Ramified Natural Theology in Science and Religion Moving Forward From Natural Theology PDF

‘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(considered merely as an ordinary historical document) and other historical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 theo-
logians. 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 dis-
tinction 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 Christian-
ity. 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 inspira-
tion 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 Founda-
tion. 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 por-

tions 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 Con-
shohocken: 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 Philoso-
phia Christi at www.epsociety.org.
• Taylor and Francis for Rodney D. Holder, ‘Explaining and Explain-
ing 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. Tradi-
tionally 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 empiri-
cal 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 argu-
ments 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 pro-
vide reasons for acceptance. In particular one needs to take account of Bibli-
cal 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 theol-
ogy 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 Pannen-
berg’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 inad-
equate; 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 gener-
ally 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 prob-
lem 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 his-
tory, 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 theolo-
gians 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 ben-
efits 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 con-
science? 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 knowl-
edge, a fact encompassing, closing in upon, absorbing, every other fact
conceivable. How can we investigate any part of any order of Knowl-
edge, 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 unphilo-
sophical. 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 natu-
ral 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 argu-
ments 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 the-
ology, 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
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 pro-
vided 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 poten-
tially, lead to arguments of natural theology: does the universe have a begin-
ning, 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 fea-
tures 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 cre-
dence 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 pub-
lished 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.
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 contra-
dict 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 knowl-
edge 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 interpreta-
tions 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 rami-
fied natural theology, namely Blaise Pascal and Bishop Joseph Butler. A sig-
nificant 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 spe-
cifically 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
forward to the present day. For Butler writing in the first half of the eight-
eenth 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. Moreo-
ver, 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 ‘viola-
tions 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 accumula-
tion 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 exist-
ence 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 Enlighten-
ment, with the result that one finds a rather bland Jesus who, as one com-
mentator 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
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 fol-
lowed 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 gos-
pels, 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 out-
side 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 them-
selves 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 con-
firmation theory, which has been introduced and deployed in earlier chap-
ters, is an invaluable tool for evaluating competing hypotheses, which each
explain the evidence in question, in a rigorous way. However, authors utiliz-
ing Bayesian confirmation theory to evaluate the evidence for the resurrec-
tion 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 fac-
tors, 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 the-
ology, so bare theism is already at least moderately likely. The McGrews
and Swinburne also vary widely in the results they derive. This chapter re-
examines the issues involved and suggests a resolution to the conflicting
methodologies and results. It shows how feeding natural theology into rami-
fied 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 fulfill-
ing Messianic prophecies and supremely his resurrection, makes it over-
whelmingly 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: Con-
structing Theories in Systematic Theology (Grand Rapids: Wm. B. Eerdmans,
1989), 71–100.
4 Pannenberg, Theology and the Philosophy of Science, 255; also Systematic The-
ology, 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 Lec-
tures 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.
11 David H. Glass and Mark McCartney, ‘Explaining and Explaining Away in Sci-
ence and Religion,’ Theology and Science 12(4) (2014), 338–361.
12 Rodney D. Holder, ‘Explaining and Explaining Away in Cosmology and Theol-
ogy,’ Theology and Science 14(3) (2016).
13 Timothy McGrew and Lydia McGrew, ‘The Argument from Miracles: A Cumu-
lative Case for the Resurrection of Jesus of Nazareth,’ in William Lane Craig and
J. P. Moreland (eds.), The Blackwell Companion to Natural Theology (Chiches-
ter: Wiley-Blackwell, [2009] 2012), 593–662.
14 Richard Swinburne, The Resurrection of God Incarnate (Oxford: Oxford Uni-
versity Press, 2003). Swinburne’s argument is summarized in Richard Swin-
burne, ‘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 puta-
tive revelation.
In this chapter I give an overview of natural theology from Thomas Aqui-
nas 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 cop-
ied 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 sci-
ence 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.
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 revela-
tion 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 Scrip-
ture. 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 require-
ment 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, can-
not be afforded to anything, but upon good Reason.’7 In the century follow-
ing 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. Pan-
nenberg 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 acces-
sible 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 mar-
riage 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 cosmologi-
cal and design arguments. The cosmological argument says that everything
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 sci-
entific 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 sci-
ences 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 physico-
theological (design). However, as Richard Swinburne notes, he failed to dis-
tinguish different forms of these.
Kant said that we can only experience phenomena; we cannot experi-
ence ‘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 appar-
ently 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 rel-
ativity. Indeed, Kant has been shown to be in error in claiming the necessity
of describing the world through Euclidean geometry because this is embed-
ded 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 per-
haps 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.
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 phe-
nomena are not like houses in that respect, and one needs a very close anal-
ogy to make the inference to design. To transfer an inference about parts of
the universe exhibiting design to the whole doing so would be dispropor-
tionate: 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 crea-
tor 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 uni-
verse, 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 con-
stitute 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. Moreo-
ver, 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 com-
plement 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 dif-
ferent 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
do just that. And against another of Hume’s points, if one universe is gener-
ated 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 athe-
istic 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 hun-
dred 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 nowa-
days. Thus he clearly considered various objections to the argument, some
of which remind one of Hume. For example, he asked whether the imper-
fection 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 pos-
sible combinations of material forms’ and whatever we found would have
had to take some configuration or other. That kind of argument is some-
times 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 dif-
ference 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 par-
ticulars 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 con-
clusion 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 infi-
nitely 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
at how the Christian doctrine of creation has been understood by the early
Church Fathers. Thus Gregory of Nyssa sees God implanting causal princi-
ples 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 fol-
lowed 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 futuro-
rum, 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 veg-
etation, 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 receiv-
ing 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
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. Echo-
ing 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 evolu-
tion. 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
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 the-
ology is provided by the inauguration in 1888 of the Gifford Lectures in nat-
ural 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 central-
ity 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 per-
sonal 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 tradi-
tional 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 choos-
ing 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
post-Second World War period, though there is still no discernible unity or
agreement. On the positive and orthodox side would be Austin Farrer, Stan-
ley 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 dog-
matic 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 dif-
ferences 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
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, ‘natu-
ral 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 dis-
cipline 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 natu-
ral theology must be brought within dogmatic theology, in its full Nicene-
Chalcedonian 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 apologeti-
cally. 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 science-
religion dialogue; both John Polkinghorne and Arthur Peacocke are exem-
plars. 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, appar-
ently 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.
McGrath also appeals to critical realism to assert, like Barth and Tor-
rance 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, the-
ology will employ methods different from the other sciences which deal with
other aspects of reality. Again, Polkinghorne would agree, though Swin-
burne 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 reli-
gions. That is to say, Christian rationality is not only tradition-dependent
but tradition-transcendent. A key reason why it can explain other reli-
gions 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 univer-
sal rationality.
Inference to the best explanation

McGrath’s third tool from philosophy is ‘inference to the best explana-
tion,’ 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 ques-
tion. 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 uni-
versal 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.
McGrath invites us to infer that the Christian faith offers the best expla-
nation for such features of the universe as the fine-tuning of physical con-
stants, 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 fail-
ure 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 Mathemati-
cal 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 meta-
phorical resource.’ However, it seems to me that Polkinghorne’s understand-
ing 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 the-
ology 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 satis-
fying 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 mul-
tiverse. Yes, says Polkinghorne, but theism explains more than a multiverse
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, Alis-
ter 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 ‘compat-
ible,’ ‘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 exam-
ple of consonance, Polkinghorne notes, referencing John Zizioulas, that the
triune God of Christian theology is ‘Being in Communion.’62 Modern sci-
ence has provided a much more relational picture of the universe than clas-
sical 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 hav-
ing 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 com-

plex phenomena, a goal to be achieved by searching for an understand-
ing 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 tradi-
tional 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,
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 Polk-
inghorne understands ‘critical realism.’ Indeed Polkinghorne’s adoption
of this, which pre-dates McGrath, also includes such aspects as the multi-
layered texture of reality, the subtle interplay between theory and experi-
ment, and the recognition that there are no facts which are uninterpreted
facts. He does not, however, endorse McGrath’s view that nature, as under-
stood 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 phi-
losophers of religion of recent times. His pioneering approach to natural
theology is entirely different from McGrath’s, and goes beyond Polking-
horne’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 explana-
tion,’ 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.
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 argu-
ments 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 the-
ory. 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, comprehen-
sive, 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 bod-
ies 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[E|H ].P[H ] (2.1)
P[H|E] =
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 prob-
ability 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 pos-
terior 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.
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 (2.2)
P E|H .P H P E| ~ H .P ~ H
What the theorem tells us is how to revise the probability of some hypoth-
esis 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 equa-
tion: 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 like-
lihood 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 E|H1 P H1
. (2.3)
P H 2 |E 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 E|H P H
. (2.4)
P ~ H|E P E| ~ H P ~ H
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
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 psychologi-
cal tests reveal people’s judgments to be. Many people confuse the likeli-
hood, 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 fine-
tuned 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
to explain any phenomenon the simplest will have the greatest prior prob-
ability, says Swinburne. This seems to be the case for science and there is
no reason for it to be otherwise when we compare metaphysical hypoth-
eses. 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 pre-
cise 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 prob-
abilities. 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 sub-
jective 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 prob-
ability 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 argu-
ments 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 the-
ology 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 tradi-
tional 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 fine-
tuning, for a Creator behind the universe, and historical evidence for the
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 perspec-
tive. 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 cri-
teria of simplicity, etc. in a similar way to experts in the scientific commu-
nity, 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 criti-
cal realist approach to Polkinghorne and a more rationalist, possibly ultra-
rationalist 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 pro-
vides. Explanations are of different kinds. And they do not necessarily
exclude each other. God provides a personal explanation, e.g. for the exist-
ence 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
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 explana-
tions begin to run out. Indeed the whole story is open to a personal explana-
tion 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 over-
flowing 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 pur-
pose 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.
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 expla-
nation 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 state-
ment 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 actu-
ally 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 Chris-
tian 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 resur-
rection 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 disil-
lusioned. 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.
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 reli-
able 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: Hen-
drickson, 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 intro-
duction 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 Concern-
ing 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 Wil-
ckens, Revelation as History, trans. D. Granskou (New York: Macmillan, 1968),
138. (First German edition, Offenbarung als Geschichte, Göttingen: Vanden-
hoeck 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.
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.
18 Gottfried Wilhelm Leibniz, ‘On the Ultimate Origination of Things,’ in Rob-
ert 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 Univer-
sity 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.
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 Febru-
ary 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 Lec-
tures 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 Rus-
sell 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: Harper-
SanFrancisco, 2005), 71.
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 Vol-
ume 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 Rap-
ids, 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 Polking-
horne, 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.
68 Most notably in Swinburne, Existence of God.
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 (Cam-
bridge, 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.
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 cosmo-
logical argument for the existence of God, in one form or another. The sec-
ond 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 evi-
dence 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 depend-
ence of HA on HI is strong enough, it may imply that HA provides addi-
tional 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. Like-
wise, it should be distinguished from Leibniz’s argument based on the prin-
ciple 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 prem-
ises, 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 cos-
mological 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 begin-
ning 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.
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 pro-
posal 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 uni-
verse 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
odd philosophical position he calls ‘model-dependent realism’ (despite say-
ing 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 say-
ing 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 geom-
etries 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 philoso-
pher 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 expan-
sion, 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 multi-
verse 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.
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 alterna-
tive 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 con-
cept (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-Robertson-
Walker) 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 contro-
versial. 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
compared with a complex physical universe and therefore a priori more
probable.18 And, as we shall see when discussing the fine-tuning, the uni-
verse 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 crea-
tion still stands. In biology, a form of design argument is undermined
by the theory of evolution, namely separate design of individual organ-
isms. However, design of the laws of nature that underlie evolution is
not undermined. Here, if there were good reason to deny that the uni-
verse 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 cosmo-
logical 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 so-
called ‘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 Noth-
ing,19 Krauss redefines the notion of nothing so that the quantum vacuum
can be identified with nothing. This really is sleight of hand. To philoso-
phers, ‘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
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 quan-
tum 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 anni-
hilated 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, infla-
tion needs to occur for just that tiny fraction of a second and then to be
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 cosmo-
logical 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 bil-
lion 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,
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 explana-
tion. 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 cre-
ated 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 elo-
quently 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 won-
derful 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 can-
not 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.
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 uni-
verse’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 astro-
physicist 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 chem-
istry 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 reli-
gious belief as illusory.26
As I say, there are a host of these examples of fine-tuning which I cannot dis-
cuss in detail, although some more examples will crop up later. The cosmol-
ogist Paul Davies puts it like this: ‘Like the porridge in the tale of Goldilocks
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. Chris-
tians 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 relation-
ship 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 obvi-
ously 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 iso-
lated 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
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 fun-
damental 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 fundamen-
tal 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 uni-
verses 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 mul-
tiverse’).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 struc-
tures.’ 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
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, ostensi-
bly solving what was also a problem with the standard Big Bang.
That sounds wonderful, but there were some serious problems with infla-
tion. 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 cha-
otic 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 ulti-
mate building blocks of matter are not point-like particles but tiny, one-
dimensional 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
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 theo-
rists, 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, any-
thing 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 Infla-
tion 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 consider-
able 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, theologi-
ans, 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 mul-
tiverse. 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
produce cognizing subjects, not that of one universe in particular.33 The fine-
tuning 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 crea-
tive capacities, so physical reality might be much larger than a single universe.
The idea would be that God expresses his infinite creativity, rather than sim-
ply and purposively creates a single universe with life in it. Collins thinks that
creating a multiverse through a single physical universe-generating mecha-
nism, 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 multi-
verse 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 pro-
ponent 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 mul-
tiverse 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 prin-
ciple 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
starts to feel at home when familiar physics can be applied to the uni-
verse, 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 mathemati-
cians think infinitely many universes are ruled out because of the para-
doxes. 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 nor-
mally 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 land-
scape 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 pre-
dictability in general based on the existence of order in the universe. Sup-
pose some unexplained feature arises in the universe. Instead of trying to
explain it rationally using ‘normal’ science, the temptation is now to say,
‘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 Nel-
son 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 intel-
ligent 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 lat-
est observations indicate that Λ probably takes a very small, but posi-
tive, 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 bil-
lions 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
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 observ-
ing 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 uni-
verse 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 uni-
verse. John Barrow makes just this point.45 Either of two options is pos-
sible. 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 begin-
ning. 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.
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 spe-
cial 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 uni-
verses 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 fine-
tuned, 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 hypoth-
esis 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 some-
one 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 mul-
tiverse 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 uni-
verses. 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
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 Princi-
ple.’48 If there were no observers, there would be no one to ask the ques-
tion, ‘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 typi-
cal 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, ‘sig-
nificantly 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 technol-
ogy. At a billion-fold, ECAs could not evolve at all. Thus one has maxi-
mal 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
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 argu-
ment, one could estimate that there are maybe 1022 times as many observ-
ers 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 lat-
ter would be completely unexplained. Add in Mawson’s criticism of the
maximal multiverse, that universes in which induction fails to work over-
whelmingly 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
Question 4. Does the multiverse hypothesis depend on design?
Unless one adopts the wildly anti-Occamite maximal multiverse, a particu-
lar 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 explana-
tion, and that design is an option. Some physicists, philosophers, and
theologians—including Richard Swinburne, Robin Collins, John Polk-
inghorne, 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 expla-
nation 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 explana-
tion, 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 meas-
ure, 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 E|H .P H
P H|E (3.1)
P E|H .P H P E| ~ H .P ~ H
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 charac-
terized 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 hypoth-
esis ~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 over-
flowing 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
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 math-
ematically, P[E|H] >> P[E|~H]. Remembering that by the laws of probabil-
ity 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 necessar-
ily 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 uni-
verse 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 begin-
ning 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 for-
mulation 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
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 uni-
verse like ours than that this is what we would observe given a multiverse.
Given the complexity and contingency of a multiverse, including the num-
ber 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 com-
parison 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
P[H | E ] 123
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’
123
probabilities are washed out
by the monstrously huge number 1010 .
Rewriting equation (3.1) with E′ substituted for E we obtain
P E |H .P H
P H|E (3.5)
P E|H .P H P E | ~ H .P ~ H
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
have been assuming, the one equating to the God of classical Christian the-
ism. But, as Swinburne argues, the hypothesis of many gods is not simple
like the single, unique God hypothesis.56 In particular, it violates the prin-
ciple of Occam’s razor, that ‘entities are not to be multiplied beyond neces-
sity.’ 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 pub-
lished 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 Sci-
3 David H. Glass, ‘Can Evidence for Design be Explained Away?’ in Jake Chan-
dler 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, Crea-
tion 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 Con-
temporary 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 Singu-
larity,’ Physical Review Letters 72(21) (1994), 3305–3308. The theorem was
extended from applying to inflationary models to include ‘brane’ cosmologies
in Arvind Borde, Alan H. Guth, and Alexander Vilenkin, ‘Inflationary Space-
times 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) (Janu-
ary 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 Uni-
versity 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.), Crea-
tion 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 Rea-
sons 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).
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 Multi-
verse? 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.
53 Swinburne, Existence of God, 6, 133–152.
54 Swinburne, Existence of God, 96–109, 165, 185–188.
55 Davies, Goldilocks Enigma, 249.
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 rea-
son 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 peo-
ple, 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 revela-
tion, 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 particu-
larities 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 the-
ology, 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 plu-
ralism. 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. There-
fore 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 stand-
ard 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 determin-
ist 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 gen-
eral and minimal knowledge of God available to all human beings simply
through being human, then different religions may agree about that. Per-
haps 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 argu-
ments 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 Plant-
inga (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 the-
ology (most especially, we may add, in his ‘Nein!’ to Emil Brunner).5 Barth
offered instead what Dietrich Bonhoeffer described as a ‘positivism of rev-
elation.’ 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
again into the open air of intellectual discussion with the world,’ says Bon-
hoeffer, 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 knowl-
edge. 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 theol-
ogy 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
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 exam-
ined 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 dif-
ference between Kraemer and Pannenberg is that what for Kraemer is a
premise of faith (the Christological starting point) is for Pannenberg a con-
clusion 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 eschato-
logical 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 reli-
gions thus resembles that of a very different Christian thinker, C. S. Lewis,
for whom other religions contain truths and hints, for example in their sto-
ries 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 theol-
ogy, each religious tradition which claims to provide knowledge of God
must have something to say about the different claims of the other reli-
gions. 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
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. Schel-
lenberg 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 reli-
gions at their face value. The fact that Islam makes no claim that Moham-
med 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 actu-
ally 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 impor-
tant example, the resurrection depends for its truth on the earlier historical
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 Japa-
nese koans, a literary aesthete who toyed with 1st-century deconstruc-
tionism, 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 his-
torical ‘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 any-
one would have wanted to crucify such an attractive moral teacher.’26 Like
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 crucifix-
ion, 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 appar-
ent?’ 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 conjec-
ture; 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 inappropri-

ate 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 redemp-

tion 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.
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 cen-
tury. 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 cruci-
fied 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 suf-
ferings 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 trans-
figured 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 state-
ment 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
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 lan-

guage, 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 resur-
rection 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 common-
sense, 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 plural-
ism, he also claims, building on the work of Alasdair McIntyre, that the
Christian faith offers not just a tradition-specific rationality, but a tradi-
tion-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
of God. As Pannenberg puts it, ‘the divine reality which is the subject of reli-
gious 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 disci-
pline, 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 main-
tained three principles of historical enquiry. The first, the principle of criti-
cism (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 phi-
losophers 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 crite-
rion 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 (Kor-
relation, but also called Wechselwirkung, interconnectivity).45
By means of the principle of analogy, that is, by virtue of the historical-
critical method he adopted, Troeltsch was able effectively to deny the res-
urrection 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 English-
speaking world, Troeltsch’s position resembles that of David Hume, espe-
cially 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,
utilizing Bayesian probability theory, of which Hume was apparently igno-
rant.48 I return to this topic in Chapter 6.
Pannenberg too refutes this a priori reasoning, while remaining fully com-
mitted 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 histo-
ricity 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 Holo-
caust 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 sur-
vivors 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 con-
tested 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
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 compo-
nent 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 halluci-
nations 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 see-
ing 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’ resurrec-
tion 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 develop-
ment 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
In fact we can see something like the principle of correlation or intercon-
nectivity 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 meta-

physical 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 pri-
macy 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 histori-
cal evidence. That means that we should not make any prior assumption of
the correctness or accuracy of Christian claims, but nor should we exclude
Christian claims, such as the existence of God or the occurrence of mira-
cles, 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 anal-
ogy 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’ experi-
ence 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 seri-
ous 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 methodolog-
ical in nature, applying Bayesian confirmation theory to the notion of mira-
cles 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.
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:
5 Karl Barth, ‘No!,’ Response to Emil Brunner, ‘Nature and Grace,’ in Emil Brun-
ner and Karl Barth, Natural Theology, trans. Peter Fraenkel, with an introduc-
tion by John Baillie (London: Geoffrey Bles/Centenary Press, 1946), 65–128.
6 Dietrich Bonhoeffer, Letters and Papers from Prison, enlarged edition, ed. Eber-
hard Bethge (London: SCM Press, 1971), 378. See my discussion in Rodney
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 Uni-
versity 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 edi-
tion, Wissenschaftstheorie und Theologie, Frankfurt am Main: Suhrkamp, 1973).
10 Pannenberg, Theology and the Philosophy of Science, 273.
(Edinburgh: T & T Clark, 1991), 128.
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: Wolf-
hart 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 Christi-
anity 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 Rich-
ard 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.
33 Irenaeus, Adversus Haereses, I.XXIV.4, in Alexander Roberts and James Don-
aldson (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 (Lon-
don: 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 Theolo-
gie,’ in Gerhard Sauter (ed.), Theologie als Wissenschaft (München: Chr. Kaiser
Verlag, 1971), 105–127. Originally published in Studien des rheinischen Predi-
gervereins, 1898.
46 Troeltsch, ‘Über historische und dogmatische Methode,’ 108.
[1748] 1975).
48 Rodney D. Holder, ‘Hume on Miracles: Bayesian Interpretation, Multiple Testi-
mony, 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).
(Edinburgh: T & T Clark, 1994), 360–361.
50 Richard Bauckham, Jesus and the Eyewitnesses: The Gospels as Eyewitness Tes-
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006), 499.
51 Bauckham, Jesus and the Eyewitnesses, 493.
54 Daniel P. Fuller, ‘The Resurrection of Jesus and the Historical Method,’ Journal
of Bible and Religion 34(1) (1966), 18–24, 22.
lative Case for the Resurrection of Jesus of Nazareth,’ in William Lane Craig and
ter: Wiley-Blackwell, [2009] 2012), 593–662, 624.
56 N. T. Wright, The Resurrection of the Son of God: Volume 3 of Christian Ori-
gins 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), dis-
cusses both natural theology and revelation and does so in terms of evi-
dence 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 theol-
ogy 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 theol-
ogy 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, how-
ever, dated in view of the fact that they preceded the rise of historical-
critical study of the Bible, which occurred alongside Darwinism in the
nineteenth century; also they precede the development of Bayes’s theo-
rem 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 discover-
ies 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 Pas-
caline), which probably explains the naming of a modern computer pro-
gramming 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, Pas-
cal 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 eradi-
cated 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 Saint-

Cyran find many echoes in the Pensées. ‘First among these,’ writes Cruick-
shank, ‘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 pas-
sage 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 dur-
ing 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
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 uncer-
tainty. 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.
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 pro-
tests 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 cen-
tury, 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 experi-
ence, 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 prophe-
cies show that that these truths are certain and thus prove that Jesus is
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 abso-
lutely 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 Chap-

ter 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 funda-
mental 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 com-
manded it. In a word, the difference is so great that, if Mahomet fol-
lowed 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 per-
formed 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,
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 spiritu-
ally. . . “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
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 suscep-
tible 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 crea-
tion and flood stories were metaphorical or in any way non-literal (though
we noted St Augustine’s more nuanced interpretation of creation in Chap-
ter 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. Moreo-
ver, 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 over-
thrown, so many changes of every kind, while the Church which wor-
ships 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
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 evi-
dence, 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 Dan-
iel 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 Artax-
erxes 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 com-
monly accepted view in Sacy’s (and Pascal’s day) was that Jesus was born in
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 mon-
archies, 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 exam-
ple, we now know that the universe began, not 6000 years ago but 13.8 bil-
lion 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 under-
stand that from the going forth of the word to restore and build Jeru-
salem 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; desola-
tions 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
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 indi-
cate that they are looking centuries or millennia beyond the period to
which chaps. 8 and 10–12 refer. . . . The passage refers to the Anti-
ochene 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 exe-

getical, 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 reveal-
ing 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 typo-
logical 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
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 Stan-
hope, 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 com-
mended 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. Glad-
stone 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 sub-
ject 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
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 counter-
arguments 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 demon-
strative proof:
Probable evidence is essentially distinguished from demonstrative by

this, that it admits of degrees; and of all variety of them, from the high-
est 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 presump-
tion 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 mean-
ing for probability here could be read as resembling the modern, Bayes-
ian 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
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 argu-
ment, 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 quan-
tum 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. Moreo-
ver, Butler’s argument has an analogy with Pascal’s wager—it would be pru-
dent 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 accumu-
lated 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 dis-
satisfied with the evidence of religion.
(Introduction, 10)
From this quotation, we see that Butler supports natural theology and vari-
ous 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 gov-
ernance 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
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, sec-
ondly, 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 govern-
ment of it, we see things, in a great degree, unlike one another; and
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 cer-
tainly seem to be against nature. However, again, he appeals to our igno-
rance 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 mirac-
ulous 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 experi-
enced 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
Butler that sufficient evidence can make the miracle probable (see Chap-
ter 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 argu-
ment 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 con-
cept 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 pre-
sumption 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 Chap-
ter 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 the-
ology, 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 Concern-
ing 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 com-
pletion of prophecy, in the common acceptation of the words. But these
two are its direct and fundamental proofs: and those other things, how-
ever 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 his-
torical evidence of the miracles of Christ and the apostles, as of the com-
mon 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
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 prof-
fered, 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 let-
ters which came from the hand of Paul (Romans, 1 Corinthians, 2 Corinthi-
ans, Galatians, Philippians, 1 Thessalonians, and Philemon); the remainder
are disputed—they could be Pauline, written by a disciple of Paul, or pseu-
donymous.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 con-
sidered, 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 Mahometan-
ism 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 reli-
gion’ of Christianity in forsaking their existing religion, friends, and com-
munity. 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
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 ‘Princi-
ple 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 preju-
dice, 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, indo-

lently to say, that there are so many principles, from whence men are
liable to be deceived themselves and disposed to deceive others, espe-
cially 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).
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 conscious-
ness 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 develop-
ment of life and consciousness, (3) an inductive version of the cosmo-
logical 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 sug-
gestion 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 com-
ponents 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 can-
not find it absurd to suppose that such a being will occasionally inter-
fere 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.
Mitchell makes an important point following his reference to Mackie:
Prophecy fares no better, if it is understood as it must be for the pur-

poses 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 inde-
pendent (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 ante-
cedent assumptions about the existence and activity of God are them-
selves 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 particu-
lar 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.
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 applica-
ble 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 his-
tory 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 his-
tory, 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 per-
son in the nation, as appears from the New Testament, and is an acknowl-
edged 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 ‘Sav-
iour 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 gen-
eral 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 Testa-
ment, 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 noth-
ing 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
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 seem-

ing 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 forerun-

ner 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.
7 Jon Elster, ‘Pascal and Decision Theory,’ in Nicholas Hammond (ed.), The Cam-
bridge Companion to Pascal (Cambridge: Cambridge University Press, 2003),
53–74.
8 Elster, ‘Pascal and Decision Theory,’ 70.
9 Cruickshank, Pascal: Pensées, 35–36.
12 Krailsheimer, ‘Introduction,’ in Pascal, Pensées, xxv.
13 David Wetsel, ‘Pascal and Holy Writ,’ in Hammond (ed.), The Cambridge Com-
panion 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 Constitu-
tion 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.
[1748] 1975), 113–114.
23 John Locke, An Essay Concerning Human Understanding, edited with an intro-
duction by P. H. Nidditch (Oxford: Oxford University Press, [1689] 1979),
IV.XV.5, 656.
24 Penelhum, Butler, 94.
27 See Penlhum, Butler, 177.
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 Rob-
erts 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 con-
troversially, 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 appa-
ratus 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 atten-

tion, “That no testimony is sufficient to establish a miracle, unless the
testimony be of such a kind, that its falsehood would be more miracu-
lous, 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 else-
where 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 enti-
tled Miracles. From the observed regularity of nature Hume concludes that
there is ‘uniform experience’ against miracles. He thus concludes that mira-
cles are the most improbable occurrences and that it is always more prob-
able 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 let-
ter 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 tes-
timony 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 defini-
tion 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 tempo-
rum 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.
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. How-
ever, 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 prob-
ably 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 Chris-
tian thinkers have thought about miracles. Indeed, the modern concept of
‘laws of nature’ post-dates Augustine and Aquinas and the Bible itself. How-
ever, 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 cel-
ebrated essay on miracles from a Bayesian perspective. Much of this discus-
sion 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 argu-
ment of multiple testimony, understood either in the sense of several inde-
pendent 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
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 prob-
ability 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 funda-
mental theorem of probability formulated by the Revd Thomas Bayes and dis-
cussed 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 his-
torical 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 testi-
mony given that it did not, the miracle is said to be confirmed. The defini-
tion 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 mira-
cle 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 testi-
mony is false to be less than the prior probability of the miracle’s occurrence.
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 foot-
note: ‘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 [ T |M ] P [ M ]
P [ M|T ] =
P[T]
Or, more helpfully:
P T |M P M
P M|T (6.1)
P T |M P M P T | ~ M P ~ M
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; << is much less than; ≈ is
approximately equal to; → implies; ↔ implies and is implied by (or ‘if
and only if’).
One very straightforward consequence of equation (6.1) is that
P[T |M] > 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] << 1
P ~ M 1
With these assumptions we obtain the approximate relationship
P M
P M|T (6.2)
P M P T | ~ M
This latter expression is greater than or less than 0.5 as P[M] is greater than
or less than P[T|~M]. Our theorem can be stated succinctly thus (always
implicitly assuming that P[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
Whether or not Hume meant this, and there is some dispute among phi-
losophers 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 tes-
timony for which P[M] > P[T|~M] as ‘miraculous’ and testimony for which
P[M] < P[T|~M] as ‘non-miraculous,’ though clearly the terms are relative
to the miracle in question; in Hume’s terms the former would be ‘testimony
the falsehood of which would be more miraculous than the miracle which it
endeavours to establish.’
Independent testimonies to a single miracle

A point that seems to have been overlooked in the earlier discussion of
Owen31 and Sobel,32 but noted by Earman33 and Sorensen,34 is that of mul-
tiple testimonies to miracles. If we have several independent testimonies to
a miracle, or independent testimonies to different miracles, it can be the
case that the probability that a miracle has occurred is greater than 0.5,
even though no testimony considered individually was as miraculous as the
miracle.
Let me give an anecdote from my time in parish ministry. I imagined that
not many people would believe me if I told them that, if you threw eggs
into the air, they would bounce. One Easter I asked my church congregation
how many believed that eggs bounce when thrown into the air. Just one
or two hands went up. I then asked one boy to tell the congregation what
happened when I threw an egg into the air in his garden a couple of days
earlier. He said, ‘It bounced.’ I then asked the congregation if they believed
that eggs bounce, and this time more hands went up. I asked a girl what
happened when I threw an egg into the air in her garden. ‘It bounced,’ she
said. Yet more hands went up, and when I asked a third witness, nearly all
the hands went up. That is how evidence from testimony mounts up and can
be overwhelming.
Earman has noted that, with a sufficient number of independent wit-
nesses, one could establish any miracle as effectively certain. If Tn is
taken to mean that there are n independent witnesses, and for simplic-
ity we assume that each testifies truly with probability p = P[T|M] and
falsely with probability q = P[T|~M], then, assuming independence, we
obtain
P[T n | M]P[M]
P[M|T n ] =
P[T n | M]P[M] + P[T n |~ M]P[~ M]
P[T | M]n P[M] (6.4)

=
P[T | M]n P[M] + P[T |~ M]n P[~ M]
(utilizing the independence condition)
1 (6.5)
=
 P [ ~ M ]   q  n
 
1+ 
  
 P [M ]   p
Now even for a minimally reliable witness it will be the case that p > 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 expres-
sions 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 combin-
ing independent probabilities) decreases very rapidly, soon becoming com-
parable 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 mira-
cle’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 com-
bined 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.
One can imagine, given that there are very many reported miracles, iter-
ating 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 scepti-
cism. The wise man will not believe any testimony that a specific miracle
took place. However, this does not rule out the wise man, equally ration-
ally, 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 testi-
monies, 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 probabili-
ties 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)
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 inde-
pendence 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[T2 |M1 ∧ M2 ]P[M2 | M1 ]

P[M2 |M1 ∧ T2 ] = (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 occur-
rence 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 ]
P[M2 |M1 ∧ T2 ] = (6.11)
P[M2 |M1 ] + P[T2 | ~ M2 ]P[~ M2 |M1 ]
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 occur-
rence 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 gen-
eral, 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] << 1. From this it
follows that P[~M2|M1] = 1 − P[M2|M1] ≈ 1, and hence
P M2 |M1
P[M2 |M1 T2 ] (6.12)
P M2 |M1 P T2 | ~ M2
The value of P[M2|M1 ∧ T2] then depends on the relative magnitudes of
P[M2|M1] and P[T2|~M2], both of which are likely to be small numbers.
Just for illustrative purposes, let us take sample values as follows:
P[M2] = P[M1] = 10–6

P[T2|~M2] = P[T1|~M1] = 10–3
P[M2|M1] = 10–4
Then
P[M2|T2] = P[M1|T1] ≈ 10–3 and P[M2|M1 ∧ T2] ≈ 0.09
Hence, from equation (6.9),
P[M1 ∨ M2|T1 ∧ T2] ≈ 10−3 + 10−3 − 0.09 x 10−3 = 1.91 x 10−3
Thus, the probability that at least one miracle occurred is indeed greater
than the probability that either miracle occurred individually, contra Schles-
inger (N.B. equation (6.6) would have given 1.999 x 10−3).
Of course the above analysis can in principle be extended to more than
two independent testimonies. Noting that there are very many reports of
miracles down the ages, we suggest that it may well be the case that the
combined force of the testimony for individual miracles leads to the prob-
able occurrence of at least one miracle.
If we make the simplifying assumption that a particular miracle depends
primarily on the testimony to it, and ignore the lesser dependence on other
miracles, we can generalize equation (6.6) to n miracles as follows:
P[M1 ∨ M2 ∨ M3 … Mn |T1 ∧ T2 ∧ T3 …Tn ]

= 1 −( 1 −P[M1 |T1 ]) ( 1 −P[M2 |T2 ]) ( 1 −P[M3 |T3 ]) …( 1 −P[Mn |Tn ])
Analogously to the last section, for simplicity let us assume that all the
P[Mi|Ti] are identical and let p = P[Mi|Ti] for all i. Then
P[M1 ∨ M2 ∨ M3 … Mn |T1 ∧ T2 ∧ T3 …Tn ] = 1 −( 1 −P[Mi |Ti ])

n
= 1 −( 1 − p )
n
As n → ∞, (1 – p)n → 0. Hence P M1 M2 M3 Mn |T1 T2 T3 Tn 1.

In a similar vein to the last section, there will be a finite value of n
such that the probability that at least one miracle occurred will be greater
than ½.
In principle the analysis could also be extended so as to calculate the
probability that at least m out of n miracles occurred, and again, for any
given m, there will be a finite value of n such that this probability too is
greater than ½.
Impact of the Bayesian formulation on Hume’s argument

If Hume’s argument is to be retrieved in the light of the above, then he must
be interpreted to mean:
1 For any particular miracle and single testimony to that miracle, the
probability that the witness gives a false report is always greater than
the prior probability of the miracle’s occurrence.
2 The combined independent testimony for any single miracle cannot
yield a probability at indifference level (i.e. probability 0.5) or greater
for that miracle’s acceptance.
3 All the evidence for all miracles when combined cannot yield a prob-
ability at indifference level or greater that at least one miracle occurred.
A major charge levelled against Hume’s argument, pressed for example by

Armstrong,35 is that he is begging the question. If we are to impose upon
Hume the interpretations 1-3, then it seems to me that the force of the
charge becomes overwhelming. But can Hume possibly mean this? The fol-
lowing passage from Part II of Enquiry, X, is relevant here:
there is not to be found, in all history, any miracle attested by a suf-

ficient number of men, of such unquestioned good sense, education,
and learning, as to secure us against all delusion in themselves; of such
undoubted integrity, as to place them beyond all suspicion of any design
to deceive others; of such credit and reputation in the eyes of mankind,
as to have a great deal to lose in case of their being detected in any false-
hood; and at the same time, attesting facts, performed in such a public
manner, and in so celebrated a part of the world, as to render the detec-
tion unavoidable: All which circumstances are requisite to give us a full
assurance in the testimony of men.36
This passage does seem to consider the possibility of combined testimony

from ‘a sufficient number of men’ for a single miracle, but rules out the idea
that such combined testimony has ever existed. However, this passage does
seem to be particularly question-begging, dismissing the possibility by fiat.
Moreover it does not consider the possibility that non-miraculous testimo-
nies for different miracles might result in a high probability that at least one
of them has occurred.
As Mackie, who approves of the criteria put forward by Hume in this
passage, notes,37 it is a matter of controversy whether they have ever been
met. In fact Hume cites a number of examples of attested miracles, but does
not attempt to provide the detailed examination of their circumstances of
occurrence which would be required to establish his claim. It should be
noted that I am interpreting the above passage, and Hume’s essay in general,
as asserting only that eyewitness testimony has never in actual fact been suf-
ficient to establish the credibility of a miracle, rather than the stronger claim
that it is impossible in principle that it do so. Whilst the general tenor of
Hume’s essay would seem to indicate that he means the former, on occasions
he somewhat confusedly seems to lapse into the latter, e.g. in dismissing the
hypothetical resurrection of Queen Elizabeth I. (N.B. he does not discuss the
more pertinent example of the resurrection of Jesus.)38 By indicating how
the falsehood of a single testimony should be interpreted, and how many
testimonies should be combined, the Bayesian analysis shows the stronger
claim to be false and seriously undermines the weaker claim.
While agreeing with Hume’s general principle that one should weigh the
unlikelihood of the miracle against the unlikelihood that the witness is mis-
taken or dishonest, Mackie adds the rider that ‘the likelihood or unlikeli-
hood, the epistemic probability or improbability, is always relative to some
body of information, and may change if additional information comes
in.’39 Just so. One may have ‘non-miraculous’ evidence for a miracle which
yields a probability for the miracle’s occurrence of less than 0.5. Further
non-miraculous evidence may arise which when combined with the exist-
ing evidence raises that probability above 0.5. Alternatively, as Mackie
observes, one may start with two independent witnesses whose agreement
that a particular miracle has occurred is hard to explain unless the miracle
has indeed occurred. These points are the stronger for the realization of
how the probabilities combine. The probability that a number of reliable
independent witnesses are all mistaken or dishonest is the product of the
individual probabilities that each is wrong, and this will in general be very
small.
Now consider the third possibility suggested here, i.e. that the combined
evidence for all miracles might yield a probability greater than 0.5 for the
occurrence of at least one (even though none individually is shown to be
probable, despite the above). Hume’s argument is further undermined since
the rational acceptance of miracles would then be established. Sorensen was
correct in noting that, in order to refute all argument from miracles, Hume
needs to establish a much greater scepticism than merely showing that the
correctness of any report of a miracle is improbable.40
The above would seem to indicate that multiple testimonies pose a severe
challenge to Hume’s argument. The combined force of the points regarding
multiple testimony for a single miracle and testimony for several miracles
is to produce a strong cumulative case against Hume. But there is still the
further question which must be addressed, regarding the importance of tes-
timony to miracles—does testimony for miracles provide evidence for the
existence of God? This we now address.
Does testimony for miracles provide evidence for the
existence of God?
A number of authors have argued that testimony for miracles does indeed
provide evidence for the existence of God. Here we are using the expression
‘provides evidence for’ in the sense of ‘supports’ or ‘confirms.’ Mathemati-
cally we require P[G|T] > P[G], where P[G] is the prior probability of God’s
existence and P[G|T] is the probability that God exists given testimony T
for miracle M. Note that this is a much weaker requirement than making
the existence of God probable, when we would require P[G|T] > 0.5. Swin-
burne 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 viola-
tion 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 fol-
lows, 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 symbol ‘≳’ here represents ‘marginally greater than.’
Statement 1, and hence the step from 1 and 2 to 3, is uncontroversial, fol-
lowing 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 testi-
mony for miracles is evidence for the existence of God. He begins with
Bayes’s theorem in the form
P M|T P[G | M T ]
P G|T (6.13)
P M|G T
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 M|T P[G | M T ] P M|G ~ T

. . (6.14)
P G| ~ T P M| ~ T 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 (pro-
vides 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 argu-
ably such circumstances did obtain both for Moses and Jesus. There is an
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. Simi-
larly, 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 mira-
cles. 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 M|T P[G | M T ]
P G|T (6.15)
P M|G T
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 T |M G P[M | G]
P M|G T
P T |M G P M|G P T | ~ M G P[~ M | G]
P M|G
(6.16)
P M|G P T | ~ M
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 mira-
cle’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
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 ‘mod-
erately 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 M|G P[T |~ M]
P G|T P M|T P G|M . (6.17)
P M|G
P T |M P M 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 M|G P T | ~ M
P G|T P G. (6.18)
P M P T | ~ M
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]
1 (6.19)
P M P T | ~ M
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 testi-
mony 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
1 (6.20)
P M
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] << P[T|~M], and furthermore P[M|G] << P[T|~M],
then P[T|~M] dominates both numerator and denominator of (6.19), and T
provides virtually no evidence for G.
There is a plausible in-between case in which P[M] << P[T|~M], but
P[M|G] >> P[T|~M]. In that case the testimony for the miracle, though not
strong (in the sense of overcoming the prior improbability of the miracle),
would nevertheless provide good evidence for G because if the miracle did
occur it would very greatly enhance the probability of G. In this case we
would have
P M|G
P G|T P G. (6.21)
P T | ~ M
For example, if we had
P[M] ≈ 10−6
P[T|~M] = 10−3
P[G] = 10−8
P[M|G] = 10−1 (N.B. remember that we have assumed that P[M|G] is not
particularly high)then we should obtain
P[G|T] ≈ 10−6,
a significant enhancement over the prior probability P[G].

Earman asserts, by analogy with his analysis of multiple testimonies to
a single miracle, that the existence of God can be made probable by accu-
mulating testimony to sufficiently many miracles. This will be the case if
the occurrences of the miracles can be treated with a degree of independ-
ence, as they plausibly could be in the light of the above critique of Schles-
inger’s argument. Note that Earman has also assumed that each testimony
to a miracle is substantial evidence for the existence of God, in the light of
the multiple testimony argument for individual miracles (he assumes that
P[T|~M] << P[M] for each M so that condition (6.19) reduces to (6.20)).
On the other hand, if we have other evidence for God’s existence, we
may want to use that instead to remove our prejudice that miracles were as
unlikely as anything could be, and then to argue, from testimony to miracles
associated with the person of Jesus Christ, to the more particular claims of
Christianity. In other words, rather than ramping up further the arguments
of natural theology for God’s existence, we may want to move to ramified
natural theology in considering the evidence for particular miracles which
would in their turn be evidence for claims Christianity makes about Christ.
In Bayesian terms, if we have other evidence for the existence of God, we
can use that to argue for a higher prior probability that a miracle occurs in
the first place; we can then use the evidence for testimony to obtain a large
posterior probability for the miracle’s occurrence; and we can thus obtain a
large posterior probability for specifically Christian claims, including ulti-
mately the claim that God became incarnate in Christ, as testified by the
miracles associated with him. It is in that direction we turn in the next few
chapters.
Summary
As noted by a number of authors, Hume’s argument about miracles can be
well formulated in terms of Bayesian probability theory. The comparison
between the quantities P[M] and P[T|~M] seems to be the most useful, and
what most authors use in practice.
The possibility of multiple testimonies significantly affects Hume’s argu-
ment, both in the case of a single miracle and that of independent testimony
for many miracles. Earman’s argument that multiple testimonies for a sin-
gle miracle could make the miracle probable is supported. The argument
that independent testimonies to many miracles could show that at least one
probably occurred, put forward by Sorensen, has been challenged by Schles-
inger. The latter argued that miracles are not independent. Whilst this may
well be true, it has been shown here that this is not sufficient to invalidate
the argument. Provided miracles are not totally dependent, i.e. all stand or
fall together regardless of the weight of testimony to them, the probability
that at least one has occurred increases as the number of miracles considered
increases. In principle this could make the occurrence of at least one prob-
able, though which had occurred would remain unknown. By extension,
one could make the occurrence of at least m miracles probable by consider-
ing a high enough number n of candidates for miracles.
I suggest that these considerations amount to a strong cumulative case
against Hume’s argument.
Finally, conditions under which testimony for a miracle might provide
evidence for the existence of God were explored in somewhat more detail
than by other authors. Testimony for a miracle would provide good evi-
dence for the existence of God if either of two conditions hold. The first of
these is that the testimony is very strong, for example it represents the com-
bined testimony of a number of independent witnesses (thus overcoming
the inherent improbability of the miracle’s occurrence). The second is that,
although the testimony were still relatively weak, and did not overcome the
inherent improbability of the miracle attested to, the miracle would, if it
occurred, very significantly enhance the probability of God’s existence (to
be precise: the probability of the miracle’s occurrence given the existence of
God would have to be greater than the probability that the witness would
testify to the miracle if it had not occurred, which in turn would be greater
than the prior probability of the miracle). Either way, the accumulation of
the testimony to many miracles could in principle make the existence of God
probable, if the miracles are treated as occurring with a suitable degree of
independence.
On the other hand, we have intimated that an alternative strategy may
be to argue first to the existence of God from evidence other than miracles,
in the manner of ‘bare’ natural theology. Then one could utilize the high
probability for God’s existence derived in this way to further the project of
ramified natural theology, i.e. one could argue to the particularities of the
Christian faith from testimony to relevant miracles. We consider this pos-
sibility in more detail in the following chapters.
Notes
1 This chapter is based on my paper 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.
[1748] 1975), 115–116.
3 C. S. Lewis, Miracles: A Preliminary Study (Glasgow: Collins, Fount Paperbacks
edition, [1947] 1977), 106.
4 R. J. Berry, ‘What to Believe about Miracles,’ Nature 322 (1986), 321–322.
5 Hume, ‘Of Miracles,’ 114.
6 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), VI.13.23, 194.
7 St Augustine, Literal Meaning of Genesis, VI.13.24, 195.
8 St Augustine, The City of God (De Civitate Dei) XXI.8, in Philip Schaff (ed.),
Nicene and Post-Nicene Fathers, First Series, vol. 2 (Peabody, MA: Hendrick-
son, 1994), 459–460.
Spottiswoode; New York: McGraw Hill, 1963), 1a, 105.5.
10 Aquinas, Summa Theologiae, 1a, 105.6.
11 Aquinas, Summa Theologiae, 1a, 110.4.
12 J. L. Mackie, The Miracle of Theism: Arguments for and against the Existence
of God (Oxford, Oxford University Press, 1982), 25–26.
13 John Earman, ‘Bayes, Hume, and Miracles,’ Faith and Philosophy, 10 (1993),
293–310.
14 John Earman, Hume’s Abject Failure: The Argument Against Miracles (New
York and Oxford: Oxford University Press, 2000).
15 Roy A. Sorensen, ‘Hume’s Scepticism about Miracles,’ Analysis, 43 (1983), 60.
16 George N. Schlesinger, ‘Miracles and Probabilities,’ Noûs, 21 (1987), 219–232.
versity Press, 2004), ch. 12.
18 Richard Otte, ‘Schlesinger and Miracles,’ Faith and Philosophy, 10 (1993),
93–98.
19 David Owen, ‘Hume Versus Price on Miracles and Prior Probabilities,’ in Rich-
ard Swinburne (ed.), Miracles (London: Macmillan, 1989); reprinted from David
Owen, The Philosophical Quarterly, 37 (1987), 187–202.
20 Hume, ‘Of Miracles,’ 114.
21 Hume, ‘Of Miracles,’ fn., 115.
22 Earman, ‘Bayes, Hume, and Miracles,’ 296, 302–306.
23 Holder, ‘Hume on Miracles,’ 51–52.
24 Schlesinger, ‘Miracles and Probabilities,’ 224.
25 Owen, ‘Hume Versus Price on Miracles and Prior Probabilities.’
26 Philip Dawid, and Donald Gillies, ‘A Bayesian Analysis of Hume’s Argument
Concerning Miracles,’ Philosophical Quarterly 39 (1989), 57–65.
27 Schlesinger, ‘Miracles and Probabilities.’
28 Jordan Howard Sobel, ‘On the Evidence of Testimony for Miracles: A Bayesian
Interpretation of David Hume’s Analysis,’ Philosophical Quarterly 37 (1987),
167–186; and Jordan Howard Sobel, ‘Hume’s Theorem on Testimony Sufficient
to Establish a Miracle,’ Philosophical Quarterly 41 (1991), 229–237.
29 Peter Millican, ‘ “Hume’s Theorem” Concerning Miracles,’ Philosophical Quar-
terly, 43 (1993), 489–495.
30 Earman, ‘Bayes, Hume, and Miracles.’
31 Owen, ‘Hume Versus Price on Miracles and Prior Probabilities.’
32 Sobel, ‘On the Evidence of Testimony for Miracles’ and Sobel, ‘Hume’s Theorem
on Testimony Sufficient to Establish a Miracle.’
33 Earman, ‘Bayes, Hume, and Miracles’ and Earman, Hume’s Abject Failure.
34 Sorensen, ‘Hume’s Scepticism about Miracles.’
35 Benjamin F. Armstrong, ‘Hume on Miracles: Begging-the-Question Against
Believers,’ History of Philosophy Quarterly 9 (1992), 319–328.
36 Hume, ‘Of Miracles,’ 116–117.
38 Slupik argues that Hume’s essay can be read self-consistently as making only the
weaker claim: Chris Slupik, ‘A New Interpretation of Hume’s “Of Miracles”,’
Religious Studies 31 (1995), 517–536.
40 Sorensen, ‘Hume’s Scepticism about Miracles.’
41 Swinburne, Existence of God, 234.
42 Schlesinger, ‘Miracles and Probabilities.’
43 Otte, ‘Schlesinger and Miracles.’
44 Schlesinger, ‘Miracles and Probabilities,’ 232.
7 Ramified natural theology
and evidence for Christian
claims about Christ
Jesus’ miracles
Introduction
In this chapter and the next four chapters I shall discuss and evaluate the
evidential force of Biblical material which bears on the divinity of Christ.
This will include matters to do with Jesus’ life and teaching, including the
miracles he is said to have worked; prophecies in the Old Testament about
the Messiah, apparently fulfilled in Christ; the reasons for his death and
the meaning attributed to it; and his purported resurrection. I shall steer
a course here between a naïve literal acceptance of all the Bible says and a
total scepticism, taking account of historical-critical study, but essentially
providing good reasons for the overall integrity, in particular of the gospels.
There will also be some reference to Paul’s epistles. Regarding the issue of
religious pluralism, which I highlighted in Chapter 4, I shall not be able to
provide a comprehensive comparison of truth claims. However, confining
myself to the Abrahamic faiths, I shall give some further brief discussion of
what is said in the Qur’an about Christ, and also allude to certain extra-
Biblical Jewish sources.
The present chapter will focus particularly on the miracles attributed to
Jesus. It will result in material to be combined with the evidence for the
fulfilment of prophecy and for the resurrection, discussed in the next three
chapters, resulting in a cumulative argument for Christ as God incarnate in
the final chapter.
Historical criticism and the integrity of the gospel accounts

If we take the New Testament at face value then clearly Jesus worked many
miracles. However, before we can be definitive we need to take a look at his-
torical criticism of the Bible. This began after the time of Pascal and Butler,
whom we considered in Chapter 5, and thus needs to be taken into account
in a modern day assessment—although of course we cannot do justice to
this vast field in such a short space. Thus I shall describe some of the major
figures involved but inevitably others will be ignored. I aim at least to dis-
cuss a representative selection of scholars.
Ramified natural theology and Christ 129
Our approach to the study of the Biblical texts could be likened to the
philosophical position of ‘critical realism’ adopted towards science by such
figures as John Polkinghorne and Alister McGrath whom we met in Chap-
ter 2. In the field of historical study of the New Testament, this position
has been espoused by such major scholars as James Dunn and Tom Wright,
whom I refer to in this chapter. It is a mediating position between naïve real-
ism and idealism, so neither, as in positivism, overemphasizes the objectivity
of what is known nor, conversely, sees knowledge as purely subjective or
relative.1 Wright’s definition is instructive:
Over against both of these positions, I propose a form of critical realism.

This is a way of describing the process of “knowing” that acknowledges
the reality of the thing known, as something other than the knower
(hence “realism”), while also fully acknowledging that the only access
we have to this reality lies along the spiralling path of appropriate dia-
logue or conversation between the knower and the thing known (hence
“critical”).2
The historical-critical method ostensibly treats the Bible in the same way as
other ancient literature. However, it has often come to Biblical study with
pre-conceived notions of what can or cannot happen in history. Thus it has
often approached the Biblical text with a ‘hermeneutic of suspicion,’3 in con-
trast to Richard Swinburne’s ‘Principle of Testimony,’ which we shall meet
in Chapter 9. For the former attitude the starting point is one of scepticism,
whereas for the latter it is that one ought normally to believe what others
tell us, while acknowledging human fallibility. Not surprisingly, then, when
historical research on the Bible was actually done in the eighteenth and
nineteenth centuries, its import was largely negative, and there has been a
strand of that kind of sceptical approach within Biblical scholarship up to
the present day. In Chapter 4 we met Ernst Troeltsch and his principle of
analogy, whereby miracles are ruled out by fiat, and we shall see shortly that
Troeltsch was far from alone in his view. However, we saw in Chapter 6 in
critiquing David Hume’s similar stance that belief in miracles can in princi-
ple be justified given enough reliable testimony.
The rise of the historical-critical method, which could be described as
the scientific study of the Bible, began in Germany in the late eighteenth
and early nineteenth centuries. Many of the early historical-critical schol-
ars were imbued with the Enlightenment elevation of reason to deriving all
truth, but often with a particular view of reason, e.g. that truths can only be
known through rational thought and not through history. Thus Gotthold
Ephraim Lessing asserted that there was an ‘ugly broad ditch’ between the
necessary truths of reason and the contingent truths of history.
The so-called ‘Quest of the Historical Jesus’ began with H. S. Reima-
rus. Like Lessing and Hume, Reimarus denied that human testimony to
a past event was sufficient to make it credible if that event appeared to
130 Ramified natural theology and Christ
be contradicted by present-day experience. As we saw in Chapter 4, that
idea was elevated to the ‘principle of analogy’ by Troeltsch and others. The
idea of the Quest was to get back to the simple Galilean teacher whom the
gospels and the church had elaborated into a supernatural redeemer. In this
programme all miracles, including the resurrection, were to be rejected.
There is already a tension here, between pure rationalism, that reason
alone can deliver all truths (history can at best confirm the truths of reason
for Lessing), and the empiricism of Hume and others whereby it is experi-
ence which gives us truth. However, as we have seen, Hume likewise rejected
miracles, if on the rather different, but nevertheless a priori, grounds that
any experience would be fraudulent or mistaken. Oddly enough, the scien-
tific laws of nature, which are contingent, could not give truth according
to Lessing; and, on the other hand, they could not be corrected if Hume
followed his own precepts logically and disallowed exceptions to what is
currently believed to be the sequence of cause and effect.
Reimarus’ view was that the New Testament writers were frauds and the
gospels full of contradictions. Not only were miracles rejected as ahistori-
cal but so was the very idea of revelation. The Biblical writers effectively
invented a ‘Christ of faith’ as a substitute for the ‘Jesus of history’ (these
terms were first introduced by Reimarus) who died a failure with his expec-
tations of Messiahship in tatters, as supposedly evidenced by his cry of der-
eliction on the cross. Reimarus’ work gained notoriety in Germany when
published anonymously through Lessing between 1774 and 1778 as the last
of the so-called Wolfenbüttel Fragments.
Another famous landmark was the publication of David Friedrich Strauss’s
Life of Jesus Critically Examined in 1835.4 Strauss regarded the Gospels as
so imbued with myth (the Fourth Gospel above all) that it is virtually impos-
sible to recover any historical core of the life of Jesus at all. The miracles
that Jesus performs and the ontological claims about his person made in the
tradition—that he is God-man—are ruled out of court. The religious genius,
or God-consciousness of Jesus, was deemed, at least in the first edition of
Life of Jesus, not to be unique. In England, George Eliot’s 1846 translation
of Strauss’s Life met with fierce criticism, and even Eliot herself was repelled
by Strauss’s dissection of the ‘beautiful story of the Resurrection.’5
In a new version of Life of Jesus published in 1864 Strauss pointed to an
ideal of moral perfection and God-consciousness exemplified by Christ, but
even then this is something that humanity as a whole can develop further.
Later still, Jesus was described as a ‘spiritual fanatic.’
In the twentieth century Karl Barth presented a number of challenges to
Strauss. For example, he wrote as follows:
Is it not a fact that the New Testament records are useless as “sources”
of a pragmatically comprehensible picture of a man and of a life? For it
is from the very first word that they seek to be something quite different,
namely testimonies to a “superhuman being,” corresponding feature by
feature to the prophecies of the Old Testament, a being whose image
must defy all historical reconstruction.
Is it not a fact that a “historical Jesus” established behind the so-
called sources, and therefore quite independently of the witness of the
New Testament, can only be comprehended as such if we remove those
predicates of his which are essential to this witness: his consciousness
of himself as the Messiah of Israel and as God’s eternal Son, his proc-
lamation of the kingdom of God and expectation that he would come
again, and his resurrection from the dead? Is it not a fact that the sen-
timental, moralizing description of character which is indispensable to
the establishment of this figure has nothing at all to do with the faith of
the Apostles?6
Furthermore, the goal of historical research can only be a relative Christ, a

Christ ‘who could on no account be the Word that became flesh, executing
God’s judgment upon us and challenging us ourselves to make a decision.’
Barth says that Strauss did for history what Ludwig Feuerbach had done
for psychology in arguing that God is a projection from our minds. His
verdict is scathing: ‘Proper theology begins just at the point where the dif-
ficulties disclosed by Strauss and Feuerbach are seen and then laughed at.’7
On the other hand, rather than being laughed at, perhaps the Enlighten-
ment analysis should be challenged on its own grounds of rationality. It
seems to me that the fault is glaring: if one starts from naturalistic presup-
positions, naturalistic conclusions follow. This is precisely what Feuerbach,
Strauss, Troeltsch, and others did.
Thus it may well be that the negative import of this historical-critical
research was a result of the methods employed, and, contra Lessing, it may
also be the case that certain historical events can have universal validity.
Generally, following the Enlightenment, historical Jesus research made no
allowance for the supernatural. This was the case with Strauss, who indeed
is criticized by Stephen Neill and Tom Wright precisely because he did not
actually engage in historical study but went to the text with his preconceived
notion that all must be myth.8 It was also the case with Troeltsch, and his
principle of analogy. Troeltsch ruled out miraculous events in Biblical times
on the grounds that there was no analogy with present events. Such meth-
ods are bound to lead to a non-supernatural Jesus because they start from
premises that rule out the supernatural. But this need not be so, and more
open-minded premises might well lead to other conclusions, as is evidenced
by more recent trends in historical Jesus research. Moreover, as I argued in
Chapter 6, evidence from testimony can in principle make it probable that a
miraculous event occurred, and that Hume’s famous argument against mira-
cles, from which much of the subsequent anti-miraculous prejudice came, is
fundamentally flawed.
Like Strauss, F. C. Baur, founder of the Tübingen school, also rejected
the historicity of John’s gospel. Baur’s presuppositions also led him astray,
in large part owing to the influence of G. W. F. Hegel. Hegel proposed that
movements begin with a thesis, but as they develop opposition forms in
terms of an antithesis. Finally, the two divisions come together in a synthe-
sis. In the case of the development of Christianity there was Judaic (Petrine)
Christianity, Pauline Christianity, and finally their reconciliation in the
Catholic church. This reconciliation did not take place until at least the mid-
dle of the second century. John’s gospel represents, not history, but the idea
of a church composed equally of Jew and Gentile. It must therefore be of
very late composition. In Mark there is almost no opposition between Jew
and Gentile so that too must date to not earlier than the mid second century.
Matthew is more Jewish and Luke more Gentile so Matthew is earlier and
to be preferred, but Mark is later and depends on both. Even so, Matthew
is largely mythical in character. Baur’s order of composition of the synop-
tic gospels is in direct conflict with the modern near consensus of Marcan
priority with Mark dated approximately to the late AD 60s or early 70s.9
These developments in Germany became known in England as a result of
the publication in 1860 of Essays and Reviews. This volume caused vastly
greater consternation than the publication the year before of Darwin’s Ori-
gin of Species, whose much readier assimilation we discussed in Chapter 2.
Indeed, historian Ieuan Ellis describes Essays and Reviews as touching off
‘the greatest religious crisis of the Victorian age.’10 It was condemned by
both Houses of the Convocation of Canterbury. This proceeding inevitably
only fuelled the sales which rose to a level twenty times that of Origin. As
Neill and Wright note, the situation was grave, giving rise to near panic,
since if the Christian faith is just an idea, and not solidly grounded in his-
tory, then it is very different from the faith that had been believed by the
church for the first 1800 years.11
A milder version of Jesus’ life was published anonymously in England by
J. R. Seeley under the title Ecce Homo. This book focussed on Jesus solely
as a man and moral philosopher, bypassing issues of theology, miracle, and
the identity of Jesus. It too attracted large sales, aided by the anonymity of its
author, rather paralleling in this respect Robert Chambers’ early work on evo-
lution, Vestiges of the Natural History of Creation, published in 1844. It also
drew condemnation. The Earl of Shaftesbury rather graphically described
Ecce Homo as ‘the most pestilential book ever vomited from the jaws of
hell’!12 This seems an exaggeration (and indeed, Shaftesbury came to regret
it) since Seeley did not deny Jesus’ divinity or that Jesus performed miracles;
rather, through the brief references he made, one is led to believe that he was
orthodox on these matters and was just reserving discussion to a later date.
There is a curious modern echo of the criticism which Seeley received.
Tom Wright reports that some reviewers of his Jesus and the Victory of God
accused him of not being interested in, or not believing in, Jesus’ resurrec-
tion, because he excluded this topic from that volume. The irony is that this
was only because there was too much to say, and his subsequent book on
the resurrection ran to over 800 pages.13
It seems to me that, just as Darwin’s theory of natural selection led not to
a refutation of design but to a reworking of natural theology (witness Fred-
erick Temple whom we discussed in Chapter 2, and who also contributed to
Essays and Reviews), so also the project of ramified natural theology needs
to engage with historical criticism of the Bible. Indeed, just as what we learn
from the natural sciences impacts natural theology by enhancing our picture
of God the Creator, so serious engagement with historical science can posi-
tively impact our understanding of God the Redeemer in Christ and thus
ramified natural theology—but more of that in a moment.
Once it had arrived on the scene, the critical method could not just be dis-
missed (as in Shaftesbury’s and others’ cavalier responses), and the Church
could not simply retreat to more naïve views of the Bible’s inspiration. Quite
the opposite: historical research could be done better. And help was at hand.
Baur’s elaborate theories were comprehensively demolished by Cambridge
Biblical scholar J. B. Lightfoot. In particular Lightfoot did remarkable tex-
tual work on the Epistles of the Apostolic Fathers Ignatius of Antioch and
Clement of Rome. Lightfoot established which of the writings of these that
had come down to us were authentic and dated them fairly precisely, Igna-
tius to the years leading up to his martyrdom in AD 115 and Clement to
about AD 96. These letters showed no knowledge of any lasting conflict
between Peter and Paul, and made many allusions to earlier Christian writ-
ings, including the gospels and St Paul, even if there are few direct quota-
tions: the very late date for these is therefore falsified.14
Adolf von Harnack in Germany was deeply impressed by Lightfoot’s
work. He also came to some surprisingly conservative conclusions about the
New Testament writings. Thus in Luke the Physician he identified Luke as
the companion of Paul and the author of both Luke and Acts, with the latter
being written during Paul’s captivity in Rome. He upheld early dates for all
three synoptic gospels and an especially early date for ‘Q,’ the putative com-
mon source, with Mark, for Matthew and Luke. Nevertheless Harnack’s
theology was radical. Thus he asserts that the Gospel has been distorted by
Hellenistic culture. For Harnack, Jesus is Christianity, but this statement
implies no doctrine about Jesus. A Christology is no part of the proclama-
tion of Jesus of Nazareth, and his own message is not self-referring.15
Alister McGrath writes:
If it can be shown, on historical grounds, that some aspect of Christian

doctrine does not naturally and legitimately arise from Scripture, it may
reasonably be regarded as of questionable validity. For Harnack, the
doctrine of the incarnation was a case in point: he regards it as being
the direct result of Greek metaphysics, with no place in the gospel itself.
This assertion of Harnack is highly questionable, as McGrath himself notes.

Surely the patristic writers were grappling with Scripture in formulating
their doctrines, and Scripture includes much that is incarnational, e.g.
John 1:1 and Philippians 2:6ff to name just two quite distinct strands of
thought in Scripture. Moreover, incarnation was not such an easy concept
for the Greeks as von Harnack implies, but actually ran counter to their
presuppositions about the transcendence of God.
For Harnack the gospel is not about Christology but soteriology, and con-
cerns only God the Father and how a person might find spiritual strength,
joy, and peace by finding him. The gospel has been encumbered with a husk
of accretions and one needs to peel these away to get to the kernel, and then
the task is to re-embed the gospel in our own culture. The scriptures give
the main features of Jesus’ teaching, they describe how his life issued in the
service of his vocation, and they describe the impression he made on his
disciples.16
Harnack sees Christianity taking many forms over the centuries—the
gospel as proclaimed by Jesus himself, by Paul, Augustine, Luther—none
of which can be absolutized. What he aims to do is to embed the gospel in
the culture of our time. Nineteenth century Liberal Protestantism produced
a variety of Christologies. In 1925 Harnack remarked that the irreducible
core was that Jesus was the ‘mirror of God’s fatherly heart.’ Catholic mod-
ernist George Tyrell’s devastating verdict on Harnack’s picture of Jesus was
stated thus: ‘The Christ that Harnack sees, looking back through nineteen
centuries of catholic darkness, is only a reflection of a liberal Protestant face,
seen at the bottom of a deep well.’17 This assessment could equally well be
applied to most of the other figures we have considered.
What I have been describing hitherto has gone under the rubric ‘The
Quest of the Historical Jesus’ and culminated in a devastating critique by
Albert Schweitzer in his 1906 book of that title.18 The purely ethical Liberal
Protestant Jesus was erroneous as it completely ignored the eschatologi-
cal dimension of the gospels. But for Schweitzer that meant that Jesus was
a fanatical eschatological prophet announcing the immediate coming of
God’s Kingdom which never happened. Hence, Jesus ‘threw himself on the
wheel’19 and, in light of the, to us, alien genre of apocalypticism, he ‘comes
to us as One unknown.’20
One major reaction to Schweitzer came from Karl Barth who stressed
the apocalyptic in Jesus’ preaching in order to emphasize his otherness, in
contrast to the bourgeois Jesus of Liberal Protestantism. We noted Barth’s
reaction to Strauss above. His rejection of natural theology, which we dis-
cussed briefly in Chapter 2, was motivated chiefly by his attitude toward
Liberal Protestantism.
A new development in the twentieth century was ‘form criticism,’ pio-
neered by Rudolf Bultmann and associates. Form criticism examined the
oral tradition of the church, the way in which the form of the message of
the apostles was passed on and what it meant for the life of the community.
The form critics supposed that there was a long period of this oral tradition
before the gospels were written down and during this period huge modifi-
cations took place to fit the developing life and thought of the church. The
gospels could not be treated as historical or biographical and only acciden-
tally contained any material going back to the man Jesus of Nazareth. They
were concerned solely with the Christ of faith.
One reason for the rejection of the gospels as biographical was that they
did not give the kind of detail a modern biography would—details such as
chronology and the appearance and development of the personality of the
character portrayed. Graham Stanton has pointed out, however, that they do
conform to ancient standards of biographical writing, with a focus on deeds
and words of the subject.21 Stanton also notes arguments against wholesale
changes made to the tradition through oral transmission. For example, he
cites Jesus’ restraint in expressing his identity, in contrast to the post-Easter
disciples’ loud proclamation of it. He also cites the retention of some details
embarrassing to that proclamation, such as Jesus’ anger in Mark 1:41,
contained in a few early manuscripts but omitted in the majority and also
omitted in Matthew and Luke.22 Morna Hooker accepts the priority of the
manuscripts in which Jesus is described as ‘moved with anger’ (ὀργίσθεις)
and their later amendment to ‘moved with compassion’ (σπλαγχνισθεὶς) pre-
cisely because a scribe is more likely to amend a manuscript in that direction
than the other way round.23 Another instance cited by Stanton is the accusa-
tion in Mark 3:21 that Jesus was mad, this being unparalleled in Matthew
and Luke.24 Moreover, he cites some scholars such as B. Gerhardsson whose
researches into Rabbinic Judaism highlight techniques by which traditions
are passed on as ‘memorized text.’25
For Bultmann, however, what he sees as the mythical elements of the New
Testament need to be interpreted for the twentieth century scientific world.
Influenced by Martin Heidegger, Bultmann embarks on a demythologiza-
tion programme whereby the New Testament is reinterpreted existentially.
The eschatology of Jesus is now to be interpreted as concerning the now
of human existence. Utilizing Heidegger’s language, Bultmann argues that
we are challenged to ‘authentic existence.’ New Testament unbelieving and
unredeemed existence is ‘inauthentic existence’; believing and redeemed
existence is ‘authentic existence.’ These ideas float free of any historical
connection to Jesus of Nazareth, and the New Testament does not give us
propositions about the world. As McGrath rightly points out, ‘Bultmann
thus substitutes a twentieth-century myth for its first-century equivalent.’26
A contemporary critic was Dietrich Bonhoeffer. He writes that the reduc-
tionist programme of Bultmann to drop the ‘mythological’ elements and
reduce Christianity to its ‘essence’ is quite mistaken. On the contrary, echo-
ing Barth, he writes: ‘My view is that the full content, including the “mytho-
logical” concepts, must be kept—the New Testament is not a mythological
clothing of a universal truth; this mythology (resurrection etc.) is the thing
itself. . . .’27
More recent followers of Bultmann, though differing from him in various
ways, include representatives of the so-called ‘Jesus Seminar.’ This is a group
of North American scholars who meet together to vote on the authenticity
of sayings of Jesus, with an elaborate scoring system to summarize their
differing judgments. They make some highly controversial premises which
are often confused with conclusions, e.g. that oral tradition is fluid and that
Jesus actually spoke very few of the sayings attributed to him.28 They regard
Mark (considered by most scholars to be the first gospel to be written—
see below) as almost entirely fictitious. Burton Mack sees apocalyptic and
Christological references to Jesus as Son of God or Son of Man as inven-
tions which do not go back to the real Jesus. Instead Jesus is something like
a Cynic sage. It is as if there were, in Gerd Theissen’s memorable phrase, a
‘Committee for Misleading Later Historians’ in first century Palestine.29 And
Tom Wright’s devastating verdict is that this programme is a myth itself,
invented to combat strands of American Christian culture obsessed with
apocalypticism. It builds castles in the air, having no supporting foundation
of evidence but constitutes merely ‘an exercise in creative imagination.’30
Dominic Crossan, perhaps the most notable member of the Jesus Seminar
and one to whom Wright is more sympathetic, sets out an ‘inventory of the
Jesus tradition,’ in which several apocryphal works, such as the first edition
of the Gospel of Thomas, are held to pre-date the canonical gospels. Wright
describes this move as ‘extremely shaky’ and contradicting almost all of the
Jesus research of the last 250 years.31 Like Mack, Crossan sees Jesus as a
Cynic with a social message and rejects apocalyptic. Interestingly, however,
this radical scholar does recognize that some of Jesus’ healing miracles are
historical. Thus he states, classifying examples according to his inventory:
Jesus was both an exorcist and a healer: I take 121 Beelzebul Con-
troversy [1/2], 110 A Leper Cured [1/2], 127 Sickness and Sin [1/2],
and 129 A Blind Man Healed [1/2] as not only typically but actually
historical. His vision of the Kingdom was but an ecstatic dream with-
out immediate social repercussions were it not for those exorcisms and
healings. Those latter were what the Kingdom looked like at the level
of political reality.32
A major response to Schweitzer, Bultmann and their followers is the inter-

pretative approach of Wolfhart Pannenberg. Pannenberg does not reject
apocalyptic but uses the apocalyptic content of Jesus’ preaching to interpret
his life and especially his resurrection. We met Pannenberg in chapters 1
and 4 when discussing the motivation for ramified natural theology, and
shall return to him again in Chapter 10 when discussing the resurrection
specifically. For now, let us remind ourselves that, for Pannenberg, Christian
theology is based on an analysis of publicly accessible history. In opposi-
tion to Lessing and other Enlightenment thinkers, he argues that revela-
tion is a universal historical event recognized and interpreted as an ‘act of
God.’ Whereas Bultmann sought to demythologize the New Testament, the
mythological elements are the hermeneutical key for Pannenberg. To read
the New Testament correctly we require an understanding of first century
apocalyptic, not twentieth century existentialism—surely a correct insight.
Importantly for Pannenberg, revelation is not completely apprehended at
the beginning, but only at the end of history. Nevertheless, that revelation
of God is publicly accessible. In addition, the Christ-event is not revealing
in isolation, but in the context of the history of God’s dealings with Israel.
Whereas the Enlightenment imposed an alien interpretative framework (i.e.
Enlightenment rationalism) on the history of Jesus, the only valid frame-
work is that of apocalyptic. The principle of analogy can be no more than
a working tool. As such it can validate the cross, since crucifixions were
common in the first century. But it cannot be used as a universal rule and
exclude unique events—and Pannenberg has the resurrection especially in
mind—a priori. For Pannenberg, the resurrection could be interpreted by a
Jew as confirming the pre-Easter activity of God, which is what we are more
concerned about in the present chapter.
We have seen that the negative results of historical-critical study are cor-
related with the naturalistic presuppositions which are assumed as a start-
ing point. We have also begun to see that more rigorous historical study
without such presuppositions can give quite different results. Nevertheless,
in the first half of the twentieth century theologians often followed Reima-
rus and others in separating the Jesus of history from the Christ of faith,
and so made faith immune from historical criticism and particularity. We
have begun to see, however, that in the latter half of the twentieth century
a return of interest in God’s revelation in history, as the shackles of the
Enlightenment are cast off. In Germany, Pannenberg was an important pio-
neer with his Revelation as History published in 1961.33
From the 1970s a much more optimistic view of the historicity of the
gospels has evolved generally, with many notable scholars involved.34 This
has often been tied to placing Jesus in his Jewish context. Interestingly, Brit-
ish scholars have often been at the forefront of resistance to the radical
scepticism of the German school. For example, Anthony Harvey was fully
aware of German critical scholarship. However, the constraints imposed
politically, culturally, and religiously on Jesus’ ministry, and his ability to
challenge the practices of his day, made real sense in the historical context
of the time and rendered the New Testament accounts much more reliable
than the Germans had given them credit for.35 There are many other schol-
ars taking such a more positive line, some of whom we have already met,
such as Pannenberg and Tom Wright, and we shall have more to say about
them in later chapters.
Dating of the gospels

Now let us look briefly at what is the near consensus today about the com-
position and dating of the gospels, from a close examination of the text.
The first three gospels (the ‘synoptics’) are clearly related, since they have
much material in common, although the same stories are often repeated
with variation. By placing these gospels side by side we can glean several
facts, from which some natural inferences can be drawn. First, virtually the
whole of Mark is repeated in Matthew and Luke. Secondly, there is a further
body of material, largely sayings of Jesus, which Matthew and Luke have in
common, but which are not found in Mark. Thirdly, each of Matthew and
Luke has a certain amount of material unique to himself. Careful study of
the ordering of events in these three gospels, and the variations in accounts
of the same events, has led to the near consensus of scholarly opinion that
Mark is the first of the gospels to be written, and that, in addition to Mark,
Matthew and Luke had a written collection of Jesus’ sayings, known as ‘Q,’
to hand when they were writing (‘Q’ stands for the German word Quelle
meaning ‘source’). Mark is generally put at being written a few years either
side of AD 70 with Matthew and Luke at about AD 80–90. Through ‘Q,’
however, we may well be getting back to material written down for church
instruction as early as AD 50.
One of the most significant books of New Testament scholarship in
recent years is Jesus and the Eyewitnesses by Richard Bauckham.36 From
a very detailed and meticulous study of the background, including such
features as the names of the disciples compared with the contemporary
frequency of such names, Bauckham concludes that the synoptic gospels
report eyewitness testimony, and even (as Bauckham admits, unfashion-
ably) that John is eyewitness testimony. He agrees with Vincent Taylor
writing in the 1930s, that if the form critics were right, ‘the disciples must
have been translated to heaven immediately after the Resurrection.’ On
the contrary, eyewitnesses continued to be around for many years preach-
ing and teaching and available to those who sought information. Bauck-
ham cites another New Testament scholar, Martin Hengel, to the effect
that, contrary to the form-critics’ assertions, the tradition about Jesus did
not circulate anonymously but was related to the memory and preaching
of particular individuals. This Bauckham views as ‘historically very prob-
able.’37 Bauckham states this:
The Gospels were written within living memory of the events they
recount. Mark’s Gospel was written well within the lifetime of many of
the eyewitnesses, while the other three canonical Gospels were written
in the period when living eyewitnesses were becoming scarce, exactly
at the point in time when their testimony would perish with them were
it not put into writing. This is a highly significant fact, entailed not by
unusually early datings of the Gospels but by the generally accepted
ones.38
This point about the eyewitnesses still being present in the early Church as
authoritative sources of their traditions at the time the gospels were written
enables us to affirm with a high degree of probability the essential reliability
of the gospel records.
Jesus’ mighty acts
If we go to the text of the gospels without the naturalistic prejudice of many
historical-critical scholars it is hard to escape the conclusion that Jesus did
indeed perform ‘mighty acts.’ Certainly there are numerous examples given
in the gospels of miraculous healings and a smattering of so-called ‘nature
miracles.’ Graham Stanton points out that people in the first century were
not gullible or credulous as they are sometimes portrayed. That is in line
with C. S. Lewis’s comment, repeated by Tom Wright, that Joseph was not
disturbed by Mary’s pregnancy because he did not know where babies come
from, but because he did.39
A difference between the gospel miracles and others in ancient literature is
how restrained those of the gospels are. Stanton notes the more magical type
of miracle recorded in the apocryphal Infancy Gospel of Thomas, in which
the child Jesus makes birds out of clay and brings them to life, and instantly
reassembles Mary’s broken water pitcher40—indeed, this ‘gospel’ attributes
many very bizarre miracles to Jesus, often quite out of character to the Jesus
we know from the authentic first century gospels. The Infancy Gospel of
Thomas is generally dated mid to late second century and stems from the
Gnostic heresy. Curiously the Qur’an records Jesus as performing miracles
but includes among them Jesus making a bird out of clay, which would seem
to indicate Mohammed’s acquaintance with the apocryphal material (Sûrah
V.110). It should, however, be noted positively that the same Sûrah men-
tions canonical miracles such as the healing of a man born blind and of a
leper, and raising of the dead. Interestingly too, Mohammed himself did not
claim the power to perform miracles. Geoffrey Parrinder writes:
Isḥāq Ḥusaini says, “the Qur’ān enumerated the miracles of Christ

which were signs of his prophecy . . . It is noteworthy that Muhammad
did not attribute miracles to himself. The Qur’ān is his only miracle . . .
All these miracles Christ produced by the will of God in order to con-
vince those who doubted his mission.”41
Mark opens his gospel with a succession of healing miracles, including that
of Peter’s mother-in-law. These acts of Jesus are at one with his teaching.
Mark’s comment is that Jesus acts ‘with authority.’ Matthew presents the
miracles in a similar way to Mark but is keen to see them as the fulfilment
of Old Testament prophecy. Luke portrays Jesus himself as a prophet (espe-
cially in the programmatic passage in Lk 4), and thus aligns Jesus’ miracles
with those of Elijah and Elisha. In John’s gospel the miracles are signs, mani-
festing Jesus’ glory and his relationship to the Father and leading people to
faith.
Both Stanton42 and Wright see the miracles such as the blind receiving
their sight, the lame walking and the poor having good news preached to
them as heralding the coming age of God’s kingly rule prophesied in such
passages as Isaiah 35:5–6 and 61:1ff. For Wright this is characteristically
expressed as ‘the coming of Israel’s god to save and to heal,’43 and ‘the resto-
ration to membership in Israel of those who, through sickness or whatever,
had been excluded as ritually unclean.’44 Wright also notes that they are
seamlessly integrated into the gospel narratives as a whole, which would
most likely not be the case if they were added to the tradition by those
‘interested in telling stories of a Hellenistic-style wonder-working figure.’45
Indeed, far from resembling such figures, ‘it is far more likely that, as a mat-
ter of history, “the miraculous activity of Jesus conforms to no known pat-
tern.” ’46 Apart from occasional exceptions, Jesus’ practice ‘seems to have
been distinct from that of his contemporaries, both Jewish and pagan.’47
Whilst noting that Jesus’ mighty works are not offered as proof of his
divinity (and indeed Jesus refuses to perform miracles for that purpose (e.g.
Mk 8:11–13), Wright states: ‘More thoroughgoing recent history has been
coming to the conclusion that we can only explain the evidence before us
if we reckon that Jesus did indeed perform deeds for which there was at
the time, and may well be still, no obvious ‘naturalistic’ explanation . . . ’48
Indeed, as noted above, it is the case that since the 1970s a much more
positive assessment of the reliability of the gospels as history has been
developing.
Wright argues that if one had a preconceived notion that miracles do not
occur then one might still be able to come up with psychosomatic explana-
tions for the healings and mythical explanations for the nature miracles.
‘But,’ he goes on, ’we must be clear that Jesus’ contemporaries, both those
who became his followers and those who were determined not to become
his followers, certainly regarded him as possessed of remarkable powers.
The church did not invent the charge that Jesus was in league with Beelzebul
[Mk 3:20–30 and parallels]; but charges like that are not advanced unless
they are needed as an explanation for some quite remarkable phenomena.’49
The evidence points to Jesus having performed the mighty acts described,
and that is what brings forth these charges.50 Wright notes that most schol-
ars, from widely differing backgrounds, now accept this.51 At the very least
Jesus’ contemporaries, both friends and foes, believed he did perform mira-
cles, and this leads Wright to use the language of best explanation: ‘the best
and simplest explanation of this is that it was more or less true.’52
Christopher Rowland makes similar comments to Wright:
To eliminate the miraculous evidence in the Gospels merely because it

offends our modern understanding of the way in which the universe
functions, is to be like the interpreters of the last century, who supposed
that all miraculous material was to be consigned to the mythological
embroidery of the early Church; it is to ignore the weight of the New
Testament evidence, not to mention that from an abundance of secular
sources.53
Rowland makes the point that this does not necessarily imply a naïve
acceptance of every recorded miracle, and that some (e.g. the coin in the
fish’s mouth in Matt 17:24ff) may be legendary accretions. Nevertheless,
like Wright, he sees that ‘the miraculous element in the Gospels is an impor-
tant element in the case made for Jesus’ proclamation of the nearness of the
kingdom. Without it there would have been little doubt that his claim would
have seemed an empty one.’54
Rowland notes that Jesus’ enemies did not question his miracle-working
and cites B. Sanhedrin 43a in support. As we noted in Chapter 4 this con-
tains the accusation that Jesus practised sorcery:
On the eve of the Passover Yeshu [Jesus] was hanged. For forty days
before the execution took place, a herald went forth and cried, “He is
going forth to be stoned because he has practised sorcery and enticed
Israel to apostacy. Any one who can say anything in his favour, let him
come forward and plead on his behalf.” But since nothing was brought
forward in his favour he was hanged on the eve of the Passover!55
Rowland concludes:
To exclude the miracles from the gospel tradition makes it impossible to

see why Jesus should have posed anything like the threat that he seemed
to have done to the religious leaders in Jerusalem. The evidence suggests
that they were not dealing merely with a deviant teacher (though there
were elements of that reaction to him) but a claimant to divine power
who had in various ways authenticated his right to be the prophet of
God by his mighty acts.
Morna Hooker, Emeritus Lady Margaret’s Professor of Divinity at Cam-

bridge, also notes that Jesus was charged with being in league with Beel-
zebul, and cites another place in the Rabbinic literature where Jesus is
accused of practising magic. The exact quotation from B. Sanhedrin 107b
is: ‘And a Master has said, “Jesus the Nazarene practised magic and led
Israel astray.” ’56 Hooker’s verdict is: ‘There is no reason to doubt that Jesus
performed exorcisms.’57
Anthony Harvey made the same point, citing B. Sanhedrin 43a as did
Rowland. Harvey also cites a famous passage of the first century Jewish
historian Josephus (AD 37—c. 100), whom one would not expect to be
sympathetic to Christianity. Josephus describes Jesus as a ‘wise man’ and a
‘doer of wonderful works.’ Harvey comments:
That he was a “wise man”, that is, a teacher, is a fact . . . That he was a
“doer of paradoxical works”—for which our word is miracle-worker—
is equally well attested: it is basic to the whole gospel tradition about
him and is implied by subsequent Jewish attacks on him, which repre-
sent him as a sorcerer.58
Harvey has here rendered the Greek of the passage, παραδόξων ἔργων
ποιητής, literally as ‘doer of paradoxical works’ and notes that παράδοξον
is a common term for ‘miracle’ in Hellenistic Judaism. In the New Testa-
ment this same word appears in the plural in Luke 5:26. When Jesus heals
the paralysed man lowered through the roof and claims to forgive sins, the
crowd says, ‘We have seen strange things [παράδοξα] today.’
The full passage in Josephus reads thus:
Now, there was about this time Jesus, a wise man, if it be lawful to
call him a man, for he was a doer of wonderful works—a teacher of
such men as receive the truth with pleasure. He drew over to him both
many of the Jews, and many of the Gentiles. {He was [the] Christ;} and
when Pilate, at the suggestion of the principal men amongst us had
condemned him to the cross, those that loved him at the first did not
forsake him, {for he appeared to them alive again the third day, as the
divine prophets had foretold these and ten thousand other wonderful
things concerning him}; and the tribe of Christians, so named from him,
are not extinct at this day.59
This passage is controversial, and many scholars believe that it has been
interpolated by later Christian scribes. However, James Dunn notes that
there is a broad consensus that Josephus did actually write the passage, and
that it is only those clauses which I have put in curly brackets which are
later additions.60
Hooker notes that the miracles Jesus’ wrought are generally categorized
as healing miracles (including the exorcisms) and nature miracles, though of
course in the healing miracles Jesus is just as much controlling nature. The
function of the miracles is primarily to demonstrate who Jesus is, namely
the Son of God, and to authenticate his message that the Kingdom of God is
breaking into the world through his ministry. Jesus comes with the author-
ity of God in his teaching and his mighty works. This is evident, for exam-
ple, with the stilling of the storm, when the disciples cry, ‘Who then is this,
that even wind and sea obey him?’ (Mk 4:41). Their minds would no doubt
be directed to such Old Testament passages as Psalm 107, where it is Yah-
weh who stills the storm. Moreover, the miracles are not designed to invoke
faith, but happen in response to faith, and are often not understood even by
his closest disciples. Note that Jesus was unable to perform mighty works in
his own country because of their lack of faith (Mk 6:1–6).61
Wright argues that, while most historians are indeed satisfied that Jesus,
and others, performed cures and other extraordinary deeds which had no
naturalistic explanation, Christian apologetics has moved on from deploying
these as ‘proof’ of anything, such as validating a ‘supernatural’ dimension to
the world or demonstrating the Christian religion to be the true one. Apolo-
gists are now concerned with issues of meaning of the events, both for Jesus
and his followers. Nevertheless, it does seem to me that it is important that
the events described in the gospels actually happened in history. One does
not necessarily have to accept every single instance recorded and it may
seem that some events—the earthquake accompanied by the rising of many
from their tombs at the time of Jesus’ crucifixion, recorded only in Matthew,
for example—might have a legendary character. But the overwhelming sense
is that Jesus did indeed perform mighty acts. And it is also the case that such
events do point to Jesus’ special relationship with the God who made the
world and exercises his will over it.
When it comes to issues of meaning, I have already noted the fulfilment
of prophecy motif and that of restoring the excluded to Israel. The latter is
in line with what we know, not just from the Torah but from first century
AD Jewish notions of purity, e.g. in Qumran a maimed Jew could not be a
full member of the community.62 And when Israel was restored the whole of
creation would be restored, as indicated in numerous Old Testament refer-
ences. Thus through Abraham ‘all the families of the earth will be blessed’
(Gen 12:3); and Israel will be restored and his servant become ‘a light to the
nations’ (Is 49:6).
Although Wright is cautious in concluding from the miracles that Jesus
must be divine—indeed he sees Jesus’ ministry as more to do with prophecy
and the inauguration of God’s rule—he nevertheless sees the question of
Jesus’ identity arising naturally out of his words and deeds. And in the gos-
pels this comes to a head with the growing perception of the disciples and
the dialogue at Caesarea Philippi, culminating in Peter’s confession, ‘You are
the Christ, the Son of the living God’ (Mk 8:27–30 and parallels).
The results of our foray into historical criticism in New Testament
scholarship would seem to be in line with the conclusion put forward by
F. F. Bruce some years ago that ‘no matter how far back we may press our
researches into the roots of the gospel story, no matter how we classify the
gospel material, we never arrive at a non-supernatural Jesus.’63 The state-
ment of Gerd Theissen and Annette Merz is also illuminating in making the
comparison between Jesus and other putative miracle workers: ‘Nowhere
else are so many miracles reported of a single person as they are in the Gos-
pels of Jesus.’64
I argued in Chapter 4 that not all religious claims can be true and that
one needs to bring reason and evidence to bear on evaluating competing
claims. Our very brief references to later Rabbinic literature and the Qur’an
confirm that Jesus worked miracles, even if these references are either hos-
tile or add material from a late non-canonical source. Inference to the best
explanation of the data of the gospels points to the truth of the Christian
claims I have examined. To say that Jesus performed miracles is to offer an
explanation for the data that is vastly simpler, and more comprehensive by
far in scope, than to posit the fantastic reconstructions of the ultra-Liberal
critics. In subsequent chapters I apply Bayesian confirmation theory to the
data but for now just provide a preliminary outline summarizing the data
we have considered.
Formalizing the argument: a preliminary outline

We have seen in Chapter 6 that multiple testimonies can mount up rapidly
and make it probable that a purported miracle occurred. Moreover, we also
saw that testimony to many miracles can make it probable that at least one
occurred. Here is a summary of what we have learnt from our discussion of
the historical critical study of the New Testament:
1 In the gospels Jesus is described as performing so many miracles that

one cannot count them. Mark’s gospel begins with Jesus healing a man
with an unclean spirit, which one might call healing of a mental illness.
He then heals Simon Peter’s mother-in-law of a fever. The same even-
ing ‘he healed many who were sick with various diseases, and cast out
many demons’ (Mk 1:34). The next day he went throughout Galilee,
preaching and casting out demons. Then he healed a leper. A few days
later he heals a paralytic. In Chapter 3 he heals a man with a withered
hand. And so it goes on. What we have is a great many witnesses to a
great many miracles, albeit their testimony reported by the author of the
gospel, not directly by themselves. I should add that there is very good
reason to believe that Mark learns what he does from Peter in Rome,
as such scholars as Hooker65 and Bauckham believe,66 since this is what
the early second century writer Papias tells us (and 1 Pet 5:13 can be
invoked in favour of this).
2 There is no counter-evidence. Indeed, even hostile sources or those not
likely to be sympathetic admit that Jesus performed miracles. I have
mentioned the following:
a The scribes from Jerusalem ascribing Jesus’ miracles to Beelzebul
(Mk 3:22);
b The near-contemporary Jewish historian Josephus, scholars gen-
erally agreeing that the terms ‘wise man’ and ‘doer of wonderful
works,’ which he uses to describe Jesus, are genuine and not the
result of Christian interpolation;
c The Babylonian Talmud describing Jesus as a magician;
d The Qur’an describing Jesus’ miracles of healing and raising the
dead, and even the apocryphal turning of clay into a bird.
This all provides powerful evidence for the truth of the Christian claim that
Jesus performed miracles. Inference to the best explanation would lead one
to see, with Tom Wright and others, that the simplest, most elegant and
natural explanation, with the widest scope, is that Jesus did indeed perform
miracles.
The argument can be formalized in Bayesian terms: there are a large num-
ber of reported miracles, and there are also a large number of testimonies
to these miracles. However, we shall leave doing this until we have con-
sidered other pieces of evidence for the truths of Christian claims, i.e. the
evidence that Jesus fulfilled prophecy and the evidence for his resurrection.
One important issue to resolve remains the vexed question of what the prior
probability should be that a miracle would be performed. In Chapter 6 we
presented an argument that in principle multiple testimonies could outweigh
a very small initial probability that a miracle occurred but not that it does so
in practice. We also intimated that if we had good evidence from elsewhere
for the existence of God, that would remove the prejudice that a miracle is
inherently a priori highly improbable. By enabling us therefore to assign a
much higher prior probability to the occurrence of miracles than before, we
could then derive a much higher posterior probability for the occurrence of
miracles. Once we have compared the approaches of Lydia and Timothy
McGrew and Richard Swinburne to the evidence for the resurrection in
Chapter 10 we shall be in a position to return to this issue of the prior prob-
ability and bring together all the evidence we shall have considered in a final
overall assessment.
Notes
2 N. T. Wright, The New Testament and the People of God: Volume 1 of Christian
Origins and the Question of God (London: SPCK, 1992), 35.
3 Wright, New Testament and the People of God, 128–129; also N. T. Wright, The
Resurrection of the Son of God: Volume 3 of Christian Origins and the Question
of God (London: SPCK, 2003), 19; Marcus Borg and N. T. Wright, The Meaning
of Jesus: Two Visions (New York: HarperCollins, [1999] 2007), 18.
4 David Friedrich Strauss, The Life of Jesus Critically Examined (Philadelphia:
Fortress, [1835–1836] 1972).
5 Stephen Neill and Tom Wright, The Interpretation of the New Testament 1861–
1986, second edition (Oxford: Oxford University Press, 1988), 18, including fn. 1.
6 Karl Barth, ‘Strauss,’ in Karl Barth, Protestant Theology in the Nineteenth Cen-
tury, new edition, trans. John Bowden (London: SCM, 2001), 551.
7 Barth, ‘Strauss,’ 551.
8 Neill and Wright, Interpretation of the New Testament, 13–20.
9 A good summary is given in Neill and Wright, Interpretation of the New Testa-
ment, 20–29.
10 Ieuan Ellis, Seven against Christ: A Study of ‘Essays and Reviews’ (Leiden: Brill,
1980), ix; quoted in Ian Hesketh, ‘Behold the (Anonymous) Man: J. R. See-
ley and the Publishing of Ecce Homo,’ Victorian Review 38(1) (Spring 2012),
93–112, 93.
11 Neill and Wright, Interpretation of the New Testament, 34.
12 Daniel Pals, ‘The Reception of “Ecce Homo,” ’ Historical Magazine of the Epis-
copal Church 46(1) (1977), 63–84, 77; also quoted in Hesketh, ‘Behold the
(Anonymous) Man,’ 100.
13 Wright, Resurrection of the Son of God, xv–xvi.
14 Neill and Wright, Interpretation of the New Testament, 35–59.
15 A. E. McGrath, The Making of Modern German Christology, second edition
(Leicester: Inter-Varsity Press, 1994), 89–98.
16 McGrath, Modern German Christology, 93–96.
17 George Tyrell, Christianity at the Crossroads (London: Longmans, Green and
Co., 1910), 44; quoted in McGrath, Modern German Christology, 98.
18 Albert Schweitzer, The Quest of the Historical Jesus: A Critical Study of its Pro-
gress from Reimarus to Wrede, trans. W. Montgomery, preface by F. C. Burkitt
(London: Adam and Charles Black, 1910).
19 Schweitzer, Quest of the Historical Jesus, 368–369. However, this passage about
Jesus’ throwing himself on the wheel is omitted in later editions.
20 Schweitzer, Quest of the Historical Jesus, 401.
21 Graham N. Stanton, The Gospels and Jesus (Oxford: Oxford University Press,
1989), 18.
22 Stanton, Gospels and Jesus, 157.
23 Morna D. Hooker, The Gospel According to St Mark (London: A & C Black,
1991), 78–81.
26 McGrath, Modern German Christology, 157.
27 Dietrich Bonhoeffer, Letters and Papers from Prison, enlarged edition, ed. Eber-
hard Bethge (London: SCM Press, 1971), 329.
28 N. T. Wright, Jesus and the Victory of God: Volume 2 of Christian Origins and
the Question of God (London: SPCK, 1996), 32.
29 Gerd Theissen, The Shadow of the Galilean (London: SCM Press, [1986]
1987), 66.
30 Wright, Jesus and the Victory of God, 40.
32 John Dominic Crossan, The Historical Jesus: The Life of a Mediterranean Jewish
Peasant (New York: HarperCollins, 1991), 332, and, for the inventory, Appendix
1, 427–450; Wright quotes Crossan in Wright, Jesus and the Victory of God, 57.
33 Wolfhart Pannenberg, ed., with Rolf Rendtorff, Trotz Rendtorff, and Ulrich Wil-
ckens, Revelation as History, trans. D. Granskou (New York: Macmillan, 1968).
(First German edition, Offenbarung als Geschichte, Göttingen: Vandenhoeck
and Ruprecht, 1961).
34 See, for example, Richard T. France, ‘Historical Jesus, Quest of,’ in Alister E.
McGrath (ed.), The Blackwell Encyclopedia of Modern Christian Thought
(Oxford: Blackwell, 1993), 260–266, 264.
35 Anthony E. Harvey, Jesus and the Constraints of History: The Bampton Lec-
tures, 1980 (London: Duckworth, 1982).
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006).
39 C. S. Lewis, Miracles: A Preliminary Study (Glasgow: Collins, Fount Paperbacks
edition [1947] 1977), 50; Wright, Jesus and the Victory of God, 186–187, n. 160.
41 Geoffrey Parrinder, Jesus in the Qur’ān (London: Sheldon Press, 1965), 83, and
footnote 1.
46 Wright, Jesus and the Victory of God, 189, quoting Harvey.
53 Christopher Rowland, Christian Origins (London: SPCK, 1985), 146.
54 Rowland, Christian Origins, 147.
55 The Babylonian Talmud, ed. Rabbi Dr Isidore Epstein (London: The Soncino
Press, 1935), Seder Niziḳin, Sanhedrin I, 43a, 281.
56 The Babylonian Talmud, ed. Rabbi Dr Isidore Epstein (London: The Soncino
Press, 1935), Seder Niziḳin, Sanhedrin II, 107b, 736, fn. 2. The footnote states
that the quotation is in the uncensored edition, and notes that the story is anach-
ronistic since it places Jesus in the time of King Jannai (Alexander Jannaeus, c.
104–78 BC).
57 Hooker, Gospel According to St Mark, 62.
58 Harvey, Jesus and the Constraints of History, 98, 59 n. 106.
59 Josephus, The Antiquities of the Jews, xviii.3.3 (63–64), in The Works of Jose-
phus, trans. William Whiston (Peabody, MA: Hendrickson, 1987), 480.
61 Hooker, Gospel According to St Mark, 71–75.
63 F. F. Bruce, The New Testament Documents: Are they Reliable? sixth revised
edition (Downers Grove, IL: IVP Academic, [1981] 2000), 29.
64 Gerd Theissen and Annette Merz, The Historical Jesus (London: SCM Press,
1997), 290; quoted in Richard Swinburne, The Resurrection of the God Incar-
nate (Oxford: Oxford University Press, 2003), 86.
66 Bauckham, Jesus and the Eyewitnesses, 202–239.
8 Jesus and prophecy
What is prophecy and how does Jesus fulfil it?

The claim running through the New Testament is that Jesus fulfils the proph-
ecies of the Old Testament. Just one example would be when the risen Jesus
tells the two disciples on the Emmaus road, ‘O foolish men, and slow of
heart to believe all that the prophets have spoken! Was it not necessary that
the Christ should suffer these things and enter into his glory.’ The passage
goes on, ‘And beginning with Moses and all the prophets, he interpreted to
them in the scriptures the things concerning himself’ (Lk 24:25–27). As we
saw in Chapter 5, like the miracles Jesus wrought, fulfilment of prophecy
was a major strand of evidence for the truth of Christian claims for Pascal
and Butler. Again, we have to tread more carefully in the light of historical
criticism of the Bible, which post-dates these writers. And at the same time,
we need to be critical of the critics in so far as naturalistic assumptions
underlie modern claims that prophecies, like miracles, either simply cannot
occur or are much too improbable to be believed.
One point to notice is that in any case prophecy is not simply or mainly
about predicting the future, though that may be an element of it. The
prophets were concerned to speak truth to power, to point to idolatry and
injustice perpetrated by rulers, priests, and people. These would inevitably
bring catastrophe, such as conquest by foreign powers or failure of harvests.
A useful definition is given by John Barton and John Muddiman:
“Prophecy,” like “wisdom,” is something of a catch all term covering a

wide diversity of material. Its basic form is the oracle: a (usually) short,
pithy saying in which the prophet either denounces some current evil,
or predicts what YHWH will do in the immediate future as a response
to human conduct.1
Thus prophecy, in the first instance, is directly related to the situation or

context in which it is given. However, C. S. Lewis has some instructive com-
ments to make about events far into the future, which the prophets them-
selves could, in normal circumstances, have no notion of, yet seem to be
Jesus and prophecy 149
prophesied. He says that prophecies, although (as per Barton and Muddi-
man) generally having a meaning for the contemporary audience, may also
carry ‘secondary meanings’ which the prophet is unaware of, yet which are
entirely in tune with what the prophet says. Lewis gives a secular analogy
with Plato, who postulated a perfectly righteous man treated as a mon-
ster of wickedness—bound, scourged, and ‘impaled’2 (my translation of the
Republic actually has him ‘crucified,’3 which is what the literal translation
‘impaled’ would equate to later in Roman times).4 Resemblance to the story
of Christ here is no accident, says Lewis, but a sign of Plato’s deep wisdom
and understanding. Likewise—one might say a fortiori—the insights of the
prophets are due to their being in touch with reality at the deepest level.
Indeed, one might argue, with Richard Swinburne, that something like the
crucifixion and resurrection of Christ must happen, or will happen with
significant probability, in a universe created by God.
There is a remarkable passage in Wisdom of Solomon 2:12–20 in the
Apocrypha which goes further than Plato. This is ungodly men reasoning
unsoundly:
Let us lie in wait for the righteous man, because he is inconvenient to us

and opposes our actions; he reproaches us for sins against the law, . . .
he calls the last end of the righteous happy, and boasts that God is his
father. Let us see if his words are true, and let us test what will happen
at the end of his life; for if the righteous man is God’s son, he will help
him, and will deliver him from the hand of his adversaries. Let us test
him with insult and torture, that we may find out how gentle he is, and
make trial of his forbearance. Let us condemn him to a shameful death,
for, according to what he says, he will be protected.
A helpful description of what prophecy means is given by Biblical scholar

Brevard Childs:
Modern biblical scholarship has repeatedly emphasized the fact that

the actual formula of Old Testament citations used to evidence New
Testament fulfillment form only a small portion of what was considered
by the New Testament writers as fulfilled in Christ. Indeed, the entire
life and teaching of Christ were molded and fashioned according to the
tradition of the Old Testament.5
Childs further argues that, if we use Biblical categories, ‘fulfilment is not in

terms of exact correspondence, if this is understood as the matching up of
two independent entities.’ Rather, it is the filling or completion of a single
reality with the intervening time span of no importance since ‘the fulfilment
is accomplished through arbitrary divine intervention.’6
Childs’ point about fulfilment being in terms of the whole life, teaching,
and, I would argue, death and resurrection of Christ will recur in what
I say below. Childs’ remarks are echoed by Dick France writing about Mat-
thew’s gospel, for which fulfilment is the central theme. Indeed the formula,
‘All this took place to fulfil what the Lord had spoken by the prophet . . .,’
occurs repeatedly in Matthew, with minor variations, and even then is far
from exhausting the fulfilment motif. France writes: ‘It is thus for Matthew
not only the explicitly predictive portions of the OT that can be seen to be
“fulfilled” in Jesus, but also its historical characters, its narratives, and its
cultic patterns, even the law itself (5:17; 11:13).’7 It is worth quoting France
more extensively:
Some [of Matthew’s references] depend on apparently superficial points

of correspondence, some on a more far-reaching typology. In many
cases it is possible to suggest several different levels of significance
depending on the degree of scriptural erudition and of shared interpre-
tive assumptions the reader is able to bring to the quotation. “Fulfill-
ment” for Matthew seems to operate at many levels, embracing much
more of the pattern of OT history and language than merely prophetic
predictions. It is a matter of tracing lines of correspondence and con-
tinuity in God’s dealings with his people, discerned in the incidental
details of the biblical text as well as in its grand design. Those who have
studied the interpretation of Scripture among other Jews at the time,
particularly at Qumran and among the rabbis, recognize that they are
on familiar ground in Matthew, sometimes in the actual interpretive
methods he employs, but also more widely in the creative ways he goes
about discovering patterns of fulfillment, ways which modern exegetical
scholarship often finds surprising and unpersuasive. But Matthew was
not writing for modern exegetical scholars, and we may safely assume
that at least some of his intended readers/hearers would have shared his
delight in searching for patterns of fulfillment not necessarily in what
the original authors of the OT texts had in mind but in what can be
perceived in their writings with Christian hindsight.8
Charlie Moule expresses a similar view succinctly thus:
And even the most literalistic ideas of Old Testament prediction and its
verification in Christ—as in the “formula quotations” in Matthew—
may be seen as symptoms, simply, of a much profounder conception of
fulfilment—namely, the discovery that on Christ converges the whole
destiny of Israel, and, through Israel, of all mankind, in its relation
to God.9
Some specific texts

The lesson from the scholars we have quoted is that, when it comes to
prophecy and fulfilment, the situation is complex, we need to be wary of
simply pointing to ‘proof texts,’ and we need to be aware that Jesus is seen
in the New Testament to fulfil the vocation of Israel in a more holistic sense.
Nevertheless, moving on to specific texts, while keeping in mind that wider
context of fulfilment, it is undoubtedly the case that some passages in the
Old Testament do seem to correspond in a quite remarkable way to specific
events in the life, death, and (for the time being, let us say, supposed) res-
urrection of Jesus. The so-called ‘Suffering Servant’ passages in Isaiah are
an example in point. These are embedded in the middle section of Isaiah,
chapters 40–55, generally considered by modern scholars to date from the
exilic period, a century and a half or so after Isaiah of Jerusalem, and thus
labelled ‘Deutero-Isaiah’ or ‘second Isaiah.’
It is worth noting in passing that this division of Isaiah is itself a result of
historical critical scholarship. There are many features of the middle section,
and indeed the final section, which distinguish them from the first part of the
book ascribed to the prophet Isaiah. Especially, it is quite clear that the con-
text of Deutero-Isaiah is the period of the Babylonian exile (589–539 BC)
rather than that of Jerusalem in the second half of the eighth century BC.
One thing which this insight implies for prophecy is that King Cyrus is not
named all those years ahead as the ruler who allows the captives to go back
to their own country: he is already at the gates, in the current history of the
time. Claus Westermann believes that this is still genuine prophecy, written
before the event, given that things did not turn out exactly as predicted (cf.
above quotation from Childs). Thus the Babylonian gods did not have to
be loaded on animals, because Cyrus allowed them to remain in place and
indeed reverenced them himself. Still, ‘This is a particularly fine example of
the fact that prophetic utterances were sometimes not fulfilled exactly, with-
out this in any way detracting from their intrinsic significance.’10 The fact
that the exiles were allowed to return home is confirmed by an extra-Biblical
source, the so-called ‘Cyrus Cylinder’ held in the British Museum.
Let us begin with Isaiah 52:13–53:12, the fourth and final Servant Song.
Here is the major part of it:
He was despised and rejected by men; a man of sorrows, and acquainted

with grief; and as one from whom men hide their faces he was despised,
and we esteemed him not. Surely he has borne our griefs and carried our
sorrows; yet we esteemed him stricken, smitten by God, and afflicted.
But he was wounded for our transgressions, he was bruised for our
iniquities; upon him was the chastisement that made us whole, and with
his stripes we are healed. All we like sheep have gone astray; we have
turned everyone to his own way; and the Lord has laid on him the iniq-
uity of us all. He was oppressed, and he was afflicted, yet he opened not
his mouth; like a lamb that is led to the slaughter, and like a sheep that
before its shearers is dumb, so he opened not his mouth. By oppression
and judgment he was taken away; and as for his generation, who con-
sidered that he was cut off out of the land of the living, stricken for the
transgression of my people? And they made his grave with the wicked
and with a rich man in his death, although he had done no violence,
and there was no deceit in his mouth. Yet it was the will of the Lord to
bruise him; he has put him to grief; when he makes himself an offering
for sin, he shall see his offspring, he shall prolong his days; the will of
the Lord shall prosper in his hand; he shall see the fruit of the travail of
his soul and be satisfied; by his knowledge shall the righteous one, my
servant, make many to be accounted righteous; and he shall bear their
iniquities. Therefore I will divide him a portion with the great, and he
shall divide the spoil with the strong; because he poured out his soul to
death, and was numbered with the transgressors; yet he bore the sin of
many, and made intercession for the transgressors.
(Is. 53:3–12)
In its original context the identity of the Servant is a matter of debate. Is

it the prophet himself who is writing the songs, is it the people of Israel as
a whole, or is it another mysterious individual? Westermann writes of the
Servant Songs in general:
The cryptic, veiled language used is deliberate . . . The veiled manner of

speaking is intentional, and to our knowledge much in them was meant
to remain hidden even from their original hearers . . . On principle, their
exegesis must not be controlled by the question, “Who is this servant of
God?” Instead, we must do them justice by recognizing that precisely
this is what they neither tell nor intend to tell us. The questions which
should control exegesis are: “What do the texts make known about
what transpires, or is to transpire, between God, the servant, and those
to whom his task pertains?”11
When we come to the New Testament, pace Westermann, it is hard to avoid

the question, ‘Who is this servant of God?’ Indeed, this is precisely the ques-
tion posed by the Ethiopian eunuch, whom Philip meets reading this pas-
sage, in Acts 8: ‘About whom, pray, does the prophet say this, about himself
or about someone else?’ (Acts 8:34). We read further, ‘Then Philip opened
his mouth, and beginning with this Scripture he told him the good news of
Jesus.’ (Acts 8:35).
As we have indicated, the passage, indeed all the Old Testament passages
one might quote, and also the miracles we have discussed, should really be
considered in the context of Jesus’ whole life, ministry, death, and supposed
resurrection. In a nutshell, Jesus proclaimed the Kingdom of God, and that
this kingdom was being established in and through himself. Expectations
that God would bring about his kingly rule were widespread in the first cen-
tury and were centred on the coming of the Jewish Messiah. However, these
expectations were largely understood in terms of the Messiah overthrowing
the Roman occupying power. Jesus fulfilled these expectations, rooted in
the Old Testament, in a totally surprising way. In particular, he fulfilled the
prophecies concerning the Suffering Servant in Isaiah as a suffering Messiah
rather than conquering hero.
Following Jesus’ multiple healings in Matthew 8:14–16 (paralleling
Mk 1:29–34) comes one of Matthew’s numerous fulfilment formulae: ‘This
was to fulfil what was spoken by the prophet Isaiah, “He took our infirmi-
ties and bore our diseases.” ’ (Mt 8:17). Indeed, the Suffering Servant pas-
sages are frequently quoted in the New Testament.
It is certainly the case that Jesus was ‘despised and rejected.’ Many of the
phrases used in Isaiah 53 correlate strongly with Jesus’ crucifixion—stricken,
smitten by God, afflicted, wounded, oppressed. We read in Mark’s gospel
that he was scourged—hence ‘stripes.’ What would have been surprising
for the original readers of this passage are the reasons why the Servant suf-
fered. It would have been natural in Deutero-Isaiah’s culture to assume that
the Servant had been smitten by God for his own sins. The revolutionary
aspect of the passage is that he ‘bore our griefs and carried our sorrows,’
that he was ‘wounded for our transgressions,’ and ‘bruised for our iniqui-
ties.’ Moreover, the chastisement which he suffered makes us whole, and
his stripes bring us healing. All this fits with how Jesus described his own
mission and what the New Testament writers say about him. Ultimately,
whether one believes that Jesus died for the sins of the world depends on
what one makes of the whole story of his life, death, and most importantly
resurrection. If the resurrection really occurred, and this we come to in the
next two chapters, then that vindicates the claims Jesus and his followers
made about himself and his mission.
The claim that Jesus’ death is an atonement for sin is much more promi-
nent in the epistles than the gospels. Nevertheless, it is made by Jesus him-
self in two places in Mark’s gospel, the earliest to be written.12 Following
the request by James and John for places of honour in the Kingdom, Jesus
rebukes them, saying: ‘ . . . whoever would be first among you must be slave
of all. For the Son of man also came not to be served but to serve, and to
give his life as a ransom for many’ (Mk 10:44–45). At the institution of the
Lord’s supper, which very significantly is a Passover meal commemorating
the liberation of Israel from slavery in Egypt, and takes place on the eve of
the crucifixion, Jesus says of the cup, ‘This is my blood of the covenant,
which is poured out for many’ (Mk 14:24).
That Jesus died as an atonement for sin is a major theme of Paul’s epistles,
which pre-date the gospels, so this is clearly a very early understanding. In
the very important passage in 1 Corinthians 15, which we consider in more
detail in later chapters, Paul writes this: ‘For I delivered to you as of first
importance what I also received, that Christ died for our sins in accordance
with the scriptures. . . ’ As I explain in Chapter 9, Paul received what he
describes in this chapter from the apostles within at most six years from
Jesus’ crucifixion and supposed resurrection, so the idea that Christ died for
our sins is there from the beginning.
As we progress through Isaiah 53 features resembling the story of Jesus
become even more striking. We read, ‘He was oppressed, and he was
afflicted, yet he opened not his mouth. . . ’ (Is 53:7). In Mark 14:60–61a we
read: ‘And the high priest stood up in the midst, and asked Jesus, “Have you
no answer to make? What is it that these men testify against you?” But he
was silent and made no answer.’ When asked further if he is the Christ, Jesus
does respond in the affirmative. Similarly, when Pilate asks, ‘Are you the
King of the Jews?’ Jesus answers, ‘You have said so’ (Mk 15:2). But we read
on: ‘And Pilate again asked him, “Have you no answer to make? See how
many charges they bring against you.” But Jesus made no further answer, so
that Pilate wondered’ (Mk 15:4).
Striking again is what we read in Isaiah 53:9: ‘And they made his grave
with the wicked and with a rich man in his death, although he had done no
violence, and there was no deceit in his mouth.’13 Jesus was crucified between
two thieves. However, we are told that a respected member of the Council,
Joseph of Arimathea, requested the body of Jesus from Pilate, which he then
buried in a newly hewn rock tomb (Mk 16:42–46). Matthew, for whom, as
we have seen, prophetic fulfilment is a major theme, adds that Joseph was ‘a
rich man’ (Mt 27:57), something which is only implicit in Mark. Westermann
notes that ‘there is point for point correspondence with the Church’s confes-
sion as it is given in the Apostles’ Creed—born, suffered, died and was buried.’
Indeed, he sees the similarity in structure with the creed as more important
than isolated references to the passage here and there in the New Testament.14
Perhaps most remarkable of all is the apparent hint that the Servant will
rise from the dead: ‘ . . . when he makes himself an offering for sin, he shall
see his offspring, he shall prolong his days; the will of the Lord shall prosper
in his hand; he shall see the fruits of the travail of his soul and be satis-
fied . . . Therefore I will divide him a portion with the great, and he shall
divide the spoils with the strong; because he poured out his soul to death,
and was numbered with the transgressors; yet he bore the sin of many, and
made intercession for the transgressors.’
This is in line with the introductory part of the song, which comes at
the end of chapter 52, indicating that the Servant will be exalted after his
humiliation:
Behold, my servant shall prosper, he shall be exalted and lifted up, and
shall be very high. As many as were astonished at him—his appearance
was so marred, beyond human semblance, and his form beyond that of
the sons of men—so shall he startle many nations; kings shall shut their
mouths because of him; for that which has not been told them they shall
see, and that which they have not heard they shall understand.
(Is. 52:13–15)
To see the resurrection of Jesus foretold here is to step beyond what an Old
Testament exegete might be warranted to claim just on the basis of Old
Testament scholarship. But as systematic theologians or philosophers, we
are entitled to draw wider conclusions.
I have spent considerable space discussing the fourth Servant Song. It is
worth dwelling on another passage in Deutero-Isaiah for a moment. It is
part of the unit Isaiah 42:5–9 which follows on immediately from the first
Servant Song in Isaiah 42:1–4, but is thought not to be a direct extension
of it:15
. . . Thus says God, the Lord, who created the heavens and stretched
them out, who spread forth the earth and what comes from it, who
gives breath to the people upon it and spirit to those who walk in it:
“I am the Lord, I have called you in righteousness, I have taken you by
the hand and kept you; I have given you as a covenant to the people,
a light to the nations, to open the eyes that are blind, to bring out the
prisoners from the dungeon, from the prison those who sit in darkness.”
(Is. 42:5–7)
It was always Israel’s role to be a source of blessing to all the world, to

be a ‘light to the nations.’ This destiny goes right back to Abraham, to
whom God promised that ‘by you all the families of the earth shall bless
themselves’ (Gen 12:3). The covenant with Israel was to be universalized to
include all peoples.
It is clear from numerous New Testament references that the writers see
Jesus as universalizing or including Gentiles within the people of God. Mat-
thew’s gospel begins with a visit to Jesus’ birthplace from the wise men,
symbolizing just that inclusion which the prophets wrote about, and ending
with the risen Christ saying, ‘Go therefore and make disciples of all nations,
baptizing them in the name of the Father and of the Son and of the Holy
Spirit. . . ’ Within the gospel Jesus himself says, ‘I tell you, many will come
from east and west and sit at table with Abraham, Isaac, and Jacob in the
kingdom of heaven, while the sons of the kingdom will be thrown into the
outer darkness. . . ’ (Mt 8:11–12, Jesus indicating that many Jews would
reject him). In Luke’s gospel, the phraseology of Isaiah 42 is echoed in the
words of the old man Simeon when he sees the baby Jesus: ‘Lord, now let-
test thou thy servant depart in peace, according to thy word; for mine eyes
have seen thy salvation which thou hast prepared in the presence of all peo-
ples, a light for revelation to the Gentiles, and for glory to thy people Israel’
(Lk 2:29–32).
There is a similar statement in the second Servant Song, Isaiah 49:1–7.
Here is an extract: ‘ “It is too light a thing that you should be my servant to
raise up the tribes of Jacob and to restore the preserved of Israel; I will give
you as a light to the nations, that my salvation may reach to the end of the
earth.” ’ (Is 49:6).
As Paul and Barnabas go about the task of preaching the gospel to the
nations, and bringing salvation to the ends of the earth, they make the point
that they are fulfilling this prophecy: ‘For so the Lord has commanded us,
saying, “I have sent you to be a light for the Gentiles, that you may bring
salvation to the uttermost parts of the earth” ’ (Acts 13:47).
Noteworthy from Isaiah 42:5–9 is also this clause: ‘ . . . to open the eyes
that are blind, to bring out the prisoners from the dungeon, from the prison
those who sit in darkness.’ (Is 42:7). A further significant passage is to be
found in the third part of Isaiah (Trito-Isaiah):
The Spirit of the Lord God is upon me, because the Lord has anointed
me to bring good tidings to the afflicted; he has sent me to bind up the
brokenhearted, to proclaim liberty to the captives, and the opening of
the prison to those who are bound; to proclaim the year of the Lord’s
favour, . . .
(Is. 61:1–2)
This is particularly significant since Jesus reads this passage when handed
the scroll of Isaiah in the synagogue at Nazareth in the programmatic pas-
sage Luke 4:16–19. Jesus goes on to say (Lk 4:21), ‘Today this scripture
has been fulfilled in your hearing.’ Indeed, if we agree that the account
of Jesus’ ministry in the gospels is broadly reliable, as I have argued we
should, then Jesus is indeed fulfilling these prophecies and others like them.
The blind are receiving their sight, the sick and demon-possessed are being
healed, the bound are being freed from their afflictions (e.g. the kyphotic
woman whom ‘Satan bound for eighteen years’ in Lk 13:10–17) and for-
given their sins.
Dying on the cross Jesus cries out, ‘My God, my God, why hast thou
forsaken me?’ (Mk 15:34). Jürgen Moltmann has written a whole book on
the rich theology of this single verse.16 However, what concerns us for the
moment is the fact that Jesus is quoting Psalm 22. In fact, the whole psalm
resonates very strikingly with Jesus’ crucifixion. We read, in similar vein to
Isaiah 53:
But I am a worm and no man; scorned by men, and despised by the

people. All who see me mock me, they make mouths at me, they wag
their heads; “He committed his cause to the Lord; let him deliver him,
let him rescue him, for he delights in him!”
(Ps. 22:6–8)
In Mark 15:16–20 we read:
And the soldiers led him away inside the palace (that is, the praeto-
rium); and they called together the whole battalion. And they clothed
him in a purple cloak, and plaiting a crown of thorns they put it on
him. And they began to salute him, “Hail, King of the Jews!” And they
struck his head with a reed, and spat upon him, and they knelt down
in homage to him. And when they had mocked him, they stripped him
of the purple cloak, and put his own clothes on him. And they led him
out to crucify him.
Again in Mark 15:29–32 we read:
And those who passed by derided him, wagging their heads, and saying,
“Aha! You who would destroy the temple and build it in three days,
save yourself, and come down from the cross!” So also the chief priests
mocked him to one another with the scribes, saying, “He saved others;
he cannot save himself. Let the Christ, the King of Israel, come down
now from the cross, that we may see and believe.” Those who were
crucified with him also reviled him.
Of course, as with many prophecies apparently fulfilled in the New Testa-

ment, it would be possible to argue that these verses were not, after all,
based on eyewitness testimony to the events described. However, given the
whole story of Jesus, including the charge of destroying the Temple and
building another, the action of the Roman soldiers and the comments of
the passers-by seem highly plausible. It is highly likely that Jesus’ criticism
of, and action in, the Temple were, at the human level, the reason for his
crucifixion. There is also in the quotation the reference to Jesus ‘saving oth-
ers,’ very much in line with the general recognition of his power to perform
miraculous healings.
Considering the authenticity of the first passage, Morna Hooker notes
that no manuscript of Mark is without it and that the incident is found
in all four gospels. Moreover, there are interesting contemporary parallels.
For example, according to Philo Agrippa was mocked when made King
of Judaea, although it was the ‘half-wit’ Carabas who was dressed up and
saluted as King instead. Other ceremonies featuring the dressing up of a
mock king are recorded, even including the king’s sacrificial death. Hooker
concludes: ‘The existence of these parallels does not undermine the his-
toricity of this particular incident; rather they suggest that it was entirely
natural.’17 Hooker notes also a parallel with Isaiah 50:6 (part of the third
Servant Song, Is 50:4–9): ‘I gave my back to the smiters, and my cheeks to
those who pulled out the beard; I hid not my face from shame and spit-
ting.’ She does not regard the reference to Jesus being spat at as being influ-
enced by the Isaiah passage.18 Nevertheless, from our point of view it is yet
another striking parallel.
Hooker thinks that the three-fold mockery in the second passage sug-
gests elaboration.19 Nevertheless, she thinks that the derision of passers-
by is ‘probable enough,’ even though Mark’s description of them shaking
their heads (a gesture of contempt) echoes our passage in Psalm 22:7 and
also Lamentations 2:15.20 That is, it is historically probable that Jesus was
mocked in the way the Psalm describes.
Another significant verse in Psalm 22 is Psalm 22:18: ‘they divide my
garments, and for my raiment they cast lots.’ In Mark’s account of the cru-
cifixion we read: ‘And they crucified him, and divided his garments among
them, casting lots for them, to decide what each should take’ (Mk 15:24).
About this verse Hooker writes:
Mark’s account clearly echoes the wording of Ps. 22.18 . . . This does
not necessarily mean that the incident has been created out of the quota-
tion, though the reference to casting lots may be an elaboration based
on the psalm. Nevertheless, Mark would probably not have mentioned
this particular incident at all if he had not seen in it the fulfilment of
scripture.21
I have dwelt a lot on the suffering of Christ as fulfilling the expectation

for God to act as prophesied in the Hebrew Scriptures. This is important
because it was a neglected aspect of prophecy in the first century. The Jews
were widely expecting God to act, but much more through a Messiah who
would act decisively to overthrow the Romans. There are contemporary
writings which support that kind view, and Messianic claimants who tried
to do just that. The Old Testament also speaks of a descendant of David
who will bring in God’s Kingdom and defeat his enemies, though it is at this
point that Jesus’ fulfilment is in a manner contrary to expectations.
That the one to come is to be a descendant of David may be illustrated
from a number of Old Testament passages. For example, in Isaiah 11:1–5
we read:
There shall come forth a shoot from the stump of Jesse, and a branch
shall grow out of his roots. And the Spirit of the Lord shall rest upon
him, the spirit of wisdom and understanding, the spirit of counsel and
might, the spirit of the knowledge of the fear of the Lord. And his
delight shall be in the fear of the Lord. He shall not judge by what his
eyes see, or decide by what his ears hear; but with righteousness he shall
judge the poor, and decide with equity for the meek of the earth; and he
shall smite the earth with the rod of his mouth, and with the breath of
his lips he shall slay the wicked. Righteousness shall be the girdle of his
waist, and faithfulness the girdle of his loins.
That the Lord’s future anointed king will win victory in battle is also sup-
ported by various Old Testament passages. For example, in Psalm 2:1–9 we
read:
Why do the nations conspire, and the peoples plot in vain? The kings of
the earth set themselves, and the rulers take counsel together, against the
Lord and his anointed, saying, “Let us burst their bonds asunder, and
cast their cords from us.” He who sits in the heavens laughs; the Lord
has them in derision. Then he will speak to them in his wrath, and ter-
rify them in his fury, saying, “I have set my king on Zion, my holy hill.”
I will tell of the decree of the Lord: He said to me, “You are my son,
today I have begotten you. Ask of me, and I will make the nations your
heritage, and the ends of the earth your possession. You shall break
them with a rod of iron, and dash them in pieces like a potter’s vessel.”
Psalm 110:1 reads: ‘The Lord said to my Lord: “Sit at my right hand, till
I make your enemies your footstool.” ’
A vitally important passage, in view of Jesus’ self-description, is the Son of
Man passage in Daniel 7. What is truly remarkable is how, in a completely
unexpected and even shocking way Jesus fulfils both kinds of prophecy,
that of a King-Messiah and a Suffering Servant. Daniel has a vision in which
various beasts arise and are overthrown, and a final king arises of whom we
read, when Daniel is given the interpretation:
He shall speak words against the Most High, and shall wear out the
saints of the Most High, and shall think to change the times and the
law; and they shall be given into his hand for a time, two times, and
half a time. But the court shall sit in judgment, and his dominion shall
be taken away, to be consumed and destroyed to the end.
(Dan 7:25–26)
Earlier in the chapter we read:
I saw in the night visions, and behold, with the clouds of heaven there
came one like a son of man, and he came to the Ancient of Days and
was presented before him. And to him were given dominion and glory
and kingdom, that all peoples, nations, and languages should serve him;
his dominion is an everlasting dominion, which shall not pass away,
and his kingdom one that shall not be destroyed.
(Dan 7:13–14)
In its context the nations which are overthrown would be the Babylonians,
Medes, Persians, and Greeks. After the death of Alexander the Great, the
Greek/Macedonian Empire was divided among his generals, the part falling
to the Seleucids including Palestine. The final king, who spoke against the
Most High, etc., was the Seleucid King Antiochus IV Epiphanes, whom we
met in Chapter 5.22 At the time of Jesus, the last empire was reinterpreted as
the Romans, and the expectation was for a conquering hero to overthrow
them. The Jews were, as Tom Wright explains, still in exile, given that, even
though they had returned from Babylon many centuries before, they had
nevertheless always been under the yoke of foreign powers.23 Hence such
prophecies were still awaiting fulfilment. However that may be, Jesus’ fulfil-
ment is in quite different terms to those expectations.
Jesus’ self-identification as ‘Son of Man’ clearly indicates that he sees
himself as the one to whom dominion is to be given and whom all peoples
will serve for ever. The shocking way in which this happens is through his
bringing the kingdom into being during the course of his ministry, with his
healings, forgiving of sins, bringing in the outcasts, and establishing a revo-
lutionary ethical framework of love for enemies, going the other mile, etc.
Then follows his death and resurrection and the spread of the gospel like
wildfire to the seat of worldly power in Rome and beyond, so that today
more than two billion people do indeed worship and serve him. Of course,
the final fulfilment is yet to come, at the eschaton when all are raised from
the dead, with him as the first fruits. In terms of building a case for proph-
ecy providing evidence for the truth of Christian claims, this does of course
imply a dependence on seeing the truth of the rest of the story. It is thus
intimately tied in with the whole of Jesus’ life, death, and supposed resur-
rection, and indeed with how the immediate circle of Jesus followers went
about spreading the good news afterwards.
There are many ways in which Jesus fulfils Israel’s hopes in a more holistic
way than by simple resemblances to individual verses or passages in the Old
Testament. Thus he chooses twelve disciples, mirroring the twelve tribes of
Israel. This would seem to indicate not simply that he is the prophet like
Moses who is promised (Deut 18:15–18), but that he is somehow embody-
ing the new Israel, recreating Israel as she is meant to be, a light to the
nations.
Jesus is tempted in the wilderness for forty days, mirroring Israel’s sojourn
in the wilderness for forty years. Jesus’ victory over the Devil (however
understood) contrasts with Israel’s rebellion, and marks out Jesus’ mission
to bring about the new Exodus, or new return from exile, as Tom Wright
sees it.
In Matthew’s gospel Jesus says, ‘Think not that I have come to abolish the
law and the prophets; I have come not to abolish them but to fulfil them’
(Mt 5:17). This comes in the Sermon on the Mount (Mt 5:1–7:27), in which
Jesus radically reinterprets the Torah and mirrors Moses’ reception of the
law on Mt Sinai. This too would be in line with the identification of Jesus
as the promised prophet like Moses (although in the Lucan parallel it is the
Sermon on the Plain, Lk 6:17–49).
W. D. Davies sees Jesus as presented by Matthew to be a law-giver greater
than Moses.24 Indeed, Christ can be seen as the new Torah or fulfilment of
the Torah. Davies’ particular focus is in arguing that Paul sees Christ this
way. This idea can be seen in Matthew 11:29, 30 where Jesus says: ‘Take my
yoke upon you, and learn from me. . .,’ since taking the yoke of the Torah
was a common Rabbinic expression (e.g. Pirḳê Aboth 3.6).
One particular quotation from Matthew’s gospel, also given by Davies,
seems to go further than this. In Matthew 18:20 Jesus says, ‘For where
two or three are gathered together in my name, there am I in the midst of
them.’ An interesting comparison here is with another quotation from the
Mishnah: ‘When they sit together and are occupied with Torah, the Sheki-
nah is among them’ (Pirḳê Aboth 3.7).25 Yes, Jesus is substituting for the
Torah, but, remarkably, he seems also to embody the Shekinah, the glory
of God.
The story of the feeding of the five thousand mirrors God feeding the
Hebrews with manna in the wilderness. Of course, the authenticity of this
is often challenged, with speculation that Jesus might just have encouraged
those gathered to take out the food which they had brought and to share it.
However, this miracle is recorded in all four gospels, and the similar miracle
of feeding the four thousand is contained in Mark and Matthew. Moreover,
there is no indication in any of the texts that it is meant to be taken other
than literally. As Morna Hooker points out, and as we noted in Chapter 7,
the distinction between healings and so-called nature miracles is artificial,
since all are really nature miracles.26 Noting what I have said above about
the ubiquity of this particular miracle, and regarding it as highly probable
that John uses a different cycle of tradition from Mark, Hooker states: ‘This
widespread use suggests that the feeding miracle, whatever its difficulties
for the modern reader, was of considerable importance in early Christian
circles.’27 She goes on:
Suggestions that the numbers involved were exaggerated (Wellhausen)

or that the feeding was a symbolic one involving only tiny fragments of
food (A. Schweitzer, The Mystery of the Kingdom of God, pp. 103–6)
or that the people were persuaded to share the food which they had
brought with them (Branscomb), cannot explain the belief of all the
evangelists that this was a miraculous feeding.
The miracle (or miracles, if we include the feeding of the four thousand)
could thus be included among those I considered in Chapter 7. Hooker,
however, goes on to discuss the more important question of what it means
for Jesus’ person, such as the parallel with Moses and the manna in the wil-
derness which we noted above.
Another significant aspect of fulfilment concerns John the Baptist. Mark’s
gospel opens with the prophecy of a forerunner in Malachi and Deutero-
Isaiah (Mal 3:1, Is 40:3). Isaiah 40:3 reads: ‘Behold, I send my messen-
ger before thy face, who shall prepare thy way; the voice of one crying in
the wilderness: Prepare the way of the Lord, make his paths straight—’
(Mk 1:2–3).
Malachi 3:1 actually reads: ‘Behold, I send my messenger to prepare the
way before me, and the Lord whom you seek will suddenly come to his tem-
ple . . . ’ And at the end of Malachi the forerunner or messenger is identified
as the prophet Elijah: ‘Behold, I will send you Elijah the prophet before the
great and terrible day of the Lord comes’ (Mal 4:5).
The description of John given in Mark 1:6 is that he was ‘clothed with
camel’s hair, and had a leather girdle around his waist . . . ’ Both the hairy
mantle and the leather girdle exactly match the same items worn by Elijah
in 2 Kings 1:8, suggesting that John is indeed Elijah who is to come. In
Matthew 11:7–15 Jesus himself identifies John as the messenger of Malachi
3:1 and says (Mt 11:13–14): ‘For all the prophets and the law prophesied
until John; and if you are willing to accept it, he is Elijah who is to come.’ Of
course, this is not to be taken in a woodenly literal way; rather John is Elijah
in the metaphorical sense of being the greatest prophet and the forerunner
(in contrast to the transfiguration, when one is meant to take it that Moses
and the real Elijah appear alongside Jesus, Mk 9:2–8).
Regarding the second part of Malachi 3:1, Jesus’ cleansing of the Temple
would also appear to be a fulfilment. Indeed, as with Matthew 18:20, it would
seem that here too Jesus is identifying himself with YHWH, whose return to his
Temple was widely expected in this period. In John’s gospel this is expounded
thematically as Jesus being the new Temple, or dwelling place of God (Jn 2:13–
22), and Jesus’ followers as derivatively a new Temple too (Jn 14).28
It could of course always be argued that these similarities to what was
prophesied were engineered with the deliberate intention of making them
look like fulfilment. But another way to view it is to say that, given an
already intuited sense of vocation in both John and Jesus, and their total
immersion in the Scriptures, they would be led to carry out precisely the
kinds of actions they did, as indicated by those same Scriptures.
Tom Wright addresses the very pertinent question as to why the early
followers of Jesus regarded him as Messiah, fulfilling those Old Testament
Scriptures. Without the resurrection, they could not have done this, but the
resurrection was so counter to the expectations associated with a Messiah
that on its own it cannot explain the use of the term Messiah to describe
Jesus. Wright takes his cue for an explanation from the ‘title’ on the cross,
i.e. the charge against Jesus of claiming to be the King of the Jews, and Jesus’
action in the Temple. The two are connected, since it is Jesus’ cleansing of
the Temple which, at the human level, most probably led to his arrest.
Wright states: ‘Jesus’ action in the Temple constitutes the most obvious
act of messianic praxis within the gospel narratives.’29 The day before this
symbolic action Jesus carried out another symbolically laden action. He
entered Jerusalem triumphantly on a donkey (Mk 11:1–10). He is hailed by
the crowds: ‘Hosanna! Blessed is he who comes in the name of the Lord!
Blessed is the kingdom of our father David that is coming! Hosanna in the
highest!’ The crowds understood precisely what Jesus was doing, no doubt
aware of the prophecy of Zechariah 9:9–10:
Rejoice greatly, O daughter of Zion! Shout aloud, O daughter of Jeru-

salem! Lo, your king comes to you; triumphant and victorious is he,
humble and riding on an ass, on a colt the foal of an ass. I will cut off
the chariot from Ephraim and the war horse from Jerusalem; and the
battle bow shall be cut off, and he shall command peace to the nations;
his dominion shall be from sea to sea, and from the River to the ends
of the earth.
My point here is not, ‘how astonishing the resemblance between the proph-
ecy and Jesus’ fulfilment of it!’ Unlike his death, the circumstances of which
could not be engineered, here it is highly likely that Jesus, aware of his voca-
tion, is deliberately carrying out the action, well-knowing the Scripture and
what the crowd’s perception would be.
Wright sees Jesus’ action here as ‘royal’ and points to various Scriptures
which were understood as ‘explicitly messianic’ in this period, for example
2 Samuel 7:12–14:
When your days are fulfilled and you lie down with your fathers, I will
raise up your offspring after you, who shall come forth from your body,
and I will establish his kingdom. He shall build a house for my name,
and I will establish the throne of his kingdom for ever. I will be his
father, and he shall be my son.
In its context, the immediate offspring of David referred to here is Solomon

who built the first Jerusalem Temple. However, the verse does establish the
link between the Temple and the establishment of Solomon’s throne ‘for
ever,’ and hence may be applied to one who judges the Temple, of which
he is ruler, and seeks to build a new one. There is even a reference to the
removal of traders in the Temple in Zechariah 14:21: ‘ . . . And there shall
no longer be traders in the house of the Lord of hosts on that day.’ Jesus’
action also echoes the story of the Maccabean revolt in which Judas Macca-
beus defeats the Syrians, who under Antiochus IV Epiphanes had desecrated
the Temple, then enters the Temple and cleanses it (1 Macc 7:36–51).
Wright cites numerous ways, both symbolic actions and teaching para-
bles, in which Jesus manifests his claim to be the Messiah, especially at the
time of his symbolic action in the Temple. Here are a few of them:30
1 The cursing of the barren fig tree (Mk 11:12–14), further explained by
the riddle, spoken on Mount Zion, of prayer for the removal of ‘this
mountain,’ both representing the spiritually bankrupt Temple.
2 The parable of the tenants in the vineyard (Mk 12:1–12), also represent-
ing the Temple and its misrule, and resonating with Isaiah 5:1–7.
3 The riddle of the ‘stone which the builders rejected’ which ‘has become
the head of the corner’ and the affirmation that ‘this was the Lord’s doing
and it is marvellous in our eyes’ (Mk 12:10–11), here quoting Psalm
118: 22–23 and also relating to the stone which crushes the empires and
brings about God’s kingdom in Daniel 2:34–35, interpreted this way in
Daniel 2:44–45.
One could go on. The question arises, however, as to how exactly Jesus ful-
filled the prophecies associated with the Messiah of victory over Yahweh’s
enemies and renewing the worship of Israel in a restored Temple. To answer
that entails taking the meaning of the cross and the truth of the resurrection
seriously and to recognize what I have said earlier, that Jesus fulfilled these
prophecies in the most shocking and surprising way. We have said something
above about what Christ himself thought he was doing in making atonement.
The resurrection will be dealt with in the next two chapters. However, one
thing can be said at this stage and that is that Jesus’ prophecy of the destruc-
tion of the Temple (within a generation, taking Mk 13:30 to refer to it)31 was
fulfilled quite literally in AD 70 when the Temple was indeed destroyed by
the Romans. The episode is related by Josephus and a physical testament to
it is the ‘spoils of Jerusalem’ relief on the inside of the arch of Titus in Rome.
I could say a lot more about prophecy, just as I could have done in the last
chapter about Jesus’ miracles, but we need to move towards a more system-
atic classification in order to set up the argument in a logically persuasive
way. Hence I must refer the reader to the literature for more detail, as I end
this chapter with a brief summary and prepare the way for formulating the
argument utilizing the Bayesian framework in Chapter 11.
Formalizing an argument from prophecy: outline of the

argument
It seems to me, and this will be apparent from the foregoing, that formal-
izing an argument from prophecy is much more complex than formalizing
an argument from Jesus’ miracles, which we did in the last chapter. This is
because, as I have reiterated, the fulfilment of prophecy is more to do with
the whole of Jesus’ life, ministry, death, and resurrection fulfilling the des-
tiny of Israel than the collection of individual instances.
Having said that, it seems that there is a subset of prophecies which may
be treated in something like the way I have described for miracles, most
especially those over which Jesus had no control. In other instances it may
be that Jesus’ sense of vocation led him to act in certain ways in conformity
with Old Testament passages. In some of these and in yet further instances
fulfilment depends on accepting that Jesus performed miracles and that he
rose from the dead and thus was indeed the Messiah expected. This is in
line with the fact that the disciples only realized after Jesus’ resurrection
how he fulfilled the prophecies of the Old Testament (see Lk 24:25–27, the
disciples on the Emmaus Road, with which we began). Let us consider these
sub-categories in turn.
(a) Prophecies the result of which Jesus had no control

1 Jesus’ scourging and crucifixion fulfilling Isaiah 53:3–5.
2 Jesus’ own prediction of his death and resurrection in Mark 8:31, 9:31,
and parallels.
3 Jesus’ prediction of the future maltreatment of his disciples in Mark
13:9, fulfilled in the book of Acts.
4 Fulfilment of Isaiah 53:9. ‘They made his grave with the wicked’ is ful-
filled in the sense that Jesus was crucified as a criminal between two
thieves. Burial in the tomb belonging to Joseph of Arimathea fulfils the
clause, ‘and with a rich man in his death’ (see Mt 27:57).
5 The mockery of Jesus in Mark 15:16–20 and 15:29–32 fulfilling Psalm
22:6–8.
6 The sharing of Jesus’ garments in Mark 15:24 fulfilling Psalm 22:18 (see
earlier discussion).
7 The recapitulation of Israel’s forty-year sojourn in the wilderness. This
is fulfilled by Jesus’ forty-day temptation in the wilderness if we take
it as literally true that he was driven there by the Spirit (Mk 1:12), the
fundamental difference being that Jesus conquers temptation whereas
Israel failed. Note that all three synoptic gospels include the tempta-
tion in the wilderness and that Matthew and Luke provide more detail,
clearly informed by a different source from that in Mark. Alternatively
this fulfilment could be placed in the next section since Jesus could have
voluntarily taken himself to the desert, knowing that it would be a place
of temptation, as for the Israelites.
8 Jesus’ prophecy of the destruction of the Temple (Mk 13:1–2 plus his
parables such as the tenants in the vineyard). This was fulfilled in AD 70.
9 Isaiah 7:14 fulfilled in Jesus’ virgin birth. Of course this is controver-
sial at a number of levels, and I have not had space above to consider
it. Although it is only found in Matthew and Luke and, of course, if it
happened would pre-date the writing of the gospels by another thirty
years, these two accounts are independent, one from Joseph’s point of
view and one from Mary’s. While it does not feature in the other gos-
pels, there may be an allusion in John 8:41 when the Jews say to Jesus,
‘We were not born of fornication.’ The implication that Jesus was born
of fornication is also to be found in Jewish sources and is an accusa-
tion made by Celsus (Origen, contra Celsum, I. 28).32 It is also argued
that the Hebrew word translated ‘virgin’ in Isaiah 7:14 simply means
‘young woman.’ While that is so, and the Hebrew can mean maid or
newly married woman, the Greek version, the Septuagint, has παρθένος
(parthenos), which does mean virgin.
10 Isaiah 11:1 is fulfilled in Jesus’ Davidic descent. According to the birth
narratives in Matthew and Luke Jesus would fulfil this on his adop-
tive father Joseph’s side; his mother Mary may be a Levite if, as she is
described by Luke, she is a cousin of Elizabeth. Note, however, Jesus’
challenge to the scribes in Mark 12:35–37, in referring to Psalm 110:1:
‘David himself calls him Lord; so how is he his son?’ Whatever his
descent, Jesus is claiming to be greater than David.
(b) Prophecies where Jesus may have acted with Old

Testament passages in mind
1 Jesus ‘opened not his mouth. . . ’ fulfilling Isaiah 53:7. This is fulfilled in
Mark 14:60–61a.
2 Isaiah 11:4, ‘with righteousness he shall judge the poor, and decide with
equity for the meek of the earth,’ is fulfilled in Jesus’ ministry to the
marginalized.
3 Jesus’ calling of twelve disciples mirroring the twelve tribes of Israel.
4 Jesus’ radical reinterpretation of the law in the Sermon on the Mount
fulfilling Deuteronomy 18:15–18. Matthew 11:29, 30 and 18:20 also
support the idea of Jesus as the prophet like Moses promised by Moses,
though clearly emphasizing that Jesus is much greater than Moses.
5 In cleansing the Temple Jesus may have had Malachi 3:1 in mind. In
addition, John the Baptist as forerunner may have had such passages as
Isaiah 40:3 and Malachi 3:1 in his mind, with Jesus identifying John as
Elijah based on Malachi 4:5.
6 Jesus entering Jerusalem on a donkey in Mark 11:1–10 and parallels
fulfilling Zechariah 9:9–10.
Here I make the point that it is no less ‘fulfilment’ that Jesus knew what he
was doing.
(c) Prophecies whose fulfilment depends on the truth of

Jesus’ miracles
1 Jesus’ healing miracles in Matthew 8:14–17 and parallels fulfilling
Isaiah 53:4.
2 Isaiah 42:7, ‘to open the eyes that are blind,’ is fulfilled literally through
Jesus healing the blind (e.g. Mk 8:22–26, 10:46–52).
3 Isaiah 42:7 and 61:1–3, release of captives, good tidings for the afflicted,
binding up of the brokenhearted: these are fulfilled in Jesus’ ministry
of healing, exorcism, and forgiveness of sins, e.g. the kyphotic woman
‘bound’ by Satan for eighteen years (Lk 13:10–17).
4 The feeding of the five thousand mirrors God’s feeding the Israelites
with manna in the wilderness (perhaps only retrospectively seen as pro-
phetic, though in John’s account the witnesses clearly took it to be Mes-
sianic, Jn 6:14–15).
(d) Prophecies whose fulfilment depends on the truth of

Jesus’ resurrection
1 The claim that Jesus died for our sins (Mk 10:44–45, 14:24 but widely
elsewhere in the New Testament) fulfilling Isaiah 53:5, 8, 10. That this
was a victory not a defeat is vindicated by his resurrection.
2 The fulfilment of Israel’s destiny to be a ‘light to the nations’ (Is. 42:5–7,
49:6). The pre-Easter Jesus heals the occasional Gentile (e.g. Mk 7:24–
30), and says that ‘many will come from east and west and sit at table
with Abraham, Isaac, and Jacob in the kingdom of heaven. . . ’ (Mt
8:11–12, Lk 13:28, 29). The risen Jesus commands his disciples, ‘Go
therefore and make disciples of all nations. . .,’ and this is the task we
see them beginning to fulfil in the book of Acts.
3 The enigmatic hint at the Suffering Servant’s resurrection in Isaiah
53:10–12. Of these verses Westermann writes:
Many editors (e.g. Mowinckel) are quite sure that what is here reported
is the Servant’s resurrection from the dead. As against this it should
be noticed that this is not clearly said to be the case. There is no
doubt that God’s act of restoring the Servant, the latter’s exaltation,
is an act done upon him after his death and on the far side of the
grave. But no attempt is made to be precise or to explain.33
The uniqueness of this seeming resurrection, if rather veiled, of an indi-
vidual in the Old Testament is quite remarkable. The only other clear
prophecy of resurrection is of the final universal resurrection in Daniel 12.
4 Isaiah 11:4: ‘he shall smite the earth with the rod of his mouth, and
with the breath of his lips he shall slay the wicked;’ and Psalm 2:8–9:
‘I will make the nations your heritage, and the ends of the earth your
possession. You shall break them with a rod of iron, and dash them in
pieces like a potter’s vessel.’ These are fulfilled metaphorically by Jesus’
death and resurrection, which are seen as the victory over evil and the
vindication of Jesus’ mission.
5 Daniel 7:13–14:
I saw in the night visions, and behold, with the clouds of heaven there
came one like a son of man, and he came to the Ancient of Days and
was presented before him. And to him were given dominion and
glory and kingdom, that all peoples, nations, and languages should
serve him; his dominion is an everlasting dominion, which shall not
pass away, and his kingdom one that shall not be destroyed.
As noted earlier, this can only be seen as fulfilled in the light of Jesus’
whole life, death, and resurrection, the spread of the gospel in subse-
quent history, and indeed it must be recognized as incompletely ful-
filled until the eschaton. However, the story of Jesus’ Ascension (Acts
1:9), and various allusions to the Son of Man coming in the clouds in
the New Testament (e.g. Mk 14:62, Stephen’s vision in Acts 7:55–56),
would lend support to this prophetic fulfilment, seen in terms of Jesus’
resurrection (and ascension) vindicating his death for sins.
6 Jesus’ own clear prophecy of his resurrection (Mk 8:31) is of course
only fulfilled if he did indeed rise.
In evaluating the above, it seems that we have several prophecies over which
Jesus had no control which it is hard to dispute, such as his scourging,
mockery, crucifixion between thieves, burial in a rich man’s grave, his own
prophecy of his death (thus not including his resurrection at this stage),
his prophecy of future persecution of his disciples, and his prophecy of the
destruction of the Temple. There is also the prophecy of the virgin birth, for
which I have admittedly given just the briefest outline for justification. There
are more but in the space available it is quite impossible to be exhaustive.
It seems to me that it is very unlikely that we would have these prophecies
fulfilled in the way they are purely by accident. On the hypothesis that the
prophets had genuine insight into the purposes of God they are much more
likely to be fulfilled, especially in combination. If Jesus is the one whom God
has sent to bring about those purposes, then we should not be surprised to
see the prophecies fulfilled in him. We saw in Chapter 6 how testimony can
combine rapidly substantially to enhance the prior probability of a miracle,
and the case of prophecy is similar.
C. S. Lewis noted the analogous case of Plato having a real insight into
the reality of human nature. Anticipating the results of the next three chap-
ters, I think the multiple prophetic testimony of the various kinds we have
outlined takes us considerably beyond Plato and speaks of a deeper contact
with reality, namely to the God who did indeed send Jesus. The full for-
malization in Bayesian terms will come in Chapter 11 when I draw all the
strands of evidence together.
Notes
1 John Barton and John Muddiman (eds.), The Oxford Bible Commentary
(Oxford: Oxford University Press, 2001), 9.
2 C. S. Lewis, Reflections on the Psalms (Glasgow: Collins, Fount Paperbacks edi-
tion, [1961] 1977), 88.
3 Plato, Republic (London: J. M. Dent and Sons, 1935), II.362, 39.
4 Indeed, it is used by Philo in this sense: Philo, Flaccus, 83, in The Works of Philo,
trans. C. D. Yonge (Peabody, MA: Hendrickson, 1993), 725–741, 732.
5 Brevard Childs, ‘Prophecy and Fulfilment: A Study of Contemporary Hermeneu-
tics,’ Interpretation: The Journal of Bible and Theology 12(3) (1958), 259–271,
263–264.
6 Childs, ‘Prophecy and Fulfilment,’ 267.
7 Richard T. France, The Gospel of Matthew: The New International Commen-
tary on the New Testament (Grand Rapids, MI: Wm. B. Eerdmans, 2007), 11.
8 France, The Gospel of Matthew, 12–13.
9 C. F. D. Moule, The Origin of Christology (Cambridge: Cambridge University
Press, 1977), 105–106.
10 Claus Westermann, Isaiah 40–66: A Commentary (London: SCM Press,
1969), 180.
11 Westermann, Isaiah 40–66, 93.
12 Morna D. Hooker, The Gospel According to St Mark (London: A & C Black,
1991), 22.
13 It is possible that ‘rich’ is a corruption of another word meaning ‘wicked;’ how-
ever, the Septuagint has ‘the rich’ (plural).
15 Westermann says that scholars are unsure of the interpretation of this passage,
and the relationship of the one called in v.6 to the servant in Isaiah 42:1–4: West-
ermann, Isaiah 40–66, 98.
16 Jürgen Moltmann, The Crucified God (London: SCM Press, 1974).
22 The historical background to the book of Daniel is given, for example, in André
Lacocque, The Book of Daniel (London: SPCK, 1979).
23 This is a major them for Wright. To give just one quotation: ‘Babylon had taken
the people into captivity; Babylon fell, and the people returned. But in Jesus’ day
many, if not most, Jews regarded the exile as still continuing.’ N. T. Wright, Jesus
and the Victory of God: Volume 2 of Christian Origins and the Question of God
(London: SPCK, 1996), 126.
24 W. D. Davies, Paul and Rabbinic Judaism (London: SPCK, [1948] 1955),
147–150.
25 Davies, Paul and Rabbinic Judaism, 150.
26 Hooker, Gospel According to St Mark, 71–72.
28 See especially Peter W. L. Walker, Jesus and the Holy City: New Testament Per-
spectives on Jerusalem (Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerd-
mans, 1996), 161–200.
30 Wright, Jesus and the Victory of God, 493–510.
32 C. K. Barrett, The Gospel according to St John: An Introduction with Com-
mentary and Notes on the Greek Text, second edition (London: SPCK, 1978),
347–348.
9 Ramified natural theology
in action
Outline of the argument for the
historicity of the resurrection
of Jesus
Introduction
In this chapter I shall look at some of the reasons and evidence for the
most important miracle of all in Christian belief, namely the resurrection
of Jesus. I shall consider St Paul’s testimony in 1 Corinthians 15, what the
gospels and Acts have to say (including differences between accounts), and,
to a limited extent what certain extra-Biblical sources contribute. I shall ask
what alternative explanations for the evidence there might be, and argue
that none do justice to the texts. I shall take into account modern Biblical
scholarship, including challenging more sceptical views, though much of
my critique of the latter has been covered in previous chapters. I shall show
that it is indeed rational to believe in the truth of the resurrection, though
the logic of the argument will be presented more rigorously in the next two
chapters when I consider and apply the Bayesian approach.
The importance of the topic

In Chapter 7 I showed that the hypothesis that Jesus performed miracles
is vastly superior to the denial of this in explaining the pertinent evidence.
Indeed, given reasonable assumptions, it is overwhelmingly probable that
Jesus performed miracles, though, as indicated, I shall make that statement
more rigorous in Chapter 11. However, the most important miracle of all
for Christians is not any miracle which Jesus performed during his ministry
while walking the roads in Galilee and Judaea, but it is his own resurrection
from the dead.
The resurrection of Jesus is the sine qua non of the Christian faith. That
is to say, Christianity stands or falls on the truth or falsehood of the claim
that Jesus rose from the dead. St Paul puts this particularly clearly in his
first Epistle to the Corinthians: ‘If for this life only we have hoped in Christ,
we are of all men most to be pitied.’ (1 Cor 15:19). To this day Christians
regularly recite these words, from the Apostles’ Creed: ‘The third day he
rose again from the dead.’ A ‘demythologized’ existentialist interpretation
of the New Testament may satisfy Bultmann and his followers, but this is
Ramified natural theology in action 171
not classic Christianity and it is not how Christians in general interpret
the data.
We have already cleared aside some objections to this. We have seen that
a scientific point of view does not preclude belief in miracles per se, and we
have seen that evidence from testimony can rationally justify such belief, by
making the occurrence of a miracle or miracles probable. Indeed, much of
what we believe comes from testimony rather than from science; of course
it also comes from our own experience and from the delivery of memory.
Richard Swinburne distinguishes what he calls the ‘Principle of Credu-
lity,’ whereby I ought normally to believe my own experiences, with what
he calls the ‘Principle of Testimony,’ that it is rational for me normally to
believe what others tell me. Of course, both our own experience and other
people’s testimony can be mistaken, but we cannot sensibly operate in a
world in which scepticism is the default position.1 Richard Bauckham says
something similar: ‘Trusting testimony is not an irrational act of faith that
leaves critical rationality aside; it is, on the contrary, the rationally appro-
priate way of responding to authentic testimony.’2 Both Swinburne and
Bauckham allude to the seminal work on testimony by James Coady, for
whom ‘our trust in the word of others is fundamental to the very idea of
serious cognitive activity,’3 and to the precedent for this modern view in the
so-called ‘common sense’ philosophy of Scottish Enlightenment philosopher
Thomas Reid. We noted in Chapter 7 how the Principle of Testimony con-
trasts with the ‘hermeneutic of suspicion’ which arose in the early days of
historical-critical study of the Bible.
Now the resurrection of Jesus is on any account an exceptional event,
indeed more exceptional than any of the miracles which Jesus himself
performed—including even his own miracles of raising several individuals
from the dead, since these persons presumably went on to die in the normal
way. Philosophers say that an exceptional event needs exceptional evidence.
‘A wise man proportions his belief to the evidence,’ says Hume.4
I shall look at whether the evidence for the resurrection satisfies the phi-
losophers’ requirements in due course. But I think we already have very
good reason to see how proportioning one’s belief to the evidence works. In
Chapter 6 I gave the example of eggs bouncing. For most of my congrega-
tion a bouncing egg would count as an exceptional event and thus, philo-
sophically speaking, demand exceptional evidence. But the congregation did
indeed ‘proportion their belief to the evidence’ when witness after witness
testified to having seen eggs bounce. This example illustrated how we can
build up evidence of testimony to make a case. The more witnesses we have
to an event, saying that it happened, the more likely it is to be true that it
happened. And in chapters 7 and 8 we saw qualitatively how this might
work in practice for the miracles of Jesus and the fulfilment of prophecy. In
this chapter, we shall just recite the main evidence for Jesus’ resurrection,
again pointing to many witnesses; and then in the next two chapters we
shall look at the Bayesian formulation of the argument.
172 Ramified natural theology in action
Evidence from outside the New Testament
The evidence for Jesus’ resurrection comes overwhelmingly from the New
Testament. And within that the main evidence is to be found in the four
gospels, Matthew, Mark, Luke, and John, and from St Paul who, in fact,
wrote a large part of the New Testament in the form of letters. Before we
come to that it is worth asking if there is any evidence from outside the New
Testament, for example from Jewish or Roman historians, which confirms
or contradicts what the New Testament says.
There is very little at all which relates directly to Jesus’ resurrection. So
this section is more a summary of references which relate to Jesus or the
New Testament more generally. We saw in Chapter 7 that the one explicit
reference to Jesus’ resurrection in Josephus is generally regarded by schol-
ars as a Christian interpolation, though it is generally agreed that Josephus
described Jesus as a ‘wise man’ and a ‘doer of wonderful works.’ Elsewhere
Josephus does express some sort of belief in resurrection in general, in line
with his Jewish contemporaries, the Pharisees.5 He does also refer indepen-
dently to other events which are described in the gospels. Thus he tells us
in some detail about John the Baptist, his imprisonment and execution by
Herod Antipas.6 Then he writes about how the high priest Ananus sum-
moned the Sanhedrin to try James, the brother of Jesus ‘who was called
Christ’—James was sentenced to stoning.7 Dunn notes that few scholars
have doubted the authenticity of this particular passage.8
This story is told in more detail by the second century Jewish convert to
Christianity, Hegesippus.9 Hegesippus gives the reason for James’s death as
his refusal, at the behest of the ‘Jews and scribes and Pharisees,’ to persuade
the people against the belief that Jesus is the Christ. He did the opposite, so
that many responded, ‘Hosanna to the Son of David.’
James’s death is probably to be dated in AD 61 or 62 from the data pro-
vided by Josephus. He was the leading figure of the Jerusalem church and
had been well-respected by his fellow Jews up to this point. Tom Wright
makes the telling point that no one ever called James the Messiah, as one
might have expected if Jesus had simply been a failed Messianic pretender
who died on the cross and stayed dead. No, he is described as ‘the brother
of Jesus, who was called Christ,’ the appellation ‘Christ’ being reserved for
Jesus alone.10
Even though Christianity arose in a remote outpost of the empire, and
was still a small sect in the first century, there are a couple of references
to Christ from Roman historians. The greatest Roman historian Tacitus
(AD 56—120) writes that Nero falsely blamed the Christians for the great
fire of Rome in AD 64 and punished the Christians for it ‘with the most
exquisite tortures.’ He writes: ‘Christus, from whom the name had its ori-
gin, suffered the extreme penalty during the reign of Tiberias at the hands of
one of our procurators, Pontius Pilate.’11
Tacitus can hardly be called a friend of Christianity since in the same pas-
sage he calls the Christians ‘a class hated for their abominations.’ He further
states that the result of Nero’s punishment was that ‘a most mischievous
superstition’ was thus checked, only to break out again, ‘not only in Judaea,
the first source of the evil, but even in Rome, where all things hideous and
shameful from every part of the world find their centre.’
Dunn notes that, had this information come to Tacitus from Christian
sources, we would have expected him to put in a disclaimer by describing
Jesus as one ‘whom they called Christ’ or similar. As it is, he treats the epi-
thet ‘Christ’ as simply a surname. He would also probably have described
Jesus as crucified rather than simply ‘executed.’12
The historian Suetonius (AD 69—122), while not explicitly mentioning
the fire as the cause says this: ‘Punishment was inflicted on the Christians, a
class of men given to a new and mischievous superstition.’13
The great fire of Rome took place in AD 64. An earlier event was the
expulsion of Jews from Rome under Claudius. Suetonius states: ‘Since
the Jews constantly made disturbances at the instigation of Chrestus, he
expelled them from Rome.’14
‘Chrestus’ is probably a simple misspelling of Christus owing to the pro-
nunciation of the two words being similar, and the disturbance may have
been due to Jewish-Christian controversy. The expulsion took place in
about AD 49 and is also recorded in the New Testament in Acts 18:2, when
Paul meets Aquila and Priscilla in Corinth at that time, this couple having
come from Rome. Note that this occurs only about sixteen years after Jesus’
crucifixion. I do not claim that ‘at the instigation of Chrestus’ necessarily
implies that Suetonius thought Christ to be alive but, at the very least, it
does not contradict the resurrection.
Pliny the Younger (AD 61—113), in AD 112 writes this: ‘They [Chris-
tians] were in the habit of meeting on a certain fixed day before it was light,
when they sang an anthem to Christ as God . . . ’15 Given that Christian-
ity arose out of Judaism, which very strongly believed in one God, that
Christians worshipped Christ as God demands an explanation. That they
believed he rose from the dead would certainly provide the explanation
since it would confirm everything that he had said about himself. Moreover,
we have here a reference to Christians worshipping regularly together. This
was most probably on a Sunday (though that is not explicit from Pliny),
which would be in keeping with their belief that Jesus rose on the first day
of the week—I shall say a little more about this later. Again, Pliny is no
friend of Christianity: as governor of Bithynia he is writing to the Emperor
Trajan asking for advice on how to deal with this troublesome but now very
numerous sect.
These non-Christian authors admittedly do not tell us a great deal,
although what they do say is entirely in keeping with what we read in
the New Testament, and the last in particular chimes in nicely with the
Christian claim that Jesus rose from the dead. Indeed, while they might
describe Christianity as ‘superstition,’ none of these authors debunks the
story of Christian origins in the New Testament or the resurrection claim in
particular, nor is any other explanation offered, say for Christians worship-
ping Christ as God. They are thus entirely in line with the extra-Biblical
references we discussed in Chapter 7.
The New Testament evidence: overview

The main evidence for the resurrection of Jesus comes from the New Testa-
ment. We shall think first about the evidence from the Gospels, so let us
begin with a few remarks by way of overview. We learned in Chapter 7 of
the scholarly consensus on their dating. Given that the crucifixion took place
in April AD 33 or possibly AD 30, we can appreciate Richard Bauckham’s
point that they were written within living memory of the events recounted.
In Chapter 7 we discussed Marcan priority and the dependence of Mat-
thew and Luke on Mark, together with a hypothetical document ‘Q.’ We
also briefly noted that early second century writer Papias describes Mark as
obtaining his information from Peter in Rome. Papias says that he got this
information from John, the Elder. Peter, the chief of the disciples, was killed
in Rome for his Christian faith in AD 64 so this makes geographical and
chronological sense. Papias’ account runs as follows:
This also the presbyter [Elder] said: Mark, having become the inter-
preter of Peter, wrote down accurately, though not indeed in order,
whatsoever he remembered of the things said or done by Christ. For
he [Mark] neither heard the Lord nor followed him, but afterward,
as I said, he followed Peter, who adapted his teaching to the needs of
his hearers, but with no intention of giving a connected account of the
Lord’s discourses, so that Mark committed no error while he thus wrote
some things as he remembered them. For he was careful of one thing,
not to omit any of the things which he had heard, and not to state any
of them falsely.16
The Elder here is John, described earlier by Papias as a disciple of Jesus.

Bauckham sees Papias getting his information from John around AD 90,
when Papias could have been twenty years old or so, but writing his account
down some years later.17 Papias’ account is supported by the internal evi-
dence of Mark’s gospel, which, as Bauckham says, ‘frequently adopts Peter’s
perspective.’18
Papias also says of Matthew’s gospel: ‘So then Matthew wrote the oracles
in the Hebrew language, and everyone interpreted them as he was able.’19
This passage should, from the Greek, read, ‘Matthew put the oracles into
an ordered arrangement (συνετάξατο) in the Hebrew language . . . ’ Impor-
tantly, Papias is referring to Matthew as an eyewitness, unlike Mark, so it
is probable—at least if one takes Papias seriously and it is hard to see why
one should not—that two of our gospels (Matthew and John) come directly
from eyewitnesses. However, it is important to add that what we possess
in Matthew is in Greek, presumably a translation—and therefore at one
remove—from Matthew himself.20
Luke begins his gospel with a preface in which he sets out his aims (Lk
1:1–4). He writes of many besides himself who have set down orderly
accounts of the events that have taken place, as handed down by ‘those
who from the beginning were eyewitnesses.’ His intention, having carefully
researched matters, is to impart to Theophilus the truth of what he has
been told. His gospel and the book of Acts contain much historical material
directly verifiable from pagan sources. That includes, for example, his list-
ing of the rulers at the time of Jesus’ ministry, which reads just like similar
listings in other ancient historians, and Claudius’ expulsion of the Jews from
Rome, as noted above. Luke was not one of the twelve apostles, but, from
passages in Acts, we can identify him as a companion of St Paul on the lat-
ter’s missionary journeys. Acts is Luke’s second volume, dealing with the
spread of the church following the resurrection, though there are also more
resurrection appearances at the beginning of Acts.
As noted in Chapter 7, John’s gospel is generally reckoned to date from
the AD 90s. In many ways it is the most theologically sophisticated of the
gospels, for example identifying Jesus with the Word, or Logos (λόγος) in
Greek, which Greeks would have understood to mean the rational principle
of order behind the universe. While the dates I have given would be com-
mensurate with most New Testament scholarship, they are not cast in stone.
For example, the radical theologian and New Testament scholar John Rob-
inson argued for dating the whole New Testament before the destruction of
the Jerusalem Temple by the Romans in AD 70,21 even arguing that John
was the first of the gospels to be written.22
All the gospels contain accounts of Jesus’ crucifixion and resurrection,
though the earliest manuscripts of Mark end at the point where the women
find the empty tomb and are simply told that Jesus has risen. The best evi-
dence is to be found in chapter 15 of Paul’s first letter to the Christians in
Corinth (1 Cor 15). This was written by St Paul in about AD 55. This date is
worked out from Luke’s account of Paul’s travels around the Mediterranean,
which is found in Acts. Thus this evidence is earlier than that of the gospels.
In it Paul gives a list of witnesses to Jesus’ resurrection. He also describes his
own direct experience of meeting the risen Christ, which he repeats in another
letter, to the Galatians, and which is repeated three times by Luke in Acts.
I shall go into a bit more detail shortly on what the actual evidence is. It is
eyewitness testimony and reported eyewitness testimony. One point to clear
up at this early stage is the fact of differences between the different gospel
accounts. This disagreement is in small details such as how many women
there were at the tomb. But consider any court case. If the witnesses all agreed
in every detail one would be led to think that there was a conspiracy. Differ-
ent witnesses to the same event will generally disagree about details, so this
disagreement in the gospels is more likely to tell us that we have independent
accounts from different sources than it is to tell us that what is described is
fabricated by a group of conspirators. There are other reasons for rejecting the
latter claim which I shall come to in due course. It might also be worth say-
ing that, here as indeed in previous chapters, we are treating the Biblical texts
just like any others in the ancient world, much as has been done since the rise
of the historical critical approach—though without any anti-supernaturalist
prejudice. Thus we are not coming to them with any commitment to pre-
conceived notions of infallibility or inerrancy, which are quite unnecessary to
making the case for what I have described as their ‘general reliability.’
The New Testament evidence: chapter and verse

There are two main strands to the evidence. The first is that Jesus’ tomb was
empty on the first Easter Day. We are told that when Jesus was buried, in a
tomb belonging to Joseph of Arimathea, a great stone was placed over the
entrance and a guard put in place (Mt 27:60–66). The second main piece of
evidence is that Jesus appeared alive to a number of different people after
his death. We now set out who witnessed the empty tomb and who met the
risen Jesus, starting with the empty tomb.
Witnesses to the empty tomb

Table 9.1 summarizes the various witnesses as described in each of the gos-
pels, and we now give references to where these witnesses are named.
Gospel of Mark
MARK 16:1, 2
Mary Magdalene, Mary the mother of James (also of Joses: Mk 15:40 and
Mk 15:47), and Salome buy spices and go to the tomb.
Table 9.1 Witnesses to the empty tomb.
Empty tomb Witnesses
Mark Mary Magdalene

Mary the mother of James (Mk 16:2; cf Mk 15:40 ‘Mary, the
mother of James the younger and of Joses,’ and Mk 15:47 ‘Mary
the mother of Joses’)
Salome
Matthew Mary Magdalene
‘the other Mary’ (Mt 28:1; cf Mt 27:56 ‘Mary the mother of James
and Joseph,’ and Mt 27:61 ‘the other Mary’)
Luke Mary Magdalene
Joanna
Mary the mother of James
The other women
John Mary Magdalene
John
MARK 16:3–8
As the women are asking themselves who will roll the stone away, they
arrive to find that the stone has already been rolled away, and inside sitting
at the right they see a young man dressed in a white robe. The young man
says Jesus has risen. He says: ‘Go, tell his disciples and Peter that he is going
before you to Galilee; there you will see him, as he told you.’ They fled from
the tomb and said nothing to anyone, for they were afraid.
Gospel of Matthew
MATTHEW 28:1–7
Mary Magdalene and ‘the other Mary’ went to the tomb. There was an earth-
quake and an angel descended from heaven and rolled back the stone. ‘For
fear of him the guards trembled and became like dead men.’ The angel, whose
appearance was like lightning, and clothing as white as snow, said: ‘He is not
here; for he has risen . . . Come see the place where he lay. Then go quickly and
tell his disciples that he has risen from the dead, and behold, he is going before
you to Galilee; there you will see him.’ We are told that the women departed
quickly from the tomb with fear and great joy, and ran to tell the disciples.
As intimated earlier, acknowledging the general reliability of the gospels does
not imply that we are committed to taking every aspect of a gospel writer’s
description as historical. In Chapter 7 I mentioned in particular the story in
Matthew’s gospel of the earthquake and many rising from the dead at Jesus’
crucifixion. Here two days after that is a second earthquake recorded by Mat-
thew. Tom Wright notes the similarity that on each occasion the earthquake is
accompanied by the opening of a tomb. After discussing various options for
interpreting the first earthquake, he notes on the one hand that there is only one
source for it, which may make it historically suspect. On the other hand, while
there is allusion to Old Testament prophecies, ‘we may doubt whether stories
such as this would have been invented simply to “fulfil” prophecies that nobody
had understood this way before.’23 The same arguments presumably apply to
the second earthquake. Wright concludes that ‘it is better to remain puzzled than
to settle for either a difficult argument for probable historicity or a cheap and
cheerful rationalistic dismissal of the possibility.’24 In any case, nothing hinges
on this if there is sufficient agreed testimony to the central details of the story.
Gospel of Luke
LUKE 24:1–11
The women went to the tomb, found the stone rolled away, and went in but
did not find the body. Two men stood by them in dazzling apparel. They
imply that Jesus is risen as he told them he must. The women return and tell
‘the eleven and all the rest.’ The women are Mary Magdalene, Joanna, Mary
the mother of James, and the other women with them.
Clearly there is a great deal in common between the synoptic accounts,
notably the unequivocal place of Mary Magdalene, as indicated in Table 9.1.
The most glaring difference is that which we have alluded to above, the
earthquakes in Matthew.
Gospel of John
JOHN 20:1, 2
Mary Magdalene goes to the tomb early on the first Easter morning and
finds the stone rolled away. She reports to Peter and the disciple whom Jesus
loved (i.e. John), saying, ‘They have taken the Lord out of the tomb, and we
do not know where they have laid him.’
JOHN 20:3–10
Peter and John go to the tomb and find it empty. We are told that John ‘saw
and believed; for as yet they did not know the scripture that he should rise
from the dead.’
Appearances of the risen Christ

Table 9.2 summarizes the various witnesses to Jesus’ appearances after his
death, as described in each of the gospels, in Acts, and in St Paul’s first letter
to the Corinthians. We now give references to where these witnesses are listed.
Table 9.2 Resurrection appearances.
Resurrection Witnesses
Appearances
Long ending of Mark Mary Magdalene

Two walking in the country
The eleven
Matthew Mary Magdalene
Mary the mother of James and Joseph
The eleven
Luke Cleopas and his companion
Peter
Cleopas, his companion, plus the eleven, plus those with
them
Acts of the Apostles Apostles
Joseph called Barsabbas, surnamed Justus
Matthias
John Mary Magdalene
Ten disciples (Thomas absent)
The eleven (Thomas present)
Seven disciples in Galilee
Paul in Peter, the twelve,
1 Corinthians 15 more than five hundred brethren at one time,
James, all the apostles,
Paul himself
Long ending of Mark’s gospel
MARK 16:9–20
The earliest manuscripts of Mark end with women finding the empty tomb
and simply being told that Jesus has risen. The gospel ends very abruptly
at verse 16:8, so it looks as if the original ending was lost and someone has
added some material at the end, with resurrection appearances correspond-
ing to those in the other gospels.
In the long ending Jesus appeared first to Mary Magdalene. She went and
told the disciples and they did not believe her. Then Jesus appeared to two
walking in the country, which sounds very much like a similar story in Luke,
which we shall come to in a moment. Finally, Jesus appeared to the eleven
as they sat at table.
Gospel of Matthew
MATTHEW 28:9–10
The risen Jesus met the women, Mary Magdalene and Mary the mother of
James and Joseph, who told them to tell his disciples to go to Galilee where
they will see him.
MATTHEW 28:11–13
The chief priests bribe the guards to say that the disciples stole the body
while they were asleep.
MATTHEW 28:16–20
Jesus appears to the eleven in Galilee. He instructs them ‘to make disciples
of all nations, baptizing them in the name of the Father and of the Son and
of the Holy Spirit,’ a highly significant Trinitarian formula at the outset of
the church’s history.
Gospel of Luke
LUKE 24:13–32
Jesus comes alongside two on the road to Emmaus, one of whom is named
Cleopas. They do not recognize him at first, but they do when he breaks bread
before them. The language of Luke recalls the Last Supper. This took place on
the last night Jesus spent with the apostles before his crucifixion, and it marks
the inauguration of the Eucharist (Lk 22:19). Whilst Cleopas and his com-
panion were not at that meal they would no doubt have seen Jesus’ character-
istic action on other occasions, such as the feeding of the multitude (Lk 9:16).
LUKE 24:33–35
The two return, having seen Jesus, to ‘the eleven and those who were with
them,’ who say that the Lord has risen and appeared to Simon, that is,
Simon Peter. Thus the first appearance may have been to Peter. Cleopas and
his companion relate their own experience of meeting Jesus and recognizing
him in ‘the breaking of the bread.’
LUKE 24:36–51
Jesus appears to Cleopas and his companion with the eleven and those who
were with them, as they are narrating the Emmaus road experience. Jesus says,
‘Handle me, and see’ and eats broiled fish. He tells them to wait in the city
‘until you are clothed from on high,’ but then leads them to Bethany where
he blesses them and is carried up to heaven. It therefore looks as though the
ascension occurs on the same day, but that is presumably just an abbreviation
of Luke’s longer account in his volume 2, the Acts of the Apostles.
Acts of the Apostles

ACTS 1:3–5
Luke tells us that ‘To them [the apostles] he [Jesus] presented himself alive
after his passion by many proofs, appearing to them during forty days.’ He
tells them to wait in Jerusalem for ‘the promise of the Father,’ and says,
‘before many days you shall be baptized with the Holy Spirit.’
ACTS 1:6–12
After the forty days Jesus ascends into heaven. Then the apostles return to
Jerusalem from the mount called Olivet.
ACTS 1:13–14
The apostles are named: Peter, John, James, Andrew, Philip, Thomas, Bar-
tholomew, Matthew, James son of Alphaeus, Simon the Zealot, Judas son
of James. They devote themselves to prayer, together with ‘the women and
Mary the mother of Jesus, and his brothers.’
ACTS 1:21–26
They choose an apostle to replace Judas. The replacement had to meet the
condition that he had been witness to the resurrection. There were apparently
a number to choose from. We read: ‘So one of the men who have accompa-
nied us during all the time that the Lord Jesus went in and out among us,
beginning from the baptism of John until the day when he was taken up from
us—one of these men must become with us a witness to his resurrection.’
Two candidates are put forward from this group identified by the crite-
rion of having seen the risen Christ: Joseph called Barsabbas, surnamed Jus-
tus, and Matthias. Matthias is chosen. The important point for our analysis
is that the witnesses to the resurrection include these two and others who
have accompanied the apostles from the beginning.
Gospel of John
JOHN 20:11–18
Mary Magdalene sees two angels in white. Then she sees the risen Christ,
but at first mistakes him for the gardener. She recognizes him when he calls
her name, ‘Mary,’ and she goes to tell the disciples.
JOHN 20:19–23
Jesus appeared to the disciples, just ten of them, since Thomas was absent.
JOHN 20:26–29
Eight days later Jesus appeared to the disciples, i.e. the eleven including
Thomas.
JOHN 21:1FF
Jesus appeared in Galilee to Peter, Thomas, Nathanael, James and John (the
sons of Zebedee), and two others of the disciples. Thus there were seven
disciples all told. ‘The disciple whom Jesus loved’ is probably one of the
unnamed ones and he it is who recognizes Jesus, and says to Peter, ‘It is the
Lord!’ (v 7). The Beloved Disciple is named as the author of the gospel in
John 21:24. Bauckham persuasively argues that he is indeed the author and
is to be identified with Papias’ John the Elder rather than John the son of
Zebedee.25
1 Corinthians 15
1 CORINTHIANS 15: 3–9
Paul gives a list of appearances to the following: Cephas (i.e. Peter), ‘the
twelve,’ ‘more than five hundred brethren at one time,’ James, ‘all the apos-
tles,’ Paul himself. He begins the list by stating, ‘For I delivered to you as of
first importance, what I also received, that Christ died for our sins, that he
was buried, that he was raised on the third day in accordance with the Scrip-
tures, and that he appeared to Cephas, then to the twelve . . . ’ Thus, Paul is
passing on a tradition, i.e. teaching that he has received from others, though
his list ends with his own direct, personal experience of the risen Christ. He is
also clear that the Christian faith stands or falls on the truth of this teaching
about the resurrection: ‘if Christ has not been raised, then our preaching is in
vain and your faith is in vain . . . If for this life only we have hoped in Christ,
we are of all men most to be pitied’ (1 Cor 15:14–19).
Piecing the evidence together
Let us try and piece together all this evidence. It is clearly coming from dif-
ferent sources because there are discrepancies between the accounts and not
all the appearances are in all the accounts. But, as I said earlier, one would
expect discrepancies in detail when different witnesses report the same
events. Quite a few of these small discrepancies can easily be ironed out,
though perhaps not all. For example, Joses in Mark is probably the same
person as Joseph in Matthew; and Salome in Mark, looking on from afar
at the crucifixion and witnessing the empty tomb, might be the name of the
mother of the sons of Zebedee in Matthew, noted by him as looking on from
afar at the crucifixion (Mk 15:40 and 16:1; cf Mt 27:56). Morna Hooker
notes that this view of Salome is shared by most commentators, though one
cannot be certain of it, and it may be the case that each evangelist has picked
out women who were best known in their own communities.26
We can put the witnesses into four distinct categories:
1 The first witnesses were women, certainly to the empty tomb and quite
possibly the first witness to the risen Christ was a woman, Mary Mag-
dalene. Mary Magdalene was in fact the chief female witness. Luke
tells us that Jesus had healed her of demon possession (Lk 8:1–3), what
today we would probably call mental illness. The testimony of women
was regarded as unreliable in the first century and, regrettably, they
could not give evidence in court. That is a sad commentary on the cul-
ture of the day. It follows, however, that if one invented a story in the
first century and wanted people to believe it, one just would not have
women as the first witnesses. The fact that women are given so much
prominence by all the gospel writers makes it much more likely that
they really were witnesses, and that what they said is to be counted as
reliable. In Luke’s gospel the two on the road had been amazed by what
the women had said, but they went on to say how some of the men went
to the tomb and found it ‘just as the women had said’ (Lk 24:24).
2 We have the named male disciples of Jesus. This list comprises the origi-
nal twelve disciples apart from Judas, who betrayed Jesus. In addition,
there is Matthias who replaced Judas, and Joseph called Barsabbas and
surnamed Justus who was Matthias’ competitor for the vacancy among
the twelve. These are described as being with the disciples from the
beginning, and being eyewitnesses of the resurrection. Another named
person is James, who was Jesus’ brother.
3 There is the group of five hundred to whom Jesus appeared at one time
as recounted by Paul. Most of these were still alive when Paul wrote, but
some had died. It is no wonder the message of the resurrection spread so
quickly. There were at least two hundred and fifty of this group still around
in AD 55, when Paul was writing, a mere twenty two years after the event,
who could be asked about it and who could testify to it as eyewitnesses.
4 Finally, there is Paul himself. In the list he gives in 1 Corinthians 15
he puts himself last. He describes himself as ‘the least of the apostles’
because he persecuted the church and Christ appeared to him last, after
his ascension into heaven. He met with the risen and ascended Christ
on the road to Damascus, where he was heading to carry out further
persecution. The experience changed him completely. Paul’s Damascus
road experience is described by himself in his letter to the Galatians, and
three times by Luke in the Acts of the Apostles.
We can do some detective work with Paul’s list of resurrection appearances in

1 Corinthians 15. As noted above, he begins by saying, ‘I delivered to you as
of first importance what I also received, that Christ died for our sins in accord-
ance with the scriptures, that he was buried, that he was raised on the third day
in accordance with the scriptures. . .,’ and then he goes on to list the appear-
ances. We can actually ascertain with some accuracy just when Paul would
have received the information which he delivered to the Corinthian Christians.
On his own account, in his letter to the Galatians, Paul was in Jerusalem
three years after his conversion, the experience he had on the Damascus
road. There he saw Peter and stayed with him for fifteen days. He also saw
James during that time. Given that his conversion was probably three years
after the death of Christ at the most, and perhaps much closer to it than
that, it follows that he obtained the information from Peter and James no
more than six years after the events themselves. And there is a further inter-
esting point, namely that Paul is not just saying what happened six years
after the event, jogging his memory, which could possibly, even after such a
short time, be inaccurate. No, he is repeating a formula, a creed that must
already have been in circulation by that date, and probably put together
within a year or two of the first Easter. Already, Christians were doing what
they do today every Sunday, reciting a creed in which they affirm the resur-
rection of Jesus. The Nicene Creed recited at Holy Communion states: ‘For
our sake he was crucified under Pontius Pilate; he suffered death and was
buried. On the third day he rose again in accordance with the scriptures.’ Of
course, Paul adds to that the detail of the appearances.
My illustration about bouncing eggs shows that the more witnesses you
have the more likely it is that what is witnessed actually occurred; and we
have seen how this kind of accumulation of evidence of testimony can be
formalized using the apparatus of Bayesian confirmation theory, in Chap-
ter 6. In fact, as we have seen, it can quite rapidly become overwhelmingly
probable that the event happened. Still, as I promised, we should look at
what possible alternative explanations of the evidence there might be.
Alternative explanations
A number of alternative explanations of the data may be offered and I detail
them here:
1 The women mistook the tomb. The problem here is that they saw where
Jesus was buried. It was a new tomb, quite unmistakeable. In any case,
if they were mistaken someone would soon have corrected them.
Mark tells us (Mk 15:40) that there were women looking on from
afar at the crucifixion, among whom were Mary Magdalene, Mary the
mother of James the younger and of Joses, and Salome. In Mark 15:47
the second of these is described simply as Mary the mother of Joses, and
we are told that she and Mary Magdalene saw where Jesus was laid.
In Matthew’s account (Mt 27:55) there were many women at the
crucifixion, looking on from afar. Among them were Mary Magdalene,
Mary the mother of James and Joseph, and the mother of the sons of
Zebedee. According to Matthew 27:61 Mary Magdalene and ‘the other
Mary’ were sitting opposite the sepulchre when Joseph of Arimathea
laid the body in the tomb and rolled the stone across.
Luke says (Lk 23:49) that at the cross ‘all his acquaintances and the
women who had followed him from Galilee stood at a distance and saw
these things.’ In Luke 23:55 we are told that ‘the women who had come
with him from Galilee saw the tomb, and how his body was laid.’ In
Luke 24:10 the women are named as Mary Magdalene, Joanna, Mary
the mother of James, and the other women with them.
John’s account (19:25–26) has the following at the cross: Mary the
mother of Jesus, his mother’s sister, Mary the wife of Clopas, and Mary
Magdalene. Jesus commends his mother to the ‘disciple whom Jesus
loved’ who was also there (as we have seen, most probably the author
of the gospel: Jn 21: 24). According to John 19:42 Joseph of Arimathea
and Nicodemus laid Jesus’ body in the tomb. John does not explicitly
state that the women were at the tomb. However, all four gospels agree
that it was Joseph of Arimathea’s grave and he at least would have cor-
rected any misunderstanding about where it was.
2 The disciples stole the body. This was the first theory at the time, delib-
erately put about by the Jewish authorities (Mt 28:11–14). But if this
were so, then it would follow that the disciples would have colluded in
a lie. Some went to their deaths for the sake of belief in the resurrection.
It was a dangerous belief to speak about. Yet they spoke about it from
the housetops and were prepared to die for it.
It may be said in response that some people do die for false beliefs.
The terrorists of so-called Islamic State are prepared to die because they
believe they are doing the will of God and will be rewarded in Paradise.
But the early Christian disciples were not prepared simply to die for
a belief but for the sake of a fact, an event which they had, they said,
witnessed with their own eyes. If they had invented the story of the
resurrection, they would then be dying for something they themselves
knew to be false. It is most unlikely that anyone would be prepared to
die for something they positively knew to be false. And surely someone
would have broken ranks and revealed the fraud if this really were a
conspiracy and the disciples had colluded to spread a lie.
3 The witnesses suffered from some form of mass hallucination. This
seems incredibly unlikely. The disciples were disillusioned when Jesus
was executed on the cross. Despite what he had told them, there were
not expecting him to rise from the dead. As Wright notes, those first
century Jews who believed in the idea of resurrection envisaged a final
resurrection of all at the last day, not in the resurrection of one person
in the midst of the present age.27 Moreover, following the crucifixion,
the disciples fled in fear. One also has to consider the large number of
witnesses, as noted above. For so many people, on so many different
occasions, to be deluded into thinking they had seen Jesus alive when he
was dead, seems extraordinarily unlikely—indeed nigh on a psychologi-
cal impossibility. In addition, even to postulate the idea that the appear-
ances of Jesus alive after his death could be hallucinatory is completely
to ignore the first strand of evidence delineated above, namely the fact
that the tomb was empty. As Wright argues, it is the combined evidence
for the empty tomb and the appearances which provide the ‘explana-
tory power’ not possessed by rival explanations (see Chapter 2 for the
technical, Bayesian definition of this term).28
4 Perhaps Jesus did not really die on the cross. Perhaps he somehow sur-
vived as in the swoon and resuscitation theory briefly alluded to by
Charlie Moule, which we noted in Chapter 4. That goes against all the
documentary evidence. In particular it contradicts John’s account, in
which the guards go around breaking the prisoners’ legs to finish them
off, but do not do so when they come to Jesus because he is already
dead. We also saw earlier that Roman historian Tacitus stated that
Christ ‘suffered the extreme penalty during the reign of Tiberias at the
hands of one of our procurators, Pontius Pilate.’ There is no intimation
that he survived the ‘extreme penalty.’ In any case, it is hard to imagine
someone surviving the intense torture and cruelty of three hours nailed
to a cross and then convincing his disciples that he had risen from the
dead. No, surely James Dunn is correct, as we noted in Chapter 4: the
evidence, Biblical and extra-Biblical, puts Jesus’ death by crucifixion
high on the ‘almost impossible to doubt or deny’ scale of historical
‘facts,’ and rightly commands almost universal assent.29
There are other things which are very difficult to explain if indeed the resur-
rection had not happened and been witnessed. One is the incredibly rapid
spread of the good news of Jesus’ death and resurrection. Starting with
thousands converted in Jerusalem shortly after the events, the message of
the resurrection spread right across the Roman empire within a very short
space of time, even to Rome itself, where we learn that Peter and Paul were
martyred in the AD 60s, and where, also in the AD 60s (as noted above),
Christians were blamed for the fire of Rome by the Emperor Nero. Wright
states:
Never before had there been a movement which began as a quasi-

messianic group within Judaism and was transformed into the sort of
movement which Christianity quickly became. Nor has any similar phe-
nomenon ever occurred again. (The common post-Enlightenment per-
ception of Christianity as simply ‘a religion’ masks the huge differences,
at the point of origin, between this movement and, say, the rise of Islam
or Buddhism.)30
Also difficult to explain is the early Christian celebration of the Eucharist

on Sundays (we briefly referred to worship on Sundays above). There is a
reference to ‘breaking bread’—the term used for what Jesus did at the Last
Supper—on the first day of the week in Acts 20:7. In 1 Corinthians 16:2
Paul urges his readers, ‘on the first day of the week,’ to put aside and store
up their contributions for the saints. At the very least this indicates that the
day had already become special to Christians and was probably the day
they would worship, even if those contributions were to be stored at home.
Another passage is Revelation 1:10 which begins, ‘I was in the Spirit on the
Lord’s day . . . ’ The writer then proceeds to relate his vision. The terminol-
ogy ‘the Lord’s day’ is that by which Sunday has been known ever since.
Outside the New Testament, the very early Christian document known as
the Didache, dating from the first or early part of the second century, com-
mands: ‘But every Lord’s day do ye gather yourselves together, and break
bread. . . .’31 Justin Martyr likewise, writing in the mid second century,
describes worship and the celebration of the Eucharist ‘on the day called
“Sunday” ’ (τῇ τοῦ ἡλίου λεγομένῃ ἡμέρᾳ).32 It may be a small piece of the
jigsaw, but surely the best explanation for this new practice is that Jesus
rose from the dead on the first Easter Sunday, elevating Sunday to the most
important day for Christians, rather than the Jewish sabbath.
The evidence for the resurrection: a brief comparison with

scientific evidence
Scientists believe in their theories on the basis of evidence. As we have seen,
there is considerable evidence that Jesus rose from the dead. It is of course
a different kind of evidence from that which scientists generally consider.
It is historical evidence of the testimony of witnesses, and history cannot
be repeated. It is worth saying, however, that even in science we often do
not repeat experiments ourselves but take on trust the published results of
others’ experiments. Moreover, some sciences such as cosmology and evolu-
tionary biology have a historical character—the big bang and the evolution
of complex life forms over millions of years are unrepeatable.
Of course, testimony to the resurrection is different for another reason,
most importantly that what is asserted apparently contradicts scientific laws.
Some scientists have decided a priori that the universe is a closed system in
which only the rigid laws of physics operate and miracles are in principle
impossible, a position which we refuted in Chapter 6. It seems to me to be
unscientific to close down debate before the evidence is considered, since
science itself often throws up many surprises. Indeed, having an open mind
to follow where the evidence leads is a much more scientific attitude than
deciding beforehand what one may think can or cannot happen.
In Chapter 4 we noted Karl Popper’s criterion of falsifiability for a the-
ory to be truly scientific. Widely accepted by scientists, this means that one
could in principle make an observation or do an experiment which would
show the theory to be untrue. Sometimes it is claimed that even if evidence
were discovered that showed that Jesus did not rise from the dead, Chris-
tians would go on believing it. They are not open-minded like scientists.
The accounts we have of the resurrection seem to contradict that point. It
would be difficult for us now to falsify the resurrection, though we could just
imagine a grave being discovered with bones in it that we could definitely
identify as Jesus’ bones. But soon after the event it could certainly have been
falsified. What would Christians have done then? Well, after Paul has listed
the various witnesses to the resurrection, as we noted above, he goes on to
say this: ‘if Christ has not been raised, then our preaching is in vain and your
faith is in vain. We are even found to be misrepresenting God; If Christ has
not been raised, your faith is futile and you are still in your sins. . . . If for
this life only we have hoped in Christ, we are of all men most to be pitied.’
For Paul, the Christian faith stands or falls on the truth of the resurrection
and if it were shown to be untrue, we would be the most pitiable of people.
Paul was certainly open to the resurrection being falsified, but in fact it was
not falsified. There was massive evidence to support it.
The story of Thomas in John’s gospel is pertinent here (Jn 20:24–29).
When Jesus first appeared to the disciples, Thomas was not present. When
later the other disciples told him that they had seen the Lord, he replied that
he would not believe unless he could see Jesus’ wounds and feel them for
himself. Then, eight days later, Jesus appeared to all the eleven including
Thomas, and Thomas was given the opportunity to feel Jesus’ wounds, but,
in fact, he did not need to do so. He simply exclaimed, ‘My Lord, and my
God!’ Thomas actually had the evidence he wanted. But there is more. Jesus
does not exactly condemn Thomas for not believing until he has seen Jesus,
but he does commend those who have not seen and yet believe. He is saying,
not that Thomas should have believed without evidence, but that he should
have believed the testimony of his companions, because they were reliable
witnesses, he knew them intimately, and he knew that they had no reason to
lie to him. And we should do the same, that is, believe the reliable testimony
of those early disciples.
In the next chapter we examine Bayesian formulations of the argument
for the resurrection.
Notes
1 These principles are developed in Richard Swinburne, Epistemic Justification
(Oxford: Oxford University Press, 2001), 123–127, 141–149; and in Richard
Swinburne, The Existence of God, second edition (Oxford: Oxford University
Press, 2004), 303–315, 322–324.
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006), 5.
3 James Coady, Testimony (Oxford: Oxford University Press, 1992), vii; quoted in
Bauckham, Jesus and the Eyewitnesses, 475.
[1748] 1975), 110.
5 E.g., Josephus, The Wars of the Jews, iii.8.5 (374), in The Works of Josephus,
trans. William Whiston (Peabody, MA: Hendrickson, 1987), 656; see N. T.
Wright, The Resurrection of the Son of God: Volume 3 of Christian Origins and
the Question of God (London: SPCK, 2003), 175–181.
6 Josephus, The Antiquities of the Jews, xviii.5.2 (116–119), in The Works of
Josephus, 484.
7 Josephus, Antiquities, xx.9.1 (200), 538.
9 Hegesippus’ account is reported by Eusebius, The Church History of Eusebius,
II.23, in Philip Schaff and Henry Wace (eds.), Nicene and Post-Nicene Fathers,
second series, vol. 1 (Peabody, MA: Hendrickson, 1994), 125–128.
10 Wright, Resurrection of the Son of God, 560–562.
11 Tacitus, Annals, xv.44, quoted in C. K. Barrett, The New Testament Back-
ground: Selected Documents (London, SPCK, revised edition 1997), 15–16.
12 Dunn, Jesus Remembered, 141–142.
13 Suetonius, Nero, 16, quoted in Barrett, New Testament Background, 16.
14 Suetonius, Claudius, 25, quoted in Barrett, New Testament Background, 14.
15 Pliny the Younger, Epistles, x.96, quoted in F. F. Bruce, The New Testament
Documents: Are They Reliable? sixth revised edition (Downers Grove, IL: IVP
Academic, [1981] 2000), 122–123.
16 Papias’ account is reported by Eusebius in The Church History of Eusebius,
III.39, 172–173. Morna Hooker discusses it in Morna D. Hooker, The Gospel
According to St Mark (London: A & C Black, 1991), 5–7.
17 Bauckham Jesus and the Eyewitnesses, 18.
18 Bauckham Jesus and the Eyewitnesses, 204.
19 Eusebius, Church History, III.39, 173.
20 As discussed by Bauckham in Jesus and the Eyewitnesses, 222–230.
21 John A. T. Robinson, Redating the New Testament (London: SCM Press, 1976).
22 John A. T. Robinson, The Priority of John, ed. J. F. Coakley (London: SCM
Press, 1985).
23 Wright, Resurrection of the Son of God, 636.
25 Bauckham, Jesus and the Eyewitnesses, 420–423.
31 Didache, XIV, in Alexander Roberts and James Donaldson (eds.), Ante-Nicene
Fathers, vol. 7, rev. A. Cleveland Coxe (Peabody, MA: Hendrickson, 1994),
369–382, 381.
32 Justin Martyr, First Apology, LXVII, in Alexander Roberts and James Donald-
son (eds.), Ante-Nicene Fathers, vol. 1, rev. A. Cleveland Coxe (Peabody, MA:
Hendrickson, 1994), 163–187, 185–186.
10 On the third day he
rose again
Bayesian methodologies applied
to the resurrection of Jesus
Introduction
In the last chapter I showed that there is considerable historical evidence
in the form of eyewitness testimony to Jesus’ resurrection. This evidence
embraces the epistles of St Paul (especially 1 Cor 15), the gospels, and, to
a much lesser extent, particularly because these do not contradict the Bibli-
cal material, certain extra-Biblical sources. Bayesian confirmation theory,
which we have met earlier, 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, for example by requiring
or not requiring the input of a high probability of theism into the analysis.
They also vary widely in the results they derive. This chapter will re-examine
the issues involved and suggest a resolution to the conflicting methodologies
and results.
In this chapter, then, I compare two very different applications of Bayes-
ian methodology for evaluating the hypothesis that Jesus of Nazareth rose
from the dead, namely the approaches of Timothy and Lydia McGrew1 and
Richard Swinburne.2
The McGrews are content simply to evaluate the historical evidence for
the resurrection without considering the prior evidence for the existence of
God from natural theology. If we can infer from their analysis that the res-
urrection occurred with high probability, we shall be only a short step from
showing that a more detailed hypothesis than plain theism has high prob-
ability, namely both that God exists and has become incarnate in Christ.
And we shall have done that by bypassing considerations of ‘bare’ natural
theology altogether.
Swinburne, in contrast, argues that a high probability for theism is
required as a pre-requisite to evaluating the historical data. We can only
infer a high probability that God was incarnate in Christ and raised him
from the dead if we already have a high probability that God exists in the
first place—hence the importance for Swinburne of bare natural theology.
Moreover, in addition to what Swinburne calls the posterior historical
190 On the third day he rose again
evidence, that directly concerning the resurrection, we also require prior
historical evidence of the life and teaching of the putative God incarnate.
This was the subject of chapters 7 and 8, and indeed in those chapters also
we amassed substantial evidence.
It is a matter of considerable importance as to which of these approaches
is the more successful. Many Christians deny the value of natural theol-
ogy altogether, and in this they are often influenced by Karl Barth, whom
we have met in earlier chapters. Some of these Christians deny the value
of empirical evidence altogether, but many accept the value of empirical
evidence from the Bible, and in particular its accounts of the resurrection,
while not accepting the value of natural theology. For example, David
Wilkinson asserts that ‘the Christian theologian’ agrees with atheist Richard
Dawkins that ‘the design and cosmological arguments do not work.’ On the
other hand, while acknowledging their complexity, Wilkinson is positive in
assessing the gospel narratives as providing reliable evidence concerning the
life, death, and resurrection of Jesus.3
This chapter and the next will show how feeding the results of natural
theology into ramified natural theology can make the resulting conclusion
overwhelmingly probable.
Like both the McGrews and Swinburne, I am not able in a short space
to consider in detail all the Biblical evidence and the wealth of scholarship
relating to it, though we have seen some of it in the preceding three chapters.
Clearly, however, some more discussion is called for in order to see where
the McGrews and Swinburne are coming from. In particular it is important
to recognize that much of the argument of both the McGrews and Swin-
burne, although there is difference in detail, is predicated on what is believed
by sceptical and non-sceptical Biblical scholars alike. For example, it is gen-
erally agreed that factual claims are being made, whether veridical or not,
or whether there are legendary features of the story or not, that Jesus’ tomb
was empty on Easter Sunday and that he appeared alive physically to many
people after his death.
The approach of Timothy and Lydia McGrew to the

Biblical data
The approach and intention of the McGrews is simpler than that of Swin-
burne, so I begin with them. Their aim is simply to evaluate the historical
evidence for Jesus’ resurrection and derive a plausible value of the ‘Bayes
factor,’ also known as the likelihood ratio, which is an expression of how
much more likely this evidence is to pertain given that the resurrection actu-
ally occurred compared with its non-occurrence.
The McGrews argue for the general reliability of the gospels in what they
affirm and, as in court cases, that their reliability is not undermined by small
discrepancies in detail. I would agree with them in this and also in their
assessment that, in the nineteenth century, F. C. Baur and the Tübingen
On the third day he rose again 191
school were much too skeptical and late with regard to the dating of the
gospels; as we saw in Chapter 7, their position was largely undermined by
the work of J. B. Lightfoot in Cambridge.4 However, unlike the McGrews
I regard the modern consensus as realistic. Thus I accept Marcan priority
with Mark written probably in the sixties, Matthew and Luke-Acts in the
seventies or eighties, and John in the nineties. I am also happy to accept,
unlike the McGrews, that Matthew and Luke had both Mark and the hypo-
thetical document ‘Q’ at their disposal when writing their gospels, though
this is still debated in scholarly circles.5
I agree with the McGrews, however, that there is no reason not to take
Papias and Irenaeus at face value in what they say about gospel compo-
sition, a point emphasized by Richard Bauckham.6 I also agree with the
McGrews’ assessment of John Dominic Crossan and his Jesus Seminar col-
leagues. They are, on the one hand, far too hyper-sceptical of the reliability
of the gospels and St Paul (reverting to an unwarranted hermeneutic of sus-
picion like that of the Tübingen school). On the other hand, they are far too
fanciful in their theories and constructions of ‘the historical Jesus,’ which
seem to float entirely free of empirical evidence and seem quite perverse, for
example, in prioritizing non-canonical gospels over the canonical gospels.
We noted Tom Wright’s verdict on the Jesus seminar in Chapter 7. To
reinforce that, I give here a couple of further quotes. On Crossan, Wright
states:
Once you doubt everything in the story, and postulate a chain of events
by which someone might have taken it upon themselves to invent such
a narrative from scratch, all things are possible. But not all things are
probable.7
Referring to Crossan’s view that Jesus’ disciples knew virtually nothing

about how or why he died, except that he was crucified, Wright goes on:
I shall argue in the appropriate place that Crossan’s reading of Jesus’

death is far less likely than a reading in which the synoptic and Johan-
nine traditions, though of course replete with theological and scriptural
allusion, intend to describe things that actually happened, and basically
succeed in this intention. This hypothesis, I submit, makes better sense
of the data, and does so with far more simplicity, than the complex and
convoluted speculation offered by Crossan.8
Of course filling out this argument is a large part of Wright’s whole project.
The important point for now, however, is to note his language of which
hypothesis is more likely (probable), which makes better sense of the data,
and which is simpler. Wright’s approach is thus very much in line with the
scientific approach to evaluating hypotheses, as I have described it, and in
its application to both natural and ramified natural theology.
The mathematics of the McGrews’ approach
The McGrews utilize Bayes’s theorem as follows.
Let R = Jesus rose from the dead, and let E = the historical evidence for
this. Then Bayes’s theorem tells us that
P E|R.P R
P R|E
P E
From equation (2.2) in Chapter 2 it follows that
P E|R.P R
P R|E
P E|R.P R P E| ~ R.P ~ R
The McGrews examine three components of E, namely

W = the reports of the women regarding the empty tomb and the risen
Christ;
D = the testimony of the disciples;
P = the conversion of Paul;
so that E = W ∧ D ∧ P.
Now any actual numbers for the probabilities P[W|R] etc. can at best
be very approximate representations of how much greater we judge the
P[W|R] etc. to be than the P[W|~R] etc. That the P[W|R] etc. are indeed
much greater than the P[W|~R] etc. follows from the arguments we gave in
the last chapter when considering alternative explanations for the data. The
McGrews give the following values:
P[W | R]
= 100
P[W |~ R]
P[D | R]
= 1039
P[D |~ R]
P[P | R]
= 103
P[P |~ R]
It seems not unreasonable to suppose that the testimony of the women is

one hundred times more likely to pertain given that the resurrection hap-
pened than if it did not, and that for an individual disciple or for St Paul,
their testimony is one thousand times more likely to pertain if R is true than
if it is not. Perhaps the testimony of the women does not give quite as high
a Bayes factor since although their testimony to the empty tomb looks very
solid their testimony to seeing the risen Christ is somewhat less so (though
still very likely on the basis of the gospel texts). I think this is a reasonable
judgment given that the there are slight discrepancies in the gospels about
precisely which women are witnesses, and whether they saw the risen Lord,
and no mention of the women in St Paul (though see later).
Clearly the second of these estimates is the most controversial and is the
dominant contributor to the overall Bayes factor. To obtain it, the McGrews
have taken there to be thirteen disciples (the twelve with Matthias substi-
tuted for Judas, plus James the brother of the Lord) each with a Bayes factor
of 103, and multiplied the individual Bayes factors together, assuming inde-
pendence, an assumption we consider in more detail in a separate section
below.
The McGrews could derive an impressively high overall Bayes factor even
if they were to loosen some of their inputs. Thus they could input a smaller
Bayes factor for each disciple, perhaps to take account of Swinburne’s point
that we do not have the direct testimony of each disciple but testimony to
that testimony. But then there are other witnesses besides those included
in their analysis, namely Cleopas and his companion (possibly his wife) on
the road to Emmaus (Lk 24:13–35), Joseph called Barsabbas and surnamed
Justus who was Matthias’ competitor for the vacant apostleship (Acts 1:23),
and the five hundred to whom Jesus appeared at one time, mentioned by
St Paul in 1 Corinthians (1 Cor 15:6). We may not be able to include the
unnamed five hundred in the same way as the other witnesses, multiplying
Bayes factors for each, but their inclusion does provide substantial addi-
tional evidence. We are told that most of them were still alive when Paul was
writing 1 Corinthians (c. AD 53–55),9 but some had fallen asleep.
Assuming that all three pieces of evidence comprising E are independent
(as well as the individual testimonies Di within D) the McGrews multiply
their Bayes factors together, thus arriving at their final result
P[E | R]
= 1044
P[E |~ R]
The McGrews essentially leave it at that and do not consider the prior prob-
ability of the resurrection. However, we can do a little more analysis.
Equation (2.4) in Chapter 2 gives
P R|E P E|R P R
.
P ~ R|E P E| ~ R P ~ R
which tells us that the ratio of the posterior probabilities is equal to the
Bayes factor times the ratio of the prior probabilities. For the evidence to
make the miracle probable, i.e. P[R|E] > ½, we require
P[E | R] P R
. 1
P[E |~ R] P ~ R
If we take the first fraction to be the Bayes factor as calculated by McGrew
and McGrew, then we need
P R 1044 P ~ R ,
i.e. to a high degree of approximation, since P[~R] = 1 – P[R],
P R 1044
Hence, as long as the prior probability of the resurrection is higher than this
tiny value, it is more likely that the resurrection occurred given the evidence
than that it did not occur. Alternatively, the prior probability of the resur-
rection has to be very low indeed, less than 10−44, for it to be more likely that
the resurrection did not occur given the evidence than that it did.
Why would one want to insist on a prior as low as 10−44, unless one real-
ized that a prior lower than this was what was needed to ensure that, at the
end of the day, we could reject the conclusion that P[R|E] > ½ based on the
overwhelming evidence incorporated into the Bayes factor (on the McGrews’
construal of the evidence)? It is hard to appreciate how low a probability
this is. After all, given that the number of people who have ever lived is only
about 108 billion (1011), a probability of resurrection P[R] between 10−44
(McGrews) and somewhat less than 10−11 would be entirely compatible with
the observation that we have not seen any other one. So P[R] ≤ 10−44 is virtu-
ally equivalent to ruling out the resurrection on principle.
Given the very high probability of the resurrection based on the historical
evidence, which seems to follow from the McGrews’ analysis, it is a short
step to assign a high probability to the divinity of Jesus since, if the resurrec-
tion did indeed occur, it would amount to a vindication of the claims made
to divinity by Jesus and his followers. Let us make this argument explicit
utilizing the Bayesian framework.
Let G = God became incarnate in Jesus. Then we wish to estimate P[G|E].
We begin with the formula (cf. equation (6.15) in Chapter 6)
P R|E.P[G | R E]
P G|E
P[R | G E]
On the McGrews’ analysis, and given a prior for P[R] not inordinately low (i.e.
near or less than 10−44), P[R|E] will be very close to 1. We would also expect
P[R|G∧E] to be close to 1 since the resurrection is not less likely if we add
the evidence that God is incarnate in Christ to E. Thus P[G|E] ≈ P[G|R∧E].
And given that E virtually implies R on the McGrews’ analysis, this is surely
equivalent to saying P[G|E] ≈ P[G|R].
Saying that the resurrection happened, considered in isolation, is not
quite the same as saying that God became incarnate in Jesus. It is surely,
however, highly pertinent to calculating a potentially very high probability
for the incarnation. As we shall see shortly, Richard Swinburne goes further
and requires the ‘prior historical evidence’ of Jesus’ life and teaching as well
as the ‘posterior historical evidence’ for his resurrection in order to estimate
the probability that God was incarnate in Jesus. Indeed, and importantly,
he also requires a prior probability of the incarnation given God’s existence,
and a probability of God’s existence from natural theology.
Despite what I have said, there is a weakness in assigning even such a low
probability as I have done to Jesus’ resurrection to complete the McGrews’
analysis. Without further reflection, a question arises: ‘Why would Jesus be
raised and no one else?’ As we argued in Chapter 6, in principle the accumu-
lation of testimony can overcome any prior probability of a miracle, how-
ever low. However, again as intimated in Chapter 6, a better strategy might
be to use other evidence for the existence of God to remove the prejudice
that miracles are as unlikely as anything can be, and then bring the historical
evidence to bear. The logical approach would also entail asking what mira-
cles God is likely to authorize. What reasons would God have for becoming
incarnate, how likely is it that God would do so, and what would we expect
a priori from the life of the putative God incarnate? Having decided on these
questions, with some very rough probability estimates, we can then look at
the historical evidence and see if Jesus or anyone else in history has matched
up to those expectations. This is already to outline Swinburne’s approach,
and we now do so in more detail.
Richard Swinburne’s approach to the Biblical data

Swinburne evaluates similar historical data concerning the resurrection to
the McGrews but also includes evidence from natural theology and ‘prior
historical data,’ that concerning the life and death of Christ. Indeed, it may
come as somewhat surprising that only a relatively small fraction of his
book is devoted to the ‘posterior historical data.’10
Here I shall concentrate mainly on the posterior historical data, since
that is what is common to Swinburne and the McGrews. We saw in chap-
ters 7 and 8 that there is also a substantial body of prior historical data.
Swinburne’s evaluation is similar to the McGrews at many points, though
there are some differences. For example, Swinburne says that testimony
to testimony is less reliable than direct testimony, a point apparently not
taken into account by the McGrews. Much of the testimony we have is
the gospel writers’ or St Paul’s testimony to the testimony of others. One
might ask whether taking this into account would substantially decrease the
McGrews’ probabilities.
It seems to me that the gospel writers and Paul have little or no reason to
misreport the testimony of the eyewitnesses. Luke, for example, emphasizes
how he wants on the contrary to be truthful in giving an ordered account
of ‘the things which have been accomplished among us, just as they were
delivered to us by those who from the beginning were eyewitnesses . . . ’
(Lk 1:1–4). Moreover, in Chapter 7 we noted both the early witness of
Papias to gospel composition, and that there is good reason to suppose that
oral evidence faithfully preserves its original. Suppose, however, that we
reflect the fact that we have testimony to testimony by reducing the Bayes
factor for W, each of the 13 Di, and P (even though for Paul we have his own
direct testimony) by a factor of 10. That would imply multiplying the overall
Bayes factor by (1/10)15, i.e. 10−15, but would still yield an enormous Bayes
factor of 1029. Arguably, however, a factor of 1/10 across the board is much
too low. Indeed, it may well be that the testimony of a reliable witness to the
testimony of another person may increase one’s confidence in the testimony
of that second person. That second person is ‘vouched for’ by the report.
Swinburne, like the McGrews, argues that discrepancies in testimony to
Jesus’ resurrection do not undermine the general reliability of the New Tes-
tament accounts. He acknowledges that there are ‘somewhat different ver-
sions of to whom he appeared, and when,’ but these ‘do little to lessen the
force of that evidence unless they are too much at odds with each other.’11
He attempts some reconciliation between conflicting accounts and, although
he believes that total harmonization is not possible, there is nevertheless
considerable positive evidence that Jesus did indeed ‘appear to many.’ Dif-
ferences in the accounts can be attributed either to a little ‘theologizing’ by
an author or by his ‘having a source whose memory was not totally accu-
rate.’12 In any normal court case there will be discrepancies between the
accounts of witnesses, which do not undermine their general reliability, but
rather reinforce the fact that there has been no collusion. The evidence for
the resurrection is similar.
I agree with Swinburne that the most authoritative account of the appear-
ances is that given by St Paul in 1 Corinthians 15. Indeed, Swinburne’s
account differs only slightly (in date estimates) from mine in the last chapter
(AD 55) and Gordon Fee’s dating (c. AD 53‑55) which I quote above. Swin-
burne argues that not only is this letter very early (c. AD 56), pre-dating the
gospels, but also comparison with St Paul’s letter to the Galatians (cf. last
chapter) enables us to track back to when the apostle received the informa-
tion about the appearances which he lists—and it is very close to the events
themselves, indeed as close as six to eight years. Moreover, Paul seems to
be repeating a formula which was already in currency at the time he met
Peter and James when he visited Jerusalem those six to eight years after the
crucifixion.
Wolfhart Pannenberg, hardly a conservative commentator himself, and
well aware of the weight of liberal Biblical scholarship, also prioritizes 1
Corinthians among the sources in his classic book in which he defends the
resurrection of Christ as a historical event.13 Legendary elements may have
come into the gospel accounts, but 1 Corinthians 15 is straight report-
ing. This may, however, only license belief in the resurrection as visionary
experiences—albeit objective and veridical—of the risen Lord, as seems to
be the case for Paul himself, rather than physical, bodily appearances.
Pannenberg goes on to argue, however, that the gospel kerygma of Jesus’
resurrection could hardly have got going unless the tomb were empty, tes-
timony to which comes in the gospels. That it was empty was agreed upon
both by the disciples and by the Jews who were already accusing the disci-
ples of having stolen the body. To me, as to both the McGrews and Swin-
burne, this strengthens the view that the appearances to the disciples, other
than Paul, were bodily, physical appearances, rather than, as Pannenberg
maintains, objective visions of the truly risen Christ in heaven. Where all
these scholars agree is that purely subjective visions are totally inadequate to
explain the data, which includes the fact that the disciples were not expect-
ing Jesus to rise and were utterly broken following his crucifixion.
Taking account of the appearances listed by St Paul in 1 Corinthians 15
and those described in the gospels, Swinburne allows some minor ‘theolo-
gizing’ to deal with the discrepancies and offers a plausible reconstruction
of to whom and where (both Jerusalem and Galilee) the appearances took
place, with a plausible assignment of the order in which the appearances
occurred. Swinburne believes the appearances were to Mary Magdalene,
Peter, Cleopas and his companion, the eleven, the five hundred, and James
the brother of the Lord. Like the McGrews, he draws attention to the fact
that the testimony of women was deemed unreliable and inadmissible in the
first century, and so not something which would be invented. Thus St Paul
may omit the appearance to Mary because he intended to compile a list of
witnesses which would be formally authoritative. Swinburne also refers to
the choosing of Matthias as a ‘witness’ to replace Judas, a point of impor-
tance to the McGrews.
Swinburne also brings into play evidence that the tomb was empty and
evidence that the Eucharist began very early on to be celebrated on a Sun-
day, the day of resurrection. I agree with Swinburne that the gospel writers
wished to convey that the resurrection was a physical, bodily resurrection.
The fact that St Paul did not refer to the empty tomb is simply because he
assumed it. If the tomb were not empty, St Paul would then have had to
explain why the resurrection was still a fact, in opposition to those who
proclaimed that there was no resurrection and who would have had ammu-
nition if the tomb still contained Jesus’ bodily remains. Both Swinburne and
the McGrews cite Peter’s sermon at Pentecost (Acts 2:29–32) as asserting
a physical resurrection of Jesus in contrast to David, whose ‘tomb is with
us to this day,’ presumably with the bones of David still there. Another
relevant passage would be Acts 10:34–43 in which Peter again speaks of
being one of the eyewitnesses and eating and drinking with Christ after his
resurrection.
Swinburne examines alternative theories to that of the reality of Jesus’
bodily resurrection, as we did in the last chapter, and finds them all highly
improbable in comparison, just as do the McGrews. The idea that Jesus did
not die on the cross has insuperable problems, e.g. that a survivor of cru-
cifixion would be so enfeebled that he would be in no position to convince
others that he had risen from the dead. It is highly implausible that the dis-
ciples mistook the tomb, especially since it was identifiable as a new tomb
belonging to Joseph of Arimathea, and the synoptic gospel writers are at
pains to emphasize that the women had already seen where the tomb was
(Mt 27:61; Mk 15:47; Lk 23:55). Various theories that the body was stolen
are highly implausible, since it could then have been produced to scotch
the church at birth—and of course, there would be no explanation for the
appearances. The fact that the disciples were prepared to die for their belief
in the resurrection is an argument against their conspiring together to steal
the body or, indeed, to invent the story. In any case, as the McGrews point
out, they were prepared to die, not for an ideology (which some might die
for) but for the sake of an empirical fact—and nobody dies for a factual
falsehood.
Swinburne gives much less attention than the McGrews to the mass hallu-
cination theory, merely noting that ‘visions shared by a number of witnesses
are very hard indeed to document.’14 I agree that this is utterly unlikely since
the disciples were far from psychologically disposed to believe that Jesus
was risen—quite the opposite: they were frightened and disillusioned after
his crucifixion. And, as the McGrews say, since Jesus apparently interacted
with the disciples collectively in numerous ways, to assert that their ‘parallel
polymodal hallucinations were seamlessly integrated is simply a nonstarter,
an event so improbable in natural terms that it would itself very nearly
demand a supernatural explanation.’15
The mathematics of Richard Swinburne’s approach

Richard Swinburne’s Bayesian approach is both more complex and more
comprehensive than that of the McGrews. He is interested not just in deriv-
ing a Bayes factor, or even a posterior probability that Jesus rose from the
dead given the historical evidence pertaining to that. He wishes to derive an
estimate for the overall, actual probability that Jesus Christ is God incarnate
who rose from the dead. For this purpose he regards it as necessary to have
a prior probability which takes account of evidence from natural theology
that there is a God, and evidence that, or reasons why, God is likely to act
this way. We need, he says, evidence from the life of Jesus suggesting that he
is divine (the prior historical evidence), in addition to the evidence specific
to the resurrection (the posterior historical evidence), and we need evidence
that there is no evidence for any other prophet comparable to that which
there is for Jesus.
Swinburne acknowledges that ‘we might be able in principle to multiply
evidence about the reliability of the witnesses or kinds of witness with whom
we are concerned . . . Then maybe the detailed historical evidence would
be so strong, despite the fact that such a Resurrection would have been a
unique exception to natural laws, that Jesus had risen that the balance of
probability would favour the latter.’ However, in contrast to the McGrews,
he goes on to say, ‘But this simply is not going to happen,’ and this is why
we need the evidence that God exists from natural theology, which provides
‘general background evidence crucially relevant to our topic.’16
Regarding the existence of God Swinburne says:
For if there is no God, the ultimate determinant of what happens in

the world is laws of nature, and for someone dead for thirty-six hours
to come to life again is (with immense probability) a clear violation of
those laws and so impossible. This is for the reason that Hume gave—
that some regularity operated on very many known past occasions is
evidence that it is a law of nature and so operated on this occasion too,
and so that Jesus did not rise.17
However, this is to ignore the problem of induction, also highlighted by

Hume, which rather negates Hume’s argument against miracles. Many past
instances of a regularity are no guarantee of future instances. The hypoth-
esis that the probability of a person rising from the dead lies between 10−44
and 10−11 is compatible with the observation that we have not seen one, as
noted above. To say that it is impossible on the basis of the evidence is to
beg the question. Scientific laws do not cause things; they merely catalogue
regularities.
There is an analogy here with the way science sometimes operates. The
steady-state theory of cosmology postulated a rate of creation of matter, in
violation of the law of conservation of mass-energy, way below the observ-
able threshold and so did not contradict observation of this regularity. In
the event the steady-state theory was much later found to be in conflict
with other evidence which came in from galaxy counts and the microwave
background radiation. The currently highly popular string theory postulates
entities whose size is way below the observational threshold, and whose
postulated curled-up dimensions are also well below the observational
threshold.
Nevertheless, Swinburne is right that one would not need a Bayes factor
anything like as high as 1044 if one had background evidence from natural
theology. The McGrews take a rather conservative line on New Testament
scholarship, even doubting the existence of ‘Q,’ the hypothetical document
used along with Mark as a source by Matthew and Luke in writing their
gospels. Swinburne is more accommodating to liberal Biblical scholarship
than the McGrews. This could still enable him to conclude that the res-
urrection is probable because he adds in the data from natural theology,
even if his Bayes factors are much lower. However, I do not think this is
the dominant factor in explaining the difference between the McGrews
and Swinburne, since both parties base their analysis on what is generally
agreed across the spectrum of Biblical scholarship—e.g. that the early dis-
ciples claimed to have seen the risen Lord, whether that claim is veridical
or not—and they both consider alternative hypotheses. As we shall see, the
greatest difference stems from the McGrews’ assumption of independence of
the different strands of evidence, i.e. their doing what in the above quotation
Swinburne thinks will not work, multiplying evidence together.
Swinburne suggests three reasons why God, if he exists, might become
incarnate: to make atonement for human sin; to identify with human suffer-
ing; and to give us moral teaching and a moral example, aiding us in this by
founding a community to encourage and pass on his teaching (a ‘church’).
Sometimes it sounds as though Swinburne is arguing a priori about what
God, if he exists, would do, or even about what God would be obliged to
do (e.g. in writing of ‘the life required of an incarnate God,’18 my italics).
In reality we only think about what God may have done in the light of the
New Testament evidence (a point Swinburne himself acknowledges).19 Nev-
ertheless, Swinburne is right that he needs some estimate of the probability
that God would become incarnate for the reasons he presents, whether these
reasons are derived a priori or after the fact, given his lower estimate than
the McGrews for what the posterior historical probabilities are.
Swinburne argues cogently and in detail that natural theology takes one
beyond simply the existence of God as supreme being to more detailed facts
about God. I argued myself in Chapter 3 that God as a necessary being
explains why there is a universe, which is contingent. Swinburne too argues
for God’s necessity.20 Moreover, God is the simplest explanation for why
there is a universe. Swinburne argues that God as truly simple would have
no limitations on his power or knowledge, since if there were limits these
would need explanation as to why they were as they were and not greater or
lesser limits.21 God would also be perfectly free and perfectly good. Indeed,
Swinburne argues that God’s goodness follows from his being free, and his
doing only those actions for which there are good reasons.22 It seems to me
that, if God made the universe with the express intention of its bringing
forth intelligent creatures with a moral sense, then he would indeed have
good reasons to become incarnate as Swinburne says.
Regarding what Biblical scholarship has to say about these matters Swin-
burne writes:
New Testament scholarship is, however, divided about whether the evi-
dence is such as one would expect if and only if Jesus proclaimed that
his life and death was an atonement for sin; and on the whole it claims
that the evidence is not such as would be expected if Jesus believed
himself divine.23
Jesus would need to believe he was divine, according to Swinburne, in order

that we should come to know the reasons why God has become incarnate in
him. Swinburne gives good reasons for rejecting the particular liberal view
that Jesus did not so believe, but the prevalence of the liberal view presum-
ably colours to some extent his assignment of probability to us finding the
evidence we do given that God becomes incarnate (though, as stated above,
I do not think this is the dominant factor). I also disagree with the liberal
view as expressed by Swinburne.
In the preceding chapters I have focused on several aspects of Jesus’ self-
understanding: as Messiah; as Son of Man; as the one fulfilling the Old
Testament scriptures; and as the one bringing in God’s rule, substituting
himself and his embryonic ‘church’ for the Temple, and inaugurating the
Eucharist. Does this amount to believing himself God incarnate? Despite
this language of incarnation being applied by the church to Jesus after the
resurrection, I think there are undoubtedly intimations in his own prior
words and deeds.
As we saw in Chapter 8, for W. D. Davies Christ is the new Torah or
fulfilment of the Torah. His main point is that this is how Paul sees Jesus,
but in making his argument Davies draws attention to Matthew 18:20:24
‘For where two or three are gathered together in my name, there am I in the
midst of them.’ Davies compares this verse with a verse from the Mishnah
(Pirḳê Aboth 3.6): ‘When they sit together and are occupied with Torah, the
Shekinah is among them.’
The Rabbinic view was that the glory of God, the divine Presence, was
among them when Jews were studying the Torah. It would seem that Jesus
is identifying himself not just with the Torah, but with the Shekinah, a quite
remarkable claim.
Let us now examine the mathematics of Swinburne’s approach in detail.
As noted above, Swinburne divides the historical evidence e into three dif-
ferent components, like the McGrews, but in Swinburne’s case only one of
them relates to the historical evidence of Jesus’ resurrection itself:
e1 = evidence of the life of Jesus (the prior historical evidence);

e2 = historical evidence of the resurrection of Jesus (the posterior histori-
cal evidence—this is Swinburne’s equivalent of the McGrews’ E);
e3 = evidence that there is no evidence for any other prophet such as there
is for Jesus;
so that e = e1 ∧ e2 ∧ e3.
Swinburne also defines f = f1 ∧ f2 ∧ f3 where fi = ei with ‘unnamed prophet’
instead of Jesus, i.e. the fi for i = 1, 2, 3 represent respectively: (1) the prior
historical requirements for this unnamed prophet being God incarnate are
satisfied; (2) the posterior historical requirements are satisfied; and (3) evi-
dence that neither set of requirements is satisfied for any other prophet such
as they are for the unnamed prophet.
Swinburne next introduces the hypotheses
h1 = God became incarnate in Jesus;

h2 = Jesus rose from the dead;
and h = h1 ∧ h2.
Swinburne also very importantly includes in his analysis
k = evidence of the existence of God from natural theology;

t = the hypothesis ‘God exists’ (theism);
c = the hypothesis ‘God became incarnate’ (in the sense formulated at
the Council of Chalcedon, and equivalent to G in my analysis of the
McGrews).
t and c are intermediate hypotheses which Swinburne needs to assess before

calculating the probability of h.
In his writing on the resurrection Swinburne utilizes a value P[t|k] = ½
arguing that this is reasonable based on his work in natural theology.
I would regard this as an underestimate. In this book I have examined only
a narrow subset of arguments for the existence of God, namely the cosmo-
logical and design arguments, with particular reference to modern cosmol-
ogy. Theism turns out to be vastly superior to the universe existing as just
a brute fact and to the popular multiverse argument, and superior to other
alternatives such as polytheism, or a weak god (as we saw above a god of
limited powers would not be simple). Theism is also superior, I would argue,
to John Leslie’s ‘extreme axiarchism,’25 a Neoplatonist position according
to which the universe is brought about by its own ‘ethical requiredness,’
since Platonic ‘ideas’ cannot cause anything. Swinburne himself produces
many other arguments to build a cumulative case. However, P[t|k] = ½ will
provide a good basis from which to proceed.
Based on his assessment of whether God, if he exists, would become
incarnate for the reasons given earlier, Swinburne then assigns a probabil-
ity P[c|t] = ½, again a reasonable figure on which to base the subsequent
analysis, though I might assess it as rather lower. I am not quite as sure
as Swinburne that God has to become incarnate, though it is hard to see
what better way there could be of achieving the ends described above, and
I agree with Swinburne that, given his attributes of omniscience, omnipo-
tence, and moral perfection, God is most likely to want to achieve those
ends.26 This probability is clearly unaffected if we add k as given, so that
P[c|t ∧ k] = P[c|t] = ½.
It follows that P[c|k] = P[c|t ∧ k]. P[t|k] = ½ x ½ = ¼
Now we want to incorporate the evidence f that we have about a certain
unnamed prophet into the overall evidence for c, as well as k:
P f |c k.P[c | k]
P c|f k (10.1)
P f |c k.P c | k P f | ~ c k.P[~ c | k]
At this point Swinburne suggests, in a similar vein to the McGrews, that it

is much more likely that we would have the historical evidence f (especially
f2, which is closest to the McGrews’ E) if God became incarnate than if he
did not. This is both because God is likely to crown the life of the prophet
who is God incarnate with a super-miracle, and because the combination
of evidence we have is much less likely to pertain if he did not. Swinburne
offers a conservatively low estimate of the probability of the evidence f given
God’s becoming incarnate and a value one hundred times lower if he did not
(both probabilities also conditioned on k):
P[f|c ∧ k] = 1/10
P[f|~c ∧ k] = 1/1000
Substituting these values and those for P[c|k] and P[~c|k] = 1 - P[c|k] = ¾
into the formula for P[c|f ∧ k] we obtain
1 1
.
P c|f k 10 4
1 1 1 3
. .
10 4 1000 4
i.e.
100
P c|f k 0.97
103
At this point Swinburne moves swiftly to his final conclusion. The evidence
we have is greater than f, namely e, that the prophet concerned is Jesus. It
follows that P[c|e ∧ k] ≈ 0.97. Furthermore, given c, e, and k, it would be
immensely improbable that God became incarnate in a prophet other than
Jesus or that this prophet’s ministry culminated in any other way (by any
other ‘super-miracle’) than the resurrection. Hence P[h|e ∧ k] will not be
very different from P[c|e ∧ k], so we can say
P[h|e ∧ k] ≈ 0.97
The crucial input for Swinburne in the above is the ratio (likelihood ratio
or Bayes factor)
P[ f | c k]
102
P[ f |~ c k]
In the McGrews’ analysis this is most closely comparable to the following

ratio:
P[E | R]
≈ 1044
P[E |~ R]
In Swinburne’s notation this ratio here is of course not quite the same. The
evidence is a subset of his total evidence and the hypothesis is a part of his
total hypothesis:
P[e2 | h2 ]
P[e2 |~ h2 ]
Of course Swinburne makes further adjustments to cohere with the logic

of his argument. Thus, as we have seen, initially he supposes historical evi-
dence similar to that available for Jesus to be available for an unnamed
prophet; he makes his hypothesis ‘God becomes incarnate’ rather than
‘Jesus rose from the dead;’ and he also conditions on background knowl-
edge from natural theology. However, I do not think these considerations
are what make for the fundamental difference in likelihood ratios, although
in the next chapter I discuss a possible caveat that God may alternatively
send a Messiah figure who is not God incarnate. I discuss this issue further
in the next chapter by following through Swinburne’s logic with somewhat
more detailed evidence, including that from chapters 7 and 8.
In the meantime, if, for the sake of argument, we assume that Swin-
burne’s fi, especially f2 within f, can be broken down in a similar way to the
McGrews’ components of E, and that the Bayes factors for Swinburne, even
though differently defined, take similar values to those of the McGrews, and
with the McGrews’ assumption of independence, then equation (10.1), with
Swinburne’s value for P[c|k], would yield something of the order
P[c|f ∧ k] ≈ 1 – 3 x 10−44
a figure very close indeed to unity. As above, it is a short step from this to
P[h|e ∧ k] ≈ 1 – 3 x 10−44
In fact, all one needs in order to derive either Swinburne’s or the McGrews’
conclusion is these likelihood ratios (Bayes factors), not the individual terms
P[f|c ∧ k] and P[f|~c ∧ k], or P[E|R] and P[E|~R]. This was already evi-
dent in my analysis above for the McGrews, and can be seen for Swinburne
simply by dividing the numerator and denominator of equation (10.1) by
P[f|~c ∧ k]. Indeed, it seems to me that a much better quantity to seek is
the ratio rather than the individual values. A person can much more easily
answer the question ‘How much more likely is it that we find the historical
evidence we do given that God became incarnate and that we have evidence
for God from natural theology than if God did not become incarnate and
we had the evidence of natural theology?’ (Swinburne) or ‘How much more
likely is it that we find the evidence we do given that the resurrection hap-
pened than if it did not?’ (McGrews). Then, taking Swinburne’s approach,
the historical evidence might pertain with low probability but the ratio of
the probabilities of getting it if God exists and becomes incarnate to getting
it if God did not exist or did not become incarnate could be large (and simi-
larly for the McGrews).
We saw above that the McGrews obtained the enormous value of their
ratio of Bayes factors by assuming independence of the evidence components
W, D, and P which comprise E, and in particular by assuming independence
within D of the testimony of the thirteen disciples. Swinburne does not par-
allel this in his calculation, although he does argue that finding both com-
ponents of evidence f1 and f2 in a single prophet is extremely unlikely unless
God became incarnate. It seems to me that breaking the evidence down in a
more detailed way into its components and accumulating the probabilities,
as do the McGrews, is both legitimate and bound to lead to a very high
value indeed for the likelihood ratio (Bayes factor). While Swinburne does
indeed consider the different components of the posterior historical evidence
(though they comprise perhaps a surprisingly small fraction of his book), he
does not do what the McGrews do in combining the probabilities (indeed,
as quoted above, he affirms that this is possible in principle but denies that
we shall be able to do it in practice).
The McGrews’ independence assumption

The McGrews get an enormously large Bayes factor chiefly because of the
assumption of independence of the individual pieces of evidence, so that
we can multiply the Bayes factors together. It is easy enough to concur with
them that the three strands, W, D, and P are independent of each other, but
what about the individual Di within D?
The McGrews argue that dropping the independence assumption may
in fact enhance the case for the resurrection rather than diminish it. Here
I give my own version of their argument. Let us for the sake of argument
limit ourselves to the testimony of two disciples D1 and D2, from which
generalization to thirteen disciples follows straightforwardly. Independence
means that
P[D1 ∧ D2|R] = P[D1|R]. P[D2|R]

P[D1 ∧ D2|~R] = P[D1|~R]. P[D2|~R]
The Bayes factor is obtained by simply dividing these two expressions:

P[D1 D2 | R] P[D1 | R] P[D2 | R]
.
P[D1 D2 |~ R] P[D1 |~ R] P[D2 |~ R]
In reality we have instead, if for example D1 depends on D2:
P[D1 ∧ D2|R] = P[D1|D2 ∧ R]. P[D2|R]

P[D1 ∧ D2|~R] = P[D1|D2 ∧ ~R]. P[D2|~R]
The question is: how does conditioning on D2 as well as R or ~R affect

the first factor on the right hand side in each case? That is, how does
P[D1|D2 ∧ R] compare with P[D1|R] and how does P[D1|D2 ∧ ~R] compare
with P[D1|~R]?
Suppose the resurrection really happened, i.e. R is true. Suppose D2 testi-
fies to the truth of R and D1 hears of it. Is D1 less or more likely to testify
to R? It seems to me that D1 is at least as likely to testify as before; indeed
D1 may be encouraged to testify to his own experience, which may or may
not have happened on the same occasion as D2’s experience. However, if D1
hears that D2 has been put in prison or even killed for testifying to R, then
P[D1|D2 ∧ R] may be less than P[D1|R], i.e. the probability when D1 does
not have this information. But what happens in the other case, if the resur-
rection did not happen, i.e. if ~R pertained?
Suppose that D1 and D2 engage in a conspiracy to spread the false rumour
that Jesus has been raised, i.e. ~R pertains but D1 and D2 each affirm R. Sup-
pose further that first D2 testifies that the resurrection happened when it did
not, and then that D1 hears that D2 has testified falsely and has been impris-
oned or killed for doing so. The McGrews, it seems to me rightly, argue
that in this case D1 is much less likely to affirm or continue to affirm the
resurrection than in the case where the resurrection really happened. That
is, P[D1|D2 ∧ ~R] will be much less than P[D1|~R] in the case when there is
a conspiracy and when, as is very probable in reality, it is very dangerous to
be part of the conspiracy.
The more accurate expression, without the assumption of independence,
for the Bayes factor is
P[D1 D2 | R] P[D1 | D2 R] P[D2 | R]

.
P[D1 D2 |~ R] P D1 | D2 ~ R P[D2 |~ R]
and from our analysis it follows that
P[D1 D2 | R] P[D1 | R] P[D2 | R]

.
P[D1 D2 |~ R] P[D1 |~ R] P[D2 |~ R]
since, while P[D1|D2 ∧ R] may be somewhat less than P[D1|R], P[D1|D2 ∧ ~R]
is likely to be proportionately very much less than P[D1|~R].
Thus the conspiracy or collusion theory adversely affects the probability
of the testimony given no resurrection much more than it does the probabil-
ity of the testimony given the reality of the resurrection. Hence the assump-
tion of independence underestimates the Bayes factor for the combined
testimony, and the McGrews’ argument is correct.
Conclusion
We have examined two different approaches to the evidence for Jesus’ res-
urrection. That evidence is certainly very powerful, but our authors differ
on just how powerful. For the McGrews it is overwhelmingly powerful.
Swinburne is much more cautious, and so has to back up his probability esti-
mates with evidence from natural theology. Swinburne is also more nuanced
in taking account of other evidence one has about Jesus, that is, evidence
of his life and teaching. If this were taken into account in the McGrews’
analysis it would have bearing—indeed, a positive bearing—on the prior
probability that Jesus in particular would rise from the dead. However, one
might still need to take account of the prior probability of getting this par-
ticular life and teaching from an individual.
It is always the case with the application of Bayes’s theorem, as with any
mathematical algorithm, that what you put in determines what you get out,
so is it a question of ‘You pays your money and you takes your choice’? Of
course the numbers can only ever be the roughest estimates. Nevertheless,
I think the answer to the question is no, because it seems reasonable to
accept the McGrews’ analysis of independence of evidence, and that drop-
ping the independence assumption is worse for naturalistic alternatives than
it is for the hypothesis that the resurrection occurred.
As a final calculation, suppose we dilute each of the McGrews’ Bayes fac-
tors for the disciples, and their factor for the women, by a factor of 10 to
take account of the fact that we are getting reports of eyewitness testimony,
rather than direct eyewitness testimony. Let us also, for good measure, dilute
the Bayes factor for St Paul by a factor of 10, even though in this case we are
getting direct evidence from the apostle himself. The overall dilution will be
1015. This results in a Bayes factor of 1029 instead of 1044 as we saw above.
Let us input this Bayes’s factor into Swinburne’s equation (I am essentially
suggesting that conditioning the somewhat differently expressed historical
evidence on the incarnation happening or not gives comparable results for
the Bayes’s factor to conditioning on the resurrection having occurred or
not). The result is
P[c|f ∧ k] ≈ 1 – 3 x 10−29
so that
P[h|e ∧ k] ≈ 1 – 3 x 10−29
i.e. it is overwhelmingly probable that God was incarnate in Christ and

Christ was raised from the dead.
Arguing against the McGrews perhaps someone will claim that, apart
from evidence from natural theology, one should adopt a prior probability
that Jesus rose which is below 10−29 (we have seen that below 10‑11 might
be reasonable). An important issue here is, however, why Jesus in particu-
lar should rise from the dead and that consideration might make for a low
probability apart from any other information. Arguing against Swinburne
a person would have to say that, given the evidence from natural theology
the probability that God would exist and become incarnate is extremely
low—indeed such a person would have to adopt a prior below 10−29 for
God existing and becoming incarnate. This latter position seems highly
implausible.
Perhaps someone will not even contemplate looking at the evidence for
the resurrection without some prior justification for believing there is a God.
Perhaps someone would think the prior probability of Jesus becoming incar-
nate is extremely low. Then Swinburne’s argument would come into play.
Any reasonable estimate of the probability of God’s existence on the basis of
natural theology, and that this God would become incarnate for the reasons
Swinburne gives, even when combined with only a cautious and moder-
ate estimate of the Bayes factor for the historical testimony, would lead to
a very high probability of the truth of the resurrection of God incarnate,
but combined with anything like the McGrews’ Bayes factor would make
the truth of the resurrection near certain. Thus we can concur with Tom
Wright’s conclusion, towards the end of his monumental tome The Resur-
rection of the Son of God:
We are left with the secure historical conclusion: the tomb was empty,
and various “meetings” took place not only between Jesus and his fol-
lowers (including at least one initial sceptic) but also, in at least one case
(that of Paul, possibly, too, that of James), between Jesus and people
who had not been among his followers. I regard this conclusion as com-
ing in the same sort of category, of historical probability so high as to
be virtually certain, as the death of Augustus in AD 14 or the fall of
Jerusalem in AD 70.27
It follows that Christians can recite the creed with a very high degree of
confidence: ‘The third day he rose again from the dead. . . .’ But more in the
next chapter as we draw together all the threads of the arguments made so
far.
Notes
lative case for the Resurrection of Jesus of Nazareth,’ in William Lane Craig and
ter: Wiley-Blackwell, [2009] 2012), 593–662.
2 Richard Swinburne, The Resurrection of God Incarnate (Oxford: Oxford Uni-
versity Press, 2003). Swinburne’s argument is summarized in Richard Swin-
burne, ‘The Probability of the Resurrection of Jesus,’ Philosophia Christi 15(2)
(2013), 239–252; and, less technically in Richard Swinburne, Was Jesus God?
3 David Wilkinson, Science, Religion, and the Search for Extraterrestrial Intel-
ligence (Oxford: Oxford University Press, 2013), 123–124, cf. 154.
4 Stephen Neill and Tom Wright, The Interpretation of the New Testament 1861–
1986, second edition (Oxford: Oxford University Press, 1988), 34ff.
5 See, for example, N. T. Wright, Jesus and the Victory of God: Volume 2 of Chris-
tian Origins and the Question of God (London: SPCK, 1996), 64.
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006).
9 Gordon D. Fee, The First Epistle to the Corinthians (Grand Rapids, MI: Wm. B.
Eerdmans, 1987), 15.
10 Swinburne, Resurrection of God Incarnate, Part III, 145–198.
13 Wolfhart Pannenberg, Jesus—God and Man (London: SCM Press, [1968] 2002),
83–104. Originally published as Grundzüge der Christologie (Gütersloher Ver-
lagshaus Gerd Mohn: Gütersloh, 1964).
15 McGrew and McGrew, ‘The Argument from Miracles,’ 626.
16 Swinburne, Resurrection of God Incarnate, 29–30.
17 Swinburne, ‘Probability of the Resurrection,’ 241.
18 Swinburne, Resurrection of God Incarnate, 55ff.
20 In refining his position, Swinburne sees God the Father as ‘ontologically neces-
sary’ with the Son and Spirit ‘metaphysically necessary’: Swinburne, Was Jesus
God? 15, 31.
23 Swinburne, ‘Probability of the Resurrection,’ 245.
24 W. D. Davies, Paul and Rabbinic Judaism (London: SPCK, [1948]1955), 150.
25 John Leslie, Value and Existence (Oxford: Blackwell, 1979), 6.
26 Of course a probability of ½ does not go with God being obliged to become
incarnate. However, Swinburne does come close to saying this for his second
reason for God’s becoming incarnate, i.e. to identify with human suffering: this is
plausibly ‘a unique best act.’ See Swinburne, Resurrection of God Incarnate, 50.
27 N. T. Wright, The Resurrection of the Son of God: Volume 3 of Christian Ori-
gins and the Question of God (London: SPCK, 2003), 710.
11 Towards a fuller picture
The fruits of ramified natural
theology
Putting the pieces together

In the last four chapters we have amassed a very substantial amount of
historical evidence for the claims of Christianity. We have also examined
the different approaches of Lydia and Timothy McGrew and of Richard
Swinburne to the evidence. In this chapter I propose to revisit my own pres-
entation of the evidence and synthesize an overall assessment of the truths of
Christian claims. In line with Swinburne’s broader programme and detailed,
logical assessment, I shall now include background knowledge from natural
theology. Let this background knowledge = k and let it include, as Swin-
burne does, not just the fruits of arguments such as the cosmological and
design arguments but also the (indisputable) facts of human wrongdoing
and suffering.
Let t = the hypothesis that God exists. We have seen that Swinburne
assigns P[t|k] = ½. I tend to think that this is an underestimate and that
P[t|k] is much closer to 1. Nevertheless all that matters for the subsequent
argument is that P[t|k] is reasonably high.
Let c = the hypothesis that God becomes incarnate. I agree with Swin-
burne that God, if he exists, has very good reasons for becoming incarnate.
He will have created us with intentions and purposes, and it is very likely he
will want to bring about atonement for human sin and to share in human
suffering. Becoming incarnate and doing something like what Jesus purport-
edly did in these respects seems perhaps the most appropriate way of doing
this, even though there may be other ways and I am not quite so comfort-
able as Swinburne with the language of ‘what God is obliged to do.’ Thus
I tend to think that Swinburne’s assignment of a probability of ½ to God’s
becoming incarnate is an overestimate, though I think it still quite likely.
With Swinburne I agree that this probability is unaffected by assuming all
the data of natural theology so I would conclude that P[c|t ∧ k] is somewhat
less than ½.
Since P[c|k] = P[c|t ∧ k].P[t|k] my two assessments above will combine to
give a probability not very different from Swinburne’s value of ¼. However,
Towards a fuller picture 211
again an exact value is not required, only an assessment that this is only
moderately low rather than extremely low.
If God becomes incarnate what might one reasonably expect? It seems to
me that the sort of things we have been considering in the previous chapters
with reference to Jesus would be quite likely. Thus the human who is God
incarnate would reasonably be expected to excel in every way the character
traits and behaviours of all other humans by leading a perfect human life.
It would be quite likely that he would work miracles, especially miracles of
healing, and would bring forgiveness and light to blighted lives. He might
well share in human suffering. Indeed this would be the human life lived
by God incarnate and so not be docetic—he would share in normal human
limitations and feel hunger, thirst, and the breadth of human pain. It is
likely that he would see his life as an offering of atonement for human sin.
And it might well be the case that God would give intimations or herald his
intention of becoming incarnate and doing these sorts of things ahead of
time, i.e. the person who is God incarnate might well fulfil prophecy. The
idea that the grave could not hold him and that he would rise from the dead
would also be made much more likely for the person who is God incarnate
than for other humans. As Swinburne puts it, God might well crown his life
with a ‘super-miracle,’ and resurrection from the dead would seem to be
singularly appropriate for that. My point, in line with Swinburne’s, is that
the prior probabilities, before one considers the historical evidence, that a
person works miracles, fulfils prophecy, and rises from the dead are raised
very substantially if that person is God incarnate. These probabilities still
may not be very high, but there is now no reason to suppose them ultra low.
Let us try and put these ideas into the Bayesian formalism. It is worth say-
ing that a few equations cannot capture all the subtlety of the arguments,
many of which have been presented in the preceding chapters, and even then
just in outline. As we have emphasized before, Bayesianism simply helps one
formulate an argument logically.
Let J′ denote an unnamed prophet for whom there is evidence that he ful-
fils criteria which one would likely expect to be fulfilled by God incarnate.
We need to evaluate the evidence to ascertain whether or not the criteria
are indeed fulfilled. I suggest that the criteria include the working of many
miracles, especially healing miracles, the fulfilment of prophecy in the sense
that the life of God incarnate was heralded long before his actual coming,
and the crowning of his life by what Swinburne calls a ‘super-miracle,’ and
I deem the most appropriate super-miracle to be his resurrection because it
seems to me quite reasonable to expect that death could not contain God
incarnate. Finally, it should be the case that no other prophet apart from
J′ fulfils the criteria. Of course these are the criteria which I see evidenced
in Christ but, like Swinburne and following Swinburne’s logic, we abstract
from Christ initially to see what would result in the case of some unnamed
prophet for whom there is evidence that he meets the criteria.
Let J1′ = the hypothesis that J′ performed substantial numbers of mira-
cles, both more in number and more suited to what God incarnate would
do, than any others who might be candidates for divinity. Now, as noted
in chapters 6 and 7, a single testimony does provide evidence in favour of
the reported miracle. Thus, for testimony T alone, Bayes’s theorem gives us
P[ J1′ | T ∧ k] P[T | J1′ ∧ k] P[ J1′ | k]

= . (11.1)
P[~ J1′ | T ∧ k] P[T |~ J1′ ∧ k] P[~ J1′ | k]
A single testimony provides evidence in favour of the reported miracle, espe-

cially if the witness is reliable, because one is more likely to receive testi-
mony if the miracle happened than if it did not. Hence the likelihood ratio,
which is the first fraction on the right hand side of equation (11.1), will be
greater than 1, i.e. P[T|J1′ ∧ k]/P[T|~ J1′ ∧ k] > 1. It follows from this that
P[J1′|T ∧ k] > P[J1′|k], i.e. the evidence of the testimony has enhanced the
probability that Jesus performed the miracle. Single testimony to a single
miracle may not do this very much, and it may be that this is insufficient to
overcome the prior improbability of J′ performing the miracle. Nevertheless
it provides what Swinburne calls a good C-inductive argument.
In fact one may already have a good P-inductive argument from what I say
below about the second ratio here, but the next step will almost certainly
ensure a good P-inductive argument. Note that in Chapter 6 we did not
include conditioning on k and we made certain approximations to mimic
Hume’s argument, viz., with our present notation, P[T|J1′] ≈ 1, P[J1′] << 1,
P[T|~J1′] << 1, and P[~J1′] ≈ 1. Then whether P[J1′|T] is greater than or less
than ½ depended on which of P[J1′] or P[T|~J1′] is greater. We shall shortly
see how conditioning on k makes a decisive difference.
Now let Tn = the combined testimony of many witnesses to many mira-
cles, n witness testimonies in total, similarly to the testimonies to Jesus’ pur-
ported miracles as discussed in Chapter 7. Then, analogously to equation
(11.1), we have
P[ J1′ | T n ∧ k] P[T n |J1′ ∧ k] P[ J1′ |k]

= . (11.2)
P[~ J1′ | T ∧ k]
n
P[T | ~ J1′ ∧ k] P[~ J1′ |k]
n
In Chapter 6 we made very simplifying assumptions about a somewhat dif-

ferently defined Tn. There we assumed that all the testimonies to a single
miracle to be independent and we treated separately the testimony to many
miracles. Here we have a much more complicated equation to derive, which
we could do in principle, to account for the multiplicity of miracle accounts
and of witnesses. There would also be the caveat to take into account that
what we are evaluating may not for the most part be direct reports from
eyewitnesses but one remove from that. Thus we are assuming for the sake
of argument that the situation for J′ is very like that for Jesus, where the
accounts are mediated by the gospel writers—though we saw that while
Mark most probably got his information from Peter in Rome, John may
well in fact have been an eyewitness, and it is possible that Matthew too was
an eyewitness on the evidence of Papias.
Having made this point about the greater complexity than what we dealt
with in Chapter 6, I do not think that more detailed mathematics would add
greatly to the discussion. This is because the principle is clear that our new
Tn will provide hugely greater support to hypothesis J1′ that did T alone,
even if diminished slightly by the second-hand nature of most of the report-
ing. As in the case of Jesus, we assume that many of the witnesses were still
around when the accounts were written down.
Thus the first fraction in equation (11.2) will be very much greater than 1:
P[T n |J1′ ∧ k]
>> 1
P[T n | ~ J1′ ∧ k]
Here the impact of conditioning on k does not materially affect either the
numerator or the denominator since the driving influence on whether there
is testimony will be whether the miracle happened. Taking k into account
implies that it is quite likely that there is a God. That would make the
numerator even larger than otherwise, since God is likely to ensure that
there is witness testimony to miracles if they occurred. On the other hand it
would be deceptive of God to allow overwhelming testimony if the miracles
did not happen, so the denominator would be even lower than otherwise.
Where conditioning on k makes a crucial difference is in the second fac-
tor on the right hand side of equations (11.1) and (11.2). We can estimate
P[J1′|k] as follows:
P[J1′|k] = P
[J1′|(c ∧ t) ∧ k] . P[(c ∧ t)|k]
+ P[J1′|~(c ∧ t) ∧ k] . P[~(c ∧ t)|k](11.3)
Now P[(c ∧ t)|k] = P[c|k] since c can only be true if t is true, and we argued
above that this is only moderately low, Swinburne’s value of ¼ being not
unreasonable.
Now if God becomes incarnate it is moderately likely that he will perform
miracles so P[J1′|(c ∧ t) ∧ k] will be moderately high. That gives a moder-
ate value to the first term in equation (11.3). It is far from impossible that
were God not to become incarnate but did exist then he would send some
prophet, not God incarnate, who would perform miracles (perhaps the
expected Jewish Messiah did not have to be God incarnate). So the second
term in equation (11.3) will also make a significant contribution. Overall
it would appear that P[J1′|k] would take a hugely higher value than P[J1′]
which would be the prior probability if one did not take into account the
evidence of natural theology.
P[~J1′|k] might still be more probable than not, but it would seem that the
first ratio in equation (11.2) would substantially outweigh the second so that
P[ J1′ | T n ∧ k]
>> 1
P[~ J1′ | T n ∧ k]
It would then follow that P[J1′|Tn ∧ k] is very close to 1. What this means
is that having the combined testimony of many witnesses to many miracles
supposedly wrought by this unnamed prophet plus the evidence of natural
theology makes it virtually assured that the prophet did in fact perform
many miracles.
For the sake of the logic of the argument we next look at the hypothesis
J2′ that J′ rose from the dead, and the evidence E for this. We assume the lat-
ter to be comparable to the evidence we have for Jesus, i.e. we assume that
E comprises the multiple testimony of many witnesses. As for the case of J′’s
miracles, we also condition on k:
P[ J2′ | E ∧ k] P[E|J2′ ∧ k] P[ J2′ |k]

= .
P[~ J2′ | E ∧ k] P[E| ~ J2′ ∧ k] P[~ J2′ |k]
Our argument here is similar to that for J1′. If we ignore conditioning on k
in the first ratio on the right hand side, then we have the same Bayes factor
as the McGrews did for Jesus. As we saw, the McGrews included several
strands of evidence in E: the witness of the women, of the disciples, and of
St Paul. They put numerical values into each component and got a very large
Bayes factor as we saw. I agreed with their way of treating the different com-
ponents of the evidence as independent, and, while the actual numbers are
of course debatable, I agree that a very large Bayes factor is justified. Add-
ing the conditioning on k will not make much difference to this assessment.
Similar again to our argument for J1′, conditioning on k in the second
factor on the right hand side radically changes the value from when k is not
conditioned on. This is because a resurrection is highly improbable unless
there is a God but the existence of God, and, further, his incarnation, is
made hugely more probable by k, even if to the extent of making the latter
only moderately low. We could expand P[J2′|k] in the same way as we did
P[J1′|k] in equation (11.3) and make a similar assessment. Thus again the
final upshot for the posterior probability ratio is that
P[ J2′ | E ∧ k]
>> 1
P[~ J2′ | E ∧ k]
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
prophecies would be purely accidental or illusory. In general the prophe-
cies 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 proph-
ecies 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 asso-
ciated 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 proph-
ecy. 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[T n ∧ E ∧ EP |J3′ ∧ k] P[ J3′ |k]

= .
P[~ J3′ | T n ∧ E ∧ EP ∧ k] P[T n ∧ E ∧ EP | ~ J3′ ∧ ] P[~ J3′ |k]
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 prophe-
cies 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 pos-
sibility 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 confi-
dently conclude that P[J4′| E4 ∧ k] ≈ 1.
It looks extremely likely, then, on the basis of the evidence, that J′ per-
formed 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
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 remark-
able 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 proclaim-
ing 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 Swin-
burne’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 some-
what, 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 informa-
tion about the prophet’s identity will not make any difference to the prob-
ability 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.
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 Univer-
sity 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 & Spot-
tiswoode, and New York: McGraw Hill, 1963).
Armstrong, Benjamin F., ‘Hume on Miracles: Begging-the-Question Against Believ-
ers,’ 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 edi-
tion (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 Testi-
mony (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 Singular-
ity,’ 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: Hod-
der and Stoughton, 1974).
Bruce, F. F., The New Testament Documents: Are They Reliable? sixth revised edi-
tion (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 Hermeneu-
tics,’ 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, Philo-
sophical, 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? (Lon-
don: 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 Con-
cerning 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 Compan-
ion 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: Hendrick-
son, 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:
Glass, David H., and McCartney, Mark, ‘Explaining and Explaining Away in Sci-
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) (Janu-
ary 2012), 6–7.
Guth, Alan H., ‘Inflationary Universe: A Possible Solution to the Horizon and Flat-
ness 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: Cam-
bridge 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 Univer-
sity 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 Testi-
mony, 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 & Sci-
ence 45(2) (November 1981), 8–12.
Hume, David, An Enquiry Concerning Human Understanding, Section X, ‘Of Mira-
cles,’ 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: Pen-
guin, [1779] 1990).
Irenaeus, Adversus Haereses, I.XXIV.4, in Alexander Roberts and James Donald-
son (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: Scot-
tish Academic Press; Macon, GA: Mercer University Press, 1986).
Josephus, ‘The Antiquities of the Jews and The Wars of the Jews,’ in William Whis-
ton (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 Uni-
versity 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 introduc-
tion 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 Tun-
ing 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: Athe-
ist Fundamentalism and the Denial of the Divine (London: SPCK, 2007).
McGrew, Timothy, and McGrew, Lydia, ‘The Argument from Miracles: A Cumula-
tive 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 Uni-
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 Rich-
ard 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. Rob-
inson 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 Clay-
ton (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 Wil-
ckens, 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: Hendrick-
son, 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 (Chi-
cago: University of Chicago Press, 1974).
Polkinghorne, John, Science and Christian Belief: Theological Reflections of a Bot-
tom-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 (Cam-
bridge: 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 Strat-
egies 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 (Lon-
don: 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:
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 Contem-
porary 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: For-
tress, [1835–1836] 1972).
Susskind, Leonard, The Cosmic Landscape: String Theory and the Illusion of Intel-
ligent 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 Univer-
sity Press, 2003).
Swinburne, Richard, The Existence of God, second edition (Oxford: Oxford Uni-
versity 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 Jerusa-
lem (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
Witham, Larry, The Measure of God: Our Century-Long Struggle to Reconcile Sci-
ence and Religion, The Story of the Gifford Lectures (San Francisco: HarperSan-
Francisco, 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 biblical data 195 – 198, 207; Richard

Swinburne’s Bayesian approach
Alston, William 19, 65, 78 198 – 205, 207 – 208
alternative hypothesis 37, 41 – 42, 55 Bayesian framework for miracle
Analogy of Religion (Butler) 82, 93 arguments 110 –113, 145; aggregation
analogy, principle of 75 – 77, 129, 137 of testimonies 114–118; impact
Antiochene crisis 92 on Hume’s arguments 118–119;
apostles 82, 87, 99, 178, 180–181 independent testimonies 113–114
apparent revelations 78 Bayes’s theorem 25 – 27, 59, 110
a priori reasoning 76 belief in God 19, 68, 102, 121
Aquinas, St Thomas 10, 12, 38, 44 – 45, Berry, R. J. 107
65, 108 – 109 best explanation, language of 21 – 24,
Artaxerxes’ order, dating of 91 28, 58, 66, 143 – 144, 170
atheism 2, 30, 66, 68, 75, 85 Bhaskar, Roy 20, 21
Augustine, St 10, 16, 18, 83, 108 – 109 Bible 6, 31, 82, 86, 98, 102, 129, 133
Big Bang 5, 29, 36, 38, 46, 48 – 49, 67
Bacon, Francis 10 – 11 Bohmian theory 67
Barth, Karl 18, 31, 130 Bonhoeffer, Dietrich 67 – 68
Basilides 73 boundary condition of universe 39
Bauckham, Richard 76, 138, 171, Brief History of Time, A (Hawking) 39
174, 191 Bussey, Peter 40
Bayes factor 7, 26, 190, 192–193, 204–207 Butler, Joseph 5–6, 93–104; analogy
Bayesian confirmation theory 4, 7, 25, 59 definition 98; analogy with Pascal’s
Bayesian evaluation of multiverse wager 93–94, 95; appeal to ‘common
hypothesis 60 – 62 practice’ 101; arguments in terms of
Bayesian formulation, of Christian claims probability 94–95, 104; assumption
210; background knowledge 210–211; of deist position 98; as Chaplain 93;
fulfilment of prophecy 211, 214–217; cumulative case 94, 101; fulfilment
healing miracles, testimonies to of prophecy 103; on miracles and
212–214; ‘super-miracle’ (Swinburne) prophecies 96–98, 101–102; on moral
202–203, 211, 214, 217 governance of universe 95–96; on notion
Bayesian formulation, of resurrection of ‘incredibility’ 100; observations by
arguments 189; McGrews’ approach analogy of nature 101; positive evidence
189 – 195, 206 – 207; McGrews’ of Christianity 98–99; as Preacher at
independence assumption 205 – 206; Rolls Chapel 93; on religious pluralism
Richard Swinburne’s approach to 99–100; revelation 102–103
Index 229
Calvin, John 10, 65, 68, 75 Daniel, book of 91 – 92, 103, 159,
causal principles 16 163, 167
chaotic inflation 49 Darwin’s theory of evolution by natural
childhood theism 75 selection 3, 15 – 16, 132, 133
Christ, Jesus: humanity and sufferings Davies, Paul 46 – 47
73; male disciples of 182; mighty Dawood, N. J. 73
works 139 – 144; place in world deism 23, 85
history 69; see also Jesus, death on deists 6, 82, 95, 97 – 98
cross; miracles; resurrection of Jesus design argument 12, 17, 24, 36, 42, 47,
Christ 202, 210
Christian belief 1, 11, 18, 74, 78, 95 ‘Deutero-Isaiah’ 151, 153, 155, 161
Christian claims 8, 90; arguments for divine creation of universe: Christian
5; evidence for 78 – 79, 145; truth of doctrine of creation 44 – 45; and
5; see also Bayesian formulation, of no boundary proposal 41 – 42; and
Christian claims; resurrection of Jesus quantum vacuum creation 43 – 44; see
Christ also multiverse hypothesis
Christian doctrine of creation 1, 16, 41, divine design 36; fine-tuning in absence
44 – 45 of 55 – 56; multiverse compatibility
Christian faith 11, 21, 22, 31, 66, 74, with 55; multiverse hypothesis
87, 181 dependence on 58; TOE and 47
Christian God 1, 30, 59, 121 divine reality 69, 75
Christian interpolation in Josephus Dunn, James D. G. 71, 129, 142,
141 – 142, 144, 172 172 – 173, 185
Christianity 5, 70, 82, 87, 93 – 94,
132 – 133, 173; claims, historical Earman, John 109
evidence of 68; death of Jesus 70 – 74; ECAs see embodied conscious agents
presuppositions 69, 74 – 76; tradition Ecce Homo (J. R. Seeley) 132
transcendent rationality 21 Eliot, George 130
Christian natural theology Elster, Jon 86
(McGrath) 20 embodied conscious agents 56 – 57
Christian revelation 10 empirical law of causality 12
Christian theology 2, 10, 23, 69, 71, Enlightenment 6, 129, 131, 137
78, 136; see also natural theology Enquiry Concerning Human
Christian Trinitarian orthodoxy 21 Understanding (Hume) 110
‘Christ of faith’ 130 epistemic probability 26, 94 – 95
Christology 133 – 134 epistles of St Paul 99
classical universe 40 Essays and Reviews 132
Collins, Robin 51, 56 – 57 eternal inflation 49
‘common sense’ philosophy 171 Euclidean 4-space 39 – 40
conditional probability 25 evidence 2, 37; categories of 23; in
‘consonance’ concept 23 Glass and McCartney’s formulation
convergent evolution, phenomenon of 9 37; independent 37; for miracles 6;
correlation, principle of 75 scientific theories 25
cosmic background radiation 30, 67 existence of God 65, 111; arguments
cosmological argument 11 – 12, 35, for 6; metaphysical proofs for
37 – 38, 42,44, 50, 59 – 60, 101, 190 (Pascal) 84; miracle testimony as
cosmology 40, 45, 53, 186; see also evidence for 120 – 125; natural
modern cosmology knowledge and 11; probabilities
critical realism 20 – 21, 24, 129 relevant to 27; ruled out by fiat
criticism, principle of 75 74 – 75; traditional arguments for
Crossan, Dominic 136 24 – 25, 31, 36; see cosmological
crucifixion see Jesus, death on cross argument; design argument
Cruickshank, John 84 expanding universes, space-times of 40
230 Index
‘Explaining and Explaining Away in Gospel of Thomas 139; Jesus’
Science and Religion’ (Glass and mighty works 139 – 144; of John
McCartney) 36 178, 181, 184, 187; of Luke
explanations: for fine-tuning 47 – 48; 177 – 178, 179 – 180, 182, 184;
provided by God 29 – 32 of Mark 176, 177, 179, 184; of
explanatory power 21, 25, 102, 185 Matthew 177, 179, 184; miracles
eyewitness testimony see testimony 107, 139 – 144
Grand Design, The (Hawking and
faith 11, 31 Mlodinow) 39 – 40
Feuerbach, Ludwig 75, 131 Grand Unified Theory 57
four-dimensional space 39 great fire of Rome 172, 173
Frazer, J. G. 18 Gregory of Nyssa 16
fulfilment of prophecy, texts on GUT see Grand Unified Theory
150 – 164; Isaiah 11:1 – 5 158; Isaiah
42 155; Isaiah 42:5 – 9 156; Isaiah Harman, Gilbert 21
49:1 – 7 155; Isaiah 50:6 157; Isaiah Harnack, Adolf von 133 – 134
52:13 – 53:12 151 – 152; Isaiah 53 Hartle-Hawking theory 39 – 40
153, 154; Isaiah 53:9 154; Jesus’ Hartle, James 39 – 40
crucifixion 153 – 154; Luke 4:16 – 19 Harvey, Anthony 141 – 142
156; Malachi 3:1 161; Malachi 4:5 Hawking, Stephen 39 – 42
161; Mark 1:2 – 3 161; Mark 1:6 Hebblethwaite, Brian 68, 69 – 70
161 – 162; Mark 14:60 – 61a 154; Hegel, G. W. F. 18, 131 – 132
Mark 15:16 – 20 156 – 157; Mark Hegesippus 172
15:29 – 32 157; Matthew 5:17 160; Heidegger, Martin 135
Matthew 8:14 – 16 153; Matthew Hellenistic Judaism 142
11:29 – 30 160; Matthew 18:20 Hengel, Martin 138
160 – 161; Paul’s epistles 153; Psalm historical-critical method 76; see also
2:1 – 9 158 – 159; Psalm 22:18 158; gospel accounts
Psalm 110:1 159 – 160; ‘Suffering historical enquiry, Troeltsch’s principles
Servant’ passages in Isaiah 151 of 75
Holocaust, prima facie incredibility 76
Gabriel (angel) 91 Hooker, Morna 135, 141, 142, 144,
Gifford Lectures in natural theology 18 157 – 158, 161, 182
Glass, David 4, 36, 41–42, 43–44, 55–58 Hoyle, Fred 44, 46
God incarnate, posterior probability Hume, David 3, 11, 75 – 76, 94 – 95,
for truth of 7, 203 – 204, 207, 106; criticism of natural theology
216 – 217 12 – 13; definition of miracle 6, 106,
God of Christian theology 23; creation 108 – 109, 111; miracles argument,
and design by 58; explanations Bayesian framework 111 – 113;
provided by 29 – 32; hypothesis 25, miracles argument, Bayesian
27, 62; as ‘necessary being’ 45; formulation impact on 118 – 119;
universe’s dependence on 44 – 45 ‘uniform experience’ against
gospel accounts, historical-critical miracles 107
study of 128 – 137; Baur 131 – 132;
Feuerbach 131; form criticism imaginary time 39 – 40, 41
134 – 135; Harnack 133 – 134; Infancy Gospel of Thomas 139
Heidegger 135; Hellenistic culture inference to the best explanation see
and 133; John’s gospel 131 – 132; best explanation, language of
Lightfoot 133; Pannenberg 136 – 137; inflation, theory of 48 – 49
Reimarus 129 – 130; Schweitzer 134, initial hypothesis 37, 41, 43, 55
136, 161; Seeley 132; Stanton 135; interconnectivity, principle of 75
Strauss 130 – 131 Irenaeus 73
gospels 7, 128; dating of 137 – 138; Islam 6, 70, 72 – 74, 87, 99, 139
historicity of 6, 137; Infancy Israel 155
Index 231
James’s death (brother of Jesus) 172 McCartney, Mark 4, 36, 41 – 42,
James, William 18 43 – 44, 55 – 58
Jansenism 83 McGrath, Alister 4, 19 – 22, 28, 74,
Jeremiah’s prophecy 91 129, 133
Jesus and the Eyewitnesses McGrew, Timothy, and McGrew,
(Bauckham) 138 Lydia 7, 77, 79, 145, 189, 190 – 198,
Jesus’ death on cross 70 – 74; evidential 199 – 200, 201, 202, 204 – 206, 208,
criteria for 71; idea that a substitute 210, 214
crucified 73 – 74; Islam’s denial of Meier, John P. 71
72 – 73; liberal Jesus not a threat miracles 86, 98, 170; aggregation
71 – 72; Mark’s account of 158; see of testimonies to 109, 114 – 118,
also Christ, Jesus; resurrection of 212 – 214; Bayesian framework
Jesus Christ 109 – 113, 212 – 214; credibility
‘Jesus Seminar’ 135 – 136 of 99, 106; definition of 6, 106,
Jews, expulsion from Rome 173 108 – 109, 111; exorcisms 141;
Judaic (Petrine) Christianity 132, 133 healings and nature miracles 139,
Judaism 6, 30, 71, 78, 135, 142, 173, 142; independent testimonies to 109,
185–186; see also Rabbinic literature 113 – 114; Infancy Gospel of Thomas
139; Jesus’ purported, Biblical
kalām cosmological argument 37–38, 40 accounts of 6, 139 – 145; prior
Kant, Immanuel 3, 9, 12 probability of 110; and prophecies 5,
Kateregga, Badru 72 6, 86 – 90, 101, 166; St Augustine on
Kingsley, Charles 3, 9, 16 – 17 108; testimony as evidence for God
‘knock-down’ proof 23 existence 109, 120 – 125; ‘uniform
knowledge of God: available to human experience’ against 107; see also
beings 67; and natural theology resurrection of Jesus Christ
10 – 12; revelation 66; sources of 1, Miracles (Lewis) 107
65 – 66 ‘model-dependent realism’ 39 – 40
Kraemer, Hendrik 69, 74 modern cosmology 4, 17, 46, 202;
Krauss, Lawrence 42 – 43 fundamental issues raised by 36;
recent developments in 48 – 50
Leibniz, Gottfried Wilhelm 9, 14, 15, modern natural theology 13, 19 – 29
38, 42, 44 Mohammed 70, 73, 87, 139; see also
Lessing, Gotthold Ephraim 129, 130, Islam
131, 136 motivated belief 23
Lewis, C. S. 7, 69, 107, 139, 148 – 149, Moule, C. F. D. (Charlie) 73 – 74
168, 215 M-theory 49
Liberal Protestantism 5, 6, 134 multiverse hypothesis 22 – 23, 36,
Life of Jesus Critically Examined 47 – 48; Bayesian evaluation of
(Strauss) 130 – 131 60 – 62; Christians welcoming 50 – 51;
Lightfoot, J. B. 133 compatibility with divine design 55;
likelihood ratio see Bayes factor vs. design by God 58; fine-tuning
Locke, John 11, 31, 94 – 95 problem solving 55 – 57; independent
logical probability 28 evidence for 57; Mawson’s criticism
Lorentzian geometry 40 of 53, 55, 57; need for design 58;
observer selection principle and 56;
MacIntyre, Alasdair 21 problems for 51 – 55; Tegmark’s Level
Mack, Burton 136 IV 48, 53, 56, 57; truth of 57
Mackie, J. L. 101 Muslim religion see Islam
Magdalene, Mary 176, 176, 177, 178,
178, 179, 181, 182, 184, 197 naturalistic assumptions 2, 6, 131
Mahometanism 99; see also Islam natural knowledge of God 5, 10 – 11
male disciples of Jesus 182 natural theology 1, 3, 4, 5, 9, 67,
Mawson maximal multiverse 53, 56, 57 74; definitions of 19 – 20, 22, 65;
232 Index
knowledge of God and 10 – 12; and physical law, design of 15
motivated belief 22 – 24; in practice physico-theology 3, 12
11 – 12; rejection of 18, 67 – 68, 83; Plantinga, Alvin 20, 67, 68 – 69, 75
vs. revealed theology 19, 68, 78 Plato 7, 10, 149, 168
natural theology, design arguments of Pliny the Younger 173
12 – 19; Aquinas 10, 12; Augustine Polanyi, Michael 21, 23 – 24
16; Barth’s rejection of 18, 67 – 68; political ideologies 66
Darwinian evolution 15 – 16; Gifford Polkinghorne, John 4, 20, 21, 22 – 24,
Lectures 18; Hume’s criticism 13 – 14; 28 – 29, 129
Kant’s criticism 12; Kingsley 16 – 17; ‘positivism of revelation’ 67
Leibniz 14; Paley 14 – 15; Temple 17 posterior probability 7, 25, 145
Newman, John Henry 2 – 3 presuppositions 74 – 76
New Testament: criteria for assessing primary and secondary causes,
authenticity 71; historical critical distinction between 108 – 109
study of 128 – 137 ‘Principle Credulity’ 171
New Testament evidence for the ‘Principle of Testimony’ 129, 171
resurrection 171–173, 189; alternative prior probability 7, 25, 27 – 28, 60, 62,
explanations of 183–186; comparison 100, 110, 111, 118, 120, 122, 123,
with scientific evidence 186–187; male 145, 168, 193, 194, 198, 207
disciples of Jesus 182; overview of probability 5, 25 – 26; conditional 25;
174–176; Paul’s list of resurrection epistemic 94 – 95; logical 28; posterior
appearances 181, 182–183; testimony 7, 25, 145; prior 7, 25, 27 – 28, 60,
of women 182–184; witnesses to 62, 100, 110, 111, 118, 120, 122,
empty tomb 176, 176–178; witnesses 123, 145, 168, 193, 194, 198, 207;
to Jesus’ appearances after his death subjective 28
178, 178–181 properly basic beliefs 20, 68, 75
Nicene-Chalcedonian formulation 20 prophecy 86 – 87, 143; definition of
‘no boundary proposal’ 39 – 41 148 – 150; feeding miracle 161;
‘nothing,’ concept of 42 – 43 formalizing argument from 164 – 168,
214 – 216; fulfilment of 149 – 150,
observer selection principle 56 211, 214 – 217; healings 153; Israel’s
Origin of Species, The (Darwin) 16, 132 hopes 155, 160; Jesus’ crucifixion
153 – 154, 156, 157; Jesus’ self-
Paley, William 3, 9, 14 – 15 description as Son of Man 159 – 160;
Pannenberg, Wolfhart 1–2, 11, 31, John the Baptist 161 – 162; and
68–69, 74–76, 78, 136–137, 196–197 miracles 166; mockery of Jesus
Papias 7, 144, 174, 181, 191, 156 – 157; proclaiming Kingdom of
196, 213 God 152; Suffering Servant 151 – 153,
Pascal, Blaise 5 – 6, 82 – 93; approach 155; texts on fulfilment of 150 – 164
to prophecy 86 – 91; arguments putative revelation, arguments or
for superiority of Christianity 86; evidence for validity of 67
conversion experience 83; emphasis
on man’s total wickedness 83 – 84; on quantum theory, interpretations of 5,
metaphysical proofs 84; predicting 66–67
timing of birth of Jesus 90 – 91; on quantum vacuum creation 42 – 44
religious pluralism 87 – 88; repudiates ‘Quest of the Historical Jesus’ 129 – 134
arguments from nature 83, 84 – 85; Qur’an: Jesus’ death 72, 73; Jesus’
wager 85 – 86 miracles 139
Pauline Christianity 132, 133
Peacocke, Arthur 20, 50 – 51, 55 Rabbinic literature 73, 135, 141,
Penelhum, Terence 93 – 94 160 – 161, 201
Penrose, Roger 57 rationality: theological 1 – 3; tradition
Pensées (Pascal) 83 – 93 transcendent (McGrath) 21; universal
Petrine (Judaic) Christianity 132, 133 (McGrath) 21
Index 233
real time 39 – 40 testimony 6, 11, 76, 95, 99, 100,
reason and revelation 65 – 66 106 – 107, 109 – 110, 113, 118 – 119,
Reid, Thomas 171 121, 122 – 126, 129, 138, 144,
Reimarus, H. S. 129 – 130 170 – 171, 175, 177, 186, 187,
religions see religious pluralism 192 – 193, 195, 196, 197, 205 – 207,
religious pluralism 5, 24, 31, 66 – 70, 212 – 213
74, 86, 87, 99, 128; atheism 66; testimony of women 182, 183
Christianity 67; ‘positivism of theism 4, 5, 13, 22, 30, 36
revelation’ 67 Theissen, Gerd 136, 143
resurrection of Jesus Christ 2, 7, theistic explanation 29 – 32
70, 74, 76 – 78, 106, 137, 170; theological rationality 1 – 3
see also Bayesian formulation, of theology 1 – 2, 21, 69 – 70
resurrection arguments; Christ, ‘theory of everything’ 47, 49
Jesus; Jesus, death on cross; New Through the Looking Glass (Carroll)
Testament evidence 42 – 43
Resurrection of the Son of God, The TOE see ‘theory of everything’
(Wright) 208 Torrance, Thomas 5, 19, 20, 22
revealed theology vs. natural theology 68 tradition transcendent rationality 21
revelation of God 1, 5, 10, 65 – 66, 137 Trinitarian-Chalcedonian Christian
Rolls Chapel 93 belief system 74; see also Nicene-
Rowland, Christopher 140 – 141 Chalcedonian formulation
Trinity 20
Sanhedrin 43a 73, 141 Troeltsch, Ernst 5, 75, 79, 129, 130, 131
Sanhedrin 107b 141 truth of Christianity 2 – 3
Schellenberg, J. L. 70
Schleiermacher, Friedrich 1 – 2 universal rationality 21
Schweitzer, Albert 134, 136, 161 universe 4, 9, 22, 30; animal analogy
science, contradictory theories in 66 – 67 13; beginning of 4, 36, 38, 39 – 40;
scientific explanations 29 – 30, 32 41 – 42; boundary condition of
Smith, Quentin 40 39; Christian doctrine of creation
Soskice, Janet 44 44 – 45; dependence on God 44;
Stanton, Graham 135 design on basis of theism 13 – 14; as
steady-state theory 5, 30, 44, 67, 199 divine creation 41 – 42, 44, 45; fine-
Strauss, David Friedrich 130 – 131 tuning of 17 – 18, 22, 36, 46 – 48;
string theory 49–50, 51, 52, 53, 66, 199 kalām cosmological argument
subjective probability 28 37 – 38, 40; multiverse hypothesis
Suetonius 173 and 47 – 48; ‘no boundary’ proposal
‘super-miracle’ (Swinburne) 70, 203, for 39 – 41; observers and 56 – 57;
211, 217; Bayesian formulation ordered and structured 57;
202 – 203, 211, 214, 217 probabilities relevant to 27 – 28;
Sûrah IV 72 – 73 quantum vacuum creation 42 – 44;
Sûrah V 139 ‘theory of everything’ and 47;
Susskind, Leonard 49 – 50 ultimate explanation of 45
Swinburne, Richard 4, 7, 9, 12, 13, Universe from Nothing, A (Krauss) 42
21, 24 – 29, 31, 41 – 42, 70, 102, 171;
approach to biblical data 195 – 198, Vestiges of the Natural History of
207; Bayesian approach 198 – 205, Creation (Chambers) 132
207 – 208 Vilenkin, Alexander 40
Tacitus 172 – 173 Westermann, Claus 151 – 152,

Tegmark’s Level IV multiverse 48, 53, 154, 167
56, 57 Westminster Confession of Faith 18
Tegmark, Max 48 witnesses to empty tomb 176, 176 – 178;
Temple, Frederick 3, 9, 17 gospel of John 178, 184; gospel
234 Index
of Luke 177 – 178, 184; gospel of gospel of John 181; gospel of Luke
Mark 176 – 177, 184; gospel of 179 – 180; gospel of Mark 179; gospel
Matthew 177, 184 of Matthew 179, 184
witnesses to Jesus’ appearances after his Wright, N. T. (Tom) 71, 72, 77, 132,
death 178, 178 – 181; acts of apostles 136, 139 – 140, 142 – 143, 159, 160,
180–181; 1 Corinthians 15 181; 162, 163, 177, 185–186

Ramified Natural Theology in Science and Religion Moving Forward From Natural Theology PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ramified Natural Theology in Science and Religion Moving Forward From Natural Theology PDF

Uploaded by

Copyright:

Available Formats

‘Probabilistic natural theology is the production of arguments from the most

general features of the universe to the probable existence of God. Ramified

‘Distilling decades of deep reflection on the consonance between the

Science and the Truthfulness of Beauty

Against Methodology in Science and Religion

Interreligious Perspectives on Mind, Genes and the Self

God, Evolution, and Animal Suffering

Ramified Natural Theology in Science and Religion

1 Theology as a scientific discipline: the place of natural

2 Reason and religion: the role of natural theology 9

3 Natural theology and modern cosmology: the

4 Moving on from natural theology: why we need

5 Pascal’s Pensées and Butler’s Analogy: foreshadowing

6 The rationality of belief in miracles 106

7 Ramified natural theology and evidence for Christian

8 Jesus and prophecy 148

9 Ramified natural theology in action: outline of the

10 On the third day he rose again: Bayesian

Natural theology, as traditionally conceived, is concerned with what we

I am grateful to the following for permission to republish articles or por-

• Oxford University Press for Rodney D. Holder, ‘Hume on Miracles:

Introduction: theological rationality

Is not the being of a God reported to us by testimony, handed down

Moreover, theology undergirds every other branch of knowledge:

Newman concludes his chapter on theology as a branch of knowledge with

Religious doctrine is knowledge, in as full a sense as Newton’s doctrine

Whether this assessment of theology as the pursuit of truth, in like manner

Outline of the book

Traditionally “natural theology” has commonly meant something like

Natural theology is an area of intellectual enquiry with a long, if chequered,

To conclude, therefore, let no man out of a weak conceit of sobriety, or

Natural theology: from Aquinas to the present day

Historically, Hume’s arguments seem not to have precipitated the downfall

What is touched by this doctrine [of Evolution] is not the evidence of

Here Kingsley is prescient. Modern cosmology has discovered how indeed

Three modern advocates of natural theology40

Alister McGrath: natural theology redefined41

Tradition transcendent rationality

Inference to the best explanation

John Polkinghorne: natural theology and motivated belief

One may define natural theology as the attempt to learn something of

That is a classic definition, which implies, contra Torrance and McGrath,

The realistic aspiration is that of attaining the best explanation of com-

Richard Swinburne and confirmation theory

Here H represents some hypothesis being evaluated against evidence E. The

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

If one wants simply to compare the probability of H with the probability of

Summary: McGrath, Polkinghorne, and Swinburne

What kind of explanation does God provide?

Question 1. Is the alternative hypothesis HA incompatible with the initial

Question 2. How likely is it that the alternative explanatory hypothesis HA

Question 3. Is the alternative explanatory hypothesis HA known to be true?

Question 4. Does the explanatory hypothesis HA depend on HI?

Question 5. Is E taken as providing evidence for HI in the first place (i.e.,

What do we make of the beginning?

1 Everything that begins to exist has a cause of its existence.

There would be no singularities at which the laws of science broke

So long as the universe had a beginning we could suppose that it had

Question 1. Is the no boundary proposal and/or lack of a beginning to the

Question 2. How likely is it that the no boundary proposal or a universe

Question 3. Is the no boundary proposal/absence of a beginning known to

Question 4. Does the no boundary proposal/absence of a beginning to the

Question 5. Is the existence of the universe taken as evidence for divine

2 Reason and religion: the role of natural theology 9

6 The rationality of belief in miracles 106

8 Jesus and prophecy 148

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

P[H|E] >> P[H ] (3.2)