Top Banner
Philosophy of Religion Gregory W. Dawes Department of Philosophy University of Otago This work is licensed under the Creative Commons Attribution-Share Alike 3.0 License.  You are free to copy and distribute the work, provided you attribute the original material to its author, include the name of the University of Otago, and redistribute it under the same terms.

DAWES-Philosophy of Religion

Apr 14, 2018



Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Page 1: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 1/205

Philosophy of Religion

Gregory W. Dawes

Department of PhilosophyUniversity of Otago

This work is licensed under theCreative Commons Attribution-Share Alike 3.0 License.

 You are free to copy and distribute the work,provided you attribute the original material to its author,

include the name of the University of Otago,and redistribute it under the same terms.

Page 2: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 2/205

Page 3: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 3/205

Table of Contents

A Word to the Reader..................................................................... 7

A Brief Glossary................................................................................. 9

Five Possible Responses to “God Exists”....................................... 13

What is Religion?.............................................................................. 1

1.1 Definitions of Religion............................................................................2

1.2 Theism.....................................................................................................4

Theism and the Religions................................................................. 7

2.1 Restricted and Expanded Theism....................................................... 8

2.2 Defending a Particular Faith..............................................................11

Varieties of Belief............................................................................ 13

3.1 Everyday Uses of “Belief”................................................................... 13

3.2 Belief Without Arguments...................................................................16

The Presumption of Atheism.......................................................... 194.1 Proving God Does Not Exist................................................................19

4.2 The Burden of Proof.............................................................................20

4.3 Merely a Presumption........................................................................21

4.4 Why Presume Atheism?......................................................................23

Reports of Miracles......................................................................... 25

5.1 The Idea of a Miracle......................................................................... 26

5.2 Are Miracle Reports Credible?..........................................................27

Ontological Arguments................................................................. 35

6.1 St Anselm’s Ontological Argument...................................................35

6.2 Alvin Plantinga’s Modal Ontological Argument............................. 41

Cosmological Arguments.............................................................. 47

7.1 Can anything be the cause of itself?...............................................49

7.2 Is the existence of the universe a contingent fact?....................... 50


Page 4: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 4/205

7.3 Do all contingent facts have a cause?........................................... 51

7.5 Bringing Explanation to an End......................................................... 54

Teleological Arguments................................................................. 59

8.1 Traditional Design Arguments............................................................59

8.2 The Fine-Tuning Argument................................................................. 65

Logical Arguments from Evil.......................................................... 75

9.1 Logical and Evidential Arguments....................................................75

9.2 The Logical Argument........................................................................ 76

9.3 Theistic Defences................................................................................ 77

Evidential Arguments from Evil...................................................... 8310.1 Paul Draper’s Hypothesis of Indifference.......................................83

10.2 The “Sceptical Theism” Defence.................................................... 86

Atheism as a Moral Imperative..................................................... 95

11.1 Ivan’s Rebellion................................................................................. 95

11.2 Why Rebel?........................................................................................98

11.3 “God’s Ways Are Not Our Ways”....................................................99

Faith and Reason: An Evidentialist View.................................... 103

Faith and Self-Authentication..................................................... 107

13.1 The Aquinas-Calvin View ............................................................... 107

13.2 The Bootstrapping Problem............................................................ 110

Evidentialism Revisited................................................................. 113

14.1 Obligations in Matters of Belief.......................................................114

14.2 Evidentialism Defined..................................................................... 115

14.3 Three Questions...............................................................................117

14.4 The Challenge to Evidentialism...................................................... 119

Reformed Epistemology.............................................................. 125

15.1 Christian Faith as Basic Belief.........................................................125

15.2 Christian Faith as Warranted Belief............................................... 127

15.3 Christian Faith as Undefeated Belief.............................................129

15.4 Against Reformed Epistemology....................................................132


Page 5: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 5/205

Religious Experience.................................................................... 139

16.1 What is Religious Experience?....................................................... 139

16.2 Perceiving God............................................................................... 142

Prudential Arguments.................................................................. 149

17.1 Pascal and His Wager.................................................................... 149

17.2 William James and the Will to Believe.......................................... 152

17.3 Richard Swinburne’s Pragmatic Grounds.................................... 155

The Kantian As If........................................................................... 161

18.1 The Autonomy of Ethics..................................................................161

18.2 The Postulates of Practical Reason...............................................163

18.3 A Kantian View of Faith................................................................... 164

18.4 A Brief Evaluation............................................................................ 165

The Wittgensteinian View............................................................ 167

19.1 A Preliminary Statement.................................................................167

19.2 D. Z. Phillips on Religion...................................................................168

Theological Non-Realism............................................................. 173

20.1 The Value of Religion......................................................................173

20.2 The Function of God-Talk............................................................... 174

20.3 The God of Abraham, Isaac, and Jacob.................................... 175

20.4 Why Not Simple Atheism (or Secular Humanism)?......................177

Works Cited................................................................................... 179


Page 6: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 6/205

Page 7: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 7/205

A Word to the Reader 

 Analytic philosophy is not, of course, either a method or a doctrine;

it is a tradition and an attitude.

(Donald Davidson)

The Coursebook you have in your hands serves as the textbook forPHIL 229/239, Philosophy of Religion. (Students will receive a sepa-rate Course Outline – at no extra cost – containing practical informa-tion regarding such matters as internal assessment.) The Course-

book follows the outline of the lectures – each chapter corresponds toat least one lecture – and contains the material I shall discuss inclass. To gain the most from your course of study, you should readthe material relating to each class before attending. That will meanreading about seven pages (on average) for each session. Incidental-ly, having read this material, don’t think you can safely miss theclass. Students have done this in the past and have failed the courseas a result. Attendance at class is vital. Those students who do bestin philosophy are generally those who have raised questions and tak-

en part in our class discussions.Philosophy is difficult. So don’t be alarmed if, on first reading thematerial, you have only the faintest idea what it is about. Don’t pan-ic! Take it slowly. Read the material once, go away and have a coffee,then come back and read it again, highlighting the main points.Then write down some questions and bring them to class. By thetime you have done this and attended the corresponding lecture, theissues should be reasonably clear. If they are not, contact me, so wecan discuss it. Don’t let yourself get behind in your reading and com-

prehension. And keep in mind that for the exam, you will have achoice of topics. You don’t have to achieve equal mastery of all the is-sues dealt with in the course.

I have written this Coursebook so that it forms one long argu-ment, and I make no secret of where I think that argument leads. You don’t have to agree with me. University lecturers often say thisand students rarely believe them. But it is true. We welcome dis-agreement, particularly in philosophy. But of course what we wel-


Page 8: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 8/205

come is reasoned disagreement. You’ll need to produce an argumentin support of your view. In exams and assignments you will bemarked on the quality of your reasoning, not on whether you agreewith your lecturer. And I hope we can have some lively discussion.There is only one requirement here, namely that the discussion beconducted in a respectful manner. Charles Pigden tells the story of the time the Department was visited by the great Karl Popper. Onemember of staff, Pavel Tichý, presented a seminar paper which at-tacked one of Popper’s favourite ideas. In Charles’s words,

Popper had recently proposed a definition of closeness to truth, whichwas intended to explicate the intuitive idea that one false theory canbe closer to the truth than another. Tichý demolished this definitionwith a proof that on Popper’s account all false theories are equally farfrom the truth, finishing in a typically downright manner: “I concludethat Popper’s definition is worthless.” There was a pause as everyoneawaited the response of the notoriously temperamental Popper. Whenit came it was remarkably gracious: “I disagree with only one word of this paper – its last word. No definition can be worthless, when it pro-vokes such a devastating criticism. I hope that Dr Tichý will join mein this project, and produce a better definition than mine.”

That’s the spirit in which we aim to do philosophy at Otago.


Page 9: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 9/205

A Brief Glossary

Much discussion in the philosophy of religion revolves around the ex-istence of God. There are, of course, many conceptions of God or thegods. But as I shall explain in a moment, when philosophers use theword, they normally mean the omniscient (all-knowing), omnipotent(all-powerful) and morally perfect being of Jewish, Christian, andMuslim monotheism. As you begin reading these debates, the follow-ing definitions of commonly used terms may be useful.

1. AgnosticismThe agnostic is one who holds that we do not know whether Godexists. A weak form of agnosticism [let’s call it  Agnosticism1]holds that we do not in fact know whether or not God exists, al-though the question is resolvable in principle. (We can conceive of what the evidence would look like. The problem is merely that theevidence we have is not decisive.) A strong agnosticism [let’s call it Agnosticism2] goes further and affirms that we could never knowwhether or not God exists. (We could never be in possession of the

kind of evidence that would resolve the question.)

2. Atheism

The broadest definition of atheism (that offered by Antony Flew:see the “Presumption of Atheism” below) suggests that atheismcovers any position which is “not theism,” that is to say, not belief in God. The atheist in this sense of the term is not necessarily op-posed to belief in God; she simply doesn’t share that belief. (Onthis definition, oddly enough, agnosticism would be a subspecies of 

atheism.) Flew calls this negative atheism. But of course most whocall themselves atheists are opposed to belief in God, for a varietyof reasons. They are at least reasonably confident that there is noGod. (In Flew’s terms, their position is that of  positive atheism,“positive” in the sense that they are making some assertion aboutGod’s non-existence.) But the reasons they have for holding to thisview can differ in significant ways. I shall distinguish the various


Page 10: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 10/205

forms of positive atheism as follows. (For a slightly different expo-sition, see “Five Possible Responses” below.)

 Atheism(1): A first group of (positive) atheists refuse to an-swer the question “Does God exist?” since they believed that nomeaning can be attributed to this proposition. (One version of this view held that only propositions that could be tested empir-ically are meaningful, and – some people held – “God exists”cannot be tested empirically.) So the question “Does God exist?”is equivalent to “Does the Smyrglebub exist?” It’s just nonsense.

 Atheism(2): A second group of (positive) atheists accept that“Does God exist?” is a meaningful question, but they insist thathis existence can be ruled out a priori because the very conceptof God is incoherent. They argue, for instance, that there couldnever exist a being who is both supremely good and completelyfree. So we do not need to examine the evidence which theistsproduce in favour of his existence. Theists have condemnedthemselves by coming up with such a crazy idea.

 Atheism(3): A third group of (positive) atheists accept thatsuch a being as God could exist, but argues that the evidence

demonstrates or suggests he does not. Those who believe thatGod’s non-existence can be demonstrated hold that there issome proposition about the world – such as “Evil exists” – which is simply inconsistent with the proposition that God ex-ists. One or the other must be false, and since we are more con-fident of the existence of evil (for instance) than the existence of God, we can conclude that God does not exist. (See my discus-sion of the “logical problem of evil.”) Other atheists hold thatthere is some fact about the world which renders the existence

of God unlikely. (See the “evidential problem of evil.”)

3. Henotheism

Henotheism is the recognition of one God for the purposes of devo-tion and worship without denying the existence of other divine be-ings.


Page 11: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 11/205

4. Monotheism

Monotheism is the belief that there exists just one God, generally

thought of an omnipresent, omnipotent, omniscient, supremelygood being, who created the universe and maintains it in exis-tence. When philosophers talk about theism, they generally meanmonotheism, although on a broader definition of theism polytheistare theists, too. (See theism below.)

5. Theism

There is a sense in which everyone who believes in the existenceof spiritual beings (Greek theoi “gods”) to whom one can pray and

who can influence human affairs is a theist. Theists in this sensemay be divided into monotheists and polytheists, believers in oneGod and in many gods. But in practice when philosophers talkabout theism, they are using the word in a narrower sense. Theyuse “theism” to refer to what I call “classical monotheism” or “clas-sical theism,” namely the belief that there exists an omnipresent,omnipotent, omniscient, supremely good being, who created theuniverse and keeps it in existence. This should be distinguishedfrom other forms of (mono-)theism such as “process theism” (see

below).6. Panentheism

Panentheism is the belief that while there is some distinction be-tween God and the world, everything is “within God,” in a waywhich makes God and the world interdependent. Some forms of process theism (see below) are panentheistic.

7. Pantheism

Pantheism is the belief that everything is God or that the naturalworld is in some sense divine.

8. Polytheism

Polytheism is belief in many gods, regarded as supernatural be-ings to whom one can pray and who influence human affairs. Apolytheist’s gods, it should be noted, are not necessarily omni-scient, omnipotent, and supremely good.


Page 12: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 12/205

9. Process theism

Process theism is a variety of monotheism which differs from clas-

sical theism in holding that the divine nature is not unchanging(immutable), but is in a process of development along with theworld. In practice, this allows process theists to deny that God isomnipotent in any traditional sense. Divine power, on this view, ispersuasive rather than coercive.

10. Tritheism

Belief in three Gods (or, perhaps more appropriately, three gods)This terms is sometimes used in discussions of the Christian doc-

trine of the Trinity (the belief that God exists in three ways: Fa-ther, Son, and Holy Spirit), which mainstream Christians insist isa variety of monotheism. However, at least one Christian group(the Church of Jesus Christ of Latter-Day Saints / Mormons) do infact seem to be tritheists.


Page 13: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 13/205

Five Possible Responses to “God Exists”

(I am assuming the question is to be answered by an appeal to reason,not by an appeal to faith; see Part Three: Faith and Reason.)

1. The proposition is meaningless. [ Atheism(1)]2. (a) The proposition is definitely false, because the idea of God

can be demonstrated to be incoherent. [ Atheism(2a)]

(b) The proposition is  probably false, because there are goodreasons to believe that the idea of God is incoherent.

[ Atheism(2b)]3. (a) The proposition is definitely false, because there exists avalid deductive argument from indisputably true premisesleading to the conclusion “God does not exist.” [ Atheism(3a)]

(b) The proposition is  probably false, because there exists avalid deductive or inductive argument, from premises that wehave good reason to think are true to the conclusion “God doesnot exist.” [ Atheism(3b)]

4. (a) The proposition is definitely true, because there exists a

valid deductive argument from indisputably true premisesleading to the conclusion “God exists.”) [Theism(1)]

(b) The proposition is probably true, because there exists avalid deductive or inductive argument, from premises that wehave good reason to think are true to the conclusion “God ex-ists.”) [Theism(2)]

5. We cannot know if the proposition is true or false

(a) since the evidence is evenly balanced [ Agnosticism(1)] or

(b) since we could never have the kind of evidence that couldsettle the issue. [ Agnosticism(2)]


Page 14: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 14/205

Page 15: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 15/205

Chapter One

What is Religion?

It is notoriously difficult to produce an adequate definition of reli-gion. There are many reasons for this. The first is that almost anyproposed definition leaves out some set of beliefs and practices thatwe normally regard as religious. For instance, if we assume that reli-gion involves belief in gods, what about Theravada Buddhism? MostBuddhists believe in gods, but you do not need to believe in gods to

be a Buddhist (Williams and Tribe 2000: 4–5). A second difficultyarises from the fact that our modern concept of a “religion” has cul-turally specific origins. It emerged within the worlds of Judaism,Christianity and Islam and its applicability to other cultural con-texts is debatable. Is there any such thing as “Hinduism,” for in-stance, or did Western observers invent the term as a way of singlingout and then lumping together a series of quite diverse Indian cul-tural traditions? A third, related difficulty has to do with the use of the term “religion.” In the academic study of religions, it is used to

refer to a particular set of beliefs and practices (e.g. “Christianity”),comparable to other, parallel sets of beliefs and practices (e.g. “Is-lam”). The problem is that there are other, quite different ways inwhich the term is used (Smith 1978: 15–50). And there is a sense inwhich no serious believer regards his religion as merely “a religion,”comparable to that of other people. He is more likely to regard otherpeoples’ religions either as mere human inventions, of no religiousvalue, or as corruptions of his own divinely-revealed truth. MedievalEuropean thinkers, for instance, tended to regard Islam, not as an-

other religion parallel to Christianity, but as a heresy (Daniel 1993:209–13). And of course in reality religions are not self-sufficient andclearly-defined units. Religious traditions overlap, have fuzzy edges,and are dependent on each other in complex ways.


Page 16: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 16/205

What is Religion?

1.1 Definitions of Religion

Despite these difficulties, many attempts have been made to offerdefinitions of religion. There are at least four classes of definition onoffer (Clark and Byrne 1993: 4–6). A first category is that of experi-

ential or attitudinal definitions. These seek what is common to reli-gions in some alleged experience of the divine or – in more detachedtreatments – in the attitude of believers towards what is thought of as divine. Rudolf Otto (1869–1937), for example, who was a Chris-tian theologian and scholar of religions, believed the heart of religion

to be the experience of “the holy,” as a mysterium tremendum et fascinans – an awe-inspiring but attractive mystery [Otto 1950: 5– 7].) Similarly, Émile Durkheim (1858–1917), one of the founders of sociology, thought that the common feature of all religions was a cer-tain attitude towards sacred things, which are thought of as beingradically different from profane realities. Communication with thesacred can be dangerous, and those who wish to enter the world of the sacred must first be transformed (Durkheim 1915: 38–40).

 A second category is that of  substantive definitions of religion,

which focus on the content of religious beliefs. E. B. Tylor (1832– 1917), for instance, who was the founder of the field of anthropology,described religion simply as “belief in Supernatural Beings” (Tylor1913 1.424). This view dropped out of favour for a time, but has re-cently been revived by Robin Horton, who has written on African re-ligions and who unashamedly describes his approach to religion as“neo-Tylorian” (Horton 1993: 53). But a new generation of scholarshas come to be interested in the content of religious beliefs, this timefrom the point of view of cognitive science. A leading figure here is

Dan Sperber, whose ideas have been taken up by thinkers such asJustin Barrett, Pascal Boyer, and Scott Atran. As we shall see in amoment, philosophers have always been predisposed to think of reli-gion in terms of religious belief, and this course will take for granteda substantive definition of religion.

 A third category is that of  functional definitions of religion, whichdefine religion in terms of the role religious beliefs and practices play


Page 17: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 17/205

What is Religion?

in people’s lives. Their question is not so much “What is religion?” as

“What do religious beliefs and practices do?” What function do youthey serve, in the lives of individuals or society as a whole? A wellknown definition of religion which is, broadly speaking, functional isoffered by the anthropologist Clifford Geertz. A religion, he writes, is

(1) a system of symbols which acts to (2) establish powerful, perva-sive, and long-lasting moods and motivations in men by (3) formu-lating conceptions of a general order of existence and (4) clothingthese conceptions with an aura of factuality so that (5) the moodsand motivations seem uniquely realistic. (Geertz 1993: 90)

 A perhaps more clearly functional definition (but one which focus-es on religious beliefs) is offered by the philosopher Keith Yandell. Areligion, Yandell writes, is “a conceptual system that provides an in-terpretation of the world and the place of human beings in it, basesan account of how life should be lived given that interpretation, andexpresses this interpretation and lifestyle in a set of rituals, institu-tions, and practices” (Yandell 1999: 16.)

 All these attempts at definition follow one of two paths (Alston

1967: 140–41). Some seek to identify what philosophers call neces-sary and sufficient conditions, that is to say, features that are com-mon to all religions and characteristic only of religion. Others try toidentify what we might call the essence of religion: the central fea-ture that makes something religious. But there is a third path, advo-cated by many recent thinkers. These argue that the best we canachieve is a  family-resemblance approach to religion. (See, for in-stance, Alston 1967: 141–43.) The family-resemblance view arguesthat we are justified in calling these diverse cultural phenomena “re-

ligions” even if there is no single quality they all have in common.What unites religions is “a complicated network of similarities over-lapping and criss-crossing,” to use Ludwig Wittgenstein’s phrase(1958: §66 [32]). In this respect, we may compare the term “religion”to the term “game.” It seems there is no single feature that is com-mon to all games. You can’t define a game as something played forpleasure (some are played for profit), or as being competitive (someare played by oneself), or as having winners and losers (consider a


Page 18: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 18/205

What is Religion?

game of make-believe).Yet if we regard activity A as a game (it might

be our paradigm case of a game, in the sense of an undisputed appli-cation of the term), we may regard activity  B as a game if it is insome way similar to  A. Activity C may be then regarded as a gameby virtue of its similarity to B, and so on.

There is something to be said for each of these approaches. Whichyou favour will depend on what you are trying to do. If you feel theneed to justify the use of the term “religion” for phenomena as differ-ent as Confucianism and Orthodox Judaism (for example), thenWittgenstein’s discussion of  family resemblances is a good place to

start. Anthropologists, because they are interested in the role playedby religion in particular societies, are more likely to adopt a  func-

tional definition. Philosophers, in turn, have a different set of inter-ests. They are interested in the justification (or at least the rationali-ty) of religious beliefs. This leads them in the direction of a substan-

tive definition of religion. Religion, or at least the particular form of religion they are studying, is defined in terms of a certain set of be-liefs.

1.2 Theism

The form of religion that philosophers of religion generally take astheir starting-point – this course will be no exception – is what Ishall describe as classical theism. Classical theism – henceforth sim-ply “theism” – is a philosophical statement (what some philosopherscall a “rational reconstruction” [Reichenbach 1938: 5–6]) of the un-derlying principles of the three great Middle-Eastern religions: Ju-daism, Christianity, and Islam. More precisely, theism is a formula-tion of the doctrine of God that is to be found in the classical thinkersof all three traditions. One thinks, for instance, of Moses Mai-monides (1135–1204) in Judaism, Thomas Aquinas (ca. 1225–74) inChristianity, and Abu Hamid Muhammad al-Ghazali in Islam(1058–1111). What is this foundational theistic doctrine? It issummed up by Richard Swinburne as the view that there exists agod who is


Page 19: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 19/205

What is Religion?

a person without a body (i.e. a spirit), present everywhere, the cre-

ator and sustainer of the universe, a free agent, able to do every-thing (i.e. omnipotent), knowing all things [omniscient], perfectlygood, a source of moral obligation, immutable, eternal, a necessarybeing, holy, and worthy of worship. (Swinburne 1979: 2)

There exist Jewish, Christian, and Muslim thinkers who wouldnot subscribe to aspects of this definition – process theologians, forinstance, deny that God is immutable – but it does represent the his-torically dominant view of God within these traditions. But of courseJews, Christians, and Muslims believe in much more than the exis-

tence of God. (Jews, for instance, believe the Torah to have been giv-en to Moses on Mt Sinai, Christians believe Jesus to be the Son of God, and Muslims believe Muhammad to be the last and greatestprophet.) How does theism relate to the rest of what they believe?That’s our next topic.


Page 20: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 20/205

Page 21: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 21/205

Chapter Two

Theism and the Religions

We have seen that there are various ways of defining religion. Butwhich of these shall we be employing here? What is the “religion”that philosophers study? What we will be analysing in this course isnot any particular religious tradition. We will not be looking at what

it means to be, for instance, a Jew, a Christian, or a Muslim. (If youare interested in that, enrol for a course in Religious Studies.) As Inoted earlier, what the philosophy of religion takes as its startingpoint is a kind of  rational reconstruction of the central beliefs thatunderlie those three monotheistic faiths. It sets these beliefs in or-der, makes them as consistent as possible, and draws out their impli-cations. The resultant structure of beliefs is what the philosopherrefers to as classical theism or simply theism.

The core theistic belief – the one that will be the particular focus

of this course – is belief in God. What do we mean by “God” in thiscontext? I’ve already given you Richard Swinburne’s account of whathe, as a Christian philosopher, understands by “God.” So it might behelpful to offer here Antony Flew’s definition. It is very similar toSwinburne’s, but comes from one who was less sympathetic to reli-gion. “The word God,” Flew writes, means “a Being which is unique,unitary, incorporeal, infinitely powerful, wise, and good, personal butwithout passions, and the maker and preserver of the universe”(Flew 1966: 2.7 [28]). It is important to realise that theists and athe-

ists are not in disagreement about the meaning of the term “God.”Swinburne and Flew agree on what God would be like, if he existed.What they argue about is whether he exists.

But why begin with this concept of God? Why take “theism” to beour paradigmatic set of religious beliefs? Does this not betray a sim-ple cultural bias? Is it not yet another example of ethnocentrism,perhaps Eurocentrism? After all, Judaism, Christianity, and Islamare not the only religions of the world; there are lots of interesting al-


Page 22: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 22/205

Theism and the Religions

ternatives on offer. (Theravada Buddhism would be, perhaps, the

most fascinating for a philosopher.) Well, to some degree thesecharges are justified. Philosophers of religion are not great intellec-tual travellers; they don’t go far from home. But in this case thereare good reasons for staying at home. For better or for worse, ourown culture has been deeply shaped by Christian beliefs. And Chris-tianity remains the dominant form of religion in a country such asNew Zealand. When people discuss, for instance, the role of religionin public life, it is normally Christian theism that they have in mind.The other reason for sticking with (Christian) theism is that it repre-

sents a clearly-defined position which has been the subject of philo-sophical scrutiny for more than a thousand years. So there is a longtraditional of philosophical engagement here from which we can ben-efit. So, given that we to start somewhere, this seems as good a placeas any.

2.1 Restricted and Expanded Theism

I have defined theism in terms of belief in a creator God and have setout a couple of descriptions of what we mean by “God.” You may

have noticed how minimal such definitions are when compared to therich bodies of doctrine we find in actual religions. Judaism, for in-stance, certainly takes for granted the existence of such a God, but itgoes on to speak of God’s choice of Abraham and his descendants, thegiving of the Torah (Law) to Moses on Sinai, the revelations madethrough the prophets throughout Israel’s history and the varying re-sponses that Israel made to these revelations. Christianity builds onthese ideas to speak of the dire state of human beings, entrapped bysin, the sending of the Son of God in human flesh to deal with sinand its consequences, the founding of the Church and the sending of the Holy Spirit, as well as the promised consummation of this planin the end-times. Islam has its own account of God’s dealings withhumanity, culminating in the sending of the prophet Muhammad asthe last and greatest of God’s messengers, who came with guidancefor humanity. To be a theist has traditionally been thought to be a


Page 23: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 23/205

Theism and the Religions

necessary condition of being a religious Jew, a Christian or a Muslim,

but it is by no means a sufficient condition.Must a Jew, Christian, or Muslim subscribe to the doctrine I am

calling theism? Could you not be a non-theistic Jew, Christian, orMuslim? Well, there are those who argue that you do not need to be atheist in order to be a Jew or a Christian. (As it happens, I know of no Muslim thinker who argues this.) There are two possible ways of being a non-theistic Jew, Christian, or Muslim. The first is to contin-ue to maintain that God exists, but to deny that he has the at-tributes traditionally ascribed to him. This might involve, for in-

stance, denying that God is omnipotent or omniscient. (Alfred NorthWhitehead’s metaphysics is often drawn on here, in support of whatI have already referred to as “process theism.”) Or she might deny

that there exists any such supernatural being, while still believingthat talk about God serves some useful function and should be pre-served. (I shall discuss this second position later under the headingof “Theological Non-realism.”)

Both options are very much minority views. As for the first, mosttheologians believe that you cannot be a Jew, Christian, or Muslim

without believing that God is omnipotent, omniscient, and morallyperfect. For a lesser deity, they believe, would not be worthy of wor-ship, nor would he be able to perform the deeds attributed to him insacred scripture (such as creating the world or redeeming us fromsin). Could a less than omnipotent deity create the world e nihilo

(from nothing)? Would a less than morally perfect being be worthy of our worship and unconditional devotion? I’m not going to try to an-swer these questions here; I shall merely note that many religiousthinkers have answered “No.” As for the second, non-realist option,

this has been embraced by a very small number of free-thinkingJews and Christians. So even if one could be a Jew, Christian, orMuslim without being a theist, most Jews, Christians, and Muslimsare, in fact, theists. And in defining theism as I have done, I ammerely following the example of Jewish, Christian, and Muslimthinkers.


Page 24: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 24/205

Theism and the Religions

So much for the idea that you must be a theist in order to be a be-

lieving Jew, Christian, or Muslim. The chief point I wish to makehere is a different one. It is that Jews, Christians, and Muslims arenot merely theists. They believe in God, but they believe in lots of other things as well. William Rowe has highlighted this point bymeans of a distinction between restricted theism and expanded the-

ism. Theism as I have defined it corresponds to Rowe’s “restrictedtheism.” It is the belief that there exists an omnipotent, omniscient,perfectly good being who created the world. Expanded theism is thesame belief “conjoined with certain other significant religious claims,

claims about sin, redemption, a future life, a last judgment, and thelike” (Rowe 1990: 161). So expanded theism takes a variety of forms,depending on the religious tradition with which one is dealing. WhileI shall normally use the unqualified term “theism” in the discussionsthat follow, it should be understood in the sense of Rowe’s restricted

theism. And in practice, the only form of expanded theism we shallbe dealing with is Christian theism, whose doctrines are appealed toby the proponents of “Reformed Epistemology.”

This distinction has an interesting implication. After examining

the arguments, you might be convinced by one or more of the tradi-tional proofs for the existence of God, but still have no reason to joinone of our actually-existing religious communities. In Rowe’s termi-nology, you might be convinced of the truth of  restricted theism, butremain unconvinced regarding any particular form of expanded the-ism. You might hold, for example, that the traditional proofs of God’sexistence show there is a God, but they do not support belief in this

 particular one. So whether or not one can be a Christian (for exam-ple) without being a theist, it does seem possible to be a theist with-

out being a Christian (Jew or Muslim). This option, or something re-sembling it, is sometimes referred to as “deism.” It is a historicallysignificant position – one can argue that many of the founding fa-thers of the United States were deists rather than traditional Chris-tians – but once again it is a minority view.


Page 25: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 25/205

Theism and the Religions

2.2 Defending a Particular Faith

Contemporary discussions in the philosophy of religion sometimes gobeyond the traditional proofs of the existence of God. The focus of thephilosophy of religion remains arguments in support of theism ingeneral (i.e. restricted theism), but a number of contemporary Chris-tian philosophers have developed arguments in support of Christian

belief (i.e. one form of expanded theism). Such arguments are putforward both by those who hold to a traditional, realist understand-ing of religious language (who believe that God exists) and those who

hold to a non-cognitivist (or non-realist) view. Among the formergroup are Richard Swinburne, Alvin Plantinga and William Alston. Among the latter are thinkers such as D. Z. Phillips and Don Cupitt.

So when we approach the work of a theistic philosopher, it is im-portant to understand just what he or she is trying to do. Is she de-fending theism in general or Christianity in particular? Of course, if you could produce a defence of Christianity, it would entail a defenceof theism. If Christianity is true, then God exists. So if it is rationallydefensible to profess the Christian faith, it is it rationally defensible

to believe in God. (In Rowe’s terms, expanded theism entails restrict-ed theism.) But to defend Christian belief (rather than restricted the-ism) will require a distinctive style of argumentation.

Insofar as contemporary Christian philosophers defend the Chris-tian faith, their arguments  go beyond the traditional proofs of God’sexistence. If these arguments succeed, they support adherence to aparticular faith in a way in which merely proving God’s existencewould not. But some of these philosophers, such as Plantinga and Al-ston, do not go as far as their predecessors. Their predecessors

thought they could demonstrate the truth of the Christian faith. (Inour own day, Richard Swinburne shares this aim.) But thinkers suchas Alston and Plantinga defend a weaker claim, namely, that a per-son may be acting rationally in holding to her Christian commit-ment. It is important to distinguish these two aims and to under-stand just what it is a particular philosopher is attempting toachieve.


Page 26: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 26/205

Page 27: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 27/205

Chapter Three

Varieties of Belief

In the philosophy of religion, we examine the arguments that havebeen offered for and against particular religious beliefs. Above all,we examine arguments for and against belief in God. But beforelaunching into these debates, there are still some issues that need tobe clarified. The most fundamental has to do with “belief.” If  p repre-

sents any proposition – a statement about a possible state of affairs(e.g. “There is a coffee cup on my desk.”) – then what does it mean to“believe p”? And what makes my belief that  p a  justified belief, onethat I have sufficient reason to regard as true? Do all our beliefsneed to backed up by evidence, in order to be justified? Or are someof our beliefs justified in some other way? And what do religious be-lievers mean when they refer to “faith”? How is religious faith relat-ed to other forms of belief? Is religious faith really a matter of “be-lieving what you know ain’t so,” as Mark Twain suggested? Could

you “believe what you know ain’t so”? Does that idea make anysense?

3.1 Everyday Uses of “Belief”

I shall come back to the question of religious faith – what philoso-phers and theologians call the question of “faith and reason” – inPart Three. But let me begin by looking at several different ways inwhich we use the terms “belief” and “faith.” In everyday life, we oftenuse the phrase “I believe” to indicate a degree of assent to a proposi-

tion that falls short of knowledge. “Is Heather in London? I believeso” suggests I have some reason to think Heather is in London, but Iam not confident about my assertion. I lack the evidence whichwould enable me to affirm, without any qualification, “Heather is inLondon.” On this view, “I believe there is a God” would be equivalentto “I have good (although not decisive) reasons to think there is aGod.” As we shall see shortly, this is not unrelated to at least one


Page 28: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 28/205

Varieties of Belief

view of religious faith. But it is not the way in which philosophers

customarily use the term “belief.”When philosophers speak of people’s “beliefs” they are generally

using the word in a broader sense. To say “I believe  p” in this philo-sophical sense is simply to say that I hold it to be true. It is not to ex-press any doubt about its truth. When a philosopher says “I believethere is a pen on the table”, he is probably about to demonstrate apoint in epistemology (the philosophy of knowledge). He is not indi-cating that he is uncertain about the existence or location of the pen.When we discuss the arguments for and against the existence of 

God, we are using the word “belief” in this broad, philosophicalsense. What we are asking is: Do we have good reason to believethere is a God? Do we have good reason to hold that the proposition“God exists” is true?

Is this ordinary, philosophical sense of “believe” what religiouspeople mean by “faith”? Well, sometimes it is. There exists what Ishall call an evidentialist view of religious faith, one of its first expo-nents being John Locke (1632–1704). According to the evidentialistview of faith, religious faith has a number of components. It has, for

instance, an affective component (it engages our emotions) and it de-mands that we adopt a particular way of life. But it also has its cog-nitive component, that is to say, it claims to be a kind of knowledge. And this cognitive component is nothing other than “belief” in the or-dinary philosophical sense. This means that the strength of one’s be-lief in God is, or ought to be, directly proportioned to the evidence. If there is good evidence that God exists, the appropriate response forsomeone whose cognitive faculties are functioning properly is to be-lieve in God. And if there is good reason that God has revealed cer-

tain propositions, then there is just that much reason to believe thatthose propositions are true.

This evidentialist view of faith is philosophically unproblematic,since it suggests that religious faith is to be assessed on the same cri-teria as any other form of belief. Much of this course – Parts One andTwo – involves studying arguments for and against the existence of God. In studying these arguments, I shall be taking for granted an


Page 29: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 29/205

Varieties of Belief

evidentialist view of faith. I shall be assuming that belief in God is

 justified only to the extent, and precisely to the degree, that it is sup-ported by evidence. Most philosophers make a similar assumption.Indeed they find it hard to make sense of any other way of thinkingabout religious belief. There is a sense in which this evidentialistview of faith does away with the distinction between faith and rea-son. Religious faith, properly understood, is simply “an assent found-ed on the highest reason” (Locke 1846: 510).

The problem is that this evidentialist view of religious faith doesnot correspond to what most religious believers mean by “faith.” For

religious believers customarily speak as if faith were something dis-tinct from reason. Let me illustrate this by making a prediction. If you tell people that you are studying arguments for and against theexistence of God, before long someone will say “But isn’t that a mat-ter of faith?” This response suggests that belief in God is indepen-dent of the arguments for and against his existence. And to say, asbelievers often do, that we know something “by faith” is to suggestthat faith is a means of knowing that is distinct from reason. Butdoes this make any sense? Are there legitimate claims to knowledge

which do not rely on reason? Can we know something in the absenceof evidence that it is true? As I mentioned a moment ago, I’ll be ad-dressing this issue – the appeal to faith as distinct from reason – inPart Three.

There exists, by the way, a third view of religious faith. This holdsthat faith is a commitment which is made in the absence of knowl-edge. (In this context, I shall adopt the traditional understanding of knowledge as “justified true belief,” leaving aside the various objec-tions that have been raised to this notion in the wake of the work of 

Edmund Gettier [1963].) On this third view, religious faith is not ameans of attaining justified true belief. It is a way of life embracedand maintained in the absence of justified belief. More precisely, it isa way of life embraced and maintained in the absence of the kind of evidence that would constitute a claim to knowledge. This seems tobe what Immanuel Kant (1724–1804) had in mind when he spoke of “denying knowledge, in order to make room for  faith” (Kant 1929:


Page 30: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 30/205

Varieties of Belief

29 [Bxxx]). I shall also examine this view of religious faith later, un-

der the heading “Living As If God Existed” (Part Four).

3.2 Belief Without Arguments

Could one have a philosophically-defensible belief in God withouthaving arguments in support of that belief? If you are the kind of ra-tionally-minded person who studies philosophy, you may think thisis an outrageous idea. You may think that all our beliefs are (or atleast ought to be) held on the basis of sound arguments. But in prac-tice, of course, we do not hold all our beliefs on the basis of explicit

reasoning. It is hard to see how we could. Perhaps the majority of our beliefs are not based on arguments at all, but we nonetheless re-gard them as rationally defensible. I believe that I am sitting at mycomputer writing, that there is a coffee cup on my desk, that mydaughter is asleep downstairs, that I have just been thinking aboutPascal’s wager (see Part Four), that I am currently writing aboutfaith and reason, and so on. All of these seem to be reasonable beliefs – I don’t feel I am acting irrationally in holding them – but I wouldfind it difficult to produce arguments in support of their truth. In-

deed I would feel it odd if someone asked me to do so.There are a number of classes of beliefs of which this is true, a

number of cases in which we believe  p in the absence of any argu-ments that (directly) support the truth of  p. The two that are of par-ticular importance in the philosophy of religion are what are called“basic beliefs” and “beliefs held on testimony.” The second of these – beliefs held on testimony – is relatively unproblematic. Does e=mc2 

represent the rate at which matter can be transformed into energy? Ibelieve so, although I would have not the faintest idea how to demon-strate its truth. I have it on good authority that it is true. I cannotgive you the reasoning involved, but I believe the reasoning could begiven by someone else, who knows more physics than I do.

There is, of course, a sense in which I do believe this on the basisof evidence. I have reasons to believe the trustworthiness of thesources from which I have gained the information. It is true that inthe formation of such beliefs, such reasons may have little or no role


Page 31: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 31/205

Varieties of Belief

to play. We seem to have a built-in disposition to believe what others

tell us, unless we have some reason not to. But if asked to justify thisbelief, we could, perhaps, give reasons why we should believe whatwe have been told. The important thing to note is that the relevantreasons do not directly support the truth of the proposition believed;they do so only indirectly, by supporting belief in the trustworthinessof our source of information.

The really controversial category of beliefs held in the absence of arguments is that of basic beliefs. This has become a hot topic withinthe philosophy of religion in recent years, as a result of the work of 

 Alvin Plantinga (to which I shall return in Part Three). But it’s alsoa hot topic within epistemology in general. (Epistemology is thephilosophical discipline that discusses the antrue of knowledge,which in Greek is epistēmē .) The paradigmatic case of a basic belief is a perceptual belief: one that arises spontaneously from sense per-ception in ordinary circumstances. (I believe there is a water bottleon the desk because it appears to me that there is a water bottle onthe desk, and I have no reason to think this appearance is an illu-sion.) Such beliefs are basic beliefs not in the sense that they lack ev-

idential support – the evidence that supports my water-bottle belief is the appearance of the water bottle – but in the sense that they arenot the result of inferences; they do not arise as a result of a processof reasoning that begins with other beliefs. So a better name forthem might be “non-inferential” beliefs.

Many philosophers argue that such beliefs have prima facie justi-fication, that is to say, they are justified “at first sight.” (Here’s aparallel legal usage of this Latin phrase: the police will initiate aprosecution against someone if there is  prima facie evidence of his

guilt.) Prima facie  justification means it is rational for us to main-tain such beliefs if we have no reason not to do so, i.e. no evidence tobelieve they are false. I shall be coming back to this question of “ba-sic beliefs” later, when I discuss what I shall call “evidentialism.”The question I shall be addressing is this, “Could religious beliefs bebeliefs of this kind, that is to say, beliefs which both arise sponta-neously and which enjoy prima facie justification?” If so, there would


Page 32: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 32/205

Varieties of Belief

be no need for the believer to produce arguments in support of her

faith. Rather, it would be up to the atheist to show that the belief isdefeated. (This would reverse what the next chapter will describe as“the presumption of atheism.”) I shall be discussing this possibilityin Part Three.


Page 33: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 33/205

Chapter Four 

The Presumption of Atheism

I’m a great believer in putting forward testable claims. So let memake a prediction, which you can test for yourselves. If you tell peo-ple that you are studying arguments for and against the existence of God, eventually someone will say to you with an air of triumph, “Butyou can’t prove God doesn’t exist!” This raises two questions. Firstly,is this true? Secondly, does the atheist need to prove that God does-

n’t exist?

4.1 Proving God Does Not Exist

Could one prove that God doesn’t exist? In principle, yes. If it couldbe shown that theism involves a logical contradiction – if it is eitherinternally incoherent or there exists some indisputable fact aboutthe world which is simply inconsistent with belief in God – then onecould prove that God does not exist. The two most promising argu-ments here have to do with the attributes of God and the problem of 

evil. (I shall be examining the latter in Part Two.) If we could showthat theism did involve some logical contradiction, then we wouldhave demonstrated, not so much that God does not exist, but that abeing answering to this description could not exist.

But if theism is not logically flawed in one of these two ways, thenits falsity cannot be proven, in the sense of demonstrated beyonddoubt. Qualified claims regarding the non-existence of things, suchas “There are no keys in my pocket,” can be proven (“Let’s have alook and see”). But you cannot prove an unqualified non-existence

claim, such as “There are no fairies.” After all, even if we have not to

date found any objects answering to the description of a fairy, andeven if their existence is unlikely, we may be wrong. Indisputable ev-idence of the existence of fairies may crop up tomorrow. As Karl Pop-per writes, “we cannot search the whole world to establish thatsomething does not exist, has never existed, and never will exist”(Popper 2002: §15 [48]; see also §22 [70–71].)


Page 34: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 34/205

The Presumption of Atheism

However, to say that we cannot prove an unqualified non-exis-

tence claim is not to say very much. To say that we cannot be certainthere are no fairies is not to say that it would be reasonable to be-lieve in fairies. We have no evidence that there are fairies and quitea bit of evidence which suggests they are unlikely to exist. (Their ex-istence is improbable, given everything else we know about theworld.) Faced with such evidence we would not be acting rationally if we did believe in fairies. Our belief in fairies would not be a  justified

belief or (to use my preferred formulation) we would not be entitled

to act on this belief. The fact that we cannot demonstrate the falsity

of a proposition doesn’t mean we are entitled to believe it is true. Soeven if we cannot demonstrate the non-existence of God, this doesnot hand any kind of victory to the theist.

4.2 The Burden of Proof

The more serious question raised by the objection with which I beganis: On whom does the burden of proof fall? As I have just noted, athe-ists are unlikely to say they have proven (in the sense of demonstrat-ed beyond doubt) that God does not exist. But they may claim we

have insufficient evidence to believe he does, and some reason to be-lieve he does not. Theists, for their part, sometimes claim to haveproven that God exists, but more often they, too, are making an evi-dential claim. Given the evidence for and against, they argue, it ismore likely than not that God exists. (It is worth noting that boththeist and atheist in this discussion accept David Hume’s principlethat the wise man should “proportion his belief to the evidence”[Hume 1902 §88 (110)].)

Who’s right? That’s the question that lies at the heart of our dis-cussion. In answering the question, it is important to decide whomust bear the burden of proof. Or, to put this a little differently, it isimportant to decide what our default position is. Is it theism or athe-ism? If the atheist cannot make a convincing case against theism,does theism win by default? Or if the theist cannot make a convinc-ing case in favour of theism, does atheism win by default?


Page 35: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 35/205

The Presumption of Atheism

 A widely held view, which until recently was without serious chal-

lenge, is that atheism would win by default. The atheist does notneed to shoulder the burden of proving that God does not exist, orthat we have insufficient reason to believe in his existence. This posi-tion is often referred to as the presumption of atheism, although theterm “atheism” is perhaps misleading in this context. The parallelhere is to the legal presumption of innocence. Just as a person ac-cused of a crime should be considered innocent until proven guilty, soatheism (in the sense of non-belief in God) should be our default posi-tion, one which we hold until we have been given reason to believe. If 

the presumption of atheism holds, then the proper response to “Youcan’t prove that God doesn’t exist” is: “But I don’t need to.”

This may seem like hair-splitting, but the issue is not a trivialone. The existence of apparently gratuitous evils, for instance, offersa powerful argument against theism. (See Part Two.) But let’s sup-pose that the theist could demonstrate that the existence of suchevils is not such a problem after all. The existence of such evils is notonly logically compatible with the existence of God; it does not evenconstitute decisive evidence against belief in his existence. If this

were demonstrated, where would this leave the debate? If the pre-sumption of atheism holds, the answer is: No further ahead. If weare interested in the truth of the proposition “God exists,” then thefailure of the atheist’s arguments would not hand a victory to thetheist. The default position would still be atheism (i.e. non-belief).That default position would hold until the theist had produced somepositive evidence in support of her belief.

4.3 Merely a Presumption

Let’s examine the presumption of atheism more closely, because atfirst sight it might seem unfair. It might appear to put the atheist ina “Heads I win, tails you lose” position. So what arguments can beproduced in its defence? The philosopher who has most clearly artic-ulated and defended the presumption of atheism is Antony Flew. Inwhat follows I shall be taking his arguments as a starting point, butat times I shall go beyond them.


Page 36: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 36/205

The Presumption of Atheism

Let’s begin by what Flew means by “atheism.” Just what is the po-

sition which, he argues, should win by default? Flew insists that inthis context “atheism” should not be taken to entail the denial of God’s existence. An atheist is merely someone who does not believein God. If one wants to distinguish the two positions, Flew writes,then the former position, entailing a denial of God’s existence, mightbe called  positive atheism, while the latter position might be callednegative atheism. It is negative atheism that should represent ourdefault position. How does this differ from agnosticism? It differsfrom agnosticism, says Flew, because the agnostic has already con-

ceded, implicitly at least, that it is possible that God could exist. Butthe atheist does not concede even that much. He does not concede,for instance, that the idea of God is coherent.

This might seem a slippery distinction, but I think it is a real one.The negative atheist does not say, “I believe there [probably] is noGod.” What he says is, “I do not believe there is a God.” In the case of negative atheism, what is being denied is the belief; the case of posi-tive atheism, it is the existence of God. It may be useful to set thisout with some simplified logical symbols. If  B is “I believe” and G is

“there is a God,” and ¬ is a negation sign (“it is not the case that”),then what the negative atheist says is

¬ B(G),

while the positive atheist says is


Since the negative atheist assumes neither the existence of Godnor the coherence of the concept, to argue for the presumption of (negative) atheism is to place a double burden on the shoulders of the theist. The theist must “first, … introduce and defend his pro-posed concept of God; and, second, ... provide sufficient reason for be-lieving that this concept of his does in fact have an application”(Flew 1972: 31). It is important to realise that the presumption of atheism is a  procedural matter, to do with the burden of proof. Itmay be that theists can meet this two-fold challenge and prove (insome broad sense of that term) what needs to be proven. But this


Page 37: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 37/205

The Presumption of Atheism

“does not even begin to show that the burden was wrongly placed

upon their shoulders” (Flew 1997: 411).We should note that Flew is here using the term “atheism” very

broadly. On this usage, agnosticism would be a species of atheism.This is not how the term “atheism” is normally used. Most of thosewho would describe themselves as atheists are positive atheists.They believe that we have good reasons not to believe in God. Theyare not merely “not theists.” Those theists who object to Flew’s prin-ciple may be doing so because they see it as a stronger principle thanit actually is. In fact, Flew’s principle offers as a default position a

particularly strong form of agnosticism. But for better or worse, the“presumption of atheism” is now a phrase in general use. So I shalluse it myself, but you should keep in mind how broadly “atheism” ishere being defined.

4.4 Why Presume Atheism?

What arguments can be offered in defence of the presumption of atheism? A first approach might be to argue that since the theist’sbeliefs include an extra belief , one that the atheist does not hold, it is

up to the theist to produce some reason why this belief should beadopted. But to say this is to do little more than restate the principlefor which we are arguing. So let’s try to make the argument a littleclearer. Flew expresses this view by way of a legal principle. In Latinit reads: ei incumbit probatio qui dicit, non qui negat: the onus of proof lies on the one who affirms, not on the one who denies (Flew1972: 35). There are two ways in which this principle could be inter-preted. The first is to regard it as a principle governing discussionsin general. The problem here is that in a debate even contrary viewscould both be expressed as affirmations. Flew’s example is, “Thatthis house affirms the existence of God” and “That this house takesits stand for positive atheism” (Flew 1972: 36). In this case, both par-ties are affirming something, so the legal principle does not apply.

 A second way to regard the principle is to regard it as a prudentialprinciple regulating the work of explanation. If one takes this line,the presumption of atheism is merely another expression of what is


Page 38: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 38/205

The Presumption of Atheism

often called “Ockham’s razor” (after William of Ockham [ca.1285– 

1347]), namely entia non sunt multiplicanda praeter necessitatem:entities should not be multiplied beyond necessity. This seems a saneand sensible principle. It is not without its detractors, but to defendit would take me too far afield. If one accepts that this is a soundprinciple, then one of its implications is the presumption of atheism.It is the theist who is introducing an extra entity into our explana-tions of the world. Ockham’s razor suggests that she be called uponto show us why this is necessary. This assumes, of course, that the-ism is being defended as the best explanation of some fact about the

world. That this is not the only way in which theism is defended willbecome clear when I discuss religious experience.

However, Flew’s defence of the presumption of atheism is a differ-ent one. Like the legal presumption of innocence, he argues, the pre-sumption of atheism is a policy which we adopt. When we adopt apolicy it is normally in pursuit of some goal. In the case of the legalpresumption of innocence, the goal is to ensure that innocent peopleare not convicted. To justify the policy, you would need to argue forthe importance of this goal. In the case of the presumption of athe-

ism, the goal is that of acquiring knowledge. But at least according toour traditional idea of knowledge, knowledge differs from mere belief (even true belief) in that the person having knowledge has grounds

for her belief. If she believes some proposition  p that happens to becorrect, but she has no grounds to believe it is true, we would notnormally say that she knows  p. So if our aim is to achieve knowledge,the onus must lie on the person who affirms a belief to producegrounds for its truth. (Remember that the atheist is here affirmingnothing. “Atheism” in this context is nothing more than a lack of be-



Page 39: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 39/205

Chapter Five

Reports of Miracles

In Parts One and Two of this course, I shall be assuming an eviden-tialist view of faith. I shall be assuming that a rational belief in Godrequires some kind of argument or evidence. Not all philosophersagree, and I’ll look at their arguments later. But since the majorityof philosophers – theists and atheists alike – have held to this evi-dentialist view, it seems a reasonable starting point. So what argu-

ments are there that might support belief in God? What reasonsmight there be for believing that God exists? One reason would sure-ly be that there exist reliable reports of miracles. If miracles occur,and if miracles are violations of the laws of nature, then there mustexist some agent capable of overriding those laws (a “supernatural”agent). Such an agent would have at least some of the characteristicsof the God of Judaism, Christianity, and Islam.

Reports of miracles play an important role in the history of reli-gion. Traditionally, they serve to bolster support for the claims of 

particular religious traditions. Thomas Aquinas, for instance, re-gards the fulfilment of prophecy, the worldwide spread of the faith,and the occurrence of miracles as the three most powerful argumentsin support of the Christian faith (SCG 1.6, 1.9). In later RomanCatholicism miracles play so important a role that the first VaticanCouncil (1870) made belief in them an article of faith. As a Catholicyou could be excommunicated if you deny (a) that miracles occur, (b)that they can be known for certain, or (c) that they serve to prove“the divine origin of the Christian religion” (DS 3034; Flew 1967:

348). Even the Protestant Reformer, John Calvin, despite his suspi-cion of Roman Catholic miracles, regards the miracles related inScripture as among the “external evidences” of biblical authority(ICR 1.8.5–6).


Page 40: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 40/205

Reports of Miracles

5.1 The Idea of a Miracle

The classic modern definition of a miracle is that offered by DavidHume. A miracle, Hume writes, is “a transgression of a law of nature

by a particular volition of the Deity, or by the interposition of some

invisible agent” (Hume 1902 §90 [115n.1]). More concisely, a miracleis “a violation of a law of nature”(Hume 1902 §90 [114]). But if youwish to avoid the idea that a miracle “violates” a “law,” I am happy todo so. You might prefer the definition of a miracle offered by a con-temporary Christian apologist, namely “an event which lies outside

the productive capacity of nature” (Craig 1986: 29). Of course, thisraises an epistemic question: How could we know that a particularevent meets this description? That’s not an easy question to answer.But there seems nothing incoherent about the idea of a miracle sodefined. In other words, miracles are not, as philosophers say, “logi-cally impossible.”

The reason one might want to avoid the Humean definition of amiracle – a violation of a law of nature – is that at first sight it doesface some logical difficulties. It certainly does so if one regards the

“laws of nature” as merely descriptive, as universal statements aboutthe way things occur, for it is hard to see what it could mean for suchlaws to be violated. The occurrence of an event that apparently vio-lated a law of nature (in this sense) would merely show that the pur-ported law was not a law after all. The purported miracle would notbe a miracle at all, but merely a counter-instance calling for revisionof the so-called law. However, the difficulty disappears if one thinksof the laws of nature as expressing some kind of necessity (or at least propensity), an association which is “built into” (as it were) the very

nature of things. A miracle is a violation of the causal relationswhich – if nature were left to itself – would govern the course of events, by way of the intervention of a supernatural cause.

But this raises a second objection. There is a widespread viewamong physicists that at the level of quantum mechanics causal rela-tions are not deterministic. We cannot say what will happen on alloccasions; all we can say is what will happen on a certain percentage


Page 41: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 41/205

Reports of Miracles

of occasions. It is hard to make sense of the idea of a violation of a

 probabilistic law, which specifies that on a small percentage of occa-sions a highly improbable outcome will occur. But this in principle

problem seems to disappear in practice. Events may be undeter-mined at the quantum level, but at the level of mid-sized physical ob- jects the classical laws of nature work perfectly well. It may be that,strictly speaking, such laws merely enjoy an extraordinarily high de-gree of probability. But the degree of probability they enjoy is so highthat we can, in practice, regard it as a certainty. Other than on thealleged occasions of divine intervention, that is to say, when miracles

are alleged to occur, the laws that miracles would violate are withoutknown exception. So even in the face of this objection the Humeandefinition of a miracle may well be defensible.

5.2 Are Miracle Reports Credible?

In what follows, then, I shall be assuming that one can give a coher-ent definition of a miracle. If so, then the occurrence of events thatcould only be understood as miracles would lend support to the ideathat there exists a “supernatural” being, one who is capable of inter-

vening in the workings of nature. It does not follow that this beingwould have the other attributes of God – he may, for instance, bemerely very powerful rather than omnipotent – but a reliable reportof a miracle would be at least the beginning of an argument for the-ism.

But has there ever been a reliable report of a miracle? Could weever be justified in believing that a reported miracle occurred? Theclassic response to this question is that offered by David Hume, whoclearly thought that his argument settled the issue, at least for thosecapable of appreciating it. As he wrote, “I flatter myself, that I havediscovered an argument ... which, if just, will, with the wise andlearned, be an everlasting check to all kinds of superstitious delu-sion, and consequently, will be useful as long as the world endures”(Hume 1902 §86 [110]). For many thinkers, whether religiously-in-clined or not, Hume’s discussion has indeed been decisive. But from


Page 42: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 42/205

Reports of Miracles

the moment of publication Hume’s conclusions have also been hotly

contested. So what is Hume’s argument? Has it withstood criticism?

5.2.1 The Strengths of Hume’s Argument

Hume was a historian as well as a philosopher. Today he is bestknown for his philosophical writings. But in his own lifetime he wasbest known for his six volume History of England (1754–62). This isnot an irrelevant fact, because Hume approaches the question of mir-acles with the instincts of an historian. (See Flew 1967: 350–53.)What does the historian do when faced with documents which pur-

port to tell of past events? If he is a good historian, he doesn’t just“cut and paste” material from his sources. Rather, he acts in themanner of the detective interviewing witnesses to a crime. (The par-ticular analogy is that of R. G. Collingwood [1961: 266–8]; compareHume’s comments about the judge [1902 §95 [122].) Witnesses arenotoriously unreliable. Similarly, the historian knows that those whowrote his documents may be unreliable. So the historian has an obli-gation to weigh the evidence.

How does the historian go about weighing the evidence? He has a

report that claims that a certain event occurred. He must take thisreport seriously. But he must also take two other factors into ac-count. The first is the  prior probability such an event should occur,that is to say, the probability that it should occur considered inde-pendently of the reports that it did occur. The basis on which hewould normally make such judgements of probability is his knowl-edge of what has occurred in similar circumstances on other occa-sions. The knowledge of what has occurred in similar circumstanceswill also inform the historian’s second judgement, which has to do

with the reliability of the reports which testify to the event. In theseparticular circumstances, how likely is it that his informants are ly-ing or mistaken? If, for instance, they contradict each other, or if they have a vested interest in the truth of what they relate, the prob-ability that they are deceived or deceiving may be high.

These relatively straightforward reflections give us the basic ele-ments of Hume’s argument. Part One of his essay deals with the first


Page 43: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 43/205

Reports of Miracles

 judgement, that of the prior probability that a miracle occurred. Part

Two deals with the second judgement, that of the reliability of thetestimony to the miracle. Let’s now see if we can reconstruct his ar-gument in a defensible form.

The argument of Part One of Hume’s essay begins with the veryidea of a miracle. Its starting-point is the observation that “nothingis esteemed a miracle, if it ever happen in the common course of events” (Hume 1902 §90 [115]). This seems uncontroversial. If a mir-acle is an exception to the regular workings of the natural world,then it must be something unusual. It must not be an event that reg-

ularly occurs in similar circumstances. If it were, it would be practi-cally “indistinguishable from a law of nature” (Gaskin 1988: 160); itwould not be recognizable as a miracle. Let’s try to spell out whatthis means. Given any regularly-occurring set of circumstances (let’scall it C), there is a regular outcome (let’s call it R), which we knowabout on the basis of past instances. (For instance, our regular expe-rience is that when someone dies and is buried, he remains dead.) If the same circumstances were to give rise to a miraculous outcome(let’s call it M, such as a person rising from the dead), then to be seen

as a miracle this would have to occur only very rarely. The propor-tion of miraculous to non-miraculous outcomes must be very small. As J. C. A. Gaskin notes, the more regularly an alleged miracle oc-curred, the more likely we would be to seek a natural explanation of the event in question.

What follows? If we ask ourselves, before examining the reliabilityof our reports, which outcome, R or M, was the more likely, the an-swer can only be R. We would have to assess the prior probability of M as very low. (I am talking here of epistemic probability, that is to

say, vis-à-vis our background knowledge.) Can we assign a figure tothis probability? It has recently been alleged that Hume is makingan elementary error here: he thinks that if our experience of R hasbeen without exception, then the prior probability of M is zero (Ear-man 2000: 22–32). If this charge is justified – I shall come back to itin a moment – it would mean (ironically!) that Hume as placing toomuch confidence in the power of inductive reasoning. (The irony here


Page 44: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 44/205

Reports of Miracles

is that Hume elsewhere claims to be a sceptic about inductive rea-

soning.) This would explain why Hume sometimes seems to say thatmiracles are impossible (that is to say, have a probability of zero).But for the moment, let’s take a more charitable interpretation of Hume’s position. Let’s assume that uniform past experience againstan alleged miraculous event does not show it to be impossible. Itmerely shows its prior probability to be very low. This is all Hume’sargument requires.

But low prior probability is not enough. If we did have a reportwhose reliability could not reasonably be doubted, then the fact that

the reported event had a low prior probability would not justify itsrejection. It would mean only that the evidence that it did occurwould have to be very reliable. So Part Two of Hume’s essay dealswith the reliability of testimony to the occurrence of a miracle. Humeargues that it is all but impossible that the evidence in support of amiracle claim could outweigh its prior improbability. He offers fourreasons. Firstly, there has never been a miracle report that clearlyexcludes the possibility that those who framed it were deceiving ordeceived. Secondly, human beings have an appetite for wonders

which, when combined with “the spirit of religion,” will almost in-evitably lead to fabrications. Thirdly, miracle reports are chieflyfound “among ignorant and barbarous nations” and they diminish innumber as one approaches the present. Fourthly, the miracle reportsof the different religions, however well established they might ap-pear to be, undermine each other’s authority. All these claims arecontestable, to a greater or lesser degree, but it is not vital to Hume’scase that each argument be convincing. It would be a rash critic whowould not recognise the possibility of deception or deceit when it

comes to miracle reports. And even a small chance that a miracle re-port may be unreliable will be enough to discredit it.

Why is this? This question brings us to the heart of Hume’s argu-ment. As wise people, who proportion their belief to the evidence,when would we be justified in accepting a report of a miracle? Only

when it would be more improbable that the report be false than that


Page 45: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 45/205

Reports of Miracles

the miracle occurred. As Hume writes, with clear allusion to the res-

urrection of Jesus,

when anyone tells me, that he saw a dead man restored to life, Iimmediately consider with myself, whether it be more probable,that this person should either deceive or be deceived, or that thefact, which he relates, should really have happened. ... If the false-hood of his testimony would be more miraculous, than the eventwhich he relates; then, and not till then, can he pretend to com-mand by belief or opinion. (Hume 1902 §91 [116])

The force of the argument will become clear if we assign some fig-ures to the probabilities involved. Any such figures will be admitted-ly arbitrary, but I will take mine from one of Hume’s more intemper-ate recent critics, John Earman. Let’s say that on past experiencethe prior probability of the reported miracle occurring is one in a mil-lion. Let’s say, too, that we have a highly reliable witness to the al-leged miracle, one who has been reliable on ten occasions for everyone that he has been mistaken. So the prior probability of his beingmistaken on this occasion in 1/11. If weigh up the two probabilities,we find that the probability of the reported event occurring is farlower than that of the witness being mistaken. So we should rejectthe miracle report. Unless we were faced with testimony of truly un-precedented reliability – testimony whose falsehood would be mirac-ulous, as Hume puts it – we may assume that miracle reports arefalse.

5.2.2 Evaluating Hume’s Argument

The key question here is: Are there any circumstances in which a fol-lower of Hume might be forced to concede the possibility of a

miracle? Towards the end of his essay, Hume suggests there are not.Even if miracles are not logically impossible, there are no circum-stances in which we could rationally believe a miracle report. If wereceived a report of an event that was truly without analogy (that isto say, truly a miracle), the event would be ipso facto so improbablethat we may assume the report to be false. The only extraordinaryevents we should accept, on the basis of apparently undeniable testi-


Page 46: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 46/205

Reports of Miracles

mony, are events that are analogous to events of which we already

have knowledge. (This seems to be the point of Hume’s contrast be-tween the eight-day darkness and the alleged resurrection of QueenElizabeth. Firm testimony to the first may be accepted; even firmtestimony to the second may not.) But of course if the reportedevents are analogous to events of which we already have knowledge,then ipso facto they are not miracles. This is surely a defensibleview. Its major weakness is that it is very difficult to draw a line be-tween these two sorts of events. Just when is a reported event sowithout analogy that we should reject it? The notion of analogy is

here very vague. But of course this is a problem facing the theist aswell, the problem of deciding just which extraordinary events are

miracles, which events are (as we might say) “outside the productivecapacity of nature.” (For a discussion of how this might be done, seeDietl 1968.) The Problem of Probability

There is, however, a problem with Hume’s argument, which has todo with how one thinks about probability. Hume apparently assesses

the prior probability of an event by counting the number of previousknown instances of that event. But this seems wrong-headed. It cer-tainly flies in the face of scientific practice. As John Earman writes,

scientists not uncommonly spend many hours and many dollarssearching for events of a type that past experience tells us havenever occurred (e.g. proton decay). Such practice is hard to under-stand if the probability of such an event is flatly zero, and theprobability of the putative law asserting the non-occurrence of thistype of event is unity. (Earman 1993: 297)

 Yet Earman’s comment also indicates a way forward. On what ba-sis do scientists search for previously unobserved events? Presum-ably because they have a well-confirmed theory which predicts thatthis event will occur at some time, in these circumstances. The likeli-hood of the event occurring is dependent on the probability of thetheory that predicts it. Applied to reports of miracles, and adoptingthe probabilistic framework of Hume’s argument, this suggests that


Page 47: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 47/205

Reports of Miracles

the prior probability of a miracle is dependent on the prior probabili-

ty of the theistic hypothesis. To put this another way, it would bereasonable to accept a hypothesis that involved a miracle only if wealready had reason to believe that there was a miracle-working God.

 You might argue that even on these grounds, Hume’s argumenthas some force. For as we have seen, what interested him was theuse of miracle reports to establish the claims of some particular(Hume 1902 §99 [127]). If you are using miracle reports to provethere is a God, you cannot use the existence of God as a reason to be-lieve the miracle report. That would be begging the question. But

Hume’s way of calculating probabilities has introduced some confu-sion into his argument. Multiple, Independent Witnesses

 Another factor that Hume did not take into account is the potentialpower of multiple, independent, and individually reliable witnesses.The mathematician Charles Babbage, in the ninth  Bridgewater

Treatise (1838), developed this idea. Babbage accepts the central ar-gument of Hume’s essay, which he argues “never has, and never will

be refuted” (cited in Earman 2000: 203). But he then turns it againsthim. Babbage concedes that we can assign a very low prior probabili-ty to the allegedly miraculous event. He suggests, for instance, thatthe odds against a dead man being raised to life could be set at200,000,000,000 to 1. So we would need very strong testimony to out-weigh it. Babbage then goes on to calculate the probability that sev-eral independent, reliable witnesses will give the same false testimo-ny. He estimates, for instance, that if you have two independent wit-nesses, each of whom is reliable 99 times out of 100, the chance of 

their giving the same false testimony is 1/10,000. As the number of independent witnesses grows, this figure increases very quickly. It isnot too long before we reach the point where it is less likely thattheir testimony is false than that the event occurred. Babbage calcu-lates that the odds against just six such independent witnesses giv-ing false testimony to be 1,000,000,000,000 to 1. In the case of the al-leged resurrection, this is five times as great as the improbability of the event which they relate. In these circumstances, a wise man,


Page 48: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 48/205

Reports of Miracles

who proportions his belief to the evidence, will accept the miracle re-

port.The follower of Hume, it seems to me, has little choice but to ac-

cept this argument. But he could still argue that in practice this con-dition is never fulfilled. (This, of course, is the point of Part Two of Hume’s essay.) It is worth noting here how cautious we are in other,non-religious contexts. We are reluctant to believe that multiple wit-nesses to an alleged, highly improbable event are both independentand reliable. If you don’t believe me, then consider the analogouscase of alien abduction reports. We have many instances of such re-

ports. But does that lead you to take them seriously? If not, why not? You will probably deny that the required conditions apply. The ap-parently independent witnesses are not in fact independent (suchstories are now part of our general culture) and what they allege tohave experienced is such that it is unlikely to have occurred. But thiswould need to be argued for each case of an apparently miraculousevent, if there existed many apparently independent witnesses. If this is correct, then Hume’s argument does not permit a a priori dis-missal of all miracle reports, but it does give us some important cri-

teria by which we can assess them.


Page 49: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 49/205

Page 50: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 50/205

Ontological Arguments

which summarises the theology that Anselm had set out in more de-

tail in his Monologion (“a speech made to oneself”) (Anselm 1995:93–94). As the name Proslogion suggests, it was written in the formof a prayer, addressed to God.

We believe that you are something than which nothing greater canbe thought. So can it be that no such nature exists, since “The foolhas said in his heart, ‘There is no God’” [Psalm 14:1; 531]? Butwhen this same fool hears me say “something than which nothinggreater can be thought,” he surely understands what he hears;and what he understands exists in his understanding, even if he

does not understand that it exists [in reality]. For it is one thingfor an object to exist in the understanding and quite another tounderstand that it exists [in reality] ... And surely that than whicha greater cannot be thought cannot exist only in the understand-ing. For if it exists only in the understanding, it can be thought toexist in reality as well, which is greater. So if that than which agreater cannot be thought exists only in the understanding, thenthat than which a greater cannot be thought is that than which agreater can be thought. But that is impossible. Therefore, there is

no doubt that something than which a greater cannot be thoughtexists both in the understanding and in reality. (Anselm 1995:100)

The argument may be set out systematically in a more conciseform (here adapted from Le Poidevin 1996: 19–20):

(1) God exists either in our minds alone, or in reality also.

(2) Something that exists in reality is greater than somethingwhich exists in the mind alone.

(3) If God existed in our minds alone, we could think of somethinggreater, namely something that really exists. [From (2).]

(4) We cannot, however, conceive of anything greater than God.[By definition.]

From (1) – (4) it follows that

(5) God does not exist in the mind alone, but in reality also.


Page 51: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 51/205

Ontological Arguments

What can we say about this argument? Well, if its premises were

acceptable, it might appear to be a sound argument. But in fact only(4) is uncontroversial, and that’s because it is simply a matter of defi-nition. Premise (3) depends on (2). So we need to look carefully at (1)and (2). Premise (1), at least as currently worded, depends on theidea that something can be said to “exist” in the mind. But this is amurky idea. We may, perhaps, say that there exist representations of things in the mind. (Contemporary philosophers of mind are perhapsmore ready to concede this than their forebears.) But if we are tomake any sense of this idea, we need to examine the content of such

representations. For if a representation is a mental state, it is an in-

tentional state. It has an object, to which it makes reference. Thatobject may be either fictional or real. So premise (1) of Anselm’s ar-gument may be re-expressed as follows:

(1*) God is either a fictional object or a real object.

But this brings us to premise (2), which may be restated as

(2*) A real object is greater than a fictional object.

But is this true? Let’s take an example. A unicorn is a fictional ob- ject. As such, it has certain attributes. (These are, I suppose, estab-lished by convention, since no such creatures exist. But even fictionalcreatures have certain characteristics.) Each of us has enoughknowledge of what makes a unicorn a unicorn to recognise one whenwe see one. Many of the unicorn’s attributes it shares with a horse,while its distinctive attribute is its single horn. If you wanted to de-scribe a unicorn, you could do so briefly by saying “a horse-like crea-ture with a single horn.” But let’s say a real unicorn was discoveredtomorrow and you wanted to describe this real unicorn. Assumingthat it shares the attributes of the fictional unicorn, what further at-tribute (if any) would it have by virtue of being real? Would you de-scribe the real unicorn as “a horse-like creature with a single hornthat exists”? Or does that sound a little odd? Is “existing” a attribute,a property that you might predicate of something? Is “existing” reallycomparable to “having a single horn”?


Page 52: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 52/205

Ontological Arguments

Criticism of the ontological argument has centred on this ques-

tion: the question of whether existence is a property or (if you prefer)a predicate. Let me look at the classic response to Anselm’s argu-ment first, that which suggests that existence cannot be regarded asa property. I shall then examine a more recent response, one whichconcedes that in certain circumstances it can.

6.1.1 If Existence is Not a Predicate...

The key figure here is Immanuel Kant, who discussed the ontologicalargument in his Critique of Pure Reason (1781, 1787). Kant ap-

proaches this question in a variety of ways. But I shall cut straightto the most persuasive of his arguments. Let’s begin with Kant’sown words.

“Being” is obviously not a real predicate; that is, it is not a conceptof something that could be added to the concept of a thing. It ismerely the positing of the thing, or of certain determinations, asexisting in themselves. Logically, it is merely the copula of a judg-ment. The proposition, “God is omnipotent,” contains two concepts,each of which has its object – God and omnipotence. The small

word “is” adds no new predicate, but only serves to posit the predi-cate in relation to the subject. If, now, we take the subject (God)with all its predicates (among which is omnipotence), and say“God is,” or “There is a God,” we attach no new predicate to theconcept of God, but only posit the subject in itself with all its pred-icates, and indeed posit it as being an object that stands in relationto my concept ... Otherwise stated, the real contains no more thanthe merely possible. A hundred real thalers [a German silver coin]do not contain the least coin more than a hundred possible thalers

... By whatever and by however many predicates we may think athing ..., we do not make the least addition to the thing when wefurther declare that this thing is. (Kant 1929: 504–5 [a598-599;b626-627])

Kant is here making an important point. It is obscurely expressedbecause he has run together two different uses of the word “is”(Mackie 1982: 46), namely the “is” of predication and the “is” of exis-tence. The first is used to attribute some characteristic to something


Page 53: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 53/205

Ontological Arguments

(“the book is boring”), while the second is used to state that some-

thing exists (“there is a boring book”). But it is precisely these twouses that – Kant insists – need to be distinguished. For Kant is argu-ing that one should distinguish the characteristics that somethinghas (“horse-like creature with a single horn”) from the judgementthat this thing exists (“My goodness, there’s a real unicorn!”).

Modern logicians mark the distinction by way of logical notation,by using what is called the “existential quantifier” (∃). So the ordi-nary-language sentence “A unicorn exists,” could be symbolised as(∃x )(x is a unicorn), which may be read “There exists an x such that x 

is a unicorn,” or perhaps (∃x )(x  is horse-like being with a singlehorn). But if this symbolism frightens you, don’t worry. In everydayEnglish we mark this same distinction by using the phrase “There isan X ,” which must be completed with a description (or term indicat-ing a description).

Kant’s immediate target is Descartes’s formulation of the ontologi-cal argument. In his fifth meditation, Descartes argues as follows:

(1) It is of the essence of God to possess all perfections.

(2) If God did not exist, he would not possess all perfections.Therefore

(3) God exists.

 Adapting Descartes’s own illustration, we might say that exis-tence is as much part of the essence of God as having three sides is of the essence of a triangle. The key assumption here is that existenceis indeed a property, indeed a complete-making property (a perfec-tion). So if it is of the essence of God to possess all perfections, he

must also “possess” existence. (If that last phrase sounds odd, it isprobably because we do not in fact think of existence as a predicate.)But although Descartes’s version of the ontological argument wasKant’s target, his argument is equally applicable to Anselm’s formu-lation. For Anselm’s argument also assumes that existence is a prop-erty, indeed a “great-making” property, without which what wemight call “God” would not really be God at all. So it is tempting toward off all forms of the ontological argument by wielding as a talis-


Page 54: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 54/205

Ontological Arguments

man the Kantian phrase “Existence is not a predicate.” (The same

phrase could perhaps ward off other evils, such as talk of Being witha capital B, so that we can commit to the flames books with titlessuch as  Being and Time [see Hume 1902 §131 (165)]. But that’s an-other issue.)

6.1.2 If Existence Can Be a Predicate ...

However, as is often the case in philosophy, the issue as not as sim-ple as it first appears. J. L. Mackie, himself no friend to the ontologi-cal argument, has raised an interesting objection to this Kantian re-

sponse. He argues that the use of the existential quantifier cannot do justice to many of our ordinary-language sentences about particularindividuals. Among his examples are “David Pears exists,” “This treeexists,” “I exist,” and “God exists” (Mackie 1976: 253; 1982: 47). Insuch sentences, he argues, “exist(s)” is being used as a predicate. Sothe doctrine that “existence is not a predicate” is simply untenable.This is a difficult question, which I don’t wish to try to settle. To givethe ontological argument as much credit as possible, let’s assumethat Mackie is correct and that the above criticism fails. Let’s as-

sume that there could exist an individual whose essence includes ex-istence, that is to say, one that has existence as one of its essentialproperties. Following Mackie, let’s call this individual X . What wouldfollow?

Well, it follows that the sentence “The  X does not exist” contra-dicts itself. The subject of the sentence, “the X,” possesses the proper-ty of existing. So to say “the X does not exist” is equivalent to saying“the X which exists does not exist.” This is the point that Anselm andhis followers have picked up. If we grant for the sake of the argu-

ment that existence can be a predicate, this seems to be correct. But,Mackie argues, we do not contradict ourselves if – taking our leadfrom the existential quantifier and its ordinary-language equivalent – we reformulate the sentence as “It is not the case that there is an X ” or “There are no X s” (Mackie 1982: 48). This may be paraphrasedas, “There is no being which has existence as one of its properties.”There is no self-contradiction here. In other words, even if God’s


Page 55: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 55/205

Ontological Arguments

essence did include existence, it would by no means follow that God

must exist. What would follow is the much weaker conclusion that if God existed, then he would exist necessarily. If there did exist such abeing – one who had existence as a property – he would always haveexisted and he could not cease to exist. (This is a weaker sense of “necessary existence” than the ontological argument requires, but itmay be the most it can warrant.)

6.2 Alvin Plantinga’s Modal Ontological Argument

 An apparently more powerful version of the ontological argument

has been offered by a contemporary American philosopher, AlvinPlantinga. Plantinga’s argument involves some of the insights of modal logic, which deals with the notions of necessity and possibility.(Modal logic is so called because originally it dealt with the modes, orways, in which a proposition may be said to be true, either necessari-ly or contingently.) A popular way of approaching this topic is by wayof the idea of  possible worlds. There are various forms of possibilitythat may be invoked here, but the most widely used is that of logical

possibility. Note that this is a much broader category than that of 

physical possibility. It may be physically impossible – in violation of the laws of nature – that pigs should fly. But it is not logically im-possible. There is no contradiction in the idea of a flying pig, as thereis (for example) in the idea of a square circle. A possible world is astate of affairs that is possible in this broad, logical sense: it can bedescribed in a way that is consistent, that avoids self-contradiction.The concept of a possible world is attractive because it makes it easi-er to work with the difficult concept of necessity. It enables us to saythat a necessary truth is one that is true in every possible world.(Does this help? Some philosophers, such as Jonathan Harrisonthink not [Harrison 1999], but this is not a question I can discusshere. I shall use “possible worlds” talk, despite the confusions it cancreate, because Plantinga does so.)


Page 56: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 56/205

Ontological Arguments

6.2.1 St Alvin’s Argument

Plantinga begins by defining a couple of notions (see Plantinga 1975:111). The first is that of “maximal greatness.” A being has “maximalgreatness” only if it has “maximal excellence” in every possibleworld. What is “maximal excellence”? It is the property of enjoying“omniscience, omnipotence, and moral perfection” in every world.Plantinga then employs these notions to construct his modal ontolog-ical argument, which may be set out as follows.

(1) There is a possible world in which maximal greatness is instan-

tiated. [ = There is nothing contradictory about the idea of max-imal greatness.]

(2) Necessarily, a being is maximally great only if it has maximalexcellence in every possible world. [By definition.]

(3) Necessarily, a being has maximal excellence in every possibleworld only if it has omniscience, omnipotence, and moral per-fection in every world. [By definition.]

From (1) – (3) it follows that

(4) Maximal greatness is instantiated in every possible world.

What this means will (I hope) become clear if I add a couple of 


(5) Maximal greatness entails having the qualities of God in everypossible world. [From (2) and (3).]

(6) The actual world is a possible world. [What is actual must bepossible.]

The conclusion follows, namely that(7) God exists in the actual world.

What Plantinga has apparently done here is to deduce the exis-

tence of God-from the mere possibility that God might exist. This is avery neat trick and at first sight he seems to have pulled it off. Afterall, all Plantinga asks of us is that we should concede the mere possi-

bility of maximal greatness, and it might seem unreasonable of the


Page 57: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 57/205

Ontological Arguments

atheist not to do this. (Whether it would be unreasonable is a ques-

tion to which I shall return.)It is true that many atheists have denied the possibility of maxi-

mal excellence, by arguing that the qualities attributed to God – om-niscience, omnipotence, and moral perfection – could not exist withinthe one being. This is an important question (Himma 2001), for if itis right, there is no (logically) possible world in which God exists. (Aclassic modern discussion of the incoherence of the idea of God canbe found in Kenny 1979.) But given the mysteriousness of the notionof God, it is difficult to make a watertight case for the incoherence of 

the divine attributes. As Richard Swinburne notes (1993: 73, 305),the fact that our language about God must be, at least in part, ana-logical, makes it impossible directly to demonstrate the coherence of theism. But by the same token it makes it impossible directly todemonstrate the incoherence of theism. This is, perhaps, the reasonwhy theology is such a slippery beast, why – as Albert Schweitzer

once remarked –  “there is no position so desperate that theology can-

not find a way out of it” (Schweitzer 1981: 116). So let’s give the the-ist the benefit of the doubt and concede that the idea of an omni-

scient, omnipotent, and morally perfect being is not self-contradicto-ry. What, then, what can we say? Has St Alvin done what St Anselmfailed to do?

6.2.2 A No-Maximality Argument

Faced with Plantinga’s argument as with Anselm’s, many philoso-phers feel instinctively that something is wrong. But what is it? Toanswer this question, I want to look at an intriguing counter-argu-ment, one offered by Plantinga himself. At the very end of his discus-

sion of the modal ontological argument, Plantinga concedes that onecould construct a parallel argument, leading to the conclusion thatGod’s existence is impossible. All one needs is to admit the possibilityof some property whose existence is incompatible with that of maxi-mal greatness. The simplest candidate here is the property of  no-

maximality, “the property of being such that there is no maximallygreat being” (Plantinga 1974: 218). There seems to be a possible


Page 58: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 58/205

Ontological Arguments

world in which this property is instantiated. In other words, it is pos-

sible that God might not exist. (The ontological argument, if success-ful, would shut down this possibility, but its success cannot be as-sumed for the sake of constructing the argument!) But if the proper-ty of no-maximality were instantiated, it would of course exclude thepossibility that there exists a being enjoying maximal greatness. Inthis case, God could not exist. One can set out the argument as fol-lows:

(1) No-maximality is possibly exemplified.

(2) If no-maximality is possible, then maximal greatness is not.Therefore

(3) Maximal greatness is impossible.

But there is a still clearer way of expressing Plantinga’s counter-argument. If God is, by definition, a being who exists in every possi-ble world, then if there is one possible world in which he does not ex-ist, then he does not exist in any possible world. In a word, if God’snon-existence is possible, his existence is impossible.

Now this is an important concession. Indeed, as Mackie writes, “itis to Plantinga’s credit that he draws attention to the no-maximalitycounter-argument, for if we failed to consider this his ontological ar-gument would be insidiously attractive” (Mackie 1982: 61). What isPlantinga’s response? He concedes that his ontological argumentfails as a piece of “natural theology,” since it does not begin with apremise that any reasonable person would be forced to accept. Theopponent is free to start with a different premise, or perhaps to re-main agnostic about the possibility of a being enjoying maximal

greatness. But, Plantinga continues, if the theist chooses to beginwith the premise that favours theism, there is nothing “improper,unreasonable, [or] irrational” about this choice. Well, perhaps thereisn’t. One response is that such a choice does rather beg the ques-tion, particularly if it is being made (as Plantinga admits) for thesake of theology. After all, his opponent can offer at least one inde-


Page 59: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 59/205

Ontological Arguments

 pendent reasons for accepting a no-maximality premise, namely that

this is more parsimonious.

6.2.3 Accepting the Initial Premise

In fact, however, a more serious objection can be raised. Plantinga’smodal ontological argument and his parallel, no-maximality argu-ment both begin with premises which seem unobjectionable. Bothpremises, one is tempted to think, could be true. There might exist aGod, and there might not. Fair enough. This doesn’t, it seems, com-mit one to very much. After all, it is true that it might be raining

outside, and it is true that it might not be raining. Since these aremerely modal truths, regarding possibilities, they are not mutuallyexclusive. But the conclusions of Plantinga’s ontological and no-max-imality arguments are certainly incompatible. In the one case, theconclusion is that God exists, in the other that he does not. Both can-not be true. What does this mean? Well, assuming there is no errorin the logic, the premises must also be mutually exclusive. One can-not have two valid arguments beginning with true premises whicharrive at conclusions, only one of which can be true. (Or at least, I

don’t think you can. Are there any logicians out there to correct me?)If this line of reasoning is correct, it follows that only one of the

two initial premises can in fact be true, although we may not be ableto tell which of them it is. We could opt for the premise of the modalontological argument – “there is a possible world in which maximalgreatness is instantiated” – simply on the grounds that we cannotshow it to be false. But if we do, we have to face the fact that thereexists another premise, which also cannot be shown to be false,which leads to the opposite conclusion. At the very least this should

weaken our confidence in the choice we have made, which turns outto be something of a gamble (a wager, perhaps).

There may be other difficulties with the modal ontological argu-ment, but I don’t want to pursue these issues here. The question onwhich I wish to focus is: Why should we accept the first premise of Plantinga’s ontological argument, which is not as innocuous as itmay appear at first sight. To accept this premise is not merely to ac-


Page 60: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 60/205

Ontological Arguments

cept the possibility of an omnipotent, omniscient and morally perfect

being. It is to accept the possibility of an omnipotent, omniscient andmorally perfect being whose existence is necessary. We might be pre-pared to concede that there could exist a being who is omnipotent,omniscient and morally perfect. But why should we make the furtherconceptions that such a being might exist necessarily? It is true thatthere is nothing obviously self-contradictory about this idea. But sowhat? Does logical possibility entail actual (metaphysical) possibili-ty? (See van Inwagen 1998.) In any case, there is nothing obviouslyself-contradictory about the no-maximality premise either. So why

should we prefer one to the other? Indeed, if we take seriously what Ishall later call “the presumption of atheism,”we should reject the ini-tial premise of Plantinga’s ontological argument until we are givensome reason to take it seriously.

In a word, the modal ontological argument is fiendishly clever, butas Plantinga himself concedes it is unlikely to satisfy anyone whodoes not already accept its conclusion. And such a person has noneed of an argument.


Page 61: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 61/205

Chapter Seven

Cosmological Arguments

It was Immanuel Kant who suggested the threefold classification of arguments I am using here: ontological, teleological and cosmologi-cal. The ontological argument, as we have seen, begins from the veryidea of God. The teleological argument begins from some particularfact about the world – Kant called this the “physico-theological” ar-

gument – and suggests that God is required as the explanation of this fact. What I want to examine here is the cosmological argument,that which begins (as Kant wrote) from the “experience of existencein general” (Kant 1929: 500 [a591; b619]).

This threefold categorization is attractive, but in some ways mis-leading. Kant’s three headings are umbrella terms: each covers a va-riety of arguments. There is, for instance, not just one cosmologicalargument. There are a series of arguments that go by that name, putforward by a range of Jewish, Christian, and Muslim philosophers.

What do they have in common? Little more than this: they are all a

 posteriori arguments in support of the contention that the world as a

whole has a cause. (See Craig 1980: 10.) To keep things manageable,I shall offer for analysis just one form of this argument, which is myown invention. But it is the strongest form of the argument that Ican devise and it will serve to illustrate the assumptions which un-derlie any form of cosmological argument.

Before I do so, however, let me note one important qualification.The argument I am about to present, like any cosmological (or teleo-

logical) argument, is not yet an argument for the existence of God. Itis an argument for the existence of a cause of the universe. It takesus only part of the way towards demonstrating the existence of God.It would not show that this cause is a personal being, or even that itis one being (Mascall 1966: 70, 70n.1), let alone that it is a being whois infinite in power, all-knowing, and morally perfect. So an argu-


Page 62: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 62/205

Cosmological Arguments

ment for the existence of God based on the cosmological argument

would need to have two parts.

The first part, depending on which version we pursue, is an argu-ment to establish the existence of a first cause, necessary being, ora being that accounts for the existence of the world. The secondpart is an argument to establish that the being established in thefirst part is God, that is, has the properties associated with thetheistic concept of God. (Rowe 1998: 6)

We should not underestimate the difficulty of this second task, forit suffers from a serious objection, first set out clearly by DavidHume. Hume argues, in what appears to be a variant on “Ockham’srazor,” that in positing a cause of some observable effect we shouldnever ascribe to that cause “any qualities, but what are exactly suffi-cient to produce the effect” (Hume 1902: §105 [136]). Now the uni-verse would appear to be a finite entity, or collection of entities. If weaccept Hume’s principle, it would appear to rule out positing an infi-nite cause to explain this finite effect. (Intriguingly, even Thomas Aquinas seems aware of this difficulty: see his ST 1a, 2.2, obj. 3.) The

theist may be able to evade this objection, by arguing, for instance,that an infinite deity is a “simpler” hypothesis that a finite deity(Swinburne 2004: 97–98). Perhaps. That would require a further dis-cussion. All I can do here is to indicate that Hume’s principle is afurther problem facing the theist.

For the moment, let’s have a look at my sample cosmological argu-ment. If sound, it would not demonstrate the existence of God, but itwould demonstrate the existence of an entity that has at least two of the traditional attributes of God, namely a being who is distinct from

the universe and who is its creator. To this extent it would be a valu-able argument for the theist.

(1) The existence of the universe is a contingent fact.

(2) All contingent facts have a cause.

(3) Nothing can be the cause of itself.



Page 63: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 63/205

Cosmological Arguments

(4) The universe has a cause that is distinct from itself.

This appears to be a valid argument, but is it sound? There arethree assumptions on which it rests, corresponding to its threepremises. The first is that the existence of the universe is a contin-gent fact. The second is that all contingent facts have a cause. Thethird is that nothing can be the cause of itself. Are these assump-tions justified?

7.1 Can anything be the cause of itself?

Let me begin with the last premise, which seems the least problem-atic. It states that nothing can be the cause of itself. At first sight,this seems entirely reasonable. To be responsible for its effect, acause needs to be distinct from the effect. One might even argue thatthis is a conceptual fact about causes. To suggest otherwise is to failto understand what it means for something to be a cause. Of course,it would also rule out the traditional formulation which says of Godthat he is causa sui (cause of himself). But perhaps this is a priceworth paying if it lends support to belief in God’s existence.

One philosopher, Quentin Smith, does argue that for what ap-pears to be an alternative view. In particular, he argues for the viewthat “the reason the universe exists is that it caused itself to exist”(Smith 1999). This appears to entail the unlikely idea that the uni-verse is indeed causa sui (cause of itself). But what Smith is actuallyarguing is slightly different, namely that the existence of each partof the universe, or each state of the universe at a particular time, canbe explained by reference to other parts or temporal states of theuniverse. Given the possibility of instantaneous causation – a cause

that occurs at the same time as its effect – it follows, Smith argues,that we can explain the beginning of the universe without referenceto anything other than the universe itself.

I’m not sure what to make of this argument. It seems to run upagainst the common-sense objection first put forward by Abu NasrMuhammad al-Farabi (AD 870–950), who argued that “a series of contingent beings which would produce one another cannot proceedto infinity or move in a circle” (Craig 1980: 83). But then perhaps


Page 64: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 64/205

Cosmological Arguments

common sense is not a good guide here. In any case, Smith’s argu-

ment has force only against those theists who invoke “big bang” cos-mology, which suggests that the universe had a beginning. For if oneaccepts that everything that begins to exist has a cause of its exis-tence, this might appear to lend support to the idea of a creator God.It doesn’t seem to have any force against that form of cosmologicalargument which suggests that the very existence of a universe re-quires explanation.

7.2 Is the existence of the universe a contingent fact?

Let’s turn to the second assumption. Is the existence of the universea contingent fact? To say that it is a contingent fact is to say that theuniverse might not have existed. But how could we know that to betrue? Defenders of theism often point out that the idea of the non-ex-istence of the universe does not lead to a contradiction. (It is not likethe idea of a square circle.) The non-existence of the universe istherefore logically possible. Since its non-existence is conceivable,they argue, its existence requires explanation. But is that last stepvalid? What follows from the statement that some state of affairs is

“logically possible”? Very little, I would suggest. After all, what islogically possible may be actually impossible, in the sense that itcould never occur. (A flying pig is not logically impossible, but it doesseem impossible on other grounds.) So while it may be true that wecan speak of the non-existence of the universe without contradictingourselves, it does not follow that the universe, in fact, might not haveexisted.

However, there may be other grounds for believing that the exis-tence of the universe is a contingent fact. You might, for instance, bepersuaded by those cosmological theories that suggest the universehad beginning, some 15 billion years ago. (Of course, if the initialevent, the “Big Bang,” was the beginning of time, then the universedid not have a beginning in time; the beginning of the universe wasthe beginning of time. But it was still a beginning.)


Page 65: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 65/205

Cosmological Arguments

7.3 Do all contingent facts have a cause?

So let’s assume, for the sake of the argument, that the existence of the universe is a contingent fact. Is this a fact of which we could ex-pect to have an explanation? Defenders of the modal cosmological ar-gument assume that any contingent fact does (in principle) have anexplanation. But to assume this is to assume some version of whatLeibniz described as the “principle of sufficient reason” (PSR).

7.4.1 The Principle of Sufficient Reason

In Leibniz’s classic formulation, the principle of sufficient reasonreads as follows.

Our reasonings are based on two great principles, that of contra-

diction, in virtue of which we judge that which involves a contra-diction to be false, and that which is opposed or contradictory tothe false to be true.

 And that of sufficient reason, by virtue of which we consider that

we can find no true or existent fact, no true assertion, without

there being a sufficient reason why it is thus and not otherwise,

although most of the time these reasons cannot be known to us.(Leibniz 1995 §31–32 [217]).

It is important to note that all forms of modal cosmological argu-ment require the PSR. Without it we could rest content with the ex-istence of the universe as a brute fact, that is to say, a contingentfact that has no explanation. After all, the cosmological argument in-volves an odd kind of explanatory claim, one that has to do not justwith particular beings and events but with the universe in its entire-ty. Only some form of the PSR would justify pushing the demand forexplanation to this point. Of course, if one posits the existence of Godas a result, the same principle can then be applied to him. Theremust also be a reason why God exists. What is it? The traditional an-swer (as found, for instance, in Aquinas) is that it is of the very na-ture of God to exist. His essence includes his existence. But thismerely takes us back to the ontological argument and its fallacies. (Ishall come back to this point shortly.)


Page 66: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 66/205

Cosmological Arguments

Now in one respect Leibniz is surely right. Something resembling

the PSR underlies much of our reasoning, both everyday and scien-tific. Faced with a unexplained phenomenon, we do not first ask,“Does it have an explanation?” We assume it has an explanation andstrive to discover what it is. If we come across events that apparentlylack a sufficient reason – perhaps free actions, conceived in a liber-tarian manner, or the behaviour of certain kinds of subatomic parti-cles – we often feel there is something unsatisfactory about this. (Al-bert Einstein seems to have felt this way about the notion of indeter-minacy in quantum mechanics.) But in the face of apparent coun-

terexamples, few contemporary philosophers would defend the PSRin the strong form in which it is expressed by Leibniz. Some rejectthe principle altogether. They regard it as merely a pious hope (moretechnically, a “regulative principle”), which guides our enquiry, butinsist we have no reason to expect it will turn out to be true. Othersprefer to defend a weaker form of the PSR.

 Various weaker forms are on offer. One is the “causal principle”articulated and defended by William Lane Craig, “Everything thatbegins to exist has a cause of its existence” (Craig and Smith 1993:

57). This version is of particular significance in discussions of BigBang cosmology, where a key issue is what one means by a “cause.”William Rowe opts for a broader version of the PSR, carefully word-ed: “For every kind of being such that beings of that kind can becaused to exist ... there must be a sufficient reason for the existenceof each being of that kind and for the general fact that there exist be-ings of that kind” (Rowe 1998: xviii). A similarly cautious principle,but still wider in scope, has recently been offered by AlexanderPruss: “If  p is a true proposition and possibly  p has an explanation,

then p actually has an explanation” (Pruss 2004: 166). But whateverform of the PSR you may opt for, there are two questions that maybe asked about it. How do we know it is true and how broad is itsscope?


Page 67: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 67/205

Cosmological Arguments

7.4.2 Why should we believe it?

The first question has to do with the status of this principle. Is it anecessary truth, known a priori? If so, one would expect it to be ei-ther self-evident (“if a is larger than b and b is larger than c, then a

is larger than c”) or analytic (“brothers are male siblings”)? But itseems to be neither. As Hume argued in his Treatise of Human Na-

ture (1.3.3), we can apparently conceive of a event without a cause. And as Quentin Smith writes,

a case in point is the universe. I can conceive of the universe exist-

ing at a certain time t (say the time of the Big Bang singularity),such that (a) there is no time earlier than t, (b) nothing else existsat t, (c) nothing timelessly exists that causes the universe to beginto exist, and (d) there are no closed timelike curves whereby ‘later’parts of the universe cause the universe to exist at t. (Craig andSmith 1993: 182)

 You may find that you cannot imagine such a thing. You may findits occurrence astonishing, or the theory that it had occurred deeplyimplausible, but it is not evidently impossible. But perhaps the PSR

is what Kant would have called a synthetic a priori truth, which isneither analytic nor known by experience. (If such things exist, anexample might be “no body is at the one time red and green allover.”) But the classic (Kantian) justification for the existence of such truths is that they represent the necessary conditions of any ex-perience. But again this does not seem to be true of the PSR. Wecould experience something coming into existence without a cause.Indeed sometimes it is suggested that we have done so. So if the PSRis true, it would appear to be a contingent truth, known a posteriori,

and justified (if it is justified at all) only to the degree to which it hasbeen borne out by experience.

The answer we give to the first question will determine what an-swer we give to the second, which has to do with the scope of thePSR. If we regard our chosen form of the PSR as a contingent truth,known a posteriori, then two things follow. First, even if the principlehad been justified by our experience to date – and it is not clear that


Page 68: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 68/205

Cosmological Arguments

it has – we may yet come across something that lacks a sufficient

reason. (This is the familiar problem of induction.) Second, the cos-mological argument would involve applying the PSR to a reality (theuniverse) which is very different from those to which it is normallyapplied. We have, for instance, no experience of universes cominginto existence. We know of just one universe. So what a posteriori

grounds could we have for arguing that the universe could not comeinto existence without a cause? From this point of view the conclu-sion of the cosmological argument is open to the sceptical objection sowonderfully expressed by Hume (and echoed, less eloquently, by


Though the chain of arguments which conduct to it were ever sological, there must arise strong suspicion, if not an absolute assur-ance, that it has carried us quite beyond the reach of our faculties,when it leads to conclusions so extraordinary, and so remote fromcommon life and experience. We are got into fairy land, long erewe have reached the last steps of our theory; and there we have noreason to trust our common methods of argument, or to think thatour usual analogies and probabilities have any authority. Our line

is too short to fathom such immense abysses. (Hume 1902 §57[72])

This is not, perhaps, a decisive objection, but it is one that shouldmake us cautious about what we claim to be able to know. (Is humili-ty not itself a religious virtue?)

7.5 Bringing Explanation to an End

 At what conclusion have we arrived? If we accept (a) that the exis-

tence of the universe is a contingent fact, (b) that all contingent factshave a cause, (c) that nothing can be the cause of itself, and (d) thatthe cause of the universe can be shown to have the other attributesof God ... well, under these conditions my cosmological argumentlooks like a pretty good argument for the existence of something re-sembling God. But does it fulfill all its promises? In particular, doesit bring our explanatory quest to an end? Theologians have often as-sumed that it does, for the God whose existence they posit is thought


Page 69: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 69/205

Cosmological Arguments

of as a necessary being, one who couldn’t not exist. Everything else

requires explanation, they argue, but his existence does not. But isthis a tenable idea? Is the idea of a necessary being a coherent one? And does positing the existence of such a being represent the “ulti-mate” explanation which so many people seek?

7.5.1 A Necessary Being?

The idea of a necessary being is a puzzling one. For it seems to entailsomething like Anselm’s idea that existence is a predicate, a perfec-tion. God, on this account, is the one being whose essence includes

existence, indeed whose essence is his existence. But the difficultiesfacing Anselm’s ontological argument recur in this context. Even if we concede that existence can be a predicate, even if we concede thatGod’s essence includes existence, it does not follow that God must ex-ist. This was J. L. Mackie’s criticism of the ontological argument, asdiscussed earlier. Even if it is of the essence of God to exist, it doesnot follow that he does in fact exist.

Why does this matter? After all, the theist who employs the cos-mological argument is not arguing from a definition of God to the ex-

istence of God. That would be the ontological argument and he re- jects it. He is arguing from the existence of the universe to the exis-tence of a being who explains it. Richard Dawkins suggests that toexplain the machinery of life “by invoking a supernatural Designer isto explain precisely nothing.” Why? Because it “leaves unexplainedthe origin of the Designer” (Dawkins 1988: 141). But this is surelywrong. Many of our most successful explanations raise new puzzlesand present us with new questions to be answered. As Peter Liptonremarks, “a drought may explain a poor crop, even if we don’t under-

stand why there was a drought; ... the big bang explains the back-ground radiation, even if the big bang is itself inexplicable” (Lipton1991: 24). So the fact that we apparently cannot explain God’s exis-tence does not necessarily undermine its explanatory power. Butthat’s not the question I’m considering here. My question is: Wouldinvoking the existence of God bring explanation to an end?


Page 70: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 70/205

Cosmological Arguments

It would apparently do so only if God’s existence requires no ex-

planation. And it would require no explanation if the theologianswere right and God were a necessary being, one that could not not

exist. The problem I have been highlighting is that it is difficult tomake sense of this idea. And even if one can make sense of this idea,even if one concedes that there could exist a being whose essence in-cludes existence, it does not follow that this being must exist. So toposit a being whose essence includes existence is not to bring theregress of explanations to an end. The existence of such a beingwould still require explanation. This is the element of truth in the

much-criticised suggestion by Kant that the cosmological argumentdepends on the ontological (Kant 1929: 512 [A611 B639]).

The theistic philosopher Richard Swinburne freely admits that itis logically possible, (i.e., conceivable) that God might not have exist-ed. The existence of God, he argues, is “an ultimate brute fact”(Swinburne 1993: 277), one which might have been other than it is,but which is without explanation. But Swinburne goes further. Heargues not only that the existence of God is without explanation; hesuggests that it cannot have an explanation. If God is a necessary be-

ing in the weaker sense of this term – one who, if he exists at anytime, exists at all times – then no agent can be responsible for his ex-istence (Swinburne 1993: 265). Perhaps this is true, although I amnot entirely convinced by Swinburne’s arguments. (By parity of rea-soning, if the universe existed necessarily in the same weak sense,then it could not have a Creator. Would Swinburne accept this con-clusion?) In any case, the question “Why does God exist?” is not a sil-ly question, as theists sometimes suggest it is.

7.5.2 A “Possible Worlds” AlternativeCould a contemporary theologian not avoid this difficulty? Could henot simply redefine what it means for God to be a necessary being?He does not need to speak of existence as if it were a predicate. Anselm’s ontological argument requires such talk, but Aquinas’s cos-

mological argument does not. He could simply suggest that God is anecessary being in the sense of one that exists in every possible


Page 71: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 71/205

Cosmological Arguments

world. But this strategy would meet another obstacle, one highlight-

ed by Robin Le Poidevin (1996: 40–41). What we are trying to ex-plain is what we are assuming to be a contingent fact, namely the ex-istence of the universe. A contingent fact is one that is the case onlyin some possible worlds. But a necessary being (as here defined) ex-ists in all possible worlds. So merely positing the existence of such abeing fails to pick out what is distinctive about this world (in which auniverse exists) as compared to other possible worlds (in which nouniverse exists). (Remember that “world” here means simply a possi-ble state of affairs.)

We could avoid this difficulty by postulating some contingent factabout this necessary being. Perhaps God freely chose to create theworld so it would produce creatures whose actions would be morallysignificant. We would then have to show that the existence of a beingacting with this intention is the best explanation of the world as it is.(This would be no mean feat, particularly given the existence of evil.)But then our initial problem recurs. Insofar as the intention ascribedto God is a contingent fact (as it must be, if it is to explain), it invitesfurther explanation. It invites the question, “Why did God choose to

act in this way?” The regress of explanations continues. If the free-dom of God is understood in a strong sense, as libertarian freedom,the theist may argue his decision to create the world is a brute fact,that is to say, a contingent fact that has no “sufficient reason.” Thiswould bring an end to the regress of explanations. But if you are af-firming brute facts as your stopping-point, the existence of the uni-verse itself would be a less controversial one.

What moral can we draw from this discussion? Merely positingthe existence of a necessary being does not bring the regress of expla-

nations to an end. Indeed it seems that the task of explanation cannever be brought to an end, except by affirming some fact that can-not itself be explained. There can be no ultimate explanation, even if that explanation posits the existence of God.


Page 72: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 72/205

Page 73: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 73/205

Chapter Eight

Teleological Arguments

Kant’s third type of argument for the existence of God is the teleolog-ical argument (from the Greek telos, meaning a purpose, goal, orend) or “argument from design.” An argument of this kind beginsfrom our observation of the “order, beauty, and complexity of things”(Wilkerson 1976: 150) and arrive at the conclusion that this can onlybe accounted for if there exists a designer. In fact, such arguments

are better described as “arguments to design” or “arguments  for de-sign,” for if there is design, then by definition there must be a de-signer. (One can describe biological organisms as “designed by natu-ral selection,” but this is a metaphorical use of the term.) The ques-tion is: Do we need to regard such features of the world as the workof a personal agent, that is to say, one acting intentionally?

8.1 Traditional Design Arguments

The high-point of the teleological argument was reached in the eigh-teenth century. It was a time when many people were struck by thecomplexity of the natural order that was gradually being revealed bythe modern sciences. The best-known example of an extended designargument is the Natural Theology of William Paley (1743–1805),first published in 1802. And the best known eighteenth-century criti-cism of such arguments is to be found in David Hume’s  Dialogues

Concerning Natural Religion, first published in 1779. The design ar-gument received what appeared to be a fatal blow with the work of Charles Darwin (1809–82), whose Origin of Species came out in1859. But in our own day it has undergone a remarkable revival. Therevival has occurred as a result of developments in two very differentfields of science, namely cosmology and molecular biology. Fromphysics and cosmology, proponents of design draw arguments aboutthe so-called “fine-tuning” of the universe, which (they argue) wasdestined from its earliest period to produce life. From molecular biol-ogy, they draw arguments about what they call the “irreducible” or


Page 74: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 74/205

Teleological Arguments

“specified” complexity of the basic units of organic life, a complexity

which they argue could only be the result of intelligent (or, more pre-cisely, intentional) design.

8.1.1 William Paley’s Design Argument

William Paley’s work, Natural Theology seems the most appropriatestarting point for our discussion. It was certainly an influential book – even the young Charles Darwin found it impressive – and (asRichard Dawkins notes) it was “informed by the best biological schol-arship of his day” (Dawkins 1988: 5). Paley made it very clear what

needs to be explained, namely the organised complexity of the natu-ral world. There is certainly something here which requires explana-tion. If you reject his theistic explanation, you will need to offer an-other. Within Paley’s book the best passage with which to start (if only because it is the most famous) is that with which it begins,namely the analogy of the watch.

In crossing a heath, suppose I pitched my foot against a stone, andwere asked how the stone came to be there: I might possibly an-swer, that for any thing I knew to the contrary, it had lain there

for ever ... But suppose I had found a watch upon the ground, andit should be inquired how the watch happened to be in that place; Ishould hardly think of the answer which I had before given, that,for any thing I knew, the watch might have always been there. Yetwhy should not this answer serve for the watch as well as for thestone? ... For this reason, and for no other, viz. that, when we cometo inspect the watch, we perceive (what we could not discover inthe stone) that its several parts are framed and put together for apurpose, e.g. that they are so formed and adjusted as to produce

motion, and that motion so regulated as to point out the hour of the day; that, if the different parts had been differently shapedfrom what they are, of a different size from what they are, orplaced after any other manner, or in any other order, than that inwhich they are placed, either no motion at all would have beencarried on in the machine, or none that would have answered tothe use that is now served by it. (Paley 1825: 3)


Page 75: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 75/205

Teleological Arguments

On finding such a mechanism, what could we conclude regarding

its origin? The only defensible conclusion, writes Paley, is that

the watch must have had a maker; that there must have existed,at some time, and at some place or another, an artificer or artifi-cers, who formed it for the purpose which we find it actually to an-swer; who comprehended its construction, and designed its use.(Paley 1825: 4)

By analogy, then, we should arrive at the same conclusion whenfaced with biological mechanisms such as the eye, which seem

exquisitely shaped to achieve a particular purpose, in this case,sight. Most of Paley’s work consists of citing instances of such mech-anisms.

 A latter-day Paley could mount a stronger case than did his prede-cessor, for we know so much more than he did about the complexityof the natural world. (As Richard Dawkins writes, “how Paley wouldhave loved the electron microscope!” [Dawkins 1988: 15].) In fact, itis the discovery of the internal mechanisms of the cell which has giv-en rise to one of the recent revivals of the design argument. So I

shall not follow Paley through his numerous examples and illustra-tions. Any textbook of biology will provide the raw material thatcould be taken to support his claims. But before moving on, weshould note a couple of points about Paley’s watch analogy. Firstly, itis an analogy. Paley’s argument rests on the idea that the mecha-nisms of the natural world are in some relevant sense analogous toobjects we already know to have been designed. Secondly, we maydistinguish two things about the watch that Paley sees as requiringexplanation. The first is that the watch serves a purpose that is use-

ful to humans (telling the time). The second is that it is exquisitelyadapted to fulfil that purpose: any alteration to its internal structurewould render it useless. The two features are, of course, closely con-nected, but they are not identical.

8.1.2 David Hume on Design

David Hume is often credited with having undermined the argumentfrom design. But of course he, too, lived before Darwin. In common


Page 76: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 76/205

Teleological Arguments

with his contemporaries ( pace Dawkins 1988: 6), he, too, was struck

by the “order, beauty, and complexity” of nature. And at the end of the day he had no alternative, naturalistic explanation that seemedin any way compelling. Hume’s most important work on this topic,his  Dialogues Concerning Natural Religion, are not like Plato’s dia-logues, where Socrates’s opponents are mere foils for the master’sbrilliance. In Hume’s Dialogues, opposing positions are strongly rep-resented and it is not always clear to which position Hume himself adheres. For the most part his position seems to be that of the philo-sophical sceptic, Philo. What surprises many of Hume’s readers is

that at the end of the  Dialogues (Part XII), after having producedmany powerful arguments against natural theology, Philo acceptsthat the world does offer us evidence of design.

 A purpose, an intention, a design strikes everywhere the mostcareless, the most stupid thinker; and no man can be so hardenedin absurd systems, as at all times to reject it. That nature does

nothing in vain, is a maxim established in all the schools, merelyfrom the contemplation of the works of nature, without any reli-gious purpose; and, from a firm conviction of its truth, an

anatomist, who had observed a new organ or canal, would never besatisfied till he had discovered its use and intention. One greatfoundation of the Copernican system is the maxim, that nature

acts by the simplest methods, and chooses the most proper means

to any end; and astronomers often, without thinking of it, lay thisstrong foundation of piety and religion. This same thing is observ-able in other parts of philosophy: And thus all the sciences almostlead us insensibly to acknowledge a first intelligent Author; andtheir authority is often so much the greater, as they do not directly

profess that intention. (Hume 1993: 116–17)Philo’s acceptance is hedged about with qualifications. He is quick

to insist that there is nothing that can be said about the cause of thisdesign, other than that it “bears some remote analogy to human in-telligence” (Hume 1993: 129). Even if one concedes the legitimacy of the inference to a designer, it remains true that “the universal causeof All” may be “vastly different from mankind, or from any object of 


Page 77: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 77/205

Teleological Arguments

human experience and observation” (Hume 1993: 68 [Part V]). The

cause of this design may not be an infinite being, for a less than infi-nite creator could have created the world. He may not be perfect, foreven if the world shows signs of perfection, it may be the result of along process of trial and error. He is surely not morally perfect, giventhe evil that exists in the world. Indeed, if there a deity, he seems tobe amoral, indifferent to the welfare of his creatures. But we may noteven assume that there is just one author of everything that exists.The world may be the product of a committee of deities.

So a strong case can be made for the idea that Hume was not an

atheist in our modern, everyday sense, namely one who affirms thatthere is no God. (In fact, as I noted earlier, many atheists are moremodest in the claims they make.) Hume may have been regarded byhis contemporaries as an atheist, but in his day it was not difficult toearn that title. If Philo’s views express Hume’s position, he wouldperhaps be better described as an agnostic. (Of course, that term hadnot yet been coined; it is first found in the work of T. H. Huxley[1825–95].) With regard to religion as with regard to other matters,Hume is anxious to point out the limits of our knowledge. In Parts

 VI–VIII of the Dialogues, Philo offers a range of increasingly outra-geous scenarios regarding the origins of the universe. The point isnot that these are being offered as viable alternatives to traditionaltheism. Rather, it is that once we being to speculate on these mat-ters, we can draw no certain conclusion. At the end of Part V Philosums up his argument.

In a word, Cleanthes, a man who follows your hypothesis, is able, per-haps, to assert, or conjecture, that the universe, sometime, arose fromsomething like design: But beyond that position he cannot ascertain

one single circumstance, and is left afterwards to fix every point of histheology, by the utmost licence of fancy and hypothesis. The world,for aught he knows, is very faulty and imperfect, compared to a supe-rior standard; and was only the first rude essay of some infant Deity,who afterwards abandoned it, ashamed of his lame performance; it isthe work only of some dependent, inferior Deity; and is the object of derision to his superiors: it is the product of old age and dotage insome superannuated Deity; and ever since his death, has run at at


Page 78: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 78/205

Teleological Arguments

adventures, from the first impulse and active force, which it received

from him. (Hume 1993: 71)

It follows that “a total suspense of judgment is here our only reason-able resource” (Hume 1993: 88–89 [Part VIII]).

What is interesting is that despite this profound scepticism, Humedoes not dismiss outright the inference to design. In the absence of any compelling naturalistic theory of how the “order, beauty, andcomplexity” of nature came about, Hume (or at least Philo) concedesthat whatever brought it about may indeed bear some resemblanceto human intelligence. Hume did demonstrate the limits of the de-

sign argument, by showing how little it could actually prove. He cer-tainly made it impossible for such an argument to act as the founda-tion of a system of religion, which was (of course) the intention of Pa-ley and his predecessors. In Hume’s day, this was already an impres-sive achievement. But he could not entirely discount the power of what we might call “the design inference.” It was only with the workof Darwin that there emerged a naturalistic alternative to the reli-gious hypothesis. One could be an atheist before 1859, but at the costof leaving some important questions unanswered. It was only Dar-

win, it seems, who made it possible to be “an intellectually fulfilledatheist” (Dawkins 1988: 6).

Despite Hume’s criticisms and Darwin’s theory, the design argu-ment has not gone away. And just as it was the scientific develop-ments of the eighteenth century that lent it support in its heyday, soit is the scientific developments of our time that are contributing toits revival. It is not surprising that a new argument should emergefrom the discoveries of the biological sciences, in the form of “intelli-gent design theory.” After all, Paley drew most of his examples from

the world of living organisms. I shall not be discussing the intelligentdesign argument here, since the issues it raises are essentially thesame as those discussed by Hume. What is perhaps more surprisingis that a new form of design argument should emerge from a quitedifferent area of science, namely physics and cosmology. This is whatis known as the “fine-tuning” argument and it does merit closerscrutiny.


Page 79: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 79/205

Teleological Arguments

8.2 The Fine-Tuning Argument

The basis of the so-called “fine-tuning” argument is a fact about theuniverse, the significance of which as been highlighted by recent de-velopments in cosmology. It appears that the values of certain “cos-mic parameters” (to use Neil Manson’s phrase) are contingent. Theycould have been other than they are. But if they had been other thanwhat they are, by even a very small margin, the universe could nothave sustained life or (in a weaker form of the argument) would havebeen much less favourable to life. In this sense the universe could be

said to have been “fine-tuned” for life. How can we explain this fact,it is argued, if not by positing the existence of a celestial tuner, a cos-mic designer?

8.2.1 The Form of the Argument

There are many such cosmic parameters that are cited in the litera-ture. Robin Collins offers a list of six, which he regards as “solid cas-es of fine-tuning” (Collins 2003: 180). First, there is the cosmological

constant (customarily represented as L), which is a measure of the

force which determines both the expansion and contraction of space.Second, there is the strong nuclear force, which keeps the protonsand neutrons together in the atom. Third, there are the various fac-tors involved in the conversion of helium to carbon and oxygen with-in stars, the key issue here being the balance that must be struck be-tween carbon and oxygen production. Fourth, there is the differencein mass between neutrons and protons. If the mass of the neutronwere even fractionally higher, stars would not be able to convert hy-drogen to helium. If it were fractionally lower, the universe would be

composed of little more than helium. Fifth, there is the weak force,which controls radioactive decay. If this were not within a relativelynarrow range of values, the ratio of neutrons to protons producedwithin the first seconds of the Big Bang would have produced a uni-verse much less favourable to life. Sixth, there is the strength of  gravity, which must also fall within relatively narrow limits if life aswe know it is to survive. What is striking here, of course, is not


Page 80: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 80/205

Teleological Arguments

merely the precise value of each constant, but their coincidence, pro-

ducing a universe capable of supporting beings like ourselves.The scientific accuracy of such claims is a question for a physicist,

not a philosopher. (And I freely admit I would have difficulty under-standing the answer.) All I can do here is to is to examine the argu-ment to design that is constructed on this basis. In its simplest form,the argument claims that these values (and their coincidence) wouldbe much less surprising  if the universe were fine-tuned to producelife by some intelligent agent than if it were not. If these values werethe result of chance, this would represent an incredible coincidence.

But if these values were the result of design, they are what youmight expect. (Proponents of the argument assume that “chance” and“design” are the only alternatives, but I’ll leave that aside for the mo-ment.) If the evidence we have is less surprising on the design hy-pothesis than on the chance hypothesis, then the evidence  favours

the design hypothesis over its rival.That was an informal presentation of the fine-tuning argument.

Let me set out its central claim more formally. The argument restson a judgement of probability. More precisely, it rests on a judge-

ment of comparative probability, the probability that the parametersin question have the life-supporting values they do (let’s call this factE), given the design hypothesis (D), over against the probability thatthey have the values they have, given the chance hypothesis (C). Theclaim is that the probability of E given D is greater than the proba-bility of E given C. Using a customary set of symbols, one can ex-press this as follows:

Pr(E∣D) > Pr(E∣C).

There are two sets of probabilities to be calculated here, namelyPr(E∣D) and Pr(E∣C). There are, as we shall see, difficulties associat-ed with both calculations.

But there is a more important feature of the argument, which isoften overlooked. It is that even if we had demonstrated this centralclaim, we would not yet have demonstrated the  probable truth of D. All we would have shown is that the observations in question are


Page 81: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 81/205

Teleological Arguments

more likely on the basis of D rather than C. (To use a more technical

term, this is sometimes known as a “likelihood argument.”) Thiswould lend some support to D, but it would not (yet) have shown thatD is true. The point may be illustrated with a simple example fromElliot Sober. When you hear noises in the attic, the hypothesis that“there are gremlins up there bowling” would certainly explain thenoises (Sober 2003: 29). If it were true, it would make the noises lesssurprising. But we would not normally conclude that the gremlin hy-pothesis is (probably) true. (We shall see why in a moment.) Similar-ly, even if we had shown that the evidence (E) is unsurprising on the

basis of hypothesis (D), we would not have shown that the hypothe-sis (D) is true. The observation would lend some support to the de-sign hypothesis, over against its rivals. But we would not haveshown that the universe is in fact the result of design.

To do that, we would need to show that the probability of D, giventhe evidence, is greater than that of C. This means reversing theterms within the brackets, so that they read not E∣D but D∣E, andnot E∣C but C∣E. The resulting formula is:

Pr(D∣E) > Pr(C∣E).

To see if this were true we would need to take into account (amongother things) the prior probability of D, that is to say, its probabilitygiven everything else we know. A low prior probability for the hy-pothesis can swing the case. To take Sober’s example, the probabilitythat you would hear the noises in the attic (N) if there were gremlins(G) – Pr(N∣G) – is high. But the probability that there are gremlinsin the attic – Pr(G∣N) – is low, since we have other reasons to believethere are no gremlins. The gremlin hypothesis fails to be a  plausible

hypothesis.To note that the fine-tuning argument could not, in itself, estab-

lish the probable truth of theism is not to say that it is worthless. Itcould form part of a useful, cumulative case for theism. But by itself the best it can do is to lend some support to the design hypothesis(McMullin 1993: 381–82). The question is: Does it succeed in doingeven that?


Page 82: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 82/205

Teleological Arguments

8.2.2 The Anthropic Selection Effect

One response to this argument is simply to deny that there is anymystery to be resolved, to argue there is nothing surprising aboutthe facts on which the argument is built. One could, of course, try todo this on scientific grounds, by showing that the parameters inquestion had to be more or less as they are. (It seems this has not yetbeen done, but I shall come back to the possibility that it might yetbe done.) What I am interested in here is a  philosophical response,one that points out the “selection effect” involved. The problem hereis a simple one, which has to do with how data are gathered or obser-vations are made. Let me take a simple example, based once againon one offered by Elliot Sober. Let’s say you were to catch a sampleof fish from Otago Harbour and you found that all the fish were morethan ten centimetres in length. Would this observation lend supportto the hypothesis that all the fish in the harbour are more than tencentimetres in length? Well, it would depend on how you caught thefish. If your net had holes so wide that fish fewer than ten centime-tres in length escaped, your sample would be worthless (for this pur-

pose). It would lend no support to the hypothesis. In these circum-stances, what other result would you expect?The analogy here has to do with our very presence as observers

within the universe. As we have seen, the probability (or “likeli-hood”) calculation at the heart of the fine-tuning argument is

Pr(E∣D) > Pr(E∣C).

But this way of presenting this issue does not take a key fact into ac-

count. It is the fact that if the universe were not such as to support

life, we would not be here to observe it. Let’s call this fact A.

(A) If we exist, then E must be as it is.

For completeness we can add the fact that we exist (W). So a moreaccurate presentation of the fine-tuning argument would be

Pr(E∣D & A & W) > Pr(E∣C & A & W).


Page 83: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 83/205

Teleological Arguments

But is this true? No, it is not. For the probability of E (given the

truth of W and A) is 1.0, and it remains 1.0 no matter which hypoth-esis is being considered. To put this in terms of our formula,

Pr(E∣D & A & W) = Pr(E∣C & A & W) = 1.0.

In other words, given that we exist “the constants must be right,regardless of whether the Universe was produced by intelligent de-sign or by chance” (Sober 2003: 44).

This is an expression of what Brandon Carter called the “weak an-thropic principle” (WAP), namely that “what we can expect to ob-

serve must be restricted by the conditions necessary for our presenceas observers” (Sober 2003: 44–45). A more precise term would be theanthropic selection effect. It is a “What else would you expect?” prin-ciple, which at first sight dispels the mystery that the fine-tuning ar-gument is intended to solve. A less esoteric example of the fallacymight help. One could argue that if the force of gravity were greaterthan it is, we wouldn’t be able to stand upright. Does it follow thatthe force of gravity was “fine-tuned” for our sake? Of course not. Toclaim that it was would be to put the cart before the horse. (I am

grateful to a former student for this illustration.)But not everyone accepts this argument. Richard Swinburne, for

example, offers what he considers to be a counterexample, based onan analogy from John Leslie.

On a certain occasion, the firing squad aim their rifles at the pris-oner to be executed. There are twelve expert marksmen in the fir-ing squad, and they fire twelve rounds each. However, on this oc-casion, all 144 shots miss. The prisoner laughs and comments thatthe event is not something requiring any explanation because if 

the marksmen had not missed, he would not be there to observethem. But of course, the prisoner's comment is absurd; the marks-men all having missed is indeed something requiring explanation;and so, too, is what goes with it – the prisoner's being alive to ob-serve it. (Cited in Sober 2003: 46)

This seems intuitively plausible, but does the analogy fit? ElliotSober believes it does not. He responds by pointing out the difference


Page 84: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 84/205

Teleological Arguments

between the situation of the prisoner and that of a bystander. The

prisoner, he argues, suffers from a blindspot. “A proposition p,” hewrites, “is a blindspot for an individual S just in case, if p were true,S would not be able to know that p were true” (Sober 2003: 48). Or,to put this differently, “if p is a blindspot for S, then if S makes anobservation to determine the truth value of p, the outcome must bethat not-p is observed” (Sober 2003: 48). It follows that the by-stander, but not the prisoner, is able to use his observations to arguethat the survival of the prisoner happened by design rather than bychance.

How does this apply to the fine-tuning argument? In the case of the fine-tuning argument,  p is the proposition, “The constants werewrong.” What is special about the fine-tuning argument is that weare all in the position of the prisoner; we all suffer from a blindspot.Whatever observation we make, it will inevitably support the conclu-sion not- p. A bystander could, perhaps, argue from the value of thesecosmic parameters to the existence of a designer. But we cannot doso, since we are not bystanders. Swinburne’s analogy appears per-suasive only if we adopt the viewpoint of the bystander. Or, to put

this another way, the prisoner’s proper response should have been,“Given that I am able to make observations, I must be alive, whethermy survival was due to intelligent design or chance” (Sober 2003:47). Only this response would recognise the existence of a selectioneffect.

8.2.3 The Probability of E, Given Design

Is Sober correct? I’m not sure. (What do you think?) But let’s giventhe fine-tuning argument the benefit of the doubt and assume he is

wrong. Let’s assume that the fine-tuning phenomenon is not just apseudo-problem, the product of a selection effect. Let’s assume, too( pace Manson 2000), that its proponents can produce an adequate,non-question-begging description of the phenomenon to be explained(and that it still requires explanation!). At the heart of the fine-tun-ing argument, as we have seen, is the following claim:

Pr(E∣D) > Pr(E∣C).


Page 85: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 85/205

Teleological Arguments

What is important to appreciate is that the design hypothesis does

not win by default. It is not enough to say that E is highly improba-ble, given chance. The theist must argue that E is more probable,given design. I have already discussed the difficulties inherent in es-timating Pr(E∣C). Can we estimate Pr(E∣D), the probability thatthese cosmic parameters should have the value they do, given the de-sign hypothesis?

Proponents of the fine-tuning argument often assume that we can.More precisely, they simply take it for granted that the probability of E is high, given design. The assumption is that if there were a God,

who intended to create beings such as ourselves, then he would setup the constants of physics so that they would have the values weobserve. But would he? Some opponents of design arguments arguethat we simply cannot answer this question. The posited divineagent is so unlike any agent with which we are familiar that we haveno idea how he might be expected to act. (See, for instance, Sober2003: 38.) But if they are correct, if we cannot estimate Pr(E∣D), thenthe fine-tuning argument (indeed any design argument) collapsesimmediately. For all we know, it may be the case that its value is

lower than that of Pr(E∣C). I disagree with this view. I think we can

estimate Pr(E∣D), on the assumption that the divine agent is actingrationally, that is consistently with his posited intentions and hischaracter. (This rationality assumption, by the way, characterisesall intentional explanations – those which appeal to the purposes of an agent – and theistic explanations are intentional explanations.)But what is the result? How likely is E, given D?

The question here is: If we assume that the posited designer isomnipotent, omniscient, and morally perfect, and if E is the history

of the universe as we know it, does E represent the way in which wewould expect him to create beings like ourselves? This is the ques-tion which Hume once posed: “Is the world considered in general,and as it appears to us in this life, different from what a man or sucha limited being would, beforehand, expect from a very powerful, wise,and benevolent Deity?” (Hume 1993: 107 [Part XI]). I won’t try to an-swer that question here. But then I haven’t seen any proponent of 


Page 86: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 86/205

Teleological Arguments

the fine-tuning argument answering it, either. There are at least two

issues which would have to be taken into account here. The first isthe problem of evil, particularly in the light of Darwin’s account of human origins, where suffering seems to be a necessary condition of the emergence of sophisticated forms of life. The second is that evenif the initial conditions of the universe were fine-tuned to producelife, the emergence of sophisticated life-forms was by no means guar-anteed. Would God have left so much to chance? I don’t know. (Onthis topic, see Quentin Smith’s comments in Craig and Smith 1993:200–1.) But again, these are questions which need answering.

What, then, can we make of the fine-tuning argument? It is diffi-cult to know precisely what “fine-tuning” might mean and, insofar aswe do understand it, the phenomenon in question may not be as puz-zling as it appears at first. It may even be a non-issue, the product of a selection effect or of the units of measurement with which it is de-scribed (see Manson 2000). If there is a real question here, if it is im-probable that these cosmic parameters should have taken the precisevalue they have by chance, then it may also be improbable that theyshould have taken the precise value they have by design, at least on

the assumption that the designer is omnipotent, omniscient, andmorally perfect. Is the former (the chance hypothesis) more improba-ble than the latter (the design hypothesis)? Since it seems impossibleto quantify one or the other, I really don’t know. But if neither hy-pothesis renders what we observe probable, the most rational re-sponse is surely to refrain from adopting either, in the hope that abetter alternative will emerge.

One such alternative is the “multiverse” scenario. This is the ideathat there exist at present, or have existed in succession, many uni-

verses, causally disconnected from one another, with different lawsof nature. Given a sufficient number of such universes, it would benot be surprising that one like ours has emerged. Is this a better al-ternative? On the face of it, it seems both ad hoc and extravagant, afairly blatant violation of Ockham’s razor. (I’m almost tempted to saythat one God is better than many universes.) One would certainlywant to have independent evidence of its truth. And if the posited


Page 87: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 87/205

Teleological Arguments

universes are causally disconnected from our own, it is hard to see

what form that evidence could take. But to discuss this option fur-ther would take us too far afield. My point is merely that if there is aso-called “fine-tuning” problem, then none of the currently availablesolutions seems terribly satisfactory.


Page 88: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 88/205

Page 89: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 89/205

Chapter Nine

Logical Arguments from Evil

So far I have been looking at the reasons theists can offer for belief in God. That they are obliged to offer such reasons emerges fromwhat I called “the presumption of atheism.” But as we approach theproblem of evil, we see the atheist go on the offensive. Since ancienttimes, the existence of apparently senseless and horrendous evils,

both in the sense of moral evil and in the sense of human and animalsuffering, has been recognized as an argument against belief in anall-powerful and benevolent deity. David Hume attributes the follow-ing formulation of the argument to Epicurus (341–270 BC), althoughit does not seem to correspond to anything that ancient Greekphilosopher wrote. “Epicurus’s old questions [about God] are yetunanswered. Is he willing to prevent evil, but not able? then he isimpotent. Is he able, but not willing? then he is malevolent. Is heboth able and willing? whence then is evil?” (Hume 1993: 100).

9.1 Logical and Evidential Arguments

In its basic outline, that is the argument from evil. But we need to bemore precise. There are two ways in which this argument can be de-veloped. The first is sometimes referred to as the logical argument

 from evil. It suggests that there is some kind of incompatibility be-tween the existence of God and that of evil. It simply could not betrue, the argument runs, that God exists and that evil exists. I shallexamine one form of this argument based on a famous 1955 article

by J. L. Mackie. If a logical argument from evil were to be successful,it would be a knock-down argument, a proof that settled the ques-tion once and for all. The best-known response to this argument isthe freewill defence, of which Alvin Plantinga has recently offered anew form. This is not as convincing as many of its advocates assume,but it does indicate the kinds of difficulties the logical argumentfaces.


Page 90: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 90/205

Logical Arguments from Evil

 A second argument for atheism is sometimes referred to as the ev-

idential argument from evil. In the following chapter, I shall look atthe most convincing form of this argument, that offered by PaulDraper. Draper sets up his argument by comparing the explanatorypower of theism with that of an alternative, naturalistic hypothesis,which he calls the “hypothesis of indifference.” The most popular the-istic response to the evidential argument from evil is what is some-times called sceptical theism. I shall briefly explore two forms of thisposition, that put forward by Stephen Wykstra and that offered byPeter van Inwagen, and assess their adequacy.

9.2 The Logical Argument

Let’s turn first to the logical argument from evil and the theistic de-fences that have been offered in response. What has come to be re-garded as the classic statement of this argument is to be found in a1955 article by J. L. Mackie. In its simplest form, Mackie writes, theargument goes as follows:

(1) God is omnipotent.

(2) God is wholly good.(3) Evil exists.

 As it stands, however, these three propositions are not obviouslyinconsistent. Some additional premises must be added to construct acomplete argument. Mackie’s suggestions are the following, at firstsight uncontroversial claims:

(4) A good being always eliminates evil as far as it can.

(5) There are no limits to what an omnipotent being can do.

Since an omnipotent being is capable of eliminating evil, from these

it follows that

(6) An good omnipotent being eliminates evil completely.

Then, given (1) and (2) it follows that

(7) There is no evil.


Page 91: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 91/205

Logical Arguments from Evil

But (7) contradicts (3). So what have we found? If we join together

two statements of theistic belief (1 and 2) with two apparently un-controversial further premises (4 and 5), the result contradicts anapparently indisputable fact about the world (3). What kind of con-tradiction are we dealing with here? It seems that what we are deal-ing with are statements whose conjunction is necessarily false, thatis to say “false in every possible world” (Plantinga 1974: 165). In oth-er words, it just could not be the case that all seven propositions aretrue. Since the existence of evil is more obvious than the existence of God, it is the latter that has to go.

9.3 Theistic Defences

Or is it? What other premises might be challenged? The theist couldtry to deny the existence of evil. There is a long-standing theologicaltradition that sees evil as merely a privation, the absence of good-ness ( privatio boni). But whatever one makes of that as a metaphysi-cal claim, it won’t take us very far in this context. After all, the factthat darkness is the absence of light does not mean that there are nodark rooms or dark objects. By analogy, what one would have to deny

is the existence of  evils, that is to say, instances of suffering or of moral wrongdoing. This is not something that many theists are pre-pared to do. (Mary Baker Eddy, the founder of the Christian Sciencemovement, is perhaps an exception.)

9.3.1 The Freewill Defence

 A more promising line would be to challenge one of the additionalpremises (4 and 5). A simple response would be to note that (4) re-quires modification, for there may be goods that cannot be attained

without permitting some degree of evil. A modified version wouldread

(4a) A good being always eliminates evil as far as it can, so long as

that evil is not a necessary condition of a greater good.

If the theist can suggest goods that God might seek and that wouldrequire the toleration of a certain amount of evil, the logical argu-ment from evil will be defeated. The most common suggestion, dating


Page 92: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 92/205

Logical Arguments from Evil

back to St Augustine (AD 354–430), has been that free will repre-

sents such a good. The existence of creatures enjoying some degree of moral autonomy is a great good, but one that could not be had with-out God’s willingness to tolerate some degree of moral evil (and itsnon-moral consequences). This is what has become known as the freewill defence.

9.3.2 Mackie’s Initial Response

The problem (for the theist) is that there is an obvious response tothe freewill defence. It is that God could have created beings who al-

ways freely chose the good. As Mackie writes,if there is no logical impossibility in a man’s freely choosing thegood on one, or on several, occasions, there cannot be a logical im-possibility in his freely choosing the good on every occasion. Godwas not, then, faced with a choice between making innocent au-tomata and making beings who, in acting freely, would sometimesgo wrong: there was open to him the obviously better possibility of making beings who would act freely but always go right. Clearlyhis failure to avail himself of this possibility is inconsistent with

his being both omnipotent and wholly good. (Mackie 1955: 209)So if the freewill defence is to work, this possibility must be exclud-ed. This would mean challenging proposition (5). More precisely, forthe freewill defence against the logical argument from evil to work,one must show that there exists a possible world in which God couldnot have created beings who always freely chose the good. For re-member, at the heart of the logical argument from evil is the ideathat the conjunction of propositions (1) to (7) is false in every possibleworld. So let’s replace (5) with

(5a) There is a possible world in which even an omnipotent beingcould not eliminate all evil.

If this were true, the logical argument from evil would be defeated.

9.3.3 Plantinga’s Freewill Defence

But is there such a possible world? Yes, says Alvin Plantinga. Inwhat has become a famous expansion of the freewill defence,


Page 93: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 93/205

Logical Arguments from Evil

Plantinga has argued that there could exist a possible, although

tragic, condition called transworld depravity. Plantinga illustratesthis possible state of affairs by reference to the character of CurleySmith, a fictional mayor whose great weakness is taking bribes. If Curley suffers from transworld depravity, it means that his charac-ter (technically, his essence) is such that in whatever world he exist-ed, he would freely choose to take a bribe.

Now it is possible that not only Curley, but everyone should sufferfrom transworld depravity. If we all suffered from transworld de-pravity, then what would follow? Since this depravity is a

transworld depravity – one that would be realised in any possibleworld – there would be no possible world in which God could elimi-nate evil. This depravity would “fill up” (as it were) the space of pos-sible worlds. What does this show? Assuming that our decisions arefree (in a strong, libertarian sense), this argument shows that (5a)above is true. There is a possible world – one in which transworld de-pravity prevails – in which not even God could eliminate all evil.

But what about natural evils, such as the suffering caused byearthquakes, diseases and floods, which do not appear to be the re-

sult of human free choices? Surely this modified freewill defence doesnot cover them? Well, says Plantinga, it could. Once again, to defeatthe logical argument from evil, all one needs is a possible state of af-fairs. Well, here is one. The natural evils we experience could becaused by the free choices of  non-human beings, namely the fallenangels led by Satan. Once again, this is part of a traditional Chris-tian theodicy, which has its own problems (Hick 1966: 367–69). ButPlantinga is merely offering a defence, so he feels he can ignore thosedifficulties (which in any case he is inclined to minimize). If this sce-

nario were true, then the existence of natural evils also fall withinthe scope freewill defence. Since this is also a possible state of af-fairs, it follows that the logical argument from evil fails.

9.3.4 Mackie’s New Response

Or does it? Mackie responds to Plantinga’s argument in his posthu-mously published The Miracle of Theism (1982). He recognises that


Page 94: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 94/205

Logical Arguments from Evil

the logical argument from evil cannot be considered a conclusive

refutation of theism, because there is always a possibility of adjust-ing the additional premises it requires so as to avoid contradiction(Mackie 1982: 176). But he insists that this has not yet been done. Inparticular, he argues vigorously against Plantinga’s expandedfreewill defence. Once again, Mackie’s argument rests on the ideathat there is nothing contradictory about the idea that human be-ings, while remaining free, would always choose the good. So if Godis omnipotent and if omnipotence implies the ability to do anythingthat is logically possible, God could have created human beings who

always freely chose what was right. (Presumably God could have cre-ated angels who did the same.) It is true that this argument seemsless controversial if one adopts a compatibilist rather than a libertar-ian view of freedom. But it does not depend on any particular view of human freedom, as Plantinga’s suggestions regarding transworld de-pravity show. After all, Plantinga is a libertarian, but suggests thata person’s essence might be such that he always freely chose what isevil. So a libertarian should be able to accept that a person’s essencemight be such that she always freely chose what is good.

It is true that if you hold to a libertarian view of freedom, Mack-ie’s argument does not automatically go through. It requires a fur-ther premise: the idea that at the moment of creating a world Godwould know that the individuals he was creating would freely choosethe good. Luis de Molina (1535–1600) referred to this as “middleknowledge,” insofar as it stands “between” God’s knowledge of whatis necessary and what is caused by his choices. (Contemporary dis-cussions speak of “counterfactuals of freedom,” statements aboutwhat a particular person would freely choose given particular situa-

tions.) The possibility of middle knowledge is a hotly contested idea,but it happens to be one that Plantinga accepts. From a theist’s pointof view, you can understand why. The rejection of this idea trans-forms the omnipotent God of traditional theism into a being whotook “literally, a hell of a risk” in creating free creatures (Mackie1982: 176).


Page 95: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 95/205

Logical Arguments from Evil

What about Plantinga’s apparent coup de grâce against the logical

argument from evil, namely the possibility of universal transworlddepravity? The existence of universal transworld depravity would

place limits on God’s omnipotence. It would mean that there is a pos-sible world in which God could not have created beings that alwayschose the good. If so, the logical argument fails. Well, says Mackie,under what conditions would transworld depravity be actualised?The existence of transworld depravity would imply that, prior to theact of creation, God was faced with a limited range of creaturelyessences from which to choose. For it seems that he did not have the

choice of creating beings who would always freely choose the good.But since such creatures are not logically impossible, what couldhave imposed such a limitation on an omnipotent God? Mackie con-cludes that Plantinga’s expanded freewill defence “is simply incoher-ent” (Mackie 1982: 174).

In his two presentations of his argument, Plantinga seems awarethat this is a possible objection. Could God not have created otherpeople, who do not suffer from transworld depravity? His response is:“Perhaps. But then again, perhaps not” (Plantinga 1974: 187; cf.

1975: 49.) But in the discussion that follows he offers no reason tothink not. He merely reiterates the possibility of universaltransworld depravity and argues that, if  this possibility were real-ized, God could not have created a world in which creatures alwaysfreely chose the good. It is true that William Lane Craig has an argu-ment that might aid Plantinga at this point. His argument suggeststhat even if it possible for God to create a particular individual whoalways choose the good, it may not be possible for him to create aworld in which all individuals always choose the good. For their co-

existence may not be possible (Craig 1989). Well, that’s true; it maynot be possible, but this needs to be shown. There is no obvious inco-herence in the idea of such a world, so it is up to the theist to showthat not even God could bring it about.


Page 96: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 96/205

Logical Arguments from Evil

9.3.5 Beyond Mackie’s Response

It is true that, in and of themselves, Mackie’s reflections do not un-dermine Plantinga’s argument. One could posit a possible scenario inwhich God  freely chooses to create beings who would suffer fromtransworld depravity. This would also undermine the logical argu-ment from evil. But this solution comes at a cost. For it would surelycast doubt on the idea that God is perfectly good. This would pose not just an evidential problem for Plantinga, having to so with whetheruniversal transworld depravity is likely to be the case. It poses a log-

ical problem. There is an important principle here. When presentedwith a theistic defence, we should ask what its consequences wouldbe if it were true. If those consequences are themselves inconsistentwith theism, it fails as a defence. This seems the only way of avoid-ing silly defences. (Here’s one: “There is a possible world in whichGod was having a bad day and simply failed to notice that the worldhe was creating would contain evil. So there is a possible world inwhich evil and God coexist. So the logical argument from evil fails.”)If the doctrine of transworld depravity is inconsistent with the doc-

trine of God’s goodness, it is a silly defence, even if it is offered as amere logical possibility. There are other difficulties with Plantinga’sargument – Graham Oppy, for instance, argues that God would notbe constrained, when choosing to creating a world, by counterfactu-als of freedom (Oppy 2004: 63–67) – but this one seems to me to befatal. Until it is answered, we should not assume we have seen thelast of the logical argument from evil.


Page 97: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 97/205

Chapter Ten

Evidential Arguments from Evil

Even if the logical argument from evil could be defeated, there re-

mains another weapon in the atheist’s arsenal: the evidential argu-

ment from evil. In its simplest form, this argument suggests that the

existence of evil is evidence against the existence of God, which out-

weighs any evidence  for his existence. Given such evils as we seeabout us, the balance of probabilities is on the side of the non-exis-

tence of an all-powerful and morally perfect God. But that’s a very

informal presentation of the argument. Let’s look at a more formal

one. The best of these (in my view) is that of Paul Draper.

10.1 Paul Draper’s Hypothesis of Indifference

The basic form of any evidential argument from evil is the idea thattheism, or what we might call “the theistic hypothesis,” cannot ex-

plain the evil we find in the world. To the question “If God exists,why is there such evil in the world?” the theist (it is argued) can offerno satisfactory answer. Draper argues that the traditional discus-sions of this issue have overlooked the fact that the theistic hypothe-sis is best tested over against its competitors. (Although Draper doesnot put it in these terms, this is implicit in the idea that theism maybe regarded as the best explanation of some phenomenon.) In thecase of the existence of evil, there is an alternative, which Drapercalls the hypothesis of indifference (HI).

HI: neither the nature not the condition of sentient beings onearth is the result of benevolent or malevolent actions performedby non-human persons. (Draper 1989: 332)

The facts to be explained here are not just those regarding sufferingand evil; they include the existence, distribution and quantity of pleasure as well as of pain. So we may now ask: Which hypothesis,the theistic hypothesis or the hypothesis of indifference, better ex-


Page 98: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 98/205

Evidential Arguments from Evil

plains these facts? Or, as Draper puts it, on the basis of which hy-

pothesis are these facts less surprising ?

10.1.1 Arguments for HI

It is easy to give an intuitive answer to this question, but rathermore difficult to offer a precise one. Draper’s argument centres onthree facts which he argues are less surprising on the basis of HIthan on the basis of theism. The first relates to the fact that painand pleasure have both biological functions and moral significance.But the actual distribution of pain and pleasure seems to reflect only

one of these, namely their biological functions. This is unsurprisingon the basis of HI. An indifferent universe or creator will (by defini-tion) have no concern for the moral value of pleasure and pain. But if God existed, one would expect that the distribution of pleasure andpain would reflect their moral value. God would have a reason to cre-ate pleasure even if it served no biological function and a reason notto permit pain unless it served a sufficient moral purpose.

The second fact is that sentient beings who are not moral agents(non-human animals and very small children) also experience plea-

sure and pain. This is unsurprising given HI, for the pleasures andpains that Draper includes in this category serve a useful biologicalfunction. If HI is true, then biological function is the only significantconsideration. But the suffering  of non-moral agents is surprisinggiven theism. If theism were true, one would expect God to permitonly those pains that can play a moral role in people’s lives. Whywould God permit beings who are not moral agents – who enjoy nei-ther free will nor the possibility of moral growth – to suffer pain?

The third fact is that sentient beings experience pains and plea-

sures that seem to have no biological function. Now for the reasonsgiven above, the existence of  pleasures that serve no biological func-tion would be unsurprising on theism and surprising on HI. But theexistence of  pains that serve neither a biological function nor (itseems) a moral function is surprising on the basis of theism. Whatabout HI? Well, the the existence of pains and pleasures that serveno useful biological function is probably an epiphenomenon. It is a


Page 99: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 99/205

Evidential Arguments from Evil

consequence of either the breakdown of biologically useful systems

(pathological pain) or from the functioning of biologically useful sys-tems in extreme situations (biologically appropriate pain). For in-stance, it is biologically useful that beings feel pain when exposed tofire, but that function breaks down when the fire is severe enough tokill. The existence of such – sometimes unpleasant and morally re-grettable – epiphenomena is unsurprising given HI. After all, nature(on this view) simply doesn’t care. But it is surprising given theism,both for the reason given above and because one would expect an om-nipotent and morally perfect creator to arrange things better.

10.1.2 Theodicies and Probability

What about theodicies? Do these not raise the probability of theismto the point that it might be a serious rival to HI? Draper considersthe free will theodicy, which is the most popular. He notes first of allthat the freedom to perform right and wrong actions does not of itself entail the existence of pain. “God could have given humans freedomwithout permitting pain” (Draper 1989: 341). So something more isneeded than the assertion that God wanted to create free creatures.

It might be the idea that having created free creatures, God couldnot prevent them inflicting pain. Or it might be the idea that Godmight use pain to influence people to perform morally right actions.The problem is, once again, that the actual distribution of painseems indifferent to such considerations, a fact that is less surprisinggiven HI than given theism. Draper considers one further develop-ment of the free will theodicy: the idea that the existence of pain wasnecessary for humans to have morally significant choices. But if Godexisted, one would expect him, like a good parent, to proportion our

capacity for evil-doing to our moral development. Once again, thefacts about evil-doing are less surprising on the basis of HI than onthe basis of theism.

10.1.3 A Likelihood Argument

Note that Draper’s argument is what I earlier called a “likelihood ar-gument.” (See my discussion of “teleological arguments” in chapterseven.) It compares the likelihood that the world would be as we ob-


Page 100: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 100/205

Evidential Arguments from Evil

serve it to be (O), given theism (T), with the likelihood that it would

be as we observe it to be, given the hypothesis of indifference (HI). Itdoes not pretend to settle the question of which is probably true,which would involve taking into account the  prior probability of each. More technically, the argument claims that

Pr(O∣HI) > Pr(O∣T).

It does not claim that

Pr(HI∣O) > Pr(T∣O).

Nonetheless, if it is successful, Draper’s argument provides us withevidence against theism and in support of the hypothesis of indiffer-ence.

10.2 The “Sceptical Theism” Defence

There are two theistic responses to Draper’s argument, which aresometimes combined. The first is Alvin Plantinga’s ‘Reformed Episte-mology,” which I shall discuss later. The second is what is sometimescalled “sceptical theism.” I shall use this phrase to refer to any posi-

tion that wards off the argument from evil by pointing to the limita-tions of our knowledge. It is not a new position. What some philoso-phers call “the mystery defence” was familiar to generations of Ro-man Catholic schoolchildren: their more difficult questions regardingthe faith were regularly met with the response, “it’s a mystery.” Thegood sisters who taught them were, of course, merely echoing thewell-established theological view, namely that the content of re-vealed truth often exceeds the capacity of human understanding. Be-cause you are accepting these truths on the authority of God, you can

believe them without fully understanding them. (I shall come back tothis point shortly, when I discuss the nature of religious belief.)Within the philosophical literature, many of the themes of scepticaltheism were first outlined in an article by Ninian Smart (1961), towhich both J. L. Mackie (1962) and Antony Flew (1962) offered re-sponses. But it is only in recent years that it has been developed in asystematic way.


Page 101: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 101/205

Evidential Arguments from Evil

10.2.1 Varieties of Sceptical Theism

Sceptical theism is an umbrella term: there are several different po-sitions which it covers. We may distinguish these by distinguishingthree kinds of alleged knowledge regarding which the sceptical theistmay be sceptical. The first has to do with the  greater goods thepreservation of which may justify the existence of evil. Given God’somniscience and our cognitive limitations, it is argued, it is surely tobe expected that God will know of goods that we are unaware of,goods that can be attained only by permitting evil. In support of thisview, Stephen Wykstra has coined the clever term CORNEA, forCondition of Reasonable Epistemic Access. CORNEA reads:

On the basis of cognized situation s, human H is entitled to claim“It appears that p” only if it is reasonable to believe that, given hercognitive faculties and the use she has made of them, if  p were notthe case, s would likely be different than it is in some way dis-cernible by her. (Wykstra 1990: 153)

In some instances (such as sense perception in ordinary circum-stances or memory) this condition is met and the phrase “it appears

that” has some epistemic significance. But why should we assumethat it does so in this situation? Why should we pay any attention tothe atheist’s claim that there appears to be no good that would out-weigh a certain evil? Even if there were such goods, we would be inno position to know about them.

 A second form of scepticism has to do with the kinds of world thatGod might have created in order to preserve those greater goods.Such scepticism is sometimes described as modal scepticism, since ithas to do with the possibilities open to even an omnipotent creator.

Peter van Inwagen is the best-known exponent of this position. VanInwagen denies that we can make reliable, intuitive judgementsabout what is possible in “circumstances remote from the practicalbusiness of everyday life” (van Inwagen 1998: 70). A good case can bemade for the reliability of  certain everyday modal judgements. Weknow, for instance, that the table placed under the window couldhave been placed a metre to the left. Experience shows that we are


Page 102: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 102/205

Evidential Arguments from Evil

often successful in making such judgements. And given what know of 

our cognitive mechanisms and their evolutionary history that suc-cess is unsurprising. We evolved precisely in order to deal with suchsituations. But why should we assume that our capacity for modal judgements goes beyond these everyday matters? How can we expectto know, for example, what kind of a world an omnipotent deitycould have created?

 Van Inwagen identifies a third form of scepticism, which he callsmoral scepticism (van Inwagen 1991: 151). This is related to but notidentical with Wykstra’s scepticism regarding goods. Van Inwagen’s

moral scepticism has to do with our ability to make judgements re-garding the comparative moral value of different states of affairs.For instance, if God were continually to intervene to prevent suffer-ing, this would introduce a disorder into the world. It regularitywould be disrupted. In suggesting that God should do so, the atheistassumes that the suffering of non-moral agents is a greater defectthan such irregularity. But how does he know this?

10.2.2 Sceptical Scenarios

How does van Inwagen make use of these sceptical observations?Given both modal and moral scepticism, van Inwagen can offer a va-riety of theistic scenarios that would each account for evil. It may bethat each scenario needs to be not improbable given theism (al-though van Inwagen vacillates on this one), but the sceptical theistdoes not need to show that they are  probable. He need only showthat given these possible scenarios, we cannot be certain about ourability to make such judgements. Van Inwagen’s method is parallelto that of the lawyer who presents the jury with a range of possible

scenarios, each of which would exonerate his client. His aim wouldbe not to suggest that any of them is true, but to convince them thathis client cannot safely be convicted.

 Van Inwagen’s favoured theistic scenario involves two posits. (Infact, it involves three, but I’ll just examine two here.) The first isthat a world that contained less suffering would be massively irregu-lar, in the sense of having no consistently-applicable laws of nature.


Page 103: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 103/205

Evidential Arguments from Evil

We cannot say this is not the case (modal scepticism). The second is

that being massively irregular is a defect in a world that outweighsthe defect of containing this much suffering. We cannot say this isnot the case (moral scepticism). So we have reason not to accept theatheist’s claim that God could have created a world in which sentientbeings did not suffer (or suffered only as a result of their own wrongactions). Responding to Draper’s argument, van Inwagen denies thatwe can compare the probability of the suffering we see on the basis of HI with its probability on the basis of theism. We simply do notknow what its probability on the basis of theism is.

10.2.3 The Implications of Sceptical Theism

Now there is a lot to be said for scepticism. But there is also a lot tobe said against it, especially from the theist’s point of view. The firstkind of scepticism – that regarding goods and their connection withevils – is a double-edged sword. It may work as a defence, wardingoff the atheist, but it leaves the theist without any explanation forthe facts of evil (a theodicy). This is true of sceptical theism in all itsforms. It may succeed in fighting off  external challenges to theism

but at the cost of undermining any attempt to solve the problem of evil that is internal to theism. Of course, modal and moral scepticismalso offers an explanation for the existence of this internal problem.If we are unable to make reliable judgements about God and evil,then it is not surprising that the theist cannot adequately accountfor the existence of evil. But as Draper points out, the theist’s failurewould be even less surprising if theism were false, for the problemwould be revealed as a pseudo-problem.

One could also extend sceptical theistic arguments to the judge-

ments involved in the traditional proofs of the existence of God. VanInwagen himself criticises the ontological argument on thesegrounds (van Inwagen 1991: 152, 1998: 67). But the cosmological ar-gument is also vulnerable. It, too, relies on modal intuitions regard-ing the contingency of the world and the need for a necessary beingto undergird its existence. But if our intuitions about God and evilare unreliable, surely the same could be said of our intuitions re-


Page 104: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 104/205

Evidential Arguments from Evil

garding contingency and necessity? And does this scepticism not cast

doubt on our ability to affirm that God is the “cause” of all that ex-ists? (This is not an everyday use of “cause.”) If the theist wants toemploy any of these arguments – it may be that van Inwagen doesnot – there had better be some limits to her scepticism.

More seriously, one could argue that the sceptical theist is not be-ing consistent in his employment of these sceptical principles. Afterall, van Inwagen himself is offering what he judges to be a  possible

theistic scenario. But if you are true to your sceptical principles, youshould not be able to make this judgement. Here is another theistic

scenario: God would have liked to create a world without evil, but a

world without evil could contain no beauty and the value of beauty

outweighs that of evil. On the basis of the sceptical theist’s argu-ments, for all we know this too might be true. (What do we know of such deep metaphysical connections between beauty and evil?) If vanInwagen regards this as less probable given theism than the scenariohe offers, then he is allowing himself some degree of modal knowl-edge. So why should he deny similar modal knowledge to his oppo-nents?

Still more seriously, if it is true that we can make no judgementsregarding logical possibility, it is hard to know what meaning can beattributed to the theistic belief that God is omnipotent. Traditional-ly, this has been taken to mean that God can do anything that is log-ically possible. But if we can give no meaning to this latter phrase,what sense does it have to speak of divine omnipotence? And if wecould not judge how an omnipotent, omniscient and morally perfectbeing might be expected to act, then any kind of religious explana-tion becomes impossible. It may be that a consistent modal and

moral scepticism would lead to agnosticism (or perhaps to Flew’s“negative atheism”), unless, of course, you believe you are in posses-sion of a divine revelation. But then how would you know this? (SeePart Three.)


Page 105: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 105/205

Evidential Arguments from Evil

10.2.4 A Merely Ad Hoc Defence

So far, my criticism is merely that sceptical theism may have un-palatable consequences for its advocates. It does not show that itfails to do what it sets out to do, namely undermine the argumentfrom evil. However, there are other arguments against sceptical the-ism, which suggest that it does fail to undermine the atheist’s argu-ments. With regard to moral scepticism, for instance, Michael Tooleyhas suggested that there are judgements in this area that we cansafely make, even given our limited knowledge. For instance, we can judge that God would be justified in inflicting suffering on an indi-vidual only if that suffering were likely to benefit that same individ-ual. This is a relatively uncontroversial application of a widely-ac-cepted principle of justice (Tooley 1991: 112–13). So the atheist’s con-clusion that the suffering of non-moral agents is incompatible withthe existence of God, even given the free will defence, would seem jus-tified.

But I can offer a more general criticism, one that addresses itself to the general strategy of offering merely possible theistic scenarios

(i.e. defences). If you accept Draper’s way of setting up the evidentialargument from evil, then this strategy appears worthless. To explainwhat I mean, let me refer to two cases offered by van Inwagen asparallels to the sceptical theism. The first case is that of the ancientGreek atomist faced with objections to his claim that the air is madeup of tiny particles. The second is the modern searcher for extrater-restrial intelligent life faced with objections to her belief that suchlife exists. What sort of objections might each of these face? The an-cient atomist might be faced with the objection that if the air were

made of tiny particles, eventually it would all settle on the earth.Similarly, the searcher for extraterrestrial life (the “extraterrestrial-ist”) might be faced with the objection that if intelligent extraterres-trials existed, we would expect to have detected signals indicatingtheir existence.

How might the atomist or extraterrestrialist respond to these ob- jections? Van Inwagen argues that he could legitimately construct adefence, a scenario that might be true, for all anyone knows. The


Page 106: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 106/205

Evidential Arguments from Evil

atomist might speculate that his atoms are covered with long spikes

that keep them at a distance from one another. The extraterrestrial-ist might speculate that any intelligent extraterrestrial civilizationthat developed would destroy itself before being able to send out suchsignals. The more such defences each could produce, the better. Sincefor all we know (or, in the case of the atomism defence, for all an an-cient Greek knew), such things might be true, no judgement can bemade regarding the probability of what we observe on the basis of the given theory. It is therefore rational to maintain the belief.

But is it? Surely we should make such judgements on the basis of 

what we have reason to believe is likely to be true, not on the basis of mere  possibilities? All that such defences show is that we may bewrong. But this is not the issue. The atheist need not deny that his judgements are fallible. Van Inwagen’s suggestions would certainlybe unacceptable in assessing a scientific hypothesis. For what is headvocating? Nothing other than the elaboration of  ad hoc auxiliaryhypotheses to save a theory. He is suggesting that we extend the the-ory in ways that protect it from refutation, in the absence of any in-dependent evidence. The fact that van Inwagen is offering these de-

fences as mere possibilities only highlights the ad hoc nature of hisstrategy.

Now it may be that at times scientists do this. But most philoso-phers of science would regard it as a temporary expedient, defensibleonly if it lent support to a research programme that was proving suc-cessful in other respects. So the recourse to defences of the sort vanInwagen advocates might be defensible if theism were proving to bea fruitful explanatory hypothesis in comparison with its naturalisticrivals. But who would be prepared to argue that? Note, too, that the

atomism example is misleading, since we cannot help but view it inthe light of our later knowledge. Given his epistemic situation, theancient Greek atomist may have been acting irrationally in main-taining his theory, certainly if this involved such ad hoc devices. Thefact that we can see that both his theory and his ad hoc defence wereclose to the truth is neither here nor there. So by showing the worth-


Page 107: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 107/205

Evidential Arguments from Evil

lessness of such defences, van Inwagen’s analogies actually lend sup-

port to Draper’s argument.

10.2.5 The Reformed Epistemology Escape-Hatch

Why is van Inwagen apparently unconcerned by such criticisms? It isprobably because he does not accept Draper’s way of setting up theargument. He does not believe that the theist needs to argue for herposition. If so, this makes the theist’s task much easier. If theism isnot being defended as a hypothesis that must show its value by doingsome useful explanatory work, then my objections would be under-

cut. If the theist can claim to know her beliefs about God to be trueby some other means, perhaps by faith, or by way of some basic formof knowledge inaccessible to the atheist (who can be presumed to beblinded by sin or lacking divine grace), then how much simpler her job is. Then a defence would suffice to undercut the atheist’s argu-ments. I say this to highlight the attraction of the position known as“Reformed Epistemology,” to which I shall return in Part Three. Ishall argue at that point that this particular escape-hatch leadsnowhere.


Page 108: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 108/205

Page 109: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 109/205

Chapter Eleven

Atheism as a Moral Imperative

Let me now turn to a quite different response to the existence of evil,namely that of rebellion against God. I say “rebellion against God”because the rebel does not necessary deny God’s existence. What hedoes do is to affirm that if God were to exist, he would be morallynegligent, even wicked, to have created a world which contains such

evils. It follows, the rebel argues, that we have a moral obligation to

rebel against God. It is often assumed, in popular discourse, that“atheism” goes hand in hand with immorality. (Some dictionary defi-nitions of atheism preserve this meaning.) But what this responsehighlights is that those who rebel against God, may be motivated bymoral considerations. A good person, they argue, will reject theclaims of religion because those claims are morally abhorrent. Evenif there were reason to believe God existed, we should not submit tohis will.

11.1 Ivan’s Rebellion

The classic starting point for any discussion of what Peter Fosl(1997) calls “the moral imperative to rebel against God” is FyodorDostoevsky’s novel The Brothers Karamazov. A key moment in thenovel is a dialogue between the elder brother Ivan and the younger Alyosha (Part Two, Book Five, Chap. 4). In the course of this dia-logue, Ivan presents his devout brother, a young monk, with an al-most unbearable account of the suffering of children. (Incidentally if you want an updated version, which – at least if you are a parent – isalmost too painful to read, try the first couple of pages of Peter Fosl’sarticle.) Here’s an extract from Ivan’s words.

‘A little girl of five was abused by her parents, “decent and re-spectable people, well educated and cultured.” ... Those educatedparents subjected that poor little five-year-old to every conceivabletorture. They beat her, whipped her, kicked her till she was black


Page 110: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 110/205

Atheism as a Moral Imperative

and blue, all for no reason. Finally, they thought of the ultimate

punishment; they shut her up all night in the outside privy, in thecold and frost, because she wet herself at night (as if a five-year-old, sleeping soundly like an angel, could excuse herself in time) – for this, they smeared her face with excrement and forced her toeat it, and it was her mother, her mother who did this to her! Andthat mother slept unconcernedly at night, oblivious to the sobs of the poor child shut up in that foul place! Can you understand sucha thing: that small child, unable even to comprehend what is beingdone to her, in the dark and the cold of that foul place, beating her

little panting breast with her tiny fists, sobbing, weeping humbletears of bloodstained innocence, praying to “Dear Father God” toprotect her – do you understand this obscenity, my friend, mybrother, my holy and meek monk, do you understand why such anobscenity should be so necessary, and what is the point of it?’ (Dos-toevsky 1994: 303)

Ivan then offers a still more horrifying account of a child tornapart by dogs in front of his mother because he had offended a richlandowner. After running through a variety of answers to his ques-

tion – accounts of why such evils must exist for the sake of somegreater good – Ivan suggests that even if such theodicies could beshown to be true he would reject them.

‘Oh, Alyosha, I’m not blaspheming! I understand how the universewill shake when heaven and earth all unite in a single paean of praise, and all that lives and has lived will cry out, “You are just,O Lord, for your ways are revealed to us!” When the mother em-braces the murderer whose dogs tore her son apart, and all threeshall cry out weeping, “You are just, O Lord” – that, of course, will

be the summit of all knowledge, and all will be explained. Buthere’s the snag; that’s just what I can’t accept. ... You see, Alyosha,perhaps it really will happen like that, and I shall live to see it orbe resurrected, and then perhaps I too, seeing the mother embrac-ing the child’s torturer, will cry out in unison with them, “You are just, O Lord,” but it will be against my will. While there’s still timeI want to guard myself against this, and therefore I absolutely re- ject that higher harmony. It’s not worth one little tear from one


Page 111: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 111/205

Atheism as a Moral Imperative

single little tortured child, beating its breast with its little fists in

its foul-smelling lock-up, and praying with unexpiated tears to its“Dear Father God!” No, it’s not worth this ... if the suffering of children is required to make up the total suffering necessary to at-tain the truth, then I say here and now that no truth is worth theprice. ... Tell me honestly, I challenge you – answer me: imaginethat you are charged with building the edifice of human destiny,whose ultimate aim is to bring people happiness, to bring thempeace and contentment at last, but that in order to achieve this itis essential and unavoidable to torture just one little speck of cre-

ation, that same child beating her breasts with her little fists, andimagine that this edifice has to be erected on her unexpiated tears.Would you agree to be the architect under those conditions? Tellme honestly!’ ‘No, I wouldn’t agree,’ said Alyosha quietly. (Dosto-evsky 1994: 307–8)

The point is so powerfully made that further comment seems su-perfluous. Nonetheless, let me try to spell out Ivan’s view. Ivan is notan atheist. As he himself says, “it’s not that I don’t accept God, Alyosha; I’m just with the utmost respect, handing Him back my

ticket” (Dostoevsky 1994: 308). Rather, he is a rebel against God. Hedoes not deny that God exists; he does not even deny that God haspermitted evil for the sake of some greater good. He simply finds thisway of acting intolerable. He suggests that no person of any moralsensitivity could agree to act as God has apparently acted.

Ivan Karamazov’s response to evil was that of rebellion against aGod whose existence he acknowledges. Is this atheism? Well, it is notatheism in the sense of denying the existence of a supernatural cre-ator. But it is atheism in the sense of denying the existence of a

morally perfect creator. For to rebel against God on moral grounds isto deny that whatever god exists is worthy of worship. This, in turn,is equivalent to affirming that the Christian (Jewish, and Muslim)God, who by definition is perfectly good, does not exist. As MichaelTooley points out, the argument from evil for the non-existence of God may be converted at will into an argument from evil for thewickedness (or at least indifference) of God (Tooley 1991: 103–4). So


Page 112: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 112/205

Atheism as a Moral Imperative

perhaps these are not different arguments after all. However, what

is worth noting about Ivan’s protest is that it starts off preciselywhere the defences and theodicies end. It accepts that the defencesare successful and that the theodicies may be true. But then it asks,“O.K., but what does this tell us about God?” If you want to knowwhy so many atheists are passionate in their rejection of theodicies,this may be a good place to start.

11.2 Why Rebel?

Let’s try to spell out Ivan’s reasoning. Why did he feel compelled to

rebel against God? Are these good reasons? Could such an apparent-ly shocking attitude as rebellion – one that Christianity attributes tothe devil himself – be justified?

11.2.1 Ends and Means

One issue that apparently bothers Ivan has to do with ends andmeans. Most theodicies come perilously close to suggesting, if they donot in fact suggest, that God treats human beings as a means to anend. Either he is prepared to sacrifice the happiness of individuals

for the alleged greater good of some abstract ideal, such as the exis-tence of human freedom, or he is prepared to sacrifice the happinessof some individuals for the sake of the happiness of others. Some the-ist philosophers are prepared to bite this particular bullet and justifysuch a course of action, Peter van Inwagen being among them (seeLevine 2004: 160). But many atheists would argue that it would notbe morally admirable for God to act in this way. No moral agent – “not even ... God” (Kant 1997: 109 [5.131]) – should treat human be-ings simply as means to an end.

While Kant, for instance, restricted this principle to rational crea-tures, one could strengthen the rebel’s arguments by extending it tonon-human animals. Free will and moral growth – the alleged“greater goods” posited by theodicies – are goods only for those whocan enjoy them. All but one species of sentient being apparently can-not. So if God were to permit animal suffering for the sake of a goodthat only humans can enjoy, this could easily be seen as a morally


Page 113: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 113/205

Atheism as a Moral Imperative

reprehensible choice (Tooley 1991: 111). “Perhaps,” the theist may

respond, “this was unavoidable. Animal suffering is the inevitablecost of human freedom.” Well, then, it would have been better forGod not to create a world at all than to create one in which suchgoods could be achieved only at such cost.

There do, however, exist ethical theories that would not take asstrong a position as Kant does on this issue. Some forms of utilitari-anism, for instance, hold that we might be justified in killing or caus-ing suffering to one or a few innocent people in order to save a largernumber from suffering or death. (This seems to be the intuition be-

hind the common “trolley” scenarios, in which people are asked if they would be prepared to divert a runaway trolley in such a waythat it would kill one person instead of five.) Many ethicists wouldsay that, in the tragic situation in which these are the only two alter-natives, this would be the correct action to take. One might, howev-er, wonder about why an omnipotent God would be faced with sotragic a choice, but this, perhaps, takes us back to the free will de-fence.

11.2.3 Justifying Individual Evils

There is another problem with theistic responses from evil which – while certainly not unavoidable – constitutes a trap into which thetheist can easily fall. Traditional responses to the argument fromevil, such as the free will defence, argue that there exists somegreater good to permit which God must tolerate evil. Even if thissuggestion is not morally problematic in itself, as Ivan suggests it is,it easily gives rise to another idea which does seem problematic. Thisis the idea that for every individual instance of evil in the world,

there exists some greater good which justifies it.The temptation of making this suggestion is one to which no less a

thinker than Richard Swinburne has succumbed.

It may be urged that, despite the good ends that its actual or possibleoccurrence serves, there is too much evil in the world... With the ob- jection that, if there is a God, he has overdone it, I feel considerableinitial sympathy. The objection seems to count against the claim that


Page 114: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 114/205

Atheism as a Moral Imperative

there is a God. But then I reflect that each bad state or possible bad

state eliminated eliminates one actual good... Suppose that one lessperson had been burnt by the Hiroshima atomic bomb. Then therewould have been less opportunity for courage and sympathy; one lesspiece of information about the effects of atomic radiation, less people(relatives of the people burnt) who would have had a strong desire tocampaign for nuclear disarmament and against imperialist expan-sion. And so on. (Swinburne 2004: 263–64)

To suggestions of this kind, Michael Levine reacts with horror, ar-guing that they are morally repugnant.

The fact that such views have not drawn harsh criticism from con-temporary Christian analytic philosophers of religion indicates both alack of interest by the wider philosophical community and also howinbred contemporary Christian philosophy of religion, and much of the criticism directed at it, has become. If van Inwagen and Swin-burne were political figures, there would be protesters on the street. Imean this literally and not polemically. After all, what they havedone is to offer not just a  prima facie, but an ultimate justification forthe holocaust and other horrors. (Levine 2000: 107)

It seems hard to disagree. If each case of horrific injury as a result of nuclear weapons brings about some greater good, which more thancompensates for it, then it’s hard to see why we should seek to pre-vent the injury. Indeed, perhaps we would even be justified in inflict-ing it. But if this is the logical outcome of Swinburne’s position,something has gone very wrong indeed.

11.2.3 The Silence of God

There is a third issue that is troubling about theistic defences andtheodicies. It is the apparent silence of God in the face of suffering.Theists often compare God to a loving parent. They point out thateven the most loving parent may permit her child to suffer if that isnecessary for the child’s welfare. But what parent would do so with-out making it clear to the child why this must be done? And whatparent worthy of the name would not stay with the child during hersuffering, offering constant affirmations of her love and concern? Butas William Rowe notes in a recent interview, this is not the case with


Page 115: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 115/205

Atheism as a Moral Imperative

God (if he exists). “Countless numbers of human beings go through

periods of horrendous suffering without any awareness at all of God’sbeing consciously present to them” (Rowe 2004: 18). The silence of God in the face of suffering suggests that if there is a God, he lacksthe qualities we would expect of any human parent. What parentwhose child was suffering would abandon the child and go on holi-day? But as Rowe notes, “it is as though God has been on holiday forcenturies ... while his creatures undergo extensive suffering andpainful deaths” (Rowe 2004: 18).

11.3 “God’s Ways Are Not Our Ways”

One common theistic response to such arguments is to suggest that“God’s ways are not our ways”: the divine qualities of justice andmercy are inscrutable. We cannot convict God of injustice or indiffer-ence, for whatever such terms mean when applied to God, they can-not mean what they would mean in ordinary use. This is anotherkind of sceptical theism, but one that is sceptical regarding what wecan affirm of God. It is true that by our standards God may appearwicked or indifferent. But God’s standards, it is suggested, are not

our standards. That this is a common response is indicated by aGoogle web search I performed on 26 April 2006, for the biblicalphrase “my thoughts are not your thoughts, neither are your waysmy ways” (Isa 55:8). It was found on 52,600 web pages. The mostcommonly accessed of these pages used this phrase to explain whyGod does not act as we might expect a loving God to act. Why, for in-stance, did he allow Hurricane Katrina to cause so much damage in August 2005? Why are my prayers apparently not answered? Whydoes God allow innocent people to suffer? The answer to all thesequestions? It is because “his ways are not our ways.”

The best response to this suggestion was offered some 150 yearsago by John Stuart Mill in his review of a book by H. L. Mansel(1820–71) on the limits of reason in matters of religion.

If in ascribing goodness to God, I do not mean what I mean bygoodness; if I do not mean the goodness of which I have someknowledge, but an incomprehensible attribute of an incomprehen-


Page 116: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 116/205

Atheism as a Moral Imperative

sible substance, which for aught I know may be a totally different

quality from that I love and venerate – and even must ... be insome important particulars opposed to this – what do I mean bycalling it goodness? and what reason have I for venerating it? ... Tosay that God’s goodness may be different in kind from man’s good-ness, what is it but saying, with a slight change of phraseology,that God may possibly not be good? (Mill 1979: 102)

It is true that if there is a God, then by the very nature of the casehe will be a mysterious being, whom we cannot hope fully to under-stand. But what bothers Mill is that theists appeal to the mysteri-

ousness of God in order to attribute behaviour to God that in a merehuman being we would consider reprehensible. As he writes, “thatwe cannot understand God; that his ways are not our ways; that wecannot scrutinize or judge his counsels ... have often ... been tenderedas reasons why we may assert any absurdities and any moral mon-strosities concerning God, and miscall them Goodness and Wisdom”(Mill 1979: 90). After criticising such views, Mill concludes with amarvellous rhetorical flourish.

If, instead of the “glad tidings” that there exists a Being in whomall the excellences which the highest human mind can conceive,exist in a degree inconceivable to us, I am informed that the worldis ruled by a being whose attributes are infinite, but what they arewe cannot learn, nor what are the principles of his government,except that “the highest human morality which we are capable of conceiving” does not sanction them; convince me of it, and I shallbear my fate as I may. But when I am told that I must believethis, and at the same time call this being by the names which ex-press and affirm the highest human morality, I say in plain terms

that I shall not. Whatever power such a being may have over me,there is one thing which he shall not do: he shall not compel me toworship him. I shall call no being good, who is not what I meanwhen I apply that epithet to my fellow creatures; and if such a be-ing can sentence me to hell for not so calling him, to hell I will go.(Mill 1979: 103)


Page 117: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 117/205

Atheism as a Moral Imperative

Mill’s point is crystal clear. If you can salvage the goodness of God

only by redefining “goodness” so that the word no longer has its usu-al meaning, you are not speaking in a straightforward manner. AsMill writes, “to assert in words what we do not think in meaning, isas suitable a definition as can be given of a moral falsehood” (Mill1979: 102).


Page 118: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 118/205

Page 119: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 119/205

Chapter Twelve

Faith and Reason: An Evidentialist View

So far, I have been examining the reasons that can be offered in sup-port of religious belief and the reasons that can be offered against be-lief. But many people feel instinctively that such appeals to reasonsare beside the point. Religious faith, they feel, is not based on the ar-guments that can be offered for or against belief. It is not like a sci-entific hypothesis, a theory about the world for which evidence can

be produced. Religion, they say, rests on faith, rather than reason. Itdoesn’t need evidence in its support.

Let’s deal with one aspect of this claim straightaway. Religiousfaith, as traditionally understood, has various dimensions. One of these dimensions is that of trust: the believer has faith in God in thesense of trusting in God. (There is a similar sense in which I can besaid to have faith in my partner: I trust her.) And this element islacking in the case of a scientific hypothesis. The scientist does nottrust in electrons in the sense in which the believer trusts in God.

But of course one cannot trust in God unless one believes that Godexists. So religious faith also has an unavoidably cognitive dimen-sion: it involves a claim to knowledge. The question to be asked hereis: Does it make any sense to make a claim to knowledge in the ab-sence of sufficient evidence that what you believe is true? More gen-erally, what is the relationship between religious faith and reason?

What I want to discuss today is a view of the relation betweenfaith and reason that many philosophers find attractive. I shall de-scribe it as an “evidentialist” view of religious faith, which we can

find in the work of the seventeenth-century philosopher John Locke.Put briefly, it is the view that while religious belief goes beyond whatcan be demonstrated by reason, it does so legitimately only when itis supported by reason. A useful starting point will be Locke’s dis-tinction between faith and reason, which has to do with the source of what is believed.


Page 120: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 120/205

Faith and Reason: An Evidentialist View

Reason, therefore, here, as contradistinguished to faith, I take to be

the discovery of the certainty or probability of such propositions ortruths which the mind arrives at by deduction made from such ideas,which it has got by the use of its natural faculties; viz. by sensation orreflection. Faith, on the other side, is the assent to any proposition,not thus made out by the deductions of reason, but upon the credit of the proposer, as coming from God, in some extraordinary way of com-munication. This way of discovering truths to men, we call revelation.(Locke 1846: 4.18.2)

This is, as it happens, a traditional distinction. When we say that

we know a proposition p on the basis of reason, it is because we cangrasp the evidence that leads us to believe that p is true. But thereare other things we can know to be true on the basis not of reasonbut of authority. To take my earlier example, I may not be able to of-fer you the evidence that e=mc2 (e being energy, m mass, and c thespeed of light), because I don’t know sufficient physics to do so. But Ican still believe it to be true on the authority of those scientists whounderstand the theory on which it is based. Religious faith, saysLocke, involves a similar belief in authority, the difference being that

in this case the authority is that of God. And on divine authority wecan believe things for which we have no direct evidence; their beingdivinely revealed is indirect evidence of their truth.

Of course, if we knew that something had been revealed to us byGod, this would be an excellent reason for believing it, if we also hadreason to believe that God was omniscient and morally perfect. (Sucha being could be neither deceiving nor deceived.) And if we knewwith certainty that something was revealed by God, we would alsoknow with certainty that it was true, even if we could offer no other

arguments in its support. If we know with certainty that the Bible,for instance, or the Qur’an was revealed by God, then we would knowwith equal confidence that whatever it says must be true. But, saysLocke, we need reason to believe that the Bible (or, we might add,the Qur’an) is revealed by God. And our confidence in the truth of what is revealed can never be greater than our confidence both thatit has been divinely revealed and that we have understood that reve-


Page 121: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 121/205

Faith and Reason: An Evidentialist View

lation correctly. As Locke writes, when we believe something on the

basis of an assumed revelation

we must be sure that it be a divine revelation, and that we under-stand it right: else we shall expose ourselves to all the extravagan-cy of enthusiasm, and all the error of wrong principles, if we havefaith and assurance in what is not divine revelation. And there-fore... our assent can be rationally no higher than the evidence of its being a revelation, and that this is the meaning of the expres-sions it is delivered in. If the evidence of its being a revelation, orthat this is its true sense, be only on probable proofs, our assent

can reach no higher than an assurance or diffidence, arising fromthe more or less apparent probability of the proofs. (Locke 1846:4.16.14)

So while faith goes beyond reason, in that it tells us (on the au-thority of God) of things we could never know otherwise, it also restson reason, in that we must have reason to believe that God has spo-ken and that we have understood what he has said. Analogously,while I may never understand the physics behind Einstein’s famousequation, I can believe it to be true, but that belief is legitimate only

if I have reason to trust those authorities who have told me it is true.What is interesting is that this view undercuts, in a certain sense,

the very distinction from which it begins, namely that between faithand reason. There is still a distinction here, a distinction between be-lieving those propositions for which I have direct evidence and be-lieving those propositions for which I have only indirect evidence,namely the authority of the person telling me they are true. But if we should believe what such a person tells us only when we havereason to believe she is reliable, then of course we are believing onthe basis of reason: the reasons we have for believing she is reliable. And this means that all the beliefs to which we are entitled are be-lieved on the basis of reason. (This is why I call such a view “eviden-tialist.”) Locke himself seems to accept this, writing that while hetreats of faith “as it is ordinarily placed, in contradistinction to rea-son,” it fact it is nothing other than “an assent founded on the high-est reason” (Locke 1846: 4.16.14). To put it more concisely, on Locke’s


Page 122: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 122/205

Faith and Reason: An Evidentialist View

view “reason must be our last judge and guide in everything” (Locke

1846: 4.19.14). Are you surprised this is a popular view amongphilosophers?


Page 123: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 123/205

Chapter Thirteen

Faith and Self-Authentication

I have just examined John Locke’s “evidentialist” view of religiousfaith. While it might appear very reasonable, it is not the dominantview of faith within the history of Christian thought. What I want tolook at now is the dominant view. I shall call it the “Aquinas-Calvin”view because variants of it are to be found in the work of the late me-

dieval Catholic theologian Thomas Aquinas (ca. 1225–74) and in thewritings of the sixteenth-century Protestant Reformer John Calvin(1509–64). According to the Aquinas–Calvin view, religious faith is aform of knowledge by testimony, the testimony in question beingthat of God himself. This is, of course, something that John Lockealso accepts. What makes the Aquinas-Calvin view different fromLocke’s “evidentialist” view is a further belief: the belief that this di-vine testimony is thought to provide its own warrant. It is, if youlike, “self-authenticating.”

13.1 The Aquinas-Calvin View

Let’s look more closely at what this entails. According to the Aquinas-Calvin view, there exists a clear distinction between whatwe might call religious faith and simple belief . Religious faith is not

based on rational insight into the truth of the matters revealed. In-deed on the traditional Roman Catholic definition of faith, put for-ward at the first Vatican Council (1870), this is expressly excluded.

The Catholic Church declare faith to be a supernatural virtue bymeans of which we believe those things that have been revealed byhim to be true, not on the basis of the intrinsic truth of the matter

seen by the natural light of reason, but on the authority of God

himself revealing , who can neither deceive nor be deceived. (DS3008; emphasis mine)


Page 124: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 124/205

Faith and Self-Authentication

Religious faith, on this view, involves accepting something on the

authority of God himself. If reason is given any role here, it is largelyrestricted to providing ex post facto (retrospective) evidence that thepropositions believed do indeed come from God and are therefore tobe accepted. More precisely, the role of reason is only to prepare theway for faith and to lend support to an act of faith already given(Aquinas, SCG 1:6) or a disposition of faith that already exists(Calvin, ICR 1.8). Arguments for the existence of God or the fact of revelation are not the basis of faith; they are merely (to use a tradi-tional term) the  praeambula fidei, the “preliminaries of faith” or

(better still) the “presuppositions of faith” (McGrath 1972: 135).The distinction is already clear in the work of Aquinas. Aquinas

argues that we do have arguments in support of the belief that it isGod who is revealing these things and that God is trustworthy. Butfor Aquinas such rational considerations are not what motivates theact of faith, properly so-called. In fact, Aquinas argues that if a per-son’s faith were based upon such rational considerations, it wouldlose its merit. For Aquinas, the virtue of faith is to be clearly distin-guished from what we might call “simple belief,” even if that belief is

a belief in divine revelation. Religious faith involves a rightly-direct-ed will rather than merely an enlightened intellect (ST 2a 2ae 2.10).One who has faith in the proper, theological sense accepts what Godhas said because of her love for God. Aquinas concedes that accord-ing to the New Testament the devils themselves “believe and trem-ble” (James 2:19). So apparently the devils, too, have faith, in somesense of that word. But the faith they have is (as it were) “simple be-lief.” It lacks the rightly-directed will, the willing acceptance of whatGod has revealed out of love for God, that is characteristic of reli-

gious faith (ST 2a 2ae 5.2).What kind of knowledge is the act of faith thought to yield? Well,

if you were to accept something on the authority of God, you woulddo so with a degree of confidence greater than that which any otherkind of knowledge can offer. After all, what more trustworthy au-thority could you ask for? Such a belief would be more certain thanany of our other beliefs, even if what is believed remains puzzling or


Page 125: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 125/205

Faith and Self-Authentication

mysterious. This is another feature of the traditional view of faith

that may seem odd at first sight. The believer can quite consistentlysay: “I am certain that p is true” –  p being some proposition thoughtto be divinely revealed – “although I am not sure what  p means.” In-deed Roman Catholic theologians often spoke of what they called“the obscurity of faith.” Since what is revealed exceeds the power of human reason, it remains only partially intelligible, even though itstruth is certain. This attitude becomes intelligible (even if not defen-sible) if one remembers that “faith” in this context is equivalent to“belief on the basis of divine testimony.”

If the act of faith is thought to yield certainty about what is be-lieved, then what about the reasons (if there are thought to be any)that support the act of faith? What kind of knowledge are theythought to yield? Do they yield the same certainty? Apparently not.Even Aquinas seems to accept that reason cannot yield decisive evi-dence of the fact of revelation. Have a look at the following passagefrom Aquinas.

 As regards ... man’s assent to the things which are of faith, wemay observe a twofold cause, either of one of external inducement,such as seeing a miracle, or being persuaded by someone to em-brace the faith. Neither of these is a sufficient cause, since of thosewho see the same miracle, or hear the same sermon, some believe,and some do not. Hence we must assert another internal cause,which moves man inwardly to assent to matters of faith. Thereforefaith, as regards the assent which is the chief act of faith, is fromGod moving man inwardly by grace. (ST 2a 2ae 6.1)

To translate Aquinas into contemporary language, the evidencethat can be offered in support of faith is such that not everyone whohears it is convinced by it. Something else must therefore be needed,namely the grace of God, which produces the religious act of faith.Calvin, too, seems to admit that what he describes as the “externalevidences” for the authority of Scripture fall short of proving thecase, in the sense of excluding the possibility of doubt (ICR 1.8.1). In-terestingly enough, more recent Roman Catholic authors, bound byChurch teaching to highlight the role of reason, do speak of certainty


Page 126: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 126/205

Faith and Self-Authentication

in this context. Rational reflection, they argue, can lead to certainty

regarding the existence of God and the fact of divine revelation. Butthey are inclined to qualify the claim. One textbook, for instance, in-sists that it is “moral certainty,” which excludes all “prudent doubt,”and that it may arise from a coming-together of arguments that in-dividually are merely probable (Tanquerey 1921 II §168 [97–98]).

13.2The Bootstrapping Problem

 You will already have noticed something odd about this traditionalview of faith. The reasons that might lead one to faith are apparentlyless than 100% convincing. But faith nonetheless is thought to giverise to a certainty that precludes any possibility of doubt. Apparent-ly, the believer is here claiming a certainty for faith that goes beyondthe confidence that reason can provide in support of its claims. Butas Locke points out, this view seems indefensible. If faith dependedentirely on the reasons that could be offered in its support, then itcould not have the degree of certainty claimed. The entire structureof religious faith, with its bold claims regarding authority and cer-tainty, appears to be built on sand.

The believer might reply that the certainty of faith stems from asource that is (in principle) independent of such reasons, namely theauthority of God. The believer is certain about her faith, not on ac-count of the reasons that could be offered in its support, but becausewhat she believes comes to her with the authority of God himself.But this merely shifts the question back one step. If we ask the be-liever, “O.K. You’re certain of this because you believe God has re-vealed it. But what makes you think God has revealed it?”, he has (itseems to me) only two possible replies. He could cite the argumentsthat lend support to the belief that this is a divine revelation. Butthis would mean abandoning the claimed certainty of faith and re-placing it by some greater or lesser degree of probability. It wouldmean abandoning the idea that faith is a mode of knowledge some-how distinct from reason. (As we’ve seen, this is Locke’s option.)Since believers are often reluctant to do this, their more common re-sponse is to “bootstrap” the believer’s sense of certainty: to base the


Page 127: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 127/205

Faith and Self-Authentication

certainty of his belief on the very revelation in which he believes.

The believer’s reply would then resemble the following:

I believe in the doctrine of creation [e.g.] on the authority of God.

How do I know that God has revealed it? On the authority of God.

On this view, religious faith is regarded as at least partially “self-au-thenticating.” I say “at least partially self-authenticating” becausesome versions of this view allow that reason offers partial, indepen-dent support to the act of faith. But as we have seen, even this viewinsists that a faith dependent on reason is not properly speaking re-

ligious faith. Religious faith believes certain propositions on the au-thority of God on the authority of God. (This is not a typographicalerror.) The authority of God is simultaneously that which (id quod)and that by virtue of which (id quo) one believes (Hervé 1935 III§321 [354]).

The circularity of this view might seem to be the Achilles’ heel,not just of the Protestant system, as David Friedrich Strauss sug-gested in 1840, but of this traditional, “bootstrapping” view of faithin whatever form it is expressed. But of course this fairly obvious dif-

ficulty has not escaped the theologians. Theologians in the traditionof Aquinas do offer what they see as a solution. Their solution is tosay that it is by one and the same act that one believes in the propo-sition revealed and the fact of revelation (Hervé 1935 III §318 [350];Garrigou-Lagrange 1965: 74). I myself can’t see how this resolves theproblem. Perhaps what is being claimed is that there is no discursiveargument (and therefore no circular argument) involved in the act of faith. But as a defence this would not work, for the circularity is in-herent in the nature of the claim. Whether or not the believer en-

gages in a circular process of argumentation, she is assuming the re-liability of her alleged belief-forming mechanism and it is the relia-bility of that mechanism that needs to be demonstrated.

 As we shall see shortly, there exists a recent variant on this tradi-tional, “bootstrapping” view of faith, namely the “Reformed Episte-mology” of Alvin Plantinga. I shall argue shortly that this is hope-less, that Plantinga’s thoroughly externalist view of knowledge sim-


Page 128: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 128/205

Faith and Self-Authentication

ply misses the point. But before I defend this cavalier dismissal of 

Plantinga’s work, I need to address some broader epistemologicalquestions, about so-called “basic beliefs.” And I shall do so by lookingagain at the question of “evidentialism.”


Page 129: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 129/205

Chapter Fourteen

Evidentialism Revisited

The major divide among philosophers of religion today – other thanthat between theists and non-theists – is that between evidentialists

and fideists. I’ve already described John Locke’s evidentialist view of religious faith: it suggests that we are entitled to hold a belief only if we have ( at least what would appear to us to be) adequate reasons

for that belief. Fideists, by way of contrast, claim that we can be enti-tled to hold certain beliefs even the absence of such evidential sup-port. Alvin Plantinga’s Reformed Epistemology is the clearest exam-ple of a contemporary philosophical fideism, but William Alston’s ap-peal to non-inferential beliefs arising from religious experiencewould also qualify. (Alston holds that we are entitled to act on ourreligious beliefs, even though we cannot offer a non-circular argu-ment in support of the reliability of the mechanisms that give rise tothem.) I’ll be examining these views shortly; for the moment let’s

look more closely at evidentialism.I am understanding evidentialism to be a claim about how an in-dividual ought to behave in matters of belief. It has to do with how anindividual ought to behave because if we do have obligations in mat-ters of belief, then what they entail will be person and context-de-pendent. Let’s say, with David Hume, that we each have a duty to“proportion our belief to the evidence” (cf. Hume 1902 §87 [110]).This may be a common human obligation, but what it entails for youmay not be the same as what it entails for me, and what it entails for

me may differ from occasion to occasion. There are at least two rea-sons for this. First, different individuals have different levels of abili-ty and therefore differing capacities to examine the evidence. Second,the consequences of acting on a false belief can vary from person toperson and from occasion to occasion. If those consequences of beingwrong are serious – in particular, if they could cause hardship or suf-fering to others – then this imposes a correspondingly serious obliga-


Page 130: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 130/205

Evidentialism Revisited

tion. And the more serious the consequences of our being wrong, the

longer and more careful that process of enquiry ought to be.

14.1 Obligations in Matters of Belief

There are those, however, who question whether it makes any senseto speak of “obligations in matters of belief.” The first worry is thatsuch an idea could be understood as assuming a libertarian view of freedom: the idea that when we act freely, then whatever we do, wecould have done other than we actually did. This is a controversialview of freedom, although talk of having obligations in matters of be-lief may still make sense even given a weaker, compatibilist view.What about determinism: the view that human beings are free inneither of these senses? On this view, it may not make sense to talkabout moral obligation, but the determinist could still argue that cer-tain actions in matters of belief are desirable. He could hold, for in-stance, that those actions are desirable that are most likely to lead toknowledge: weighing the evidence, making careful judgements, andso on. So even a determinist can make deontological judgements –  judgements about how one ought to act – although they are judge-ments about the moral value of actions rather than the moral re-sponsibility of individuals.

But a question remains. In what sense is belief an action? Therehas been much discussion of the position known as belief volun-

tarism, which suggests that we can choose what we believe. Thisidea seems to be mistaken. Beliefs are not, and cannot be, directlychosen. We typically form beliefs spontaneously as a result of follow-ing an argument or having a certain kind of experience. But if this istrue, what sense does it make to speak of our having obligations inmatters of belief? One could respond to this observation in a coupleof ways. One is to focus on the processes by which we form or examine

our beliefs, which do seem to be under our control, even if the beliefsthemselves are not. We cannot, it seems, choose directly to believe  p,but we can choose to examine the evidence and the arguments forand against p. So even if we form our beliefs spontaneously, we canchoose to critically examine them. All this is true. But it seems to me


Page 131: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 131/205

Evidentialism Revisited

that talk of obligations makes most sense when it refers to acting on

our beliefs, in the sense of taking them as premises in our practicalreasoning. When we act on a belief, we not only hold a proposition tobe true, but we take it to be true in deciding how we are to behave.What are our obligations in this respect? This is the question towhich I shall understand evidentialism to be an answer.

We can, however, ask what kind of obligations these are. One canthink of them as ethical obligations. If I am a doctor, for instance, Ihave an ethical obligation to my patients to pay attention to theavailable evidence before prescribing a course of drugs. But obliga-

tions in matters of belief do not always have such a clear ethical di-mension. So it might better to think of the obligation in question – the “ought” if you like – as of a different kind. Richard Feldman, forinstance (2000: 676), suggests it is best regarded as a “role ought.”(“If you are a competent pianist, you ought to be able to play Beeth-oven’s Moonlight Sonata.”) On this view, there are ways in whichyou ought to act, as a knower. Alternatively, one could regard it asthe kind of “ought” that is characteristic of  practical reason. (“If youwant to become a competent pianist, you ought to practise daily.”)

On this view, if you want to obtain knowledge, this is how you oughtto act in order to attain your goal. But I don’t need to settle this par-ticular discussion here.

14.2 Evidentialism Defined

It has been customary for philosophers discussing this question tospeak about justified beliefs or (better still) justified acts of believing.But the use of the term “justification” in this context is potentiallymisleading. It risks conflating the question of epistemic obligation(how we ought to act) with that of evidential support (whether cer-tain evidence does, in fact, support a belief). These issues are bestkept distinct; I shall discuss how they relate to one another in a mo-ment. For this reason, I prefer to focus on the notion of doxastic enti-

tlement. On my preferred definition of evidentialism,

a person is entitled to act on a belief if (and only if) (a) she has doneall that she is obliged to do in order to discover the truth of the mat-


Page 132: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 132/205

Evidentialism Revisited

ter, and (b) there are matters of which she is aware that she would,

on reflection, consider to be adequate grounds for her belief.

 A few comments should be made about this definition. First of all, itfocuses on acceptance, rather than merely belief, acceptance beingunderstood as the act of taking a belief as a premise in one’s practi-cal reasoning (Cohen 98: 368) Roughly speaking, when accept a be-lief when we choose to act on it. Secondly, it is what is sometimescalled an internalist condition: the matters that are the basis of theperson’s belief must be matters of which she is aware. Finally, the“on reflection” clause is to leave open the possibility of  tacit confir-

mation (see below): the idea that we can have grounds for our belief  – considerations that have actually brought about or confirmed ourbelief – on whose evidential force in this context we have never re-flected.

How does the question of entitlement relate to that of evidentialsupport? A remark made by Alvin Plantinga might help clarify therelation. Plantinga discusses what he calls the de jure question re-garding belief (“are you within your rights in believing what youdo”). This corresponds to my question of entitlement. Plantinga refers

to this as a “deonotological” question, which is fine: deontologicalquestions have to do with how we ought to act. But he, too, risks con-fusing the issues by using the term “justification.”

Justification needn’t detain us for long. There should be littledoubt that Christian belief can be and probably is (deontologically) justified, and justified even for one well acquainted with Enlight-enment and post-modern demurrers. If your belief is a result of the inward instigation of the Holy Spirit, it may seem obviouslytrue, even after reflection on the various sorts of objections thathave been offered. Clearly, one is then violating no intellectualobligations in accepting it. No doubt there are intellectual obliga-tions and duties in the neighbourhood; when you note that othersdisagree with you, for example, perhaps there is a duty to pay at-tention to them and to their objections, a duty to think again, re-flect more deeply, consult others, look for and consider other possi-ble defeaters. If you have done these things and still find the belief 


Page 133: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 133/205

Evidentialism Revisited

utterly compelling, however, you are not violating duty or obliga-

tion – especially if it seems to you, after reflection, that the teach-ing in question comes from God himself. (Plantinga 2000: 252–53;see also 101–2)

To put this in my terms, what Plantinga is saying is that such Chris-tians are entitled to act on their beliefs.

Now Plantinga may be right. Many Christians may be entitled tohold and act on their religious beliefs. They would be, on my defini-tion, if (a) they have done all that they are obliged to do in order todiscover the truth of the matter and (b) they have what would (on re-

flection) seem to them adequate reason for their beliefs. But whilethis is an important observation, it leaves some vital questions unan-swered. In particular, it doesn’t follow from the fact of an individual’sentitlement either that her beliefs are true or that the reasons thatseem to her adequate reasons are in fact adequate. The mere fact of entitlement doesn’t tell us a lot.

Having noted that a Christian might well be entitled to her belief,Plantinga’s response is to try to shift the discussion to the questionof whether Christian belief is true. What he argues, as we shall see

shortly, is that if it is true, then it is warranted. (By “warranted”here he means produced by a reliable causal mechanism.) Now itmay be the case that if Christian belief is true then it is warranted.But we can still ask whether the available evidence supports theclaim that it is true. Plantinga fails to answer this question. Hepoints to all kinds of considerations and experiences that a Christianmight understand to support her belief. Fair enough. These maymake for entitlement. But the key question then is: Should she un-derstand these considerations and experiences to support her belief?

14.3 Three Questions

What does all this mean? It means that there are three questionshere that need to be distinguished. Applied to Christian faith, theyare as follows.

(1) Is a particular Christian entitled to act on her Christian belief?


Page 134: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 134/205

Evidentialism Revisited

This is Plantinga’s de jure question. As the form of the question sug-

gests, it permits of no general answer. The answer we give will de-pend on the obligations, abilities, and situation of the subject. Oneperson might be entitled to act on her Christian belief because shehas been told that these beliefs are true by people whom she thinksare reliable authorities. That may be the most that this particularperson can be expected to do. She may have neither the ability northe time nor the need to do anything else. But it would not be suffi-cient for a highly educated person who has, for instance, the respon-sibility of being a teacher. Such a person ought to investigate the

truth of these beliefs for herself.There exists, however, a second question. It has to do not with the

situation of the individual believer, but with the objective relationbetween the available evidence and the proposition believed.

(2) Do we have, as a matter of fact, adequate reasons to take what

Christians believe to be true?

This is the question that most courses in the philosophy of religionaddress. The “adequate reasons” can be of various kinds. We could,

for instance, have good reason to believe some other propositionsfrom which Christian beliefs could be derived. (Many cosmologicalarguments, as we have seen, are of this kind: they are deductive ar-guments.) Or it may be that Christian beliefs are the best availableexplanation of some fact, or range of facts, about the world. (Somedesign arguments, for instance, are best thought of as “inferences tothe best explanation.”) Or we may discover that Christian beliefs arebrought about by a reliable mechanism, one that normally producestrue beliefs. But however one conceives of this evidential support, it

is an objective matter, which does not vary from one individual to an-other. I shall refer to this as the de justificatione question (regarding justification), which I distinguish sharply from the de jure question(regarding entitlement).

The two questions, namely (1) and (2) – the de jure question of en-titlement and the de justificatione, epistemic question of justification – are easily confused, since they are closely related. After all, one


Page 135: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 135/205

Evidentialism Revisited

might readily agree that a person is entitled to act on a belief if she

has done all she is obliged to do to investigate its truth and it seemsto her that she has sufficient reason to accept it. But this leads natu-rally to the de justificatione question, which has to do with whethershe is correct. When is a belief (in fact) supported by the evidenceand does this particular belief enjoy such support?

 Above and beyond these two questions, there exists a third ques-tion, which is:

(3) Are Christian beliefs true?

This is what Plantinga calls the de facto question. While Plantingaseems to regard it as the most important question – since if Chris-tianity is true, it is warranted – it seems the least helpful of thethree questions. How could we answer it, except by looking for thereasons we have to regard these beliefs as true? So in practice ques-tion (3) is indistinguishable from question (2), the de justificatione

question. To ask if Christian beliefs are true is (in practice) to ask if we have sufficient reason to regard them as true. And to this ques-tion, as we shall see, Plantinga’s talk about “warrant” offers no an-

swer.It follows that the most interesting question in the philosophy of 

religion is undoubtedly the traditional one, namely question (2),which is the de justificatione question. What should a person consid-er to be adequate reasons for holding a religious belief? What kind of evidential support should she be seeking? And do those reasons ex-ist?

14.4 The Challenge to Evidentialism

Let’s come back to evidentialism, which rests on the idea that wehave some kind of obligation in matters of belief, perhaps the obliga-tion to ensure that the beliefs on which we act really are supportedby the evidence. A person is entitled to act on her belief when shehas met her obligations. What would it take to undermine the claimthat we have obligations of this kind? What it would need is a coun-


Page 136: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 136/205

Evidentialism Revisited

terexample. We would need to find a situation in which two condi-

tions apply:(a) people are acting on a belief that seems to lack evidential sup-

port, and

(b) not even an evidentialist could say that they acting irresponsi-

bly in doing so.

The best candidate for an apparent counterexample of this kind isour belief in the reliability of sense perception (BRSP). It is true thatthis is not so much a conscious belief, as an assumption on which we

act every day. But I am assuming that even such an assumptioncounts as a belief.

When discussing what they call “basic” or non-inferential beliefs,philosophers often cite those beliefs that we form spontaneously as aresult of sense perception. (I form the belief that there is a coffee cupon my desk. I do so not as a result of any inference from other beliefsI hold, but as a result of the experience of apparently seeing a coffeecup on the desk.) This is fair enough: these are indeed non-inferen-tial beliefs. But they are not beliefs that we hold without what ap-

pears to us to be adequate reason. In normal circumstances – goodlighting, reliable eyesight, and so on – I feel I am entitled to take theexperience of apparently seeing a coffee cup as veridical. The percep-tual experience is all the reason I need to act on the belief to which itgives rise.

But if my belief in the coffee cup on the desk is not without (whatseem to me to be) adequate grounds, the situation is not so clear withregard to the assumption on which I am relying: that of the reliabili-ty of my vision in these circumstances. It is this assumption (BRSP)

that might seem to offer a challenge to evidentialism, as I have de-fined it.

Given that we can describe this assumption as a belief, is it a be-lief that we hold without adequate reason? There are two possibili-ties here. One possibility is that, sometimes at least, our BRSP reallyis held without adequate reason. Take, for instance, a philosopherwho has read William Alston’s  Perceiving God. He has become con-


Page 137: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 137/205

Evidentialism Revisited

vinced that Alston is correct: that our practice of forming beliefs as a

result of sense perception is a doxastic practice that cannot be justi-fied in a non-circular manner. We might think that sense perceptionis veridical because it has been apparently veridical in the past, buthow do we know it has been veridical in the past? Our appeal to pastcases of sense perception merely begs the question. Such a philoso-pher might conclude that we lack adequate reasons to believe in thereliability of sense perception. But we might judge that even he con-tinues to act responsibly in accepting the everyday deliverances of his senses. He is entitled to act on the belief that there is a coffee cup

on the desk. Yet if he is entitled to act on this belief – a belief that hebelieves lacks adequate reason – would this not be a counterexampleto the evidentialist claim?

One interesting response is that offered by John Bishop and Im-ran Aijaz. It is that when it comes to sense perception, the questionof obligation does not arise. It does not arise because we simply haveno choice about taking sense perception as reliable. We cannot, inpractice, do otherwise. Seriously to doubt the deliverances of oursenses, in all circumstances, would be a path to madness. And if we

cannot do otherwise, then we cannot be said to have any obligationsin this respect. As a general principle, we must be entitled to dowhat we cannot avoid doing (“ought implies can”). It follows thatwhat Plantinga calls the de jure (“Are we entitled?”) question aboutbasic perceptual beliefs does not arise as a genuine question bearingon how we live our lives: acting confidently on their truth is “hard-wired” into us. “You might just as well ask whether we are withinour property rights when we take in oxygen with every breath”(Bishop and Aijaz 2004: 121).

This seems to be an excellent response. It does not, of course, offeran adequate reply to the sceptic. It does not show that the deliver-ances of the senses are, in fact, reliable. So the epistemic question(the de justificatione question, which my question [2]) remains unan-swered. But if the de jure question has to do with when we are enti-tled to act on a belief, then Bishop and Aijaz’s reply seems entirelyappropriate. If the person who accepts Alston’s arguments is acting


Page 138: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 138/205

Evidentialism Revisited

responsibly in continually to regard sense perceptual as veridical,

this is no counter-example to the evidentialist thesis. The question of entitlement cannot arise when a person cannot do anything else.

 Yet there is another possibility. This argument has assumed thatwe do hold our belief in the reliability of sense perception (BRSP)without adequate evidence. But is this true? It may be that we do

have adequate reasons for our BRSP, even though we rarely if everreflect on what those reasons are. Does this idea make sense? It ismy contention that it does make sense, if we employ something likeJonathan Adler’s notion of tacit confirmation. Adler argues that a be-

lief is tacitly confirmed (or corroborated) when it is bundled with an-other belief which has survived an explicit test. I believe, for in-stance, that cows exist. This belief is bundled with the idea that milkcomes from cows. So whenever I see look to see if there is milk on thesupermarket shelf and discover that there is, then my belief thatcows (continue to) exist is tacitly confirmed. Here’s another example,from Adler himself:

If I recall a canoe trip with a friend, and I want to be told thename of the rental service that we used, I would call the friendand ask him whether he remembered it. In asking, I take forgranted that the trip took place. But in calling, I risk his failing torecall the trip, which subjects to falsification what I take for grant-ed. Of course, one reason he might not recall the trip is simplythat his memory failed him. But another might be a failure of mymemory. So in calling the friend I risk undermining my own belief.The result of the questioning is that I may be led to doubt eitherthat I took the trip at all or that I took it with this friend. Sincethese background beliefs risk falsification, if they are not falsified

they are confirmed, however unintended, effortless, and minor.(Adler 2002: 163–64)

Let’s express this point more formally. Adler’s idea arises from acombination of two familiar theses (Adler 1990: 559–70). The first isthat “statements are not tested in isolated”; the second is that “a hy-pothesis is confirmed to the extent that it ran the risk of being falsi-fied [and was not].” The conjunction of these two entails that “many


Page 139: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 139/205

Evidentialism Revisited

more statements receive confirmation from a successful test than the

focal hypothesis under investigation. … [But] of course, these back-ground beliefs are not viewed as being tested at these times, so theresulting confirmation is a mere by-product of enquiry, going unno-ticed and unrecorded.”

Tacit confirmation seems to involve a kind of tacit reasoning: weare reasoning, but we are not aware of (in the sense of reflecting onthe fact that) we are reasoning. Assuming – as many psychologistsdo (see, for example, Dijksterhuis and Nordgren 2006) – that we doengage in tacit reasoning, what kind of tacit reasoning might confirm

our BRSP?The reasoning involved, I would suggest, is a kind of “abductive”

reasoning. C. S. Peirce (1839–1914) came up with this term to de-scribe reasoning of the following form:

Fact E is observed.If hypothesis H were true, E would be a matter of course.Therefore, we have reason to believe E.

The conclusion here is rightly cautious, since if it were “Therefore, E 

is true,” it would be a blatant case of the fallacy of affirming the con-sequent. (Compare: “The path is wet. If it had been raining, the pathwould be wet. Therefore, it has been raining.” This is, as a moment’sreflection will show, an invalid argument.) So the most we can say isthat abductive reasoning lends some support to a conclusion. Itcould, for example, form part of an “inference to the bestexplanation,” in which it is coupled with the idea that none of theavailable alternative hypotheses explains E as well as H does.

There are various ways in which sense-perception could be being

tested tacitly by means of abductive reasoning. I shall explore whatseems the most promising of these later (16.2.3), and argue that itdoes not involve circular reasoning, as William Alston (for example)alleges (16.2). If I am correct, then we may be aware of matters thatwould (on reflection) constitute adequate reasons for our belief in thereliability of sense perception (BRSP). That belief has been tacitlyconfirmed by a host of experiences. And if this is true, then eviden-


Page 140: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 140/205

Evidentialism Revisited

tialism – the view that we should act on beliefs only if these are ade-

quately supported by evidence – remains undefeated.


Page 141: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 141/205

Page 142: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 142/205

Chapter Fifteen

Reformed Epistemology

I have already made reference, on several occasions, to AlvinPlantinga’s “Reformed Epistemology.” It is now time to look moreclosely at this currently influential view. It involves three claims, allof which have to do with not just theism but Christian theism. Thefirst is that the Christian faith can be a form of basic belief; the sec-

ond is that it is warranted; the third is that it is undefeated.

15.1 Christian Faith as Basic Belief

Plantinga’s starting-point is the idea that Christian faith may be,and often is, a form of basic (non-inferential) belief. He freely admitsthat we cannot produce arguments that would demonstrate the truthof the Christian faith. To this extent Plantinga is in agreement withthe atheist. But this does not bother him, since he assumes that if Christian faith is a form of properly basic belief, it needs no support-ing arguments. As he writes, if Christian faith is a form of basic be-lief (and is properly so), then

the believer is entirely within his intellectual rights in believing ashe does even if he doesn’t know of any good theistic argument (de-ductive or inductive), even if he doesn’t believe that there is anysuch argument, and even if in fact no argument exists... It is per-fectly rational to accept belief in God without accepting it on thebasis of any other beliefs and propositions at all. (Plantinga 1981:42)

Let’s look at this a bit more closely. Rather than saying that belief in God may be a form of basic belief ( properly basic belief, to usePlantinga’s earlier terminology), it would be more accurate to saythat various beliefs about God may be properly basic. These includebeliefs such as the following:

God is speaking to me.


Page 143: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 143/205

Reformed Epistemology

God has created all this.

God disapproves of what I have done.God forgives me.God is to be thanked and praised.

Each of these entails the existence of God, but of course they entailmuch more than that. Hence Plantinga’s proposal, if successful, of-fers support not just to belief in God (that God exists), but to an in-definite number of beliefs about God. So with one deft move, Plantin-ga can eliminate the gap between theism and Christianity, a gapthat troubles so many other would-be defenders of the faith. All he

needs to do is to argue that the Christian, on reading the Bible, canform the properly basic belief that this or that biblical teaching is thevery word of God. This is precisely what Plantinga’s “extended Aquinas / Calvin model” allows him to argue, as we shall see in a mo-ment. If the Christian reads 2 Tim 3:16 (“all Scripture is inspired byGod”), and believes this to be true in a properly basic way, then inthat very moment, as if by magic, the whole body of biblical teach-ings receives justification.

But I’m leaping ahead. What does it mean to say that such Chris-

tian beliefs may be basic beliefs? What are the circumstances inwhich such beliefs are formed? Well, on Plantinga’s account it wouldseem that almost any circumstance can trigger the formation of sucha belief about God. It may be perceiving the majesty or delicate beau-ty of nature, it may be a feeling of guilt or of forgiveness, or it may bethe fact of being in grave danger. In the case of distinctively Chris-tian beliefs, the conviction of their truth normally arises on the occa-sion of reading Scripture or hearing these beliefs proclaimed.

We read Scripture, or something presenting scriptural teaching, orwe hear the gospel preached, or are told of it by parents, or en-counter a scriptural teaching as the conclusion of an argument (orconceivably even as an object of ridicule), or in some other way en-counter a proclamation of the Word. What is said simply seemsright; it seems compelling; one finds oneself saying, “Yes, that’sright, that’s the truth of the matter; that is indeed the word of theLord.” ... And I may also think something bit different, something


Page 144: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 144/205

Reformed Epistemology

about that proposition: that it is a divine teaching, or revelation,

that in Calvin’s words it is “from God.” (Plantinga 2000: 250)

This conviction may arise suddenly, or it may arise slowly. It mayeven arise with regard to a belief one has had since childhood. But“in each case,” Plantinga writes, “there is the reading or hearing, andthere is the belief or conviction that what one reads or hears is trueand a teaching of the Lord” (Plantinga 2000: 251).

What is important is that on Plantinga’s account the belief soformed is truly basic, in the sense of non-inferential. It is not theproduct of a process of argumentation, however swift or unconscious.The experiences in question are the occasions of belief, not premises

in an argument that leads to belief. As Plantinga writes,

one doesn’t argue thus: I am aware of the beauty and majesty of the heavens (or of my own guilt, or that I am in danger, or of theglorious beauty of the morning, or of my good circumstances):therefore there is such a person as God. The Christian doesn’t ar-gue: “I find myself loving and delighting in the good things of thegospel and inclined to believe them; therefore they are true.”These would be silly arguments; fortunately they are neither in-voked nor needed. (Plantinga 2000: 330)

There is no “argument from religious experience” here, althoughsome kind of experience is certainly involved. But it may be simplywhat Plantinga calls a “doxastic experience”: the experience of enter-taining a proposition one believes to be true (Plantinga 2000: 183).(Contemplating the proposition 3+ 2 = 5  feels different from contem-plating the proposition 3+ 2 = 6.)

15.2 Christian Faith as Warranted Belief

So much for Christian beliefs as basic beliefs. In what sense maythey be said to be  properly basic beliefs? At the heart of Plantinga’sposition is the view that if what Christians believe is true, then theirbeliefs have warrant. It is in this respect that Reformed Epistemolo-gy resembles the “bootstrapping” view of faith that I examined earli-er – in which religious faith is thought to provide its own evidential


Page 145: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 145/205

Reformed Epistemology

support – since you will believe that the Christian faith has warrant

only if you already believe it is true. What warrant do Christian be-liefs have, according to Plantinga? There are two sources of warrantat work here. The first is what John Calvin called the sensus divini-

tatis, the “sense of divinity,” a tendency to form belief in God that isimplanted in the hearts of every human being. The second is whatPlantinga calls “the internal instigation of the Holy Spirit” (abbrevi-ated IIHS, which corresponds to what John Calvin called the “inter-nal testimony of the Holy Spirit”). This refers to the work of the HolySpirit in the heart of the believer, which produces a firm conviction

regarding the truths God has revealed. While the sensus divinitatis

is an innate cognitive mechanism, the IIHS is something supernatu-ral: it is a special work of God. Strictly speaking, it is not a cognitivefaculty at all, as the sensus divinitatis would be, but an external in-fluence. But let’s leave that aside for the moment.

How do we know of the existence of such mechanisms (if indeedwe do know of them)? By way of the Christian faith, of course. Noone but a Christian would believe that such mechanisms exist. Doesthis not make Plantinga’s appeal to the existence of such a warrant

question-begging and circular? Plantinga argues that it does not. Itis true that the warrant exists only if Christianity were true. If one’sbelief in the truth of Christianity depended on the warrant – if thewarrant were the premise of an argument in support of the truth of Christianity – this would lead to a fatal circularity. But if belief inthe truth of Christianity is a form of basic belief , this accusation of circularity has no force, since the believer is not producing an argu-ment in support of her faith. And if there is no argument, there canbe no circular reasoning.

Note that for Plantinga the Christian’s belief may be warranted (if Christianity is true) whether or not she knows about this warrant.What we see at work here is Plantinga’s thoroughly externalist con-ception of warrant. Whether a belief has warrant is an objective fact,pertaining to how the belief is produced. Plantinga has spent twobooks developing and discussing his conception of warranted belief.Here is his formal definition of warrant.


Page 146: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 146/205

Reformed Epistemology

 A belief has warrant for a person S only if that belief is produced

in S by cognitive faculties functioning properly (subject to no dys-function) in a cognitive environment that is appropriate for S ’skind of cognitive faculties, according to a design plan that is suc-cessfully aimed at truth. (Plantinga 2000: 156)

 At the heart of this idea is that of “a design plan successfullyaimed at truth.” The underlying idea here is that the physical organsand other systems that constitute the person each have a  proper

 function: there are ways in which they supposed to work, in order todo the job for which they are designed. As Plantinga notes, this sense

of “design” need not, of itself, imply that we were “created by God orsome other conscious agent” (Plantinga 2000: 154). Plantinga con-cedes that in principle a natural process of evolution could equip ourminds and bodies with proper functions, although in fact he believesthat there is no adequate naturalistic account of proper cognitivefunction. As we have just seen, what Plantinga calls the “Aquinas /Calvin model” offers the believer warrant for her belief. But sincethat warrant has to do with the proper functioning of these cognitivemechanisms, it does not depend on the believer knowing about the

 Aquinas / Calvin model. It follows that the believer may be warrant-ed in her belief while having “no second-order beliefs at all” aboutwhat she believes (Plantinga 2000: 179).

15.3 Christian Faith as Undefeated Belief

Plantinga claims that this view of Christian faith undercuts what hecalls the de jure (as opposed to de facto) objections to belief. De facto

objections are those that argue that certain central Christian beliefs

are false, either because they are incoherent and therefore could notpossibly be true or because there exists some fact that is incompati-ble with their truth. (The argument from evil, Plantinga argues, maybe a de facto objection, as is the argument that the doctrine of theTrinity is incoherent.) De jure objections, on the other hand, have todo with the rationality of Christian faith.

These are arguments to the effect that Christian belief, whether ornot true, is at any rate unjustifiable, or rationally unjustified, or irra-


Page 147: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 147/205

Reformed Epistemology

tional, or not intellectually respectable, or contrary to sound morality,

or without sufficient evidence, or in some other way rationally unac-ceptable, not up to snuff from an intellectual point of view. (Plantinga2000: ix)

The de jure arguments suggest that the Christian is in some way vio-lating her epistemic duties in believing as she does, at least if she isadequately informed of the facts. Plantinga claims that the de jure

objection would have force only if one had proven that the Christianfaith is not, in fact, true, since if it is true, it has warrant. (This iswhere the sleight of hand occurs. Can you spot it? See below.) What

the atheist requires is a de facto objection. He needs to show that theChristian faith is not, in fact, true.

The problem for the atheist is that it is practically impossible topersuade the Christian of the falsity of her belief, if we begin fromthe assumption that her faith is a form of properly basic belief. Whyis this? Well, let’s look at some potential defeaters. First of all, thereare naturalistic explanations of religion, similar to those offered byKarl Marx and Sigmund Freud, that suggest that religion is theproduct of cognitive dysfunction. If these seemed more likely to be

true that the Christian’s claim regarding warrant, would this not un-dercut Christian faith? No, says Plantinga. It would do so if you hadembraced the Christian faith as one explanatory hypothesis amongmany, to be judged on the basis of its explanatory power. But if yourbelief in God is a form of basic belief, you are not holding it on thebasis of its explanatory power. “Hence the fact that there are betterexplanations of some range of phenomena (if there are) does not sofar cast any doubt on belief in God” (Plantinga 2000: 371).

What about historical judgements of probability, of the kind that

(as Hume argued) apparently undermine belief in miracles? Thesepose no difficulty for the Christian. Since (from Plantinga’s point of view) she already knows (in a basic way) that Christianity is true,she has no reason to exclude this knowledge from her historical judgements. Only something like an outright contradiction betweena well-established historical claim and what is known by faith (a re-butting defeater) might overturn that faith.


Page 148: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 148/205

Reformed Epistemology

What about the fact of religious pluralism? Does it not seem im-

probable that among the variety of the world’s religions only this onewill be true? Perhaps it does seem improbable, even to the Christian,against the background of her other beliefs. But this objection

(like so many others) seems to make sense only if the believer, tobe rational, must hold her Christian beliefs on the basis of theirrelations to other beliefs she has – or, at any rate, only if thoseChristian beliefs are probable with respect to those beliefs. On of the main burdens of this book, however, is that the believer can beperfectly rational in acquiring some of her beliefs in the basic way

 – not on the basis (probabilistic or otherwise) of other beliefs.(Plantinga 2000: 442)

The Christian’s source of warrant is independent of those other be-liefs, so the fact that what she believes might seem antecedently im-probable is neither here nor there. After all, the fact that it mightseem improbable that I should win Lotto is no reason for me not tobelieve it when I see my numbers come up.

But what about the existence of horrifying and apparently gratu-itous evils in the world? Surely this is a defeater for Christian faith?No, because insofar as this is a probabilistic argument – that a worldwith this degree of suffering is more probable on the assumption thatthere is no God than on the assumption that there is – the same con-siderations can be brought into play. An evidential challenge can bemounted against many of our beliefs. But if we have good indepen-dent reasons for holding those beliefs, that does not make them anyless rational.

My friend has a cat named Maynard; I believe that Maynard is a

cat and also (as my friend reports) that Maynard likes cookedgreen beans; the latter, however, is much more likely on the ... al-ternative hypothesis that Maynard is a Frisian, or possibly aFrenchman; so the belief that Maynard is a cat is evidentiallychallenged for me. I believe (naturally enough) that you are a hu-man being; you and I are on a walk in the woods, however, so Ialso believe that you are in a forest; of course that proposition isvastly more likely on the ... alternative hypothesis that you are a


Page 149: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 149/205

Reformed Epistemology

tree; so the belief that you are a human being is evidentially chal-

lenged for me. (Plantinga 2000: 475–6)

But perhaps, Plantinga concedes, the problem of evil is not a mat-ter of an argument. Perhaps it is simply that belief in the non-exis-

tence of God is a properly basic belief that arises when confrontedwith the sheer horror of the world. The mechanism that producesatheism would then be “a sort of inverse sensus divinitatis” (Plantin-ga 2000: 484). To reject belief in God when confronted with extremeand apparently senseless suffering show that one’s cognitive facul-ties are functioning properly. No, says Plantinga, because that is pre-cisely what the Christian will not accept. She has an account of whatit means for our cognitive faculties to be functioning properly – Plantinga’s version of this is the Aquinas / Calvin model – and onthis account proper function does not include giving up belief in Godwhen confronted with horrendous evils.

I needn’t continue. This will be enough to convey Plantinga’sstrategy. Once it has been established that Christian belief can be awarranted basic belief, then only the strongest sort of defeater, thatis to say, one that arises from a contradiction between Christianfaith and some other well-established proposition, has any chance of defeating that faith. Any lesser objection – an argument which wouldmerely reduce the likelihood that some Christian belief was true – can be ignored on the grounds that Christian belief is a form of basicbelief. It is not the result of an argument and does not rely on theweighing of evidence.

15.4Against Reformed Epistemology

There are many objections that can be offered to this line of argu-ment. The most decisive, in my view, arises from the “Great Pump-kin objection,” first raised by Plantinga himself.

If belief in God is properly basic, why can’t  just any belief be prop-erly basic? Couldn’t we say the same for any bizarre aberration wecan think of? What about voodoo or astrology? What about the be-lief that the Great Pumpkin returns every Halloween? Could I


Page 150: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 150/205

Reformed Epistemology

properly take that as basic? And if I can’t, why can I properly take

belief in God as basic? ... If we say that belief in God is properlybasic, won’t we be committed to holding that just anything, ornearly anything, can properly be taken as basic, thus throwingwide the gates to irrationalism and superstition? (Plantinga 1981:48)

In other words, does Reformed Epistemology not loosen the con-straints on what we consider warranted belief to the point wherethey mean practically nothing at all?

15.4.1 Plantinga’s ResponsePlantinga responds to this objection in Warranted Christian Belief .First, he reminds us that not just any set of beliefs have the propertyhe is claiming for Christianity, that of being warranted if true. Yethe immediately concedes that Christianity may not be the only con-tender: there may be other beliefs which have this property. As hewrites, “probably something like that is true for the other theistic re-ligions: Judaism, Islam, some forms of Hinduism, some forms of Bud-dhism, some forms of American Indian religion” (Plantinga 2000:

350). These sets of religions beliefs would also be warranted if true.It is tempting to conclude from this remark that, by virtue of Plantinga’s arguments, all these incompatible sets of beliefs wouldbe warranted, a fact (if it were a fact) that would destroy his case.But as he himself points out, a set of beliefs is warranted only if true

and since these cannot all be true, they cannot all be warranted. He,of course, believes the Christian account is true and therefore thatonly this account has warrant. And his belief that the Christian ac-count is true is a basic belief. End of story.

If this were the point of the Great Pumpkin objection, Plantinga’sresponse would perhaps be adequate. But it is not, since it missesthe implications of the objection. The force of this objection is notthat any set of bizarre beliefs could in fact be warranted on the basisof Plantinga’s arguments; it is that the holders of such beliefs couldclaim their beliefs to be warranted on the basis of Plantinga’s argu-ments. They may of course be wrong. Indeed most of them would


Page 151: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 151/205

Reformed Epistemology

have to be wrong. But that’s not the point. It is the fact that anyone

could claim warrant in this way, for almost any set of beliefs, that isworrying. The conditions under which this could occur are not verydemanding. First, your belief would have to be arrived at in a non-in-ferential way. As soon as you are presented with these beliefs, a con-viction regarding their truth arises in your heart. Secondly, amongthe things thus believed there would need to be the existence of amechanism guaranteeing the truth of that belief. It doesn’t matterhow bizarre the belief may appear to be, that is to say, how improba-ble it might seem to be against the background of the rest of our

knowledge. The fact that it is a basic belief exempts you from havingto assess its probability vis-à-vis our other beliefs. So it is very hardto defeat.

 Are you not convinced? Then consider the following statement of my warranted, basic Zeeoplean faith. The Zeeopleans are members of a highly-advanced alien civilization, from a far-off part of the uni-verse. For some years now, one of their spacecraft has been in sta-tionary earth orbit, directly above New Zealand. Its presence cannotbe detected because, like the Zeeopleans themselves, it occupies an-

other dimension of what we think of as space–time, one unknown toearthly science. (This is how the Zeeopleans have explained it to me,but in talking about “dimensions,” they are speaking metaphorically,to accommodate the truth to the weakness of my human intellect.)They are kindly disposed towards humanity and have chosen me tobe their spokesman on earth. So you should listen carefully to me.But I should warn you. If you don’t listen to me, the Zeeopleans haveauthorized me to use violence against you. While they are kindly dis-posed, they are prepared to allow some people to suffer, even to suf-

fer horribly, for the good of humanity as a whole.How do I know this to be true? I have no arguments, at least no

arguments that you would see as convincing, in support of this belief.It simply came to me one night as I was out walking, looking at thestars, and I immediately formed a conviction of its truth. So it is abasic belief. What makes me think it is a warranted basic belief?Well, one of the things the Zeeopleans have revealed to me is that


Page 152: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 152/205

Reformed Epistemology

when I was three years old they abducted me for a time – my parents

thought I was lost in a shopping centre – and implanted in my braina reliable cognitive mechanism. (I think of it as the sensus alieno-

rum.) When a belief occurs to me, this mechanism enables me imme-diately to discern if it is a message from the Zeeopleans. Why do Ialone have this set of beliefs? Well, it’s because I alone have the sen-

sus alienorum. I’m not proud of this, in fact, I feel humbled that theZeeopleans chose me as their messenger. Why did they choose a mes-senger, especially one from a far-off country, rather than revealthemselves immediately to everyone? Yes, this does seem odd, in the

sense of antecedently improbable. But I know (in the sense of havinga warranted basic belief) that it is true and, after all, lots of oddthings turn out to be true. In any case, who am I to understand theirintentions?1

 Are my arguments in support of my Zeeoplean faith different inkind from Plantinga’s? Apparently not. It might be argued that myZeeoplean faith could not withstand the various defeaters that mightbe brought against it. Perhaps it wouldn’t. But bear in mind that Icould deploy the same strategy against potential defeaters as does

Plantinga. Whenever someone offered evidence against my view – evidence which made it improbable, given everything else we know,that the Zeeopleans exist – I could simply reply that my belief in theZeeopleans is not based on such evidence. It is a basic belief, one thatdoes not arise by inference from my other beliefs. So the fact that mybeliefs render my Zeeoplean faith “evidentially challenged” mattersnot a whit, for the truth of that faith is evident to me, independently

1 When I first came up with this example, I thought it was entirely fictional.

But I have since learned about Nancy Lieder, who claims that a cognitivemechanism was implanted in her brain when she was younger, allowing herto be channeling messages from a group of aliens called “Zetans,” i.e. ZetaReticulans, from the star Zeta Reticuli. (My daughter, aged 10, suggestedthat it should be called Zeta Ridiculi, but in fairness, Zeta Reticuli is a realstar.) Those messages speak of a disaster that will shortly occur on theearth because of a pole shift caused by the close approach of a large plane-tary object. Needless to say, Nancy has a website and followers:


Page 153: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 153/205

Reformed Epistemology

of any other beliefs I hold. Such a strategy would make my Zeeo-

plean faith all but unassailable. But if this is the case, then surelysomething has gone badly wrong. To understand what has gonewrong, we need to look more closely at what Plantinga is doing.

15.4.2 A Fundamental Confusion

It is important to appreciate that Plantinga’s argument is not ad-dressed to what I have called the question of doxastic entitlement(the de jure question, as he calls it). Doxastic entitlement is an inter-nalist notion: it has to do with the agent’s state of mind. But

Plantinga’s account of warrant is externalist. It has to do with theexistence of a reliable belief-forming mechanism; it does not requirethat the believer have reason to believe it exists. But then I cannotsee what question it is intended to answer, for it addresses neitherwhat I have called the de justificatione question (that of evidentialsupport) nor the de facto question (that of truth). In fact, at the endof the day, his argument appears trivial.

To understand why this is the case, let’s look again at Plantinga’scentral claim, which is that if Christian faith is true, then it is war-

ranted. Keep in mind that we know of the warrant in question (if wedo know of it) only by way of the Christian faith. That the Christianfaith is warranted, in other words, is itself one of the things Chris-tians believe. What this means is that Plantinga’s central claim re-duces to a proposition of the same form as the following.

If the cat is large and the cat is black, then the cat is black.

This is, of course, true, but it is (as philosophers say) triviallytrue. Its conclusion tells us nothing more than we knew already.

To see how this applies to Plantinga’s view, let (“a and b and c andd and e” ) represent the conjunction of all Christian beliefs, with (“dand e”) being the two of these beliefs that relate to warrant, namelybelief in the sensus divinitatis and the internal testimony of the HolySpirit. Plantinga’s central claim – that if Christian faith is true, thenit is warranted – can now be expressed as follows:

If (a & b & c & d & e), then (d & e).


Page 154: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 154/205

Reformed Epistemology

It is a mystery to me why Plantinga should regard this as a matter of 

any philosophical interest.He argues, of course, that it is a fact of some significance, for it

undercuts the atheist’s de jure objection. The atheist must demon-strate that the Christian faith is not true, since if it is true, it haswarrant. But this is simply beside the point. There is no point inChristian faith having warrant unless the Christian can know it haswarrant. And what the atheist can argue is that neither Plantinganor any other Christian knows this. No Christian, the atheist mightsay, has adequate reason to regard the Christian faith as true (and

thus warranted).Plantinga’s response would surely be that the Christian need not

produce such a reason, for her faith is a matter of basic belief. But if one starts from an evidentialist view of belief, this will not do. Notevery belief that we form spontaneously is a belief for which we haveadequate evidence. Even basic beliefs, I would argue, are in need of  justification; we need reason to believe that they are the product of areliable mechanism. (A recognition of this fact lies behind philosophi-cal responses to scepticism.) The atheist’s argument is that no Chris-

tian can produce the kinds of reasons required. This is the very ob- jection that Plantinga claims to have disposed of, and I fail to seethat all his talk about warrant does anything to address it.

Here, in fact, lies the fundamental confusion of Reformed Episte-mology. Plantinga speaks as though a story about a possible warrant – about a mechanism by which the belief could be produced – conferson Christian belief some kind of doxastic entitlement or epistemic justification. But the fact that there exists some story about how abelief is produced tells us nothing of any philosophical interest, un-

less there is evidence to suggest that this story is true. Plantinga of-fers no evidence of this kind; indeed he denies that such evidence isneeded. But he offers us no reasons why we should accept his story. At the end of the day, therefore, the discussion is right back where itbegan.


Page 155: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 155/205

Chapter Sixteen

Religious Experience

So much for arguments, but what about religious experience? Formany religious people, their faith rests not on arguments, but on ex-perience. They have, they report, had some experience that convincesthem God exists. That people do have such experiences is beyonddoubt. The key question for the philosopher is: What evidential force

(if any) do such experiences possess, either for the person herself orfor others?

16.1 What is Religious Experience?

The first problem is to know what is meant by “religious experience,”which is a very murky phrase, used in a wide variety of ways. Let’sstart with the term experience. I shall take it here to refer to anykind of conscious mental state. The subject whose mental state it ismay or may not believe it to have an external-world object. This al-lows us to distinguish two kinds of experience. When the subjectdoes not believe her experience to have an external-world object, wemay describe it as a subject–state experience. Examples of subject– state experiences are feeling nauseated or dizzy or a generalized feel-ing of anxiety ( Angst) or happiness. In such cases I am certainly con-scious of something, but what I am conscious of is – I would freelyadmit – nothing more than my own internal state. When the subjectdoes believe her experience to have an external object, we may de-scribe it as a subject–object experience. Examples of subject–objectexperiences are seeing a tree, smelling coffee, or hearing a voice. Of course, these experiences may not be reliable: the subject’s beliefsabout their source may be false. One can hallucinate trees, the aro-ma of coffee, and voices. The important thing is that such experi-ences are believed to be experiences of something outside of the self.Subject–state experiences, on the other hand, are thought of as expe-riences of something within the self.


Page 156: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 156/205

Religious Experience

We normally assume that our subject–object experiences will be

mediated by way of the senses: sight, hearing, touch, taste, smell.But what about alleged experiences of God? It is true that sometimesthese are thought of as sensory experiences. However, there is some-thing problematic about the idea of a sensory experience of God. Asensory perception of some this-worldly divine action, if one were tooccur, seems intelligible. (One thinks of Moses’ reported vision of thebush that burned but was not consumed.) But there is something oddabout the idea of an alleged sensory perception of a spiritual (i.e.non-physical) being. So those who claim that there exist subject–ob-

 ject experiences of God must show that it makes sense of speak of anexperience “of God.”

How could they do this? Well, they could defend this idea by argu-ing for the possibility of extrasensory perception (ESP). Or they mayargue that our consciousness of God is the result of a supernatural – presumably miraculous – action on God’s part, whereby he presentshimself to human consciousness, whether by way of sensory percep-tion or not. But neither solution seems terribly adequate. We have noreliable evidence that any form of ESP is possible, let alone one

which has God as its object. And to suggest that God could work amiracle may be true. But why would he create us in such a way thata miracle was required? These difficulties surely establish a pre-sumption against any claim to have “experienced God.” But suffi-cient evidence may overturn the presumption. If we had sufficientreason to do so, it may be rational to believe that there exist experi-ences of God even if we have no account of how this occurs.

16.1.1 Implied Explanatory Claims

I have argued that what I am calling “experiences” include both sub- ject–state and subject–object experiences. What would make such ex-periences religious? In practice, people regard a wide range of experi-ences as religious. But when people refer to religious experience inthe context of arguing for belief in God, they speaking of experiencesthat are thought to have a particular kind of authority. How can weaccount for that perceived authority? My suggestion (which I cannot


Page 157: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 157/205

Religious Experience

defend here) is this. Those who appeal to religious experience as evi-

dence for God’s existence are making an (explicit or implicit) ex-planatory claim. By labeling an experience religious and by suggest-ing that it has evidential force, they imply there is something aboutthis experience that cannot be adequately explained without refer-ence to divine action.

16.1.2 The Varieties of Religious Experience

If we do define religious experiences in this way – as experiencesthat the subject believes cannot be adequately explained without ref-

erence to God – then what kind of experiences may these include?Well, there are two possibilities, corresponding to the two types of experience identified above. The religious experience may be a sub-

 ject–state experience, such as a feeling of profound joy or a deepsense of peace. Such an experience may profoundly change a person’sbehaviour (a psychological fact), but it is not at all clear what eviden-tial weight it could carry. Such experiences could have various caus-es, many of them entirely natural. Under what circumstances coulda subject–state experience have evidential force? It would do so only

if it could not be explained by reference to natural causes, or if a reli-gious explanation of its cause was demonstrably superior to any nat-ural explanation. But in these circumstances the religious experiencein question becomes simply another fact about the world that appar-ently requires God for its explanation. Its evidential force is no moreor less than that of any other alleged miracle. The argument from re-ligious experience now becomes an inference from the occurrence of amiracle to the existence of a supernatural agent capable of causingthe miracle. We have already discussed such arguments and they

need not detain us here. Alternatively, the religious experience may be a subject–object ex-

perience, that is to say, an alleged experience of a supernatural agent(such as God, an angel, or the risen Jesus). Such an experience maybe thought to involve sensory perception, if it is believed that the di-vine agent in question has a physical body. (The post-resurrectionJesus, for instance, is often thought to have a physical body, albeit


Page 158: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 158/205

Religious Experience

mysteriously transformed.) Or it may be thought to involve some

kind of extrasensory perception. (You may share my scepticism aboutextrasensory perception, but let’s leave that aside.) If you are lookingfor evidence for the existence of God, evidence that differs from thetraditional theistic arguments, these alleged experiences of Godwould seem to be the most promising place to start. Remember thatmany of our perceptual beliefs are basic beliefs, that is to say, beliefsthat enjoy a prima facie justification, even in the absence of inferen-tial support. If there were experiences of God, could they not giverise to a basic belief in God, one that required no further support?

In my discussion of evidentialism, I’ve tried to cast doubt on thevery idea of an (evidentially) basic belief, but let’s assume for the mo-ment that I’m wrong and that there do exist such beliefs. After all,many philosophers do hold that some beliefs – such as our belief inthe reliability of sense perception – have a special status. We are,they argue, entitled to hold such beliefs until presented with reasonto think they are false. (They have, such philosophers argue,  prima

 facie justification.) If this were the case, then could beliefs arisingfrom an apparent experience of the divine not fall into the same cate-

gory? Might the believer not be entitled to hold such religious beliefs,if she has had an experience of this kind?

16.2 Perceiving God

The most sophisticated defender of this view is William Alston. Al-ston’s central claim is a very simple one. It is that religious experi-ences – understood as a subject–object experiences that have God astheir putative object – can contribute to the justification of a person’sbelief in God. Such a person may be acting rationally in taking Godto be the object of that experience.

More precisely, Alston argues that there exist differing doxastic

 practices, different ways in which we do in fact form our beliefs.These include sense perception, reliance on memory, rational reflec-tion, and reliance on religious experience. The key issue for Alston isthat we cannot establish the reliability of any of these practices in anon-circular manner. He illustrates this point with reference to what


Page 159: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 159/205

Religious Experience

he calls our “sense perceptual doxastic practice” (SP). On the one

hand, this is surely the best-established of all our doxastic practices.On the other hand, Alston argues, we cannot demonstrate its relia-bility in a non-question-begging fashion. All our arguments in sup-port of SP either fall short of demonstrating that reliability or atsome point take granted it for granted.

If this is true, then our epistemic situation seems desperate. Wecannot prove the reliability of any of the ways we form our beliefs.What are we to do? The only reasonable course of action, he argues,is to accept the outputs of our established doxastic practices. We can-

not think at all without adopting some doxastic practice. And any al-ternative practices we came up with would almost certainly sufferfrom the same limitations. So it is perfectly reasonable for us to ac-cept the practices we already have and to choose to regard their out-puts as reliable. But among our established doxastic practices aremystical perceptual practices, some of which are well establishedand sophisticated. Alston is particularly interested in what he calls“Christian mystical perceptual doxastic practice” (CMP). If it is ratio-nal for us to accept our other doxastic practices, he argues, it is ratio-

nal to accept CMP. It is up to the sceptic to overturn this presump-tion of reliability. He must show why this particular well establishedand widely accepted practice should not be considered a rationalchoice.

Let’s look at this conclusion more closely. The rationality that Al-ston is claiming for CMP (along with SP and our other doxastic prac-tices) is a  practical rationality. It has to do with how we may bestlive. If we follow Alston’s advice, we are not accepting these practicesbecause we know they are reliable. We are accepting these practices

because, given our cognitive limitations, this seems the best (or per-haps the only) thing we can do. So in what sense could the beliefsthat emerge from such practices be considered  justified beliefs? Al-ston himself distinguishes two sense of the term justification. Thefirst is a strong sense, in which justification implies “likelihood of truth.” To use the term “justification” in this strong sense, to saythat appearance A justifies belief  B for subject S , is not merely to say


Page 160: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 160/205

Religious Experience

that S is acting rationally in believing B on the basis of A. It is to say

that appearance A is a reliable (truth-indicative) sign of the truth of belief  B. But this latter is precisely what one cannot do on the basisof Alston’s argument, as he himself concedes. The most we could sayis that S is acting rationally in believing B on the basis of  A. This is amuch weaker sense of justification, which I have argued is bettercaptured by the term “entitlement.” But even with this qualification, Alston’s argument would still an important one. It would lend sup-port to the idea that someone who has had a religious experience mybe entitled to hold her religious beliefs as a result of this experience.

There are, however, at least three points at which Alston’s posi-tion is vulnerable to criticism. The first has to do with the allegedparallels between SP and CMP. The second is related to the problemof religious diversity. This will lead us to a third point, regarding thevery idea of a doxastic practice.

16.2.1 Are SP and CMP relevantly analogous?

 Alston’s argument rests on the idea that there exist significantanalogies between SP and CMP. But we need to ask just how far

these analogies extend. If there are significant disanalogies, then theargument will be weakened. It may mean that we should not extendto the results of CMP the prima facie justification that we (might) ac-cord to the results of SP. Alston, of course, recognizes the existence of disanalogies, which he believes do not damage his case. But at leastone of them is more serious than he seems to realize. It is the factthat CMP is a voluntary doxastic practice, while SP is not. Even if itis true that there is no non-circular way of demonstrating the relia-bility of sense perception, our adherence to SP is not a matter about

which we have any choice. We are so constructed that we cannotavoid acting as if sense perception were reliable. As philosophers, wemay consider it regrettable that we cannot give a non-circular justifi-cation of our practice. But that is no reason for compounding the evil.It is no reason to adopt another doxastic practice that cannot be jus-tified in a non-circular fashion, when we know we can survive with-out it.


Page 161: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 161/205

Religious Experience

There is another fact about CMP that lends weight to this objec-

tion. It is the fact that CMP is far from firmly established. As it hap-pens, it is not at all clear what makes a doxastic practice “estab-lished” or whether a clear line can be drawn between “established”(and thus respectable) and “non-established” (and thus disreputable)practices. But even if these questions could be answered, Alston him-self notes that some doxastic practices are more firmly establishedthan others and that, in the case of a conflict, the more firmly estab-lished practice is to be preferred. What does it mean for a practice tobe “more firmly established”? Alston writes:

I don’t have a precise definition, but it involves such componentsas (a) being more widely accepted, (b) having a more definite struc-ture, (c) being more important in our lives, (d) having more of aninnate basis, (e) being more difficult to abstain from, and (f) itsprinciples seeming more obviously true. (Alston 1991: 171)

Even if there is no conflict between SP and CMP, by at least some of these criteria CMP is a weakly established practice. This seems to beall the more reason for not embracing CMP if we do not need to do

so.16.2.2 What About Religious Diversity?

These objections to Alston’s position seem to me to be fatal. Yet thereis a still more serious objection which – as he himself notes – arisesfrom the fact of religious diversity and the way in which Alston re-sponds to this objection reveals another weakness in his position.The mystical perceptual doxastic practice (MPs) found in differentreligious seem incompatible. The beliefs that they produce cannot allbe true. It may be that at least one of them is reliable, that it yieldstrue beliefs. But the vast majority must be unreliable, yielding falsebeliefs. So how can I tell which MP is reliable? What non-question-begging reason can I give for thinking that my MP is reliable whileother MPs are not? Let’s say, for the sake of the argument, that I cangive none. Let’s assume that my Christian faith rests on an assump-tion of the reliability of CMP alone. In committing myself to CMP, Iwould be committing myself to a doxastic practice when – statistical-


Page 162: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 162/205

Religious Experience

ly speaking – is almost certainly unreliable. It follows that “it cannot

be rational to engage in CMP; and by the same reasoning it cannotbe rational to engage in any other particular form of MP” (Alston1991: 269).

This seems a knock-down argument. How does Alston respond toit? He appeals to what we might call the “incommensurability” of MPs. Within a particular doxastic practice it is clear how competingclaims might be resolved, since each practice has its own criteria of assessment. But when it comes to comparing different doxastic prac-tices, when it comes to deciding between them, things are not so sim-

ple. In this situation there are no common criteria of assessment.Different MPs have different criteria for constructing and evaluatingbeliefs on the basis of mystical experience. One cannot use the crite-ria which are specific to a particular practice to adjudicate disputesbetween them.

 According to Alston, the lack of common criteria of assessmenttakes the edge off the religious diversity argument. If we knew of some way of choosing between competing MPs, then the critic wouldbe right. It would not be rational simply to assume the reliability of 

my own. But we have no agreed criteria. It follows, in Alston’s words,that “the only rational course for me is to sit tight with the practiceof which I am master and which serves me so well” (Alston 1991:274). It is important to note that this argument applies to all the fol-lowers of any established MP. Not only Christians, but followers of other religions may be acting rationally in choosing to rely on differ-ent MPs. Once again, we see how modest Alston’s conclusions are.He has not shown that CMP is a reliable doxastic practice. At bestwhat he has shown is that the person who adheres to CMP, assum-

ing its reliability, is acting rationally in so doing. Yet even this modest conclusion is open to criticism. A first criti-

cism has to do, once again, with the fact that MPs are voluntary

practices. We are not forced to choose one form of MP over another,in the manner in which we are forced to rely on SP. Now the reli-gious diversity argument suggests that the vast majority of MPs areunreliable and that it would be only by a lucky chance that our was


Page 163: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 163/205

Religious Experience

the correct one. If we wish to avoid error – admittedly at the risk, as

William James reminds us, of missing out on truth – the most ratio-nal course of action would be to avoid adopting (or to abandon re-liance on) MPs altogether. Alston does not consider this option.

 A second, more serious criticism has to do with the alleged incom-mensurability of doxastic practices. It may be that there are doxasticpractices that lack common criteria of assessment. But this is not al-ways true of competing religions. Jews, Christians, and Muslims, forinstance, share a common set of traditions against which competingclaims may be judged. In the case of Jews and Christians, these are

embodied in a common set of Scriptures. There is a practical,hermeneutical difficulty here – that of deciding what these sacredtexts mean – but this is a an intra-practice and well as an inter-prac-tice difficulty. It follows that in this case I am not justified in assum-ing the reliability of my existing practice, on the grounds that I haveno option. Given competing practices and the existence of at leastsome common criteria of judgement, I am obliged to argue that mypractice is reliable.

16.2.3 An Evidentialist View of BRSP

There exists a final way of arguing against Alston’s view, which con-tests one of his assumptions. This is that there is no non-circularway of defending a doxastic practice such as belief in the reliabilityof sense perception (BRSP). I have argued earlier (14.4) that our ev-eryday assumption of the reliability of sense perception could bebased on a kind of tacit reasoning, namely abductive reasoning.What is important here is that such reasoning need not be circular,as Alston alleges. Those who, like Alston, allege that epistemic circu-

larity is unavoidable in this context make claims like the following:It has been pointed out that reliance on sense perception enables us(1) to successfully predict and thereby (2) to exercise considerable con-trol over the course of events. Furthermore (3) when several indepen-dent investigators use sense perception to explore the physical envi-ronment they generally come up with the same answer. It has beenfelt that these facts testify to the reliability of sense perception. …[But] how do we know that predictions formed on the basis of obser-


Page 164: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 164/205

Religious Experience

vationally based principles are often correct? By looking to see

whether things came out as predicted, or by using instruments to de-termine this, which in turn … How do we know that different ob-servers generally agree on what is before them? By listening to whatthey say. Once more we have to rely on sense perception to gather thedata that are being used in the argument for its reliability. (Alston1986: 6–7)

The problem with this line of argument is that it assumes that senseperception is unitary: that there is just one thing that we are depend-ing on when we depend on “sense perception.” But in fact we have

five senses, and can check the deliverances of one sense by those of another. As Alan Musgrave points out, Macbeth said it all:

Is this a dagger which I see before me,The handle toward my hand? Come, let me clutch thee:I have thee not, and yet I see thee still. Art thou not, fatal vision, sensibleTo feeling as to sight? Or art thou but A dagger of the mind, a false creation,Proceeding from the heat-oppressed brain?(William Shakespeare, Macbeth, Act II, Scene 1) (Musgrave 2009: 12)

What Macbeth is doing here is checking his apparent visual percep-tion of a dagger by attempting to grasp it. When he finds that he can-not touch the dagger that he appears to see, he decides that his visu-al impression is, perhaps, a hallucination. Since he is not relying onvisual perception in order to test visual perception, there exists atleast one non-circular way in which the reliability of sense percep-tion can be tested.


Page 165: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 165/205

Page 166: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 166/205

Chapter Seventeen

Prudential Arguments

I turn now to examine a rather different approach to religious com-mitment. This suggests that a religious commitment – a commitmentto live according to the dictates of a particular faith – can be justifiedon something other than evidential grounds. More precisely, it sug-gests that such a commitment can be based on something other than

a claim to knowledge. The commitment in question may have someevidence in its support; one might want to argue that at the veryleast, it shouldn’t be made in defiance of the facts. But on this view areligious commitment can legitimately go well beyond what the evi-dence warrants. I will examine just one form of this view, which sug-gests that there may be good  prudential grounds for making a reli-gious commitment. The three thinkers whose work I shall examineare Blaise Pascal (1623–62), William James (1842–1910), and ourown contemporary, Richard Swinburne (1934– ). All three write as if 

their arguments supported belief in God. But I shall argue that theysupport something rather different, namely a commitment to live as

if God existed.

17.1 Pascal and His Wager 

The first to articulate a prudential or pragmatic view of religiouscommitment seems to have been the seventeenth-century philoso-pher Blaise Pascal. Pascal articulates his view of faith in the contextof his reflections on reason and religion, which lead him to the con-

clusion that the existence of God cannot be proved by reason. It isnot that there is no evidence in support of God’s existence; Pascal be-lieves that there is. The problem is that it is apparently equally bal-anced by evidence against belief in God. So we are in a situation inwhich reason cannot decide the issue. Are we then free to withholdour decision? No. By the nature of the case, we must be either believ-ers or unbelievers. (Pascal would presumably have considered agnos-


Page 167: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 167/205

Prudential Arguments

ticism a type of unbelief.) So what should we choose? Well, as people

accustomed to weighing the odds, we should make a wager. What areour options? What do we have to gain or lose?

The options, as Pascal offers them, may be set out in the form of atable, as found on the following page. Examining the table, we maybegin to calculate our best bet. Let’s say we put our money (as itwere) on the existence of God by becoming believers. Then if we arecorrect, we gain an infinity of happiness (in heaven) at the cost of afinite amount of pleasure (that which was forbidden or made impos-sible by his faith). If we are wrong, we gain nothing, but all that we

lose is a finite amount of pleasure. In other words, by betting on theexistence of God we have an infinite amount to gain and only a finiteamount to lose. But what if we were to put our money on the non-ex-istence of God, by opting to remain atheists or agnostics? Then if weare correct, we gain a finite amount of pleasure (the pleasure wewould not enjoy as believers). But if we are wrong, we face an eterni-ty of pain. In other words, if we bet on the non-existence of God, wehave a finite amount to gain and an infinite amount to lose. So therational option – rational in the prudential sense of calculating bene-

fits and risks – would be to opt for belief.

God exists. God does not exist.

I believe

that God


1. I live according to a

religious faith, which

involves a finite burden,

but leads to infinite

happiness (heaven).

3. I live according to a

religious faith, which

involves a finite burden,

and death will be the

end of my existence.

I do not


that God


2. I live as an unbeliever,

which brings finite

pleasures, but leads to

an infinity of pain (hell).

4. I live as an unbeliever,

which brings finite

pleasures, and death will

be the end of my



Page 168: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 168/205

Prudential Arguments

Pascal’s argument is surely valid, but it is not clear that it is

sound. In other words, it rests on assumptions both about God andabout belief that are, to say the least, questionable.

Firstly, Pascal assumes a simple choice between unbelief andChristian faith. But Christian faith is not the only choice on offer. Itmay have been the only “live option” – to use William James’s phrase – for a seventeenth-century thinker, but objectively speaking that isneither here nor there. There exist a variety of religious options, avariety of faiths, some of which promise equally horrific conse-quences for those who make the wrong choice of religious faith.

(Some versions of Islam promise hell for Christians as well as foratheists, just as some versions of Christianity promise hell for Mus-lims, and so on.) In this situation, a prudential calculation such asPascal urges becomes indescribably more complex, perhaps impossi-ble.

Secondly, Pascal assumes that a faith chosen on this basis wouldbe acceptable to God. But many people, atheists and believers alike,have found this implausible. To such people William James’s com-ment has seemed apposite: “if we ourselves were in the place of the

Deity, we should probably take particular pleasure in cutting off be-lievers of this pattern from their infinite reward” (James 1966: 91).In other words, it is at least a moot point whether any God worthy of worship would be impressed by the kind of calculating self-interestthat Pascal seems to be urging.

Thirdly, Pascal’s argument assumes that in some sense we canchoose to believe (a topic often dealt with under the heading of “belief voluntarism”). On the face of it, the idea that I can choose to believesomething directly seems implausible. (Give it a go. Try to believe – 

actually believe, not just entertain the idea – that there was a catsitting on my desk as I wrote these words.) But even if I cannotchoose (directly) to believe something, I can choose to act in a waythat will probably give rise to a certain belief. In this situation, I amchoosing to believe something indirectly, by acting as if it were true.In fact, this is precisely what Pascal is suggesting, as the followingpassage reveals.


Page 169: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 169/205

Prudential Arguments

 You would like to attain faith, and do not know the way; you

would like to cure yourself of unbelief, and ask the remedy for it.Learn of those who have been bound like you, and who now stakeall their possessions ... Follow the way by which they began; byacting as if they believed, taking the holy water, having massessaid, etc. Even this will naturally make you believe, and deadenyour acuteness. (Pascal, Pensées §233)

 Yet as that last phrase suggests, this same advice can be read rathercynically. There is evidence that acting in certain ways will, overtime, give rise to a tendency to believe, or even an actual sense of 

conviction. The anthropologist Tanya Luhrmann testifies to this,based on her fieldwork within the Wicca (modern-day witchcraft)movement. As she writes, “the once-non-magician begins to do whatmagicians do, and begins to find magical ideas persuasive because hebegins to notice and respond to events in different ways” (Luhrmann1989: 12). So if you want to become a believer, Pascal’s advice wouldprobably work. But what would I be saying if I were deliberately toadopt it? “I have no reason to think these things are true (and somereason to think they are not), but I am choosing to act in such a way

that I may yet come to believe them.” Is this course of action defensi-ble? If there is, as philosophers fondly imagine, an “ethics of belief,”this hardly seems a praiseworthy way to act.

17.2 William James and the Will to Believe

 A more satisfactory defence of a religious commitment lacking evi-dential justification was provided in 1896 by the American philoso-pher William James. James’s essay is entitled “The Will to Believe.”James’s argument is that there are circumstances in which it is rea-sonable for us to make a decision to believe that is based, not on theevidence available to us, but on what he calls our “non-intellectualnatures” (James 1966: 95). In fact, James notes, non-intellectual fac-tors determine many if not most of our beliefs. We are predisposedtowards certain beliefs and consider others outside the pale (as itwere) because of what we desire and hope for, which is itself partly aproduct of the society in which we live. In many cases we are unable


Page 170: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 170/205

Prudential Arguments

to give entirely satisfactory reasons in support of our most heartfelt

convictions. The idea that such beliefs are the product of evidenceand argument is a kind of rationalist oversimplification. Is this a badthing? Not always, James argues. His argument hinges on the ideathat there are circumstances in which non-intellectual factors maylegitimately sway our judgement. As he writes,

our passional nature not only lawfully may, but must, decide anoption between propositions, whenever it is a genuine option thatcannot be decided on intellectual grounds; for to say, under suchcircumstances, “Do not decide, but leave the question open,” is it-

self a passional decision – just like deciding yes or no – and is at-tended with the same risk of losing the truth. (James 1966: 95)

Such a decision, he argues, is justified under three conditions:firstly, when the options between which we must choose are live op-tions (ones we find to some degree attractive and could in fact em-brace), secondly, when the decision is forced rather than avoidable(where, in a choice between a and b, we lose a chance of possessingthe truth by not choosing) and, thirdly, when what is involved in the

decision is momentous, of real significance for my life, rather thantrivial. (Presumably, another condition – at least implicit in James’sposition – is that the evidence does not decisively favour one positionrather than the other.)

 All three conditions are fulfilled, James argues, in the case of reli-gion, with the further consideration that if the religious hypothesiswere true, it would require precisely the sort of decision here beingadvocated. Why? Because in all relationships between personal be-ings a certain “trusting spirit,” a willingness to take the first step in

faith, is required if a relationship is to be established. Why shouldthis be less so of our relationship with the gods than of our relation-ships with each other? To hold ourselves aloof from religious faith inthe manner of the sceptic is to succumb to the fear of being deceived.Would it not be better to yield to our hope that it might be true? Ineither case we are following an inclination of our passional nature,but the sceptic is acting in such a way as to forever exclude the possi-bility of obtaining this particular kind of truth.


Page 171: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 171/205

Prudential Arguments

17.2.1 Some Difficulties

This is a powerful and in many ways persuasive argument. In my judgement it is second only to Richard Swinburne’s defence of a reli-gious commitment, to which I shall come in a moment. But James’sposition is not without its difficulties. For instance, the claim thatsince the sceptic is also following his passional nature, the believer is justified in doing the same embodies an obvious fallacy. For the scep-tic and the believer are urging two very different courses of action.The sceptic is urging us to proportion our belief to the evidence,while the (Jamesian) believer is urging us to overlook the insufficien-cy of the evidence. The fact that both may be motivated by analo-gous, “passional” concerns is neither here nor there. The question is:Which course of action is the more responsible?

In addition, some of the same considerations offered against Pas-cal’s wager may be offered against James’s argument. In particular,James’s argument, like Pascal’s, assumes that we are being offered achoice between being religious (which James defines in a particularlybroad and inclusive way) and not being religious. But this does not

correspond to the complex reality of human religiosity. The actualchoice is between not being religious and choosing one of a series of actual religions, which offer competing and mutually exclusive vi-sions of reality. Could our “passional nature” legitimately choose oneof these doctrinal schemes in preference to another? James himself,for all his sympathy for religion in general, seems to have been un-able to commit himself to any particular form of religion, regardingmost of the doctrines on offer as “absurd” (Fuller 1996: 634–5).

17.2.2 Commitment without Belief

But the point I wish to focus on here is a different one. For James’sessay raises once again the question of whether we can choose to be-lieve. We may now put this question more acutely, not from the pointof view of psychology but rather that of epistemology. Could suchprudential arguments support belief in the proper sense of the word(that is to say, holding a proposition to be true) or merely someweaker attitude, such as acceptance? Neither Pascal nor James make


Page 172: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 172/205

Prudential Arguments

this distinction, which corresponds to my distinction between believ-

ing in God and living as if God existed. It is quite possible to acceptsomething, to proceed as if it were true, without committing oneself to its truth. If you ask a scientist, for instance, if he believes thequantum theory of radiation to be true, he may squirm a bit. Hewould be more likely to say that he accepts the theory: he believesthat it gives us “some insight into the structures of the world” (seeMcMullin 2002: 275) or at least that it is the best theory available.

There is a lot to be said for regarding a religious commitment inthis way, as a type of acceptance that falls short of an affirmation of 

truth. Robert Audi refers to this as a “non-doxastic faith” (Audi 1991:223). Despite Audi’s valiant attempts to compare his view with thatof Aquinas, it does not correspond to any of our traditional views of faith. In fact, the phrase “non-doxastic faith” appears to be some-thing of an oxymoron. If it is non-doxastic, in what sense is it faith?So if you are to advocate such a commitment, it would be better toavoid the terms “belief” and “faith.” By using such terms, all three of our thinkers – Pascal, James, and Swinburne – muddy the waters.

17.3 Richard Swinburne’s Pragmatic Grounds

 A stronger defence of religious commitment is that offered byRichard Swinburne. Swinburne’s starting point is a general distinc-tion between two different kinds of belief. How are we best to charac-terize a belief that p, he asks, where p is any proposition? It is, Swin-burne argues, best thought of not simply as a belief that  p is true,but as a belief that  p is more probably true than some alternative.Sometimes the alternative may be simply not- p. But on other occa-sions, it will be q, r, or s. For instance, the belief that New Zealandwill win the next rugby World Cup may be a belief that it is moreprobable than not that New Zealand will win. But it might be the be-lief that it is more probable that New Zealand will win than thatSouth Africa or Australia or England will win. In both cases, there isa contrast-class, as it were. But the evidential considerations will bedifferent as the contrast-class varies.


Page 173: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 173/205

Prudential Arguments

This distinction is particularly important then it comes to practi-

cal rationality, that is, reasoning about how I am to act. If I want toachieve a particular goal and I am faced with more than one possiblecourse of action, what would be the rational course of action to fol-low? Presumably that which appears more likely than any other toreach the goal. If I want to drive to Queenstown, for example, andI’m faced with two roads, I should follow that which I believe is morelikely to get me there.

17.3.1 Acting on an Assumption

So far, so good. But Swinburne makes a further point. He arguesthat when I want to reach a goal, but I do not have sufficient evi-dence to believe that any course of action will lead me there, I may justifiably act on an assumption as opposed to a belief . Indeed it maybe reasonable (in the sense of practical rationality) to act on an as-sumption even if I have reason to believe that this assumption is not

true. As Swinburne writes,

a man in an underground cave may believe that none of the sever-al exits lead to the surface. He may nevertheless take a certain

exit, because only by taking some exit has he any chance of achiev-ing his purpose of getting to the surface. We may say of him thatalthough he does not believe that this exit leads to the surface, heis acting on the assumption that it does. To act on an assumptionthat p (or to act as if  p) is to do those actions which you would do if you believed that p. (Swinburne 1981: 31)

It would not be reasonable for me to act in this way if I believedthere was no chance that one of the paths led to the surface. Butwhile there exists some chance (however small) that one of them will

do so, it may be reasonable to act on the assumption that it does. Thegreater my desire to reach the surface – it may be, for instance, be-cause I am trapped – the more reasonable it would be to act on thisassumption. (Although Swinburne does not extent the analogy tothis point, it seems consistent with his argument to do so.)

The parallel to religious faith should be becoming clear. Religions,Swinburne argues, can be regarded as a means of attaining certain


Page 174: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 174/205

Prudential Arguments

goals, which are spelled out in the beliefs of the religion in question.

One of these is, of course, that of attaining salvation, however that isenvisaged. So our choice is which of these differing religious paths tofollow. But to make a reasonable choice, all we need to believe is thatpath  A (e.g. the Christian one) is more likely to be true – and thusmore likely to lead to these goals – than path  B, C ,  D, or E . Swin-burne refers to this as a “weak” as opposed to a “strong” belief. A“strong” belief, by way of contrast, would be the belief that A is morelikely to be true than not- A.

 As it happens, when it comes to belief in God, Swinburne believes

that the evidence warrants a strong belief. In other words, he be-lieves that the existence of God is more probable than not, and hasdevoted a book (The Existence of God) to proving this thesis. But inFaith and Reason he argues that religious faith requires a much low-er evidential threshold. As he writes, “all that is necessary for weakbelief in the Christian Creed is a much less probable belief – say,putting it loosely, that there is a significant probability that there isa God” (Swinburne 1981: 175). How high need this probability be? Apparently not very high, for religious faith is primarily a matter of 

trust in God and this trust can be based on a mere assumption.Swinburne at one point offers the following revealing analogy.

 A man may put his trust in something which, on balance, he doesnot believe to exist. A man in prison may be told that he will berescued by “The Big Chief” from the outer yard of the prison, if hecan get there at night. On balance the prisoner does not believethis rumour; he does not think there is any such Big Chief. Butthe rumour has some plausibility; and the prisoner has no otherhope of escaping. So he steals a file, files away the bars of his cell,

and squeezes through the cell window to get into the outer yard of the prison. He is liable to be punished when all this is discovered,unless by then he has succeeded in escaping. The prisoner is notinappropriately described as putting his trust in the Big Chief.(Swinburne 1981: 167)

Or, as Swinburne writes in his conclusion, “a man who prized salva-tion far above everything else, would pursue it despite having a be-


Page 175: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 175/205

Prudential Arguments

lief that it was very unlikely that pursuit of any religious way would

attain it” (Swinburne 1981: 199).

17.3.2 Critical Comments

There are various criticisms that could be made of Swinburne’s posi-tion. For instance, William Alston (1994) takes issue with Swin-burne’s general conception of belief, arguing that we often form be-liefs (and are acting rationally in doing so) without forming any judgement as to their comparative probability. But Alston also ad-mits that this objection is not necessarily fatal to Swinburne’s view

of  religious belief, which he regards as offering some valuable in-sights. However, the question I wish to focus on is a different one. Itis the question I also asked about the views of Pascal and James,namely: Can this argument vindicate belief in the proper sense – holding a proposition to be true – or does it vindicate somethingweaker, namely living as if a proposition were true?

On Swinburne’s view of faith, the answer given will vary fromcase to case. It will depend on what degree of evidential as opposedto prudential support a particular religious commitment can claim.

To the extent that the commitment in question depends on evidentialconsiderations, it can support belief in the proper sense. If I believethat  p (and act on that belief) because the Bible tells me that  p istrue and I have good reasons to believe that the Bible is a reliable au-

thority, then there is nothing to distinguish my religious faith fromany other form of justified belief, in this case, belief on the basis of testimony. Such a belief is defensible if (and only if) I have good rea-sons to believe that the Bible is reliable. But insofar as Swinburne isdefending a religious commitment in the absence of adequate eviden-

tial considerations, he cannot claim that it yields knowledge. It maybe rational – in the sense of practical rationality – to act as if  p weretrue even in the absence of sufficient reasons to believe that it is. Butin these circumstances I cannot say that I know  p to be true.

Of the various accounts of faith which I am surveying, Swin-burne’s is surely the most attractive to the philosopher, of whateverreligious persuasion. Even more than William James’s, it would ap-


Page 176: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 176/205

Prudential Arguments

pear to lend support to the rationality of making a particular reli-

gious commitment. But we should also keep in mind the all-too-hu-man tendency (on which Pascal relies) to pass from acting as if some-thing were true to forming the conviction that it is true. Religious be-lievers of this stripe may begin with the best of intentions, keepingin mind that their religious commitment is based largely on pruden-tial rather than epistemic grounds. But it is all too easy for them toforget this and, over time, to claim a degree of certainty in matters of faith which cannot be defended. In light of this tendency, Swin-burne’s account, like Pascal’s, looks like an invitation to systematic

self-deception. In a word, the course of action which Swinburne de-fends is fraught with dangers, dangers of which he seems insuffi-ciently aware.


Page 177: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 177/205

Page 178: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 178/205

Chapter Eighteen

The Kantian As If

 A subtly different but not unrelated approach to religious faith isthat associated with the Immanuel Kant. Kant’s view of religiousfaith does not correspond to any of the positions we have studied. Herejects the force of the traditional arguments (ontological, cosmologi-cal, and teleological) for the existence of God, on broadly sceptical

grounds. His reading of David Hume, he tells us (1977: 5 [4.260]),“interrupted” his “dogmatic slumber” by alerting him to the limits of speculative reasoning. Kant’s first Critique  –  that of  Pure Reason  – was dedicated to demonstrating these limits. It argues that when wetry to deal with matters that lie outside the phenomenal world of space and time, purely theoretical reason fails. It can make no sub-stantive claims about realities which lie beyond all possible experi-ence. It is this view that leads Kant to be severely critical of the tra-ditional proofs of the existence of God.

 You might expect this to lead Kant to an agnosticism that resem-bled Hume’s. But it does not. His strategy, as it puts it, was that of “denying knowledge, in order to make room for faith” (Kant 1929: 29[Bxxx]). What Kant wishes to defend is a very particular kind of faith, one based on what he calls “practical reason.” In Kant’s secondcritique – the Critique of Practical Reason – he argues that the exis-tence of God, along with that of freedom and immortality, is one of the three “postulates” of practical reason. What does this mean? Is ittrue? And what would implications would it have for the rationality

of a religious commitment?

18.1 The Autonomy of Ethics

Practical reason, as Kant understands it, has to do with how weought to act, that is to say, with morality. But to assert a connectionbetween morality and religion is to invite misunderstanding. Peopleoften assume that we require religious sanctions in order to act


Page 179: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 179/205

The Kantian As If

morally, or that we require a divine revelation in order to know how

to act morally. But neither view seems defensible. To do the rightthing in anticipation of a reward or out of fear of punishment seemsa poor kind of moral action. Virtue, as we commonly say, ought to beits own reward. And if we tie morality too closely to the will of God,this leads to further problems. If we say, for instance, that we knowwhat is moral only by way of some divine revelation – perhaps theten commandments – then what about those humans who, throughno fault of their own, have not heard of this revelation? Are they un-able to act morally? And if we say that what makes something

morally right is that it conforms to God’s command, then what if Godcommanded us to torture small children? Would that be morally ac-ceptable, under these circumstances? Philosophical discussion of thatissue goes back to Plato’s Euthyphro dilemma, the central questionof which is “Do the gods love holiness because it is holy, or is it holybecause the gods love it?”

For the most part, Kant argues vigorously for the independence of morality from religion in both of the above senses. Morality, he ar-gues, does not need religion to provide us with the motivation we

need to live well. Nor does it need religion to spell out the content of the moral law. Obedience to the moral law should be motivated bynothing other than the recognition of its demands; practical reasonalone can tell us what those demands are. A key principle here isKant’s categorical imperative: “act only according to that maxim bywhich you can at the same time will that it should become a univer-sal law” (Kant 1959: 39 [4.421]). This is the measure of whether ornot something is morally acceptable. But although morality does notneed religion, there remains a very close connection between morali-

ty and religion. It is to be found in what morality presupposes, whatit takes for granted. What does morality take for granted? Kant ar-gues that it takes for granted three things: freedom, immortality,and the existence of God. These are what Kant refers to as the postu-lates of practical reason.


Page 180: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 180/205

The Kantian As If

18.2 The Postulates of Practical Reason

The first and most obvious of these postulates is freedom. The morallaw only makes sense if human being are, at least in part, indepen-dent of the cause–effect relationships that constitute the worldknown by science. It makes sense to say that this is how we ought toact only if we have a choice about how we act. Or to put this anotherway, as moral agents we must act as if we are free even if we do notknow that we are free. (Freedom is here being understood in thestrong, libertarian sense.) What about immortality and the existence

of God? Well, they are postulates in a derived, secondary sense. AsKant writes, the idea of God and that of the immortality of the soul

are not conditions of the moral law but only conditions of the nec-essary object of a will determined by the law, that is, of the merepractical use of our pure reason; hence with respect to these ideaswe cannot affirm that we cognize and have insight into – I do notmerely say the reality but even the possibility of them. But theyare, nonetheless, conditions of applying the morally determinedwill to its object give to it a priori (the highest good). Consequently

their possibility in this practical relation can and must be as-sumed, although we cannot theoretically cognize and have insightinto them. (Kant 1997: 4 [5:4]).

That is, I’m sorry to say, about as clear as Kant’s writing ever gets.So let’s see if we can unpack the meaning of that difficult passage.

What Kant appears to be saying is that we must assume the reali-ty of immortality and of the existence of God because otherwisemoral action would be in vain. Why is this? Well, moral action has anatural goal; it is, of its very nature, directed toward the achieve-

ment of some particular state of affairs. Kant refers to this as the“highest good.” Note how careful Kant has to be here. He does notwant to say that the highest good is the motivation of our moral ac-tions, since this would endanger his insistence on the autonomy of morality. Kant insists that “morality really has no need of an end[goal] for right conduct” (Kant 1998: 34 [6:4]). What he appears to besaying is that the highest good is the state of affairs that would be


Page 181: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 181/205

The Kantian As If

created as a result of acting in accord with the moral law. What is

that state of affairs? For the individual, it is the combination of “virtue and happiness”; for the world as a whole, it is “happiness dis-tributed in exact proportion to morality” (Kant 1997: 93 [5:110]). Ourmoral action, Kant argues, take for granted the possibility of achiev-ing such a goal.

The key question then becomes: What are the conditions underwhich the highest good could be achieved? The fulfilment of virtue isholiness, a complete conformity of the individual’s dispositions withthe moral law. But no one is capable of this complete conformity: it is

an ideal, towards which we strive, but only (if you like) “asymptoti-cally.” This aspect of the highest good therefore requires immortali-ty, as the condition of an endless striving towards holiness. The high-est good also involves happiness in proportion to virtue. But there isnothing corresponding to this is nature, so the only way it could beachieved is by the actions of a supernatural being, namely God. Itfollows that “it is morally necessary to assume the existence of God”(Kant 1997: 105 [5:125]).

18.3 A Kantian View of Faith

The “morally necessary” assumption of the existence of God under-girds what Kant understands as religious faith, a faith that respectsthe limits of human reason. Does this faith involve a claim to knowl-edge? Can the believer be said to know that God exists? Apparentlynot, since the theoretical reasoning that would support such a con-clusion cannot do the job. Assuming that the idea of God is coherentand that belief in God is not inconsistent with any other of our justi-fied beliefs, then theoretical reason cannot demonstrate God’s non-

existence. This is a vital point, since the practical rationality of reli-gious faith would be undermined if there existed logical problems of this kind. But for the reasons given in the first Critique, we cannotbe said to have an objective justification of belief in God. The justifi-cation of our belief in God is practical not theoretical; it is our needto act as if God existed.


Page 182: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 182/205

The Kantian As If

William Lad Sessions sets out Kant’s view of the practical justifi-

cation involved in a more formal fashion. Where S is some agent (or“subject”), then

S ’s belief that p is practically justified [when]

a. S believes  p [the belief in question excluding the possibility of knowledge]

b.  p’ s truth is a logically necessary condition for the achievementof some logically possible end, A, of S,

c.  A is practically necessary for S , and

d. S ’s grounds for (a) are (b) and (c). (Sessions 1980: 464)

18.4 A Brief Evaluation

Kant’s view of religious faith differs in important ways from otherdefences of what I am calling “living as if God existed,” such as thoseof Pascal, James, and Swinburne. Not only is it more upfront in itsrenunciation of any claim to knowledge, but it does not represent thereligious attitude as a choice. We may have a choice about whetherto act morally – if we are free, then presumably we do have such achoice – but if we act morally we have no choice but to act as if Godexisted. This may or may not issue in an explicit religious faith, al-though for the rational subject, it ought to do so. One could arguethat on a Kantian view of religion even an atheist could be said to beliving a devout life, if she obeys the moral law, although Kant him-self would surely disagree: he makes very critical remarks aboutatheism.

Let me offer just one critical comment. Kant assumes that moralaction would be pointless unless the highest good can be achieved.But this seems wrong, as J. L. Mackie points out. Even given Kant’sview of morality, we do not need to assume that “the full realization”of the highest good is possible (Mackie 1982: 109). All we need to as-sume is that we can go some distance towards achieving it. And it isat least arguable that there is no need to posit the existence of Godin order to make this less demanding assumption.


Page 183: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 183/205

Page 184: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 184/205

Chapter Nineteen

The Wittgensteinian View

In the decades following his death, the followers of Ludwig Wittgen-stein (1889–1951) developed a distinctive approach to the philosophyof religion. This “Wittgensteinian” approach to religion was particu-larly popular in the 1950s and 1960s. It no longer receiveswidespread support. But it still has its proponents, the best-known of 

whom is D. Z. Phillips. It is of interest here for a couple of reasons.Firstly, it takes us further in the direction of a thoroughly non-real-ist interpretation of religious language, although in a way thatseems confused. Secondly, it has much in common with what peoplesometimes describe as “postmodernism.” As an intellectual move-ment, postmodernism may now be all but dead, but the epistemic rel-ativism which it represented was already present in some versions of Wittgenstein’s philosophy, and it remains an ever-present dangeragainst which we should be forewarned.


Page 185: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 185/205

19.1 A Preliminary Statement

What do I mean by the Wittgenstein view of religion? A good place to

start is with a negative statement of the position, since it is easier tosay what proponents of this position reject than to say what they af-

 firm. A useful starting-point is the short article (actually an extend-ed book review) published in 1962 by George Hughes, for many yearsProfessor of Philosophy at Victoria University in Wellington. Hughesis taking issue with a view put forward by C. B. Martin, that reli-gious concepts are not so much false as simply confused. Hughes’s re-sponse is that any set of statements would appear confused if onewere to judge them by way of an alien logic, applying criteria of con-

sistency and coherence that are not appropriate in this particularfield. If religion is to be properly understood, Hughes argues, wemust “allow the actual use of religious terms and statements to de-termine their logic, rather than trying to force an alien logic uponthem” (Hughes 1962: 214).

What does this mean? I shall assume that Hughes is not sayingthat religious thought is exempt from the requirements of logic, inthe sense of the principles of correct reasoning. He was, after all, alogician himself, the co-author of a well-known textbook on modal

logic. I am assuming that he intends “logic” here to be taken in awider sense, namely the criteria appropriate for evaluating state-ments in a particular field. Hughes’s complaint is that critics of reli-gion often apply inappropriate standards of criticism.

Now it seems beyond dispute that different categories of proposi-tion should be evaluated in different ways. If someone said, “Abor-tion is evil” and I were to respond by asking for biological evidence, Iwould be committing a category mistake. Ethical statements, what-ever they do express, do not express the kinds of facts that can be

discovered under a microscope. (Of course, things that can be discov-ered under a microscope might contribute to shaping our ethical judgements, but that’s another matter.) But one can give an accountof how it is that ethical assertions differ from statements regardingobservable facts. As you would expect in philosophy, there exist arange of such accounts and a subject, called metaethics, devoted toexamining them. So it would seem reasonable to ask the Wittgen-steinian for an account of how religious propositions differ from, forinstance, those found in the sciences. Critics of religion, for instance,

Page 186: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 186/205

assume that the proposition “God exists” is a statement of fact: it as-serts (among other things) that there is an omnipotent, omniscientand benevolent being who created the world. As an alleged state-ment of fact, it can be either true or false. But if “God exists” is nottrying to state a fact, it would be obviously inappropriate to treat itas if it were. So what it is doing?

19.2 D. Z. Phillips on Religion

For an answer to this question we may turn to the work of D. Z.Phillips, who more than any other philosopher has developed and de-fended a Wittgensteinian approach to religion.

19.2.1 Meaning and Use

 At the heart of Phillips’s position is the idea that the meaning of a

religious utterance is identical with (or can be understood as identi-cal with) the role that it plays in the life of the believer. Let’s take anexample. Phillips holds that “metaphysics, magic and religion aremisunderstood if they are treated as mistakes or blunders concern-ing the facts” (Phillips 1976: 111). (As it happens, Phillips believesthere are significant differences between metaphysical statementsand magical and religious ones, but that need not concern us here.)With regard to religious statements, what it might mean to say thatGod cares for us all the time (at least when such an utterance is “notsuperstitious”)? The answer, Phillips writes, depends on the contextof the utterance.

What believing that God cares for us all the time amounts tovaries with different circumstances. The circumstances bring outthe force of the belief... Suppose one witnesses a child falling over-

board. One is unsure of one’s chances of saving him or of survivingthe effort to do so. Suddenly, one may say to oneself, “Jump! Trustin God!” This expression need not be connected with a belief thatsome supernatural agency is going to guarantee the safe return of either the child or oneself. No, it is a matter of not putting oneself first, weighing up pros and cons. One gives oneself to what has tobe done. It is this giving of oneself without reserve – trusting it – which gives force to the expression “Trust in God” in this context.(Phillips 1976: 111–12)

Page 187: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 187/205

Now there are two ways of understanding this idea. At times,Phillips seems to be saying that religious language should never beconstrued as making factual claims. It is never used to refer to a su-pernatural being who might be expected to save you. But a weakerinterpretation of his words is possible. It is that religious languageneed not be used to make factual claims. Yet it may be that some-times it is understood in this way, not only by critics of religion, butby believers themselves.

It is the second of these two interpretations that seems more accu-rately to express Phillips’s view. The expression “Trust in God,” weare told, need not be understood as referring to a supernatural agent.What Phillips is interested in spelling out is how religious utterancesmight be used in a manner other than to state facts. So despite hisprotestations, Phillips is not merely engaged in the descriptive taskof asking “how language is in fact used” (Phillips 1976: 41). He is en-gaged in the  prescriptive task of specifying which uses of religiouslanguage can be regarded as meaningful. His conclusion? That reli-gious utterances may be meaningful when used expressively, as indi-cating a particular attitude. But they make no sense when under-stood factually, as assertions about supernatural entities.

19.2.2 Religion and Superstition

The mention of “superstition” in this context is telling. Phillips’s in-terpretation, he himself notes, assumes that the utterance in ques-tion is not superstitious. What is he suggesting here? Some peoplemay in fact understand “Trust in God” to imply that there exists asupernatural agent who will come to your aid. But, Phillips suggests,such a usage would be superstitious rather than religious. This be-comes clear in another passage in the same book, where Phillips dis-cusses the “intellectualist” theory of religion put forward by the fa-

ther of modern anthropology, E. B. Tylor. He has already rejectedTylor’s idea that religious beliefs are “mistaken hypotheses.” But onthe last page he notes:

nothing which has been said in this chapter has meant to denythat practices called magical or religious may be shown to be su-perstitious. One could imagine circumstances where almost all theactions we have described could bear a significance which wouldmerit that description. The point is that there is a wide range of examples where this cannot be said. (Phillips 1976: 41)

Page 188: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 188/205

If Phillips were making the strong claim that religious utterancesare never understood to make factual claims about supernatural be-ings, this would be evidently false. Many religious people do under-stand religious utterances in this way. But in fact he is making theweaker claim that religious utterances are sometimes used in amerely “expressive” way, to indicate (for instance) one’s attitude to-wards a particular course of action.

19.2.3 The Functions of Language

The problem here is that even Phillips’s weaker claim seems mis-leading. Phillips writes as though the only thing a religious utter-ance (properly so called) does is to express a particular attitude. But

this is far too narrow a conception of meaning. Utterances which con-vey very different meanings may serve the same expressive function.To take the example given above, faced with the need to save thechild, an atheist would presumably not say to himself “Trust in God!”He might think “Go on! You can do it!” or even “Come on! Show thoseChristians that atheists can act altruistically too!” These three utter-ances may all have the same function: that of committing the personunreservedly to a course of action. But of course they have very dif-ferent meanings and, more importantly, represent very different mo-

tivations. Phillips’s practice of interpreting a religious utterance interms of  just one of the roles it plays in the life of the believer ob-scures these differences.

But it gets worse. At least some of the religious utterancesPhillips discusses are propositional in form. They have the form of statements that make assertions about reality. (“Trust in God!” is animperative, but “It is God’s will” at least has the appearance of astatement of fact.) Now it is of course true that such utterances mayhave many functions. They may have expressive functions, as

Phillips suggests. They may even have performative functions, in thesense of bringing about the states of affairs to which they refer. (Sec-ular examples include “I bid five dollars” pronounced at an auction or“This bridge is now open” pronounced by a visiting dignitary. A reli-gious example would be “I absolve you from your sins” when pro-nounced by a Roman Catholic priest.) But such utterances are notwithout descriptive content. Indeed their expressive or performativefunctions are dependent on their descriptive content. It is because “Ibid five dollars” has descriptive content of a very specific sort (which

Page 189: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 189/205

is different from “I bid three dollars”) that it can function as it doesin an auction. It is because “It’s God’s will” has descriptive contentthat it can function expressively in the life of the believer. To set ex-pressive and descriptive uses of language in opposition is not only a“false antithesis” (Mackie 1982: 223); it ignores the way the first isdependent on the second.

Page 190: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 190/205

Chapter Twenty

Theological Non-Realism

Let’s turn to a final form of religious commitment “beyond belief.” Itis the view that while the proposition “God exists” is almost certainlyfalse, it is a useful fiction, a “myth we have to have” (Cupitt 1980:166). I shall describe this view as theological non-realism, since it re-sembles the non-realist views of the theoretical entities of science

held by Bas van Fraassen and others. It is a theological non-realisminsofar as it argues that there are no supernatural entities such asgods, angels and demons. The language which apparently refers tosuch things should be taken to have a different function. The sameposition is sometimes referred to as religious non-cognitivism, where“cognitive” means “pertaining to knowledge.” For my purposes, Ishall regard these two terms as interchangeable.

The best known contemporary proponent of theological non-real-ism is Don Cupitt. Cupitt’s starting point is is a vigorous critique of 

the evidential basis of traditional Christian belief. Traditional the-ism, Cupitt argues, is no longer intellectually defensible. We do nothave sufficient reason to believe that there exists a God who corre-sponds to the description offered by our monotheistic traditions. AsCupitt puts it, “we are and have been for many generations in a posi-tion where not one single religious doctrine (of the sort that mentionssupernatural beings, events and causes) can be established by a rep-utable intellectual method. Not one” (Cupitt 1980: 82).

20.1 The Value of Religion

If this is true, why not be an atheist simpliciter? Because, Cupitt ar-gues, there are values that our religious traditions embody whichcan continue to form the basis of a much-needed spirituality. Whatare these values? Cupitt lists six.


Page 191: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 191/205

Theological Non-Realism

It is good that one should appraise oneself and one’s life with an un-

conditional religious seriousness that tolerates no concealment orself-deception.

It is good that one should cultivate meditation and contemplativeprayer, and especially the inner fortitude and resilience needed tocombat evils of all kinds.

It is good that one should come to transcend the mean defensive egoand learn absolute disinterestedness and purity of heart.

It is good that one should commit oneself to existence in religioushope and receptivity to grace.

In spite of all the ugliness and cruelty in the world, it is good that oneshould at least sometimes experience and express cosmic awe,thanksgiving and love.

It is good that such values as these should not only be cultivatedin and for oneself, but that they should shape our attitudes to-wards other people and be expressed in our social life. (Cupitt1980: 82)

Insofar as they can help preserve and encourage these attitudes, Cu-pitt argues, religious practices should be maintained. But they

should be reinterpreted in non-realist sense.

20.2 The Function of God-Talk 

What does the term “God” mean, from a non-realist perspective?What does it refer to, if it is not an omnipotent, omniscient, morallyperfect being? The term “God” functions for Cupitt as a kind of asymbol. What we say about God is not true of any supernatural be-ing. We may speak of God as if he were a being, but to do so is tospeak mythically. Now if a myth embodies a falsehood, it might be

argued that we should simply abandon it. But Cupitt argues thatthis particular myth is a necessary myth, one that we need to have.Why do we need this myth? We need the myth of God because thetraditional attributes of God point to the requirements of the spiritu-al life. “The doctrine of God,” as Cupitt puts it, “is an encoded set of 

spiritual directives” (Cupitt 1980: 101). So the function of the term“God” is that of “a unifying symbol that eloquently personifies and


Page 192: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 192/205

Theological Non-Realism

represents to us everything that spirituality requires of us” (Cupitt

1980: 9).Cupitt’s view may be made clear by way of one of his own exam-

ples, namely the divine attribute of omniscience. Cupitt, of course,would insist that there is no omniscient being. But the doctrine of di-vine omniscience tells us something about ourselves; more precisely,it tells us something about the requirements of the spiritual life.What is the spiritual life, for Cupitt? It has to do with human auton-omy, with our ability to act freely and generously, having been freedfrom the superficial cravings that so easily dominate our life. Or as

he puts it,

the highest and central principle of spirituality (the religious require-ment...) is the one that commands us to become spirit, that is, precise-ly to attain the highest degree of autonomous self-knowledge and self-transcendence. To achieve this we must escape from “craving” or “car-nal lusts” and the false ego thereby created, and we must seek perfectpurity of heart, disinterestedness, quiet and recollected alertness, andso on. (Cupitt 1980: 9)

To achieve this state requires a constant self-scrutiny. No aspects

of one’s life are to be keep hidden: all must be brought into the open,examined, and set right. It is this truth – a truth about the spirituallife, not about a supernatural being – that is symbolized by the doc-trine of divine omniscience. Cupitt makes the point that in the bibli-cal writings, divine omniscience is not a matter of God “accumulatinginformation for its own sake” (Cupitt 1980: 86). Rather, God’s knowl-edge of us is primarily practical and ethical. He knows not merelyour external behaviour but the secrets of our hearts.

20.3 The God of Abraham, Isaac, and Jacob

On the one hand, Cupitt is quite open about what he is doing here.He accepts that his non-realist reinterpretation of talk of the super-natural is just that, a deliberate reinterpretation of language that inthe past was often understood in a realist way. In Cupitt’s words,

the realists will point out that religious language in prescientifictimes undoubtably did have a cosmological and descriptivist strain


Page 193: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 193/205

Theological Non-Realism

in it. Of course it did. Modern realism naturally appeals to and

builds upon that observation, and I must admit that in so far as Iam abandoning such claims as no longer defensible I am departingfrom tradition. In mitigation of this confessed heterodoxy, though,I plead that there is no merit in clinging to untruths. (Cupitt 1980:125)

Here Cupitt differs from the Wittgensteinians, who sometimesspeak as though language about God were never intended to be de-scriptive and referential. But that Wittgensteinian claim is clearlyfalse. To admit that such language was originally intended to be de-

scriptive, but that we cannot use it that way – this seems a muchmore defensible option. While Cupitt is reinterpreting the theistictradition, he is not misinterpreting it, as the Wittgensteinians so of-ten appear to be doing.

On the other hand, Cupitt insists that his non-realist views arenot entirely new. They can find some support within the Christiantradition. This claim can be understood partly as a strategic move. If a theologian is to persuade his fellow-believers to adopt the newstandpoint for which he is arguing, he must stress the continuity be-

tween what we might call the old faith and the new (Cupitt 1980:75). But Cupitt’s claim to be in partial continuity with the traditionalso has some basis in fact. Look again at Cupitt’s comments on thedoctrine of divine omniscience. What he is highlighting is the gap be-tween what Pascal called “the God of the philosophers” and “the Godof Abraham, Isaac, and Jacob.” The concept of God that is actuallyemployed in religious circles is not identical with the “theologicallycorrect” description of God that lies at the heart of philosophical the-ism (Barrett and Keil 1996: 219–47). As Cupitt suggests, the God of 

the Bible has just those attributes that the biblical authors felt wererequired for him to fulfil his religious functions. The remaining at-tributes were added by philosophically-minded believers, who judgedthat these were the attributes such a deity required if he were to ful-fil his religious function. A God who could scrutinize the humanheart must by that very fact be able to know other things as well, tothe point of omniscience. But much of what the philosophical theolo-


Page 194: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 194/205

Theological Non-Realism

gian thereby affirms about God is religiously redundant. It has no re-

ligious function.If one no longer needs to affirm that God exists, the religiously re-

dundant attributes of God can be stripped away. One can return to aconcept of God that is closer to the God of Abraham, Isaac, and Ja-cob, although construed in a non-realist fashion. If you are affirmingthe existence of an actual entity, then your statements about himneed to be consistent. And they need to be reconcilable with the restof our knowledge. But if you are using the term “God” as a symbol of the spiritual life, these requirements fall away. You can enjoy a

great freedom in your use of religious language. Our concept of Godneed not be coherent, if we are using it merely instrumentally.

20.4 Why Not Simple Atheism (or Secular 


There is something to be said for Cupitt’s position, but it does raisesome critical questions. The first is whether theological non-realismis a psychologically tenable position. To whom would theological non-

realists pray and would their God-language maintain its motivation-al power? And if they continue with religious practices, would theynot run the danger of deceiving others? More seriously, perhaps,would they not be in danger of deceiving themselves? After all, asPascal noted, religious practices have a way of leading to religiousbelief, a belief not based on any evidence but on the mere psychologi-cal power of the practices. Perhaps the adoption of a theologicallynon-realist position would be worth these risks, if the practice of reli-gion achieved some goals which could not be achieved by some other

means. But is this true? Would not an openly avowed atheism, orperhaps (more positively) what E. O. Wilson (2005) has called “sci-entific humanism” be preferable?


Page 195: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 195/205

Page 196: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 196/205

Works Cited

 Adler, Jonathan E. 1990. “Conservation and Tacit Confirmation.” Mind

99: 559–70.

 ——— . 2002. Belief’s Own Ethics. Cambridge, MA: MIT Press.

 Alston, William P. 1967. “Religion.” The Encyclopedia of Philosophy

edited by Paul Edwards, 7. 140–45. New York: Macmillan.

 ——— . 1986. “Epistemic Circularity.” Philosophy and Phenomenologi-

cal Research 47: 1–30.

 ——— . 1991.  Perceiving God: The Epistemology of Religious Experi-

ence. Ithaca, NY: Cornell University Press.

 ——— . 1994. “Swinburne on Faith and Belief.” Reason and the Chris-

tian Religion: Essays in Honour of Richard Swinburne edited by AlanG. Padgett, 21–37. Oxford: Clarendon Press.

 Anselm of Canterbury. 1995. Monologion and Proslogion with the

Replies of Gaunilo and Anselm translated by Thomas Williams. Indi-

anapolis, IN: Hackett Publishing. 

 Aquinas, Thomas. 1920 [= ST : Summa Theologiae]. The “Summa Theo-

logica” of St Thomas Aquinas translated by the English DominicanFathers. London: Burns, Oates & Washbourne.

 ——— . 1924 [= SCG]. The Summa Contra Gentiles translated by theEnglish Dominican Fathers. London: Burns, Oates & Washbourne.

 ——— . 2004 [=  De Veritate]. Quaestiones Disputatae de Veritate q.14,translated by Alfred J. Freddoso, University of Notre Dame. .htm&aquinas.5 .html 

 Audi, Robert. 1991. “Faith, Belief, and Rationality.”  Philosophical Per-

spectives 5: Philosophy of Religion edited by James E. Tomberlin,213–39. Atascadero, CA: Ridgeview Publishing.


Page 197: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 197/205

Works Cited

Barrett, Justin L. and Frank C. Keil,. 1996. “Conceptualizing a Nonnat-

ural Entity: Anthropomorphism in God Concepts.” Cognitive Psychol-

ogy 31: 219–47.

Bishop, John and Imran Aijaz. 2004. “How to Answer the de jure Ques-tion about Christian Belief.” International Journal for Philosophy of 

Religion 56: 109–29.

Calvin, John. 1961 [= ICR]. Institutes of the Christian Religion translat-ed by Ford Lewis Battles. Library of Christian Classics. Vols. 20 &21. London: SCM.

Clark, Peter B. and Peter Byrne. 1993. Religion Defined and Explained.London: Macmillan Press & New York: St Martin’s Press.

Cohen, L. Jonathan. 1989. “Belief and Acceptance.” Mind 98: 367–89.

Collingwood, R. G. 1961. The Idea of History [1946]. Oxford UniversityPress.

Collins, Robin. 2003. “Evidence for Fine-Tuning.” God and Design: The

Teleological Argument and Modern Science edited by Neil A. Manson,178–99. London: Routledge.

Craig, William Lane. 1980. The Cosmological Argument from Plato to

Leibniz. Library of Philosophy and Religion. London: Macmillan.

 ——— . 1986. “The Problem of Miracles: A Historical and TheologicalPerspective.” Gospel Perspectives. Vol. 6: The Miracles of Jesus editedby David Wenham and Craig Blomberg, 9-48. Sheffield: JSOT Press.

 ——— . 1989. “‘No Other Name’: A Middle Knowledge Perspective onthe Exclusivity of Salvation through Christ.” Faith and Philosophy 6:

172–88.Craig, William Lane and Quentin Smith. 1993. Theism, Atheism, and

 Big Bang Cosmology. Oxford University Press.

Cupitt, Don. 1980. Taking Leave of God. London: SCM.

Daniel, Norman. 1993. Islam and the West: The Making of an Image

[1960]. Oxford: Oneworld.


Page 198: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 198/205

Works Cited

Davidson, Donald. 1985. “Preface.”  Actions and Events: Perspectives on

the Philosophy of Donald Davidson edited by Ernest LePore and Bri-an P. McLaughlin, ix. Oxford: Basil Blackwell.

Dawkins, Richard. 1988. The Blind Watchmaker. [1986.] London: Pen-guin.

Denzinger, Henricus and Adolfus Schönmetzer S.J. 1976 [= DS]. Enchi-

ridon Symbolorum Definitionum et Declarationum de Rebus Fidei et

Morum. 36th edition. Freiburg im Breisgau: Herder.

Dietl, Paul J. 1968. “On Miracles.” In Contemporary Philosophy of Reli-

 gion edited by Steven M. Cahn and David Shatz, 146–53. New York:Oxford University Press.

Dijksterhuis, Ap and Loren F. Nordgren. 2006 “A Theory of UnconsciousThought.” Perspectives on Psychological Science 1: 95–109.

Dostoevsky, Fyodor. 1994. The Karamazov Brothers [1880] translatedby Ignat Avsey. World’s Classics. Oxford University Press.

Draper, Paul. 1989. “Pain and Pleasure: An Evidential Problem for The-ists.” Noûs 23: 331–50.

Durkheim, Émile. 1915. The Elementary Forms of the Religious Life

[1912] translated by Joseph Ward Swain. London: George Allen &Unwin.

Earman, John. 1993. “Bayes, Hume, and Miracles.” Faith and Philoso-

 phy. 10: 293–310.

 ——— . 2000. Hume’s Abject Failure: The Argument Against Miracles.

Oxford University Press.

Feldman, Richard. 2000. “The Ethics of Belief.”  Philosophy and Phe-

nomenological Research. 60: 667–95.

Flew, Antony. 1962. “Are Ninian Smart’s Temptations Irresistible?” Philosophy 37: 57–60.

 ——— . 1966. God and Philosophy. Philosophy at Work. London:Hutchinson.


Page 199: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 199/205

Works Cited

 ——— . 1967. “Miracles.” In The Encyclopedia of Philosophy edited by

Paul Edwards, 5: 346–53. New York: Macmillan.

 ——— . 1972. “The Presumption of Atheism.” Canadian Journal of 

 Philosophy. 2: 29–53.

 ——— . 1997. “The Presumption of Atheism.” In A Companion to Phi-

losophy of Religion edited by Philip L. Quinn and Charles Taliafer-rro, 410–16. Blackwell Companions to Philosophy. Oxford: Blackwell.

Fosl, Peter S. 1997. “The Moral Imperative to Rebel Against God.” Cogi-

to 11: 159–168.

Fuller, Robert C. 1996. “‘The Will to Believe’: a Centennial Reflection.” AAR Review Essay. Journal of the American Academy of Religion 64:633–50.

Garrigou-Lagrange, Reginald, O.P. 1965. The Theological Virtues: Vol. 1

 – On Faith: A Commentary of St Thomas’s Theological Summa Ia

IIae qq. 62, 65, 68: IIa IIae, translated by Thomas a Kempis ReillyO.P. St Louis, MI: B. Herder.

Gaskin, J. C. A. 1988. Hume’s Philosophy of Religion. 2nd edition. Li-brary of Philosophy and Religion. London: Macmillan.

Geertz, Clifford. 1993. The Interpretation of Cultures [1973]. London:Fontana.

Gettier, Edmund L. 1963. “Is Justified True Belief Knowledge?”  Analy-

sis 23: 121–23.

Harrison, Jonathan. 1999. “The Impossibility of ‘Possible’ Worlds.” Phi-

losophy 74 : 5–28.

Hervé, J. M. 1935. Manuale Theologiae Dogmaticae. Paris: Berche etPagis.

Hick, John. 1966. Evil and the God of Love. London: Macmillan.

Himma, Kenneth Einar. 2001. “The Ontological Argument.” In The In-

ternet Encyclopedia of Philosophy.


Page 200: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 200/205

Works Cited

Horton, Robin. 1993. Patterns of Thought in Africa and the West: Essays

on Magic, Religion and Science. Cambridge University Press.

Hughes, G. E. 1962. “Critical Notice: Religious Belief by C. B. Martin.” Australasian Journal of Philosophy 40: 212–19.

Hume, David. 1902. Enquiries Concerning Human Understanding and

Concerning the Principles of Morals [1777] edited by L. A. Selby-Bigge. 2nd edition. Oxford: Clarendon Press.

 ——— . 1993. Principal Writings on Religion, including Dialogues Con-

cerning Natural Religion [1779] and The Natural History of Religion

[1757]. Oxford World’s Classics. Oxford University Press.

James, William. 1966. Essays in Pragmatism. Hafner Library of Clas-sics. New York: Hafner.

Kant, Immanuel. 1929. The Critique of Pure Reason [1787] translatedby Norman Kemp Smith. London: Macmillan Press.

 ——— . 1959. Foundations of the Metaphysics of Morals [1785] translat-ed by Lewis White Beck. Library of Liberal Arts. New York: Macmil-lan.

 ——— . 1977. Prolegomena to Any Future Metaphysics [1783] translat-ed by Paul Carus; revised by James W. Ellington. Indianapolis:Hackett.

 ——— . 1997. Critique of Practical Reason [1788] translated by MaryGregor. Cambridge Texts in the History of Philosophy. CambridgeUniversity Press.

 ——— . 1998. Religion within the Boundaries of Mere Reason and Other

Writings translated by Allen Wood and George di Giovanni. Cam-bridge Texts in the History of Philosophy. Cambridge UniversityPress.

Kenny, Anthony. 1979. The God of the Philosophers. Oxford: ClarendonPress.


Page 201: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 201/205

Works Cited

Leibniz, Gottfried Wilhelm. 1995. “The Principles of Philosophy, or, the

Monadology” [1714]. In  Philosophical Essays edited by Roger Ariewand Daniel Garber, 213–25. Indianapolis: Hackett Publishing.

Le Poidevin, Robin. 1996.  Arguing for Atheism: An Introduction to the

 Philosophy of Religion. London: Routledge.

Levine, Michael P. 2000. “Contemporary Christian Analytic Philosophyof Religion: Biblical Fundamentalism, Terrible Solutions to a Horri-ble Problem, and Hearing God.” International Journal for Philosophy

of Religion 48: 89–119.

 ——— . 2004. porary Christian Analytic Philosophy of Religion” [2000].In Faith and Reason: Friends or Foes in the New Millennium? editedby Anthony Fisher O.P. and Hayden Ramsey, 134–70. ACU Series 1. Adelaide: ATF Press.

Lipton, Peter. 1991. Inference to the Best Explanation Philosophical Is-sues in Science. London: Routledge.

Locke, John. 1846.  An Essay Concerning Human Understanding. 13thedition. London: Thomas Tegg.

Luhrmann, T. M. 1989.  Persuasions of the Witch’s Craft: Ritual Magic

in Contemporary England. Cambridge, MA: Harvard UniversityPress.

Mackie, J. L. 1955. “Evil and Omnipotence.” Mind 64: 200–12.

 ——— . 1962. “Theism and Utopia” Philosophy 37: 153–58.

 ——— . 1976. “The Riddle of Existence.” Proceedings of the Aristotelian

Society Supplementary Volume 50: 247–265.

 ——— . 1982. The Miracle of Theism. Oxford: Clarendon Press.

Manson, Neil A. 2000. “There Is No Adequate Definition of ‘Fine-Tunedfor Life.’” Inquiry 43: 341–52.

Mascall, E. L. 1966. Existence and Analogy [1949]. London: Libra/Dar-ton, Longman & Todd.


Page 202: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 202/205

Works Cited

McGrath, P. J. 1972. “Faith and Reason.” In Faith: Its Nature and

Meaning. Papers of the Maynooth Union Summer School 1970 editedby Paul Surlis, 133–49. Dublin: Gill & Macmillan.

McMullin, Ernan. 1993. “Indifference Principle and Anthropic Principlein Cosmology.” Studies in History and Philosophy of Science 24: 359– 89.

 ——— . 2002. “A Case for Scientific Realism” [1984]. In  Philosophy of 

Science: Contemporary Readings. Routledge Contemporary Readingsin Philosophy edited by Yuri Balashov and Alex Rosenberg, 248–81.

London: Routledge.Mill, John Stuart. 1979. “An Examination of Sir William Hamilton’s

Philosophy” [1872]. In Collected Works of John Stuart Mill edited byJohn M. Robson, vol. 9. Toronto: University of Toronto Press.

Musgrave, A. E. 2009. “Experience and Perceptual Belief.” In Rethink-

ing Popper Boston Studies in the Philosophy of Science Volume 272,editd by S. Parusnikova and R. S. Cohen, 5–19. Dordrecht: Springer.

Oppy, Graham. 2004. “Arguments from Moral Evil.” International Jour-

nal for Philosophy of Religion 56: 59–87.

Otto, Rudolf. 1950. The Idea of the Holy [1917]. 2nd edition. London: Ox-ford University Press.

Paley, William. 1825. Natural Theology – or Evidence of the Existence

and Attributes of the Deity Collected from the Appearances of Nature

[1802]. In The Complete Works of William Paley edited by Robert Ly-nam, 3.1–322. London: George Cowie et al.

Pascal, Blaise. 1932.  Pensées [1670] translated by W. F. Trotter. Every-

man’s Library 874. London: J. M. Dent.

Phillips, D. Z. 1976. Religion without Explanation. Oxford: Basil Black-well.

Plantinga, Alvin. 1974. The Nature of Necessity. Clarendon Library of Logic and Philosophy. Oxford: Clarendon Press.


Page 203: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 203/205

Works Cited

 ——— . 1975. God, Freedom and Evil. Essays in Philosophy. London:

George Allen & Unwin.

 ——— . 1981. “Is Belief in God Properly Basic?” Noûs 15: 41–51.

 ——— . 2000. Warranted Christian Belief . New York: Oxford UniversityPress.

Popper, Karl R. 2002. The Logic of Scientific Discovery [1935]. RoutledgeClassics. London: Routledge.

Pruss, Alexander R. 2004. “A Restricted Principle of Sufficient Reason

and the Cosmological Argument.” Religious Studies 40: 165–79.Reichenbach, Hans. 1938. Experience and Prediction: An Analysis of the

Foundations and the Structure of Knowledge. Chicago, IL: Universityof Chicago Press.

Rowe, William L. 1990. “Evil and the Theistic Hypothesis: A Responseto Wykstra” [1984]. In The Problem of Evil edited by Marilyn McCord Adams and Robert Merrihew Adams, 161–67. Oxford UniversityPress.

 ——— . 1998. The Cosmological Argument [1975]. New York: FordhamUniversity Press.

 ——— . 2004. Interview with Nick Trakakis.  Philosophy Now. August/September: 16–18.

Russell, Bertrand. 1944. “My Mental Development.” In The Philosophy

of Bertrand Russell edited by Paul A. Schlipp, 3–20. Library of LivingPhilosophers. Evanston, IL: Northwestern University Press.

Schweitzer, Albert. 1981. The Quest of the Historical Jesus. 1st edition

[1906] translated by W. Montgomery (1910). London: SCM.

Sessions, William Lad. 1980. “Kant and Religious Belief”  Kant-Studien

71: 455–68

Smart, Ninian. 1961. “Omnipotence, Evil and Supermen.”  Philosophy

36: 188–95.


Page 204: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 204/205

Works Cited

Smith, Quentin. 1999. “The Reason the Universe Exists Is That It

Caused Itself to Exist.” Philosophy 74 : 579–86.

Smith, Wilfred Cantwell. 1978. The Meaning and End of Religion

[1962]. London: SPCK.

Sober, Elliott. 2003. “The Design Argument.” In God and Design: The

Teleological Argument and Modern Science edited by Neil A. Manson,27–54. London: Routledge.

Swinburne, Richard. 1979. The Existence of God. Oxford: ClarendonPress.

 ——— . 1981. Faith and Reason. Oxford: Clarendon Press.

 ——— . 1992. Revelation: From Metaphor to Analogy. Oxford: Claren-don Press.

 ——— . 1993. The Coherence of Theism. Revised edition. Clarendon Li-brary of Logic and Philosophy. Oxford: Clarendon Press.

 ——— . 2004. The Existence of God. 2nd edition. Oxford: ClarendonPress.

Tanquerey, A. 1921. Synopsis Theologiae Dogmaticae ad Mentem S.

Thomae Aquinatis Hodiernis Moribus Accommodata. 18th edition.Paris: Desclée.

Tooley, Michael. 1991. “The Argument from Evil.” In  Philosophical Per-

spectives 5: Philosophy of Religion edited by James E. Tomberlin, 89– 134. Atascadero, CA: Ridgeview Publishing.

Tylor, Edward B. 1913.  Primitive Culture. 5th edition. London: JohnMurray.

van Inwagen, Peter. 1991. “The Problem of Evil, the Problem of Air, andthe Problem of Silence.” In  Philosophical Perspectives 5: Philosophy

of Religion edited by James E. Tomberlin, 135–65. Atascadero, CA:Ridgeview Publishing.

 ——— . 1998. “Modal Epistemology.” Philosophical Studies 92: 67–84.


Page 205: DAWES-Philosophy of Religion

7/30/2019 DAWES-Philosophy of Religion 205/205

Works Cited

Wilkerson, T. E. 1976.  Kant’s Critique of Pure Reason: A Commentary

 for Students. Oxford University Press.

Williams, Paul and Anthony Tribe. 2000.  Buddhist Thought: A Com-

 plete Introduction to the Indian Tradition. London: Routledge.

Wilson, Edward O. 2005. “Intelligent Evolution: The Consequences of Charles Darwin’s ‘One Long Argument.’” Harvard Magazine Novem-ber–December 2005: 29–33.

Wittgenstein, Ludwig. 1958.  Philosophical Investigations. 2nd edition,translated by G. E. M. Anscombe. Oxford: Basil Blackwell.