Professional Documents
Culture Documents
‘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
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.
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
Bibliography219
Index228
Preface
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
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.
9 Newman, Idea of a University, 38.
10 Newman, Idea of a University, 50.
11 David H. Glass and Mark McCartney, ‘Explaining and Explaining Away in 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.
10 Reason and religion
The knowledge of God and natural theology1
When it comes to knowledge of God, it has traditionally been thought that
there are two kinds of such knowledge: natural knowledge which is what
we know about God purely by being human and the knowledge of God
which is given through God’s special revelation. The former comes under
the domain of ‘natural theology,’ and a typical way of understanding this
term is given by James Barr:
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:
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
12 Reason and religion
which exists has a cause for its existence. Therefore there is a cause for the
universe’s existence. The design argument appeals to the ordered structure
of the universe as requiring explanation. The merits of different forms of
these arguments have been the subject of philosophical debate for centuries
and we shall return to them.
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
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
Reason and religion 17
positive is Kingsley’s statement in his lecture of 1871 entitled ‘The Natural
Theology of the Future’: ‘We knew of old that God was so wise that He
could make all things; but behold, He is so much wiser than even that, that
He can make all things make themselves.’29 Indeed Alister McGrath sees the
‘neat slogan’ about God choosing ‘to make all things make themselves’ as
‘nothing more than a creative reworking of Augustine’s notion of rationes
seminales.’30
Frederick Temple, who was consecrated Archbishop of Canterbury in
1896, could, like Kingsley, see that the design argument of natural theology
was still valid, though revised from the form it took in William Paley. Echo-
ing Kingsley he wrote:
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
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.
Reason and religion 21
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.
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,
24 Reason and religion
unique infallible method, but involves judgments made by persons. That
means that complete objectivity is unattainable. A favourite quotation of
Polkinghorne’s from Polanyi sets out the aims of Personal Knowledge: ‘The
principal purpose of this book is to achieve a frame of mind in which I may
hold firmly to what I believe to be true, even though I know that it may
conceivably be false.’65
Ultimately for Polkinghorne judgments of truth can only be made from
inside the community—be that the scientific fellowship or the church. This
is Anselm’s ‘faith seeking understanding’ and, says Polkinghorne, ‘there is
no Archimedean point of detachment from which judgment can be made;
insight is only gained through participation.’66 This is all part of how 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.
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
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
Reason and religion 27
a ‘P-inductive argument.’70 These are in contrast to deductive arguments,
which make a hypothesis certain, but work in neither science nor theology.
Let me give an example of Bayes’s theorem in practice. One dark and
rainy night Sarah witnesses a hit-and-run accident. The car which leaves
its victim behind is a taxi and Sarah tells the police that it was a blue taxi.
There are two taxi firms in the town, one using green taxis and the other
blue taxis. The green taxi firm is the dominant one with 85% of the taxis.
Sarah is tested to see if she can correctly identify the colours of taxis under
similar conditions to the night in question, and she is right 80% of the time.
The jury is asked to judge whether it was indeed the driver of a blue taxi, as
reported, who committed the crime.71
In this example the prior probability that the culprit was driving a blue
taxi is 15%, i.e. 0.15; the likelihood that it was blue, i.e. the probability that
it was reported blue if it was blue, is 0.8; the likelihood that it was green is
0.2. The correct answer to this problem contradicts what most 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
28 Reason and religion
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.
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.
Reason and religion 33
12 See Richard Swinburne, The Existence of God, second edition (Oxford: Oxford
University Press, 2004), 58, n. 4.
13 David Hume, Dialogues Concerning Natural Religion, ed. J. M. Bell (London:
Penguin, [1779] 1990).
14 Hume, Dialogues, part 2.
15 Hume, Dialogues, part 5.
16 Hume, Dialogues, part 6.
17 Hume, Dialogues, part 7.
18 Gottfried Wilhelm Leibniz, ‘On the Ultimate Origination of Things,’ in 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.
23 Paley, Natural Theology, 12.
24 Paley, Natural Theology, 207.
25 Holder, God, the Multiverse, and Everything, 36.
26 Gregory of Nyssa, Apologetic Treatise on the Hexemeron, quoted in McMullin,
‘Introduction: Evolution and Creation,’ in Ernan McMullin (ed.), Evolution and
Creation (Notre Dame, IN: University of Notre Dame Press, 1983), 12.
27 See St Augustine, The Literal Meaning of Genesis (De Genesi ad Litteram),
Ancient Christian Writers, vol. 1, translated and annotated by John Hammond
Taylor SJ (New York and Mahwah, NJ: Paulist Press, 1982), translator’s note 67
to Book IV, 252–254.
28 Darwin Correspondence Project Database, letter no. 2534, accessed 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.
34 Reason and religion
40 The author is grateful to the Ian Ramsey Centre and the International Society for
Science and Religion, joint organizers of the ‘God and Physics’ conference held
in Oxford in July 2010 to celebrate John Polkinghorne’s 80th birthday, for the
opportunity to present a short paper. The latter half of this chapter is a modified
and extended version of the paper given at the conference.
41 For more on McGrath, see Holder, The Heavens Declare, 169–231.
42 William Alston, Perceiving God: The Epistemology of Religious Experience
(Ithaca, NY: Cornell University Press, 1991), 289.
43 See Alister E. McGrath, The Open Secret: A New Vision for Natural Theology
(Oxford: Blackwell, 2008), 7; Alister E. McGrath, A Scientific Theology 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.
Reason and religion 35
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.
76 Swinburne, Existence of God, 35–45.
3 Natural theology and modern
cosmology
The cosmological and design
arguments
Introduction1
Modern cosmology tells us that the universe began with a gigantic explosion
we now call the Big Bang, out of which all the galaxies, stars, and planets
we observe today evolved over a period of 13.8 billion years. It seems to me
that modern cosmology raises two fundamental issues which take us beyond
the science into the realm of metaphysics and, I shall argue, theology. The
first relates to the question of the beginning and to the traditional 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:
With this framework in mind, let us proceed to examine the issue of the
ostensible beginning of the universe.
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.
Natural theology and modern cosmology 39
The Hartle-Hawking ‘no boundary’ proposal
One such theory is due to the late Stephen Hawking and his colleague James
Hartle. According to the Hartle-Hawking theory, as one goes back towards
the beginning space-time gets ‘smoothed out’ and time becomes like a space
dimension. Time itself is then imaginary in the technical sense of complex
numbers. This four-dimensional space has no boundary or edge. This is what
Hawking called the ‘no boundary proposal.’ It appears in his first book, A
Brief History of Time;6 and it reappears in his later book, co-authored with
Leonard Mlodinow, The Grand Design.7
Hawking seemed to agree with Craig that if the universe had a beginning
then it would need God to create it. He thought that his no boundary 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:
And again:
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
40 Natural theology and modern cosmology
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.
Natural theology and modern cosmology 41
Glass and McCartney’s questions to the no boundary
proposal
In Glass and McCartney’s notation, suppose the evidence E is the existence
of the universe, our initial hypothesis HI is divine creation, and our alterna-
tive hypothesis HA is the no boundary proposal.
“I see nobody on the road,” said Alice. “I only wish I had such eyes,”
the King remarked in a fretful tone. “To be able to see Nobody! And
at that distance too.” . . . “Who did you pass on the road?” the King
went on, holding out his hand to the Messenger. . . . “Nobody,” said
the Messenger. “Quite right,” said the King: “this young lady saw him
too. So of course Nobody walks slower than you.” “I do my best,” the
Messenger said in a sullen tone. “I’m sure nobody walks much faster
Natural theology and modern cosmology 43
than I do!” “He can’t do that,” said the King, “or else he’d have been
here first.”
(Lewis Carroll, Through the Looking Glass, ch. vii)21
Just as the King mistakes the absence of any person for a person called
Nobody, so Krauss et al. ontologize the concept of ‘nothing,’ turning it from
the absence of anything at all to a very sophisticated something.
Well, this nothing of Krauss, Hawking, and others is certainly not nothing
in the sense that it would trouble a theist. These scientists simply have not
proposed a model in which the universe arises spontaneously from literally
nothing without the need for God.
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
Natural theology and modern cosmology 47
and the three bears, the universe seems to be “just right” for life, in so many
intriguing ways.’27
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
48 Natural theology and modern cosmology
rich complexity including living creatures that we see? Given the infinite
variety of outcomes we can imagine, it is desperately puzzling why the only
possible set of laws gives a universe with human beings in it. So, given this
massive puzzle, this seems a not very convincing explaining away of the
need for divine design.
There is a less all-embracing version of strategy 1, namely that the 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.
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.
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
60 Natural theology and modern cosmology
to relate to him. Thus it would seem that a universe is much more likely to
exist given God as prior cause than if there is no cause.
It follows that P[E|H] is much larger than P[E|~H]. Expressing this 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[E | H ] P H
. (3.3)
P[M | E ] P[E | M] P M
P[H | E ] 123
1010 (3.4)
P[M | E ]
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
62 Natural theology and modern cosmology
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-
ence and Religion,’ Theology and Science 12(4) (2014), 338–361.
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
Natural theology and modern cosmology 63
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).
64 Natural theology and modern cosmology
38 Martin Rees, New Perspectives in Astrophysical Cosmology, second edition
(Cambridge: Cambridge University Press, 2000), 138.
39 Martin Rees, Our Cosmic Habitat (London: Weidenfeld & Nicolson, 2001), 164.
40 T. J. Mawson, ‘Explaining the Fine Tuning of the Universe to Us and the Fine
Tuning of Us to the Universe,’ Royal Institute of Philosophy Supplement 68
(2011), 25–50.
41 Nelson Goodman, Fact, Fiction, and Forecast (New York: Bobs-Merrill, 1946).
42 D. K. Lewis, On the Plurality of Worlds (Oxford: Blackwell, 1986).
43 Steven Weinberg, ‘Living in the Multiverse,’ in Carr (ed.), Universe or 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.
52 Swinburne, Existence of God, 133–135.
53 Swinburne, Existence of God, 6, 133–152.
54 Swinburne, Existence of God, 96–109, 165, 185–188.
55 Davies, Goldilocks Enigma, 249.
56 Swinburne, Existence of God, 145–147.
4 Moving on from natural
theology
Why we need ramified natural
theology
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
68 Moving on from natural theology
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:
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:
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
70 Moving on from natural theology
Hebblethwaite argues that ‘Biblical interpreters must keep their eyes open
to the other scriptures and rival revelation-claims in world religion.’ He
moreover recognizes the problem here that a single scholar cannot do justice
to all these revelation claims. As a simple illustration of this he says that he
can learn Hebrew and Greek for Biblical study, but really ought to learn
Sanskrit, Pali, and Arabic for in-depth study of other traditions.19
Richard Swinburne argues that it is not necessary to go down the route of
detailed familiarization with other religions. He was accused by J. L. 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.
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
72 Moving on from natural theology
many modern scholars now (e.g. E. P. Sanders, Anthony Harvey, and Tom
Wright), Dunn sees the ‘cleansing of the Temple,’ which we shall discuss in
Chapter 8, as the primary cause for Jesus’ arrest by the Temple authorities
and execution by Pontius Pilate.27
One would therefore need very strong grounds for denying the 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.
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
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
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
74 Moving on from natural theology
have been more who distinguish between the mortal being who died
and a transcendental being who was unscathed: these are the docetists,
alike in Christianity and in Islam. There have also been those who hold
that Jesus did not die but only swooned: the resuscitation theorists. But
the great majority, if they agree on nothing else, will allow that Jesus
of Nazareth lived and truly died. It was a death, and took place when
Pontius Pilate was prefect of Judaea. So much may be established as a
historical event, as objectively as the death of Julius Caesar.35
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
Moving on from natural theology 75
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,
76 Moving on from natural theology
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 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
78 Moving on from natural theology
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:
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
Moving on from natural theology 79
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.
2 St Thomas Aquinas, Summa Theologiae, Blackfriars edition (London: Eyre &
Spottiswoode; New York: McGraw Hill, 1963), 1a. 2, 2.
3 This chapter is based on my article in that issue: Rodney Holder, ‘Why We Need
Ramified Natural Theology,’ Philosophia Christi 13(2) (2013), 271–282.
4 John Polkinghorne, Quantum Theory: A Very Short Introduction (Oxford:
Oxford University Press, 2002), 88–89.
5 Karl Barth, ‘No!,’ Response to Emil Brunner, ‘Nature and Grace,’ in Emil 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
80 Moving on from natural theology
Holder, The Heavens Declare: Natural Theology and the Legacy of Karl Barth
(West Conshohocken: Templeton Press, 2012), 90–96.
7 Brian Hebblethwaite, The Problems of Theology (Cambridge: Cambridge 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.
11 Wolfhart Pannenberg, Systematic Theology, vol. 1, trans. Geoffrey W. Bromiley
(Edinburgh: T & T Clark, 1991), 128.
12 Pannenberg, Systematic Theology, vol. 1, 51.
13 Wolfhart Pannenberg, Basic Questions in Theology, vol. 1, trans. George H.
Kehm (London: SCM Press, 1970), 60. Originally published as Grundfragen
Systematischer Theologie: Gesammelte Aufsätze (Göttingen: Vandenhoeck and
Ruprecht, 1967).
14 Carl E. Braaten, ‘The Place of Christianity Among the World Religions: 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.
Moving on from natural theology 81
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.
47 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).
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).
49 Wolfhart Pannenberg, Systematic Theology, vol. 2, trans. Geoffrey W. Bromiley
(Edinburgh: T & T Clark, 1994), 360–361.
50 Richard Bauckham, Jesus and the Eyewitnesses: The Gospels as Eyewitness Tes-
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006), 499.
51 Bauckham, Jesus and the Eyewitnesses, 493.
52 Bauckham, Jesus and the Eyewitnesses, 500.
53 Pannenberg, Systematic Theology, vol. 2, 362.
54 Daniel P. Fuller, ‘The Resurrection of Jesus and the Historical Method,’ Journal
of Bible and Religion 34(1) (1966), 18–24, 22.
55 Timothy McGrew and Lydia McGrew, ‘The Argument from Miracles: A 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, 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)
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)
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)
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.
86 Pascal’s Pensées and Butler’s Analogy
A more detailed analysis of Pascal’s Wager, including some objections, is
to be found in Jon Elster’s chapter in The Cambridge Companion to Pascal.7
Besides some technical issues to do with infinities, Elster notes the ‘many
gods’ argument. As we see below, Pascal is aware of the problem of religious
pluralism and argues for the superiority of Christianity, but these arguments
would need to be made to, and accepted by, the person challenged to make
the Wager, so that his choice is a binary one between believing in the true
God and rejecting him. A further difficulty with Pascal is squaring any kind
of argument with his doctrine of predestination, which despite Pascal’s 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
Pascal’s Pensées and Butler’s Analogy 87
divine. In him and through him, therefore, we know God. Apart from
that, without Scripture, without original sin, without the necessary
mediator, who was promised and came, it is impossible to prove 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)
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,
88 Pascal’s Pensées and Butler’s Analogy
regarding Scripture, ‘I admit that there are obscurities as odd as those of
Mahomet, but some things are admirably clear, with prophecies manifestly
fulfilled.’ (L. 218). He cites this obscurity, again making an important point
about non-collaboration between the gospel writers which we shall also
return to in later chapters: ‘Thus all the most obvious weaknesses are really
strengths. Example: the two genealogies of St Matthew and St Luke. What
could be clearer than that there was no collaboration?’ (L. 236).
Pascal is adamant that the prophecies foretell the coming of Christ and
are demonstrative of Christian truth. For example, he writes:
If a single man had written a book foretelling the time and manner of
Jesus’s coming and Jesus had come in conformity with these prophecies,
this would carry infinite weight. But there is much more here. There is
a succession of men over a period of 4,000 years, coming consistently
and invariably one after the other, to foretell the same coming: there is
an entire people proclaiming it, existing for 4,000 years to testify in a
body to the certainty they feel about it . . .
(L. 332)
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)
The hypothesis that the Apostles were knaves is quite absurd. Follow
it out to the end and imagine these twelve men meeting after Jesus’s
Pascal’s Pensées and Butler’s Analogy 89
death and conspiring to say that he had risen from the dead. This means
attacking all the powers that be. The human heart is singularly 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
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)
The prophecies, even the miracles and proofs of our religion, are not of
such a kind that they can be said to be absolutely convincing, but they
are at the same time such that it cannot be said to be unreasonable to
believe in them. There is thus evidence and obscurity, to enlighten some
90 Pascal’s Pensées and Butler’s Analogy
and obfuscate others. But the evidence is such as to exceed, or at least
equal, the evidence to the contrary, so that it cannot be reason that
decides us against following it, and can therefore only be concupiscence
and wickedness of heart. Thus, there is enough evidence to condemn
and not enough to convince, so that it should be apparent that those
who follow it are prompted to do so by grace and not by reason, and
those who evade it are prompted by concupiscence and not by reason.
(In Part III, L. 835)
Here it is not a question of absolute proof but rather of the weight of 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
Pascal’s Pensées and Butler’s Analogy 91
the 4000th Year of the World, meaning that the creation occurred in 4000
BC. A simple sum now gives the date of Christ’s crucifixion:
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
92 Pascal’s Pensées and Butler’s Analogy
oracle in Jeremiah 25:1, from 605 BC to 171 BC, which is the year of the
murder of the second ‘anointed one,’ namely the High Priest Onias III. This
occurred during the reign of Antiochus IV ‘Epiphanes,’ and is described in 2
Maccabees 4. Given the scholarly consensus that Daniel was written in the
Antiochene period, what looks like a prophecy may be considered to be a
vaticinium ex eventu, as the early pagan critic Porphyry saw it.
Does all this mean that we cannot relate the prophecies of Daniel to the
time of Christ? As we shall see in Chapter 8 this is by no means the case.
John Goldingay offers a way to do this. Referring to Daniel 9:24–27, he
says:
The detail of vv 24–27 fits the second-century B.C. crisis and agrees
with allusions to this crisis elsewhere in Daniel. The verses do not 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 . . . .
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
Pascal’s Pensées and Butler’s Analogy 95
it represents a real problem for Hume, since there is an analogy between
not believing in ice and not believing in miracles. Locke himself makes the
point that belief for the prince will be wholly dependent on testimony, and
then even ‘the most untainted Credit of a Witness will scarce be able to find
belief.’23 I argue in Chapter 6, in line with Butler’s cumulative style of 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:
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
96 Pascal’s Pensées and Butler’s Analogy
from a baby in the womb to an adult human, or a worm metamorphosing
into a fly, with that from life to a future state after death (I.I.2, 3). Such
arguments may not be as compelling to us today as Butler found them. He
also sees revelation as confirming natural religion with regard to the moral
governance of the world (e.g. II.VII.32).
Butler sees miracles and prophecies, first, as giving natural religion a
firmer base, since various miracles he lists which Jesus purportedly worked
would support his teaching (II.I.7). But then, in like manner to Pascal, 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)
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
Pascal’s Pensées and Butler’s Analogy 97
therefore ought not to wonder at such unlikeness between things visible
and invisible.
(II.II.5)
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
100 Pascal’s Pensées and Butler’s Analogy
aware that enthusiasts have done this for the ‘most idle follies imaginable.’
However, there is a distinction between dying for opinions and for facts:
And if the apostles and their contemporaries did believe the facts, in
attestation of which they exposed themselves to sufferings and death;
this their belief, or rather knowledge, must be a proof of those facts: for
they were such as came under the observation of their senses.
(II.VII.11)
Also important, though of less weight, is the fact that the martyrs of the next
generation suffered the same fate. They were not themselves eyewitnesses
but had full opportunity to inform themselves of the facts (II.VII.12).
Butler recognizes objections to his argument, anticipating in some ways
the critique of Hume which was to come just a few years later. Thus the case
could be weakened because it arises from ‘enthusiasm’ or the ‘incredibility’
of what is attested or by contrary testimony. The notion of ‘incredibility’
sounds like the concept of a low prior probability. However, Butler does not
accept the incredibility of New Testament miracles and sees enthusiasm as a
far-fetched explanation, in contrast to the ‘direct, easy, and obvious account
of it, that people really saw and heard a thing not incredible, which they
affirm sincerely and with full assurance, they did see and hear’ (II.VII.13).
Even recognizing prejudice and bias in human affairs, testimony is still ‘a
natural ground of assent’ and this assent ‘a natural ground of action’ (II.
VII.16). Here he is anticipating something like Richard Swinburne’s ‘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:
(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.
102 Pascal’s Pensées and Butler’s Analogy
Mitchell makes an important point following his reference to Mackie:
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:
Notes
1 A. J. Krailsheimer, ‘Introduction,’ in Blaise Pascal, Pensées, trans. A. J.
Krailsheimer (London: Penguin, revised edition, 1995), xiii.
2 To be found in Pascal, Pensées, trans. Krailsheimer, Section III, 255–281.
3 L. Lafuma (ed.), Pascal: Oevres complètes (Paris: Éditions du Seuil, Collection
l’Intégrale, 1963).
4 Pascal, Pensées, trans. A. J. Krailsheimer.
5 John Cruickshank, Pascal: Pensées (London: Grant & Cutler Ltd, 1983), 39–40.
6 Cruickshank, Pascal: Pensées, 18.
Pascal’s Pensées and Butler’s Analogy 105
7 Jon Elster, ‘Pascal and Decision Theory,’ in Nicholas Hammond (ed.), The 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.
10 Cruickshank, Pascal: Pensées, 75.
11 Cruickshank, Pascal: Pensées, 74.
12 Krailsheimer, ‘Introduction,’ in Pascal, Pensées, xxv.
13 David Wetsel, ‘Pascal and Holy Writ,’ in Hammond (ed.), The Cambridge 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.
21 Penelhum, Butler, 90–91.
22 David Hume, An Enquiry Concerning Human Understanding, Section X,
‘Of Miracles,’ in the first part of L. A. Selby-Bigge (ed.), Enquiries 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), 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.
25 Penelhum, Butler, 100.
26 Penelhum, Butler, 175.
27 See Penlhum, Butler, 177.
28 Penelhum, Butler, 96–97.
29 David Hume, Dialogues Concerning Natural Religion, ed. J. M. Bell (London:
Penguin, [1779] 1990).
30 See, for example, Joseph A. Fitzmyer, S. J., ‘Introduction to the New Testament
Epistles,’ in Raymond E. Brown, S. S., Joseph A. Fitzmyer, S. J., and Roland E.
Murphy, O. Carm (eds.), The New Jerome Biblical Commentary (London and
New York: Geoffrey Chapman, 1989), 768–771, 770.
31 ‘The First Epistle of Clement to the Corinthians,’ ch. XLVII, in Alexander 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.
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.
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:
For Augustine, the laws of nature, as we discern them, are simply God’s
normal way of working, but any way he worked would in reality be in
accordance with nature, since ‘nature is what He has made.’ Moreover, in
another place Augustine argues that God, the author of nature, can change
the natures of objects so that they exhibit altered properties.8
St Thomas Aquinas (1225–1274) is similar to Augustine, but makes a
distinction between primary and secondary causes. God is the primary cause
of all things—he it is who causes them to exist and gives them the powers
they have to act. But creatures can and do act as secondary causes through
the powers with which they have thus been endowed by God.9 Aquinas
further believes that God is involved intimately in the natural world and
he appeals to Job 10:11: ‘Thou hast clothed me with skin and flesh; thou
hast put me together with bones and sinews.’ Just like Augustine he sees the
development of a human person from the womb to adulthood in terms of
God’s action, but God acting in and through the processes with which he
has endowed nature.
The rationality of belief in miracles 109
For Aquinas, as for Augustine, this does not limit God’s action because
God freely created the secondary causes and they are subject to him. 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
110 The rationality of belief in miracles
Schlesinger’s version of the argument. Again, this topic warrants further
consideration.
In this chapter I establish the Bayesian procedure to be followed, noting
some of the related work done elsewhere, before proceeding to examine the
issues of multiple testimonies and whether testimony for miracles provides
evidence for the existence of God.
Hume writes in the passage quoted of subtracting what he calls ‘degrees
of force,’ which sounds distinctly non-Bayesian. It seems that what he is
saying is that a miracle is an event with very low probability. By describing
testimony as miraculous he presumably means that there has to be a very
low probability that the testimony is false.
In the case when the miracle and the falsehood of the testimony to it are
equally miraculous, which we interpret to mean ‘have equal but very low
probabilities,’ Hume would ostensibly give zero ‘degree of force’ or 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.
The rationality of belief in miracles 111
Following Hume, then, we define a miracle to be ‘a violation of the laws
of nature.’20 It is true that Hume modified this simple definition in a 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
P [ T |M ] P [ M ]
P [ M|T ] =
P[T]
P T |M P M
P M|T (6.1)
P T |M P M P T | ~ M P ~ M
112 The rationality of belief in miracles
In what follows the symbols have these meanings:
> is greater than; ≥ is greater than or equal to; >> is much greater than;
< is less than; ≤ is less than or equal to; << is much less than; ≈ is
approximately equal to; → implies; ↔ implies and is implied by (or ‘if
and only if’).
P[T |M] ≈ 1
P M 1
P[T | ~ M] << 1
P ~ M 1
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):
Or, equivalently:
P M|T 0.5 P M P T | ~ M
Other authors provide interpretations of, or advances on, Hume which are
broadly equivalent to the above, especially when the régime of values taken
by the probabilities is taken into account (see, for example, Owen,25 Dawid
and Gillies,26 Schlesinger,27 Sobel,28 Millican,29 and Earman).30
The rationality of belief in miracles 113
Whether or not Hume meant this, and there is some dispute among 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.’
P[T n | M]P[M]
P[M|T n ] =
P[T n | M]P[M] + P[T n |~ M]P[~ M]
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.
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)
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.
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:
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 rationality of belief in miracles 117
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:
Then
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:
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
= 1 −( 1 − p )
n
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.
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:
P M|T P[G | M T ]
P G|T (6.13)
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
122 The rationality of belief in miracles
analogy here with our discussion of the probability of the occurrence of one
miracle given the occurrence of another. The existence of God would remove
our prejudice that miracles were as unlikely as anything could be, but would
not license our belief in any particular miracle apart from testimony. 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
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
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
The rationality of belief in miracles 123
this is necessarily the case for reasons advanced earlier. If it were the case
it would invalidate the particular approximation here, but strengthen
the overall argument that T provides evidence for G. In chapters 10 and
11, as well as the prior probability of the resurrection of Jesus, I discuss
the prior probability of God becoming incarnate and argue for a ‘mod-
erately low’ value, i.e. less than ½ but not extremely low.
3 P[T|~M ∧ G] ≈ P[T|~M].
P M|G P[T |~ M]
P G|T P M|T P G|M . (6.17)
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
124 The rationality of belief in miracles
In fact, it may well be that P[M|G] >> P[M], so that T (which is virtually
equivalent to asserting that M has occurred) will be very good evidence for
G in this case.
On the other hand, if we have only singular, non-miraculous testimony T
for M, such that P[M] << 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
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,
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
126 The rationality of belief in miracles
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.
2 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), 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.
9 St Thomas Aquinas, Summa Theologiae, Blackfriars edition (London: Eyre &
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.
17 Richard Swinburne, The Existence of God, second edition (Oxford: Oxford Uni-
versity Press, 2004), ch. 12.
18 Richard Otte, ‘Schlesinger and Miracles,’ Faith and Philosophy, 10 (1993),
93–98.
The rationality of belief in miracles 127
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.
37 Mackie, Miracle of Theism, 14.
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.
39 Mackie, Miracle of Theism, 23.
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.
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
Ramified natural theology and Christ 131
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
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
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.
Ramified natural theology and Christ 139
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:
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
140 Ramified natural theology and Christ
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:
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:
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
142 Ramified natural theology and Christ
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
Ramified natural theology and Christ 143
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
144 Ramified natural theology and Christ
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.
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.
Ramified natural theology and Christ 145
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
1 James D. G. Dunn, Jesus Remembered: Volume 1 of Christianity in the Making
(Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2003), 110.
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.
146 Ramified natural theology and Christ
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.
24 Stanton, Gospels and Jesus, 157.
25 Stanton, Gospels and Jesus, 158.
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.
31 Wright, Jesus and the Victory of God, 48.
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).
36 Richard Bauckham, Jesus and the Eyewitnesses: The Gospels as Eyewitness Tes-
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006).
37 Bauckham, Jesus and the Eyewitnesses, 7.
38 Bauckham, Jesus and the Eyewitnesses, 7.
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.
40 Stanton, Gospels and Jesus, 215.
41 Geoffrey Parrinder, Jesus in the Qur’ān (London: Sheldon Press, 1965), 83, and
footnote 1.
42 Stanton, Gospels and Jesus, 218.
43 Wright, Jesus and the Victory of God, 193.
44 Wright, Jesus and the Victory of God, 191.
45 Wright, Jesus and the Victory of God, 189.
Ramified natural theology and Christ 147
46 Wright, Jesus and the Victory of God, 189, quoting Harvey.
47 Wright, Jesus and the Victory of God, 191.
48 Wright, Jesus and the Victory of God, 186.
49 Wright, Jesus and the Victory of God, 187.
50 Wright, Jesus and the Victory of God, 190.
51 Wright, Jesus and the Victory of God, 194.
52 Wright, Jesus and the Victory of God, 194.
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.
60 Dunn, Jesus Remembered, 141.
61 Hooker, Gospel According to St Mark, 71–75.
62 Wright, Jesus and the Victory of God, 191.
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.
65 Hooker, Gospel According to St Mark, 6.
66 Bauckham, Jesus and the Eyewitnesses, 202–239.
8 Jesus and prophecy
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
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
Jesus and prophecy 155
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)
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:
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
Jesus and prophecy 157
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.
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.
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
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
Jesus and prophecy 159
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)
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.
160 Jesus and prophecy
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
Jesus and prophecy 161
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:
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
162 Jesus and prophecy
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:
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.
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
164 Jesus and prophecy
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.
Here I make the point that it is no less ‘fulfilment’ that Jesus knew what he
was doing.
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
168 Jesus and prophecy
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).
14 Westermann, Isaiah 40–66, 257.
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).
Jesus and prophecy 169
17 Hooker, Gospel According to St Mark, 370.
18 Hooker, Gospel According to St Mark, 370.
19 Hooker, Gospel According to St Mark, 371.
20 Hooker, Gospel According to St Mark, 373.
21 Hooker, Gospel According to St Mark, 373.
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.
27 Hooker, Gospel According to St Mark, 164.
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.
29 Wright, Jesus and the Victory of God, 490.
30 Wright, Jesus and the Victory of God, 493–510.
31 Wright, Jesus and the Victory of God, 364–365.
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.
33 Westermann, Isaiah 40–66, 267.
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.
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
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.
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.
178 Ramified natural theology in action
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.’
Resurrection Witnesses
Appearances
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).
180 Ramified natural theology in action
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.
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
Ramified natural theology in action 181
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).
182 Ramified natural theology in action
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’
Ramified natural theology in action 183
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.
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.
184 Ramified natural theology in action
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
Ramified natural theology in action 185
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:
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.
188 Ramified natural theology in action
2 Richard Bauckham, Jesus and the Eyewitnesses: The Gospels as Eyewitness Tes-
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.
4 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), 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.
8 James D. G. Dunn, Jesus Remembered: Volume 1 of Christianity in the Making
(Grand Rapids, MI, and Cambridge, UK: Wm. B. Eerdmans, 2003), 141.
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.
24 Wright, Resurrection of the Son of God, 636.
25 Bauckham, Jesus and the Eyewitnesses, 420–423.
26 Hooker, Gospel According to St Mark, 379.
27 Wright, Resurrection of the Son of God, 205.
28 Wright, Resurrection of the Son of God, 686.
29 Dunn, Jesus Remembered, 339.
30 Wright, Resurrection of the Son of God, 17.
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.
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
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.
192 On the third day he rose again
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
P E|R.P R
P R|E
P E|R.P R P E| ~ R.P ~ R
P[W | R]
= 100
P[W |~ R]
P[D | R]
= 1039
P[D |~ R]
P[P | R]
= 103
P[P |~ R]
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
194 On the third day he rose again
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 ,
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
On the third day he rose again 195
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.
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
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
and h = h1 ∧ h2.
202 On the third day he rose again
Swinburne also very importantly includes in his analysis
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]
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]
P[E | R]
≈ 1044
P[E |~ R]
204 On the third day he rose again
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 ]
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
On the third day he rose again 205
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).
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.
On the third day he rose again 207
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
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
1 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.
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?
(Oxford: Oxford University Press, 2008).
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.
On the third day he rose again 209
6 Richard Bauckham, Jesus and the Eyewitnesses: The Gospels as Eyewitness Tes-
timony (Grand Rapids, MI: Wm. B. Eerdmans, 2006).
7 Wright, Jesus and the Victory of God, 61.
8 Wright, Jesus and the Victory of God, 61–62.
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.
11 Swinburne, Resurrection of God Incarnate, 146.
12 Swinburne, Resurrection of God Incarnate, 148.
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).
14 Swinburne, Resurrection of God Incarnate, 185.
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.
19 Swinburne, Resurrection of God Incarnate, 35.
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.
21 Richard Swinburne, The Existence of God, second edition (Oxford: Oxford Uni-
versity Press, 2004), 96–98.
22 Swinburne, Existence of God, 99–106.
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
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.
214 Towards a fuller picture
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]
>> 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
Towards a fuller picture 215
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:
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[c | T n ∧ E ∧ EP ∧ E4 ∧ k]
P[~ c | T n ∧ E ∧ EP ∧ E4 ∧ k]
P[T n ∧ E ∧ EP ∧ E4 |c ∧ k] P[c|k] (11.4)
= .
P[T ∧ E ∧ EP ∧ E4 | ~ c ∧ k] P[~ c|k]
n
Here, in the first ratio on the right hand side, we have to compare how
likely the evidence is if God becomes incarnate (the numerator) with how
likely this is if either there is a God who does not become incarnate or there
is no God (the denominator). It may be that we are building too much into
where the evidence might lead. It is hard to imagine that we would have
Towards a fuller picture 217
this evidence, which virtually assures us that the prophet did remarkable
things and that his life was crowned by resurrection from the dead, if there
were no God at all. But it might just be that the data are compatible with
a lesser figure—perhaps somebody like a Jewish Messiah who is a 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.
218 Towards a fuller picture
Notes
1 C. S. Lewis, ‘Is Theology Poetry?’ in Screwtape Proposes a Toast and other Pieces
(London: Collins, Fontana Books, 1965, this chapter first read as a paper to the
Socratic Club in Oxford in 1944), 50.
2 Interestingly, even so eminent a Biblical scholar as Charlie Moule is drawn to this
kind of speculation when discussing the ‘ultimacy’ of Christ: C. F. D. Moule, The
Origin of Christology (Cambridge: Cambridge University Press, 1977), 143 ff.
3 Richard Swinburne, The Resurrection of God Incarnate (Oxford: Oxford Univer-
sity Press, 2003), 214.
Bibliography