Testament miracles which will be my main focus in Chapter 7. Butler argues that the most obvious interpretation of the miracle stories is that they are historically true. Although other explanations may be proffered, these may be quite far-fetched and should not be admitted simply because they are asserted—positive evidence is required (II.VII.5, 6). Butler, rightly in my view, first singles out St Paul as a reliable witness. Naturally, given his time of writing, he considers all of Paul’s letters to be genuine, whereas the consensus of modern scholars is that there is a core of seven letters which came from the hand of Paul (Romans, 1 Corinthians, 2 Corinthians, Galatians, Philippians, 1 Thessalonians, and Philemon); the remainder are disputed—they could be Pauline, written by a disciple of Paul, or pseudonymous.30 As it happens the undisputed letters give Butler and ourselves plenty to go on. In support of the authenticity of 1 Corinthians, which he mainly focuses on, Butler cites a quotation of this epistle by Clement of Rome in the latter’s own epistle to the same church.31 Paul’s epistles provide a particular proof, says Butler, for the following reason: In them the author declares, that he received the gospel in general, and the institution of the Communion in particular, not from the rest of the apostles, or jointly together with them, but alone, from Christ himself; whom he declares likewise, conformably to the history in the Acts, that he saw after his ascension. So that the testimony of St Paul is to be considered, as detached from that of the rest of the Apostles. (II.VII.7) Further, in this epistle St Paul describes his own gift of working of miracles, and that of members of the Corinthian church, ‘incidentally,’ but rebukes them ‘for their indecent use of them.’ Moreover, he depreciates them ‘in comparison of moral virtues; in short he speaks to these churches, of these miraculous powers, in the manner any one would speak to another of a thing, which was as familiar and as much known in common to them both, as any thing in the world’ (II.VII.7). Like Pascal, Butler is aware of religious pluralism, in particular of the claims of Islam. But, he says, ‘It does in no sort appear that Mahometanism was first received in the world upon the foot of supposed miracles’ (II. VII.8). Particularly significant was the way great numbers were won over to Christianity on the testimony of just a few persons ‘and them of the lowest rank’ all at once. And this is unique to Christianity. Butler notes that converts made huge sacrifices to follow the ‘new religion’ of Christianity in forsaking their existing religion, friends, and community. Moreover, he makes the point, which will be of importance in our argument in Chapter 9, that many gave their lives for their new faith. He is

100  Pascal’s Pensées and Butler’s Analogy aware that enthusiasts have done this for the ‘most idle follies imaginable.’ However, there is a distinction between dying for opinions and for facts: And if the apostles and their contemporaries did believe the facts, in attestation of which they exposed themselves to sufferings and death; this their belief, or rather knowledge, must be a proof of those facts: for they were such as came under the observation of their senses. (II.VII.11) Also important, though of less weight, is the fact that the martyrs of the next generation suffered the same fate. They were not themselves eyewitnesses but had full opportunity to inform themselves of the facts (II.VII.12). Butler recognizes objections to his argument, anticipating in some ways the critique of Hume which was to come just a few years later. Thus the case could be weakened because it arises from ‘enthusiasm’ or the ‘incredibility’ of what is attested or by contrary testimony. The notion of ‘incredibility’ sounds like the concept of a low prior probability. However, Butler does not accept the incredibility of New Testament miracles and sees enthusiasm as a far-fetched explanation, in contrast to the ‘direct, easy, and obvious account of it, that people really saw and heard a thing not incredible, which they affirm sincerely and with full assurance, they did see and hear’ (II.VII.13). Even recognizing prejudice and bias in human affairs, testimony is still ‘a natural ground of assent’ and this assent ‘a natural ground of action’ (II. VII.16). Here he is anticipating something like Richard Swinburne’s ‘Principle of Testimony,’ which we come back to in later chapters. He does not see the dangers, such as being deluded by miracles, as any different from those in ordinary affairs in which humans are deluded (II.VII.17). Moreover, that some miracle report is falsely attested does not mean that another is not genuinely attested (II.VII.18). It cannot be a general rule to dismiss testimony on the grounds of prejudice, self-deception etc., but such would need to be established for the case in question. Until some such objection were indeed established, Butler says, ‘the natural laws of human actions require, that testimony be admitted.’ He goes on: It can never be sufficient to overthrow direct historical evidence, indolently to say, that there are so many principles, from whence men are liable to be deceived themselves and disposed to deceive others, especially in matters of religion, that one knows not what to believe. (II.VII.19) A further objection is that there might be contrary testimony. However, Butler rightly observes (as we shall see in more detail in later chapters), that in the case of Christianity, there is none (II.VII.21).

Pascal’s Pensées and Butler’s Analogy 101 When it comes to prophecy, Butler likewise makes observations ‘which are suggested by the analogy of nature, i.e. by the acknowledged natural rules of judging in common matters, concerning evidence of a like kind to this from prophecy’ (II.VII.22). If part of a prophecy is obscure but part understood as the genuine result of foresight, the latter is not invalidated, just as a document written part in cypher and part in language understood is not invalidated by the part that cannot be understood (II.VII.22). Again, Butler builds a cumulative case. Thus: ‘A long series of prophecy being applicable to such and such events, is itself a proof that it was intended of them. . .,’ whereas taking the prophecies individually they may well not appear to be ‘intended of those particular events to which they are applied by Christians’ (II.VII.24—see also, for example, II.VII.30). The kind of reasoning described here, joining separate pieces of evidence together to make a cumulative case, is for Butler the same as that involved in ‘common practice’ (II.VII.30), and, as Basil Mitchell notes, this appeal to ‘common practice’ is central to Butler’s argument.32 Mitchell notes how atheist materialist J. L. Mackie concedes that such a mode of argument could in principle be successful. For Mackie, the components brought together in such an argument would be: (1) reported miracles, (2) inductive versions of the design and consciousness arguments, picking out as marks of design both the fact that there are causal regularities at all and the fact that the fundamental natural laws and physical constants are such as to make possible the development of life and consciousness, (3) an inductive version of the cosmological argument, seeking an answer to the question “Why is there any world at all?” (4) the suggestion that there are objective moral values whose occurrence likewise calls for further explanation, and (5) the suggestion that some kinds of religious experience can be best understood as direct awareness of something supernatural.33 My point in this book is that a cumulative case can be made very strong indeed, and we have already discussed the cosmological and design components of such an argument; the argument from reported miracles will be treated later in the light of developments following Butler, and will be included in an overall cumulative case in Chapter 11. One important issue, as noted above, will be whether an inclination to theism needs to precede an argument from miracles, a point Mackie makes early on in his book. If the parties to a debate already agree ‘that there is an omnipotent God, or at any rate one or more powerful supernatural beings, they cannot find it absurd to suppose that such a being will occasionally interfere with the course of nature.’34 If they do not agree, then, according to Mackie, it is ‘well nigh impossible’ for an argument from miracles to find traction.

102  Pascal’s Pensées and Butler’s Analogy Mitchell makes an important point following his reference to Mackie: Prophecy fares no better, if it is understood as it must be for the purposes of the argument, as miraculous foretelling of the future. Hence if miracles and prophecy are to form any part of the Christian apologetic, it can only be as contributing to the explanatory power of the Christian story taken as a whole. It is not that the Resurrection of Christ, as we find it narrated in the Gospels, provides the warrant for interpreting Scripture as God’s revelation of himself, but rather that, given independent (though not coercive) grounds for belief in God, it makes better sense of all the evidence to accept the inspiration of the Bible and the genuineness of the Resurrection than to deny both and adopt an entirely naturalistic account of all the matters in dispute. In so far as the Bible, when carefully considered, lends itself to this interpretation, the antecedent assumptions about the existence and activity of God are themselves strengthened. This is how Butler’s argument ought to go if it is to conform to his dominant teaching about the nature of rational belief.35 As we shall see in later chapters, the approach outlined in this quotation is very much that adopted by Richard Swinburne, and indeed we follow out its consequences in more detail in subsequent chapters. Mitchell goes on to reflect on intellectual developments subsequent to Butler, which impact on whether and how we can pursue such an apologetic strategy today. Yet, as he indicates, Butler anticipated some of the criticisms of his method, for example in arguing that we are in no position to decide beforehand what form a revelation should take and therefore in no position to rule out any purported revelation a priori (II.III.7). Butler states: And therefore, neither obscurity, nor seeming inaccuracy of style, nor various readings, nor early disputes about the authors of particular parts; nor any other things of the like kind, though they had been much more considerable in degree than they are, could overthrow the authority of the scripture; unless the prophets, apostles, or our Lord, had promised, that the book containing the divine revelation should be secure from those things. (II.III.9) What is needed here is an assessment of the results of critical study of the Bible, examining to what extent such results are secure, and whether or not the basic integrity of the Scriptures is still intact when what is secure is taken into account. While we return to this topic in later chapters, we now give another way in which Butler anticipates such critical study. Now the prophets may not have understood the sense of their predictions in the way they are taken by later generations of Christians to be fulfilled.

Pascal’s Pensées and Butler’s Analogy 103 Fulfilment of the prophecy is proved, nevertheless, says Butler, by the words, rather than by whether or not the prophet understood them in that way (II. VII.27). Moreover, it does not matter that prophecies had a meaning for the age in which they were written. Interestingly, Butler recognizes Porphyry’s claim that Daniel’s prophecies could be seen as applicable to the time of Antiochus Epiphanes. Even if that were so, he says, that does not mean that the prophecies were not also applicable to a later age, and indeed the main point of inquiry should be to ascertain whether the prophecies are applicable to Christ ‘in such a degree, as to imply foresight’ (II.VII.28). Butler, like Pascal, is a man of his time in accepting a six thousand year history of the world (II.VII.34), but this is not crucial to his argument, though his statement that the sciences confirm this is anachronous (II.VII.44). Besides an accurate portrayal of prophetic history, of the history of the Jewish nation, and, where relevant to its impact on religion, of common history, most importantly the Old Testament predicts the coming of a Messiah: ‘It foretells, that God would raise them up a particular person, in whom all his promises should finally be fulfilled; the Messiah, who should be, in an high and eminent sense, their anointed Prince and Saviour’ (II.VII.37, see also II.VII.44, 45). Butler goes on to make the significant point that ‘This was foretold in such a manner, as raised a general expectation of such a person in the nation, as appears from the New Testament, and is an acknowledged fact . . . ’ (II.VII.37). Indeed, it is now acknowledged that there was widespread expectation in Judaea of a coming Messiah in the first century AD (and beyond). Butler also says that it is prophesied that the Messiah would be rejected by his people; and that, nevertheless he would be the ‘Saviour of the Gentiles’ (II.VII.38). He cites passages from Isaiah and Malachi in support of these claims. However, since we shall return in more detail to these prophecies in Chapter 8, we defer further discussion of the details until then. Butler’s chief point is that these prophecies were indeed fulfilled, ultimately leading to Christianity being spread throughout the world (II. VII.39). Revelation supports natural religion, he says, and does not destroy any proof of the latter from reason (II.VII.43). Butler argues that there is a general coherence to the Biblical narratives, in that they relate to the ‘manners of the age’ in which they are set, and nothing arises in any age which would be improbable in the light of preceding ages (II.VII.45). Customs and events of the time, mentioned both ‘incidentally and purposely’ in the New Testament, are confirmed by the contemporary writings of profane authors (II. VII.48). And this gives credibility to the accounts of miracles, especially as these are interwoven with the ‘common’ material. Moreover, ‘There is nothing in the characters, which would raise a thought of their being feigned; but all the internal marks imaginable of their being real’ (II.VII.45). For Butler, the Biblical narratives thus possess what J. B. Phillips described as ‘the ring of truth.’36

104  Pascal’s Pensées and Butler’s Analogy Interesting for his time too, Butler is not what we even today would call a ‘fundamentalist.’ Thus he does not argue for ‘inerrancy.’ On the contrary: There may be mistakes of transcribers, there may be other real or seeming mistakes, not easy to be particularly accounted for: but there are certainly no more things of this kind in the scripture, than what were to have been expected in books of such antiquity; and nothing, in any wise, sufficient to discredit the general narrative. (II.VII.47) Like Pascal, Butler sees not only the Christian faith, but also the Jewish people’s perpetuity as evidence of fulfilment of prophecy (II.VII.50, 51, 52, 53, 54, 55). Prophecies fulfilled give credibility to those which have yet to be fulfilled (II.VII.54). Butler’s case, as I  have emphasized, is a cumulative one, and, although he frequently uses the word ‘proof,’ in reality it is based on probability. He states: This general view of the evidence for Christianity, considered as making one argument, may also serve to recommend to serious persons, to set down every thing which they think may be of any real weight at all in proof of it, and particularly the many seeming completions of prophecy: and they will find, that, judging by the natural rules, by which we judge of probable evidence in common matters, they amount to a much higher degree of proof, upon such a joint review, than could be supposed upon considering them separately, at different times; how strong soever the proof might before appear to them, upon such separate views of it. For probable proofs, by being added, not only increase the evidence, but multiply it. (II.VII.60) In summary, Butler can be seen, like Pascal, as a quite remarkable forerunner of modern ramified natural theology but, despite his sophistication in many places, needing to be updated in the light of subsequent intellectual history.

Notes 1 A. J. Krailsheimer, ‘Introduction,’ in Blaise Pascal, Pensées, trans. A. J. Krailsheimer (London: Penguin, revised edition, 1995), xiii. 2 To be found in Pascal, Pensées, trans. Krailsheimer, Section III, 255–281. 3 L. Lafuma (ed.), Pascal: Oevres complètes (Paris: Éditions du Seuil, Collection l’Intégrale, 1963). 4 Pascal, Pensées, trans. A. J. Krailsheimer. 5 John Cruickshank, Pascal: Pensées (London: Grant & Cutler Ltd, 1983), 39–40. 6 Cruickshank, Pascal: Pensées, 18.

Pascal’s Pensées and Butler’s Analogy 105 7 Jon Elster, ‘Pascal and Decision Theory,’ in Nicholas Hammond (ed.), The Cambridge Companion to Pascal (Cambridge: Cambridge University Press, 2003), 53–74. 8 Elster, ‘Pascal and Decision Theory,’ 70. 9 Cruickshank, Pascal: Pensées, 35–36. 10 Cruickshank, Pascal: Pensées, 75. 11 Cruickshank, Pascal: Pensées, 74. 12 Krailsheimer, ‘Introduction,’ in Pascal, Pensées, xxv. 13 David Wetsel, ‘Pascal and Holy Writ,’ in Hammond (ed.), The Cambridge Companion to Pascal, 162–181. 14 Wetsel, ‘Pascal and Holy Writ,’ 179–180. 15 André Lacocque, The Book of Daniel (London: SPCK, 1979), 189–199. 16 John E. Goldingay, Daniel (Dallas, TX: Word Books, 1989), 267–268. 17 Krailsheimer, ‘Introduction,’ in Pascal, Pensées, xxv–xxvi. 18 Terence Penelhum, Butler: The Arguments of the Philosophers (London and New York: Routledge, 1985), 1–3. 19 Joseph Butler, The Analogy of Religion Natural and Revealed to the Constitution and Course of Nature, ed. W. E. Gladstone (Oxford: Oxford University Press, 1895); and idem, ed. J. H. Bernard (London: Macmillan, 1900). 20 Penelhum, Butler, 89–90. 21 Penelhum, Butler, 90–91. 22 David Hume, An Enquiry Concerning Human Understanding, Section X, ‘Of Miracles,’ in the first part of L. A. Selby-Bigge (ed.), Enquiries Concerning Human Understanding and Concerning the Principles of Morals by David Hume, third edition, revised by P. H. Nidditch (Oxford: Oxford University Press, [1748] 1975), 113–114. 23 John Locke, An Essay Concerning Human Understanding, edited with an introduction by P. H. Nidditch (Oxford: Oxford University Press, [1689] 1979), IV.XV.5, 656. 24 Penelhum, Butler, 94. 25 Penelhum, Butler, 100. 26 Penelhum, Butler, 175. 27 See Penlhum, Butler, 177. 28 Penelhum, Butler, 96–97. 29 David Hume, Dialogues Concerning Natural Religion, ed. J. M. Bell (London: Penguin, [1779] 1990). 30 See, for example, Joseph A. Fitzmyer, S. J., ‘Introduction to the New Testament Epistles,’ in Raymond E. Brown, S. S., Joseph A. Fitzmyer, S. J., and Roland E. Murphy, O. Carm (eds.), The New Jerome Biblical Commentary (London and New York: Geoffrey Chapman, 1989), 768–771, 770. 31 ‘The First Epistle of Clement to the Corinthians,’ ch. XLVII, in Alexander Roberts and James Donaldson (eds.), Ante-Nicene Fathers, vol. 1, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson, 1994), 5–21, 18. 32 Basil Mitchell, ‘Butler as a Christian Apologist,’ in Christopher Cunliffe (ed.), Joseph Butler’s Moral and Religious thought: Tercentenary Essays (Oxford: Oxford University Press, 1992), 97–116, 99. 33 J. L. Mackie, The Miracle of Theism: Arguments for and against the Existence of God (Oxford: Oxford University Press, 1982), 251–252; quoted in Mitchell, ‘Butler as a Christian Apologist,’ 103. 34 Mackie, Miracle of Theism, 27. 35 Mitchell, ‘Butler as a Christian Apologist,’ 108–109. 36 J. B. Phillips, Ring of Truth: A  Translator’s Testimony (London: Hodder  & Stoughton, 1967).

6 The rationality of belief in miracles

Introduction One of the important aspects of ramified natural theology to be explored is the credibility of claimed miracles in the Bible, and supremely that of the resurrection of Jesus. Before dealing with specific cases in later chapters, here I shall argue that in principle it can be perfectly rational to believe in miracles, based on the evidence of testimony. So this is a piece of natural theology which is bringing us much closer to ramified natural theology. The Scottish Enlightenment philosopher David Hume defines, albeit controversially, a miracle to be ‘a violation of the laws of nature,’ and in his famous essay ‘Of Miracles’ he argues that it is always more probable that a witness is in error for whatever reason than that the miracle occurred. This chapter will comprise a critique of Hume, especially by arguing that Hume’s understanding of probability was deficient and bringing to bear the apparatus of Bayesian confirmation theory.1 I  shall show that in principle the accumulation of independent testimony to a single miracle can make that miracle probable and that the aggregation of testimony to many miracles can make it probable that at least one occurred. I shall also ask whether or in what circumstances testimony to miracles can provide evidence for the existence of God.

David Hume and miracles The critical passage from Hume’s ‘Of Miracles,’ in his Enquiry Concerning Human Understanding, states this: The plain consequence is (and it is a general maxim worthy of our attention, “That no testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous, than the fact, which it endeavours to establish; and even in that case there is a mutual destruction of arguments, and the superior only gives us an assurance suitable to that degree of force, which remains, after deducting the inferior.”2

The rationality of belief in miracles 107 Some philosophers and modern day scientists make categorical assertions that miracles cannot happen. Hume sometimes says that and sometimes, as in the above quotation, renders their occurrence not impossible but just too improbable. Presumably Hume is thinking that one should not believe a report of an event which falls outside the normal pattern of cause and effect. Indeed, in another place, he writes of ‘uniform experience’ being against miraculous occurrences. This is already a slightly strange position for him, since elsewhere he is critical of inductive inference, i.e. predicting future behaviour on the basis of past experience. In other words, Hume is critical of arguing that we should expect the normal pattern of cause and effect always to be followed, since we have no logical reason to think that it will. C. S. Lewis succinctly countered Hume’s argument in his own book entitled Miracles. From the observed regularity of nature Hume concludes that there is ‘uniform experience’ against miracles. He thus concludes that miracles are the most improbable occurrences and that it is always more probable that the witness was lying or mistaken. But the argument is circular for, as Lewis says: Now of course we must agree with Hume that if there is absolutely “uniform experience” against miracles, if in other words they have never happened, why then they never have. Unfortunately we know the experience against them to be uniform only if we know that all the reports of them are false. And we can know all the reports to be false only if we know already that miracles have never occurred.3 Hume has his modern counterparts. In 1984 The Times published a letter from fourteen British professors of science, including six Fellows of the Royal Society, which supported belief in the gospel miracles. The prestigious scientific journal Nature published a leading article in response in which it defined miracles as ‘inexplicable and irreproducible phenomena (which) do not occur.’ This is indeed seriously to beg the question, even more than Hume did. Fortunately, in this case the balance was redressed with a lucid attack on the position of the journal by Professor R. J. Berry, one of the original fourteen.4 In the quotation at the beginning of this passage, Hume seems to concede that, given testimony to the miraculous, it is simply more likely for the testimony to be false than for the miracle to have occurred, i.e. it is a matter of probability not certainty as to whether the miracle occurred. In other words, the ‘uniform experience’ is at least open to question, even if in Hume’s view the probability of the miracle occurring is low. It is thus Hume’s take on probability which we need to examine in more detail. At this stage, we are not addressing, with Berry et al., the gospel miracles as such, but simply examining as a matter of principle whether one should believe accounts of miracles. Thus, this is still ‘bare’ natural theology, but

108  The rationality of belief in miracles coming closer to ramified natural theology since we intend to apply what we have learned to the New Testament. If miracles are not ruled out by fiat, but are a matter of probability, then it becomes important to analyse the problem using the valid framework of probability theory, and that is what I do in the bulk of this chapter.

Hume’s definition of ‘miracle’ Hume defines a miracle to be ‘a violation of the laws of nature.’5 We can accept that for the purposes of argument, though it would not be the definition of miracle for many theologians. St Augustine (AD 354–430) says that God created what we would call the laws of nature, and acts through them, but is not bound by them: God has established in the temporal order fixed laws (certas temporum leges) governing the production of kinds of beings and qualities of beings and bringing them forth from a hidden state into full view, but His will is supreme over all. By His power He has given numbers to His creation, but He has not bound His power by these numbers.6 Augustine also says that what we might think of as miracles are not against nature. Referring to the miracle of Exodus 7:10, he writes: All serpents require a certain number of days according to their kind to be implanted, formed, born, and developed. Did Moses and Aaron have to wait all those days before the rod could be turned into a serpent? When events like this happen, they do not happen against nature except for us, who have a limited knowledge of nature, but not for God, for whom nature is what He has made.7 For Augustine, the laws of nature, as we discern them, are simply God’s normal way of working, but any way he worked would in reality be in accordance with nature, since ‘nature is what He has made.’ Moreover, in another place Augustine argues that God, the author of nature, can change the natures of objects so that they exhibit altered properties.8 St  Thomas Aquinas (1225–1274) is similar to Augustine, but makes a distinction between primary and secondary causes. God is the primary cause of all things—he it is who causes them to exist and gives them the powers they have to act. But creatures can and do act as secondary causes through the powers with which they have thus been endowed by God.9 Aquinas further believes that God is involved intimately in the natural world and he appeals to Job 10:11: ‘Thou hast clothed me with skin and flesh; thou hast put me together with bones and sinews.’ Just like Augustine he sees the development of a human person from the womb to adulthood in terms of God’s action, but God acting in and through the processes with which he has endowed nature.

The rationality of belief in miracles 109 For Aquinas, as for Augustine, this does not limit God’s action because God freely created the secondary causes and they are subject to him. However, Aquinas says that God is free to act outside the order of secondary causes, for example by producing the effects without the cause or by doing something to which the secondary causes do not pertain.10 This is probably what Aquinas has in mind when he later defines a miracle as ‘an event that occurs outside the natural run of things.’11 The Latin here is aliquid fit præter ordinem naturæ and is better understood as ‘something is done besides or beyond the order of nature’ rather than ‘contrary to the order of nature.’ Thus ‘violation of the laws of nature’ may not be how mainstream Christian thinkers have thought about miracles. Indeed, the modern concept of ‘laws of nature’ post-dates Augustine and Aquinas and the Bible itself. However, it is worth noting that ancient and pre-modern people did not need the concept of scientific laws of nature to realize that virgin women do not give birth and that dead men do not rise. In arguing that reports of miracles hail from uncivilized and remote peoples Hume seems not to have grasped this point. It is also worth noting that, whilst coming to accept such exceptional acts of God, Biblical writers also saw, as do modern theologians, that the general consistency of the processes of nature, what we would call the laws of nature, are a sign of God’s faithfulness (e.g. Acts 14:17).

Bayesian framework A number of authors in the last few decades have considered Hume’s celebrated essay on miracles from a Bayesian perspective. Much of this discussion has centred on the issues of whether Hume can rightly be considered a proto-Bayesian, how exactly to interpret him in terms of the calculus of probabilities, and hence how to judge whether his argument is correct. There has been some further discussion of whether testimony for miracles can provide evidence for the existence of God. An important point for further consideration is the impact on the argument of multiple testimony, understood either in the sense of several independent testimonies for a single miracle or of independent testimonies for a number of miracles. Atheist philosopher J. L. Mackie12 has noted the strong evidential force of two independent testimonies for a single miracle, and John Earman13 has provided some analysis of this from the perspective of the probability calculus. Indeed, since that paper Earman has produced an excellent book on Hume and Bayes.14 Also, Roy Sorensen15 has noted the possibility that combined testimony for many miracles may yield a high probability that at least one has occurred. However, this latter claim has been dismissed by George Schlesinger16 on the grounds that the occurrence of one miracle is not independent of the occurrence of any other. This latter assumption requires further examination. Both Schlesinger and Swinburne17 have argued that testimony for miracles provides evidence for the existence of God. Richard Otte18 has challenged

110  The rationality of belief in miracles Schlesinger’s version of the argument. Again, this topic warrants further consideration. In this chapter I establish the Bayesian procedure to be followed, noting some of the related work done elsewhere, before proceeding to examine the issues of multiple testimonies and whether testimony for miracles provides evidence for the existence of God. Hume writes in the passage quoted of subtracting what he calls ‘degrees of force,’ which sounds distinctly non-Bayesian. It seems that what he is saying is that a miracle is an event with very low probability. By describing testimony as miraculous he presumably means that there has to be a very low probability that the testimony is false. In the case when the miracle and the falsehood of the testimony to it are equally miraculous, which we interpret to mean ‘have equal but very low probabilities,’ Hume would ostensibly give zero ‘degree of force’ or probability to the miracle given the testimony. That is presumably because the a priori probability of the miracle and the probability that the testimony is false cancel out. This interpretation is clearly incorrect since it implies that ‘miraculously’ reliable testimony has reduced rather than increased the probability of the miracle’s occurrence. Hume’s Enquiry was published in 1748 and Bayes’s theorem, the fundamental theorem of probability formulated by the Revd Thomas Bayes and discussed in previous chapters, was published posthumously in 1763, still within Hume’s lifetime. It is not clear that Hume knew about it—certainly he did not ever take it into account. However, the more important question is not the historical but the philosophical one, namely ‘Can Hume’s argument be rephrased in Bayesian terms?’ since Bayes’s theorem is the modern and correct way to phrase Hume’s argument in probability terms. It tells us how we should revise the a priori or prior probability we might have of some hypothesis, that is, the probability based solely on background knowledge, before we take account of some specific piece of evidence for it, in the light of such evidence. On a Bayesian interpretation, provided the probability of the testimony given that the miracle occurred is greater than the probability of the testimony given that it did not, the miracle is said to be confirmed. The definition of ‘confirmation’ here is that the probability of the miracle given the testimony for it is greater than its probability before the testimony was taken into account (as we saw in Chapter 2). If the testimony is ‘non-miraculous’ in Hume’s sense, it may not be raised to a large value. However, in the case of equally ‘miraculous’ miracle and falsehood of testimony, the probability of the miracle having occurred given the testimony would be 0.5—not zero as Hume seemed to claim. In this case one is said to be indifferent between the miracle and the falsehood of the testimony.19 In order to make the miracle rationally acceptable, one requires the probability of the miracle having occurred given the testimony to be greater than the indifference level of 0.5, and for this to entail, on Hume’s reasoning, the probability that the testimony is false to be less than the prior probability of the miracle’s occurrence.

The rationality of belief in miracles 111 Following Hume, then, we define a miracle to be ‘a violation of the laws of nature.’20 It is true that Hume modified this simple definition in a footnote: ‘A miracle may be accurately defined, a transgression of a law of nature by a particular volition of the Deity, or by the interposition of some invisible agent.’21 We have seen that in Christian theology, God may not be acting ‘against nature’ in performing what we would see as miracles. However, this modification also confuses the issue: it is one thing to receive testimony to an event that may be deemed to constitute a violation of a law of nature, and quite another to assess the likelihood of a divine origin of the event (see Earman).22 In addition, including the existence of God in background knowledge may significantly affect the prior probability of the miracle. We develop the Bayesian framework as follows:23 Let M denote a specific miracle occurring at space-time location L. Let K denote background knowledge that a certain witness W was at L and made a report about what happened at L. Let T be testimony to M provided by witness W, i.e. T is W’s assertion that M occurred. Then we define the following probabilities: Let P[M] = prior probability of the occurrence of M, P[M|T] = probability that M has occurred given T, P[T|M]  =  probability that T is provided given that M has indeed occurred, P[T|~M]  =  probability that T is provided given that M has not occurred. All probabilities are further conditioned on K, though this dependence is omitted for convenience. M might be, to take Schlesinger’s example: ‘the walls of Jericho collapsed upon the blowing of the Israelite trumpets.’24 T would be the statement of a witness W, at Jericho at the time, that M had supposedly occurred (though whether in this case there was such a witness would be a moot point for Biblical scholars to grapple with). With this notation, Bayes’s theorem states: P [ M|T ] =

P [ T |M ] P [ M ] P[T]

Or, more helpfully:

P  M|T  

P T |M  P  M 

P T |M  P  M   P T | ~ M  P ~ M 

(6.1)

112  The rationality of belief in miracles In what follows the symbols have these meanings: > is greater than; ≥ is greater than or equal to; >> is much greater than; < is less than; ≤ is less than or equal to; P[T | ~ M] → PM|T ] > P[M] which can be interpreted as saying that the evidence of testimony enhances the probability that a miracle occurred. Now a miracle is, as Hume asserts, intrinsically improbable. If we also assume that W is highly reliable, then the following approximations and inequalities hold for the probabilities entering equation (6.1):

P[T |M] ≈ 1 P  M   1 P[T | ~ M]  0): P  M|T   0.5  P  M   P T | ~ M  (6.3) Or, equivalently: P  M|T   0.5  P  M   P T | ~ M  Other authors provide interpretations of, or advances on, Hume which are broadly equivalent to the above, especially when the régime of values taken by the probabilities is taken into account (see, for example, Owen,25 Dawid and Gillies,26 Schlesinger,27 Sobel,28 Millican,29 and Earman).30

The rationality of belief in miracles 113 Whether or not Hume meant this, and there is some dispute among philosophers of science about this, the comparison between P[T|~M] and P[M] seems to me the most natural to make, having a good rationale, and we shall proceed on this basis. As a convenient shorthand we shall denote testimony for which P[M] > P[T|~M] as ‘miraculous’ and testimony for which P[M]   q whereas for a highly reliable one p >> q. Either way, as n → ∞, (q/p)n → 0 and so P[M|Tn] → 1 (N.B., in mathematical as opposed to logical expressions the symbol ‘→’ means ‘tends to;’ ‘∞’ is the symbol for ‘infinity’). More significantly, as Earman points out, no matter how small P[M] one can choose n (finite) such that P[M|Tn] > 0.5. The point is that, as n increases, the factor (q/p)n (which arises because of the multiplicative law for combining independent probabilities) decreases very rapidly, soon becoming comparable with, and then falling below, P[M]. Whether or not there have been examples in history where the combined testimony of independent witnesses has enhanced the probability of a miracle’s occurrence to greater than indifference level is a matter for empirical investigation. The point is that Hume’s (already questionable) assumption that no individual testimony can be this reliable does not preclude the combined testimonies being so.

The aggregation of testimonies to several miracles Now let us consider the case of independent testimonies for two miracles. Suppose specifically that we have independent testimonies T1 and T2 for miracles M1 and M2 respectively. Now one might naïvely suppose that we can combine the probabilities for miracles M1 and M2 occurring given testimonies T1 and T2 as follows: Probability that at least one of Ml and M2 occurs = 1 – (Probability that neither M1 nor M2 occurs) = 1 − (Probability that M1 does not occur) x (Probability that M2 does not occur) i.e., using the notation of the probability calculus,





P  M1  M2 |T1  T2   1  1  P  M1 |T1  1  P  M2 |T2 



 P  M1 |T1   P  M2 |T2   P  M1 |T1  P  M2 |T2  (6.6) Here ∨, read ‘or,’ is the ‘inclusive disjunction’ symbol of propositional logic and M1 ∨ M2 means either M1 or M2 or both; ∧, read ‘and,’ is the symbol for ‘conjunction’ and T1 ∧ T2 means both T1 and T2.

The rationality of belief in miracles 115 One can imagine, given that there are very many reported miracles, iterating equation (6.6) to build up a substantial probability that at least one occurred. The occurrence of at least one miracle might well become probable (at the end of this section we show how this is so in the simplified case for which all the probabilities of the miracles given their individual testimonies are equal). Sorensen notes that Hume has at best established a case-by-case scepticism. The wise man will not believe any testimony that a specific miracle took place. However, this does not rule out the wise man, equally rationally, believing that one of a set of reported miracles, though it is not known which, did actually occur. If the analysis in the previous section is correct, then case-by-case scepticism is also refuted on the grounds of multiple testimonies, but even if this is not so, testimony for many miracles may lead one rationally to accept that at least one occurred. Schlesinger has pointed out a possible flaw in this argument, namely that it assumes that M1 and M2 are independent. Schlesinger writes the formula for combining probabilities as P[M1 ∨ M2 ] = P[M1 ] + P[M2 ] −P[M1 ∧ M2 ] = P[M1 ] + P[M2 ] −P[M2 | M1 ]P[M1 ]



(6.7)

In the independence case P[M2|M1]  =  P[M2], but Schlesinger argues that P[M2|M1] = 1. He does so on the grounds that if we know that one miracle has occurred then our reasoning to the intrinsic improbability of miracles in general is wrong, and we should instead assume that they are likely. Surely, however, they are not dependent to the extent which Schlesinger suggests. The known occurrence of a turning of water into wine at some point in space and time does not license me to believe in a levitation at some other location in space and time—even if I have testimony to the latter. Miracles do not all stand or fall together. This discussion should more accurately proceed on the basis of probabilities conditioned on the evidence of testimony (this is implicit in Schlesinger’s discussion, but needs to be made explicit). What we are really interested in is the quantity P[M1 ∨ M2|T1 ∧ T2], i.e. the probability that at least one miracle has occurred given the testimony for each. The correct, full formula for combining the probabilities is: P[M1 ∨ M2 | T1 ∧ T2 ] = P[M1 | T1 ∧ T2 ] + P[M2 | T1 ∧ T2 ] − P[M1 ∧ M2 | T1 ∧ T2 ]  P[M1 |T1  T2 ]  P[M2 |T1  T2 ]  P[M2 |M1  T1  T2 ] P[M1 |T1  T2 ] 

(6.8)

116  The rationality of belief in miracles We make the following observations with regard to this equation: 1 The occurrence of M1 may well be evidence for M2, and vice versa, as Schlesinger suggests. Hence we should retain the M2|M1, dependency. 2 T2 provides evidence for M1 indirectly through the possible occurrence of M2. However, it is surely safe to assume that this evidence is far weaker than T1. The same comment applies with indices reversed. With these assumptions (6.8) can be approximated to obtain P[M1 ∨ M2 |T1 ∧ T2 ] = P[M1 |T1 ] + P[M2 |T2 ] −P[M2 |M1 ∧ T2 ]P[M1 |T1 ] (6.9) The key factor to calculate in (6.9) is P[M2|M1 ∧ T2]. In the case of independence this is simply P[M2|T2], and the formula reduces to that of (6.6) above. We agree with Schlesinger that this is too simplistic, but disagree with him that it should be taken as unity. Both M1 and T2 are evidence for M2. P[M2|M1 ∧ T2] can be expanded using Bayes’s theorem as follows: P[M2 |M1 ∧ T2 ] =

 P[T2 |M1 ∧ M2 ]P[M2 | M1 ] (6.10) P[T2 |M1 ∧ M2 ]P[M2 |M1 ] + P[T2 |M1∧ ~ M2 ]P[~ M2 |M1 ]

We can surely make the further simplifying assumption that the occurrence of M1 has no bearing on testimony T2 for M2, given that M2 has not occurred. Then, with similar assumptions about the reliability of testimony as were made in our earlier discussion, we can approximate the above thus: P[M2 |M1 ∧ T2 ] =

P[M2 | M1 ]  P[M2 |M1 ] + P[T2 | ~ M2 ]P[~ M2 |M1 ]

(6.11)

Now if P[M2|M1] >> P[T2|~M2] P[~M2|M1], then P[M2 |M1 ∧ T2 ] ≈ 1. This would appear to be Schlesinger’s position. From this it would follow that the occurrence of at least one miracle is no more likely than the occurrence of a single one. Now we can agree that the occurrence of M1 makes M2 more likely because we no longer have near absolute scepticism about miracles in general, i.e. we would argue that P[M2|M1] >> P[M2]. However, as noted above, the occurrence of a turning of water into wine somewhere will not make the occurrence of a levitation elsewhere by any means probable; it will merely remove the prejudice we had that the latter was as intrinsically unlikely as anything could be. Thus we should still expect P[M2|M1]  0.5. Swinburne argues that testimony increases the probability that a miracle, defined as a violation of a natural law, occurred. The occurrence of a miracle in turn increases the probability that God exists. Hence testimony for miracles increases the probability that God exists. More precisely, Swinburne states: Certainly witness-reports can add to the probability that a violation occurred and so add to the probability that there is something not to be explained by natural processes. If e is merely witness-reports that a violation occurred, which are substantial evidence that a violation did occur (because the occurrence of the reports is not easily to be explained in some other way, e.g. in terms of the witnesses being misled by some non-miraculous phenomena), then e is more to be expected if a violation did occur that if it did not, and so marginally more to be expected if there is a God capable of bringing about violations than if there is not. The occurrence of much evidence of miracles is indirect evidence of the existence of God for it is evidence of the occurrence of events which natural processes do not have the power to produce. In so far as there is much such evidence, it would provide a good C-inductive argument to the existence of God.41 Swinburne does not spell out this argument in probability calculus terms. However, we might construct a probabilistic version of the argument as follows, where we substitute T for e in keeping with our own notation, and write G for the hypothesis that God exists: 1 P[T|M] > P[T|~M] ↔ P[M|T] > P[M]. 2 Suppose we have T such that P[T|M] > P[T|~M]. 1 and 2 → 3 P[M|T] > P[M] (i.e. T provides evidence for M).

4 Suppose similarly that M is such that P[G|M]  >  P[G] (i.e. M provides evidence for G).

3 and 4 → 5 P[T|G] ≳ P[T|~G]. 5 → 6 P[G|T] ≳ P[G].

The rationality of belief in miracles 121 The symbol ‘≳’ here represents ‘marginally greater than.’ Statement 1, and hence the step from 1 and 2 to 3, is uncontroversial, following straightforwardly from Bayes’s theorem. These lines merely express the fact that testimony enhances the probability of a miracle. Statement 4 is a similar expression asserting that M is a miracle which enhances the probability that God exists. The step from 3 and 4 to 5, and hence 6, is the one which is not so obvious, and requires a more careful analysis, as given below. Schlesinger42 utilizes Bayes’s theorem in an attempt to show that testimony for miracles is evidence for the existence of God. He begins with Bayes’s theorem in the form P G|T  

P  M|T  P[G | M  T ] P  M|G  T 

(6.13)

where again we have modified his notation to be consistent with ours. By replacing T with ~T in the above and dividing the two equations obtained one derives the following: P G|T 

P G| ~ T 



P  M|T 

P  M| ~ T 

.

P[G | M  T ] P  M|G ~ T  (6.14) . P[G | M  ~ T ] P  M|G  T 

Schlesinger argues that the first ratio here is greater than 1, the second may be taken as equal to 1, and the third may also be taken as 1 (for the reasons for these judgements, see Schlesinger’s paper). With Otte43 I  agree that the first two ratios are as Schlesinger describes them (and I  shall say a little more about this shortly). However, Otte has also rightly pointed out that, contrary to Schlesinger’s argument, the third ratio here cannot necessarily be equated to unity. Indeed one would expect this last ratio to be less than 1. I am much more likely to believe that the Red Sea parted if God exists and I have testimony to the event, than if God exists and I have no such testimony. Even if the miracle we are talking about is the resurrection of Jesus, and G is the Christian God, testimony will still make a difference to our degree of belief in the miracle’s occurrence. After all, no one claims that Moses, for example, rose from the dead, so my belief in God does not warrant belief in Moses’ resurrection apart from testimony. Moreover, Christians tend to argue that Jesus’ resurrection vindicates (provides evidence for?) the unique claims he makes about his relationship to God. To argue that the probability of this miracle’s occurrence is dependent only on the existence of God and independent of testimony is insufficient. Schlesinger acknowledges that at the time of the miracle it is necessary that ‘circumstances characteristic to those, which from a religious point of view demand the occurrence of a miracle, obtained.’44 The problem is that arguably such circumstances did obtain both for Moses and Jesus. There is an

122  The rationality of belief in miracles analogy here with our discussion of the probability of the occurrence of one miracle given the occurrence of another. The existence of God would remove our prejudice that miracles were as unlikely as anything could be, but would not license our belief in any particular miracle apart from testimony. Similarly, given the existence of God as background knowledge the probability of the resurrection of God incarnate would be much higher than it would apart from this background knowledge, but testimony is still required to make the resurrection probable. We see how the probability of the existence of God affects our overall probabilities for Christian claims in chapters 10 and 11. As with Swinburne’s argument, we need a more careful analysis of this question, which is akin to our discussion above of testimony for many miracles. Bayes’s theorem will again provide the necessary tool. Essentially, we wish to express P[G|T] in terms of quantities we know or can derive: P[M|T] and P[G|M]. We begin from Bayes’s theorem in the same form that Schlesinger has it: P G|T  

P  M|T  P[G | M  T ] P  M|G  T 

(6.15)

We utilize reasoning similar to that of our discussion of testimony for many miracles. Thus P[G|M ∧ T] can be approximated by P[G|M]. This is because, given M’s occurrence, testimony to M will not much further affect the probability of G. The key factor to evaluate is P[M|G  ∧  T]. The occurrence of M will depend in general on both G and T (as M2 depended on M1 and T2 above): P  M|G  T  



P T |M  G P[M | G]

P T |M  G P  M|G  P T | ~ M  G P[~ M | G]

P  M|G

P  M|G  P T | ~ M 



(6.16)

The approximation here depends on the following three assumptions: 1 P[T|M ∧ G] ≈ 1.   P[T|M ∧ G] is virtually identical to P[T|M]—if the miracle has occurred, testimony is near certain, regardless of whether there is a God or not. 2 P[~M|G] ≈ 1.   P[~M|G] = 1 − P[M|G], and we assume that the probability of a miracle’s occurrence given God, P[M|G], is higher than its prior probability P[M], but still not particularly high, testimony still being required to maximize the probability. The argument is analogous to that about two miracles in my earlier section.   Of course, arguably P[M|G] might be high in some specific case like the resurrection of Jesus, with G the Christian God, although I do not think

The rationality of belief in miracles 123 this is necessarily the case for reasons advanced earlier. If it were the case it would invalidate the particular approximation here, but strengthen the overall argument that T provides evidence for G. In chapters 10 and 11, as well as the prior probability of the resurrection of Jesus, I discuss the prior probability of God becoming incarnate and argue for a ‘moderately low’ value, i.e. less than ½ but not extremely low. 3 P[T|~M ∧ G] ≈ P[T|~M]. It is whether a miracle has occurred or not which is the dominant factor in determining whether there is testimony, not the existence of God. Substituting from equation (6.16) into (6.15), we obtain P G|T   P  M|T  P G|M .





P  M|G  P[T |~ M] P  M|G

P T |M  P  M 



(6.17)

P  M|G P G P  M|G  P[T |~ M] . P T |M  P  M   P T | ~ M  P ~ M  P M P  M|G

P M

.

P  M|G P G P  M|G  P T | ~ M  . P  M   P T | ~ M  P M P  M|G .

i.e. P G|T   P G.

P  M|G  P T | ~ M  P  M   P T | ~ M 

(6.18)

We are now in a position to write down the condition that testimony T for miracle M provides evidence for the existence of God, in the sense of making the hypothesis G more probable than its a priori value, i.e. P[G|M] > P[G]. This condition is P  M|G  P[T |~ M] P  M   P T | ~ M 

 1 (6.19)

This condition is of course always met. However, it turns out that Swinburne is correct in asserting that T may only provide marginal evidence for G over ~G. All depends on the relative values of P[M], P[M|G], and P[T|~M]. If T is very good testimony for M, e.g. it represents the combined testimony T1 ∧ T2 ∧ . . . Tn of independent witnesses with parameters such that P[M] >> P[T|~M], then condition (6.19) will approximate to P  M|G P M

 1 (6.20)

124  The rationality of belief in miracles In fact, it may well be that P[M|G] >> P[M], so that T (which is virtually equivalent to asserting that M has occurred) will be very good evidence for G in this case. On the other hand, if we have only singular, non-miraculous testimony T for M, such that P[M]  1

so that P[J2′|E ∧ k] ≈ 1. Thirdly, let J3′ be the hypothesis that this same unnamed prophet fulfilled numerous prophecies. Then ~J3′ would be the negation of J3′ so that any resemblance between J′’s actions (and ways he was acted upon) and the

Towards a fuller picture 215 prophecies would be purely accidental or illusory. In general the prophecies considered would be those in the literature of the people out of whom the prophet arises, through whom perhaps God has been working over the course of history. These would be known to the prophet’s contemporaries. Interestingly, however, C. S. Lewis sees stories of dying and rising gods in other religious contexts as also somehow prefiguring, if not prophesying, the one true dying and rising God seen in Jesus—resembling the way he sees the treatment meted out to Plato’s idealized righteous man as prefiguring Christ: The Divine light, we are told, “lighteneth every man”. We should, therefore, expect to find in the imagination of great Pagan teachers and myth-makers some glimpse of that theme which we believe to be the very plot of the whole cosmic story—the theme of incarnation, death and re-birth. And the differences between the Pagan Christs (Balder, Osiris, etc.) and the Christ Himself is much what we should expect to find. The Pagan stories are all about someone dying and rising, either every year, or else nobody knows where and nobody knows when. The Christian story is about a historical personage, whose execution can be dated pretty accurately, under a named Roman magistrate, and with whom the society that He founded is in a continuous relation down to the present day. It is not the difference between falsehood and truth. It is the difference between a real event on the one hand and dim dreams or premonitions of that same event on the other.1 Clearly, what Lewis is talking about is indeed intimation and not the kind of detail and accuracy one might expect in the case of prophecies to be found within a people ostensibly chosen by God for his special purpose of revealing himself to the nations. The latter hypothetically confirmed prophecies might comprise several categories (as in Chapter 8 we outlined them for Jesus): prophecies over which J′ has no control; prophecies which are under J′’s control, perhaps which somehow recapitulate the story of God’s relations with the prophet’s people or mankind in general; prophecies associated with miracles supposedly wrought by J′; and prophecies associated with the purported resurrection of J′. The prophecies would need to have a coherence about them, fitting the life, ministry, and vocation of the prophet. Now Tn and E, which virtually assure us that J′ performed miracles and rose again from the dead, comprise part of the evidence for fulfilled prophecy. Let EP represent all the further evidence, such as the prophetic oracles themselves and their apparent match to the evidence we have about the prophet’s life, death, and purported resurrection. Taking all the evidence, and also conditioning on k, we obtain P[ J3′ | T n ∧ E ∧ EP ∧ k] P[~ J3′ | T n ∧ E ∧ EP ∧ k]

=

P[T n ∧ E ∧ EP |J3′ ∧ k]

.

P[ J3′ |k]

P[T n ∧ E ∧ EP | ~ J3′ ∧ ] P[~ J3′ |k]

216  Towards a fuller picture Again the argument is similar to what has gone before. The first ratio on the right hand side will be very much greater than unity. We are very much more likely to get a large concatenation of evidence that J′ fulfilled prophecies if there was really something profound going on and he did indeed fulfil them, than if he did not and instead the apparent fulfilments—shown by the alignments of ancient and contemporary texts—were merely accidental. In a similar way to the preceding arguments, it is also the case that P[J3′|k] will be dramatically greater than P[J3′] since k implies a moderate chance that there is a God and that he would become incarnate, and therefore would be likely to herald his coming with prophetic oracles. We could expand P[J3′|k] as we did P[J1′|k] to bring this out more clearly. Again, too, it might still be the case that P[~J3′|k] exceeds ½, so that the second fraction on the right hand side is less than unity. Nevertheless it seems reasonable to draw the conclusion again that this is swamped by a very large first ratio. Therefore it is very probable that many prophecies were indeed fulfilled, i.e. P[J3′|Tn ∧ E ∧ EP ∧ k] ≈ 1. Finally, let J4′ be the hypothesis that there is no other prophet for whom there is all the evidence there is for J′. As Swinburne argues, if there were a God and he became incarnate there is no reason to suppose that he would do so more than once (though there remains the highly speculative possibility that he might do so in other intelligent life forms elsewhere in the universe).2 Suppose in fact that there is no evidence of testimony whatever to support ~ J4′. Call this evidence (or rather, absence of evidence) E4. Given, then, that there is no a priori reason to expect ~J4′ and no prior or posterior historical evidence (E4 is the null set) to support ~J4′, surely we can confidently conclude that P[J4′| E4 ∧ k] ≈ 1. It looks extremely likely, then, on the basis of the evidence, that J′ performed many miracles, rose from the dead, fulfilled many prophecies, and (we assume) has no rival. Now we need to see how the historical evidence, when united with k, supports the hypothesis c that the prophet was God incarnate. We have P[c | T n ∧ E ∧ EP ∧ E4 ∧ k] P[~ c | T n ∧ E ∧ EP ∧ E4 ∧ k] =

P[T n ∧ E ∧ EP ∧ E4 |c ∧ k]

P[c|k] (11.4) P[T ∧ E ∧ EP ∧ E4 | ~ c ∧ k] P[~ c|k] n

.

Here, in the first ratio on the right hand side, we have to compare how likely the evidence is if God becomes incarnate (the numerator) with how likely this is if either there is a God who does not become incarnate or there is no God (the denominator). It may be that we are building too much into where the evidence might lead. It is hard to imagine that we would have

Towards a fuller picture 217 this evidence, which virtually assures us that the prophet did remarkable things and that his life was crowned by resurrection from the dead, if there were no God at all. But it might just be that the data are compatible with a lesser figure—perhaps somebody like a Jewish Messiah who is a remarkable man but not God, rather than one who was both fully God and fully man. It seems to me that some of the evidence could tip the balance decisively in favour of incarnation. This would include the prophet’s fulfilling prophecies about the coming of God himself and of being heralded by one who himself was prophesied to prepare the way for God to come. It would include him substituting himself for the place where God is to be found and worshipped. This might well be in a veiled way. After all, going around loudly proclaiming oneself to be God incarnate might look odd, even if it were true. Some of the miracles wrought by the prophet might be such as to prompt questions about his identity, especially if they mimic actions already perceived to be of divine origin. Supremely, however, it would be the resurrection that would most likely mark out God incarnate from a non-divine prophet. Thus it is likely that the first factor on the right hand side of equation (11.4) would outweigh the moderately low value of the second (⅓ on Swinburne’s reckoning). We can thus safely conclude that P[c|Tn ∧ E ∧ EP ∧ E4 ∧ k] ≲ 1

where the symbol ‘≲’ means ‘less than or approximately equal to.’ At this point I follow the logic of Swinburne’s argument, paraphrasing him somewhat, I trust for the sake of clarity.3 The fact is, then, that we have all the evidence surveyed here. We have it for Jesus and for no one else in history. Each piece of evidence is stronger than described because it pertains to a prophet we can name, i.e. to Jesus, and to him alone. Adding this information about the prophet’s identity will not make any difference to the probability of c. Furthermore, given the evidence, both the historical evidence, now seen as applying to Jesus, and k, and given c, it would be utterly improbable that the incarnation would occur in any other prophet than Jesus or would culminate in a ‘super-miracle’ other than the resurrection. The alternative would be to imagine God intending to become incarnate, and thus to live a life, die a death, and rise again in a manner similar to the way the evidence points to Jesus doing, but doing so in another prophet. That would be for God to undertake a massive deception. The upshot of the argument is that it is highly probable that God became incarnate in the person of Jesus Christ, who lived the kind of life to which the evidence points: who performed miracles, fulfilled prophecy, and was indeed raised from the dead.

218  Towards a fuller picture

Notes 1 C. S. Lewis, ‘Is Theology Poetry?’ in Screwtape Proposes a Toast and other Pieces (London: Collins, Fontana Books, 1965, this chapter first read as a paper to the Socratic Club in Oxford in 1944), 50. 2 Interestingly, even so eminent a Biblical scholar as Charlie Moule is drawn to this kind of speculation when discussing the ‘ultimacy’ of Christ: C. F. D. Moule, The Origin of Christology (Cambridge: Cambridge University Press, 1977), 143 ff. 3 Richard Swinburne, The Resurrection of God Incarnate (Oxford: Oxford University Press, 2003), 214.

Bibliography

Alston, William, Perceiving God: The Epistemology of Religious Experience (Ithaca, NY: Cornell University Press, 1991). Aquinas, St Thomas, Summa Theologiae, Blackfriars edition (London: Eyre & Spottiswoode, and New York: McGraw Hill, 1963). Armstrong, Benjamin F., ‘Hume on Miracles: Begging-the-Question Against Believers,’ History of Philosophy Quarterly 9 (1992), 319–328. Augustine, St, The Literal Meaning of Genesis (De Genesi ad Litteram), Ancient Christian Writers, vol. 1, translated and annotated by John Hammond Taylor SJ (New York and Mahwah, NJ: Paulist Press, 1982). Augustine, St, ‘The City of God (De Civitate Dei),’ in Philip Schaff (ed.), Nicene and Post-Nicene Fathers, First Series, vol. 2 (Peabody, MA: Hendrickson, 1994), 1–511. Barr, James, Biblical Faith and Natural Theology: The Gifford Lectures for 1991 Delivered in the University of Edinburgh (Oxford: Oxford University Press, 1993). Barrett, C. K., The New Testament Background: Selected Documents, revised edition (London: SPCK, 1997). Barth, Karl, ‘No!,’ Response to Emil Brunner, ‘Nature and Grace,’ in Emil Brunner and Karl Barth, Natural Theology, trans. Peter Fraenkel, with an introduction by John Baillie (London: Geoffrey Bles/Centenary Press, 1946), 65–128. Barth, Karl, Protestant Theology in the Nineteenth Century, new edition, trans. John Bowden (London: SCM, 2001). Barton, John, and Muddiman, John (eds.), The Oxford Bible Commentary (Oxford: Oxford University Press, 2001). Bauckham, Richard, Jesus and the Eyewitnesses: The Gospels as Eyewitness Testimony (Grand Rapids, MI: Wm. B. Eerdmans, 2006). Berry, R. J., ‘What to Believe about Miracles,’ Nature 322 (1986), 321–322. Bonhoeffer, Dietrich, Letters and Papers from Prison, enlarged edition, ed. Eberhard Bethge (London: SCM Press, 1971). Borde, Arvind, Guth, Alan H., and Vilenkin, Alexander, ‘Inflationary Spacetimes Are Incomplete in Past Directions,’ Physical Review Letters 90(15) (2003), 151301-1–151301-4. Borde, Arvind, and Vilenkin, Alexander, ‘Eternal Inflation and the Initial Singularity,’ Physical Review Letters 72(21) (1994), 3305–3308. Borg, Marcus, and Wright, N. T., The Meaning of Jesus: Two Visions (New York: HarperCollins, [1999] 2007).

220  Bibliography Braaten, Carl E., and Clayton, Philip, The Theology of Wolfhart Pannenberg: Twelve American Critiques, with an Autobiographical Essay and Response (Minneapolis: Augsburg, 1988). Brown, Raymond E., S. S., Fitzmyer, Joseph A., S. J., and Murphy, Roland E., O. Carm (eds.), The New Jerome Biblical Commentary (London and New York: Geoffrey Chapman, 1989). Bruce, F. F., Jesus and Christian Origins outside the New Testament (London: Hodder and Stoughton, 1974). Bruce, F. F., The New Testament Documents: Are They Reliable? sixth revised edition (Downers Grove, IL: IVP Academic, [1981] 2000). Bussey, Peter J., ‘God as First Cause—a Review of the Kalam Argument,’ Science and Christian Belief 25(1) (2013), 17–35. Butler, Joseph, The Analogy of Religion Natural and Revealed to the Constitution and Course of Nature, ed. W. E. Gladstone (Oxford: Oxford University Press, 1895); and idem, ed. J. H. Bernard (London: Macmillan, 1900). Calvin, John, Institutes of the Christian Religion, trans. Henry Beveridge (Grand Rapids, MI: Wm. B. Eerdmans, [1559] 1989). Carr, Bernard (ed.), Universe or Multiverse? (Cambridge: Cambridge University Press, 2007). Childs, Brevard, ‘Prophecy and Fulfilment: A  Study of Contemporary Hermeneutics,’ Interpretation: The Journal of Bible and Theology 12(3) (1958), 259–271. Clement of Rome, ‘The First Epistle of Clement to the Corinthians,’ ch. XLVII, in Alexander Roberts and James Donaldson (eds.), Ante-Nicene Fathers, vol. 1, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson, 1994), 5–21. Coady, James, Testimony (Oxford: Oxford University Press, 1992). Collins, Robin, ‘Evidence for Fine-Tuning,’ in Neil Manson (ed.), God and Design: The Teleological Argument and Modern Science (London: Routledge, 2003), 178–199. Collins, Robin, ‘The Multiverse: A Theistic Perspective,’ in Carr (ed.), Universe or Multiverse? 459–480. Collins, Robin, ‘Modern Cosmology and Anthropic Fine-tuning: Three Approaches,’ in Holder and Mitton (eds.), Georges Lemaître, 173–191. Copan, Paul, and Craig, William Lane, Creation out of Nothing: A Biblical, Philosophical, and Scientific Exploration (Grand Rapids, MI: Baker Academic, and Leicester: Apollos, 2004). Craig, William Lane, The Kalām Cosmological Argument (London: Macmillan Press, 1979). Crossan, John Dominic, The Historical Jesus: The Life of a Mediterranean Jewish Peasant (New York: HarperCollins, 1991). Cruickshank, John, Pascal: Pensées (London: Grant & Cutler Ltd, 1983). Davies, Paul, The Goldilocks Enigma: Why is the Universe Just Right for Life? (London: Allen Lane, 2006). Davies, Paul, ‘Universes Galore: Where will it all End?’ in Carr (ed.), Universe or Multiverse? 487–505. Davies, W. D., Paul and Rabbinic Judaism (London: SPCK, [1948] 1955). Dawid, Philip, and Gillies, Donal, ‘A Bayesian Analysis of Hume’s Argument Concerning Miracles,’ Philosophical Quarterly 39 (1989), 57–65. Didache, in Alexander Roberts and James Donaldson (eds.), Ante-Nicene Fathers, vol. 7, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson, 1994), 369–382.

Bibliography 221 Donnelly, Peter, ‘Appealing Statistics,’ Significance 2(1) (2005), 46–48. Dunn, James D. G., Jesus Remembered: Volume 1 of Christianity in the Making (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2003). Earman, John, Bayes or Bust? A  Critical Examination of Bayesian Confirmation Theory (Cambridge, MA: MIT Press, 1992). Earman, John, ‘Bayes, Hume, and Miracles,’ Faith and Philosophy 10 (1993), 293–310. Earman, John, Hume’s Abject Failure: The Argument Against Miracles (New York and Oxford: Oxford University Press, 2000). Ellis, Ieuan, Seven against Christ: A  Study of ‘Essays and Reviews’ (Leiden: Brill, 1980). Elster, Jon, ‘Pascal and Decision Theory,’ in Hammond (ed.), Cambridge Companion to Pascal, 53–74. Eusebius, The Church History of Eusebius, in Philip Schaff and Henry Wace (eds.), Nicene and Post-Nicene Fathers, second series, vol. 1 (Peabody, MA: Hendrickson, 1994), 81–387. Fee, Gordon D., The First Epistle to the Corinthians (Grand Rapids, MI: Wm. B. Eerdmans, 1987). France, Richard T., ‘Historical Jesus, Quest of,’ in Alister E. McGrath (ed.), The Blackwell Encyclopedia of Modern Christian Thought (Oxford: Blackwell, 1993), 260–266. France, Richard T., The Gospel of Matthew: The New International Commentary on the New Testament (Grand Rapids, MI: Wm. B. Eerdmans, 2007). Fuller, Daniel P., ‘The Resurrection of Jesus and the Historical Method,’ Journal of Bible and Religion 34(1) (1966), 18–24. Gauch, Hugh, ‘The Methodology of Ramified Natural Theology,’ Philosophia Christi 13(2) (2013), 283–298. Glass, David H., ‘Can Evidence for Design be Explained Away?’ in Jake Chandler and Victoria S. Harrison (eds.), Probability in the Philosophy of Religion (Oxford: Oxford University Press, 2012), 79–102. Glass, David H., and McCartney, Mark, ‘Explaining and Explaining Away in Science and Religion,’ Theology and Science 12(4) (2014), 338–361. Goldingay, John E., Daniel (Dallas, TX: Word Books, 1989). Goodman, Nelson, Fact, Fiction, and Forecast (New York: Bobs-Merrill, 1946). Grossman, Lisa, ‘Death of the Eternal Cosmos,’ New Scientist 213(2847) (January 2012), 6–7. Guth, Alan H., ‘Inflationary Universe: A Possible Solution to the Horizon and Flatness Problems,’ Physical Review D 23(2) (1981), 347–356. Guth, Alan H., The Inflationary Universe: The Quest for a New Theory of Cosmic Origins (London: Jonathan Cape, 1997). Hacking, Ian, An Introduction to Probability and Inductive Logic (Cambridge: Cambridge University Press, 2001). Hammond, Nicholas (ed.), The Cambridge Companion to Pascal (Cambridge: Cambridge University Press, 2003). Harvey, Anthony E., Jesus and the Constraints of History: the Bampton Lectures, 1980 (London: Duckworth, 1982). Hawking, S. W., ‘The Quantum State of the Universe,’ Nuclear Physics B239 (1984), 257–276.

222  Bibliography Hawking, S. W., ‘Quantum Cosmology,’ in S. W. Hawking and W. Israel (eds.), Three Hundred Years of Gravitation (Cambridge: Cambridge University Press, 1987), 631–651. Hawking, Stephen, A Brief History of Time (London: Bantam, 1988). Hawking, Stephen, and Mlodinow, Leonard, The Grand Design: New Answers to the Ultimate Questions of Life (London: Bantam, 2010). Hebblethwaite, Brian, The Problems of Theology (Cambridge: Cambridge University Press, 1980). Hesketh, Ian, ‘Behold the (Anonymous) Man: J. R. Seeley and the Publishing of Ecce Homo,’ Victorian Review 38(1) (Spring 2012), 93–112. Holder, Rodney D., ‘Hume on Miracles: Bayesian Interpretation, Multiple Testimony, and the Existence of God,’ British Journal for the Philosophy Science 49 (1998), 49–65. Holder, Rodney D., God, the Multiverse, and Everything: Modern Cosmology and the Argument from Design (Aldershot and Burlington, VT: Ashgate, 2004). Holder, Rodney, The Heavens Declare: Natural Theology and the Legacy of Karl Barth (West Conshohocken: Templeton Press, 2012). Holder, Rodney D., Big Bang, Big God: A  Universe Designed for Life? (Oxford: Lion Hudson, 2013). Holder, Rodney D., ‘Natural Theology in the Twentieth Century,’ in Russell Re Manning (ed.), The Oxford Handbook of Natural Theology (Oxford: Oxford University Press, 2013), 118–134. Holder, Rodney D., ‘Why We Need Ramified Natural Theology,’ Philosophia Christi 13(2) (2013), 271–282. Holder, Rodney D., ‘Explaining and Explaining Away in Cosmology and Theology,’ Theology and Science 14(3) (2016), 234–255. Holder, Rodney D., and Mitton, Simon (eds.) Georges Lemaître: Life, Science and Legacy (Heidelberg: Springer, 2012). Hooker, Morna D., The Gospel According to St Mark (London: A & C Black, 1991). Howson, Colin, and Urbach, Peter, Scientific Reasoning: The Bayesian Approach, second edition (Chicago and La Salle, IL: Open Court, 1993). Hoyle, Fred, The Nature of the Universe (Oxford: Blackwell, 1950). Hoyle, Fred, ‘The Universe: Some Past and Present Reflections,’ Engineering & Science 45(2) (November 1981), 8–12. Hume, David, An Enquiry Concerning Human Understanding, Section X, ‘Of Miracles,’ in the first part of L. A. Selby-Bigge (ed.), Enquiries Concerning Human Understanding and Concerning the Principles of Morals by David Hume, third edition, revised by P. H. Nidditch (Oxford: Oxford University Press, [1748] 1975). Hume, David, Dialogues Concerning Natural Religion, ed. J. M. Bell (London: Penguin, [1779] 1990). Irenaeus, Adversus Haereses, I.XXIV.4, in Alexander Roberts and James Donaldson (eds.), Ante-Nicene Fathers, vol. 1, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson). Jaki, S. L., Lord Gifford and his Lectures: A Centenary Retrospect (Edinburgh: Scottish Academic Press; Macon, GA: Mercer University Press, 1986). Josephus, ‘The Antiquities of the Jews and The Wars of the Jews,’ in William Whiston (trans.), The Works of Josephus (Peabody, MA: Hendrickson, 1987), 27–772. Justin Martyr, First Apology, LXVII, in Alexander Roberts and James Donaldson, in Ante-Nicene Fathers, vol. 1, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson, 1994), 163–187.

Bibliography 223 Kant, Immanuel, Critique of Pure Reason, trans. J. M. D. Meiklejohn (London: J. M. Dent, [1781] 1934). Kateregga, Badru D., and Shenk, David W., Islam and Christianity: A Muslim and a Christian in Dialogue (Grand Rapids, MI: Wm. B. Eerdmans, revised edition, 1981). Kingsley, Charles, ‘The Natural Theology of the Future,’ in Westminster Sermons (London: McMillan, 1874), v–xxiii. Krauss, Lawrence M., A Universe from Nothing: Why There is Something Rather than Nothing (New York: Simon and Shuster, 2012). Lacocque, André, The Book of Daniel (London: SPCK, 1979). Leibniz, Gottfried Wilhelm, ‘On the Ultimate Origination of Things,’ in Robert Latta (trans.), The Monadology and other Philosophical Writings (Oxford: Oxford University Press, [1697] 1971), 337–351. Lewis, C. S., ‘Is Theology Poetry?’ in Screwtape Proposes a Toast and other Pieces (London: Collins, Fontana Books, 1965). Lewis, C. S., Miracles: A  Preliminary Study (Glasgow: Collins, Fount Paperbacks edition, [1947] 1977). Lewis, C. S., Reflections on the Psalms (Glasgow: Collins, Fount Paperbacks edition, [1961] 1977). Linde, A. D., ‘Particle Physics and Inflationary Cosmology,’ Physics Today 40(9) (1987), 61–68. Locke, John, An Essay Concerning Human Understanding, edited with an introduction by P. H. Nidditch (Oxford: Oxford University Press, [1689] 1979). Mackie, J. L., The Miracle of Theism: Arguments for and against the Existence of God (Oxford: Oxford University Press, 1982). Mawson, T. J., ‘Explaining the Fine Tuning of the Universe to Us and the Fine Tuning of Us to the Universe,’ Royal Institute of Philosophy Supplement 68 (2011), 25–50. McGrath, Alister E., The Making of Modern German Christology, second edition (Leicester: Inter-Varsity Press, 1994). McGrath, Alister E., T. F. Torrance: An Intellectual Biography (Edinburgh: T & T Clark, 1999). McGrath, Alister E., A Scientific Theology Volume 1: Nature (Edinburgh and New York: T & T Clark, 2001). McGrath, Alister E., A Scientific Theology Volume 2: Reality (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2002). McGrath, Alister E., A Scientific Theology Volume 3: Theory (Grand Rapids, MI: Wm. B. Eerdmans, 2003). McGrath, Alister E., Dawkins’ God: Genes, Memes, and the Meaning of Life (Oxford: Blackwell, 2005). McGrath, Alister E., The Order of Things (Oxford: Blackwell, 2006). McGrath, Alister E., The Open Secret: A New Vision for Natural Theology (Oxford: Blackwell, 2008). McGrath, Alister E., A Fine-Tuned Universe: The Quest for God in Science and Theology, The 2009 Gifford Lectures (Louisville, KY: Westminster John Knox Press, 2009). McGrath, Alister E., with McGrath, Joanna Collicutt, The Dawkins Delusion: Atheist Fundamentalism and the Denial of the Divine (London: SPCK, 2007). McGrew, Timothy, and McGrew, Lydia, ‘The Argument from Miracles: A Cumulative Case for the Resurrection of Jesus of Nazareth,’ in William Lane Craig and

224  Bibliography J. P. Moreland (eds.), The Blackwell Companion to Natural Theology (Chichester: Wiley-Blackwell, [2009] 2012), 593–662. Meier, John P., ‘Criteria: How Do We Decide What Comes from Jesus?’ in James D. G. Dunn and Scot McKnight (eds.), The Historical Jesus in Recent Research (Winona Lake, IN: Eisenbrauns, 2005), 126–136. Millican, Peter, ‘ “Hume’s Theorem” Concerning Miracles,’ Philosophical Quarterly 43 (1993), 489–495. Mitchell, Basil, ‘Butler as a Christian Apologist,’ in Christopher Cunliffe (ed.), Joseph Butler’s Moral and Religious Thought: Tercentenary Essays (Oxford: Oxford University Press, 1992), 97–116. Moule, C. F. D., The Origin of Christology (Cambridge: Cambridge University Press, 1977). Neill, Stephen, and Wright, Tom, The Interpretation of the New Testament 1861– 1986, second edition (Oxford: Oxford University Press, 1988). Newman, John Henry, The Idea of a University Defined and Illustrated: I. In Nine Discourses Delivered to the Catholics of Dublin; II. In Occasional Lectures and Essays Addressed to the Members of the Catholic University (Oxford: Oxford University Press, [1852, 1889] 1976). Otte, Richard, ‘Schlesinger and Miracles,’ Faith and Philosophy 10 (1993), 93–98. Owen, David, ‘Hume Versus Price on Miracles and Prior Probabilities,’ in Richard Swinburne (ed.), Miracles (London: Macmillan, 1989); reprinted from David Owen, The Philosophical Quarterly 37 (1987), 187–202. Page, Don N., ‘Predictions and Tests of Multiverse Theories,’ in Carr (ed.), Universe or Multiverse? 411–430. Page, Don N., ‘Multiple Reasons for a Multiverse,’ in Holder and Mitton (eds.), Georges Lemaître, 113–123. Paley, William, Natural Theology, or Evidence of the Existence and Attributes of the Deity, Collected from the Appearances of Nature (Oxford: Oxford University Press, [1802] 2006). Pals, Daniel, ‘The Reception of “Ecce Homo”,’ Historical Magazine of the Episcopal Church 46(1) (1977), 63–84. Pannenberg, Wolfhart, ‘The Revelation of God in Jesus of Nazareth,’ in J. M. Robinson and John B. Cobb, Jr. (eds.), Theology as History (New York: Harper & Row, 1967), 101–133. Pannenberg, Wolfhart, Jesus—God and Man (London: SCM Press, [1968] 2002). Originally published as Grundzüge der Christologie (Gütersloher Verlagshaus Gerd Mohn: Gütersloh, 1964). Pannenberg, Wolfhart, Basic Questions in Theology, vol. 1, trans. George H. Kehm (London: SCM Press, 1970). Originally published as Grundfragen Systematischer Theologie: Gesammelte Aufsätze (Göttingen: Vandenhoeck and Ruprecht, 1967). Pannenberg, Wolfhart, Theology and the Philosophy of Science, trans. Francis McDonagh (London: Darton, Longman & Todd, 1976). (First German edition, Wissenschaftstheorie und Theologie, Frankfurt am Main: Suhrkamp, 1973). Pannenberg, Wolfhart, ‘A Response to My American Friends,’ in Braaten and Clayton (eds.), The Theology of Wolfhart Pannenberg. Pannenberg, Wolfhart, Systematic Theology, vol. 1, trans. Geoffrey W. Bromiley (Edinburgh: T & T Clark), 1991. Pannenberg, Wolfhart, Systematic Theology, vol. 2, trans. Geoffrey W. Bromiley (Edinburgh: T & T Clark, 1994).

Bibliography 225 Pannenberg, Wolfhart, ed., with Rolf Rendtorff, Trotz Rendtorff, and Ulrich Wilckens, Revelation as History, trans. D. Granskou (New York: Macmillan, 1968). (First German edition, Offenbarung als Geschichte, Göttingen: Vandenhoeck and Ruprecht, 1961). Parrinder, Geoffrey, Jesus in the Qur’ān (London: Sheldon Press, 1965). Pascal, Blaise, Pensées, trans. A. J. Krailsheimer, revised edition (London: Penguin, 1995). Peacocke, A. R., Theology for a Scientific Age, second edition (London: SCM Press (1993). Penelhum, Terence, Butler: The Arguments of the Philosophers (London and New York: Routledge, 1985). Penrose, Roger, The Emperor’s New Mind: Concerning Computers, Minds, and the Laws of Physics (Oxford: Oxford University Press, 1989). Philo, Flaccus, in The Works of Philo, trans. C. D. Yonge (Peabody, MA: Hendrickson, 1993), 725–741. Plantinga, Alvin, Where the Conflict Really Lies: Science, Religion, and Naturalism (New York: Oxford University Press, 2011). Plato, Republic (London: J. M. Dent and Sons, 1935). Polanyi, Michael, Personal Knowledge: Towards a Post-Critical Philosophy (Chicago: University of Chicago Press, 1974). Polkinghorne, John, Science and Christian Belief: Theological Reflections of a Bottom-up Thinker (London: SPCK, 1994). Polkinghorne, John, Belief in God in an Age of Science (New Haven and London: Yale University Press, 1998). Polkinghorne, John, Faith, Science and Understanding (London: SPCK, 2000). Polkinghorne, John, ‘Where is Natural Theology Today?’ Science and Christian Belief 18(2) (2006), 169–179. Polkinghorne, John, Theology in the Context of Science (London: SPCK, 2008). Rees, Martin, New Perspectives in Astrophysical Cosmology, second edition (Cambridge: Cambridge University Press, 2000). Rees, Martin, Our Cosmic Habitat (London: Weidenfeld & Nicolson, 2001). Robinson, John A. T., Redating the New Testament (London: SCM Press, 1976). Robinson, John A. T., The Priority of John, ed. J. F. Coakley (London: SCM Press, 1985). Rowland, Christopher, Christian Origins (London: SPCK, 1985). Sarwar, Ghulam, Islam: Beliefs and Teachings, third edition (London: The Muslim Educational Trust, 1984). Schellenberg, J. L., ‘Christianity Saved? Comments on Swinburne’s Apologetic Strategies in the Tetralogy,’ Religious Studies 38 (2002), 283–300. Schlesinger, George N., ‘Miracles and Probabilities,’ Noûs 21 (1987), 219–232. Schweitzer, Albert, The Quest of the Historical Jesus: A Critical Study of its Progress from Reimarus to Wrede, trans. W. Montgomery, preface by F. C. Burkitt (London: Adam and Charles Black, 1910). Shellard, E. P. S., ‘The Future of Cosmology: Observational and Computational Prospects,’ in G. W. Gibbons, E. P. S. Shellard, and S. J. Rankin (eds.), The Future of Theoretical Physics and Cosmology: Celebrating Stephen Hawking’s 60th Birthday (Cambridge: Cambridge University Press, 2003), 755–780. Slupik, Chris, ‘A New Interpretation of Hume’s “Of Miracles”,’ Religious Studies 31 (1995), 517–536.

226  Bibliography Smith, Quentin, ‘The Wave Function of a Godless Universe,’ in William Lane Craig and Quentin Smith (eds.), Theism, Atheism, and Big Bang Cosmology (Oxford: Oxford University Press, 1993), 301–337. Sobel, Jordan Howard, ‘On the Evidence of Testimony for Miracles: A  Bayesian Interpretation of David Hume’s Analysis,’ Philosophical Quarterly 37 (1987), 167–186. Sobel, Jordan Howard, ‘Hume’s Theorem on Testimony Sufficient to Establish a Miracle,’ Philosophical Quarterly 41 (1991), 229–237. Sorensen, Roy A., ‘Hume’s Scepticism about Miracles,’ Analysis 43 (1983), 60. Soskice, Janet, ‘Creatio ex nihilo: Its Jewish and Christian Foundations,’ in David B. Burrell, Carlo Cogliati, Janet M. Soskice and William R. Stoeger (eds.), Creation and the God of Abraham (Cambridge: Cambridge University Press, 2010), 24–39. Spitzer, Robert J., New Proofs of the Existence of God: Contributions of Contemporary Physics and Philosophy (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2010). Stanton, Graham N., The Gospels and Jesus (Oxford: Oxford University Press, 1989). Strauss, David Friedrich, The Life of Jesus Critically Examined (Philadelphia: Fortress, [1835–1836] 1972). Susskind, Leonard, The Cosmic Landscape: String Theory and the Illusion of Intelligent Design (New York: Little Brown and Company, 2006). Swinburne, Richard, Epistemic Justification (Oxford: Oxford University Press, 2001). Swinburne, Richard, ‘Response to My Commentators,’ Religious Studies 38 (2002), 301–315. Swinburne, Richard, The Resurrection of God Incarnate (Oxford: Oxford University Press, 2003). Swinburne, Richard, The Existence of God, second edition (Oxford: Oxford University Press, 2004). Swinburne, Richard, ‘Natural Theology, Its “Dwindling Probabilities” and “Lack of Rapport”,’ Faith and Philosophy 21(4) (2004), 533–546. Swinburne, Richard, Was Jesus God? (Oxford: Oxford University Press, 2008). Swinburne, Richard, ‘The Probability of the Resurrection of Jesus,’ Philosophia Christi 15(2) (2013), 239–252. Tegmark, Max, ‘Parallel Universes,’ Scientific American 288(5) (May 2003), 30–41. Temple, W. F., The Relations Between Religion and Science, The Bampton Lectures for 1884 (London: Macmillan, 1885). Theissen, Gerd, The Shadow of the Galilean (London: SCM Press, [1986] 1987). Theissen, Gerd, and Merz, Annette, The Historical Jesus (London: SCM Press, 1997). Torrance, Thomas F., Space, Time and Incarnation (Edinburgh: T & T Clark, 1969; paperback edition, 1997). Torrance, Thomas F., The Ground and Grammar of Theology (Belfast: Christian Journals, 1980). Troeltsch, Ernst, ‘Über historische und dogmatische Methode in der Theologie,’ in Gerhard Sauter (ed.), Theologie als Wissenschaft (München: Chr. Kaiser Verlag, 1971), 105–127. Originally published in Studien des rheinischen Predigervereins, 1898.

Bibliography 227 Tyrell, George, Christianity at the Crossroads (London: Longmans, Green and Co., 1910). Van Huyssteen, Wentzel, ‘Theology as the Science of God: Wolfhart Pannenberg,’ in his Theology and the Justification of Faith: Constructing Theories in Systematic Theology (Grand Rapids: Wm. B. Eerdmans, 1989), 71–100. Vilenkin, Alexander, Many Worlds in One: The Search for Other Universes (New York: Hill and Wang, 2006). Walker, Peter W. L., Jesus and the Holy City: New Testament Perspectives on Jerusalem (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 1996). Weinberg, Steven, ‘Living in the Multiverse,’ in Carr (ed.), Universe or Multiverse? 29–42. Westermann, Claus, Isaiah 40–66: A Commentary (London: SCM Press, 1969). Wetsel, David, ‘Pascal and Holy Writ,’ in Hammond (ed.), Cambridge Companion to Pascal, 162–181. Wilkinson, David, Science, Religion, and the Search for Extraterrestrial Intelligence (Oxford: Oxford University Press, 2013). Witham, Larry, The Measure of God: Our Century-Long Struggle to Reconcile Science and Religion, The Story of the Gifford Lectures (San Francisco: HarperSanFrancisco, 2005). Wright, N. T., The New Testament and the People of God: Volume 1 of Christian Origins and the Question of God (London: SPCK, 1992) Wright, N. T., Jesus and the Victory of God: Volume 2 of Christian Origins and the Question of God (London: SPCK, 1996). Wright, N. T., The Resurrection of the Son of God: Volume 3 of Christian Origins and the Question of God (London: SPCK, 2003).

Index

Note: Page numbers in bold indicate a table on the corresponding page. 1 Corinthians 15 181 Alston, William 19, 65, 78 alternative hypothesis 37, 41 – 42, 55 Analogy of Religion (Butler) 82, 93 analogy, principle of 75 – 77, 129, 137 Antiochene crisis 92 apostles 82, 87, 99, 178, 180–181 apparent revelations 78 a priori reasoning 76 Aquinas, St Thomas 10, 12, 38, 44 – 45, 65, 108 – 109 Artaxerxes’ order, dating of 91 atheism 2, 30, 66, 68, 75, 85 Augustine, St 10, 16, 18, 83, 108 – 109 Bacon, Francis 10 – 11 Barth, Karl 18, 31, 130 Basilides 73 Bauckham, Richard 76, 138, 171, 174, 191 Bayes factor 7, 26, 190, 192 – 193, 204 – 207 Bayesian confirmation theory 4, 7, 25, 59 Bayesian evaluation of multiverse hypothesis 60 – 62 Bayesian formulation, of Christian claims 210; background knowledge 210 – 211; fulfilment of prophecy 211, 214 – 217; healing miracles, testimonies to 212 – 214; ‘super-miracle’ (Swinburne) 202 – 203, 211, 214, 217 Bayesian formulation, of resurrection arguments 189; McGrews’ approach 189 – 195, 206 – 207; McGrews’ independence assumption 205 – 206; Richard Swinburne’s approach to

biblical data 195 – 198, 207; Richard Swinburne’s Bayesian approach 198 – 205, 207 – 208 Bayesian framework for miracle arguments 110 – 113, 145; aggregation of testimonies 114 – 118; impact on Hume’s arguments 118 – 119; independent testimonies 113 – 114 Bayes’s theorem 25 – 27, 59, 110 belief in God 19, 68, 102, 121 Berry, R. J. 107 best explanation, language of 21 – 24, 28, 58, 66, 143 – 144, 170 Bhaskar, Roy 20, 21 Bible 6, 31, 82, 86, 98, 102, 129, 133 Big Bang 5, 29, 36, 38, 46, 48 – 49, 67 Bohmian theory 67 Bonhoeffer, Dietrich 67 – 68 boundary condition of universe 39 Brief History of Time, A (Hawking) 39 Bussey, Peter 40 Butler, Joseph 5 – 6, 93 – 104; analogy definition 98; analogy with Pascal’s wager 93 – 94, 95; appeal to ‘common practice’ 101; arguments in terms of probability 94 – 95, 104; assumption of deist position 98; as Chaplain 93; cumulative case 94, 101; fulfilment of prophecy 103; on miracles and prophecies 96 – 98, 101 – 102; on moral governance of universe 95 – 96; on notion of ‘incredibility’ 100; observations by analogy of nature 101; positive evidence of Christianity 98 – 99; as Preacher at Rolls Chapel 93; on religious pluralism 99 – 100; revelation 102 – 103

Index  229 Calvin, John 10, 65, 68, 75 causal principles 16 chaotic inflation 49 childhood theism 75 Christ, Jesus: humanity and sufferings 73; male disciples of 182; mighty works 139 – 144; place in world history 69; see also Jesus, death on cross; miracles; resurrection of Jesus Christ Christian belief 1, 11, 18, 74, 78, 95 Christian claims 8, 90; arguments for 5; evidence for 78 – 79, 145; truth of 5; see also Bayesian formulation, of Christian claims; resurrection of Jesus Christ Christian doctrine of creation 1, 16, 41, 44 – 45 Christian faith 11, 21, 22, 31, 66, 74, 87, 181 Christian God 1, 30, 59, 121 Christian interpolation in Josephus 141 – 142, 144, 172 Christianity 5, 70, 82, 87, 93 – 94, 132 – 133, 173; claims, historical evidence of 68; death of Jesus 70 – 74; presuppositions 69, 74 – 76; tradition transcendent rationality 21 Christian natural theology (McGrath) 20 Christian revelation 10 Christian theology 2, 10, 23, 69, 71, 78, 136; see also natural theology Christian Trinitarian orthodoxy 21 ‘Christ of faith’ 130 Christology 133 – 134 classical universe 40 Collins, Robin 51, 56 – 57 ‘common sense’ philosophy 171 conditional probability 25 ‘consonance’ concept 23 convergent evolution, phenomenon of 9 correlation, principle of 75 cosmic background radiation 30, 67 cosmological argument 11 – 12, 35, 37 – 38, 42,44, 50, 59 – 60, 101, 190 cosmology 40, 45, 53, 186; see also modern cosmology critical realism 20 – 21, 24, 129 criticism, principle of 75 Crossan, Dominic 136 crucifixion see Jesus, death on cross Cruickshank, John 84

Daniel, book of 91 – 92, 103, 159, 163, 167 Darwin’s theory of evolution by natural selection 3, 15 – 16, 132, 133 Davies, Paul 46 – 47 Dawood, N. J. 73 deism 23, 85 deists 6, 82, 95, 97 – 98 design argument 12, 17, 24, 36, 42, 47, 202, 210 ‘Deutero-Isaiah’ 151, 153, 155, 161 divine creation of universe: Christian doctrine of creation 44 – 45; and no boundary proposal 41 – 42; and quantum vacuum creation 43 – 44; see also multiverse hypothesis divine design 36; fine-tuning in absence of 55 – 56; multiverse compatibility with 55; multiverse hypothesis dependence on 58; TOE and 47 divine reality 69, 75 Dunn, James D. G. 71, 129, 142, 172 – 173, 185 Earman, John 109 ECAs see embodied conscious agents Ecce Homo (J. R. Seeley) 132 Eliot, George 130 Elster, Jon 86 embodied conscious agents 56 – 57 empirical law of causality 12 Enlightenment 6, 129, 131, 137 Enquiry Concerning Human Understanding (Hume) 110 epistemic probability 26, 94 – 95 epistles of St Paul 99 Essays and Reviews 132 eternal inflation 49 Euclidean 4-space 39 – 40 evidence 2, 37; categories of 23; in Glass and McCartney’s formulation 37; independent 37; for miracles 6; scientific theories 25 existence of God 65, 111; arguments for 6; metaphysical proofs for (Pascal) 84; miracle testimony as evidence for 120 – 125; natural knowledge and 11; probabilities relevant to 27; ruled out by fiat 74 – 75; traditional arguments for 24 – 25, 31, 36; see cosmological argument; design argument expanding universes, space-times of 40

230 Index ‘Explaining and Explaining Away in Science and Religion’ (Glass and McCartney) 36 explanations: for fine-tuning 47 – 48; provided by God 29 – 32 explanatory power 21, 25, 102, 185 eyewitness testimony see testimony faith 11, 31 Feuerbach, Ludwig 75, 131 four-dimensional space 39 Frazer, J. G. 18 fulfilment of prophecy, texts on 150 – 164; Isaiah 11:1 – 5 158; Isaiah 42 155; Isaiah 42:5 – 9 156; Isaiah 49:1 – 7 155; Isaiah 50:6 157; Isaiah 52:13 – 53:12 151 – 152; Isaiah 53 153, 154; Isaiah 53:9 154; Jesus’ crucifixion 153 – 154; Luke 4:16 – 19 156; Malachi 3:1 161; Malachi 4:5 161; Mark 1:2 – 3 161; Mark 1:6 161 – 162; Mark 14:60 – 61a 154; Mark 15:16 – 20 156 – 157; Mark 15:29 – 32 157; Matthew 5:17 160; Matthew 8:14 – 16 153; Matthew 11:29 – 30 160; Matthew 18:20 160 – 161; Paul’s epistles 153; Psalm 2:1 – 9 158 – 159; Psalm 22:18 158; Psalm 110:1 159 – 160; ‘Suffering Servant’ passages in Isaiah 151 Gabriel (angel) 91 Gifford Lectures in natural theology 18 Glass, David 4, 36, 41 – 42, 43 – 44, 55 – 58 God incarnate, posterior probability for truth of 7, 203 – 204, 207, 216 – 217 God of Christian theology 23; creation and design by 58; explanations provided by 29 – 32; hypothesis 25, 27, 62; as ‘necessary being’ 45; universe’s dependence on 44 – 45 gospel accounts, historical-critical study of 128 – 137; Baur 131 – 132; Feuerbach 131; form criticism 134 – 135; Harnack 133 – 134; Heidegger 135; Hellenistic culture and 133; John’s gospel 131 – 132; Lightfoot 133; Pannenberg 136 – 137; Reimarus 129 – 130; Schweitzer 134, 136, 161; Seeley 132; Stanton 135; Strauss 130 – 131 gospels 7, 128; dating of 137 – 138; historicity of 6, 137; Infancy

Gospel of Thomas 139; Jesus’ mighty works 139 – 144; of John 178, 181,  184, 187; of Luke 177 – 178, 179 – 180,  182, 184; of Mark 176, 177, 179, 184; of Matthew 177, 179, 184; miracles 107, 139 – 144 Grand Design, The (Hawking and Mlodinow) 39 – 40 Grand Unified Theory 57 great fire of Rome 172, 173 Gregory of Nyssa 16 GUT see Grand Unified Theory Harman, Gilbert 21 Harnack, Adolf von 133 – 134 Hartle-Hawking theory 39 – 40 Hartle, James 39 – 40 Harvey, Anthony 141 – 142 Hawking, Stephen 39 – 42 Hebblethwaite, Brian 68, 69 – 70 Hegel, G. W. F. 18, 131 – 132 Hegesippus 172 Heidegger, Martin 135 Hellenistic Judaism 142 Hengel, Martin 138 historical-critical method 76; see also gospel accounts historical enquiry, Troeltsch’s principles of 75 Holocaust, prima facie incredibility 76 Hooker, Morna 135, 141, 142, 144, 157 – 158, 161, 182 Hoyle, Fred 44, 46 Hume, David 3, 11, 75 – 76, 94 – 95, 106; criticism of natural theology 12 – 13; definition of miracle 6, 106, 108 – 109, 111; miracles argument, Bayesian framework 111 – 113; miracles argument, Bayesian formulation impact on 118 – 119; ‘uniform experience’ against miracles 107 imaginary time 39 – 40, 41 Infancy Gospel of Thomas 139 inference to the best explanation see best explanation, language of inflation, theory of 48 – 49 initial hypothesis 37, 41, 43, 55 interconnectivity, principle of 75 Irenaeus 73 Islam 6, 70, 72 – 74, 87, 99, 139 Israel 155

Index  231 James’s death (brother of Jesus) 172 James, William 18 Jansenism 83 Jeremiah’s prophecy 91 Jesus and the Eyewitnesses (Bauckham) 138 Jesus’ death on cross 70 – 74; evidential criteria for 71; idea that a substitute crucified 73 – 74; Islam’s denial of 72 – 73; liberal Jesus not a threat 71 – 72; Mark’s account of 158; see also Christ, Jesus; resurrection of Jesus Christ ‘Jesus Seminar’ 135 – 136 Jews, expulsion from Rome 173 Judaic (Petrine) Christianity 132, 133 Judaism 6, 30, 71, 78, 135, 142, 173, 185–186; see also Rabbinic literature kalām cosmological argument 37 – 38, 40 Kant, Immanuel 3, 9, 12 Kateregga, Badru 72 Kingsley, Charles 3, 9, 16 – 17 ‘knock-down’ proof 23 knowledge of God: available to human beings 67; and natural theology 10 – 12; revelation 66; sources of 1, 65 – 66 Kraemer, Hendrik 69, 74 Krauss, Lawrence 42 – 43 Leibniz, Gottfried Wilhelm 9, 14, 15, 38, 42, 44 Lessing, Gotthold Ephraim 129, 130, 131, 136 Lewis, C. S. 7, 69, 107, 139, 148 – 149, 168, 215 Liberal Protestantism 5, 6, 134 Life of Jesus Critically Examined (Strauss) 130 – 131 Lightfoot, J. B. 133 likelihood ratio see Bayes factor Locke, John 11, 31, 94 – 95 logical probability 28 Lorentzian geometry 40 MacIntyre, Alasdair 21 Mack, Burton 136 Mackie, J. L. 101 Magdalene, Mary 176, 176, 177, 178, 178, 179, 181, 182, 184, 197 Mahometanism 99; see also Islam male disciples of Jesus 182 Mawson maximal multiverse 53, 56, 57

McCartney, Mark 4, 36, 41 – 42, 43 – 44, 55 – 58 McGrath, Alister 4, 19 – 22, 28, 74, 129, 133 McGrew, Timothy, and McGrew, Lydia 7, 77, 79, 145, 189, 190 – 198, 199 – 200, 201, 202, 204 – 206, 208, 210, 214 Meier, John P. 71 miracles 86, 98, 170; aggregation of testimonies to 109, 114 – 118, 212 – 214; Bayesian framework 109 – 113, 212 – 214; credibility of 99, 106; definition of 6, 106, 108 – 109, 111; exorcisms 141; healings and nature miracles 139, 142; independent testimonies to 109, 113 – 114; Infancy Gospel of Thomas 139; Jesus’ purported, Biblical accounts of 6, 139 – 145; prior probability of 110; and prophecies 5, 6, 86 – 90, 101, 166; St Augustine on 108; testimony as evidence for God existence 109, 120 – 125; ‘uniform experience’ against 107; see also resurrection of Jesus Christ Miracles (Lewis) 107 ‘model-dependent realism’ 39 – 40 modern cosmology 4, 17, 46, 202; fundamental issues raised by 36; recent developments in 48 – 50 modern natural theology 13, 19 – 29 Mohammed 70, 73, 87, 139; see also Islam motivated belief 23 Moule, C. F. D. (Charlie) 73 – 74 M-theory 49 multiverse hypothesis 22 – 23, 36, 47 – 48; Bayesian evaluation of 60 – 62; Christians welcoming 50 – 51; compatibility with divine design 55; vs. design by God 58; fine-tuning problem solving 55 – 57; independent evidence for 57; Mawson’s criticism of 53, 55, 57; need for design 58; observer selection principle and 56; problems for 51 – 55; Tegmark’s Level IV 48, 53, 56, 57; truth of 57 Muslim religion see Islam naturalistic assumptions 2, 6, 131 natural knowledge of God 5, 10 – 11 natural theology 1, 3, 4, 5, 9, 67, 74; definitions of 19 – 20, 22, 65;

232 Index knowledge of God and 10 – 12; and motivated belief 22 – 24; in practice 11 – 12; rejection of 18, 67 – 68, 83; vs. revealed theology 19, 68, 78 natural theology, design arguments of 12 – 19; Aquinas 10, 12; Augustine 16; Barth’s rejection of 18, 67 – 68; Darwinian evolution 15 – 16; Gifford Lectures 18; Hume’s criticism 13 – 14; Kant’s criticism 12; Kingsley 16 – 17; Leibniz 14; Paley 14 – 15; Temple 17 Newman, John Henry 2 – 3 New Testament: criteria for assessing authenticity 71; historical critical study of 128 – 137 New Testament evidence for the resurrection 171 – 173, 189; alternative explanations of 183 – 186; comparison with scientific evidence 186 – 187; male disciples of Jesus 182; overview of 174 – 176; Paul’s list of resurrection appearances 181, 182–183; testimony of women 182 – 184; witnesses to empty tomb 176, 176 – 178; witnesses to Jesus’ appearances after his death 178, 178 – 181 Nicene-Chalcedonian formulation 20 ‘no boundary proposal’ 39 – 41 ‘nothing,’ concept of 42 – 43 observer selection principle 56 Origin of Species, The (Darwin) 16, 132 Paley, William 3, 9, 14 – 15 Pannenberg, Wolfhart 1 – 2, 11, 31, 68 – 69, 74 – 76, 78, 136 – 137, 196 – 197 Papias 7, 144, 174,  181, 191, 196, 213 Pascal, Blaise 5 – 6, 82 – 93; approach to prophecy 86 – 91; arguments for superiority of Christianity 86; conversion experience 83; emphasis on man’s total wickedness 83 – 84; on metaphysical proofs 84; predicting timing of birth of Jesus 90 – 91; on religious pluralism 87 – 88; repudiates arguments from nature 83, 84 – 85; wager 85 – 86 Pauline Christianity 132, 133 Peacocke, Arthur 20, 50 – 51, 55 Penelhum, Terence 93 – 94 Penrose, Roger 57 Pensées (Pascal) 83 – 93 Petrine (Judaic) Christianity 132, 133

physical law, design of 15 physico-theology 3, 12 Plantinga, Alvin 20, 67, 68 – 69, 75 Plato 7, 10, 149, 168 Pliny the Younger 173 Polanyi, Michael 21, 23 – 24 political ideologies 66 Polkinghorne, John 4, 20, 21, 22 – 24, 28 – 29, 129 ‘positivism of revelation’ 67 posterior probability 7, 25, 145 presuppositions 74 – 76 primary and secondary causes, distinction between 108 – 109 ‘Principle Credulity’ 171 ‘Principle of Testimony’ 129, 171 prior probability 7, 25, 27 – 28, 60, 62, 100, 110, 111, 118, 120, 122, 123, 145, 168, 193, 194, 198, 207 probability 5, 25 – 26; conditional 25; epistemic 94 – 95; logical 28; posterior 7, 25, 145; prior 7, 25, 27 – 28, 60, 62, 100, 110, 111, 118, 120, 122, 123, 145, 168, 193, 194, 198, 207; subjective 28 properly basic beliefs 20, 68, 75 prophecy 86 – 87, 143; definition of 148 – 150; feeding miracle 161; formalizing argument from 164 – 168, 214 – 216; fulfilment of 149 – 150, 211, 214 – 217; healings 153; Israel’s hopes 155, 160; Jesus’ crucifixion 153 – 154, 156, 157; Jesus’ selfdescription as Son of Man 159 – 160; John the Baptist 161 – 162; and miracles 166; mockery of Jesus 156 – 157; proclaiming Kingdom of God 152; Suffering Servant 151 – 153, 155; texts on fulfilment of 150 – 164 putative revelation, arguments or evidence for validity of 67 quantum theory, interpretations of 5, 66 – 67 quantum vacuum creation 42 – 44 ‘Quest of the Historical Jesus’ 129 – 134 Qur’an: Jesus’ death 72, 73; Jesus’ miracles 139 Rabbinic literature 73, 135, 141, 160 – 161, 201 rationality: theological 1 – 3; tradition transcendent (McGrath) 21; universal (McGrath) 21

Index  233 real time 39 – 40 reason and revelation 65 – 66 Reid, Thomas 171 Reimarus, H. S. 129 – 130 religions see religious pluralism religious pluralism 5, 24, 31, 66 – 70, 74, 86, 87, 99, 128; atheism 66; Christianity 67; ‘positivism of revelation’ 67 resurrection of Jesus Christ 2, 7, 70, 74, 76 – 78, 106, 137, 170; see also Bayesian formulation, of resurrection arguments; Christ, Jesus; Jesus, death on cross; New Testament evidence Resurrection of the Son of God, The (Wright) 208 revealed theology vs. natural theology 68 revelation of God 1, 5, 10, 65 – 66, 137 Rolls Chapel 93 Rowland, Christopher 140 – 141 Sanhedrin 43a 73, 141 Sanhedrin 107b 141 Schellenberg, J. L. 70 Schleiermacher, Friedrich 1 – 2 Schweitzer, Albert 134, 136, 161 science, contradictory theories in 66 – 67 scientific explanations 29 – 30, 32 Smith, Quentin 40 Soskice, Janet 44 Stanton, Graham 135 steady-state theory 5, 30, 44, 67, 199 Strauss, David Friedrich 130 – 131 string theory 49 – 50, 51, 52, 53, 66, 199 subjective probability 28 Suetonius 173 ‘super-miracle’ (Swinburne) 70, 203, 211, 217; Bayesian formulation 202 – 203, 211, 214, 217 Sûrah IV 72 – 73 Sûrah V 139 Susskind, Leonard 49 – 50 Swinburne, Richard 4, 7, 9, 12, 13, 21, 24 – 29, 31, 41 – 42, 70, 102, 171; approach to biblical data 195 – 198, 207; Bayesian approach 198 – 205, 207 – 208 Tacitus 172 – 173 Tegmark’s Level IV multiverse 48, 53, 56, 57 Tegmark, Max 48 Temple, Frederick 3, 9, 17

testimony 6, 11, 76, 95, 99, 100, 106 – 107, 109 – 110, 113, 118 – 119, 121, 122 – 126, 129, 138, 144, 170 – 171, 175, 177, 186, 187, 192 – 193, 195, 196, 197, 205 – 207, 212 – 213 testimony of women  182, 183  theism 4, 5, 13, 22, 30, 36 Theissen, Gerd 136, 143 theistic explanation 29 – 32 theological rationality 1 – 3 theology 1 – 2, 21, 69 – 70 ‘theory of everything’ 47, 49 Through the Looking Glass (Carroll) 42 – 43 TOE see ‘theory of everything’ Torrance, Thomas 5, 19, 20, 22 tradition transcendent rationality 21 Trinitarian-Chalcedonian Christian belief system 74; see also NiceneChalcedonian formulation Trinity 20 Troeltsch, Ernst 5, 75, 79, 129, 130, 131 truth of Christianity 2 – 3 universal rationality 21 universe 4, 9, 22, 30; animal analogy 13; beginning of 4, 36, 38, 39 – 40; 41 – 42; boundary condition of 39; Christian doctrine of creation 44 – 45; dependence on God 44; design on basis of theism 13 – 14; as divine creation 41 – 42, 44, 45; finetuning of 17 – 18, 22, 36, 46 – 48; kalām cosmological argument 37 – 38, 40; multiverse hypothesis and 47 – 48; ‘no boundary’ proposal for 39 – 41; observers and 56 – 57; ordered and structured 57; probabilities relevant to 27 – 28; quantum vacuum creation 42 – 44; ‘theory of everything’ and 47; ultimate explanation of 45 Universe from Nothing, A (Krauss) 42 Vestiges of the Natural History of Creation (Chambers) 132 Vilenkin, Alexander 40 Westermann, Claus 151 – 152, 154, 167 Westminster Confession of Faith 18 witnesses to empty tomb 176, 176 – 178; gospel of John 178,  184; gospel

234 Index of Luke 177 – 178, 184; gospel of Mark 176 – 177, 184; gospel of Matthew 177, 184 witnesses to Jesus’ appearances after his death 178, 178 – 181; acts of apostles 180–181; 1 Corinthians 15 181;

gospel of John  181; gospel of Luke 179 – 180; gospel of Mark 179; gospel of Matthew 179, 184 Wright, N. T. (Tom) 71, 72, 77, 132, 136, 139 – 140, 142 – 143, 159, 160, 162, 163, 177, 185–186