(Cornell Studies in The Philosophy of Religion) ) J.L. Schellenberg - Divine Hiddenness and Human Reason-Cornell University Press (2006)

Divine Hiddenness
and Human Reason

A volume in the series
Cornell Studies in the Philosophy of Religion

EDITED BY WILLIAM P. ALSTON
A full list of titles in the series appears at the end of the book.
J. L. Schellenberg
Divine Hiddenness
and Human Reason
Cornell University Press, Ithaca and London

\ .
Copyright © 1993 by Cornell University
All rights reserved. Except for brief quotations in a review, this book,
or parts thereof, must not be reproduced in any form without
permission in writing from the publisher. For information, address
Cornell University Press, 124 Roberts Place, Ithaca, New York 14850.
First published 1993 by Cornell University Press.
International Standard Book Number 0-8014-2792-4

Library of Congress Catalog Card Number 92-32633
Printed in the United States of America
Librarians: Library of Congress cataloging information
appears on the last page of the book.
© The paper in this book meets the minimum requirements

of the American National Standard for Information Sciences—
Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984.
For Wendy
Digitized by the Internet Archive
in 2019 with funding from
The Arcadia Fund
https://archive.org/details/divinehiddennessOOsche
Contents
Acknowledgments ix
Introduction i
Part i Framing the Argument

1 Some Epistemic Implications of Divine Love 17
2 Is a Strong Epistemic Situation in Relation to Theism
Possible? 44
3 The Reasonableness of Nonbelief 58
4 A Summation of the Case 83
Part 2 The Force o f the Argument

5 Moral Freedom and Its Requirements 95
6 The Importance of Inwardness 131
7 Investigation, Diversity, and Responsibility 168
Conclusion 208
Index 215
Acknowledgments
I am indebted to a number of individuals and institutions for the

help they have given me during the writing of this book. Richard
Swinburne commented extensively on more than one of the drafts.
His comments were invariably acute and penetrating and they have
prompted many changes (though no doubt not as many as he
would have liked). I am deeply grateful to him for his contribution
to my work. Thanks are also due to Maurice Wiles, who helped
with theological matters during the critical first stages of the proj¬
ect. His theological critique of certain initial formulations of argu¬
ments in Chapters 1 and 6 was especially useful. Terence Penel-
hum, my first teacher in the philosophy of religion, provided the
impetus for my thinking about Divine hiddenness and also dis¬
cussed with me the end result. I thank him for that and for sug¬
gesting the title. For stimulating discussion and criticism in Oxford
and in Calgary, I thank Alan Padgett, James Baird, Mark Wynn,
Terry Fach, and Michael Stoeber. I am also indebted to William
Alston and an anonymous referee for Cornell University Press,
who gave me many helpful comments and suggestions for im¬
provement, virtually all of which have been incorporated into the
final product. Absolutely essential financial support was provided
by the Alberta Heritage Fund and the Social Sciences and Human-
] Acknowledgments
ities Research Council of Canada. Finally, I thank my wife,

Wendy, and my children, Matthew and Justin, who have endured
a philosopher in the house and many different houses. Theirs is
without doubt the largest contribution of all. A special debt is
owed to Wendy for many hours at the computer and for continu¬
ally reminding me that life is more than philosophy. I dedicate this
book to her with love and gratitude.
J. L. S.
Divine Hiddenness
and Human Reason
Introduction
Many religious writers, sensitive to the difficulties in which our

evidence for God is involved, have held that God would wish (or
at any rate, permit) the fact of his existence to be obscure. God,
so it is said, is a hidden God. But upon reflection, it may well ap¬
pear otherwise. Why, we may ask, would God be hidden from us?
Surely a morally perfect being—good, just, loving—would show
himself more clearly. Hence the weakness of our evidence for God
is not a sign that God is hidden; it is a revelation that God does not
exist.1
i. Several contemporary writers have touched on this problem. See Terence

Penelhum, God and Skepticism (Dordrecht: Reidel, 1983), pp. 156-158, esp. p. 158;
John Hick, Faith and Knowledge, 2d ed. (London: Macmillan, 1988), p. 121; Ronald
Hepburn, “From World to God,” in Basil Mitchell, ed., Philosophy of Religion (Ox¬
ford: Oxford University Press, 1971), p. 178; Frank B. Dilley, “Fool-Proof Proofs
of God?” International Journal for Philosophy of Religion, 8 (1977), 19-27, 35; An¬
thony O’Hear, Experience, Explanation, and Faith (London: Routledge and Kegan
Paul, 1984), pp. 238-239; George Schlesinger, “The Availability of Evidence in
Support of Religious Belief,” Faith and Philosophy, 1 (1984), 422-427; William Al¬
ston, “Religious Diversity and Perceptual Knowledge of God,” Faith and Philoso¬
phy , 5 (1988), 445; C. Robert Mesle, “Does God Hide From Us? John Hick and
Process Theology on Faith, Freedom, and Theodicy,” International Journal for Phi¬
losophy of Religion, 24 (1988), 97; Thomas V. Morris, “The Hidden God,” Philo¬
sophical Topics, 16 (1988), 5-7, 11; Mark R. Talbot, “Is It Natural to Believe in
God?” Faith and Philosophy, 6 (1989), 160-161; and Robert McKim, “The Hidden¬
ness of God,” Religious Studies, 26 (1990), 141-143. But few of these writers have
] Introduction
In this book I argue that there is support for such claims as these.
The weakness of evidence for theism, I maintain, is itself evidence
against it. Paving the way for my argument will be a discussion of
certain important questions raised by its central claim: What reasons
are there to suppose that God would be more clearly revealed? (It is
not obvious a priori that he would, so reasons need to be pro¬
vided.) If reasons there be, what is suggested by them as to the
degree of clarity, and of accessibility, we might expect to find?
Should we look for proof—perhaps an overwhelming manifesta¬
tion—or would something less than this serve God’s purposes?
And would the evidence be made available to all, or only to those
who satisfy certain conditions beforehand? More generally, what
explications of “weak” and “strong” ought we to give here? It
seems clear enough that these questions need answering if our
problem is to be properly developed, but they have not been ad¬
dressed in previous discussions, even by those (few) who have seen
that there may be an argument for atheism lurking here. Our dis¬
cussion will seek to remedy this neglect—to state with some preci¬
sion why we might expect stronger evidence for theism to be
available if God exists, and what sort of evidence God might be
expected to provide.
The argument that will emerge from this discussion is, in broad
outline, as follows. A perfectly loving God would desire a recipro¬
cal personal relationship always to obtain between himself and
every human being capable of it. But a logically necessary condi¬
tion of such Divine-human reciprocity is human belief in Divine
existence. Hence a perfectly loving God would have reason to en¬
sure that everyone capable of such belief (or at any rate, everyone
capable who was not disposed to resist it) was in possession of
evidence sufficient to bring it about that such belief was formed.
But the evidence actually available is not of this sort (the claim that
it is “weak” is to be read simply as the claim that it is not “strong”
in this sense). The most obvious indication that it is not is that
inculpable—or as I prefer to term it, reasonable—nonbelief actually
offered suggestions as to how the weakness of theistic evidence might yield an

argument for atheism; and the remarks of those who have are sketchy.
Introduction [3
occurs.2 Hence we can argue from the weakness of theistic evi¬

dence (where this is understood as indicated), or more specifically,
from the reasonableness of nonbelief, to the nonexistence of a per¬
fectly loving God. But God, if he exists, is perfectly loving. Hence
we can argue from the reasonableness of nonbelief to the nonexist¬
ence of God.
In Part i of the book (Chapters 1-4), I fill out the argument here
briefly sketched, and detail the considerable initial support that it
can claim. I show, first, that there are indeed grounds to suppose
that a perfectly loving God would put his existence beyond reason¬
able nonbelief. I argue, further, that although it is possible for God
to prevent it, reasonable nonbelief occurs. In Part 2 (Chapters 5-7),
I examine various possible rebuttals, explicit or implicit in the liter¬
ature—arguments of Pascal, Joseph Butler, Kierkegaard, John
Hick, and others which might be thought to defeat the prima facie
case of Part 1 by showing that there are reasons for God to remain
hidden from us that override or at least offset the reasons for reve¬
lation (where “hiddenness” and “revelation” are understood in
terms of the permission and of the prevention of reasonable non¬
belief, respectively). Not all of these arguments have been given a
clear shape in the past or are applicable as they stand, and so my
procedure is to clarify them and adapt them for my purposes as I
go along. That they are in many cases less than fully developed is
at least in part due to the fact that the problem to which they must
be seen as responses—the problem of weak theistic evidence—has
not heretofore been carefully articulated. My claim in Part 2 is that
no available argument is an adequate defense against the version of
the problem that I present. The final claim of the book is accord¬
ingly that for anyone who accepts my judgments (along with cer¬
tain assumptions, to be indicated later in the Introduction), the ar¬
gument from the reasonableness of nonbelief goes through,
providing good grounds for atheism.3
2. Nonbelief is reasonable, I will stipulate, if and only if it is not the result of

culpable actions or omissions on the part of the subject.
3. It may seem to some, upon reading my argument, that it might just as well
have been developed in terms of Divine justice—that an emphasis on the implica¬
tions of love is not essential to my case. For, it may be argued, a perfectly just God
would distribute good evidence of his existence evenly, and so ensure that the op-
4] Introduction
I have suggested (and the title of this book also suggests) that the
question here raised for discussion is connected to the idea of the
hiddenness of God. But the notion of God’s hiddenness is ambig¬
uous, and so this way of referring to our topic can be misleading.
It can mislead, first of all, by suggesting that all the ways in which
God may be said to be hidden pose problems for theology. It can
also mislead in a very different way, by suggesting that the absence
of persuasive and readily available theistic evidence is not a prob¬
lem—that since theology has always insisted on God’s hiddenness,
the question I am raising can be dismissed a priori, as displaying
theological naivete. In order to avoid being misled in these ways,
we must make some distinctions. The notion of God’s hiddenness
can be interpreted in at least three ways: as referring to the obscur¬
ing of God’s existence, the incomprehensibility of God’s nature, or
our inability to detect the exact pattern of God’s activity in the
world. To the question I am raising, only “hiddenness” in the first
sense is relevant. All I seek to show is that we might expect God’s
existence to be more obvious. I am happy to allow that the same
claim made in respect of the depths of God’s nature or the exact
pattern of his activity would have much less to recommend it.
Can it now be argued, however, that theologians have always
insisted on God’s hiddenness in all three senses? I think not. Al¬
though some writers, such as Karl Barth, seem to refer to God as
hidden not only in the second and third of the above senses, but
also in the first, it is important to note that this is largely a post-
Enlightenment phenomenon.4 The majority of patristic, medieval,
portunity to acquire belief and the benefits of belief was missed by no one (see
Schlesinger, “Availability of Evidence,” 424-425). But if only God’s justice need
be saved, the theist may always point to the possibility that everything will even
out in the end—that those who fail to benefit from belief in this life will be com¬
pensated hereafter. Or else she may claim that it is not unjust to give a gift to
someone while not giving it to others, unless everyone can be shown to be entitled
to it. Such easy “outs” are, however, not possible where Divine love is concerned.
If God is loving as well as just, he is motivated to pursue personal relationship
with each of us in this life as well as in the hereafter. In the interests of a stronger
argument, therefore, I have decided to fill out the reasoning suggested by the ques¬
tion “Why, if there is a God, is his existence not more obvious?” in terms of love
rather than in terms of justice.
4. For Barth’s view, see his Church Dogmatics, vol. 2, pt. 1 (Edinburgh: T&T
Introduction [ 5
and reformation theologians affirmed that God’s existence could be

“clearly seen, being understood through the things that have been
made” (Rom. 1:20), while maintaining that “such knowledge or
demonstration could not comprehend God as he was in himself.”"
Even Martin Luther, who championed the idea of a hidden God,
wrote by way of comment on Romans 1:20 that “there must be
that which is more sublime than anything else, which is higher
than all and helps all.”* * * 5 6 According to Jaroslav Pelikan, Luther as¬
cribed to natural reason “the recognition, even apart from Scrip¬
tures, that God was almighty and his foreknowledge was infal¬
lible.”7 We may therefore conclude that the traditional emphasis of
theology on the hiddenness of God does not imply that the evi¬
dence for God’s existence must be weak, and hence does not sanc¬
tion a dismissal of our inquiry, as it might otherwise be thought to
do.
There is, however, a third way in which the description of our
problem in terms of the hiddenness of God can be misleading.
“God is hidden” (in the sense of “hidden” that concerns us) is per¬
haps most naturally construed as equivalent to
(1) God exists and has intentionally withheld (or permitted to

be obscured) strong evidence of his existence.
But (1) entails that God exists, and so it would be contradictory,

on this understanding of “God is hidden,” to speak of an argument
from the hiddenness of God against God’s existence. Of course,
those who think of the hiddenness of God as an epistemic problem
of the sort here discussed may well be understanding it solely in
terms of the absence of strong evidence for God’s existence. But it
is not clear to me that this is so; and it is even less clear that this
would be an acceptable understanding. It seems much more natural
Clark, 1957), p. 187. He writes: “That God is, lies as little in the field of our
spiritual oversight and control as what He is. We lack the capacity both to establish
His existence and to define His being.”
5. Jaroslav Pelikan, The Christian Tradition, vol. 3 (Chicago: University of Chi¬
cago Press, 1978), p. 289.
6. Martin Luther, quoted in Jaroslav Pelikan, The Christian Tradition, vol. 4
(Chicago: University of Chicago Press, 1984), pp. 165-166.
7. Ibid., p. 166.
6] Introduction
to interpret talk of God’s hiddenness as implying a commitment to

God’s existence, and to understand the problem of God’s hiddenness
as a nonepistemic problem—a theological, or perhaps pastoral,
problem of the sort that only believers may admit.8 This, at any
rate, is the understanding I will assume. Accordingly, our problem
is not to be understood as the problem of God’s hiddenness, nor
my argument as an argument from God’s hiddenness. This, how¬
ever, is not to deny that the claim that God is hidden, and sug¬
gested reasons for such hiddenness, are relevant to our problem, or
that my argument has no implications for how the idea of Divine
hiddenness should be assessed. Quite the contrary. Such claims and
such reasons may, as we have seen, function in a solution to the
problem, and if no purported solution succeeds, the claim that God
is hidden (as well as the claim that God exists) must be relin¬
quished. Human reason may, as the title of this book suggests,
have something to say about the idea of Divine hiddenness, and
what it has to say may be of great significance for the philosophy
of religion.
It is interesting to note at this juncture that the problem posed
for theology by the argument I develop may also be construed as a
special instance of the problem of evil. But it is important to be
clear about what is meant by this. Philosophers who discuss the
problem of evil usually restrict their attention to the question
whether certain events, actions, and states of affairs that we would
all naturally view as negative or destructive and so as evil (e.g.,
wrong actions, physical pain, mental anguish) provide the basis for
a strong argument against the existence of an omnipotent, omnis¬
cient, and morally perfect God (a logical argument showing the
existence of such evils to be incompatible with the existence of
God, or an empirical argument showing that their existence renders
his unlikely). The phenomena commonly discussed are, further¬
more, ones whose elimination would be a great good even if athe¬
ism were true—facts which, as we might say, the world would be
much better without whether God exists or not. But as an article
by Terence Penelhum suggests, to pose a problem of evil, an event,
8. Of course, the theological problem is closely connected to ours inasmuch as

any solution to the former might well provide the theist with a response to the latter.
But insofar as the theological problem presupposes God’s existence, the two are not
to be identified.
Introduction
action, or state of affairs need not be of this sort. ' It need only be
such that the affirmation of its existence comes into conflict with
what theists are committed to saying about the nature of God—in
particular, about the moral nature of God: it poses a problem of evil
if it seems that a morally perfect God who was also omnipotent
and omniscient could not allow it, or if it seems improbable that he
would. Hence the state of affairs discussed in this book—the rea¬
sonableness of nonbelief—though it is not a state of affairs that we
would all naturally view as negative or destructive and so as evil,
and though it is not a state of affairs the world would be much
better without even if God does not exist, may still legitimately be
viewed as posing a problem of evil if the affirmation of its exist¬
ence comes into conflict with what theists are committed to saying
about God’s moral nature. And that it does so is my claim. Hence I
seem to be in a position to claim that the problem of reasonable
nonbelief is a problem of evil. However, the “evil” to which it
refers is, as we have seen, not of the ordinary sort (for it does not
consist in pain or suffering or any other commonly recognized evil,
nor would its removal be a great good unless personal relationship
with God is possible); and so, to mark this difference, it may be
appropriate to refer to it as a special instance of the problem of evil.
It remains only to point out the sort of conflict I have in mind
here: Do I consider the problem of reasonable nonbelief to be a
special instance of the logical, or of the empirical, problem of evil?
If the former, then the claim I advance is tantamount to the claim
that
(2) If God exists and is perfectly loving, reasonable nonbelief

does not occur
is necessarily true, or equivalently, that
(3) God exists and is perfectly loving
is logically incompatible with
(4) Reasonable nonbelief occurs.
9. Terence Penelhum, “Divine Goodness and the Problem of Evil,” Religious

Studies, 2 (1967), 95-107.
8] Introduction
And in reply to such a claim, it would surely be tempting to argue,

in the spirit of Alvin Plantinga’s famous response to the logical
problem of evil, that the compatibility of (3) with (4) can quite
easily be demonstrated.10 It might be argued, for example, that
there is a possible world in which humans are constructed in such a
way that strong evidence for God’s existence causes them to con¬
form automatically to what they perceive as God’s will. If such a
world were actual, even a perfectly loving God concerned to facili¬
tate reciprocal personal relationship would surely allow reasonable
nonbelief to occur, for his doing so would be a necessary condition
of human moral freedom. Thus, it might be concluded, the claim
that God exists and is perfectly loving can easily be shown to be
compatible with the claim that reasonable nonbelief occurs: there is
a possible world (one in which humans are as described) in which
both claims are true.
Now it seems to me that the compatibility of (3) with (4) is not
so easily demonstrable as our imaginary critic imagines. For the
critic must show not only that there is a possible world in which
humans are so constituted, but also that there is a possible world in
which a loving God has made them so; and the latter possibility has
not been demonstrated. To demonstrate it, the critic would have
to show either (i) that there is a possible world containing God in
which humans are as described and in which God would not have
any inclination to seek personal relationship, or (more hopefully)
(ii) that there are possible goods which require for their existence
that humans be so constituted, and which outweigh or offset the
goods that would be sacrificed by a loving God in so making
them. But while it seems to me, for this reason, that it would be
rash to suppose that the claim that (2) is necessarily true can easily
be shown false, I will not seek to defend it. It will be more conve¬
nient for our purposes to concentrate on the question whether (2)
is true. If we do, a Plantinga-style response will be not just difficult
but irrelevant. And in any case, it will at some points in the argu¬
ment be necessary to appeal to features of the actual world which
we cannot assume are replicated in every possible world containing
10. See Alvin Plantinga, The Nature of Necessity (Oxford: Clarendon Press,
1974), chap. 9.
Introduction
human beings. Accordingly, the problem of reasonable nonbelief,

as I develop it, must be viewed as a special instance of the empirical
problem of evil: the claim at issue, properly understood, is that
because of what we have reason to believe about the connection
between Divine love and the prevention of reasonable nonbelief,
and because states of affairs for the sake of which even a loving
God might permit reasonable nonbelief to occur apparently do not
actually obtain, there is good reason to suppose that (2) is true, and
so (given the occurrence of reasonable nonbelief) good reason to
suppose that a loving God does not exist.11
In a philosophical work of this sort, one must always decide
what to assume and what to defend or oppose. It will be useful at
this point to indicate several important assumptions I will be mak¬
ing—claims for which no explicit argumentation is provided in
later parts of the book.
(1) The first of these is that belief is involuntary in the sense that
we cannot believe a proposition at a moment’s notice. If we could
decide to “believe” where formerly we had not, and our decisions
were immediately efficacious, we would know that our “beliefs”
were the result of our decisions and not determined by how things
are. But in that case we would not have any reason to suppose that
11. It may be worthwhile to point out here that there is a way of construing my
claim as a necessary truth that is equally immune to a Plantinga-style response. This
can be seen as follows. The claim that some state of affairs obtains and the claim
that it obtains in the actual world (a) are truth-functionally equivalent. (I assume
here that “the actual world” and “a” designate rigidly.) Hence we may with per¬
fect propriety rephrase (2) as follows:
(2') a includes the state of affairs consisting in God’s being perfectly

loving only if reasonable nonbelief does not occur.
But as Plantinga’s own work shows, “a-claims” like (2') are noncontingent—ei¬
ther necessarily true or necessarily false. For if a includes some state of affairs S, it
necessarily includes it. And if a does not include S, it necessarily excludes it. (See
Plantinga, Nature of Necessity, pp. 45-46, 55.) Hence if (2') is true, it is necessarily
true, which was what we wished to show. So anyone who wishes to do so may
interpret the claim for which I will be seeking to make a prima facie case as a claim
with respect to what is necessarily true: no susceptibility to Plantinga-style maneu¬
vers need thereby be incurred. But for myself, I will continue to view it simply as
the claim that if God is perfectly loving, reasonable nonbelief does not occur—a
claim which purports to be no more than true.
Introduction
what we “believed” was true and so would not really believe.12

That belief is involuntary has been convincingly argued by a num¬
ber of writers, and the force of their arguments would be conceded
by perhaps the majority of contemporary philosophers.13
(2) A second assumption is that humans have libertarian free
will. As Richard Swinburne has concisely put it, a being has free
will in this sense if “he acts intentionally and . . . how he acts is not
fully determined by prior states of the world; his choices are to
some extent up to him.”14 The claim that human beings have liber¬
tarian free will is of course highly controversial, but I will assume
it because it is accepted by most of the contemporary philosophers
and theologians to whom this work is addressed and often presup¬
posed in their arguments.
(3) I will further assume that it is constitutive of the idea of God
that God, if he exists, is unsurpassably great. As such, God is to be
described (minimally) as ultimate (i.e., the source or ground of all
existence other than his own, to whom nothing stands as a ground
of existence), personal (that is to say, one of whom agential, intel¬
lectual, and affective qualities may appropriately be predicated),
and (in some sense) all-powerful, all-knowing, perfectly good, and
perfectly loving. I will not, with one exception, offer a defense of
any of the several parts of this description, for it would, I think, be
accepted by the majority of contemporary theologians and philoso¬
phers of religion. The exception concerns the claim that God is to
be understood as perfectly loving—a claim central to my argu¬
ment. It might be thought that this is a claim only Christians have
any reason to accept. But I would deny this. Without offering any-
12. I have taken this argument from Richard Swinburne, Responsibility and
Atonement (Oxford: Clarendon Press, 1989), pp. 165-166. See also Bernard Wil¬
liams, Problems of the Self (Cambridge: Cambridge University Press, 1973), pp.
147-149. There will be more on the nature of belief in Chapter 1, second section.
13. While it is impossible, as I am suggesting here, to will belief directly, it may
be possible to will it indirectly, and so belief may be said to be indirectly voluntary.
But this is a very different notion. Getting oneself to believe, for example, that
there is a God when the evidence does not seem initially to support that belief
requires deliberate self-deception, and such self-deception—since it involves view¬
ing the evidence selectively and forgetting that one has done so—requires a consid¬
erable period of time.
14. Richard Swinburne, The Existence of God (Oxford: Clarendon Press, 1979),
P- 153-
Introduction [
thing like a complete explication of “Divine love,” I think we can
say that what usually goes by that name—at a minimum, self-giv¬
ing, unconditionally accepting, relationship-seeking love—is such
that any being who lacked it would be a being whose greatness
could be surpassed, and therefore not God. Love of the sort in ques¬
tion is clearly one of the highest manifestations of personal being;
so if God is conceived as embodying the perfections of personal
life, he must be conceived as perfectly loving.
Now it may be said that this picture of God as perfectly loving is
a picture of God peculiarly suited to my purposes and so one I will
naturally be inclined to defend. But it should be noted that very
many who are not antecedently committed to the conclusion of
my argument would nonetheless accept this premise without hesi¬
tation. As H. H. Price puts it, “What could conceivably be better
than universal and unconditional love?”15 In the unconditional love
of a Mother Teresa, for example, we see reflected a quality of be¬
ing that most of us would intuitively say must belong to God, if
God exists. For it is something to which we are inclined to ascribe
very great value indeed. When we reflect on the nature of such
love, all the power and knowledge in the world fade by compari¬
son. Hence it would seem (and I will assume) that all who espouse
a form of theism are rationally committed to the truth of the claim
that God, if he exists, is perfectly loving.
(4) A fourth assumption is that the claim that God exists is coher¬
ent. I do not wish to suggest that there are no problems of coher¬
ence which theists must resolve, but these problems will not con¬
cern us here. That some description of God falling within the
parameters laid down above is coherent is in any case assumed by
most contemporary philosophers of religion to be in principle de¬
monstrable.16 The questions that concern philosophers of religion
writing today are more commonly questions of truth than ques¬
tions of coherence.
(5) I will further assume that the relevant evidence (exclusive of
15. H. H. Price, “Faith and Belief,” in John Hick, ed., Faith and the Philosophers
(London: Macmillan, 1964), p. 5.
16. See, for example, Richard Swinburne, The Coherence of Theism (Oxford:
Clarendon Press, 1977), and J. L. Mackie, The Miracle of Theism (Oxford: Claren¬
don Press, 1982), pp. 1-3.
] Introduction
evidence adduced in this book) does not clearly favor either the
conclusion that there is a God or its denial. This assumption seems
warranted in light of recent discussions in the philosophy of reli¬
gion, which often end in deadlock, and is accepted by many, both
inside and outside the academic study of religion.17 In the Conclu¬
sion, however, I will consider the implications of our discussion
for those who would view belief in the existence of God as having
considerably more in the way of evidential support than I here al¬
low.
(6) Finally, I will assume that it is coherent to suppose that hu¬
man beings survive their deaths. This assumption, like the assump¬
tion of libertarian free will and of the coherence of theism, is made
in full recognition of its controversial status, for the sake of argu¬
ment: it too is accepted and commonly presupposed by those to
whom this book is addressed.
My description in this Introduction of the argument to follow
may seem to suggest that it is my wish to defend atheism—to
show that God does not exist. But this is not an accurate represen¬
tation of my intent; and I would like to emphasize that it is not.
The intended contribution of this book is more correctly charac¬
terized as a demonstration that there is an argument here—an argu¬
ment of considerable force, which deserves much more attention
than it has heretofore received. It is true, as noted earlier, that I will
conclude by declaring all the available counterarguments to be
bankrupt, and by suggesting that anyone who accepts my conclu¬
sion (as well as the fifth assumption above) has good grounds for a
denial of theism. But I by no means wish to rule out the possibility
that better arguments than the ones here discussed may one day be
devised. It is indeed my hope that those who read this book may
be motivated to seek to provide such arguments. Thus although I
do believe that, where there is a prima facie case, the failure of all
the arguments one can think of to rebut that case provides good
grounds for believing its conclusion, and that no available counter-
17. The deadlock is exemplified by, for example, the opposed arguments of
Swinburne’s Existence of God, and Mackie’s Miracle of Theism.
Introduction
argument defeats the case that I present, my claim should be con¬

strued by theists not as a cry of triumph but rather as a challenge,
an invitation, to find for the problem herein discussed the solution
I myself have been unable to uncover.
Part 1
Framing the Argument

[ I ]
Some Epistemic Implications
of Divine Love
We saw in the Introduction that the concept of God is the con¬

cept of a being who is, among other things, perfectly loving. But
what is it for God to be loving? More specifically, what is it for
God to love us—to love human beings? One term that may seem
important here is “benevolence”: if God is loving, he desires our
well-being. And indeed, it can hardly be denied that a reference to
benevolence must find its way into any adequate explication of
Divine love, for the sort of love that can be viewed as a perfection
of personal being is clearly other-regarding. The one who claims
that benevolence is a feature of Divine love can also point out that
this is entailed by the understanding of God’s love as agape, a self¬
giving love.1 As Robert Adams writes, “There cannot be Agape at
all without benevolence.”2 But there is more to love than a general
reference to benevolence can capture. What more there is cannot be
fully detailed here, but most important for our purposes is the (of¬
ten neglected) connection between Divine love and the seeking of
personal relationship. In the present chapter, I seek to clarify this
1. The classic treatment of the concept of “agape” is to be found in Anders

Nygren, Agape and Eros, Philip S. Watson, trans. (Philadelphia: Westminster Press,
1953). But as later discussion will suggest, Nygren’s distinction between the two
forms of love mentioned in his title may be too sharply drawn.
2. Robert Adams, The Virtue of Faith (New York: Oxford University Press,
1987), p. 190.
8] Framing the Argument
connection and to develop certain of its implications.3 In particular,

I argue that if a perfectly loving God exists, all human beings capa¬
ble of personal relationship with himself are, at all times at which
they are so capable, in a position to believe that he exists.
A Neglected Feature of Divine Love
“God seeks to be personally related to us.” In claiming that this

proposition is essential to any adequate explication of “God loves
human beings,” I am claiming that God, if loving, seeks explicit,
reciprocal relationship with us, involving not only such things as
Divine guidance, support, and forgiveness, but also human trust,
obedience, and worship. So understood, this proposition seems
obviously required. For only the best human love could serve as an
analogy of Divine love, and human love at its best clearly involves
reciprocity and mutuality. If I love you and so seek your well¬
being, I wish to make available to you all the resources at my dis¬
posal for the overcoming of difficulties in your life. But then I
must also make it possible for you to draw on me personally—to let
you benefit from my listening to your problems, from my encour¬
agement, from my spending time together with you, and so on. In
other words, I wish to make available to you the resources of an
intimate personal relationship with me. This, indeed, is part of
what is involved in self-giving. As W. H. Vanstone puts it, “the
authenticity of love must imply a totality of giving—that which
we call the giving of self or self-giving. The self is the totality of
what a man has and is: and it is no less than this that is offered or
made available in love.”4 Therefore, if I am to act toward you with
perfect benevolence, I must, it seems, seek personal relations with
you.
This emphasis seems especially appropriate in the Divine case:
personal relationship with God would immeasurably enhance our
well-being, and so God as perfectly benevolent must seek it for us.
3. In line with earlier discussion in the Introduction, claims with respect to this
connection are defended as true, not as necessarily true. Needless to say, if they are
necessarily true as well, the argument will be none the worse for it.
4. W. H. Vanstone, Love’s Endeavour, Love’s Expense (London: Darton, Long¬
man and Todd, 1977), p. 45.
Epistemic Implications of Divine Love [
Before endorsing this claim, however, we must examine it a little
more closely. What are the benefits such a relationship would
make available? We may begin by noting the ethical benefits. It
seems clear that explicit relationship with a perfectly loving God
would have a certain moral influence and make available certain
resources for dealing with the moral weakness endemic to human¬
ity; and few would deny that, were we to become ethically stron¬
ger in this way, our well-being would be enhanced. But that is not
all. If God exists and is perfectly loving, and if the life of God is
unsurpassably good, as by definition it must be, and if, in addition,
God has created us and wills our good, then the love of which we
are in some measure capable must be seen as a significant clue to
the nature of our deepest well-being. Growth in self-giving love
must, in other words, lead to a deeper realization of well-being,
and so God as perfectly benevolent will naturally seek to facilitate
it. In helping us replace self-centered patterns of activity with
agapeistic ones, he will be allowing us to share in his own life and
thus will be facilitating the achievement of an objective quality of
being which is of great worth. Now, of course, to avoid begging
the question at issue, we must assume for the moment that the
love God seeks to facilitate is a form of benevolence not necessarily
involving the seeking of personal relationship. But, even so, the
promised additional point follows. It is that if self-giving love has
this special importance vis-a-vis our well-being, then given that we
are often morally weak, we are even more obviously in a position
to benefit from outside help—help of the sort made available by
personal relationship with God.
Few religious writers would quarrel with this view. Their posi¬
tion is represented by Brian Hebblethwaite:
Forgiveness, reconciliation, peace and justice sound much the same

when advanced as ideals of life by theists and non-theists alike. But
in fact these qualities and ideals of life turn out rather differently
when they are experienced and embraced as effects of gratitude,
grace and the divine indwelling. . . .
The whole question of their practical realizability is different.
Men need the resources of God’s indwelling grace and inspiration if
these values and ideals are to be realized humanely. . . . The major-
20 ] Framing the Argument
ity of people need a power not themselves that makes for righteous¬
ness.5
Basil Mitchell and Robert Adams argue similarly:
It is, above all, the love of God which serves both as a motive and as
a reason for the love of neighbor. We should love him because he
first loved us; and we should love others because he loves them. It is
this theme preeminently which explains how it is possible for a man
to turn away from anxious self-concern and identify himself with
the interests of others, however uncongenial those others are, and
even if it runs counter to the prevailing ethos of his society.6
If one both loves God and trusts in God’s love, this will issue in an
inner peace or sense of security. And this, as many religious thinkers
have argued, will free one to take a lively interest in God’s creatures
for their own sake—to enjoy his gifts with un-self-conscious grati¬
tude and to love one’s neighbor. Here a love for God, combined
with faith in him, provides an atmosphere of gladness and security
in which a love for the creature can be encompassed.7
That a personal relationship with God would contribute ethically

to human well-being seems to me, for reasons of the sort here
mentioned, to be beyond dispute.
Turning now to the possible experiential benefits of such a rela¬
tionship, I suggest that a similar judgment is warranted. Here we
are dealing not with moral character brought about through one’s
actions, but with the quality of one’s inner life—not with what one
does or becomes, but with what happens to one. Consider, for ex¬
ample, the peace or sense of security mentioned by Adams, the joy
that may come from the conviction that one is rightly related to
what is ultimately real, the self-enrichment experienced in wor¬
ship, the experience of God’s loving presence. As Hebblethwaite
5. Brian Hebblethwaite, The Ocean of Truth (Cambridge: Cambridge Univer¬

sity Press, 1988), pp. 15-16.
6. Basil Mitchell, Morality: Religious and Secular (Oxford: Clarendon Press,
1980), p. 145.
7. Robert Adams, “The Problem of Total Devotion,” in Robert Audi and Wil¬
liam Wainwright, eds., Rationality, Religious Belief and Moral Commitment (Ithaca,
N.Y.: Cornell University Press, 1986), p. 184.
Epistcmic Implications of Divine Love
suggests, it is arguable that even the experiential aspects of our

relationships with other human beings would be significantly en¬
hanced by relationship with God: “Where the other is seen as a
child of God, and where human interpersonal relation is believed
to find its own true fulfillment in conformity to and inspiration by
the . . . love of God who made us, love occurs in a different di¬
mension and takes on other transcendent qualities.”8 These things
clearly are benefits in their own right; even if they did not
strengthen us ethically, experiential features like these would con¬
tribute greatly to human well-being.
So far we have noted certain ethical and experiential benefits of
personal relationship with God. There are, however, two more
(closely related) senses in which such relationship may be said to
contribute to human well-being. First, a personal relationship with
God must also be viewed as adding value to human life by virtue
of what it is intrinsically. To be personally related to unsurpassable
goodness is a great good in itself. Second, the one who has com¬
mitted herself to God and is growing in personal relationship with
God will come to see ever more clearly that this is so (i.e., will see
that relationship with God adds value to life by virtue of what it is
intrinsically), and thus will desire for its own sake to enter ever
more deeply into communion with God. In allowing this to take
place, God is fulfilling the desire of the individual, and hence the
individual will be happy—her well-being will be increased. She
will be so even if she does not seek to be happy. Indeed, her not
seeking personal relationship with God for the sake of the happi¬
ness it may produce is a necessary condition of her being happy in
this way.4
The considerations I have adduced show that personal relation¬
ship with God would indeed enhance human well-being; hence we
might expect a perfectly loving God, concerned for our well-be¬
ing, to make such relationship possible for us. It must now be
noted, however, that—as the argument of the previous paragraph
already suggests—a proper understanding of “God loves human
beings” requires us to view God as valuing (and so seeking) the
8. Hebblethwaite, Ocean of Truth, p. 15.

9. I am grateful to Richard Swinburne for drawing these points to my atten¬
tion.
personal relationships in question not merely because they would

be good for us but for their own sakes as well. This is persuasively
argued by Adams, who sees the denial of this view as involving a
confusion over the relation of agape and eros:
The contrast between Agape and Eros is popularly seen as a special

case of the contrast between altruism and self-interest. . . . [This]
contrast between altruistic and self-interested desires is legitimate
and useful. But it has too often been treated as a dichotomy. . . .
The mistake, in trying to force love into the dichotomy of self-
interest and altruism, is a failure to recognize a desire for a relation¬
ship for its own sake as a third type of desire that is not just a
combination or consequence of desire for one’s own good and desire
for another person’s good. . . .
Thus identification of Eros with self-interested desire for personal
relationship is in error; and so is the identification of Agape with
benevolence. The ideal of Christian love includes not only benevo¬
lence but also desire for certain kinds of personal relationship, for
their own sake. Were that not so, it would be strange to call it
“love”. It is an abuse of the word “love” to say that one loves a
person, or any other object, if one does not care, except instrumen-
tally, about one’s relation to that object.10
As Adams notes, his view apparently agrees with the biblical un¬
derstanding of Divine love:
God’s love for us is surely seen as involving a desire for certain

relationships between God and us, for their own sakes and not
merely as good for us. . . . He desires our worship and devotion. . . .
No doubt it would be possible to interpret all of this on the hypoth¬
esis that God desires to be related to us only because it would be
good for us. But I think that is implausible. The Bible depicts a God
who seems at least as interested in divine-human relationships as in
human happiness per se.u
If Adams’s claims are correct—and it seems to me that they are—

we must say that not only the one who acts benevolently without
10. Adams, Virtue of Faith, pp. 187-188.

11. Ibid., p. 189. Adams’s view that eros is included in the Divine agape finds
parallels in the writings of contemporary theologians. See, for example, Hendrikus
Berkhof, Christian Faith, Sierd Woudstra, trans. (Grand Rapids: William B. Eerd-
mans, 1979), p. 124.
seeking personal relations, but also the one who seeks personal re¬
lations only from benevolence, fails to achieve the fullness of love.
And so God, if he exists and is perfectly loving, must also desire
personal relationship with us for its own sake.
This is, it seems to me, an important point. For it allows us to
claim with full assurance that even if our well-being would be as
well served for a time by the existence of a state of affairs entailing
the absence of personal contact with God, God would not on that
account be deterred from seeking personal relationship with us.
His valuing of friendship for its own sake would in every case
prevent him from actualizing the state of affairs in question.12 By
reflecting on our own friendships with other human beings and on
the judgments we are inclined to make with respect to their value
as we grow in them, we can see a little of the intrinsic value God
(if he exists) might be expected to find in personal relationship
with us. But it is only a little. God, infinitely more knowledgeable
and capable of love than we, will, we may expect, see ever so
much more of value in personal relationship with us—the beings
he has (in biblical terms) fashioned in his own image—than we can
see ourselves.
I have been defending the claim that the proper explication of
“God loves human beings” must include the proposition “God
seeks to be personally related to us.” But two very important ques¬
tions about this claim have yet to be addressed: (i) What is the
extension of “us” here? Does this expression refer to all human
beings or only to some? (2) Would a perfectly loving God seek to
be personally related to us only at some future point, perhaps after
death, or already in this life? I will treat these questions in turn.
(1) Although some theologians in times past have seemed to fa¬
vor a contrary view, virtually all contemporary writers would say
that if God’s love is a perfect love, it must extend to everyone
equally—nothing less than unlimited love in this sense would be
worthy of God. Consider, for example, the following passages,
chosen at random:
12. Except, perhaps, insofar as the latter was viewed as an equally great good in
itself or as the necessary condition of such a good.
One thing is sure, that there is no theological justification for setting

any limits on our side to the friendliness of God towards man.13
Since God is the one God of all beings, he can be no respecter of

persons. His love is not marked by favoritism towards this individ¬
ual or that but embraces the whole of mankind.14
The Divine love is limitless. . . . [It] extends to all creatures impar¬

tially. While it is fairly absurd to speak of degrees of love, it is true
that in loving all things without restriction, God’s love is unsurpass¬
able; no other love could extend so far and be so indefectible.15
The view defended by the theologians cited seems clearly cor¬

rect: if God is perfectly loving, he must give himself wholly to his
creation, and to his whole creation. Does this imply, however, that
God seeks to be personally related to all human beings? It would
seem that it does, for perfect love, we have said, involves the seek¬
ing of personal relationship, and God, as we have just claimed,
must be said to love everyone. But are all human beings capable of
personal relationship with God? It would seem that the vast major¬
ity are, at least to some degree. For, presumably, to be capable of
personal relationship with God at a time, one must be in possession
at that time of the cognitive and affective equipment required to
hold religious beliefs and exhibit such attitudes as trust, grateful¬
ness, obedience, and worship; and this excludes few of us. Indeed,
given this understanding, the majority of us are in most of the
stages of life, from early to late, capable to some degree of personal
relationship with God.16 And if some human being, for whatever
13. Karl Barth, God, Grace, and Gospel, J. S. McNab, trans. (Edinburgh: Oliver
and Boyd, 1966), p. 50.
14. Wolfhart Pannenberg, Theology and the Kingdom of God (Philadelphia: West¬
minster Press, 1969), p. 118.
15. Keith Ward, Rational Theology and the Creativity of God (Oxford: Basil
Blackwell, 1982), p. 142.
16. For example, even individuals who have cognitive and affective powers that
are still developing (children) or impaired (the mentally handicapped) may surely
sometimes be said to have a capacity, however limited, for belief in God, grateful¬
ness to God, and so on. Indeed, as the New Testament writer suggests, children
are often the ones who provide our best examples of trusting, grateful, worshipful
attitudes in relation to God (Matthew 19:13-15; 21:16). Of course, there comes a
point when we would wish to say that a capacity for such attitudes does not exist.
reason, environmental or genetic, seems at some time utterly inca¬

pable of personal relationship with God, we cannot rule out the
possibility that God, if he exists and is perfectly loving, will at
some future point in that individual’s life, or in the hereafter, pro¬
vide him with the capacities required for this and other forms of
well-being. So, I suggest, we may suppose that God seeks to be
personally related to all human beings, for all may, at some time or
other, be given some degree of capacity for such relationship. But
to avoid the possibility of objection, let us say that God, if per¬
fectly loving, seeks to be personally related to those human beings
who at some time evince some capacity for such relationship. This
is compatible both with the view that all human beings at some
time enter into personal relationship with God and with the view
that some never do. And surely it represents the least that we can
say. Since love involves the seeking of personal relationship, if God
is perfectly loving and creates an individual with a capacity for
such relationship, he must surely seek to help him exercise it.
(2) This brings us to our second question. Would a perfectly
loving God seek to be personally related to us only at some future
point, perhaps after death, or already in this life? The correct an¬
swer, it seems to me, is “already in this life.” For the points I have
just made, if legitimate at all, would seem to support as well the
further claim that God seeks to bring about a personal relationship
with himself for human beings capable of such relationship at all
times at which they are so capable, and hence not only in the here¬
after but in this life as well. As Marilyn Adams puts it:
For each created person, the primary source of meaning and satisfac¬
tion will be found in his/her intimate personal relationship with
God. This relationship will also be the context in which a created
person can be best convinced of his/her worth, because it is the
place where God’s love for the individual is most vividly and inti¬
mately experienced. Christians naturally see it as to everyone’s ad¬
vantage to enter into this relationship as deeply as one can in this
world, as soon as possible.'7
But this point is not reached as quickly as my fairly complicated description of this
capacity might seem to suggest.
17. Marilyn McCord Adams, “Forgiveness: A Christian Model,” Faith and Phi¬
losophy, 8 (1991), 291.
Now Adams is writing from within the Christian tradition, but I

would suggest that what she and other Christians “naturally see”
to be true ought to be more widely held. A personal relationship
with a loving God could, as she writes, only enhance my well¬
being at any time at which I exist and, we might add, would not
detract from—indeed, would contribute to—my deepest well-be¬
ing and the well-being of others. For if there is a God, my deeper
well-being lies in a deeper relationship with him. And because of
its ethical contribution, such a relationship could, it seems, only
result in the enhancement of the well-being of others. We might
go on to point out once more that God would, at any time, desire
personal relationship with us for its own sake as well. A loving
God, we might expect, would bring us into existence so that he
might enter into fellowship with us—for our sakes, but for its own
sake too. We have, then, reason to suppose that there is no time at
which some human being is to some extent capable of personal
relationship with God but at which God does not wish the poten¬
tial represented by that capacity to be realized.18 God may create
beings without such capacity, but if the beings he creates have it,
then, at any time that they have it, to the extent that they do, we
may expect that he will wish them to exercise it.19 In this sense
also, God’s love must be unlimited.
Now to this, someone may be inclined to object as follows: “Al¬
though there is no reason for God to refrain from relating person¬
ally to us at some time at which we are capable of such relation¬
ship, there is no reason for him not to do so, and he therefore very
well might do so. More specifically, God need only do what he is
obligated to do, and relating personally to human beings is not
among his obligations.”
It seems to me, however, that all talk of obligation is out of
18. The fact that it is at some time a limited capacity does not imply that God
does not wish it to be exercised, for he may have—and if religious claims about
freedom are correct, has—created us this way so that we may have the opportunity
of growing in personal relationship with himself as a result of our own free choices.
19. Perhaps God will also allow some to lose the capacity for a relationship with
him or to suffer a diminished capacity because of, for example, the free actions of
others. I am thinking here of those unfortunate individuals who have been de¬
prived and/or abused in childhood and so are suspicious of everyone, incapable of
trust in man or God. (This fact may, of course, itself constitute a problem of evil,
but it is not the one that concerns us here.)
Epistemic Implications of Divine Love [ 27
place here. Even if God may legitimately be said to have moral

obligations (and some would deny this), the objector must still face
the fact that it is not the nature of obligation-fulfillment but of love
that we are exploring. Love transcends obligation. It is spontaneous
and supererogatory, and naturally seeks the well-being of its object
in relation to itself. This is true even of human love at its best.
Parents who love their children fully do their best to ensure that it
is always possible for their children to draw on the resources of
personal relationship with them. If we add to this that love seeks
personal relationship for its own sake, it seems that we have good
reason to make the all-inclusive claim in question.
A qualification must, however, be entered here. For a personal
relationship of the sort in question is not something God can bring
about on his own: God may wish to be personally related to me,
but if I choose not to respond to his overtures, personal relations
will not exist between us. Indeed, there is reason to suppose that an
emphasis on freedom is itself essential to the explication of Divine
love. Love, as John Macquarrie puts it, involves “letting-be, a re¬
spect for the otherness, freedom and individuality of the beloved.”20
And as John Hick points out, freedom is essential to personal rela¬
tionship: “In a [personal] relationship we apprehend and treat the
other person as an autonomous mind and will, a responsible and
self-directing consciousness with views and rights of his own
which must be consulted and respected—in short, as another per¬
son.”2' Hence it may seem that our claim should be that God will
bring it about that it is at all times possible for us to relate person¬
ally to him if we so choose.
The point about freedom, however, requires us to go farther
still. For a loving God, out of respect for our freedom, might well
allow us to shut him out altogether—not only to fail to respond to
his overtures, but also to put ourselves in a position where these
were no longer noticed. Such resistance of God would, of course, be
culpable, for it would involve shutting out one whom we had seen
to be our creator, and perfectly good, as well as the culpable activ¬
ity of self-deception: in exercising our freedom in this way, we
would be bringing it about through our own actions and/or omis-
20. John Macquarrie, In Search of Humanity (London: SCM Press, 1982), p. 180.
21. John Hick, Faith and Knowledge, 2d ed. (London: Macmillan, 1988), pp.
128-129.
sions that what was once seen was seen no longer. But if God is
perfectly loving, and treats us as persons, he will, we may suppose,
permit even this extent of freedom over against himself Hence the
clarified claim should read as follows: God will bring it about that,
unless we culpably put ourselves in a contrary position, it is at all
times possible for us to relate personally to him if we so choose.
Or to put it more formally,
Pi If God exists and is perfectly loving, then for any human

subject 5 and time t, if 5 is at t capable of relating person¬
ally to God, S at t is in a position to do so (i.e., can at t do
so just by choosing to), except insofar as 5 is culpably in a
contrary position at t.22
Our discussion so far, then, suggests that we have good reason

to affirm Pi. In the absence of our own attempts to bring it about
that a contrary state of affairs obtains, a perfectly loving God must
surely bring it about that we are in a position to relate personally to
himself. But some will no doubt feel that this claim requires more
in the way of defense than I have given. It may be thought, for
example, that I am asking for the beatific vision in this life. But the
“personal relationship with God” referred to here is not to be
viewed as identical with the beatific vision, although it might cul¬
minate in such an experience. As my use of the word “culminate”
already suggests, the relationship I am thinking of is to be under¬
stood in developmental terms. Were it to obtain, it would admit of
change, growth, progression, regression. It might be shallow or
deep, depending on the response of the human term of the rela¬
tion. This is, of course, what we would expect if the relationship is
conceived as a relationship between God and beings caught up in
the toil and vicissitudes of earthly life. Such a relationship belongs in
this life: now, in the midst of earthly pain and conflict, is when we
require Divine guidance, support, consolation, and forgiveness. In
light of this, as well as of the other points we have adduced, I
would suggest that there is indeed reason to suppose that a being
22. It is important to note that “capable of’ and “in a position to” are here
understood in such a way that someone might be capable of a personal relationship
with God at a time—have the requisite cognitive and affective machinery—with¬
out being in a position to exercise her capacity at that time and so enter into the
relationship.
who did not seek to relate himself to us explicitly in this life—who

elected to remain elusive, distant, hidden, even in the absence of
any culpable activity on our part—would not properly be viewed
as perfectly loving.
Let us conclude this section by inspecting briefly the theological
credentials of this view. We may do so by looking at the notion of
“salvation.” Salvation is often said to be something human beings
attain or may seek to attain. But this language of attainment, when
applied to salvation, is potentially misleading. In focusing our at¬
tention on the eschatological goal, it may obscure the fact that the
process presupposed by the attaining of that goal, and indeed the
events that initiate such a process, are also referred to by religious
writers in salvific terms. “Salvation” in theology not only refers to
the eschaton but also has application in the preeschatological life of
the believer. The faith by which one begins upon the religious way
is held to facilitate a personal relationship with God and deepened
relations with others in the here and now. Thus, if we are to employ
the language of attainment appropriately, we must say that salva¬
tion is an attainment partially realized in the present that may, as a
result of activity in the present, be ever more fully realized. Or, to
put it another way, the ultimate attainment that the hoped-for es¬
chatological events represent is not discontinuous with the attain¬
ments of the present but is a fuller, deeper, perfected version of
them. Grace Jantzen sums it up nicely:
Salvation is not (or at least not primarily) about our future destiny
but about our relationship to God and the gradual transforming ef¬
fect of that relationship in our lives. ... If religious experience is
centrally the sense of the loving presence of God, gradually helping
people to reorient and integrate their lives in accordance with their
love for him, is this not precisely what salvation is? Salvation must,
surely, be religious experience if anything ever is: not in the sense of
being a single climactic experience . . . but in the sense of a gradual
opening of all life, all of experience to the wholemaking love of
God.23
Hence theologians, too, seem committed to the affirmation of Pi.
23. Grace Jantzen, “Conspicuous Sanctity and Religious Belief,” in William J.

Abraham and Steven W. Holtzer, eds., The Rationality of Religious Belief (Oxford:
Clarendon Press, 1987), pp. 128-129.
] Framing the Argument
Some Epistemic Implications
If the preceding claims about what a loving God would do are

correct, then given an uncontroversial intervening premise, certain
interesting conclusions about theistic belief must follow. I will now
attempt to clarify these.
The intervening premise I have in mind is foreshadowed in my
earlier reference to the awareness of God. It states that a personal
relationship with God entails belief in Divine existence, that is, en¬
tails a disposition to “feel it true” that God exists.24 This claim
seems obviously true. For I cannot love God, be grateful to God,
or contemplate God’s goodness unless I believe that there is a God.
An adequate description of such attitudes and actions entails refer¬
ence to belief in propositions such as the following: “God is the
source of my being”; “God loves me”; “God is to be praised.” And
clearly, one can only believe propositions such as these if one be¬
lieves that God exists.25
It is important to note that my point here is a logical one. There
is something logically amiss in the suggestion that I could display
attitudes and perform actions of the sort in question without being
disposed to feel it true that God exists. It is not as though someone
who cannot be grateful to God or praise God because she does not
believe there is a God could do so if only she tried a little harder.
Such attitudes and actions are not just contingently difficult but
logically impossible for one who does not believe that God exists.
Since one cannot add to one’s beliefs just by choosing to (since
belief is involuntary), it follows that while in a state of nonbelief I
am not in a position to relate personally to God. But then, given
24. I am here following L. J. Cohen, who defines “belief that p” as “a disposi¬

tion to feel it true thatp” (“Belief and Acceptance,” Mind, 98 [1989], 368). Cohen’s
definition conforms quite closely, I think, to actual usage. It is at any rate very
helpful in pinpointing what I am claiming to be logically presupposed by personal
relationship with God. (More is said on the topic of the nature of belief later in this
section.)
25. As Adams puts it, “it is our highest good to be related in love to God, and
. . . we have to believe that he exists and loves us in order to be related to him in
that way” (Virtue of Faith, p. 20). Adams’s claim (as well as mine) echoes a much
older claim: “anyone who comes to God must believe that he exists” (Hebrews
11:6, New International Version).
Epistemic Implications of Divine Love [3
Pi, we can infer that God will seek to bring it about that I am
never in a state of nonbelief. More exactly, we can infer that
P2 If God exists and is perfectly loving, then for any human

subject 5 and time t, if S is at t capable of relating person¬
ally to God, 5 at t believes that God exists, except insofar
as S is culpably in a contrary position at t.
For if God will bring it about that (insofar as I am capable and

unless I resist) I am always in a position to relate personally to him,
and if the latter state of affairs obtains only if I always believe that
God exists, it follows that God will bring it about that (insofar as I
am capable and unless I resist) I at all times believe that God exists.26
Let us now try to get a little clearer about the content of P2.
What, more specifically, does belief involve? And what would God
do to facilitate belief in his existence (call this the belief that G)?
Beginning with the first question, we may note that the concept
of belief is a “graded” concept: varying degrees of belief thatp—of
disposition to feel it true that p—seem possible.27 I may believe p
weakly or firmly; I may have a weak or a strong disposition to feel
it true that p. While I believe both that I am presently in Calgary,
Canada, and that the distance between Calgary and Oxford, En¬
gland, is 5,000 miles, the former belief is stronger than the latter.
Even though both of the propositions in question are felt by me to
be true—are constitutive of what I feel to be so about the world—I
am more willing to give credence to the idea that I am wrong
about the second than to the idea that I am wrong about the first.
26. It might be held that belief is not just necessary but also su fficient to put one
in a position to enter into personal relationship with God, since anyone who has
the relevant emotional and intellectual capacities and believes can, just by choosing
to, contemplate God’s goodness, cultivate a loving and trusting attitude toward
God, and so on. But then, it might be concluded, P2 is in fact equivalent to Pi. I
have considerable sympathy for this claim, but it is not necessary, for our pur¬
poses, to endorse it. And the weaker claim, which is required, is quite obviously
true.
27. The view of belief as “graded” is discussed by Alvin Goldman in Epistemol¬
ogy and Cognition (Cambridge, Mass.: Harvard University Press, 1986). See esp. p.
324. Goldman rejects the graded notion of belief in favor of a categorical one, ac¬
cording to which the onset of belief marks the complete victory of one contender
over all others. My view, it will be noted, is a sort of compromise position.
This understanding of belief as graded may, however, seem to

present a complication for my argument. For does not a personal
relationship with God presuppose firm belief? If belief is graded,
must we not say that not only belief, but belief of a certain degree o f
strength is necessary for a personal relationship with God?
Such a concession would indeed complicate matters, but I do not
think it is required. For however weakly I believe, I am still, by
definition, disposed to see the proposition in question as true. As
far as I am concerned, it still reports the way the world is in rele¬
vant respects. There is, in other words, a categorical element in
belief. If I believe p, I must be disposed to feel it true; and if I am
not so disposed, I no longer believe it. However weak or strong
the belief, if it is a belief, it finds a place in my worldview; and
although I may be aware of alternatives to what I believe, these
will, to varying degrees, remain in the background. If alternatives
intrude too much, I must come to be uncertain whether p is true
and so cease to believe. A closely related point is that in believing
p, I do not usually think much about the fact that I am believing p.
There is commonly a shift at the moment of belief formation from
thinking about the proposition, its epistemic status, and so forth, to
thinking in terms of it. And this must also be true of weak belief,
for otherwise it could not be belief at all. If these arguments are
correct, the contrast between weak and strong belief that p is not
accurately described as a contrast between, for example, feeling it
true that (probably) p and feeling it true that (certainly) p. If I be¬
lieve that p occurrently, I have the thought that p, not thoughts
about evidence. (Or at any rate, if I have thoughts of the latter
sort, they are not mistaken for my thought that p.)
What all of this would seem to indicate is that we need not con¬
cede that belief of a certain (presumably quite high) degree of
strength is necessary for personal relationship with God. Even a
weak belief that God exists is compatible with gratefulness, love
toward God, trust, contemplation, and the like, for even a weak
belief involves a disposition to feel it true that G. If I feel, however
weakly, that it is true that there is a God, I may be moved to praise
him and to struggle with him in prayer in ways that would be
ruled out were I to, for example, be uncertain whether G was true.
Now, of course, perhaps God would wish that I normally believe

firmly. Weak belief, whether I am conscious of it or not, may
carry with it a certain measure of stress, and this may affect the
relationship. But, on the other hand, God might have reasons for
leaving me for a time in a state of weak belief and, given that firm
belief is not required for a personal relationship with God, might
very well do so.28 At any rate, relationship-related considerations
do not provide nearly as strong a presumption in favor of the con¬
trary view as they would if firm belief were necessary. I will there¬
fore continue to speak of God wishing to facilitate belief, meaning
by this some degree or other of a disposition to feel it true that G.
We must now take up the second question mentioned above:
What will God do to facilitate belief in his existence? I would sug¬
gest, as a first approximation, that God would provide evidence that
is sufficient to produce belief.29 For if belief is involuntary, then, if I
am to believe that G, there must be something or other apart from
my own choice—some evidence—on account of which I feel it to
be true that G. Now it may be objected that God could simply
bring it about that it seems to me strongly that he exists—where
the “strong seeming” is something analogous to what produces
(for most of us) such beliefs as the belief that 2 + 2 = 4 —instead of
providing evidence. But on the broad understanding I am here as¬
suming, “evidence” refers to anything that can serve as a ground of
belief, and so not only to propositions that provide the basis for
deductive and inductive inference but also to nonpropositional, ex¬
periential evidence in which belief may be directly (noninferen-
tially) grounded. Hence the objector’s point can be accommodated.
For presumably, if it seems strongly to S that p is true, S may
point to that experiential circumstance—the circumstance of it seem¬
ing strongly to S that p is true—as the ground of his belief. 3" More
28. And conversely, God might at times (as we will see) have reason to facilitate
a stronger belief.
29. What I say here is meant to apply both to the initial acquisition of belief and
(should it be retained) to its persistence.
30. This view is defended by Alvin Plantinga. See his “Reason and Belief in
God,” in Alvin Plantinga and Nicholas Wolterstorff, eds., Faith and Rationality
(Notre Dame: University of Notre Dame Press, 1983), pp. 78-79.
34 1 Framing the Argument
generally, there is nothing to prevent God, on our understanding

of “evidence,” from bringing it about that I have certain experi¬
ences instead of providing an argument that has G as its conclusion.
This is not yet the whole story, however. For it is compatible
with what we have said so far that I be led to believe on inadequate
grounds, and this is surely not to be expected: God, we might
expect, would provide evidence that adequately supports belief. Let
us now look at this a little more closely. Why should we have this
expectation? The answer, it seems to me, must be given in terms
of the requirements of resistance and the nature of God. We have
said that God would bring it about that it is only if S has put
herself in a contrary position culpably that S is not in a position to
relate personally to himself. If this is true, then, clearly, God
would not take actions to facilitate belief that left open the pos-
ssiblity of inculpable resistance. And so he would provide adequate
evidence, for if the evidence was not adequate, S might very well
come to see it as such and inculpably reject it.
Perhaps it will be replied here that God as omnipotent could
easily prevent me from ever viewing the evidence he provided as
poor (even when it was). But were God to take this route, he
would systematically deceive us by bringing it about that when¬
ever the (actually inadequate) evidence was examined, it was
viewed as adequate; and this seems incompatible with his perfect
goodness.31 This point, indeed, provides us with an independent
reason for supposing that the evidence provided would be ade¬
quate. For a perfectly good God would not permit his intentions to
be fulfilled by deceitful means. And so, it seems, he would not
permit me to believe that G on grounds I viewed as adequate, but
that did not adequately support that belief. We might go farther
and point out that even if I did not explicitly consider my grounds
or the support they provided, God would still deceive me by
31. The qualification “whenever the evidence was examined” is to be carefully

noted. 5 may believe without reflecting on her evidence and without articulating to
herself exactly how it provides support for G or how much support it provides
(how probable G is rendered by it). If we did not suppose this, we would have to
say that individuals who do not have the capacity for such reflection and evaluation
(e.g., small children) could not believe that G; and this is an implication to be
avoided. I am indebted to William Alston for drawing the need for some such
qualification to my attention.
bringing it about that they were inadequate. For in that case, my

degree of firmness in believing (whatever it was) would not corre¬
spond to the degree of support provided by the evidence; and so—
although I would not believe a false proposition about the evi¬
dence—God would have brought it about that I was, so to speak,
living out a lie. We may therefore conclude that if God provides
me with evidence, intending thereby to produce belief in his exis¬
tence, it will be evidence that not only is sufficient to produce be¬
lief but also adequately supports it.
If now we consider what sort of evidence would provide the
degree of objective support required here, I think we must say
probabilifying evidence—evidence that renders G probable.32 As
William Alston has argued, it is only if a ground renders a claim
probably true that the formation of belief on that basis would be
“desirable from the epistemic point of view.”33 But there is another
reason too: it is just not possible for anyone who considers the
evidence on the basis of which she believes to continue believing
unless it seems to her to render the proposition in question proba¬
bly true. For suppose 5 offers the following description of her
mental state: “I feel that G is true—that G conforms to the way the
world is. But the arguments for G known to me (i.e., the public
evidence) do not seem to me to favor G: neither individually nor
cumulatively do they seem to me to show G to be more probable
than its denial. Nor do I have any private evidence—any hunch or
feeling or experience—that favors a contrary view.” This descrip¬
tion seems clearly contradictory, and the contradiction is one that
S could hardly fail to see:
(i) I feel G to be true and I do not feel G to be true.
32. I do not wish to be taken as suggesting that God would ensure that his
existence was probable on the totality of the evidence that exists—whatever that
might be—or on the public evidence available—that is, on the set of propositions
that provide the premises for arguments in natural theology. All we have seen
reason to suppose (and all that the arguments immediately below suggest) is that
God would bring it about that his existence was probable on S’s evidence; and this
evidence, as we have seen, might well include not only propositions of the sort
mentioned above but certain of S’s experiences too.
33. William Alston, Perceiving God: The Epistemology of Religious Experience
(Ithaca, N.Y.: Cornell University Press, 1991), p. 74.
Hence the claim that one could consider one’s evidence and con¬
tinue to believe while not holding it to render the proposition one
believed probable entails that S could believe a contradiction. But
this does not seem to me to be something anyone could do. Hence
that claim is false. But if it is false, we do indeed have additional
reason to suppose that the evidence God would provide would be
probabilifying evidence. For only that way could God ensure
(without deception) that anyone who examined her evidence con¬
tinued to believe.34
Perhaps it will now be objected that while probabilifying evi¬
dence is necessary, it would not be sufficient. Someone might ar¬
gue for this as follows: “For those who consider the evidence, only
the belief that the evidence renders G very probable would be suffi¬
cient to produce the belief that G. And so there could be a situation
in which God provided probabilifying evidence and S saw it as
such, but in which S did not come to believe, or in which S came to
see the evidence on which she believed as probabilifying and forth¬
with ceased to believe. To avoid this, God would have either to
deceive S into supposing the evidence to be stronger than it was or
to provide stronger evidence. But obviously, if perfectly good,
God would not choose the former route. Hence he would choose
the latter. But then we cannot rest content with the claim that God
would provide evidence that rendered his existence probable. We
must, instead, say that he would provide evidence that rendered
his existence very probable.”
In response to this objection, I would suggest that we have good
reason to reject the claim about belief on which it depends. We
might first point out, with Richard Swinburne, that it is “tidier” to
suppose that belief will exist as soon as the probability of the prop¬
osition believed is perceived as greater than o. 5 than to identify the
34. I would add here that on my view, any claim to the effect that S holds that
G is not more probable than not-G and yet (irrationally) believes that G is most
charitably interpreted as ignoring private evidence held by S to favor G, or as
confusing belief with acceptance (a commitment to act-as-if some proposition is
true which does not necessarily involve belief that the proposition is true). It is of
course also possible for S, through self-deception, to lose the belief she once held
that G and not-G are at epistemic parity and come to believe that G is true. But I
can make no sense of the suggestion that S could at one and the same time hold both
that G is not more probable than not-G and that G is true.
point at which belief arises with some (inevitably arbitrary) value

of probability between 0.5 and i.35 But other reasons might also be
adduced. Suppose I believe that G has a probability of 0.6. Then I
believe that the world is to some extent in favor of G, and (it
seems) I will have at any rate a weak disposition to feel it true that
G. It is clear that I may have such a disposition when the perceived
probability is less than 1—as the objector himself admits, if I view
G as very probable, I believe that G—so why should I not believe
weakly when the probability is taken to be 0.6? Of course, if I
consider the evidence to be evenly balanced, I will be uncertain
whether G is true. But if I see G’s probability as greater than 0.5, it
seems natural to suppose that I believe it to some extent.
This claim can call in its defense an argument of Nicholas
Wolterstorff concerning (what he calls) human “belief disposi¬
tions.” Wolterstorff suggests, following Thomas Reid, that “at any
point in our lives we each have a variety of dispositions, inclina¬
tions, propensities, to believe things.”3*' He provides a number of
examples:
What accounts for our beliefs, in the vast majority of cases anyway,
is the triggering of one or another such disposition. For example,
we are all so constituted that upon having memory experiences in
certain situations, we are disposed to have certain beliefs about the
past. We are all disposed, upon having certain sensations in certain
situations, to have certain beliefs about the external physical world.
Upon having certain other sensations in certain situations, we are all
disposed to have certain beliefs about other persons. Likewise we
are all so constituted as to be disposed in certain circumstances to
believe what we apprehend people as telling us.37
And a little farther on, he writes: “In addition to the features of our
constitution thus far mentioned, we are all so constituted that upon
judging some proposition which we already believe as being good
35. Richard Swinburne, Faith and Reason (Oxford: Clarendon Press, 1981), p. 5.
36. Nicholas Wolterstorff, “Can Belief in God be Rational?’’ in Alvin Plantinga
and Nicholas Wolterstorff, eds., Faith and Rationality (Notre Dame: University of
Notre Dame Press, 1983), p. 149.
37. Ibid.
evidence for another proposition not yet believed, we are disposed

to believe that other proposition as well.”38
In the spirit of Wolterstorff s argument we may wish to claim
that when someone takes a proposition to be probable—even mar¬
ginally so—a belief disposition is triggered and the individual
comes to (in some degree) feel the proposition to be true. We
might even speculate that there is an evolutionary reason for this.
Perhaps it is necessary that we have dispositions of this sort if we
are to have, in Swinburne’s words, a “map or view of the world”
which can guide our actions.34 Whatever the case, there seem to be
introspective grounds for supposing that we do have such disposi¬
tions. And if we do, the belief that G is probable is indeed suffi¬
cient for the belief that G. Hence the argument that suggests it is
not can safely be rejected.
Now all of this (as suggested above) is of course subject to the
qualification at the end of Pi and Pi, namely, that God will at all
times leave it open to us to resist his facilitative endeavors—to re¬
sist the actions he takes to put us in a position to relate personally
with himself. Transposing this into the key of the present discus¬
sion, we must say that S will remain free to bring it about that he
is in a position incompatible with belief (and so no longer in a
position to relate personally to God) by opposing the evidence that
God provides. Since such resistance would be culpable, it follows
that S would remain free to culpably bring about the loss of belief.
Taking this point (and the clarificatory points above) into ac¬
count, we may now express Pi more fully as follows:
Pi' If God exists and is perfectly loving, then for any human
subject S and time t, if S is at t capable of relating person¬
ally to God, S at t believes that G on the basis of evidence
that renders G probable, except insofar as S is culpably in
a contrary position at f.40
38. Ibid., p. 150.

39. Richard Swinburne, The Evolution of the Soul (Oxford: Clarendon Press,
1986), p. 122.
40. It may be said that given the way P2' is phrased, we may argue not only
from reasonable nonbelief to the nonexistence of God, but also from reasonable
(but inadequately grounded) belief. For P2' states that all will believe on good evi-
P2' represents my clarified estimate as to the sort of epistemic

situation a perfectly loving God would seek to facilitate. If there is
a perfectly loving God, 5, unless prevented by her own culpable
activity, will at all the times in question find herself in possession
of evidence that renders G probable and will in some degree be¬
lieve that G. If such a situation were to obtain, 5 could only fail to
believe that G on good evidence culpably, and so (given my defini¬
tion of “reasonable”) G would be beyond reasonable nonbelief for
S. What P2' states, therefore, is that if a perfectly loving God ex¬
ists, our situation will be one in which God’s existence is beyond
reasonable nonbelief for all who are capable of a personal relation¬
ship with God at all times at which they are so capable.41 For the
sake of ease of reference and stylistic variation, I will hereafter refer
to such a situation as one in which God’s existence is beyond rea¬
sonable nonbelief, or as one in which there is evidence sufficient
for belief, or as “a strong epistemic situation in relation to theism.”
A weak epistemic situation in relation to theism I will take to be a
situation in which God’s existence is not beyond reasonable non-
belief—where, for one or more human beings capable of a per¬
sonal relationship with God at one or more times at which they are
dence, and is it not obvious that many actual believers do not? This point may be
conceded, but it does not cast into question the approach I am taking. For the
claim of the one who argues from inculpable (but poorly grounded) belief to the
nonexistence of God—who emphasizes that a loving God would provide good
evidence—depends on the claim (and arguments for the claim) that God would pro¬
vide evidence at all. The success of my argument, in other words, is a necessary
condition for the success of any argument from inculpable (but poorly grounded)
belief. To put it yet another way, the claim under consideration is really a conjunc¬
tive claim—“God will provide evidence, and the evidence God provides will be
good evidence”—the first conjunct of which must be supported by an argument of
the sort provided in this book. If this argument does not succeed, there is no
reason to suppose that the other will. And if it does succeed, the other will be
superfluous.
41. It may be objected that this conditional is not in fact equivalent to P2'. For
as we have seen, S could be fooled into believing that poor evidence was good and
so might very well find that G was beyond reasonable nonbelief for her—might
find that she could not reject it without resisting what she took to be good evi¬
dence—without it being the case that G was in fact rendered probable by her
evidence. I will simply assume, however, that what is meant by “G is beyond
reasonable nonbelief for S” is “5 can only fail to believe that G on good evidence
culpably,” in which case the equivalence holds.
so capable, evidence of the sort in question is not available. In

other words, on my definition “a weak epistemic situation obtains”
is simply the negation of “a strong epistemic situation obtains.”
In bringing the chapter to a close, I will briefly consider two
objections that might be brought against the argument of this sec¬
tion. According to the first, although it is necessary to believe that
God exists in order to be in a position to relate personally to him,
such belief is not all that is needed. One must also believe various
other religious propositions, such as “God was in Christ, reconcil¬
ing the world to himself.” But then (the objector may claim) I am
committed to arguing not only that P2' is true, but also that var¬
ious P2' counterparts (which refer to these other propositions) are
true, and this may considerably lessen the appeal of my view: the
claim that God would be interested in ensuring that all human be¬
ings at all times are apprised of the truth of all these propositions
is, to say the least, a large claim. Hence (he may say) we can argue
as follows: The view that God might reveal any one of the propo¬
sitions in question entails a similar claim in respect of all the others.
But the general claim is unacceptable. So any of the more specific
ones—including, in particular, the claim concerning G—is also un¬
acceptable.
I answer as follows. It seems that if I believe that G, I may, by
responding appropriately to this belief and to belief in the various
propositions I see to be clearly entailed or rendered probable by G,
turn it into faith and so (if there is a God) enter into personal rela¬
tionship with God. Specifically, if I come to believe that there is a
perfectly loving God, I will also believe such propositions as that I
owe my existence to God, that my well-being lies in relationship
with God, that other individuals are loved and valued by God, that
I too should seek to act toward them in loving ways. And individ¬
uals who hold such beliefs are clearly in a position to act upon
them by thanking and praising God, praying to God, cultivating a
loving disposition towards others, and so forth. In other words,
acquiring the belief in question does seem to put a person in a posi¬
tion to relate personally to God—at least if we assume (as the dis¬
cussion in the first section of this chapter surely allows us to do)
that God, if he exists and is perfectly loving, is disposed to respond
in appropriate personal ways to the activities of such a believer.
Now no doubt there are, if there is a God, many other interesting

and important religious truths not clearly entailed or rendered
probable by the proposition “God exists,” but I do not see that
awareness of these is essential to a personal relationship with God if
such a relationship is construed (as it is here) in developmental
terms. Belief in the existence of a perfectly loving God, on the
other hand, is clearly necessary to get one started in such a relation¬
ship: without it, as we have seen, explicit Divine-human reciproc¬
ity is ruled out. So I do think it appropriate to focus on belief in
God’s existence as we are doing. After this belief is acquired, many
others may follow in due course. But if God exists, and if the
individual has responded appropriately to his belief in God’s exis¬
tence, the acquisition of each such additional belief will be an event
within the relationship and not a prerequisite for it.
I have been arguing that if God exists and is perfectly loving,
humans will be given access to evidence sufficient for belief in
God’s existence. The second objection to this claim I will consider
is a theological objection. Many recent theological writers, sensi¬
tive to the presence in our world of nontheistic religious traditions,
as well as of morally sincere individuals who are nonetheless non¬
religious, have claimed that there is no reason why persons who
fail to believe that there is a God should not achieve salvation, and
(by implication) that even a perfectly loving God, concerned for
our salvation, need not be concerned to facilitate belief for every¬
one. In light of my earlier comments on salvation (which included
the suggestion that theological statements about salvation provide
confirmation for my claims), this view may seem to require a re¬
sponse. I will briefly consider the claims of one of its foremost
exponents, Karl Rahner.
Rahner develops the Catholic position that nonbelievers may ex¬
hibit implicit belief. Those who do, he (notoriously) calls “anony¬
mous Christians.”42 The following passage from the essay of that
title expresses his view succinctly:
42. See Karl Rahner, “Anonymous Christians,” in G. A. McCool, ed., A

Rahner Reader (London: Darton, Longman, and Todd, 1975), p. 211. Further refer¬
ences to this essay will be made parenthetically in the text.
No matter what a man states in his conceptual, theoretical and reli¬

gious reflection, anyone who does not say in his heart “there is no
God” . . . but testifies to him by the radical acceptance of his being,
is a believer. But if in this way he believes in deed and in truth in the
holy mystery of God . . . then the grace of this truth by which he
allows himself to be led is always already the grace of the Father in
his Son. And anyone who has let himself be taken hold of by this
grace can be called with every right an “anonymous Christian.” [P.
214]
Elsewhere in the essay, “the radical acceptance of being” (which is

also referred to as the acceptance of “transcendence” or of “limit¬
less openness” [p. 213]) seems to be spelled out in terms of moral
commitment: “We can say quite simply that wherever, and in so
far as, the individual makes a moral decision in his life . . . this
moral decision can also be thought to measure up to the character
of a supernaturally elevated, believing and thus saving act” (p.
218). Rahner also speaks here of “loving humaneness” (p. 214), and
suggests that “an atheist can be justified and receive salvation if he
acts in accordance with his conscience” (p. 221).
In response to Rahner, we may point out that such claims seem
to comport ill with views concerning God’s love to which—as the
following passages from “Anonymous Christians” show—he is
also committed.
Love in its truly personal sense is . . . the ceding and the unfolding
of one’s inmost self to and for the other in love. [God wishes to
enter] into unrestricted personal communion with man. [P. 134]
God makes a creature whom he can love: he creates man. He creates

him in such a way that he can receive this Love which is God him¬
self, and that he can and must at the same time accept it for what it
is: the ever astounding wonder, the unexpected, unexacted gift. [Pp.
186-187]
Man should be able to receive this Love . . . ; he must have a conge¬

niality for it. . . . He must have it always. He is indeed someone
always addressed and claimed by this love. For ... he is created for
it; he is thought and called into being so that Love might bestow
itself. . . . The capacity for the God of Self-bestowing personal Love
is the . . . abiding existential of man as he really is. [Pp. 187-188]
These passages seem to make my point much more effectively than

Rahner’s. Indeed Rahner admits that in “explicitness,” belief finds
“its greatest support and confidence” (p. 214). So what should we
say about the earlier claims concerning implicit belief? It would
seem that in Rahner’s case, as in the case of other writers who
address this question, such claims presuppose either that God exists
(which assumption would permit one to draw the conclusion that
implicit belief, since it is in many cases the only form of “belief’
possible, somehow falls within God’s purposes), or that theo¬
logically acceptable arguments can be adduced to show that despite
the support that exists for the contrary view, explicit awareness of
God will be absent in many cases. But these assumptions are in this
context at best inapplicable, and at worst question-begging—the
first because we are exploring an argument against God’s existence,
and the second because we have not yet come to consider whether
the support for P2' (to which it alludes) can be overridden: we
have only been concerned to elucidate that support. Certainly we
cannot assume that that support will be overridden. We may con¬
clude, therefore, that this second objection, like the first, does not
take anything away from the considerable force of the argument
for P2'.
[ 2 1
Is a Strong Epistemic
Situation in Relation
to Theism Possible?
According to the argument of the previous chapter, there is rea¬

son to suppose that a perfectly loving God would ensure that the
epistemic situation of humans in relation to theism was a strong
one, that is, would bring it about that his existence was beyond
reasonable nonbelief. But even if God apparently would wish to
bring about such a state of affairs, perhaps it would not be possible
(i.e., logically possible) for him to do so. If it would not be possi¬
ble, then of course, any argument purporting to show that such a
state of affairs will obtain if God exists collapses; for even an om¬
nipotent God cannot do what is logically impossible for him to do.
Hence it is important, in order to satisfy a necessary condition for
the success of the prima facie case I am constructing, that some
space be devoted to showing that the situation in question is indeed
possible—that there is a possible world in which God brings it
about that his existence is beyond reasonable nonbelief.
My procedure is as follows. In the first section of the chapter I
briefly examine certain general considerations that might be held to
show the situation in question to be impossible and offer certain
considerations of my own that seem to me to provide at least initial
support for the contrary view. In the second section I attempt to
provide further and conclusive support for this view by describing
in detail one way in which the possibility in question might be
instantiated.
Is a Strong Epistemic Situation Possible? [ 45
Opening Arguments
Some writers would no doubt be inclined to argue that it is not

possible for God to put his existence beyond reasonable nonbelief,
and that we can show this without examining particular examples
of situations God could bring about. It might be argued, for exam¬
ple, that the state of affairs in question is ruled out by the fact of
human finitude. Thus John Macquarrie:
Contemporary theology probably reckons more seriously with the

finitude of human existence than many theologians of the past have
done. . . . To be finite is to live in risk and uncertainty, and that this
is our life is clear to us from everyday experiences in which we have
to commit ourselves to policies of action without complete knowl¬
edge of all the relevant circumstances and still less of all the conse¬
quences that will flow from the action. Our life in this world is not
one that can be based only upon the certitude of knowledge—the
man who tries to live this way, without risk, never really lives at
all—but one that must go out in faith. This is true about the under¬
standing of our life as a whole—we see it only from our limited
standpoint and cannot know the ultimate truth about it. Thus to
demand the guaranteed certitude of rational demonstration (that
there is a God . . . ) is to refuse to acknowledge one’s own finitude.'
But our understanding of “beyond reasonable nonbehef’ does

not require that there be “certitude”; only that there be, in the ab¬
sence of resistance, some degree of belief that G. Perhaps the for¬
mer state would be difficult for finite beings to maintain (although
certain fundamentalist groups seem to belie this); we must note
that the latter does not seem at all like it in this respect. And while
it might seem impossible for beings who are radically finite to be
always convinced of the force of some “rational demonstration” of
God’s existence (although again there are apparent counterexam¬
ples), it is not at all obvious that probabilifying evidence of the sort
that my argument requires must be such as finite beings like our¬
selves could not appreciate. Hence I suggest that the argument
from finitude does not succeed.
1. John Macquarrie, Principles of Christian Theology, 2d ed. (London: SCM

Press, 1977), p. 51.
A second argument that might be considered is an argument

from Divine transcendence. Now not just any transcendence will do
here. God may be very different from creatures, but so long as he
is describable by analogy with ourselves, it is not obvious that his
existence could not be put beyond reasonable nonbelief. This is
brought out very clearly and forcefully by Swinburne in the con¬
text of a related discussion:
God cannot be totally different if he is describable with words

which we use to describe mundane things—e.g. “wise”, “good”,
“powerful”. If these words do have, however analogical, an applica¬
tion to him, God must be something like wise, good and powerful
things on earth; and in that case kinds of arguments which are ap¬
propriate to prove the existence or non-existence of wise, good and
powerful things in principle have application to proving his exis¬
tence or non-existence.2
So the argument from Divine transcendence must claim that

God is to be understood as absolutely transcendent—that talk about
created things, including ourselves, does not have even an analogi¬
cal application to him. Now if this view is correct, then it is hard
to see how there could be evidence that probabilified God’s exis¬
tence or how God could meaningfully be described as acting so as
to provide such evidence. But then, by the same token, it is hard
to see how anything meaningful could be said about God. For this
reason, I would claim, theists—who are after all committed to say¬
ing quite a lot about God—are not in a position to think of God as
absolutely transcendent; and so this second argument is no more
successful than the first.
Can we go any farther than this? Are there considerations that
suggest God could put his existence beyond reasonable nonbelief? I
think there are. The first of these is that the evidence actually avail¬
able to individuals who inculpably fail to believe does go some way
toward showing that there is a God. Individuals who doubt, for
example, may nevertheless look upon such phenomena as the or¬
der manifested in the universe and religious experience as confirm-
2. Richard Swinburne, Faith and Reason (Oxford: Clarendon Press, 1981), p.

84.
ing to some degree the proposition that there is a God. And surely
no one could reasonably deny that such phenomena do render the
existence of God at least somewhat more likely than it would oth¬
erwise be. The second point (which will be seen to complement
the first) is that our evidential situation is not the only one that
could have obtained. This follows from the obvious truth that ours
is not the only possible world. We can conceive of a great many
ways in which our world might have been different. But then,
surely, the evidential situation of humans vis-a-vis God’s existence could
have been different as well: every contingently existing thing in the
world is relevant to the question whether there is a Creator, and so
if the world changes, the evidence changes too.
These points, when taken in tandem, would seem to provide a
certain amount of initial support for the claim under consideration.
If we accept that the evidence actually available to individuals in
our world goes some way toward showing that there is a God, and
accept as well that there are innumerable ways in which the world
(and ipso facto the evidence) could have been different, then it
seems we have some reason to accept that there is a possible world
in which stronger evidence is available—in particular, in which indi¬
viduals are always in the presence of evidence sufficient for belief.
The Possibility of Religious Experience
What we have seen so far is that we have no obvious reason to

suppose that God’s existence could not be put beyond reasonable
nonbelief and some initial reason to accept the contrary view. In
this section, I attempt to provide the latter with conclusive support
by means of a particular example—an example of a state of affairs
that seems clearly possible, and that would, if it obtained, instanti¬
ate the state of affairs consisting in God’s having brought it about
that his existence is beyond reasonable nonbelief.
It may seem to some that God could quite easily put his exis¬
tence beyond reasonable nonbelief by creating a world without evil
(or with much less evil than ours contains), or by producing now
and then spectacular and overwhelming events that could not rea¬
sonably be considered anything but counterinstances to true laws
of nature. But this claim seems dubious. Presumably God could
create a world without evil or with much less evil, but while this
might remove an important obstacle to belief, it is not clear that it
would make the required positive contribution to the epistemic cir¬
cumstances of everyone concerned. In any case, I wish to show
that a strong epistemic situation in relation to theism is compatible
with the existence of much evil. Overwhelming miracles may also
be possible, but miracles are by definition rare events, and so it is
not easy to see how the sort of evidence they might provide could
be generally and at all times available, as a strong epistemic situa¬
tion requires. Furthermore, I wish to show that God need not
overwhelm us in order to elicit belief, and to prove that the de¬
scription of a state of affairs of the sort in question need not be
theologically crude, as a description involving reference to Di¬
vinely produced spectacular events would no doubt seem to many
to be.
A more fruitful approach, I suggest, would focus on the possible
epistemic contribution of religious experience.3 Experiential evi¬
dence, as we will see, could be generally and at all times available.
It is also more likely to be religiously efficacious—to stimulate a
religiously appropriate response. I may, if I am presented with a
good argument or witness a spectacular miracle, conclude that
there is a God, but if God is present to me in experience, my re¬
sponse (if I respond positively at all) is perhaps more likely to be
the personal response a loving God would desire. Related to this is
the point that it is only religious experience that makes possible the
deepest forms of personal relationship between God and human¬
kind.4 Although perhaps not necessary for such relationship, reli¬
gious experience must obviously enrich it and contribute to its
flourishing.
Suppose, then, that the world is one in which all human beings
who evince a capacity for personal relationship with God have an
experience as of God presenting himself to them, which they take
to be caused by God and which actually is caused by God present-
3. I do not assume that this is the only line of thought it would be profitable to
pursue in this context, but, as I argue, it does seem an appropriate one.
4. On this point, see William Alston, Perceiving God: The Epistemology of Reli¬
gious Experience (Ithaca, N.Y.: Cornell University Press, 1991), pp. 303-304. This
impressive book is full of insights on our topic and has stimulated my thinking in a
variety of ways.
ing himself to their experience. This experience, let us say, is non-

sensory—an intense apparent awareness of a reality at once ulti¬
mate and loving which (i) produces the belief that God is lovingly
present (and ipso facto, that God exists), (2) continues indefinitely
in stronger or weaker forms and minimally as a “background
awareness” in those who do not resist it, and (3) takes more partic¬
ular forms in the lives of those who respond to the beliefs to which
it gives rise in religiously appropriate ways (for example, the be¬
liever who pursues a personal relationship with God may describe
his experience as that of the forgiving, com forting, or guiding presence
of God). Since the experience is had as soon as a capacity for per¬
sonal relationship with God exists, we may suppose that it occurs
quite early on in the life of each individual, in particular, before
any investigations as to the existence of God have been under¬
taken. We may further suppose that any investigations subsequently
undertaken (at any rate by individuals who have not taken steps to
hide from themselves the experience and its apparent implications
for belief) fail to undermine, and indeed reinforce, the beliefs
formed by this experience. In particular, those who encounter the
problem of evil and other objections to theistic belief continue to
believe that there is a God on the basis of their experience and take
what they believe to be confirmed as they discover the universality
of the experience and the uniformity of its descriptions—as they
learn that reports of an experience very similar to their own have
come from every time and place accessible to historical inquiry.
Let us now clarify one or two aspects of this description.
(1) The experiences in question are, as I have said, nonsensory,
that is, not mediated by sensations of any sort. As Swinburne puts
it in describing experiences of this kind, “The answer to ‘What was
it about your experience which made it seem to you that you were
having an experience of God?’ will be ‘It just did. There were no
visual, auditory or any other sensations which made it seem thus to
me.”’1 I do not mean to imply, however, that no religious experi¬
ences of a sensory sort are had by individuals in the world in ques¬
tion, or that if they are, none are caused by God.
(2) When I say that at the time in question humans have an expe-

p. 252.
rience caused by God being present to them, I do not mean to suggest

that prior to this time, God has not been lovingly present to hu¬
man beings or that their experience is necessarily due to some spe¬
cial action of God at that time. Let us take up each of these points
in turn. God, if he exists and is perfectly loving, is always present
and lovingly disposed toward us, but it is only at a certain stage,
when certain capacities are in place, that humans may become sen¬
sitive to the Divine presence. Just as the health of my optic nerve is
necessary if distant objects—which are there all the time—are to
cause me to see they are there, so it is that even though God, if he
exists, is always and everywhere present, certain capacities are re¬
quired if I am to become aware of God’s presence—if God’s being
present is to cause me to perceive that God is present. And in the
world I have described, these coincide with the capacities required
for personal relationship with God.
Does this causation imply any special activity on God’s part vis-
a-vis the individuals in question? It seems to me that it does not.
Suppose that the world I have described is actual and that Francine,
an inhabitant of this world, becomes capable of personal relation¬
ship with God at time t. Although the state of affairs consisting in
Francine’s becoming aware of God’s presence might obtain at t as a
result of God’s special intention at t that Francine should become so
aware, this does not seem necessary. Given the state of affairs I
have described, it may just as well be that the laws of nature which
God has caused to operate are such that when the appropriate stage
in Francine’s development is reached, Francine, so to speak,
“switches on” to the Divine presence. On this view, God is always
lovingly disposed to Francine, and when certain developmental
conditions are satisfied, she simply becomes aware of this, that is
to say, the fact of God’s being present to Francine becomes caus¬
ally efficacious in the relevant way vis-a-vis Francine. Perhaps not
much turns on which of these conceptions is accepted. But anyone
inclined to accept a noninterventionist view of God should take
note that my description does not rule it out.6
I turn now to a defense of my judgments as to the coherence of
the description I have given and what it entails.
6. See Maurice Wiles, God’s Action in the World (London: SCM Press, 1986) for
a statement and defense of such a view.
The state of affairs I have described seems clearly possible: its

description seems perfectly coherent. Surely it could be the case
that all human beings with a capacity for personal relationship with
God become aware of God’s presence.7 If it is suggested that only
persons who already have a concept of God and a wider experience
of life experience religiously, so that religious experience presup¬
poses a period of preparation following acquisition of the capacity
in question, we may reply that this is a contingent truth. An om¬
nipotent God would be able to create us with the required mental
furniture in place or, at any rate, would be able to bring it into
existence at the time when the capacity for personal relationship
with God was acquired. It has, indeed, often been held by theo¬
logians that the idea of the Divine is implanted in us by God. Now
whether this is true or not, surely no one would deny that it is
logically possible. But if so, the problem under consideration van¬
ishes.8
What about the rest of the description? Is it coherent? It seems to
me that it is. Belief could only fail to be formed in such circum¬
stances if the subject was deterred by strong apparent counterevi¬
dence, and given the temporal priority of the experience as I have
described it, this possibility is ruled out—that is, in the first in¬
stance at least, the subject’s attention is focused on his experience
and not on other evidence. And if the experience were universal
and its descriptions uniform, and if it was ongoing in the lives of
7. Remember, we are not asking whether actual experiences of God have oc¬
curred or whether it is reasonable for persons who have actually had religious expe¬
riences to view them as providing contact with God. What we are asking is
whether it should be deemed possible for God to bring human beings to awareness
of himself.
8. Some might argue that while my points show that an apparent awareness of
God could occur early on, we have no reason to suppose that anyone could actually
recognize at this stage (or perhaps any other) the presence of a being exemplifying
the collection of properties God is said to possess. But there is considerable support
in the literature for the view that such attributes could be directly given in experi¬
ence. See, for example, Swinburne, Existence of God, pp. 267-268; Gary Gutting,
Religious Belief and Religious Skepticism (Notre Dame: University of Notre Dame
Press, 1982), pp. 153— 155; Alston, Perceiving God, pp. 17, 59-63; and George
Mavrodes, Revelation in Religious Belief (Philadelphia: Temple University Press,
1988), pp. 104-109. Mavrodes defends the notion of “basic cognitive acts” analo¬
gous to basic physical acts. Earlier in his book (p. 94), Mavrodes makes what
seems to me to be one of the central points here: “What kinds of experience there
are [is] itself a matter of experience.”
those who did not resist it, the judgment of experients with respect
to God’s presence would (barring steps taken to remove belief)
surely be maintained indefinitely and, indeed, reinforced—even in
the face of evils such as our world contains. One’s own experience
is a very powerful stimulus to belief; only under strong pressure
from outside forces will what is suggested by it be rejected.
Suppose, however, that some experients came to view evil (or
some other apparent counterevidence) as a threat. As we have seen,
there are degrees of experiential force and of belief, and so it would
be possible for the force of the experience in such circumstances to
be increased. God might allow the individuals in question to feel
his presence more strongly—or at least strongly enough to sustain
a certain degree of belief. Furthermore, if a state of affairs of the
sort I have described were to obtain, much of the additional evi¬
dence—in particular, the public evidence concerning the distribu¬
tion and uniformity of this form of experience—would seem to
confirm the judgments of the individual with respect to her experi¬
ence and its epistemic implications. This apparent confirmation,
together with what has already been described, would, it seems to
me, prevent the problem of evil or any other objection from over¬
whelming the subject’s belief (provided, of course, that she did not
take steps to hide from herself her experience and this apparent
confirmation).9
Now even if the state of affairs I have sketched is a possible one,
it may be that its description does not entail the proposition in
question, namely (when completely spelled out), “God has
brought it about that for any subject S and time t, if S is at t
capable of relating personally to God, S at t believes that G on the
basis of evidence that renders G probable, except insofar as 5 is
culpably in a contrary position at t.” In particular, it may be that
even if S in fact believes because of her evidence, including the
evidence of religious experience, that evidence does not render G
probable.
9. Perhaps it will be claimed that naturalistic explanations of the universality

and uniformity of the experience in question will cause S to doubt, at least occa¬
sionally. But given that religious experience can be forceful and that S would, as
mentioned below, have available to her an argument for design as an attractive
alternative to naturalistic explanation, this seems unlikely.
It seems to me, however, that the entailment does hold—at least

if we allow for a little filling out of the description. What the expe-
rient comes to believe in the circumstances I have described is, we
may assume, at least in the first instance basic for her. That is, the
belief is supported by experience and not held for the reason that it
is rendered probable by other propositions the subject believes.
And utilizing terminology recently popularized by Alvin Plantinga
and others, I think we can go on to say that it is properly basic, that
is, “such that it is rational to accept it without accepting it on the
basis of any other propositions or beliefs at all.”10 The subject’s
experience, in the first instance at least, is the only relevant evi¬
dence she has. Given that the experience is intense, it is perfectly
proper for her to believe that God is present to her (and ipso facto,
that God exists) and to believe this in a basic way. What follows
after is, of course, another matter, but anyone looking back on
such an experience would, I think rightly, wish to say that in the
circumstances, the proposition believed by the subject was indeed
properly basic for her. Hence it seems right to say that in the first
instance at least, the relevant evidence available to 5 is such as to
render G probable: given the nature of belief, to say that it is ratio¬
nal for S to believe on her evidence commits one to saying that it
would be rational for her, upon reflection, to view her evidence as
probabilifying; and there is no distinction between this claim and
the admission that her evidence is probabilifying. It also seems
right to say that unless she deceives herself, S believes that G. The
rejection of belief at this stage would clearly require resistance, and
because of the absence of counterevidence, such resistance would
clearly be culpable. As Swinburne puts it, “We find ourselves with
involuntary inclinations to belief; in the absence of reasons against
going along with such an inclination the rational man will do so.”11
Having established that the subject’s initial belief is adequately
supported, we must of course go on, as I have noted, to ask about
what comes after. Does the proposition in question, by remaining
10. Alvin Plantinga, “Reason and Belief in God,” in Alvin Plantinga and
Nicholas Wolterstorff, eds., Faith and Rationality (Notre Dame: University of No¬
tre Dame Press, 1983), p. 72.
11. Richard Swinburne, “Does Theism Need a Theodicy?” Canadian Journal of
Philosophy, 18 (1988), 292.
properly basic or in some other way, retain its strong support

when the subject comes to consider other evidence?
It seems to me that it may. Let us look first at the positive con¬
siderations. Of the relevant facts that are likely to come to S’s at¬
tention subsequently, two are of particular interest. The first is the
universality of the experience: others have it too, and indeed vir¬
tually every individual S believes to be honest can testify to it. The
other (as might be anticipated) is the uniformity in the detail of its
descriptions. These facts serve to reinforce S’s belief not only psy¬
chologically but epistemically as well. Now it may seem that if this
happens, S must cease to believe in a basic way. But even if S takes
these considerations into account, he need not believe on the basis
of them. A more natural understanding is that S’s beliefs about the
universality and uniformity of this kind of religious experience
cause him to trust his own experience more strongly, that is, to
believe more strongly—but still in a basic way—that God is pre¬
sent to him.
Suppose, however, that S does come to believe at least in part
on the basis of other of his beliefs. Perhaps the considerations men¬
tioned cause him to recognize the increased support for the argu¬
ment for design. Then we may wish to accept a less stringent defi¬
nition of basicality, such as Swinburne’s: “I shall call those
propositions which seem to a man to be true and which he is in¬
clined to believe, but not solely on the ground that they are made
probable by other propositions which he believes, his basic propo¬
sitions” (my emphasis).12 On this definition, even if other of the
propositions S believes form part of his grounds for believing that
God exists, his belief may also be grounded in what he takes to be
his experience of the world, and so it can appropriately be said that
G is basic for him. In any event, not much turns on whether we
say 5 believes basically or not, so long as we can say that he does
continue to believe, and that he does so with the strong legitimacy
referred to earlier. And it seems clear that given the circumstances
described, we can say this. Because of his experience and the expe¬
riences of others, S continues to believe that there is a God and (I
am arguing) is right to do so: the facts of universality and uniform¬
ity mentioned earlier provide confirmation for his judgments.
12. Swinburne, Faith and Reason, p. 20.

Is a Strong Epistemic Situation Possible? [55
This claim would appear to be strongly supported by several

recent philosophical discussions of religious experience. Swin¬
burne, for example, argues that in this context a “Principle of Cre¬
dulity” applies: “(In the absence of special considerations) if it
seems (epistemically) to a subject that x is present, then probably x
is present.”13 His examination of suggested “special considerations”
leads him to the conclusion that “a religious experience apparently
of God ought to be taken as veridical unless it can be shown on
other grounds significantly more probable than not that God does
not exist.”14 But he adds that even an apparent perception of what
is “on background evidence too improbable to believe . . . may
become credible if backed up by positive evidence that the experi¬
ence is genuine,” such as “others’ having corroborating experi-
”15
ences.
Garry Gutting, in his discussion of religious experience, also
mentions the value of corroborating experiences. Indeed, on his
view, most of the weight must be placed here. He criticizes Swin¬
burne’s principle of credulity which, as he puts it, suggests “that
the evidence of the experience is itself decisive unless there is some
overriding consideration in our background knowledge,” and ar¬
gues that “an of-X experience in general provides prima facie evi¬
dence of X’s existence only in the sense of supplying some (but not
sufficient) support for the claim that X exists. For belief in this
claim to be warranted the solitary of-X experience requires supple¬
mentation by additional corroborating experiences.”16
The point about corroboration is supported from a different an¬
gle by William Alston, who has given a good deal of attention to
the (actual) problem of religious diversity. Incompatible religious
experiential claims, he allows, significantly reduce the force of the
evidence provided by religious experience for religious believers in
the actual world. More to the point, were we not faced with “such
persistent incompatibilities,” persons forming beliefs on the basis
of religious experience “would feel much more confident” and
“would be justified in so feeling.” It can hardly be denied, accord¬
ing to Alston, that such incompatibilities as actually exist reduce
13. Swinburne, Existence of God, p. 254.

14. Ibid., p. 270.
15. Ibid., p. 271.
16. Gutting, Belief and Skepticism, pp. 148, 149.
the rationality of believing on the basis of religious experience “be¬

low what it would be if this problem did not exist.”17 The applica¬
tion to our argument should be obvious.
Now it will perhaps be argued, in response, that we have
reached this positive conclusion only by ignoring the possibility of
contrary evidence. S, we may expect, will also come upon various
objections to theistic belief in the course of her life. Perhaps most
troubling will be the problem of evil, and her own experience of
evil and knowledge of the evil experienced by others will only
make this problem seem more severe. Would not this work against
her experience of God? And if S for a short or longer time failed to
believe in the existence of a loving God because of her experience
of evil, would her failure to believe not be inculpable? Although it
may be tempting initially to do so, I suggest that we have no rea¬
son to accept this view. If the experience apparently of God were
strong enough and its corroboration by others’ experiences evi¬
dent—if 5 were to feel strongly that God was present to her (and
especially if it were to seem strongly to S that God was present to
her in comforting and sustaining ways), and if other individuals
known to S testified to very similar experiences—then if S were to
lose even weak belief in the face of evil, we would not correctly
attribute this to the force of contrary evidence having its proper
effect: we would, I think, have to conclude that S had taken a hand
17. Alston, Perceiving God, p. 275. The arguments of Swinburne, Gutting, and
Alston seem to involve the claim that the justification provided for religious beliefs
(in the actual world) by religious experience is analogous to the justification pro¬
vided for sensory beliefs by sensory experience. This analogy argument is sub¬
jected to detailed criticism by Richard Gale, in On the Nature and Existence of God
(Cambridge: Cambridge University Press, 1991), chap. 8. But Alston has replied
to Gale by pointing out that the important analogies do hold: the belief-forming
practice in question is socially established; it has a “functioning overrider system”;
no reasons for judging the practice to be unreliable are sufficient; and the practice
enjoys “a significant degree of self-support” (Perceiving God, p. 224). Hence, Al¬
ston concludes, it is rational for persons to form beliefs on the basis of religious
experience in the absence of specific overriding considerations. It seems to me that
whatever may be true of the actual world, were individuals in the possible world I
am describing to avail themselves of Alston’s premises (as well as the many argu¬
ments he offers in their support), adding to them the important points I have been
emphasizing with respect to universality and uniformity—points unavailable to
Alston, which strengthen the analogy with sense experience—they surely would
be justified in drawing his conclusion.
in the process herself, engaging in self-deception out of bitterness,

resentment, or whatever. Although we might in such circum¬
stances feel a certain sympathy for 5 and view her actions as in a
sense understandable, if we were apprised of all the facts, we
would also view them as mistaken and wrong.
These points seem to me to show that in the circumstances in
question, the relevant evidence available to 5 is not only initially,
but at all times, such that, unless culpably deterred, 5 believes that
G on evidence rendering G probable. In summary, it seems clear
that the experiential evidence described here does provide the re¬
quired support in the absence of overriding considerations, and
that God could bring it about, for all S and for all t, that there are
no overriding considerations.18 We may therefore conclude that the
proposition at issue is entailed by the description I have given. But
that description is of a possible state of affairs. Hence the proposi¬
tion itself represents a possible state of affairs. In other words, a
strong epistemic situation in relation to theism is indeed possible.14
18. Anyone who thinks that this position is incorrect may add other evidence or
postulate further Divine action, as appropriate. (For an interesting discussion of
how religious experience may interact with other grounds of religious belief, see
Alston, Perceiving God, chap. 8.) But for myself, I do not think such additions to
the picture are necessary.
19. Note that I have not argued that God should bring to our attention every
piece of evidence normally considered relevant to the question of God’s existence.
This would entail God making philosophers of religion of us all, and then the
world would be in a sorry state! Rather, I have pictured a state of affairs in which
every human being is given at least the evidence of experience described, and have
argued that given any amount of subsequent inquiry, this would be sufficient, in
the absence of resistance, to sustain belief.
[ 3 ]
The Reasonableness
of Nonbelief
Although there are possible worlds (such as the one described in

the preceding chapter) in which a strong epistemic situation in rela¬
tion to theism obtains, the actual world is not one of these. In our
world reasonable nonbelief occurs. This, at any rate, is my claim in
the present chapter. I begin by spelling out what is meant by this
claim and giving several arguments in its defense. I then go on to
consider some of the relevant views of contemporary theologians
and philosophers of religion, as well as a possible objection sug¬
gested by this discussion.
The Claim: Explication and Defense
Let us take a nonbeliever to be one who fails (for whatever rea¬

son and in whatever way) to believe that there is a God. Even
allowing for some flexibility in the interpretation of “God,” it is
clear that many human beings in the actual world fit this descrip¬
tion. There are, first of all, individuals—primarily from non-West-
ern cultures—who have never so much as entertained the proposi¬
tion “God exists” (G), let alone considered the question of its truth
or falsity. Second, there are those, from both Western and non-
Western backgrounds, who are to some extent familiar with the
idea of God, but who have never considered with any degree of
seriousness whether it is instantiated. Individuals in these two cate-
The Reasonableness of Nonbelief [5
gories exhibit what we may call unrejlective nonbelief. They are to

be contrasted with reflective nonbelievers—individuals who disbe¬
lieve or are in doubt about G as a result of reflection on its content
and some attempt to discover whether it is true or false.1 Those
who disbelieve consider this proposition to be improbable or cer¬
tainly false and so believe that not-G.2 Individuals who are in
doubt, on the other hand (to whose state I will be returning pres¬
ently), are uncertain about the truth of this proposition, believing
neither G nor not-G, typically as a consequence of believing that
epistemic parity obtains between G and its denial.3
So much for the varieties of nonbelief.4 What about reasonable
nonbelief? In the Introduction, I suggested that reasonable non¬
belief is in this context to be understood as exemplified by any
instance of failure to believe in the existence of God that is not the
result of culpable actions or omissions on the part of the subject.
The claim that reasonable nonbelief occurs is therefore the claim
that the nonbelief of at any rate some nonbelievers is not the conse¬
quence of their culpable actions or omissions—that it arises
through no fault of their own and so they are not in any sense to
blame for it. In defending this claim, I will be looking at one par¬
ticular form of nonbelief—doubt—and attempting to show that it
is sometimes inculpable. This is not to suggest that other forms of
nonbelief are not inculpable. It seems clear enough that each type is
inculpably exemplified, especially the first. But it will be conve¬
nient, for our purposes, to narrow the discussion to some particu¬
lar form of nonbelief; and it will be interesting in its own right to
learn that even where there has been reflection on G, and on the
evidence in its favor, inculpable nonbelief may remain.
1. Given present concerns, the expression “is in doubt about G” is to be prefer¬

red to “doubts whether (or that) G.” The latter is most naturally construed as “is
inclined to disbelieve G,” and this is not a meaning I wish to convey. Doubt is also
sometimes understood in such a way as to be compatible with belief, and this too is
a view from which mine must be distinguished. On the understanding assumed
here, one who doubts neither believes nor disbelieves that G. To put it another
way: doubt is identified with the point midway between belief and disbelief.
2. For the sake of simplicity, I include in the “certainly false” subcategory
those who consider G to be incoherent or meaningless.
3. There will be more on “epistemic parity” later in the chapter.
4. No doubt more varieties of nonbelief could be distinguished, but these are
the main ones and will suffice for our purposes.
So let us look more closely at the notion of doubt. As I have

already indicated, it is here explicated in terms of uncertainty about
the truth of some proposition (typically) generated by the belief
that epistemic parity obtains between that proposition and its de¬
nial.5 Inculpable doubt, therefore, obtains if such a belief is inculpa-
bly held. That is to say, for all S, if S inculpably believes that
epistemic parity obtains between G and not-G, then S is inculpably
in doubt about G.6 This is, of course, only a start, but it does show
that our first topic of discussion must be the concept of inculpable
belief.
A careful and nuanced discussion of belief and epistemic ratio¬
nality which includes a discussion of culpability and inculpability in
believing can be found in Richard Swinburne’s Faith and Reason.
Swinburne distinguishes five kinds of rationality. The first kind
(which he calls rationality]) obtains if and only if S’s “belief that p
is probable, given his inductive standards and given his evidence. .
. . A failure in respect of rationality] is a failure of internal coher¬
ence in a subject’s system of beliefs, a failure of which the subject is
unaware.” But even if a subject’s beliefs are internally coherent,
says Swinburne, something may be lacking: the subject may not be
“responding to the world in a justifiable way.” He therefore intro¬
duces a second kind of rationality (rationality2) which obtains if
and only if S’s belief that p is properly grounded in experience or
reason and arrived at by the application of correct inductive stan¬
dards. “Rationality2 is a matter of conformity to objective stan¬
dards which the believer may not recognize and may indeed ex¬
plicitly deny.”7
5. “Typically” is inserted here to allow for the possibility of doubt occurring

without being caused by a parity belief (see note 6), not to suggest that a parity
belief need not generate doubt. In my view, it must do so. Anyone tempted by a
contrary view is referred to the discussion of belief in Chapter 1, second section.
6. As this formulation indicates, I am not assuming that S’s inculpably holding
a parity belief is a necessary condition of S’s being inculpably in doubt. To suppose
that it is would be, as William Alston has pointed out to me, unduly restrictive,
forcing us to say that individuals incapable of a parity belief (e.g., small children)
cannot be in doubt. That such a belief constitutes a sufficient condition for inculpa¬
ble doubt is, in any case, all that is required for our purposes.
7. Richard Swinburne, Faith and Reason (Oxford: Clarendon Press, 1981), pp.
45, 46. References to this work in Chapter 3 will henceforth be made par¬
enthetically in the text.
The Reasonableness of Nonbelief [ 61
The distinction between rationality in these first two senses and

rationality in senses (3), (4), and (5) is an important one. It is ex¬
pressed by Swinburne in the following passage:
The rationality of both rational] and rational beliefs is a matter of

the believer’s response to present sensations and memories of the
past and to apparently self-justifying truths of reason at the time in
question. . . . However, we often feel that although a man is justi¬
fied in holding a certain belief at some time, he ought to have
looked for more evidence or checked his standards more thoroughly
at earlier times. Had he done so, he might have beliefs which were
better justified, more probable. And so, according to whether the
failure at an earlier stage was a failure by the subject’s own standards
of which he was aware, a failure by the subject’s own standards of
which he was not aware, or a failure by correct standards, we have
three further kinds of irrationality of belief. In so far as these possi¬
ble failures have been avoided, we have three further kinds of ratio¬
nality. [P. 49]
Filling this out, Swinburne writes that S has a rational belief

that p if and only if his evidence, inductive standards, and belief as
to p's probability on the evidence have been, in his own view at the
time, adequately investigated; a rational4 belief if and only if his
investigation was adequate by his normal standards; and a rational5
belief if and only if his investigation satisfied correct standards (pp.
49-54). S’s judgment as to when investigation is adequate, says
Swinburne, will depend “on four beliefs of his: (a) about the im¬
portance of the issue, (b) about the closeness to o or 1 of the proba¬
bility of his belief about the issue, (c) about the probability that
investigation will achieve something, and (d) about whether he has
other more important actions to do” (p. 53). And how much in¬
vestigation is in fact adequate will depend on what is in fact true
with respect to (a)-(d) (p. 53).
According to Swinburne, belief is culpable if it is irrational3. He
describes several ways in which such irrationality may arise: “First,
there may be a culpable failure, of which the subject is aware, to
collect enough true, representative, relevant evidence of good qual¬
ity” (p. 50). S may through culpable negligence fail to gather enough
relevant evidence or to check the reliability of the evidential propo-
sitions he is inclined to believe on the basis of experience or reason.

Sometimes this may involve self-deception: S may look only for
evidence that supports his own view and then deliberately forget
the bias of his evidence sample. S’s belief would also be irrational3
and hence culpable, says Swinburne, if S recognized that his induc¬
tive standards had not been “subjected to proper criticism” (p. 50).
And “the third reason why a belief can fail to be rational3 is that
the subject has culpably failed to check that, given his standards,
the belief is made probable by the evidence” (p. 51). So, on Swin¬
burne’s view, irrationality3, and hence culpability in believing, is a
matter of negligence of which one is aware in evidence acquisition,
the use of inductive standards, or in judging the probability of
one’s belief on the evidence one accepts. And this irrationality is
compounded, he suggests, if negligence at any of these levels is of
a sort that is deliberately self-deceptive—designed to produce in
one a belief about one’s evidence, one’s standards, or the relevant
probabilities that is recognized (at least initially) to be unwarranted.
So far, so good, we might say. But is it only if it is irrational3
that belief is culpable? Swinburne thinks so:
It is only irrationality in sense (3) which is culpable irrationality, for

it results from the subject neglecting investigative procedures which
he recognizes that he ought to pursue. Irrationality in senses (4) and
(5) are a matter of objective discrepancy between the subject’s actual
investigative procedures and either those which he normally recog¬
nizes or really adequate investigative procedures; but in so far as the
subject does not recognize these discrepancies, no blame attaches to
his conduct. Irrationality in senses (1) and (2) arises from a failure to
recognize certain things at the time in question—discrepancies
within the class of the subject’s beliefs in the case of irrationality (1),
and unjustified evidence and incorrect standards in the case of irra¬
tionality (2). But either you recognize the things in question at the
time or you do not. Recognizing is coming to believe; and if, as I
have argued, belief is a passive matter, so too is recognition. No
blame is attachable to you for things that happen to you, only for
things that you do. fP. 54]
The principle suggested here is the following: “S is culpable in

respect of some failure of hers if and only if her failure is volun¬
tary.” If Swinburne does accept this principle, the argument of the
first part of the passage (which concerns irrationality in senses (3),

(4), and (5)) can be filled out as follows: “Only voluntary neglect
of one’s epistemic obligations is culpable; unrecognized neglect is
involuntary; therefore unrecognized neglect is inculpable.” When it
comes to irrationality in senses (1) and (2), where the failures in
question are not failures of investigation which may be recognized
or unrecognized, but failures of recognition (i.e., failures to recog¬
nize certain things at the time of belief formation), Swinburne
again argues that since these failures are involuntary, no blame at¬
taches to the subject in respect of them. And of course, since the
very state of belief is involuntarily acquired, it would be inap¬
propriate to blame the believer for this as well.
In my view, Swinburne is right on these points. It seems correct
to say that only for voluntary actions could we ever legitimately be
blamed. If S is to blame for something, S must have made some
intentional contribution to it. Hence no one can justifiably be
blamed for involuntary epistemic failures or for the very fact of
belief. This argument has, however, been criticized in recent philo¬
sophical writing. In his article “Involuntary Sins,” Robert Adams
maintains that we are sometimes to blame for holding some belief
even though believing is not under our (direct) control.8 Adams’s
argument (as well as a similar argument of Plantinga) seems to be
motivated by our tendency to say that individuals who hold mor¬
ally repugnant beliefs (e.g., the belief that all Jews should be exter¬
minated) are to be blamed for holding such beliefs even if they
arose involuntarily.9 10 But as Swinburne has convincingly argued, in
reproaching people for such beliefs, we are either voicing our con¬
viction that it is an objectively bad thing that such beliefs should be
held (which is compatible with the believers themselves being
blameless) or laying blame “for past omissions to act which al¬
lowed such attitudes to develop.”'1' And as Plantinga admits, when
8. Robert Adams, “Involuntary Sins,” The Philosophical Review, 94 (1985), 3-

3i.
9. For Plantinga’s argument, see his “Reason and Belief in God,” in Alvin
Plantinga and Nicholas Wolterstorff, eds., Faith and Rationality (Notre Dame: Uni¬
versity of Notre Dame Press, 1983), pp. 34-37. Plantinga is ambivalent, however.
See esp. p. 36. (See also immediately below.)
10. Richard Swinburne, Responsibility and Atonement (Oxford: Clarendon Press,
1989), p. 34, n. 1.
blame is laid for morally repugnant beliefs per se, it may derive
from neglect of the possibility that the individuals in question are
constitutionally defective in some way and so not really subject to
blame.11
For these reasons, I suggest we need not accept Adams’s exam¬
ples as indicative of the truth of his position.12 And therefore the
claim that we are only culpable for voluntary epistemic failures lead¬
ing up to belief—in particular, for voluntarily neglecting proper in¬
vestigative procedures—can be upheld. Applying this to the ques¬
tion of inculpable religious doubt, we obtain the following:
5 is inculpably in doubt about the truth of G if (i) 5

believes that epistemic parity obtains between G and
not-G, and (2) 5 has not knowingly (self-deceptively or
non-self-deceptively) neglected to submit this belief to ad¬
equate investigation.
Having decided upon a criterion of inculpable doubt, we must

now consider the question of its application: How can we tell in
individual cases that its conditions are satisfied?
It is not difficult to see how we might have reason to say of
some individual that she believes that G and not-G are at epistemic
parity, for she may tell us that she does and we may have no rea¬
son to doubt her word.13 But it may be useful to look a little more
11. Plantinga, “Reason and Belief,” p. 36.

12. Suppose, however, that Adams is right. Then, even assuming S is not to
blame for any voluntary epistemic failure, if the belief that epistemic parity obtains
between G and not-G is morally repugnant (and perhaps also if S’s belief, as a
result, say, of laziness, fails to be rational in Swinburne’s senses (1), (2), (4), or
(5)), we may impute blame to S and claim that his parity belief is culpable. Should
this disturb us? It seems not. The belief that epistemic parity obtains between G
and not-G is, to say the least, not obviously morally repugnant, and as we will see
later, there are individuals who hold this belief but have investigated it with great
care—individuals to whom the expansion of Adams’s point in terms of laziness,
even if it is a legitimate expansion, is therefore manifestly inapplicable. Thus, even
if Adams is right (and I do not think he is), his arguments are in this context
irrelevant and so may safely be disregarded.
13. I am of course appealing here to the principle of testimony, which, as Swin¬
burne puts it, states “that (in the absence of special considerations) the experiences
of others are (probably) as they report them” (The Existence of God [Oxford:
Clarendon Press, 1979], p. 272).
closely at what is involved in holding this belief. As I understand

it, one who believes that G and not-G are at epistemic parity be¬
lieves that, given her evidence, she is not justified in holding either
proposition to be more probable than its denial—that neither is,
for her, epistemically preferable to its denial. Given this understand¬
ing, it may seem that what the parity believer believes is that G
and not-G are equally probable, but this is not necessarily so: I
may hold beliefs about G and not-G that entail the belief that nei¬
ther is epistemically preferable to the other without believing that
they are equally probable. I may, for example, believe that given
my evidence, the correct values for the relevant probabilities
(whether precise numerical or comparative) cannot be determined,H
This belief entails the belief that neither G nor not-G is epis¬
temically preferable to its denial, but it clearly does not entail the
belief that they are equally probable. Therefore, in deciding
whether some individual holds the parity belief, we must not only
look for evidence that she believes G and not-G to be equally proba¬
ble.
The second part of my criterion may appear to pose more diffi¬
cult problems. How could we ever have reason to say, upon eval¬
uating S’s parity belief, that 5 has not knowingly neglected ade¬
quate investigation? How could this information be available to us?
It seems to me that in some cases there may not be enough rele¬
vant information available. But it seems equally clear that in certain
circumstances a judgment in favor of the subject would be appro¬
priate. Again, what 5 tells us of her investigation is not to be taken
lightly. But we may also be witness to her investigation and see it
to be an exemplary one. S may have given as much or even more
time and energy to investigation than our beliefs about the issue’s
importance, probabilities in the field, the probability that investiga¬
tion will achieve something, and S’s other responsibilities suggest
is adequate. If so (and provided that S’s own beliefs on these mat¬
ters of which we have knowledge do not suggest that more inves¬
tigation is required), we will rightly judge that S has not know¬
ingly failed to pursue adequate investigation.
It may seem to some that the question whether self-deception
14. As we will see, this is in fact the more common assessment.

has occurred will be especially difficult to answer in individual

cases. But here, too, under certain conditions (which may well be
present), we would have to rule in the subject’s favor. S’s conduct
in other contexts, especially other epistemic contexts, is partic¬
ularly important. Has he shown himself to be honest, a lover of the
truth? Does he resist his wants when his head tells him he ought
not to give in to them? We may also have reason to believe that S
desires to have a well-justified belief that G or that not-G. If this is
clearly so in some particular case, then (unless there is very strong
evidence to the contrary) we may surely conclude that 5 is not self-
deceived in arriving at a parity belief. For, given such a wish, 5 is
much more likely to find ways of avoiding a parity belief than to
find ways of acquiring one.
This suggests a more general point as well. If S desires a well-
justified belief that G, or that not-G, he will arrive at a parity belief
only reluctantly and, therefore, only if careful attention to the mat¬
ter seems to him to leave him with no other option. Thus, if we
have reason to believe that 5 wishes to have a well-justified belief
that G, or that not-G, we have reason to believe not only that 5 is
not self-deceived, but, more generally, that his investigation was a
thorough one.
A final point about assessing parity beliefs as culpable or inculpa¬
ble is that, where beliefs of this sort are concerned, we may also
ask whether the propositions in questions are controversial—
whether expert opinion is divided over which is true. If this is the
case, it is likely that there is something to be said for each side’s
position and so likely that more cautious investigators will see this.
In such circumstances, instances of honest doubt are to be ex¬
pected. Now this is not to say that if G and not-G are controver¬
sial, every parity belief is automatically inculpable. However, such
controversy, where it exists, provides us with useful additional in¬
formation, allowing our judgment, when we are inclined to con¬
clude on other grounds that S’s parity belief is inculpable, to be
more confident. If such information is not available, or if the con¬
trary is clearly true, it is reasonable to be more tentative and cau¬
tious, and to ask for more of the sort of evidence described above
before making a judgment. For then we have inductive grounds
for supposing that a thorough investigation would lead S to the

confident endorsement of either G or not-G.
It is time now to apply these considerations to the question
whether inculpable doubt about the truth of G actually occurs. It is
clearly true that many individuals satisfy the first of the conditions
specified by our criterion, for, as a glance at the philosophical liter¬
ature will show, many do believe that G and not-G are at epis-
temic parity—that neither of these propositions is epistemically
preferable to its denial. (Of course, many nonphilosophers hold
this belief as well.) The term commonly associated with this view
is “agnosticism.” The agnostic claims that it is impossible to judge
on rational grounds that there is or is not a God. (She may have in
mind the present state of the evidence, allowing that the epistemic
status of G and not-G may change, or she may hold that a judg¬
ment is impossible in principle.15) Now clearly, if an individual be¬
lieves that the evidence does not allow a judgment as to whether
there is a God, she believes that neither G nor not-G is epis¬
temically preferable to its denial—that is, she holds a parity belief.
That there are persons who believe the former (i.e., agnostics) is a
truism.16 We may therefore infer with the highest degree of confi¬
dence that there are individuals who hold a parity belief.
The second condition specified by our criterion seems also to be
satisfied in many cases. Many who doubt have investigated the
question of God’s existence with great care and concern for the
truth over a period of years. To say of them that they have not
knowingly failed to pursue adequate investigation is to say too lit¬
tle: if their doubt is inculpable at all, it is strongly inculpable, that is,
their investigations are exemplary, even supererogatory, and
match in quality those of the most scrupulous of their opponents.
Now it may be thought that I have neglected the possibility of
self-deception. But although there are no doubt some cases in
which we have reason to suppose that it has occurred, in many
15. See Anthony Kenny, Faith and Reason (New York: Columbia University
Press, 1983), pp. 87-88.
16. Prominent contemporary examples include Anthony Kenny and Ronald
Hepburn. See ibid., p. 85, and Ronald Hepburn, Christianity and Paradox (London:
Watts, 1966), p. 1.
others we either do not have such reason or have good reason to

suppose that it has not occurred. There are, in particular, individ¬
uals of whom we would have to say that if they have any desire at
all with respect to this issue, it is to have a well-justified belief one
way or the other. Their longing corresponds to that of Pascal:
I look around in every direction and all I see is darkness. Nature has
nothing to offer me that does not give rise to doubt and anxiety. If I
saw no sign there of a Divinity I should decide on a negative solu¬
tion: if I saw signs of a creator everywhere I should peacefully settle
down in the faith. But seeing too much to deny and not enough to
affirm, I am in a pitiful state, where I have wished a hundred times
over that, if there is a God supporting nature, she should unequivo¬
cally proclaim him, and that, if the signs in nature are deceptive,
they should say all or nothing so that I could see what course I
ought to follow. Instead of that, in the state in which I am, not
knowing what I am or what I ought to do, I know neither my
condition nor my duty. My whole heart strains to know what the
true good is in order to pursue it: no price would be too high to pay
for eternity.17
In individuals such as the one represented here—who certainly ex¬

ist—self-deception, if it occurred, would, it seems, be much more
likely to produce belief than doubt.
As I suggested earlier, the fact that some individuals who doubt
desire to believe also gives us an independent reason for saying that
doubt is sometimes inculpable. For such persons, the parity view is
only to be arrived at after all alternatives have been exhausted. We
can infer from the fact that they strongly wish to settle the ques¬
tion for themselves one way or the other but nonetheless hold a
parity belief that their investigation was thorough, and that their
parity belief—although it may be mistaken—is not the result of
negligence.
My final point in support of the view that inculpable doubt oc¬
curs is perhaps more obvious than any other I have made, namely,
that the question of God’s existence is controversial—a question
over which expert opinion is divided. The individual who begins
17. Blaise Pascal, Pensees, A. J. Krailsheimer, trans. (Harmondsworth: Penguin,

1966), fragment 429.
an inquiry into God’s existence is faced with a plethora of argu¬

ments both for and against. There is no easy way to sort through
these arguments, and it is not obvious a priori that one side’s argu¬
ments are deficient. From this it would seem to follow that we
should expect what we in fact seem to find, namely, that scru¬
pulous doubt occurs. Of course, as I suggested above, this infor¬
mation should not on its own lead us to conclude that such doubt
occurs. But it seems to me that when taken in conjunction with
other points I have adduced (which suggest that there are individ¬
uals holding parity beliefs who pass various tests of inculpability),
it ought indeed to have this effect.
Some Contemporary Theological and

Philosophical Views
I have argued that in the actual world inculpable doubt occurs.

Many contemporary theologians appear to agree. Eberhart Jiingel,
for example, writes that “the fundamental theoretical and practical
changes of our so-called modern age have defined the contempo¬
rary awareness of the problem [of God] in a way which is espe¬
cially unsettling for theology. One simply cannot ignore the fact
that the dubiousness of talk about God has intensified.”18
Perhaps the majority of theologians would go farther still and
allow that inculpable disbelief occurs. Take, for example, the view
of theologians at Vatican II, as reported by Karl Rahner:
The council makes no reference to the traditional textbook view that

positive atheism cannot be entertained for any considerable period
of time by a fully developed person of normal intelligence without
involving blame on his part. The Council actually assumed a contrary
thesis, i.e. that it is possible for a normal adult to hold an explicit atheism
for a long period of time—even to his life’s end—without this implying
moral blame on the part of such an unbeliever. [Emphasis in the orig¬
inal]19
18. Eberhart Jiingel, God as the Mystery of the World, Darrel L. Coden, trans.
(Edinburgh: T&T Clark, 1983), p. 4.
19. Karl Rahner, “Atheism and Implicit Christianity,” in G. A. McCool, ed., A
Rahner Reader (London: Darton, Longman, and Todd, 1975), pp. 221-222. Rahner
himself, as the context indicates, was wholly in favor of this view.
Many theologians seem themselves to be less than sure about the

epistemic preferability of theism. There is a tendency to describe
the world as “religiously ambiguous.”2"John Macquarrie, referring
to the dispute between theism and atheism, writes that “the world
always remains ambiguous, and although people can argue as long
as they like, tracing their rival patterns, the case will never be es¬
tablished conclusively, one way or the other.”21 Parallel claims are
made by theologians influenced by the conceptual relativism of
such writers as Thomas Kuhn.22 James McClendon, writing about
Christian and other world views, states that “there is not among
these any which can be clearly established to be the superior of all
the others.”23 And George Lindbeck, in a similar vein, argues that
“there is no higher neutral standpoint from which to adjudicate
their competing perceptions of what is factual.”24
Lest anyone suspect that theologians are alone among Christian
writers in this regard, it should be noted that many contemporary
philosophers of religion—even quite conservative Christian philos¬
ophers of religion—appear to be making similar claims. Perhaps in
most cases these claims do not presuppose Kuhnian assumptions,
but the difficulties involved in establishing theism as epistemically
preferable to atheism are clearly recognized. Thomas Morris, for
example, has argued that “epistemically null or epistemically am¬
biguous conditions can reasonably be thought sometimes to ob¬
tain. ”25 And elsewhere he appears prepared to acquiesce in the view
that “this world of ours is at best religiously ambiguous to the
20. This phrase, as far as I know, originates in John Hick’s writings. For more
on Hick and religious ambiguity, see Chapter 5.
21. John Macquarrie, Thinking about God (London: SCM Press, 1975), p. 119.
22. See Thomas Kuhn, The Structure of Scientific Revolutions (Chicago: Chicago
University Press, 1962).
23. James W. McClendon and James M. Smith, Understanding Religious Convic¬
tions (Notre Dame: University of Notre Dame Press, 1975), p. 117.
24. George Lindbeck, The Nature of Doctrine (Philadelphia: Westminster Press,
1984), p. 11.
25. See Thomas Morris, Anselmian Explorations (Notre Dame: University of
Notre Dame Press, 1987), p. 203.
The Reasonableness of Nonbelief
inquiring observer,” suggesting that this view may pose an impor¬

tant theological problem.26
Many Christian thinkers, therefore, not only allow that those
who hold a parity belief sometimes do so reasonably, but appear to
hold this belief themselves. As I have argued, one who investigates
the evidence can believe that G only if he believes that the evidence
favors G—that G is epistemically preferable to not-G. And surely
all the individuals mentioned would claim to have investigated the
evidence and to believe that there is a God. So what should we
make of this?
As it seems to me, what we must do to solve this problem is
sharpen up an important distinction, already alluded to, namely,
the distinction between public and private evidence. When writers
of the sort I have referred to speak of ambiguity or parity (or put
forward arguments that entail claims about ambiguity or parity),
they are certainly referring to the public evidence—evidence which
is in principle available to everyone equally, (typically) reported in
the premises of the various theistic and atheistic arguments. But I
suspect that in many cases their own personal experiences, which
are at least in part private to themselves, are not included in this
judgment. Many theologians, for example, would argue that rea¬
son for believing is in some sense provided by such experiences—
that religious experience is in an important sense the ground of
faith.27 Now although discussion of these matters among theo¬
logians is not always as clear as one might like, it is natural, I
suggest, to take individuals who make such claims to be (implic¬
itly) judging, at least in part on the basis of their own experience,
that God’s existence is probable, and thus to be treating their experi¬
ence as supplementary evidence. Although they consider epistemic
parity to obtain at the public level, the private evidence available to
them (I am suggesting), perhaps in conjunction with the public
evidence, seems to them to render G more probable than not-G,
26. See the review by Thomas Morris of Anthony O’Hear’s Experience, Expla¬
nation, and Faith, in Faith and Philosophy, 2 (1985), 316.
27. See, for example, Langdon Gilkey, Naming the Whirlwind: The Renewal of
God-Language (Indianapolis and New York: Bobbs-Merrill, 1969), part 2, chaps. 3
and 4.
] Framing the Argument
and thus, all things considered, they do not consider epistemic parity
to obtain.28
That this is also the correct interpretation where the theistic be¬
liefs of many Christian philosophers of religion are concerned is
strongly suggested by their writings. Stephen Davis, who accepts
the ambiguity view, refers explicitly to the distinction between
public and private evidence and argues in favor of the rationality of
believing on the basis of the latter when the former is inconclusive.29
Some of the more recent arguments of John Hick, Alvin Plantinga,
William Alston, and others on the relations between religious ex¬
perience and the justification of religious belief also lend themselves
to this interpretation.30 The claim of these writers is that experi¬
ences apparently of God, although they may count for little when
publically reported (because they only provide a basis for weak
inductive arguments), may provide the experient with good
grounds for belief in God.31 It is argued that such experiences must
be understood by analogy with sensory experiences and the justi¬
fication of the corresponding beliefs by analogy with the justifica¬
tion of sensory beliefs. I would suggest that what this shows is that
the philosophers in question may, however implicitly, be treating
their religious experiences as relevant evidence, and may therefore
(all things considered) be believing that there is a God while con¬
tinuing to believe that the public evidence (including such proposi¬
tions as “religious experiences have been reported in various parts
of the world”) is indecisive. As George Mavrodes writes, after giv¬
ing voice to the skeptical feelings many have about accepting the
28. Indeed, we can deduce the claim that the individuals in question consider
themselves to have private evidence from the following statements, all of which
are true: (1) that anyone who investigates the support for G can only believe that G
if she believes G to be more probable than not; (2) that the individuals in question
believe that G; (3) that they believe the public evidence to support only a parity
judgment.
29. Stephen Davis, Faith, Skepticism, and Evidence (Lewisburg, Pa.: Bucknell
University Press, 1978).
30. See John Hick, An Interpretation of Religion (London: Macmillan, 1989),
chap. 13; Plantinga, “Reason and Belief’; and William Alston, Perceiving God: The
Epistemology of Religious Experience (Ithaca, N.Y: Cornell University Press, 1991).
31. It should be noted that Swinburne, who accepts a similar view, is not as
pessimistic as many others about the value of religious experience as public evi¬
dence.
public evidence provided by the religious experiences of others as

indicative of God’s existence, “I think that there is something that
we could have, and that we might have, something that is better
than that testimony, more solid somehow, richer in epistemic sig¬
nificance than any mere testimony can be. . . . That richer thing . . .
would be an experience of our own” (emphasis in the original).32
So where does this leave us? It seems clear that both theologians
and Christian philosophers of religion in many cases view the pub¬
lic evidence as indecisive and so are ready to go at least this far
with those who are in doubt. Nevertheless, as we have seen, they
seem to fall back on private evidence which they do not consider to
be indecisive, and so cannot be classified with the doubters. The
only issues between these believing theologians and philosophers
and individuals who are in doubt must therefore be the issues of
the availability and interpretation of private experiences apparently
of God. Now as we have seen, many writers (and especially theo¬
logians) seem prepared to concede the reasonableness of doubt and
so appear to accept as reasonable the claims of those who say that
they have no experiences of this sort or that such experiences as
they do have are ambiguous.33 Nonetheless, some writers—in par¬
ticular, certain latter-day Calvinian philosophers of religion—seem
prepared to argue that a form of private evidence has been made
32. George Mavrodes, Revelation in Religious Belief (Philadelphia: Temple Uni¬

versity Press, 1988), pp. 152-153. Richard Gale has recently argued against this
sort of claim (and so implicitly against my distinction between public and private
evidence—at least as justification-relevant) in On the Nature and Existence of God
(Cambridge: Cambridge University Press, 1991), pp. 287-288. According to Gale,
it is a fundamental principle of epistemology that “a cognitive [i.e., evidence- or
warrant-providing] experience’s status as evidential is observer-neutral.” But it is
surely a necessary condition of warrant in this context that one be justified in
believing that the experience occurred; and this at least is greater for the experient. In
some cases it might, indeed, be only the experient who was justified in believing
that the experience occurred. This would be so if observers had reason to believe
the experient to be a deceiver when in fact she was not, and knew that she was not.
In such a case, the experient would, other things being equal, be justified in believ¬
ing on the basis of her experience even though others were not so justified. Gale’s
general principle is therefore false and so poses no threat to Mavrodes’s claim or to
my distinction.
33. On the possibility of such claims, see Ronald Hepburn, “From World to
God,” in Basil Mitchell, ed., The Philosophy of Religion (Oxford: Oxford Univer¬
sity Press, 1971), p. 177.
available to everyone, and that where it has not brought about belief
in God’s existence, this is because of the sinful resistance of the
nonbeliever.34 We have, therefore, the following objection to the
view I am defending: doubt is never inculpable because it is always
a result of sin. It is to a discussion of this objection that I now turn.
The Calvinian Response
In his Institutes of the Christian Religion, John Calvin argues that

none of us lacks an awareness of God:
“There is within the human mind, and indeed, by natural instinct,

an awareness of divinity.” This we take to be beyond controversy.
To prevent anyone from taking refuge in the pretense of ignorance,
God himself has implanted in all men a certain understanding of the
divine majesty. Ever renewing its memory, he repeatedly sheds
fresh drops. Since, therefore, men one and all perceive that there is a
God and that he is their Maker, they are condemned by their own
testimony because they have failed to honour him and to consecrate
their lives to his will.35
Calvin’s view here is apparently shared by Alvin Plantinga. In re¬

sponse to the “Great Pumpkin” objection to his view that belief in
God is properly basic (the objection that if belief in God is properly
basic, so are many other beliefs we would wish to reject as irra¬
tional, such as the belief that the Great Pumpkin returns every Hal¬
loween), Plantinga writes that “the Reformed epistemologist may
concur with Calvin in holding that God has implanted in us a natu¬
ral tendency to see his hand in the world around us; the same can¬
not be said of the Great Pumpkin, there being no Great Pumpkin
and no natural tendency to accept beliefs about the Great Pump¬
kin.”36
Now for Calvin, an obvious corollary of the claim that God has
implanted in us “a certain understanding” of divine things is that
34. This view, it will be noted, entails that a situation much like that described
in Chapter 2 in fact obtains.
35- John Calvin, Institutes of the Christian Religion, Lewis Battles Ford, trans.
(Philadelphia: Westminster Press, i960), book 1, chap. 3, p. 43.
36. Plantinga, “Reason and Belief,” p. 78.
those who claim to be ignorant of God are “condemned”; they

have rejected God, knowing full well that he exists, and so are
blameworthy. We might expect something similar to be a corol¬
lary of Plantinga’s “natural tendency” doctrine. But he is not very
clear about this. It may be that he holds, with Calvin, that those
who do not believe in God “instinctively” are inhibited by their
own sin, having, for example, deceived themselves. On the other
hand, he may hold that those who fail to believe are inhibited by
sin indirectly in that they have inherited and now express involun¬
tarily dispositions and values inimical to belief. The latter view im¬
plies that those who fail to believe are not themselves culpable for
it, and thus gives rise to no objection to my claim. Because of this
unclarity, I suggest we look no further at Plantinga but move in¬
stead to consider the claims endorsed by certain of his Calvinian
counterparts, who quite clearly do wish to suggest that nonbelief is
due to the sin of the nonbeliever.
One such philosopher is Mark R. Talbot. In his recent article “Is
It Natural to Believe in God?” Talbot argues that Christians are
justified in asserting the contrary-to-fact conditional “Everybody
would believe in God, if it weren’t for sin,” and that “even a non-
Christian’s experience makes that claim more probable for him
than it would otherwise be.” Talbot spends some time developing
the notion of an “epistemic set,” which he defines as “a disposition
to have particular experiences, thoughts, beliefs, and so on, in par¬
ticular situations.” Using this to explicate Calvin’s view, he writes:
“To sense God’s glory in his handiwork and to feel his majesty
within us is, by this account, a matter of possessing reliable epis¬
temic sets. Since he thinks we naturally possess these sets, anyone’s
not possessing them signals, according to Calvin, his having
worked to dismantle or lose them.” Talbot clearly endorses Cal¬
vin’s view, so I think that by “if it weren’t for sin,” we can take
him to mean “if it weren’t for the personal sin of the nonbeliever.”
This is confirmed later in his essay when he states that “unbe¬
lievers” have some reason to believe that they are “resisting theistic
belief’ and that this resistance is “in various ways, blinding them.”37
37. Mark R. Talbot, “Is It Natural to Believe in God?” Faith and Philosophy, 6
(1989), 166, 168, 161, 165, 168. As far as I am aware, Talbot’s is the only piece of
For our purposes, it will be best to focus on Talbot’s argument

for the claim that Christians are justified in asserting the contrary-
to-fact conditional in question: it seems to me to display all the
errors that those who attempt a “resistance” explanation of non¬
belief tend to fall into. To do his argument justice, I must quote
Talbot at some length. The quotation divides fairly neatly into
three parts, and I will discuss it in stages corresponding to these.
The first part is prefaced by a brief discussion of Hume, who, ac¬
cording to Talbot, proposes for our belief what amounts to the
following contrary-to-fact conditional: “the laws of nature would
be observed to be perfectly regular, if it weren’t for our limited
powers of observation.”38 Talbot then goes on to argue as follows:
A. Someone is justified in asserting such a conditional if and only if

it seems true to her when she surveys all of her relevant experi¬
ence and belief. And, thus, asserting the uniformity of causal
influence is, as it should be, justified for almost all of us.
B. But, then, Christians can be justified in asserting that everybody

would believe in God, if it weren’t for sin.
For suppose someone is converted and then argues like this: “At
conversion, I realized that, up to then, I had been sinfully resist¬
ing acknowledging what I am naturally inclined to believe,
namely, that God exists, that he is my Maker, and that I ought
to make his will the law of my life. Others tell me that they
also, at conversion, realized this. Moreover, before I was con¬
verted, I wasn’t clearly aware of my resistance because of the
damage that sin had already done to my epistemic sets. There¬
fore, my experience—and that of these others—suggests that a
sinner’s judgment about moral and spiritual matters is unreli¬
able. And, therefore, my post-conversion experience justifies
writing in the contemporary literature of philosophy of religion devoted in its

entirety to defending the Calvinist view of nonbelief, and so I focus on it. Others
have defended this view more briefly or in passing; their arguments seem to me to
suffer from weaknesses similar to those I claim to find in Talbot’s argument. SeeJ.
Kellenberger, The Cognitivity of Religion: Three Perspectives (London: Macmillan,
1985), pp. 136-137, 159-160; George Schlesinger, “The Availability of Evidence
in Support of Religious Belief,” Faith and Philosophy, 1 (1984), 426; Ronald J. Feen-
stra, “Natural Theology, Epistemic Parity, and Unbelief,” Modern Theology, 5
(1988), 9—11 -
38. Talbot, “Is It Natural to Believe?” 166.
my discounting such preconversion judgments and my accepting the

claim that everybody would believe in God, if it weren’t for sin. ”
C. How can anyone justifiably object to this?

No one can claim that the believer is unjustified in asserting
this conditional because that believer’s grounds do not convince
him. That begs the question against belief by failing to come to
grips with the believer’s claim that if you don’t find her reason¬
ing convincing it may be because you are unregenerate and thus
lack a reliable epistemic set. Nor can anyone insist that the be¬
liever must factor in, as evidence relevant to her asserting this
conditional, his opinion challenging it. For then his judgment
would necessarily weaken or falsify her assertion, which begs
the question again. Nor will it do to say that adequate grounds
for justifiable assertion, in circumstances like these, must guar¬
antee the assertion’s truth. For that applies a stricter standard to
the believer than it is possible to apply to oneself.39
A. Talbot argues that someone is justified in asserting a condi¬

tional of the sort in question if and only if it seems true to her
when she surveys all of her relevant experience and belief. This
seems to imply that when individuals are considering whether they
are justified in asserting such conditionals, the experiences and be¬
liefs of persons other than themselves are not relevant. That Talbot
accepts this is confirmed later when he argues that the Christian
who is investigating the justifiability of asserting that nonbelievers
are sinfully resisting belief in God need not consider any experi¬
ences or beliefs incompatible with her own. But how, we may ask,
can such epistemic isolationism be justified? Does it not imply the
repugnant conclusion that individuals are justified in asserting vir¬
tually anything, no matter how outrageous, so long as it seems to
them that their experience supports it? Was not Hitler, on this cri¬
terion, justified in supposing that the Jews deserved extermination?44’
39. Ibid., 167.

40. It may be replied that no contrary-to-fact conditional is here being asserted.
But Talbot does not show there to be a justification-related distinction between condi¬
tionals and other propositions such that the former may justifiably be asserted under
conditions of epistemic isolation whereas the latter may not. But if a conditional is
desired, consider the following: “The Jews would worship me as a god were they not
ignorant and worthless.”
Talbot himself gives no argument for his claim. Perhaps he takes it

to follow from a proper understanding of the Humean analogue.
But if so, he is wrong. For Hume, in the passage Talbot quotes,
clearly (and correctly) suggests the relevance to his claim of the
researches of individuals who have, for example, shown the hu¬
man body to be a “mighty complicated machine” which therefore
might well “appear very uncertain in its operation” while in fact
the laws of nature are “observed with the greatest regularity in its
internal operations and government.”41 And, furthermore, as Tal¬
bot himself notes, it is apparently part of Hume’s suggested justi¬
fication for belief in causal uniformity that virtually everyone is dis¬
posed to believe in it.42 So there is no support from Hume. And no
other support is suggested. Indeed, the more it is examined, the
more Talbot’s criterion for the justified assertion of contrary-to-
fact conditionals looks tailor-made to accommodate the particular
conditional “Everybody would believe in God, if it weren’t for
sin.” The only way Talbot’s criterion escapes these criticisms is if
he includes under “relevant belief’ what seems to the inquirer to be
the case upon carefully investigating the relevant experiences and
beliefs of others. But that this is not his intent seems clearly indi¬
cated by the next part of the quotation, to which I now turn.
B. There are two ways of reading the argument of Talbot’s hy¬
pothetical convert, but on neither reading does it provide the nec¬
essary support for its conclusion. Indeed, this is true even if we
allow Talbot his epistemic isolationism. On the first reading
(which, given Talbot’s earlier arguments, seems the more natural
one), “naturally inclined to believe” is to be understood as “in¬
clined qua human being to believe,” and “sinner” as “unbeliever.”
Read this way, the argument begs the question: by asserting that
she is inclined qua human being to believe, the believer ipso facto
asserts this of every human being and so asserts that in the absence
of resistance or rebellion she and every other human being must
believe, which is to say that if there are those who do not believe,
this is to be attributed to resistance or rebellion on their part. On
this interpretation, the premise “I am naturally inclined to believe”
entails the argument’s conclusion but is logically irrelevant to the
41. Talbot, “Is it Natural to Believe?” 166.

42. Ibid., 164.
enterprise of establishing it, since it presupposes its truth. Thus, on

its first reading, the believer’s argument begs the question.
Now of course, although even an epistemic isolationist must, if
she wishes to argue from her experience to some conclusion, avoid
fallacies like that of begging the question, perhaps she need not
argue at all. It is at least initially puzzling why, given his criterion
of justification, Talbot wishes his hypothetical convert to argue:
what is to prevent her from simply asserting the conditional in
question, full stop—perhaps together with a brief explanation that
this is what she is inclined to believe? But it seems that Talbot does
want to argue, and not simply to assert his conclusion at the out¬
set. He implies, for example, that in his view the conditional in
question is not entailed by the experience of the new convert, but
that that experience instead provides certain nonconclusive grounds
for believing it to be true. (See Part C of the quotation above,
especially its penultimate sentence.) I suggest, therefore, that we
look again at the hypothetical convert’s claims in an attempt to
find a reading of them compatible with Talbot’s apparent argu¬
mentative intent. As it seems to me, the only possible such reading
involves understanding the phrase “what I am naturally inclined to
believe,” not as “what I am inclined qua human being to believe,”
but as “what I in fact am inclined to believe” (or “what I actually
believed all along”) and the term “sinner,” not as “unbeliever,”
but—and here a longer explication is unavoidable—as “one who
resists acknowledging what he or she is in fact inclined to believe. ”
Given these explications, the hypothetical convert has indeed got
an argument, but unfortunately for Talbot, only an argument for
the conclusion that anyone who (like herself) resists acknowledg¬
ing what he or she is in fact inclined to believe would believe if it
weren’t for their resistance. The conclusion that everybody would
believe in God if it weren’t for resistance is wildly unrelated to the
evidence the convert is actually in a position to adduce on the basis
of her experience. Only if some premise about everyone being nat¬
urally inclined to believe is illicitly slipped into the argument could
the universal conclusion follow. But then, as we have seen, the
argument would beg the question.43
43. I have argued that the believer’s experience gives her no good reason to
affirm the conditional in question. If this is correct, then, a fortiori, the believer’s
C. Talbot asks (with reference to the argument described under

Part B of the passage quoted above), “How can anyone justifiably
object to this?” The answer to this question should be clear by
now. But it will be useful to discuss briefly the particular objec¬
tions Talbot goes on to consider and his responses to them. These
objections are (i) that the believer is not justified in believing the
conditional because her grounds do not convince the unbeliever,
and (2) that the believer, to be justified in her claim, must factor in,
as relevant evidence, the unbeliever’s positive reasons for taking a
contrary stance. Now the rejection of (1) and (2) would seem to
follow directly from Talbot’s “isolationist” criterion of justifica¬
tion, enunciated in Part A of the passage quoted above. But here
he offers an additional reason for their rejection, namely, that they
are question-begging. The argument for this claim in the case of
(1) seems to be as follows: “Given the truth of the believer’s claim,
the unbeliever’s unwillingness to accept the grounds that support it
is to be expected. By claiming that the believer is not justified in
accepting her claim because of his dissent, the unbeliever is there¬
fore presupposing the probable falsehood of that claim (if it were
true, his dissent would not be relevant) and so is begging the ques¬
tion.” But given our earlier discussion, this argument can quite
easily be rendered ineffective. For only a foolish unbeliever would
urge the relevance of his view that the believer’s premises are false.
The proper point to make is the logically prior point that the be¬
liever’s premises, even if true, fail for reasons of logic to support her
conclusion. Given the first reading of the believer’s argument, no
real grounds have been presented (the argument begs the ques¬
tion), and given the second, the believer’s grounds, such as they
are, do not lead validly to her conclusion. Understood this way,
the first objection does not beg the question. What the unbeliever
is drawing to the believer’s attention are certain logical points. And
I presume that Talbot would not wish to go so far as to suggest
that if the believer’s claim is true, even the unbeliever’s views on
logic are incorrect. If this is right, it follows that the unbeliever’s
dissent, if it derives from logical considerations of the sort in ques-
experience fails to provide the nonbeliever with any appreciable reason for affirm¬
ing the conditional, and Talbot’s claim (ibid., 168) that it does provide such reason
is to be rejected.
tion, should provoke the believer to the realization that her claim
lacks justification.
What about the second objection? This, as will be recalled, was
that the believer must factor in, as relevant evidence, the unbe¬
liever’s positive reasons for taking a contrary view. Is this question¬
begging? Talbot’s argument in defense of an affirmative answer
seems to run as follows: “By claiming that the believer is not justi¬
fied in making her claim unless she takes his reasons for accepting
its denial into account, the unbeliever assumes that his reasoning is
not unreliable. But given the truth of the believer’s claim, the un¬
believer’s reasoning in this context is unreliable.44 Hence the unbe¬
liever, in making his objection, presupposes the falsehood of the
believer’s claim and so begs the question.” Let us suppose that the
second premise here is true—that the unbeliever’s reasoning in this
context is unreliable if the believer’s claim is true. Even then, we
may show that Talbot’s conclusion need not be accepted, for his
first premise is false. (Or rather, it is false under the interpretation
required to generate the conclusion.) The unbeliever clearly must
not assume that his reasoning is “not unreliable” in the sense of “in
fact reliable” when the proposition at issue entails the unreliability
of his reasoning. Otherwise his argument is indeed question-beg¬
ging. But why should we suppose that the unbeliever needs to as¬
sume this in order to claim that the believer is not justified in
making her claim unless she takes his reasons for denying it into
account? All he need assume is that it has not been shown that his
reasoning is unreliable—that it is at the start of the investigation an
open question whether the contrary-to-fact conditional at issue is
true or not, and that his arguments are for this reason not justifia¬
bly viewed as less likely to be reliable than the believer’s, and
should therefore be taken into account.45 It is clearly compatible
with this assumption that the believer’s claim is true (for the unbe¬
liever allows that his reasoning may be unreliable). Hence the un-
44. The context suggests that what Talbot means is that given the truth of the
believer’s claim, the unbeliever’s reasoning is of an unreliable type, that is, that the
unbeliever is qua sinner unreliable in his reasoning.
45. After all, it is not as though the reliability of the unbeliever’s reasoning is
less likely than it would otherwise be just by virtue of the question of its reliability
being raised! At the beginning of investigation there is by definition no presumption
in favor of either claim.
believer may press his objection without begging the question.

And since there is good reason to suppose that Talbot’s epistemic
isolationism must be rejected, this objection succeeds. Talbot’s be¬
liever cannot legitimately refuse to consider as relevant the reason¬
ings of those who would deny her claim.
We have seen that arguments of the sort put forward by Talbot,
which consider the relevance of the believer’s experience, fail to
provide even the believer with justification for the claim “Every¬
body would believe in God, if it weren’t for sin.” If we now rein¬
troduce the considerations adduced in the first section of this chap¬
ter (and we must remind ourselves that in view of the untenability
of epistemic isolationism and the failure of Talbot’s response to the
second objection above, even the believer must take these into ac¬
count), we can see, I think, that honest inquirers have very good
reason indeed to accept that not all failures to believe are due to the
sin of the nonbeliever, and in particular, that inculpable doubt oc¬
curs. There is, for example, good reason to suppose that some
who claim they have no private experiences apparently of God or
that such experiences as they do have are ambiguous, and who
have carefully examined the relevant arguments, finding them in¬
decisive, have no wish to be in doubt. Indeed, there are doubters
who have agonized long years over matters of faith, hoping that
belief may come to them. Why, we may ask, should this be the
case if all doubt is due to a sinful rejection of belief? Now it is
perhaps not impossible that despite such strong positive evidence
and the absence of any significant contrary evidence, all doubters
have sinfully rejected belief, but given the circumstances, that
claim is, to say the least, unlikely.461 suggest, therefore, that we are
justified in rejecting it.
It is my conclusion, then, in view of all the arguments consid¬
ered, that in the actual world reasonable nonbelief occurs. Many
writers already accept this, and so I have no quarrel with them.
And although some latter-day Calvinians may be inclined to argue
in favor of a contrary position, their views are unsustainable.
46. An additional point is that if in the face of the evidence, believers confi¬
dently claim that all doubt is due to the sinful rejection of belief, they display the
very fault they claim to find in the nonbeliever—moral insensitivity. On the unbe¬
liever’s alleged moral insensitivity, see Talbot, “Is it Natural to Believe?” 168.
[ 4 ]
A Summation of the Case
Let us pause for a moment to consider the path along which we

have come. In the Introduction we saw that
(1) If there is a God, he is perfectly loving.
Exploring this idea of Divine love further in Chapters i and 2,

we found considerable support for the following claim:
(2) If a perfectly loving God exists, reasonable nonbelief does

not occur.
Most recently, in Chapter 3, it was shown that (2)’s consequent

is false, that
(3) Reasonable nonbelief occurs.
But (3), in combination with (2), yields
(4) No perfectly loving God exists;
and from (4), together with (1), it follows that
(5) There is no God.
We have arrived, then, at an argument of considerable force

from the reasonableness of nonbelief to the nonexistence of God.
In this final chapter of Part i, I wish to consider more closely the

status of this argument (call it A), and the conditions that must be
satisfied if it is to be rebutted. We are, in other words, tying up
some loose ends and setting the stage for Part 2. On the issue of
status, my view is as follows. First, it seems clear that A commits
no error of logic: given (1), (2), and (3), we may validly conclude
(5). Equally clear, I suggest, is the truth of (1), and (3); these prem¬
ises seem more than adequately supported by the arguments I have
given in their defense. It follows that everything depends on (2)—
if it is true, A is sound. So what is (2)’s status? As it seems to me,
our arguments for (2) constitute at least prima facie evidence that it is
true: the reasons detailed in Chapter 1 for supposing that God, if
perfectly loving, will prevent the occurrence of reasonable non¬
belief are clearly sufficient to warrant the conclusion that he will do
so unless an adequate defense of that claim’s denial can be
mounted. It follows that A must be viewed as sound unless such a
defense can be given. There is, in other words, a presumption in
favor of its conclusion which is defeated only if our support for (2)
is defeated. We may think of our prima facie evidence for (2),
therefore, as conferring on A itself the status of a prima facie case.
This, at any rate, is how I will hereafter refer to it.
What, more specifically, must an argument against (2) show to
be “an adequate defense of that claim’s denial’’ and so, by exten¬
sion, a rebuttal for A? To facilitate discussion of this question, I
introduce the following stipulative definition: “p is plausible” =
df. up is such that there is at least as much reason to suppose it true
as to suppose it false.”1 2 The stipulative nature of this definition
must be carefully noted. By defining “plausible” in this way, I give
to it a narrower and weaker sense than it is commonly given. But it
will be convenient for us to have something less cumbersome than
“such that there is at least as much reason to suppose it true as to
suppose it false” to work with, and “plausible” is chosen for lack of
something better.’ Assuming this to be clear, I would suggest, as a
1. And more loosely, we may say, if p is plausible, that it is plausible to sup¬

pose that p, and, if the proposition “x is y” is plausible, that x is plausibly viewed
as y.
2. It might seem that “p is plausible” = df. “p is as probable as not” would
A Summation of the Case [85
first approximation, that we have a rebuttal for A of the sort re¬

quired if and only if we have an argument that succeeds in show¬
ing the plausibility of (2)’s denial. If the denial of (2) is shown
plausible, there is no longer a presumption in (2)’s favor or in favor
of A’s conclusion, (5), and so we are not justified in accepting A as
a sound argument; if it is not shown plausible, the presumption
remains, permitting us to conclude that A is, all things considered,
sound. This plausibility condition, it is important to note, is rela¬
tively weak, for it entails that to provide a rebuttal for A we need
show no more than that there is as much reason to suppose its cen¬
tral premise false as to suppose it true. Of course, if it could be
shown conclusively that (2) is false, we would also have provided a
rebuttal for A (for if the denial of (2) is proven conclusively, it is
plausible), but this much is not required (we can show (2)’s denial
to be plausible without proving it conclusively), although it is
required if the problem of reasonable nonbelief is to be finally
settled.
Let us look more closely now at the plausibility condition. What
must we show if we are to establish the plausibility of (2)’s denial?
To answer this question, we must know what are the conditions of
(2)’s falsity. I suggest that (2) is false if and only if there is a state of
affairs in the actual world which it would be logically impossible
for God to bring about without permitting the occurrence of at
least one instance of reasonable nonbelief,* * 3 for the sake of which
God would be willing to sacrifice the good of belief and all it en-
have served as well or better here. But some of the propositions I will be consider¬
ing seem necessarily true if true at all (e.g., propositions claiming that one state of
affairs is a greater good than another), and there is some doubt as to the appli¬
cability of terms like “probable as not” to necessary propositions: such proposi¬
tions are commonly viewed as having a probability of 1. Therefore, I have chosen
to speak in terms of reasons for and against. This language, it seems to me, is
clearly applicable to inductive evidence capturable by the probability calculus but is
appropriate as well where arguments for and against the necessity of a proposition
are concerned, since we do commonly speak of having reasons for believing some
proposition to be necessary.
3. It must be a state of affairs whose existence logically necessitates the permis¬
sion of reasonable nonbelief, for otherwise an omnipotent God could (and a per¬
fectly loving God would) bring it about without permitting the occurrence of rea¬
sonable nonbelief.
tails.4 If both these conditions are met, premise (2) of A is false; if

either one is not met, (2) (and so (5)) is true. From this conclusion,
together with the foregoing points, it follows that the denial of (2)
is shown plausible if and only if we can show to be plausible the
claim that each of the conditions just mentioned is met.
What, more specifically, must we be in a position to say about a
state of affairs necessitating the permission of reasonable nonbelief
if it is to be plausible to suppose that God would for its sake sacri¬
fice the good of belief (that the second condition mentioned above
is met)? In discussions of the problem of evil, to which this discus¬
sion is clearly related, it is sometimes said that the good for the
sake of which evil is permitted to occur must be a greater or out¬
weighing good, that is, it must be such that the world is better with
it and the corresponding evil than with neither.5 Other writers
have suggested that the good state of affairs need only be an equal
or offsetting good—in other words, such that the world would be at
least as good with it and the corresponding evil as with neither.6 It
seems to me that the latter, weaker, claim is the one we should
accept. An outweighing good, while clearly sufficient, is not nec¬
essary. For if there were a state of affairs which required for its
existence that God permit the occurrence of reasonable nonbelief
4. I say that it must be logically necessary that God permit the occurrence of
reasonable nonbelief, meaning by this, not that it must be logically necessary that
God bring about reasonable nonbelief, but that it must be logically necessary that
God bring about or allow the occurrence of reasonable nonbelief. Only by confusing
the denial of (2) (viz., “It is not the case that if God exists, reasonable nonbelief
does not occur”) with the stronger claim “If God exists, reasonable nonbelief oc¬
curs” could we be led to suppose that “permit” must here be read as “bring
about.” This point is an important one, for it implies that not only goods that
require the actual existence of reasonable nonbelief (and so God’s bringing it about
that reasonable nonbelief occurs) are relevant here, but also goods that require for
their existence only that God allow the occurrence of reasonable nonbelief (by,
e.g., leaving it up to human beings to determine whether reasonable nonbelief will
occur or not). Cf. the discussion of the Responsibility Argument in Chapter 7.
5. See, for example, Bruce Russell, “The Persistent Problem of Evil,” Faith and
Philosophy, 6 (1989), 122. As Russell points out (129), “good” and “good state of
affairs” must in the context of discussions like this be construed fairly broadly.
God might permit an evil to occur “to fulfill a duty or to satisfy some other deon-
tological requirement,” or to prevent “an even worse evil from occurring.”
6. See William L. Rowe, “The Empirical Argument from Evil,” in Robert
Audi and William J. Wainwright, eds., Rationality, Religious Belief, and Moral Com¬
mitment (Ithaca, N.Y.: Cornell University Press, 1986), p. 229.
and which constituted an offsetting good, God might very well

bring it about, and so might be kept from putting his existence
beyond reasonable nonbelief.
Allowing this, however, introduces certain complications,
which we must now sort out. For the operative phrase here is
“might very well.” We cannot suppose that if an offsetting good
requiring the permission of reasonable nonbelief presented itself,
God would automatically seek to bring it about. Because it was
only offsetting (i.e., because its value would no more than equal
the value represented by God’s putting his existence beyond reason¬
able nonbelief), he might, but also might not. So we cannot say of
the claim that God would bring about a certain (actually obtaining)
state of affairs necessitating the permission of reasonable nonbelief
that it is plausible when all we know is that the state of affairs is
plausibly viewed as an offsetting good, for then we are, given the
point just made, only in a position to say that it is plausible to
suppose that God might or might not bring it about—from which it
certainly does not follow that it is plausible to suppose he would do
so. We must have good reason to suppose that the state of affairs
constitutes an offsetting good before we can say that it is plausible
to suppose that God would bring it about. If we have good reason
to suppose this, we can straightforwardly say that God might or
might not bring it about (and not just that it is plausible to suppose
that he might or might not do so), which is to say that it is plaus¬
ible to suppose that he would: we have just as much reason to en¬
dorse this conclusion as to deny it. On the other hand, if we know
of the state of affairs that it is plausibly viewed as an outweighing
good, we have all that we need. For if it is outweighing, we may
be certain that God would bring it about;7 and therefore, if it is
plausible to suppose that it is outweighing, it is plausible to sup¬
pose that God would bring it about.
To sum up our discussion, then, the denial of (2) is shown to be
plausible and so A is rebutted if and only if (i) it is shown con¬
clusively that an offsetting good necessitating the permission of rea¬
sonable nonbelief exists, and/or (ii) it is shown to be plausible that
7. I, at any rate, will assume that we may be certain of this. In making this
assumption, it will be noted, I am only making it easier for my argument to be
rebutted.
an outweighing good requiring the permission of reasonable non¬

belief exists.8
I come now to an important objection to the claim that A consti¬
tutes a prima facie case for atheism which goes through if counter¬
vailing considerations are not uncovered. This objection is derived
from an argument of Stephen Wykstra.9 It can be developed as
follows: “A critical assumption of this chapter’s argument is that if
it turns out that all the goods we know of do not provide a rebuttal
for A, we may infer that God has no good reason to permit the
occurrence of reasonable nonbelief, and so that premise (2) of A
(and hence A’s conclusion) is true. But while this assumption, or
an analogous assumption, is appropriate in legal contexts, where
the notion of prima facie evidence originates, it is manifestly inap¬
propriate here. In legal contexts it may be assumed that if the de¬
fendant is innocent despite strong initial indications of his guilt,
signs of this are likely to become apparent to careful investigators.
But in the present case, things are very different. Because of the
weakness of our moral vision and the shortness of our cognitive
grasp, and because of God’s status as the omniscient creator of all,
we may say that if there are God-purposed goods corresponding to
the evil we experience, they are likely to be inscrutable, that is, be¬
yond our grasp. Hence the fact that we find no rebuttal for our
argument (if we do not)—our evidence—is not more to be ex¬
pected given that no God-purposed goods exist than otherwise: we
have no reason to suppose that this would be any less likely to
occur given that God-purposed goods actually exist than other¬
wise. But our evidence (supposing we have it) can justify the con¬
clusion that no God-purposed goods exist only if it is more to be
8. Correspondingly, a rebuttal attempt of this sort fails if and only if (i) there is
good reason to suppose that there is no necessary connection between the obtain¬
ing of the state of affairs in question and the reasonableness of nonbelief, and/or (ii)
there is good reason to suppose that the state of affairs does not constitute an
outweighing good and it is no more than plausible that the state of affairs consti¬
tutes an offsetting good.
9. See Stephen Wykstra, “The Humean Obstacle to Evidential Arguments
from Suffering: On Avoiding the Evils of‘Appearance’,” International Journal for
Philosophy of Religion, 16 (1984), 73-93. See also Bruce Russell and Stephen Wyk¬
stra, “The ‘Inductive’ Argument from Evil: A Dialogue,” Philosophical Topics, 16
(1988), 133-160.
A Summation of the Case [ 89
expected given that no God-purposed goods exist than otherwise,

for it is a basic principle of confirmation theory that some claim C
is more probable than not given evidence e only if e is more to be
expected, more probable, given C than given not-C. Hence our
evidence does not justify it.”
Now it might seem that to repel this objection we need only
make a distinction between God-purposed and God-justifying
goods. What we have been discussing in this chapter, it might be
said, are the latter, and all we need for our argument is to show
that if we find no goods that are even plausibly viewed as God-jus¬
tifying, we can infer that no God-justifying goods exist. But then, it
might be continued, our inference is secure against the Wykstra-
style objection. For to block this inference, the objector would
have to show not only that we might expect God -purposed goods
to be inscrutable, but also that wo/i-God-purposed goods of the
requisite sort (which could exist even if God does not) might be
expected to be inscrutable. Otherwise, it might be concluded, the
objector must beg the question in favor of God’s existence.10
But this reply will not do. As the reply itself suggests, the claim
that there are no goods of the sort in question (i.e., no God-justify¬
ing goods) entails, inter alia, that there are no God-purposed goods,
and so any evidence supporting the former claim must support the
latter as well. But if the Wykstra-style objector is right, the fact of
our inability to find the requisite goods (supposing it to be a fact)
does not support the latter claim. Hence, if the objector is correct,
it cannot support the former either.
So I think it is safe to assume that if the claim about the in¬
scrutability of God-purposed goods is justified, the inductive infer¬
ence in question is indeed blocked. But is it justified? As it seems to
me, an affirmative answer faces some serious problems. We must,
first of all, distinguish between the claim that there are possible
goods that we cannot grasp and the claim that the permission of
evil is logically necessary for the existence of such goods. It is the
second claim that the objector is apparently saying will be true if
10. Bruce Russell has recently leveled just this sort of criticism against the origi¬
nal Wykstra argument. See Russell and Wykstra, “‘Inductive’ Argument from
Evil,” 147-148, 154. For the reasons given immediately below, it does not seem to
me to succeed.
there is a God, and it seems clear that it has much less in its favor
than the first. It is to be expected, perhaps, that a God would
know of kinds of goodness that are impossible for us to under¬
stand. But why should this lead us to suppose that evils like that of
the reasonableness of nonbelief, with which we are intimately ac¬
quainted, in fact serve such goods if God exists? If there is a per¬
fectly good and loving God, he has, in creating, sacrificed his own
interests and taken on ours. We can therefore infer with some con¬
fidence that the goods many evils serve, if God exists, are human
goods—goods related to our “salvation.” But if so related, it seems
unlikely that they, or their relation to evil, should be impossible
for us to grasp. Goods of the latter sort we might expect to be
totally unrelated to the vicissitudes of human life.
Applying a point made by William Rowe, we might also point
out that the goods for the sake of which God would permit evil to
occur would likely be “goods that either are or include good expe¬
riences” of the human beings in question. God, we might suppose,
would not subject anyone to evil who was not to “figure signifi¬
cantly in the good” for the sake of which it was permitted. But
then we might expect to know of the goods evils serve, for “the
conscious experiences of others are among the sorts of things we
do know.”11
We should consider, finally, the great specificity of the claim that
reasonable nonbelief will serve inscrutable goods if God exists. The
claim that there are some evils (we don’t know which) that will
appear pointless if there is a God may not be clearly false, but
when it is made on behalf of the occurrence of reasonable nonbelief
in particular, we should be much less sanguine. Suppose that it is as
likely as not (despite criticisms like those mentioned above) that if
there is a God, at any rate some evils of human experience serve
inscrutable goods. How can we know, with regard to the occur¬
rence of reasonable nonbelief, that it is a member of this class? With¬
out independent information bearing on the question, we would
have to say, “Well, it might belong to that group, but then again it
might not.” Hence the probability that reasonable nonbelief will be
apparently pointless if God exists would seem to be at most half
11. Rowe, “Empirical Argument from Evil,” p. 244.

that of the proposition that some instances of evil will be so. But
then, given that we can surely say no more than that the latter
proposition is as probable as not, the probability of the claim that
reasonable nonbelief will be apparently pointless if God exists is
clearly too low for rational acceptance.
The Wykstra-style objection, therefore, appears not to provide
any reason to suppose that the inductive inference in question is
unjustified. Accordingly, should we turn up no consideration of
the sort required, we might indeed be justified in concluding on
that ground that none exists, and that premise (2) of A is true. It
must now be noted, however, that even if this objection were
shown to be correct, A's claim to the status of a prima facie case
would not be imperiled. For even if the evidence of our inability to
find any goods of the sort in question does not permit the inference
that there are none, and so we cannot draw this conclusion from
the failure of the arguments of Part 2 or in this way show that (2) is
true, we may still infer from the failure of these arguments that we
have no reason to deny (2); hence, given the point made earlier (from
which the Wykstra argument has momentarily diverted our atten¬
tion), should these arguments fail, we would still be in a position
to conclude that (2) (and so (5)) is true. As I have argued, there is
considerable reason to affirm (2), which establishes a presumption
in its favor. Accordingly, if the denial of (2) is not shown to be at
least plausible, we must accept the claim that (2) is true as much
better substantiated than the claim that it is false, in other words, as
more probable than not. Against this, it seems, Wykstra could
have nothing to say. His argument (even if successful) does not
provide any reason to believe that the relevant goods exist; it only
provides reason to suppose that our inability to find them is not
evidence that they do not exist, and the denial of the latter claim is
not essential to my case.
This brings me to my final point. I have been claiming that un¬
less a rebuttal for A can be provided—that is, unless it can be
shown plausible to suppose that there are countervailing considera¬
tions of the sort characterized above—its conclusion, that God
does not exist, goes through. I will be looking at various argu¬
ments that may provide such a rebuttal in Part 2. What I wish to
point out is that our discussion in this chapter has clarified both
their proper role or function and their great importance. Those

who have criticized instances of such arguments have sometimes
suggested that it is puzzling that they apparently do not succeed
and that no explanation of the absence of clear evidence for theism
can apparently be given. If my claims in Part 1 have been correct,
however, and if there is indeed a prima facie case against theism of
the sort I have described, it is clear that a much stronger conclusion
than this must be drawn if no such argument can be shown to
succeed, namely, that there is no God.12
12. This, at any rate, is the case if our assumption that neither theism nor athe¬
ism is clearly better-evidenced than its denial is accepted. There will be more on
this in the Conclusion.
Part 2
The Force of the Argument

[ 5 ]
Moral Freedom and Its
Requirements
So far in the book my concern has been to show that the argu¬
ment from the reasonableness of nonbelief is a successful prima
facie argument—that in the absence of countervailing considera¬
tions, the reasonableness of nonbelief shows the nonexistence of
God. Now in Part 2 I take up the question whether any such con¬
siderations can be adduced. As was mentioned in the Introduction,
not all of the considerations in the literature that are relevant to the
answering of this question have been given a clear shape in the past
or are applicable as they stand, and so my procedure is to clarify
and adapt them as I go along. Questions of interpretation of course
inevitably arise in a project of this sort, but I wish to make clear at
the outset that my main aim is to get as many relevant arguments
out into the open as possible, and to assess their force according to
the criterion for rebuttal set out in Chapter 4. This is not to say
that I do not attempt to determine, in individual cases, what
writers have claimed, but where their intent is unclear, I move
quickly to talk of “possible” arguments. For our purposes, it does
not ultimately matter whether some argument is actually to be at¬
tributed to a writer or not. If it is an argument suggested by his
writing which can be turned into a possible rebuttal, I will consider
it. A corollary of this procedural point is that there is here no chro¬
nological ordering of arguments or detailed historical discussion.
Where there is a reason for the ordering of arguments, it has more
96 ] The Force of the Argument
to do with logical relations holding among them than with any

historical consideration.1
In this first chapter of Part 2, I consider arguments from the
value of moral freedom. It is of course widely held that an appeal to
this good is an essential part of any credible theistic response to the
problem of evil, and so, given that the problem of reasonable non¬
belief can be construed as a special instance of this problem, it is
perhaps not surprising that free will arguments have been deployed
here as well. According to the writers whose arguments I will be
considering—John Hick and Richard Swinburne—God is con¬
cerned to ensure that our moral freedom be preserved, and thus is,
to the extent required for the preservation of this good, hidden
from us.2 Since a proper explication of “the extent required”
seems, for both writers, to entail reference to the permission of
reasonable nonbelief, their arguments are clearly relevant to our
discussion. In assessing them, I will assume that the freedom said
to be valuable is indeed a good for the sake of which God would, if
necessary, permit reasonable nonbelief to occur. For it is itself a
necessary condition of personal relationship with God, and quite
apart from this, of great intrinsic value. But I will maintain that the
permission of a weak epistemic situation in relation to theism is not
required for the existence of this good, and that there is in conse¬
quence no successful free will defense against the problem of rea¬
sonable nonbelief.
1. I do not mean to imply that historical questions are irrelevant. If through

careful investigation we discover that some historical figure x meant to say y in¬
stead of z, we may uncover a new argument. But there is obviously neither time
nor space here for the sort of detailed historical investigation that might be useful.
And in any case, by looking at all the arguments suggested by a person’s writings,
we stand a good chance of uncovering the argument actually intended, and so of
missing none.
2. Hick’s view on Divine hiddenness, which is well known and influential, is
commonly understood to involve only an emphasis on the value of cognitive free¬
dom, that is, freedom with respect to the acquisition of religious belief. But I will
argue that this interpretation—although a natural one given what Hick has to
say—is in fact mistaken. Hick is ultimately concerned for the preservation of our
moral freedom, and so discussion of his argument belongs in this chapter.
Moral Freedom and Its Requirements [ 97
Exposition of Hick’s Arguments
Hick’s claim, impressively developed in numerous writings, is

that the world is “capable of being interpreted intellectually and
experientially in both religious and naturalistic ways.” It is a “reli¬
giously ambiguous” world.3 This, apparently, is the state of affairs
he has in mind when he claims that God, if he exists, has set us at
an “epistemic distance” from himself.4 From where we are, God’s
existence is anything but obvious. The third of Hick’s famous
claims in this connection is that this situation is logically necessary
if our “cognitive freedom” in relation to God is to be preserved,
and the fourth is that this freedom must be preserved if God’s lov¬
ing purposes for us are to be fulfilled.5 These four claims—in par¬
ticular, the last two—represent Hick’s contribution to the discus¬
sion of our topic. Let us now examine them more closely.
Interpreters of Hick, no doubt influenced by his repeated use of
the phrase “cognitive freedom,” commonly suppose that his main
concern is to show the importance of freedom in respect of the
belief that there is a God. For example, Richard Swinburne, in his
Faith and Reason, associates Hick’s position with “the argument
that if there were a proof of the existence of God which became
known, those who heard it and understood it would have no op¬
tion but to believe and in that case faith would not be mer¬
itorious.” Responding to this argument, Swinburne writes that
“meritorious religious faith is not a matter merely or at all of hav¬
ing a belief that there is a God (and other beliefs also)—what mat¬
ters is how we act on our belief; what purposes, given that belief,
we seek to achieve.”6 Similar criticisms of Hick (suggesting similar
3. John Hick, An Interpretation of Religion (London: Macmillan, 1989), p. 129.

Further references to this work will be made parenthetically in the text.
4. See, for example, John Hick, The Second Christianity (London: SCM Press,
1983), pp. 47, 56. Further references to this work will be made parenthetically in
the text.
5. See John Hick, Evil and the God of Love, 2d ed. (London: Macmillan, 1985),
pp. 281-282, 287.
6. Richard Swinburne, Faith and Reason (Oxford: Clarendon Press, 1981), p.
86.
assumptions about his position) are put forward by Anthony

O’Hear and (implicitly) by Frank B. Dilley.7 8
Not only do many interpreters of Hick take him to be primarily
concerned with freely believing that God exists, many also read
him as saying, more specifically, that what this freedom amounts
to is the freedom to choose (directly) to believe (or not to believe)
that there is a God. C. Robert Mesle, for example, in a recent
article refers to Hick’s view as the view that we are “[free] to make
the voluntary cognitive choice to believe that God exists.”" A simi¬
lar interpretation is to be found in Robert McKim’s “The Hidden¬
ness of God.”9
It seems to me, however, that neither of the interpretations in
question—namely, that Hick is primarily concerned with freedom
in respect of the belief that there is a God and, more narrowly, that
on his view we are free to choose to believe or not to believe that
there is a God—is obviously correct. If Hick did make the latter
claim, his view would clearly be subject to the criticism that belief
is involuntary. But despite his frequent use of expressions that
seem at first sight to imply such a claim, I think it must be said that
an explicit and unambiguous statement or defense of it is not to be
found anywhere in his writings. On the basis of what is clearly
stated, I would venture to say that our cognitive freedom, on
Hick’s view, is more a matter of the necessity of searching for God
if we are to experience his presence and of the possibility of will¬
fully removing theological questions from our purview, than of
our freedom to choose to believe or not to believe that God exists.
And it seems to me that he is primarily concerned not with the
belief that there is a God, but with belief in God (i.e., religious
commitment, involving love and trust) and with the preservation
of our freedom in respect of it—a kind of freedom which he him¬
self refers to as moral freedom. It is because of what he takes to be a
7. See Anthony O’Hear, Experience, Explanation, and Faith (London: Routledge

and Kegan Paul, 1984), p. 239; and Frank B. Dilley, “Fool-Proof Proofs of God?”
International Journal for Philosophy of Religion, 8 (1977), 21.
8. C. Robert Mesle, “Does God Hide from Us? John Hick and Process Theol¬
ogy on Faith, Freedom, and Theodicy,” International Journal for Philosophy of Reli¬
gion, 24 (1988), 99.
9. Robert McKim, “The Hiddenness of God,” Religious Studies, 26 (1990), 150.
very close connection between the belief that there is a God and
belief in God that he sometimes focuses on the importance of free¬
dom in respect of the former.
This interpretation, of course, requires elucidation and defense. I
will begin with the point about Hick’s notion of cognitive free¬
dom.
It is Hick’s talk (apparently in connection with the acquisition of
theistic belief) of “cognitive decisions” and “acts of interpretation”
which seems to suggest the view that belief in the existence of God
is under our direct voluntary control.10 He writes in Faith and
Knowledge that “our knowledge of God ... is not given to us as a
compulsory perception, but is achieved as a voluntary act of inter¬
pretation.”11 And in An Interpretation of Religion a similar view is
expressed: “Faith as interpretation is ... a cognitive decision in
face of an intrinsically ambiguous universe” (p. 159).
We need not look far, however, to note that there appears to be
some confusion in Hick’s account over exactly what this “act of
interpretation” or “cognitive decision” amounts to; in particular,
over whether it is truly voluntary. For example, in Faith and
Knowledge (p. 102) he defines “interpretation” as a “recognition,”
“attribution,” or “perception” of “significance” (a significance at¬
taching, in the religious case, to the world as a whole). These
terms (with the possible exception of “attribution”) suggest that
interpretation is involuntary, but, despite the apparent incongruity,
Hick refers in the same breath to interpretation as an uact of recog¬
nition” (p. 102, my emphasis).12
10. Some writers not only may be influenced by Hick’s talk of cognitive deci¬
sions and the like, but may infer from his claim that religious belief must not be
forced upon us that, on his view, such belief should be (and is) under our (direct)
voluntary control. That this would be an illicit inference is clear; the view in ques¬
tion should not be the only one to occur to someone questing about for alterna¬
tives to intellectual coercion. Hick himself, as we will see, takes a less directly
voluntarist stance, but one compatible with the claim that we are not forced into
belief.
11. John Hick, Faith and Knowledge, 2d ed. (London: Macmillan, 1988), p. 121.
Further references to this work will be made parenthetically in the text.
12. Perhaps in order to soften the incongruity, Hick suggests that recognition
may be correct or incorrect, but this seems odd, to say the least. Later in the same
work, however (e.g., p. 115), it is clear that he is vacillating between a theo¬
logically committed and an ontologically neutral account.
i oo ] The Force of the Argument
It is worth pointing out that this confusion seems all-pervasive

of Hick’s discussion of interpretation in Faith and Knowledge, which
refers to recognition (or attribution) of significance as operative not
only at the religious level but at every level of our experience. Reli¬
gious significance, he writes, is only the “highest and ultimate or¬
der of significance. ” We may speak, in addition, of “natural” and
“ethical” significance (p. 113). And “in the case of each of these
three realms, the natural, the human, and the divine, a basic act of
interpretation is required which discloses to us the existence of the
sphere in question, thus providing the ground for our multifarious
detailed interpretations within that sphere” (pp. 107-108).13 But the
voluntarist position suggested here is apparently denied a few
pages later:
We cannot explain . . . how we are conscious of sensory phenomena

as constituting an objective physical environment; we just find our¬
selves interpreting the data of experience in this way. . . . Likewise
we cannot explain how we know ourselves to be responsible beings
subject to moral obligations; we just find ourselves interpreting our
social experience in this way. . . . The theistic believer cannot ex¬
plain how he knows the divine presence to be mediated through his
human experience. He just finds himself interpreting his experience in
this way. [Pp. 118-119, my emphases]
This passage, it is interesting to note, is reproduced in its entirety

in Hick’s most recent work, An Interpretation of Religion (p. 214).
So where does this lead us? I would suggest the following mod¬
est result. Although Hick wants to use libertarian language in de¬
scribing interpretation (“choice,” “act,” “decision,” etc.), he falls
into involuntarist forms of speech time and again. It is not clear
13. It may seem to some that this quotation, with its interesting distinction
between a “basic” act of interpretation which, so to speak, gets us started, and
more detailed interpretations which presuppose it, holds out a way of harmonizing
the (seemingly) incompatible emphases in Hick’s account. Perhaps the basic act of
interpretation is voluntary even if the others are not. This way of resolving the
problem is ultimately unsatisfactory, however. For Hick often recognizes no such
distinction, speaking of all our interpretive activity, both basic and nonbasic, in
involuntarist language (see, for example, Faith and Knowledge, pp. no, in, 114-
115).
Moral Freedom and Its Requirements [ ioi
that interpretation is on his view voluntary at all, and therefore,

even if belief formation (including formation of the belief that
there is a God) is being identified by him with interpretation, it
does not clearly and unambiguously follow that belief is, on his
view, under our voluntary control.
We must now note, however, that upon closer examination,
Hick’s view seems to be that interpretation and belief formation are
not to be identified, since the latter is a result of the former. This is
indicated most clearly in An Interpretation of Religion:
Propositional faith [i.e., belief] rests upon ... a distinctively

religious mode of experiencing the world and one’s life within it.
And I suggest the interpretive activity upon which this depends should be
equated with faith in its most fundamental sense. [P. 159, my em¬
phases]
How can religious experience be both powerfully convincing, leav¬

ing no room for doubt, and also an exercise of cognitive freedom in
response to ambiguity? The answer is that these phrases refer to differ¬
ent stages. Behind all conscious experience there lies a phase of un¬
conscious interpretive activity and it is here that, in the case of reli¬
gious experience, the free response to ambiguity occurs. In the
conscious experience the ambiguity has been resolved in a dis¬
tinctively religious . . . way, and the resulting experience itself may
have any degree of intensity and of compelling quality. [P. 170, n.
9, my emphasis]14
We should also note that, as these quotations already suggest,

belief in the existence of a Divine reality is on Hick’s view given in
and along with religious experience, from which it would seem to
follow that he views such belief as involuntary. Other passages
from the same work provide additional support for this interpreta¬
tion:
14. In this passage, it will be noted, the “cognitive decision” is said to be uncon¬
scious, and this is added reason to suppose that on Hick’s view, the primary “act”
of interpretation is not to be identified with the formation of belief, for belief (as he
implies) is given in and along with religious experience, and “experience” is earlier
defined by him as “a modification of the content of consciousness” (Interpretation of
Religion, p. 153).
What it is reasonable for a given person ... to believe depends in

large part upon ... his or her information or cognitive input. And
the input that is most centrally relevant in this case is religious expe¬
rience. [P. 211]
Theistic belief arises, like perceptual belief, from a natural response

of the human mind to its experiences. All that we can say of a form
of natural belief, whether perceptual, moral, or religious, is that it
occurs and seems to be firmly embedded in our human nature. [P. 214, my
emphases]1"
The observation that propositional faith is subjectively firm belief

corresponds to the powerfully convincing character of much reli¬
gious experience, leaving no room for doubt. [P. 159]
What our investigations reveal, therefore, is the following: (1)

Even if the “act” of interpretation were, on Hick’s view, identical
to the formation of belief, the conclusion that belief is, on his view,
a voluntary matter would not clearly follow, since he is ambivalent
about the voluntariness of interpretation. (2) Belief is in fact said by
Hick to follow interpretation—it is the indirect result of interpreta¬
tion—so that even if the latter is to be construed as voluntary, it
does not follow that the former is. (3) There is independent textual
evidence for the view that, whatever may be true of interpretation,
Hick considers belief to be /^voluntary (the output of religious ex¬
perience).
We have reason to suppose, then, that Hick does not hold the
view that we are free to choose to believe (or not to believe) that
there is a God. But while this helps us avoid a common misunder¬
standing of Hick’s view of cognitive freedom, it does not tell us
what the correct understanding is. He obviously thinks that we are
in some sense free with respect to belief and that this freedom is in
some sense connected to religious interpretation, but how we are
free and how this is related to interpretation we have yet to deter¬
mine.
To see what the proper understanding is, we must, I think, turn
to such passages as the following:
15. In the first part of this passage, Hick is describing a possible view, but the
context clearly indicates that he endorses it.
In order to know our environment aright . . . we have to interpret it

aright. . . . We are thus endowed with a significant measure of
cognitive freedom. Our powers of apprehension are improved and
extended not by eliminating but by deliberately perfecting their in¬
terpretive phase. We must often exert ourselves in relation to a sus¬
pected or reported or half-apprehended aspect of reality in order to
become more fully aware of it. What we can know depends in con¬
sequence, to an important extent, upon what we choose to be and
do. [Faith and Knowledge, p. 122]
The point being made here seems to be that cognitive freedom

involves the possibility of choosing, not to experience in a particu¬
lar way or to believe in the existence of the reality in question, but
to be open to its existence, to carefully and honestly follow up on
any suggestion of its existence we may find in our experience. The
point appears again a little later in Faith and Knowledge, except this
time with a more explicit reference to what we may choose not to
do:
Our rejection of moral obligations which we are unwilling to accept

does not typically take the form of a blank refusal to do what we see
to be right, but rather of an evasion at the prior stage of cognition,
the turning of a blind eye to the moral facts of the situation. We try
to exclude from our minds an obligation which is beginning to
dawn unwelcomely upon us. . . . This wilful moral blindness is an
exercise of cognitive freedom. The line on which we make our
stand is the outer defence of our personality, the frontier of aware¬
ness. . . . This frontier of the personality, which each man controls
for himself, safeguards his personal integrity and liberty in relation
to those aspects of the environment which would lay a claim upon
him. We have the primary cognitive freedom to recognize or reject
the credentials of any imperative which claims authority over us.
[Pp. 126-127]
What is important here, of course, is Hick’s application of this

point to the realm of the religious. In Faith and Knowledge he puts it
this way: the recognition of God is contingent on our “desiring to
enter into relationship with him” (p. 134). The same understanding
is evident in other writings. In his Arguments for the Existence of
God, Hick states that “the individual’s own free receptivity and
4] The Force of the Argument
responsiveness play an essential part in his dawning consciousness

of God.”16 In The Second Christianity, the knowledge of God is said
to depend on our “willingness” to live in God’s presence (p. 48).
And most recently, in An Interpretation of Religion, Hick speaks of
the exercise of cognitive freedom in terms of “the individual’s in¬
nermost choice of openness to the divine presence” (p. 159).
What a close examination reveals, then, is that cognitive free¬
dom, on Hick’s view, consists in the fact that if we are to know
God, we must first take certain steps in his direction, and that if we
do not wish to know him, we can always shut him out. Openness,
willingness, searching (and their contraries)—these represent the
volitional element with which Hick is concerned. (He connects
them to the notion of a “propensity” which may be encouraged or
thwarted, but this is perhaps dispensable.) Openness to God and a
willingness to live in his presence give rise to religious experience
and to belief; as he writes in The Second Christianity, “we can . . .
come to know God by a free response to the ambiguous indica¬
tions of his existence, a willingness to know him which then crys¬
tallizes into the experience of being in his presence” (p. 48). Our
freedom to determine whether belief is acquired is therefore indi¬
rect, consisting in our ability to allow or not to allow ourselves to
come to experience the world as mediating God’s presence and
purpose.17
With this understanding of cognitive freedom in mind, we must
now ask (moving on to the second part of my interpretation) why
Hick considers it to be important. As I noted earlier in the chapter,
Hick claims that cognitive freedom is necessary if God’s loving
purposes for us are to be fulfilled. Why is it necessary?
It seems to me that to make sense of this claim we must look
more closely at what Hick has to say about belief in God—about
the trusting, loving, obedient response of the believer to his aware-
16. John Hick, Arguments for the Existence of God (London: Macmillan, 1970), p.
114.
17. One of the merits of this interpretation of Hick is that it makes sense of his
repeated endorsement of Pascal’s claim that God “so regulates the knowledge of
Himself that He has given signs of Himself, visible to those who seek Him, and
not to those who seek Him not.” (The reference to Pascal is found in, e.g., Hick,
Faith and Knowledge, p. 141.)
ness of God’s existence. Many of Hick’s interpreters, as we saw,

take him to be primarily concerned with the preservation of our
freedom with respect to the belief that there is a God. But far from
being irrelevant or of secondary importance, as this would suggest,
belief in God is at the heart of Hick’s concern.
Of critical importance in this connection is what Hick says about
the nature of persons and personal relationships; in particular, what
he says about the conditions necessary for a personal relationship
with God. In order to bring this out clearly, I will quote at some
length from Faith and Knowledge:
The infinite nature of the Deity requires him to veil himself from
us if we are to exist as autonomous persons in his presence. For to
know God is ... to know the One ... in whose will lies our final
good and blessedness . . . , whose commands come with the accent
of absolute and unconditional demand. . . .
Clearly, to become aware of the existence of such a being must
affect us in a manner to which the awareness of other human per¬
sons can offer only a remote parallel. . . . [The believer’s] life must
become consciously reorientated towards a Being infinitely superior
to himself in worth as well as in power. There is thus involved a
radical reordering of his outlook as must be undergone willingly if it
is not to crush and even destroy the personality. . . . Only when we
ourselves voluntarily recognize God, desiring to enter into relation¬
ship with him, can our knowledge of him be compatible with our
freedom, and so with our existence as personal beings. . . .
If man is to be personal, God must be deus absconditus. . . . He
desires, not a compelled obedience but our uncoerced growth to¬
wards the humanity revealed in Christ. [Pp. 133-135]
Summing it all up a few pages later, Hick writes: “The reason why
God reveals himself indirectly—meeting us in and through the
world as mediating a significance which requires an appropriate
response on our part . . .—is that only thus can the conditions exist
for a personal relationship between God and man” (p. 140).
It would seem, therefore, that Hick is primarily concerned not
with cognitive freedom but with moral freedom—freedom with re¬
spect to belief in God. Our love, trust, and obedience must be
freely given; otherwise God’s intention to relate to us as persons
cannot be fulfilled. But this should not be taken to imply that he
views cognitive freedom as unimportant. Quite the contrary:

“Cognitive freedom . . . has a negative function, namely to protect
our finite freedom and autonomy” (Interpretation of Religion, p. 162,
my emphasis). As the passages quoted above already suggest, there
is, according to Hick, a very strong connection between belief that
and belief in, such that if one acquires the former, one must (in
some sense of “must”) also acquire the latter. Anyone who came
suddenly to believe that God exists as a result of God’s presenting
himself to her experience would on account of fear and astonish¬
ment automatically conform to what she perceived as God’s will.
Cognitive freedom is therefore, as Hick sees it, a necessary condition
of moral freedom. For we could not be forced to believe that there
is a God and yet remain free to respond or not to respond to this
belief.
The following passages also provide support for this interpreta¬
tion:
To know God is to know oneself as standing in a subordinate rela¬

tionship to a higher Being and to acknowledge the claims of that
being upon the whole range of one’s life. The act of will or the state
of willingness and consent by which one adopts the religious mode
of apperception is accordingly also an act of obedience or a willing¬
ness to obey. Thus although belief in the reality of God, and a prac¬
tical trust and obedience towards him, must be distinguished in
thought, they occur together and depend closely upon one another:
fides and fiducia are two elements in a single whole, which is man’s
awareness of the divine. [Faith and Knowledge, pp. 143-144]
God, if he is known to exist, can only be known as the One who

makes a total difference for us. For he is known as infinitely higher
than we, in worth as well as in power, and as having so made us
that our own final self-fulfillment and happiness are also the fulfill¬
ment of his purpose for us. I cannot know that such a being exists and be
at the same time indifferent to him. [Second Christianity, p. 48, my em¬
phasis]
Since Hick, as these passages indicate, considers belief in to follow

from belief that, his emphasis on the latter, which we have traced,
and on freedom with respect to it (cognitive freedom), is perfectly
compatible with my claim that he is ultimately interested in the
former, and in freedom with respect to it (moral freedom). As the

following more formal reconstruction of it shows, the movement
of his argument is a backward movement from the importance of
moral freedom to the necessity of cognitive freedom:
(1) If we were deprived of moral freedom in relation to

God—forced to obey God, to commit ourselves to him—
personal relationships between ourselves and God could
not exist.
(2) God wishes us to enter into personal relationship with

himself.
(3) So God will not deprive us of our moral freedom in rela¬

tion to himself. (From (1) and (2))
(4) Anyone forced to experience (and to believe in the exis¬

tence of) God is ipso facto forced to obey God, and so is
deprived of moral freedom in relation to God.
(5) Hence God will not force us to experience and to believe

in the existence of God, but will leave us with a measure
of freedom in this respect, that is, cognitive freedom.
(From (3) and (4))
Our answer to the question “Why, on Hick’s view, is cognitive

freedom essential to the fulfillment of God’s loving purposes for
us?” must consequently be the following. Cognitive freedom is
theologically necessary because only if it is in place can those who
experience the presence of God and commit themselves to him be
said to be morally free in relation to God. God, by making experi¬
ence of himself contingent upon a demonstrated disposition to be¬
lieve in his existence and live in his presence, creates a situation in
which those who experience God and find themselves obeying
God are by definition individuals who have shown themselves
willing to be in this position. In this way their moral autonomy,
and therefore the possibility of genuinely personal relationships be¬
tween themselves and God, are preserved.
So how does all of this bear on our discussion? Do Hick’s points
provide a possible rebuttal for the prima facie case of Part 1? Well,
given his famous claim about the connection between cognitive
freedom and religious ambiguity, mentioned at the outset of this dis-
cussion, I think Hick would endorse the following expansion of his

argument:
(6) Cognitive freedom requires that the world be religiously

ambiguous.
(7) God will therefore create a religiously ambiguous world.

(From (5)and (6))
(8) If God’s existence is beyond reasonable nonbelief, the

world is not religiously ambiguous in the required sense.
(9) Therefore, if God exists, his existence is not beyond rea¬

sonable nonbelief. (From (7) and (8))
Hick would likely concede to my argument prima facie force, ad¬

mitting the (initial) strangeness of the suggestion that support for
theistic belief could possibly stand in the way of explicit personal
relationship with God when belief is itself a logically necessary
condition of such relationship. But he would argue that since the
early availability of good evidence for theistic belief would have
the effect of removing our cognitive and so our moral freedom in
relation to God, and since this would rule out the possibility of
personal relationship with God, it is necessary for the realization of
such relationship that evidence not be provided as soon as a capac¬
ity for it exists. Belief must rather be something we come to over
time as a result of the exercise of moral freedom in investigation.
Hick’s Challenge: A Critique
Let us move on now to a discussion of objections that can be

raised against Hick’s account.18 As I pointed out in the preceding
18. One criticism of Hick I will not consider is that he reaches his conclusion
only by focusing too narrowly on religious experience. Hick himself has distin¬
guished between theoretical proofs and religious experience, arguing that only the
latter could remove human freedom (Arguments for God, pp. 105-107), and so
would seem clearly to be open to such an objection. But as I argued in Chapter 2,
there seems to be some reason to suppose that, were God to put his existence
beyond reasonable nonbelief, he would do so by means of religious experience.
And so it will be important to see whether Hick’s argument succeeds when stated
in those terms.
section, the “involuntariness of belief’ objection, which seems on a

surface reading to be an objection to which Hick’s account is sus¬
ceptible, in fact reflects only a misunderstanding of it. There are,
however, other objections that cannot so easily be turned aside.
The first of these is directed at Hick’s claim that openness to God
is sufficient for religious experience. It would seem that some who
earnestly seek to know God do not have experiences (apparently)
of him. Regardless, therefore, of whether Hick’s claim is based on
psychological or theological considerations or both, it would ap¬
pear to be false.
Suppose, however, that Hick is right on this matter and that
there is a “constant correlation” between a willingness to know
God and religious experience. As a second objection points out,
religious experience is not always coercive in the sense of creating
the “situation of a person who cannot help believing in the reality of
God.”'1' Some of those who have religious experiences—even of
those willing to know God—find them ambiguous and remain in
doubt as to the existence of God. Again, although it might seem
strange from a theological perspective that a willingness to know
God should be followed by a religious experience too weak to pro¬
duce belief (and this perhaps explains why Hick makes the connec¬
tion between experience and belief a tight one), it does seem that
this sometimes occurs.
Let us, however, grant for the sake of argument that here too
Hick’s view can be successfully defended—that a willingness to
know God leads inevitably both to religious experience and the ac¬
quisition of belief. The most fundamental objection to Hick is that
his claims with regard to the need for such a prior willingness if our
experience of God is to be compatible with our moral freedom
(premises (4) and (5) of his argument) are false, and that he has
therefore provided us with no reason to suppose that religious ex¬
perience must be withheld by God until it is shown. Even if our
situation were to be of the sort described in Chapter 2, one in
which belief is (at least in the first instance) the product of religious
experience and religious experience is available as soon as a capac¬
ity for personal relationship exists, it is not at all clear that individ-
19. Hick, Arguments for God, p. 114.

11 o ] The Force of the Argument
uals who came to believe would not be cognitively and morally

free.
In support of this, it can be argued, first of all, that there are
indirect ways in which a newly acquired belief can be resisted. One
can, if one finds the moral implications of a belief distasteful, avoid
acting upon it (moral freedom) by taking steps to remove one’s
active awareness of it or to lose it altogether (cognitive freedom).
That we are capable of self-deception is, as we saw in the previous
section, recognized by Hick, but he seems to think that it could
only operate prior to the formation of belief: once belief is
formed—at any rate, belief in the existence of God—it cannot be
successfully resisted. But there seems no reason to suppose that our
powers of self-deception are as restricted as this. It seems possible
to shut out an already formed belief—even belief in God’s exis¬
tence—as well as to fail to acquire it in the first place. As Anthony
Kenny writes in response to a similar view (to that of Hick’s) about
knowledge, “it is all too easy to shut one’s mind to what one
knows.”20 Where motives for self-deception exist (and they seem
not to be lacking in the religious case), an initial inclination to be¬
lieve can be overcome, for example, through looking at the evi¬
dence for and against again but selectively, and then taking steps to
forget having done so. This is all the more obviously possible
where (as in the circumstances described in Chapter 2) the evidence
does not render the proposition in question certain; any small mar¬
gin of negative probability can be blown all out of proportion if
one has the requisite motives.
I have argued that moral commitment need not follow upon the
formation of theistic belief because such belief can be successfully
resisted, but it is interesting to note that even if one does not take
this option and theistic belief is retained, moral commitment need
not follow. Hick’s claim (premise (4) of his argument) with regard
to the connection between believing and obeying—between cogni¬
tive and moral commitments, belief that and belief in—is unten¬
able. As Terence Penelhum writes, in direct opposition to Hick’s
view: “I can ‘know that such a Being exists and be at the same time
20. Anthony Kenny, Faith and Reason (New York: Columbia University Press,
1983), p. 77-
Moral Freedom and Its Requirements [in
indifferent to him’—if I lull myself with sufficient persistence.”21

Penelhum suggests that such “lulling” (self-deception, this time,
with respect to the moral implications of belief) could take the fol¬
lowing forms:
One way . . . might be to join an undemanding church organization

that let him [i.e., one who believes that there is a God but wishes to
avoid the moral implications of this belief] participate in cozy and
received practices that could serve as a specious sign of involvement
and lull him into complacency. This is . . . the sort of conventional
Christianity which Kierkegaard attacks most heatedly. Another way
would be to adjust his understanding of the moral demands to
which he saw he was subject so that they did not interfere much
with his worldly preferences. This is the sort of behaviour Pascal
ascribed, no doubt unfairly, to the Jesuits of his day."
Penelhum has suggested another reason as well for thinking that

theistic belief, even if retained, would not render us morally un¬
free: “Perhaps, what makes faith voluntary is not that its grounds
are inconclusive, but that even if they are conclusive, men are free
to deceive themselves and refuse to admit that they are. Faith
would be the outcome of a willingness to admit this, and faith and
knowledge need not then be exclusive at all.”23 In other words,
given that it is possible to deceive ourselves with respect to both
theistic belief and its moral implications, we are (meritoriously)
exercising both cognitive and moral freedom if, instead of giving in
to the temptation to deceive ourselves, we respond in the right way
to our beliefs.
There are, therefore, two ways in which cognitive and moral
freedom may be exercised even by one who retains theistic belief:
(i) She may deceive herself about the moral implications of her
belief. (2) She may, instead of actualizing the very real possibility
of self-deception with respect to either theistic belief or its moral
implications, embrace these beliefs gladly, willingly acting upon
21. Terence Penelhum, God and Skepticism (Dordrecht: Reidel, 1983), pp. in —
112.
22. Ibid., p. no.
23. Terence Penelhum, “The Analysis of Faith in St. Thomas Aquinas,” in Ter¬
ence Penelhum, ed., Faith (New York: Macmillan, 1989), p. 132.
11 2 ] The Force of the Argument
them in religiously appropriate ways. These points, together with

my earlier point about self-deception with respect to theistic belief
itself, suggest that a range of moral choices would remain open to
one provided with good evidence for theism.
How might Hick respond to these arguments? It seems to me
that he would be inclined to say that they hold only if we assume
that belief is arrived at inferentially, through argument, that things
are different where belief is the product of religious experience,
since it is impossible to ignore and remain unaffected by the latter.24
My arguments, he might say, are especially unconvincing if we
presuppose a situation of the sort described in Chapter 2, in which
all have forceful religious experiences as soon as the relevant capaci¬
ties are in place. Indeed, it would likely be his view that my argu¬
ments in Chapter 2 undermine those presented here. For (he might
suggest) in arguing that God should bring it about that individuals
have ongoing religious experiences and so (continuously) evidence
sufficient for belief, I have ipso facto argued (however inadver¬
tently) that God should be so close as to prevent a genuinely free
response.
To this I would respond, first of all, with a reminder of what we
have already seen, namely, that the sense of God’s presence, as I
suggested in Chapter 2 and as Hick himself states in An Interpreta¬
tion of Religion, need not always be “sharply focused,” but could
after its initial impression continue as “a general background
awareness” (p. 154). As such, it would not be inordinately intru¬
sive—incapable of being ignored or overridden. Hence self-decep¬
tion seems possible even when belief is the product of experience. I
have, in fact, deliberately made certain elements of the Chapter 2
description flexible to allow for this. The evidence of experience
there described is at all times sufficient to sustain a measure of belief
and so clearly need not be always forceful. As I have described the
experience, it is strongest when the legitimacy of belief genuinely
seems threatened. Hence there is no reason to suppose that God
must be vividly sensed as present in situations where, for example,
the moral implications of belief with respect to some particularly
difficult action are recognized. A related point, but one which
24. On this (alleged) difference, see Hick, Arguments for God, pp. 105-107.
Moral Freedom and Its Requirements [in
builds on what has already been said, is that even in a situation of

the sort described in Chapter 2, individuals would still be required
to deal with what Penelhum calls “the exigencies and distractions
of worldly existence.”25 Specifically, the evil which is so much a
part of human existence (and which, as we have shown, could very
well coexist with convincing evidence of God) would even in such
circumstances make self-deception or a grudging obedience a live
possibility and render meritorious a willing response. For there is
no reason to suppose that persons who believed under such cir¬
cumstances would feel less in the way of pain when serious illness
befell them or escape altogether the emotional havoc caused by, for
example, a death in the family.26
As far as I can see, the only way Hick gets his point, given these
considerations, is if he assumes that the religious experiences in
question are overwhelming—ones in which God is present to us
unmistakably and continuously—in such a way as to take our (moral)
breath away. He does in fact suggest at times (when arguments of
the sort I am advancing are seen as needing to be addressed) that
this is the only alternative to our actual situation of ambiguity. In
God and the Universe of Faiths, for example, he writes that
given . . . the human situation as we observe it, we form the theo¬
logical hypothesis that the meaning of human existence ... is that
God is creating finite personal beings, with a real freedom over
against himself, who may thus enter into personal relationship with
him. Suppose that such beings had been brought into existence in
the immediate “presence” of God. . . . They would not in that case
have had any real autonomy and freedom in relation to God. In
order then to possess such freedom they had to be created at ... an
epistemic distance. . . . They had to be brought into being in a
situation in which they are not automatically conscious of God. . . .
Now our actual human situation . . . fits these specifications.27
And in Faith and Knowledge he suggests that the only alternative to

ambiguity is God revealing himself to us “in the coercive way in
25. Penelhum, “Faith in Thomas Aquinas,” p. 132.

26. To avoid begging the question, I suggest we understand these afflictions as
ones that are not caused or preventable by human action, that is, as instances of
natural evil. But natural evil is, I think, quite enough.
27. John Hick, God and the Universe of Faiths (London: Macmillan, 1973), pp.
95-96.
which the physical world is disclosed to us” (p. 134). But this view
must be rejected, for it suggests that any slight change in our actual
situation with respect to the distribution and forcefulness of reli¬
gious experience would place us in the immediate presence of God,
and this is clearly false. Hick himself (at other times) recognizes
that it is false. Indeed, he argues that it is precisely to avoid this
sort of proximity to the Divine that we are created as part of a
physical universe in the first place:
The physical universe is a divine creation, determined by a purpose

which has deliberately made it an autonomously functioning sphere
in which its creator is not evident. ... Its function in relation to
ourselves is to enable finite personal life to exist in its own crea-
turely world. The Creator is so disproportionate to his creatures, as
the infinite being over against finite beings, that the two cannot ex¬
ist in the same sphere. If they were to run on the same rails of
physical existence there would be no room left for the creature!
Therefore God, who is not a physical entity but Spirit, has created
man within and as part of a physical universe in which he can live
his own proper creaturely life. For only if man is a free being exist¬
ing in his own sphere at a distance from his Creator can he make a
free response of faith and worship to that Creator. [Second Chris¬
tianity, p. 105]
This suggests that only in another life could we experience the di¬
rect presence of God (and Hick seems unsure at times whether ex¬
perience of God will be unmediated even then; see Faith and Knowl¬
edge, p. 187, and Interpretation of Religion, p. 179). In this life,
because of the physical “screen” between ourselves and God, we
cannot expect our experience to be of that kind:
The ordinary person’s religious awareness here on earth is not [a

vision of God in solitary glory]. He claims instead an apprehension
of God meeting him in and through his material and social environ¬
ments. ... In short, it is not apart from the course of mundane life,
but in it and through it, that the ordinary religious believer claims to
experience, however imperfectly and fragmentarily, the divine pres¬
ence and activity. [Faith and Knowledge, pp. 95-96]
The passage quoted earlier, therefore, presents us with a false

choice. God, according to Hick, cannot be revealed in the way in
Moral Freedom and Its Requirements [ns
which the physical world is disclosed to us. It is always “possible

for our minds to rest in the world itself without passing beyond it
to its Maker. ”2H If, as Hick suggests, it is only in the direct presence
of God that we would be morally unfree", and if, as he also states,
we can never be in the direct presence of God in this life, it follows
that we could never be morally unfree in this life, no matter what
form our experience of God might take. Of course, Hick may be
wrong here. There may well be forms of religious experience that,
even in this life, would make self-deception impossible and render
us morally unfree. But clearly experiences of this sort do not con¬
stitute the sole alternative to our actual situation. God’s presence
need not be overwhelming to be persuasive. Religious experiences
strong enough to remove the possibility of reasonable nonbelief
need not be so overwhelming as to “crush our autonomy.” Only a
quite unjustified insistence on more evidence than is needed or a
lack of imagination concerning the manner of its provision could,
it seems, lead anyone to suppose otherwise.
Hick, then, would appear to have an argument that cannot be
successfully defended. In particular, it does not provide the re¬
quired rebuttal. There is no reason to suppose that, were God to
put his existence beyond reasonable nonbelief, our cognitive and
moral freedom would be infringed, for there is no reason why we
should not exercise cognitive and moral freedom by rejecting God
or responding appropriately to him after this. For all that Hick has
told us, a personal relationship with God would still be possible
even if such a situation were to obtain.
A Variation on the Hickian Theme
Hick’s arguments, as we have seen, suggest that improvements

in our epistemic situation of the sort described in Part 1 would
entail the loss of cognitive and moral freedom in relation to God.
Given such improvements, theistic belief would be forced upon us,
and we would be forced as well to serve God. Fear and stupefac¬
tion would produce automatic conformity to what was perceived
as God’s will. A subtly different approach (but one leading to a
28. Hick, Evil and Love, p. 282.

similar conclusion) is taken by Richard Swinburne. Swinburne ar¬

gues, not that forceful religious experience, in particular, would
paralyze or shatter human personality and autonomy, but that any
too-clear indication of God’s existence would render obedience to
God eminently prudent and rational, removing temptation to do
wrong, and hence removing the freedom we now possess to form
our own characters for good or ill through our responses to tempta¬
tion (in Swinburne’s terminology, the freedom of a genuine choice
of destiny).
Now this notion of “self-constitution” is clearly related to some
aspects of Hick’s discussion, but Hick does not give much atten¬
tion to a concept central to Swinburne’s discussion, namely, that of
free will limited by desire. Like Hick, Swinburne holds that humans
are free in a sense not compatible with determinism—that is, they
“act intentionally,” and how they act is “not fully determined by
prior states of the world”—but, unlike Hick, he makes much of
the point that this freedom is and must be limited if a genuine
choice of destiny is to be possible. “Man has to struggle to act
rationally in a situation of temptation.” Yet, significantly, “if man
does have very limited free will, he seems also to have the power
to grow in the extent of his freedom, to decrease the influence over
himself of natural inclinations.” And, according to Swinburne, it is
in this capacity and in its corollary, the capacity to allow natural
inclinations to have their way, that a choice of destiny lies. Such a
choice therefore requires limited freedom. God, who is perfectly
free, does not and could not have a choice of destiny. Seeing the
good, and having no contrary desires, he must inevitably pursue it.
The freedom that humans actually possess, that is, limited free¬
dom, is the freedom they must have if they are to have the choice
of whether to “return to the level of the beasts or to move in the
direction of divinity. ”29
It is this notion of the necessity of limited freedom for a genuine
choice of destiny that provides the basis for Swinburne’s explana¬
tion of our epistemic situation. On his view, if God’s existence
were clear and evident, the desires for lesser goods essential to free-
pp. 153, 154, 157, 159.
dom of this sort would be overridden by strong desires to conform

to God’s will. This possibility did not come up in our discussion of
Hick. There it was possible to assume quite straightforwardly—
because nothing in Hick’s argument suggests otherwise—that hu¬
mans would, given clear evidence of God, continue to have desires
for bad actions and, thus, strong motives for self-deception. Swin¬
burne’s argument suggests that the structure of human desire sys¬
tems would change and, hence (implicitly), that such motives
would be greatly weakened. It is by virtue of this point, apparently
unnoticed by Hick, that Swinburne’s view requires separate treat¬
ment.30
Swinburne’s argument appears in The Existence of God as well as
in a more recent article titled “Knowledge from Experience, and
the Problem of Evil. ”31 In each case the situation envisaged is one
in which human beings become aware of the fact that there is a
God through his speaking to them at regular intervals. In The Exis¬
tence of God Swinburne spells out what he takes to be the implica¬
tions of such a situation as follows:
The existence of God would be for them [i.e., human beings] an

item of evident common knowledge. Knowing that there was a
God, men would know that their most secret thoughts and actions
were known to God; and knowing that he was just, they would
expect for their bad actions and thoughts whatever punishment was
just. Even if a good God would not punish bad men further, still
30. Hick puts so much emphasis on the need for freedom that he fails to note
that unlimited freedom would be as much a problem as no freedom at all—that it
would rule out the sort of difficult and gradual development toward personhood
that he, like Swinburne, considers to be of great value. Because of this, he is
prevented from seeing the possibility of a “prudential” argument of the kind Swin¬
burne advances. There is also an independent reason for supposing that he fails to
see this, namely, that he distinguishes between theoretical proofs and religious
experience, arguing that only the latter could remove human freedom (Arguments
for God, pp. 105-107). If he had seen the possibility of a prudential argument, he
would surely not have insisted on this distinction. For, clearly, proofs could indi¬
cate the path of prudence as well as religious experience, and (assuming that pru¬
dential reasons would have the force assigned to them by the Prudential Argu¬
ment) would therefore remove human freedom just as effectively.
31. Richard Swinburne, “Knowledge from Experience, and the Problem of
Evil,” in William J. Abraham and Steven W. Holtzer, eds., The Rationality of Reli¬
gious Belief (Oxford: Clarendon Press, 1987).
they would have the punishment of knowing that their bad actions
were known to God. They could no longer pose as respectable citi¬
zens; God would be too evident a member of the community. Fur¬
ther, in seeing God, as it were, face to face, men would see him to
be good and worshipful, and hence would have every reason for
conforming to his will. In such a world men would have little temp¬
tation to do wrong—it would be the mark of both prudence and
reason to do what was virtuous. Yet a man only has a genuine
choice of destiny if he has reasons for pursuing either good or evil
courses of action; for ... a man can only perform an action which
he has some reason to do.32
And in the article mentioned above, Swinburne puts the argument

in the following way:
I would regard my every movement as overseen by an all-knowing

and perfectly good being, namely a God who would therefore wish
me to be good, and value me and so preserve me, insofar as I was
good. The reasons for being good would be virtually irresistible: a
genuine choice of destiny would not be open to me, at least given
that men are as rational as they now are. If men were given a much
greater inbuilt depravity than they now have their choice would still
be open. By depravity I mean strong desires to do what is correctly
believed to be evil. But then such extra depravity would itself be a
great evil.33
A central notion here is that of temptation. By “temptation”

Swinburne means a “felt desire” or “natural inclination” to do
what is seen to be evil (all things considered), which “it needs an
effort of will” to stop one from acting upon. On his view, if we
knew of God’s existence, we would not be faced with temptation
32. Swinburne, Existence o f God, pp. 211-212.

33. Swinburne, “Knowledge from Experience,” p. 157. Swinburne’s point in
these passages was anticipated by Kant: “For suppose we could attain to scientific
knowledge of God’s existence, through our experience or in some other way. . . .
Then in this case, all our morality would break down. In his every action, man
would represent God to himself as a rewarder or avenger. This image would force
itself involuntarily upon his soul and his hope for reward and fear of punishment
would take the place of moral motives. Man would be virtuous out of sensuous
impulses” (Lectures on Philosophical Theology, Allen W. Wood and Gertrude M.
Clark, trans. [Ithaca, N.Y.: Cornell University Press, 1978], p. 123).
Moral Freedom and Its Requirements [ i 19
in this sense, for such knowledge would bring with it strong rea¬
sons for not doing evil—reasons stemming ultimately from desires
for our well-being, for example, the desire not to be punished.
Prudential reasons of this sort would in such a situation swing us
strongly in the direction of good action—it would require little in
the way of an act of will to master desires to do evil—and so a
genuine choice of destiny would not be open to us.34
Before I assess the force of this argument (as well as that of an¬
other, apparently independent, argument for the “no choice of des¬
tiny” conclusion to be found in the first passage quoted above),
two points should be noted. First, what Swinburne hints at in the
first passage (by saying little temptation) but does not bring out is
that, even in the state of affairs he has described, desires to perform
actions seen to be wrong would not simply vanish’, although such
desires would, if he is right, be rendered inefficacious by all the
reasons available for not doing evil (reasons stemming from strong
prudential desires), they might be expected to persist. As Swin¬
burne himself writes in The Evolution of the Soul, “the agent still
finds himself ready geared to do the action, even when he believes
it immoral,” and “it is the fate of humans that the inclinations to
act needed for enjoyment still move towards action (make the ac¬
tion easier, more natural to do) when the agent believes the action
overall bad.”35
Examples bring this out. If I strongly desire to go to the univer¬
sity bookstore to buy a volume that has just come out on the
shelves but recognize that in so doing I will fail to prepare ade¬
quately for an important lecture, my desire for the lesser good will
not simply disappear. I will do what I see as the better act (if I do it
at all) somewhat reluctantly, still feeling the pressure of my con¬
trary desire. This example can be made more directly relevant to
the present discussion as well. If I am told by someone who has
taken it upon himself to keep the book of which I desire a copy out
of the hands of the reading public that he will cause me serious
bodily harm if I go to the bookstore, and if I am certain that the
threat will be carried out if I do go, I will likely not go to the
34. Swinburne, Existence of God, p. 157.

35. Richard Swinburne, The Evolution of the Soul (Oxford: Clarendon Press,
1986), pp. 106, 107.
bookstore. But it does not follow that my desire to go will simply

vanish. It will at most be rendered inefficacious, and even as I
choose not to go in order to avoid bodily harm I may feel the
twinge that accompanies frustrated desire. What follows from this
is that when Swinburne says “little temptation,” he is best inter¬
preted as suggesting not that felt desires to do what is wrong
would in the situation in question be simply eliminated, but that
one’s overwhelming inclination would be to do what is right, so
that whatever felt desires one might have to do wrong would be
rendered inefficacious.
Another point which should be noted in this connection is that,
on Swinburne’s view, having a desire involves seeing the satisfac¬
tion of that desire as in some way a good thing, and so involves
having a reason for satisfying it: “An agent’s having a desire to do
some action is always a reason for his doing it.” This implies that
in the situation Swinburne describes, if contrary desires persisted,
as we have seen they would, human beings would continue to
have reasons for doing bad actions, although, if he is right, reasons
overwhelmed by other reasons for not doing bad actions.36
As I mentioned earlier, there seems to be a second argument as
well in the passage quoted above from The Existence of God. I find
it in the sentence following the “little temptation” claim, which
states that a man “only has a genuine choice of destiny if he has
reasons for pursuing either good or evil courses of action; for ... a
man can only perform an action which he has some reason to do.”
It is implied here that in the situation described, humans would
have no reason for performing bad actions, and that in order to
have a genuine choice of destiny, they must be capable of both
good and bad actions. The argument suggested by these claims is
the following:
(io) A person can only perform an action if he has some rea¬

son to perform it.
(n) If no reasons existed for performing bad actions, persons

would be incapable of them. (From (io))
36. Ibid., p. 115.

Moral Freedom and Its Requirements [ i2i
(12) If God’s existence were evident, persons would have no

reasons to do bad actions.
(13) So in such a situation bad actions couldn’t be done. (From
(11) and (12))
(14) But both good and bad actions must be possible if per¬
sons are to have a genuine choice of destiny.
(15) Therefore, in a situation of the sort in question, no one
would have a genuine choice of destiny. (From (13) and
(14))
But given my interim points above, it is clear that on Swinburne’s

own principles, premise (12) of this argument must be rejected.
Desires to do bad actions would not simply vanish even in a situa¬
tion of certainty of the sort Swinburne describes, and so there
would always be reasons for doing bad actions (although perhaps
reasons overwhelmed by contrary reasons). Thus, if we assume
that Swinburne’s points about desire and reason are correct, this
second argument for the conclusion that a genuine choice of des¬
tiny would in a situation of certainty not be open to human beings
is unsound. I do not have any difficulty with this assumption, and
so I think the argument is unsound.
As it seems to me, a more challenging argument is Swinburne’s
first argument, described above, which, put a little more formally,
runs as follows:
(16) In the situation in question, persons would have strong

prudential reasons for not doing wrong.
(17) Because of the strength of these reasons, it would require
little in the way of an act of will to do what was right—
there would be little temptation to do wrong, contrary
desires would be easily overcome.
(18) Where there is little temptation to do wrong, persons lack
a genuine choice of destiny.
(19) Therefore, in the situation in question no one would have
a genuine choice of destiny.
The situation referred to here is, as we have noted, one in which

God’s existence is “clear and evident,’’ and so one in which humans
know for certain that there is a God and in which whatever reasons
humans take themselves to have for doing good actions they con¬
sider themselves to certainly have. A situation in which the evidence
available is sufficient for belief (i.e., as we have seen, probabilify-
ing) is, however, not of this sort, and it is the latter situation that
must be shown to be problematic if the prima facie case of Part i is
to be rebutted. To assess the force of Swinburne’s argument in the
context of our discussion, therefore, we must take it to refer to
evidence sufficient for belief, and ask whether it is plausibly viewed
as sound when construed this way.
For the sake of argument, let us, for the moment, concede to
Swinburne the truth of (18). This leaves (16) and (17). Are these
premises plausibly viewed as true? Is it plausible to think of the
reasons in question as ones that humans would, given evidence
sufficient for belief, consider themselves to have (i.e., as pertinent)?
And are they plausibly viewed as strong enough to overwhelm
desires for bad actions (i.e., as efficacious)? The reasons in question
are, it seems, primarily reasons deriving from the belief that God
preserves human beings insofar as they do what is right, and more
specifically, that bad actions will be punished by God. Are these
reasons pertinent? And would they be efficacious? Let us begin
with the second question.
Unfortunately, Swinburne is less than clear (in the passages con¬
cerned) on how the relevant notion of punishment is to be under¬
stood. Does it, for example, imply punishment for each bad action
at the time at which it is performed, or only punishment in the
distant future, perhaps after death? And what is the nature of this
punishment? Bodily harm in the here and now or hellfire in the
hereafter or a progressively deteriorating quality of character lead¬
ing to eventual annihilation or. . . ?37 Clearly, to know whether the
reasons associated with punishment would be efficacious, we need
to know more about the form that punishment might be expected
to take. Swinburne comments: “whatever punishment is just.” But
this leaves the question wide open. ITuman beings, it seems, might
very well conceive of God as justly lenient in the moment of de-
37. This is the view Swinburne seems to support in more recent writing. See his
Responsiblity and Atonement (Oxford: Clarendon Press, 1989), pp. 180-184.
sire, and of punishment as, at worst, an afterlife affair, and hence

find themselves in a situation of temptation after all, and with a
genuine choice of destiny.
Suppose, however, that the punishment expected would be pun¬
ishment in the here and now, in the form of some sort of bodily
harm correlated with each bad action (perhaps something analo¬
gous to a severe electrical shock). This, it seems, would have quite
a strong deterrent effect. But would all temptation to do forbidden
actions be removed? Swinburne suggests that given such a state of
affairs, performing a bad action would amount to “deliberately
bringing harm upon oneself, ” and that few of us are ever tempted
to do this.38 But, clearly, one who performed a bad action in such a
situation would be acting on exceedingly strong desire, and so her
action would first of all be an action performed in satisfaction of
desire. To say that it would be an action of deliberately bringing
harm upon oneself obscures this fact. Desires may remain even
when one sees overriding reason not to satisfy them and may exer¬
cise an influence disproportionate to the weight one is inclined to
give them when dispassionately considering reasons for and against
action.39 In any case, given evidence sufficient for belief instead of
proof, one who is under the influence of desires for what is “cor¬
rectly believed to be evil” is likely to seize upon the margin of
possible error: believing, but not certain of God’s existence, or of
punishment, she may well move, through self-deception, from the
belief that God exists and will punish bad actions to (i) the belief
that God likely or as likely as not does not exist, and so that pun¬
ishment is unlikely, or only as likely as not, or (ii) the belief that
God will not punish after all, but will be tolerant of lapses from
correct behavior/’ If self-deception is still open to individuals, then
38. Swinburne, “Knowledge from Experience,” p. 157, n. 10.

39. Swinburne seems to think (see the last point of the second passage quoted
above) that our desires are not in fact of this character; that they would be so only
if God gave us “a much greater inbuilt depravity” than we in fact have. But this
rests on a judgment about the rationality of human beings which the empirical
evidence does not seem to me to support. There will be more on this later in the
section.
40. Of course, the subject may also deceive himself into thinking that the action
he desires is not after all so bad and that punishment is for this reason unlikely, or
clearly they are still in a position to yield to bad desires and so

retain a genuine choice of destiny.
Let us suppose, however, that these points can be answered and
that fear of punishment would, in the circumstances in question,
overwhelm contrary desires, greatly reducing temptation to do
wrong. It seems clear that Swinburne’s premise (17), when inter¬
preted in terms of punishment, requires that the circumstances be
circumstances of this sort (viz., ones in which the punishment ex¬
pected is punishment in the here and now, correlated with each
bad action). For as soon as we weaken the notion of punishment,
all the points I have been making—about the continuing, often
disproportionate, influence of desires, for example—come back
with redoubled force, and the notion that temptation would no
longer exist is rendered implausible. As soon as punishment is
pushed off into the future, rendered less immediate and concrete,
the force of any desires I may have to avoid punishment is reduced.
It always requires an act of will to give up short-term goods in
favor of longer-term interests. If punishment is seen as something
in the future, its deterrent effect must be greatly reduced. Then, in
addition, various further forms of self-deception become possible.
The agent may, for example, be inclined to reason as follows: “It is
not if I give in to this desire that God will punish me, perhaps even
annihilate me, but only if I persistently, to my life’s end, give in to
such desires. But of course, I do not intend to give in tomorrow,
or the next day . . .; only today. So I may perform this action
without fear of being punished.” I suggest, therefore, that it is only
if an individual believes that God’s policy on punishment implies
that a failure to do good actions will in the here and now result in
bodily harm or loss of life, that the motivating effect of his belief
can be plausibly viewed as great. Only then would good actions
involve little effort. If the reasons associated with punishment are
not given such a strong interpretation, they cannot legitimately be
viewed as efficacious.
that his desires are overpowering and so he is not responsible for his action. These
sorts of self-deception are possible even where there is certainty about God’s exis¬
tence and policies. Swinburne does not seem to me to take adequate account of
them.
Let us now turn to the first question mentioned above, concern¬

ing Swinburne’s premise (16). Are the reasons we have men¬
tioned—reasons associated with punishment—plausibly viewed as
pertinent? Are they reasons that, given evidence sufficient for belief,
human beings would in fact consider themselves to have?
It seems to me that these reasons are clearly not pertinent when
given the strong interpretation required. Why should God’s goodness
be taken to imply that he will harm us severely each time we do
wrong? It seems much more reasonable to suppose that persons
met in experience by a loving God would come to believe (cor¬
rectly) that God desired their deepest well-being, and that they
would, in consequence, be left without a clear belief about the ulti¬
mate implications of resisting the good or with the belief that God
would never refuse anyone a second chance (or, at the very most,
with the belief that only after persistently, over a long period of
time, rejecting the good, the likelihood of well-being in the hereaf¬
ter would be greatly diminished). Even if the expectation of pun¬
ishment in the strong sense were prevalent in some quarters at first,
upon further experience and reflection the understanding of hu¬
mans might be expected to mature and deepen (as it has in the
actual world under much less favorable epistemic conditions) to
the point where such views were universally rejected. Further,
those who (unreasonably) expected severe punishment to follow
each bad action would soon note that those who did not have this
expectation, and so occasionally fell into temptation and did bad
actions, were not immediately severely punished. As in our world,
the wicked would prosper, and all would soon come to the realiza¬
tion that, in the short term at least, the wicked could be expected
to continue to prosper.
The Swinburnian case fails, therefore, when stated in terms of
punishment. The plausibility of premise (17) is incompatible with
that of (16); that is, to be plausibly viewed as efficacious, reasons
associated with punishment must be interpreted in such a way as to
be clearly no longer pertinent. But perhaps Swinburne is less inter¬
ested in emphasizing the possible effects of punishment than the
passages quoted above seem to suggest, and so we must ask
whether there is some other way of understanding the prudential
reasons in question, and whether they are plausibly viewed as both
pertinent and efficacious when so understood. One such alternative

construal would involve giving more emphasis than we have so far
to a phrase appearing in the second passage, which affirms that
God, as perfectly good, will value (i.e., “think well of”) human
beings insofar as they are good." Developing this idea, Swinburne
may argue as follows: “All of us already have desires to be well
thought of by other persons, in particular, by other good persons.
We strive to do actions that will cause these others to hold us in
high regard. Of course, other human beings do not observe every¬
thing we do, and so it is possible to perform bad actions and still
be esteemed by them, but insofar as doing certain good actions is
required for them to think well of us, we desire to perform them.
If now we came in contact with God and so came to believe in
God’s existence—in the existence of a being at once unsurpassably
good and aware of all our actions and thoughts—this would give a
particular focus to the aforementioned prudential desires, resulting,
in effect, in a greatly increased desire to do good actions in every
circumstance. Desiring to be well thought of by God, whom we
knew to be aware of all our actions and thoughts, we would have
little temptation to do evil. It would be easy to do the good in
every situation, and so we would be left without a genuine choice
of destiny.”42
What should we say in response to this? Can premises (16) and
(17) of Swinburne’s argument both be accepted as plausible when
the prudential reasons in question are reasons associated with Di¬
vine approval, instead of reasons associated with punishment? Start¬
ing this time with the question of pertinence, I would suggest that
Swinburne’s case again faces difficulties. It is, first of all, not
clearly true that all or most of us have the desire to be well thought
of by good persons as opposed to others. And so seeing God to be
unsurpassably good might not create in us any particular desire to
be well thought of by him. Suppose, however, that Swinburne is
correct on this point. A more important criticism is that the con¬
ception of Divine approval at the heart of his argument seems to
41. Swinburne has indicated in conversation that “think well of’ is the meaning
(of “value”) he intends.
42. I am grateful to Professor Swinburne for drawing this alternative construal
of his passages to my attention.
conflict with the notions most of us would naturally associate with

perfect love—notions of equal regard and unconditional accept¬
ance—and so seems to be one that individuals who came to believe
themselves loved of God would reject. Now of course, presuma¬
bly few would, upon encountering a perfectly loving God, deny
that God desires humans to do good actions. (We might expect it to
be noticed, for example, that an obvious feature of love is desire
for the well-being of the beloved, and that human well-being, es¬
pecially if God exists, is best served by good actions.) The notion
of a God gladdened by our good actions and saddened by any de¬
viations therefrom is no doubt a pertinent one. And we may sup¬
pose as well that God would be viewed by individuals as taking
seriously human freedom and so as not willing to prevent humans
from doing wrong actions when they choose to do so. But none of
these beliefs implies that humans are (in what is apparently Swin¬
burne’s sense) less well thought of by God Vvhen they do bad actions.
It is compatible with these beliefs that God, if he exists and is per¬
fectly loving, views each human being as irreducibly valuable. And
this view is one that Swinburne’s notion of God thinking well of
human beings insofar as they are good seems, on its most natural
reading, to exclude.
Is there any other, more qualified, sense in which God may be
said to value (think well of) human beings insofar as they are good,
which individuals met in experience by a loving God would, plau¬
sibly, accept? Perhaps there is. Perhaps we can speak of God as
approving of our actions insofar as they are good and, therefore,
disapproving of what we do, and so in a sense (and indirectly) of
us, if our actions are bad. But this notion, to be pertinent, will have
to be spelled out in such a way as to be compatible with the view
mentioned earlier, namely, that God, if he exists and is perfectly
loving, accords to each of us a basic dignity and value which is not
altered by our actions, good or bad. Otherwise, it is not a notion
that we have any reason to believe persons met in experience by a
perfectly loving God would accept.
Let us turn now to premise (17) and the question of efficacious¬
ness. Under which, if either, of the interpretations we have distin¬
guished can the prudential reasons in question—reasons associated
with Divine approval—be plausibly viewed as strong enough to
overwhelm contrary desire? Let us take, first of all, the strongest

interpretation, according to which, if we fail to do good actions we
will in the here and now be rejected or ignored by God, and
viewed as inferior to other of his human creations. Supposing we
were provided with evidence sufficient for belief and saw this as an
entailment of our belief, would bad desires be overwhelmed? It
seems not. The consideration in question must be plausibly viewed
as giving rise to a very strong desire indeed before it can plausibly
be viewed as rendering easy those actions we all otherwise must
struggle to perform, and this requirement does not seem to be
met.43 If we take into account that God’s existence would not be
known for certain, that contrary desires would persist at least to
some extent, that God would not be physically present, like other
humans, and (as we saw in the preceding section) that even if God
chose to reveal himself in our experience, he might well bring it
about that the sense of his presence was diminished in moments of
ethical deliberation, we must surely conclude that many good
choices would remain difficult. It seems likely, for example, that
since (as Swinburne’s argument points out) we could continue to
have the high regard of our peers while doing bad actions (even if
not that of God), we would, under the pressure of contrary desire,
often be tempted to ignore God and focus on pleasing human be¬
ings instead, thus making it possible to do as we wished. Even if
this form of self-deception does not seem likely, others clearly are.
I might convince myself that my bad actions would not be re¬
peated tomorrow or the next day and that I would therefore not
really fall out of favor with God by performing them. I might
reason that God, having made me weak and subject to contrary
desire, would understand and tolerate at any rate some departures
from correct behavior. I might also convince myself that what I
wished to do was in fact legitimate, appearances to the contrary not¬
withstanding, and so remain assured of God’s favor for this reason.
In any of these ways (and no doubt, there are others), I might
convince myself that both the desire to be well thought of by God
and desires for lesser goods could be satisfied, and so find ways of
43. Of course, not all of us find the same actions difficult, but all that is needed
here is the assumption that each of us finds some actions difficult.
yielding to contrary desire without ignoring God. I suggest, there¬

fore, that even given a strong interpretation, the reasons under
consideration—reasons associated with Divine approval—are not
plausibly viewed as efficacious: there seems good reason to suppose
that given evidence sufficient for belief, they would not over¬
whelm contrary desire. And if this is true with respect to the stron¬
gest interpretation, it is true a fortiori where a weaker is concerned.
It seems, then, that the Swinburnian case is not successful. Of
the reasons mentioned, some, to be plausibly viewed as efficacious,
must be construed in such a way as to be clearly no longer perti¬
nent, and others are clearly not efficacious under any interpreta¬
tion. Since, as we have seen, there must be prudential reasons that
are plausibly viewed as both efficacious and pertinent if the Swin¬
burnian argument is to succeed, we may conclude that it does not
succeed.
Two final points. First, in evaluating arguments of the sort con¬
sidered here, we are not restricted to hypothetical examples. Their
claims may also be tested by reference to “real life” situations in
which people consider themselves to have experienced God and are
convinced of his existence and sustaining presence, and of his
moral demands. (Of course, it may be that their beliefs are mis¬
taken. But how is that relevant? The question at issue has been
“What effects might we expect assurance as to the existence of a
perfectly good and loving God to have?” and so it is not the objec¬
tive quality of the evidence these people have but the degree of
assurance it affords for them that counts.) Now such people seem
quite capable of doing what they believe to be wrong. They have
temptations, that is, they are forced to struggle with desires for
lesser goods, and occasionally yield to them. One might think in
this connection of the corruption in the television evangelism in¬
dustry: not all televangelists who end up corrupt started out that
way. But perhaps a more appropriate example is that of St. Paul,
who in Romans 7 writes, “I have the desire to do what is good,
but I cannot carry it out. For what I do is not the good I want to
do; no, the evil I do not want to do—this I keep on doing.” Surely
no one was ever more convinced on experiential grounds of God’s
existence and goodness than St. Paul. And yet, by his own admis¬
sion, bad actions continued. This (and more generally, the fact that
undeniably religious people evince weakness of will) provides my

response to Swinburne with additional support.
Finally, it should be noted that even if premises (16) and (17) of
Swinburne’s argument were to be plausible, it would not clearly
succeed as a rebuttal, since there are also reasons for rejecting (18).
Premise (18) states that “where there is little temptation to do
wrong [i.e., little temptation to fail in one’s obligations], persons
lack a genuine choice of destiny.” But this seems false. Swinburne
has neglected the connection between a choice of destiny and super¬
erogatory action. Even if his claims are plausible where obligatory
action is concerned (and I have argued that they are not), it seems
that supererogatory action must escape the net: there is no reason
for anyone to suppose, for example, that failure to perform super¬
erogatory actions would ever be punished by God. But if this is the
case, then even if persons would not in circumstances of evidence
sufficient for belief be tempted to fail in their obligations, they
might still desire not to perform supererogatory actions, and thus
have the opportunity of progressively overcoming such desires
through acts of will and, thereby, of molding their characters for the
good. Even though they might not, in such circumstances, have the
opportunity of returning “to the level of the beasts,” they would
still have a choice between moral mediocrity and moving “in the
direction of divinity,” and so a genuine choice of destiny would
remain open to them.
The Importance of Inwardness
We saw in the previous chapter that the “free will” arguments

offered by Swinburne and Hick fail to provide a rebuttal for the
argument of Part I. Evidence sufficient for belief would not shatter
human autonomy or remove a genuine choice of destiny. In this
chapter I consider arguments suggested by the writings of Pascal
and Kierkegaard—arguments whose common claim is that a cer¬
tain valuable sort of inward or subjective orientation (valuable be¬
cause essential to the proper moral and religious development of
human beings) requires for its existence that God’s existence be
obscure.1 As in the preceeding chapter, the question that will guide
our discussion is whether the arguments considered are successful
when adapted for our purposes, that is, when treated as claiming that
it is plausible to suppose that the good in question is such as a
loving God would seek to bring about even at the cost of permit¬
ting the occurrence of reasonable nonbelief, and that its existence is
plausibly viewed as incompatible with a strong epistemic situation
in relation to theism. My conclusion is that the arguments that can
be drawn from the writings of Pascal and Kierkegaard are not suc¬
cessful, when so construed.
Any investigation of our question would be derelict in its duty if
I. The “common claim” must, of course, be stated loosely to allow for differ¬
ent kinds and degrees of emphasis.
it did not consider the famous discussions of Kierkegaard and Pas¬

cal. But for the very reason of their fame, it may seem strange that
my treatment of these discussions does not precede my treatment of
Swinburne and Hick. However, as I pointed out at the beginning
of Part 2, the structure of this second part of the book is dictated
not by historical but by theoretical considerations. As it happens,
some of the arguments considered here bear interesting relations to
arguments discussed in the previous chapter. It will therefore be
convenient to have the results of that chapter’s discussion before us
as we proceed.
Reasons for Divine Hiddenness in Pascal
To understand Pascal’s doctrine of the Hidden God we must first

recognize that on his view, God was not always hidden. There was
a time, so his Pensees tells us, when the existence of God was evi¬
dent to all: “I created man holy, innocent, perfect, I filled him with
light and understanding, I showed him my glory and my
wondrous works. Man’s eye then beheld the majesty of God. He
was not then in the darkness that now blinds his sight.”2 But man
withdrew from God, and God’s hiddenness was the eventual re¬
sult. As the Pensees has it:
He could not bear such great glory without falling into presump¬
tion. He wanted to make himself his own centre and do without my
help. He withdrew from my rule, setting himself up as my equal in
his desire to find happiness in himself, and I abandoned him to him¬
self. The creatures who were subject to him I incited to revolt and
made his enemies, so that today man has become like the beasts, and
is so far apart from me that a barely glimmering idea of his author
alone remains of all his dead or flickering knowledge. [Fragment
149, p. 77]
Pascal, then, accepts the historicity of the Fall. But there are two
aspects of his view which, for our purposes, need to be made more
2. Blaise Pascal, Pensees, A. J. Krailsheimer, trans. (Harmondsworth: Penguin,

1966), fragment 149, p. 77. Further references to this work, in both text and notes,
will be made parenthetically.
The Importance of Inwardness [i33
explicit. First, humans did not initially fall away from the knowl¬
edge of God; only from a proper relation to it. On Pascal’s view, a
consciousness of God was at first retained; the Fall consisted in an
attempt to become equal with God. But the sequence of events that
followed took human beings ever farther away from God, until
today only a vestige of the original knowledge remains. Second, the
claim “God is hidden by sin,” where this is interpreted as “the
apparent weakness of theistic evidence is simply a matter of human
blindness,” would be criticized by Pascal as incomplete, for on his
view God himself contributed to the sequence of events just men¬
tioned by “inciting” the creatures to revolt: the harmony of the
created order (and, ipso facto, the evidence of its origin in God)
was intentionally weakened.3 Of course, humans have made things
worse for themselves by not responding appropriately to the signs
of God that remain in nature, and so there is a sense in which God’s
continuing hiddenness may be said to be (in part) the result of hu¬
man blindness; but it should be recognized that on Pascal’s view,
the relative weakness of the evidence provided by the created order
is fundamentally a function not of our blindness but of God’s in¬
tentional withdrawal.
The theme of Divine hiddenness can be traced not only in pas¬
sages of the Pensees which refer to the Fall but also in those dealing
with the Incarnation. In agreement with the Christian tradition,
Pascal held that God in the fullness of time responded to the Fall by
opening a “way of salvation” (fragment 149, p. 79). But he writes
that God’s wish to redeem humankind did not lead him to give a
clear revelation of himself—to attempt in this way an immediate
return to the pre-Fail state of affairs. God did make contact with
human beings, to be sure; but, in accordance with the policy of
restraint suggested by his immediate response to the Fall and its
repercussions on the created order, he came mildly—in a hidden
way:
If he had wished to overcome the obstinacy of the most hardened,

he could have done so by revealing himself to them so plainly that
3. Hence Pascal writes, “All creatures either distress or tempt him [i.e., man]
and dominate him either by forcibly subduing him or charming him with sweet¬
ness, which is a far more terrible and harmful yoke” (ibid.).
they could not doubt the truth of his essence. . . . This is not the
way he wished to appear when he came in mildness, because so
many men had shown themselves unworthy of his clemency, that
he wished to deprive them of the good they did not desire. It was
therefore not right that he should appear in a manner manifestly
divine and absolutely capable of convincing all men, but neither was
it right that his coming should be so hidden that he could not be
recognized by those who sincerely sought him. He wished to make
himself perfectly recognizable to them. Thus wishing to appear
openly to those who seek him with all their heart and hidden from
those who shun him with all their heart, he has qualified our knowl¬
edge of him by giving signs which can be seen by those who seek
him and not by those who do not.
There is enough light for those who desire only to see, and
enough darkness for those of a contrary disposition. [Fragment 149,
pp. 79-8o]
So God is hidden. Neither nature nor history proclaims him as

loudly as it might, and in each case this is not in the first instance a
matter of human blindness, but of God’s intentional withdrawal or
restraint.4 Now we must ask, why is this the case? Why, according
to Pascal, does God hide himself?
As we saw in the previous chapter, John Hick cites Pascal in
support of his own view that human beings are required to seek
God so that their moral freedom may be preserved. But it seems that
Hick’s interpretation of Pascal must be mistaken. For, as the sec¬
ond passage quoted above clearly indicates, Pascal accepts the tra¬
ditional view that human beings were morally free (in Hick’s
sense) even in the direct presence of God, and that they exercised this
freedom by rejecting God’s rule. Now Pascal does say (in the last
passage quoted) that a startlingly clear revelation would “overcome
the obstinacy of the most hardened,” but the obstinacy referred to
seems to be obstinacy with respect to the admission of God’s exis¬
tence. In other words, what Pascal appears to be saying is that God
could render us cognitively unfree. Given his earlier claims, how¬
ever, it seems that he, unlike Hick, does not consider there to be a
4. As Pascal puts it, “all things combine to establish the point that God does
not manifest himself to men as obviously as he might” (ibid., fragment 449, p.
168).
The Importance of Inwardness [135
necessary connection between cognitive and moral freedom in this

context. Hick’s repeated appeal to Pascal is therefore misleading at
best.5
So why does God hide himself? Another possible Pascalian an¬
swer, suggested by the last passage quoted above, is the Just De¬
serts Argument: “Human beings have shown themselves unworthy
and so God has deprived them of the good of his presence, leaving
them to their folly.6 7 Divine hiddenness is no more than they de-
serve.
But surely there must be more to Pascal’s doctrine of the Hidden
God than this. For the response this view attributes to God is all
too human. It is, indeed, one that contravenes the Christian ethic,
as Pascal seems himself to have recognized:
As for those who live without either knowing or seeking him, they
consider it so little worth while to take trouble over themselves that
they are not worth other people’s trouble, and it takes all the charity
of that religion they despise not to despise them to the point of
abandoning them to their folly. But as this religion obliges us al¬
ways to regard them, as long as they live, as beings capable of grace
which may enlighten them . . . we must do for them what we
would wish to be done for us in their place, and appeal to them to
have pity on themselves, and to take at least a few steps in an at¬
tempt to find some light. [Pensees, fragment 427, pp. 160-161]
If the Christian attitude of love should prompt Christians to a

more charitable response, what must we say about God, of whose
perfect love our own best love is but a dim reflection?
5. In any event, I have provided reasons (in Chapter 5) for rejecting Hick’s
Autonomy Argument and so will not need to consider it any further here.
6. As another fragment puts it, “Religion is so great a thing that it is right that
those who will not take the trouble to look for it, if it is obscure, should be
deprived of it” (Pensees, fragment 472, p. 180).
7. It might seem that a less strident and more persuasive version of this argu¬
ment is also possible: “Humans have freely chosen to reject God, and it would be
inappropriate for God to overrule this decision. So they have been left to them¬
selves. God respects their freedom.” But this argument suggests no reason for God
to intentionally withdraw, to hide himself. It may be good for God to leave us in
our blindness if this results from free choices, but why should he make things even
more obscure? To this question, free will arguments do not seem to have an an¬
swer.
There are other reasons as well for supposing that Pascal may
have had more in mind than is suggested by the Just Deserts Argu¬
ment. As Terence Penelhum points out (and as, more generally,
our discussion in the preceeding chapter should lead us to con¬
clude), God can be hidden from human eyes even when, objec¬
tively, the signs of his presence are clear:
If God hides himself in the face of human corruption he may hide

himself from human beings who are in the presence of probative
phenomena. . . . Their corruption might hinder them from heeding
the presence of these phenomena. It might hinder them from draw¬
ing from these phenomena the conclusions they proved. And it
might affect some of them so that they drew the conclusions but did
so with such reluctance that they did not submit themselves to God
and allow their knowledge of him to change their lives. In all these
cases they would be responsible for God’s remaining hidden from
them, for they would be keeping faith at arm’s length.8
As against the Just Deserts Argument, this point shows that if hu¬
mans are indeed corrupt, God would not necessarily get any better
revenge by hiding himself than by manifesting himself clearly. Of
course, perhaps in a situation of the sort Pascal describes, in which
God’s existence is overwhelmingly manifested, it would be more dif¬
ficult for us to keep God “at arm’s length” (although, as his discus¬
sion of the Fall suggests, it may be that humans could eventually
fall away from God even then). But why should such a situation be
thought to be the only alternative to hiddenness? For these reasons,
I conclude both that the Just Deserts Argument is inadequate and
that Pascal, brilliant and charitable thinker that he was, may well
have had some other explanation of hiddenness in mind. 9 But what
might this explanation be? It seems to me that an answer to this
question is latent in the following passages from the Pensees:
If there were no obscurity man would not feel his corruption: if

there were no light man could not hope for a cure. Thus it is not
8. Terence Penelhum, God and Skepticism (Dordrecht: Reidel, 1983), p. 109.

9. It will, in any event, be useful for our purposes to assume that he did, in
order to see whether there are any arguments lurking in this neighborhood that
might serve to rebut the prima facie case we are considering.
The Importance of Inwardness [ 137
only right but useful for us that God should be partly concealed and
partly revealed, since it is equally dangerous for man to know God
without knowing his own wretchedness as to know his wretched¬
ness without knowing God. [Fragment 446, p. 167]
[Christianity] teaches men then these two truths alike: that there is a
God, of whom men are capable, and that there is a corruption in
nature which makes them unworthy. It is of equal importance for
men to know each of these points. . . . Knowing only one of these
points leads either to the arrogance of the philosophers, who have
known God but not known their own wretchedness, or to the de¬
spair of the atheists, who know their own wretchedness without
knowing their Redeemer. [Fragment 449, p. 168]
“If I had seen a miracle,” they say, “I should be converted.” . . .

They imagine that such a conversion consists in a worship of God
conducted, as they picture it, like some exchange or conversation.
True conversion consists in self-annihilation before the universal
being whom we have so often vexed. [Fragment 378, p. 137]
He [i.e., man] must not see nothing at all, nor must he see enough
to think that he possesses God, but he must see enough to know
that he has lost him. [Fragment 449, p. 170]
God wishes to move the will rather than the mind. Perfect clarity
would help the mind and harm the will.
Humble their pride. [Fragment 234, p. 101]
One needs no great sublimity of soul to realize that in this life there
is no true and solid satisfaction, that all our pleasures are mere van¬
ity, that our afflictions are infinite, and finally that death which
threatens us at every moment must in a few years infallibly face us
with the inescapable and appalling alternative of being annihilated or
wretched throughout eternity. . . .
Let us ponder these things, and then say whether it is not beyond
doubt that the only good thing in this life is the hope of another life,
that we become happy only as we come nearer to it. . . .
It is therefore certainly a great evil to have . . . doubts, but it is at
least an indispensable obligation to seek when one does thus doubt;
so the doubter who does not seek is at the same time very unhappy
and very wrong. [Fragment 427, p. 157]
The view of Divine hiddenness suggested by these passages

(when interpreted in the light of Pascal’s remarks on the Fall and
on the Incarnation) is the following: “The hiddenness of God has

both a positive and a negative function, acting both as a stimulus
and as a restraint. Negatively, it prevents us from responding to
God in external, presumptuous ways, by effectively removing the
possibility of (easily acquired) knowledge of God. This restraint
was first imposed in the beginning, when, in order to put an end to
human attempts to become equal with God, and to begin to rees¬
tablish the relationship on another footing, God withdrew from
humankind and weakened the testimony of nature to his existence,
introducing death and misery into the world. But it is good for us
that it remain in effect, so that we are prevented from frustrating
God’s program of rehabilitation and restoration through yet another
inappropriate response. The danger of such a response is clear.
Witness the arrogance of philosophers who consider themselves to
have found proofs of God’s existence. They not only use such
proofs to advance their own selfishly conceived intellectual ends,
but have, in constructing them, failed to show proper respect for
the great difference between God and humans. This suggests that,
were true intellectual clarity to be restored, we would simply dem¬
onstrate once more our great capacity for presumption, ignoring
the implications for behavior of our creaturely status and God’s
holiness, and, ipso facto, failing to acquire the very attitudes of
contrition and humility required for the restoration of relationship
with God. We would be led away from a proper relationship with
God, rather than toward it. Now, of course, as has already been
suggested, God wishes not only to prevent us from responding
inappropriately to the knowledge of his existence (this would be
compatible with never revealing himself), but also to provide op¬
portunities for the acquisition of an appropriate self-conception and
humility, so that a revelation may at some point be given and
properly received, and a relationship with God restored. The hid¬
denness of God helps to promote this and so has not only a nega¬
tive but also a positive function. By withdrawing, God hopes to
awaken us to the wretchedness of life on our own, to stimulate a
recognition of the barrenness of our existence in a corrupt God¬
forsaken environment, and to prompt us to search for God with
due contrition and humility. We must be made to feel our corrup¬
tion. God’s hiddenness can help us to feel this. By doing so, it also
helps us to admit our weakness and limitations, and the impos¬

sibility of earthly fulfillment. In short, it (paradoxically) points us
Godward. Those who listen to the message and seek for God will
not be disappointed, for God has left signs of his presence (in par¬
ticular, of his presence in Christ as Mediator) which they will find.
Of course, in accordance with God’s general policy of restraint,
these signs have been hidden. But this does not diminish their
usefulness. Indeed, their usefulness is thereby increased. By pro¬
viding hidden signs, God ensures both that those who do seek will
find and that those who find will respond appropriately (i.e., in¬
wardly,10 with humility) and so enter into the relationship which
was his wish for them from the beginning.”11
If this interpretation of Pascal is correct, he has at least two addi¬
tional answers for those who ask why God hides himself. The first,
which I will call the Presumption Argument, claims that if God
were not hidden, humans would relate to God and to their knowl¬
edge of God in arrogant and presumptuous ways, and that the pos¬
sibility of developing the inner attitudes essential to a proper rela¬
tionship with God would ipso facto be ruled out. The second,
which I will call the Stimulus Argument, claims that God’s hidden¬
ness is required because it prompts human beings to recognize their
true condition and thus helps to bring about the necessary inward¬
ness, clearing the way for a revelation of God and a restored rela¬
tionship with God. Both arguments allow Pascal to say that “it is
10. Pascal writes: “a purely intellectual religion would be . . . appropriate to the

clever, but it would be no good for the people. The Christian religion . . . exalts
the people inwardly, and humbles the proud outwardly” (Pensees, fragment 219, p.
99)-
11. It may seem to some that there is an inconsistency in Pascal’s view, as I have
presented it. If human beings are as eager as he suggests to exploit the knowledge
of God, will they not avidly seek for it? But Pascal has at least two answers to this.
First, he can say that although humans would not wish to be rid of the knowledge
of God were it handed to them, they are too caught up in selfish pursuits to seek it
for themselves. Second, and more important, he can say that many of the signs
God has left of himself are in the Church and its scriptures and (because of the
nature of the Church’s message) are not only indications of God’s presence but at
the same time indictments of our wretchedness. We do not wish to be told of our
wretchedness and so, unless we have come to acknowledge it first, are not likely to
recognize or pursue these signs (even though we may play with theoretical argu¬
ments). God, that is, in hiding the signs, has ensured that they are of such a sort as
only individuals with a certain inward orientation will take to be signs.
40 1 The Force of the Argument
not only right but useful for us that God should be partly concealed
and partly revealed” (my emphasis), and so take him beyond the
Just Deserts Argument.12
One of the virtues of the Presumption Argument, in particular,
is that it provides Pascal with a response to arguments such as Pe-
nelhum’s. Penelhum argues, as we saw, that God might very well
remain hidden from corrupt human beings in a situation of good
evidence, for, desiring to escape the knowledge of God, such indi¬
viduals would likely engage in various sorts of self-deception and
thus keep God at “arms length.” Pascal’s argument responds to
such claims by questioning the assumption on which they depend,
namely, that human beings would wish to escape the knowledge of
God. On his view, corrupt humans would not wish to be rid of
God, but rather would wish to compare themselves with him and
compete with him, or use him for their own selfish purposes. If
this view is correct, a situation of good evidence is not compatible
with Divine hiddenness. Thus Penelhum’s argument, although it
may point up a weakness in the Just Deserts Argument when that
argument is viewed in isolation, does not do so when it is taken
together with the Presumption Argument; nor does it have any
force against the Presumption Argument considered independ¬
ently.13
It must now be noted, however, that even if the Stimulus Argu¬
ment and the Presumption Argument, as so far developed, were to
be judged sound, no answer to our problem would be forthcom¬
ing. All these arguments suggest is that God has a reason for with¬
holding good evidence from those humans whose present actions
and motives are such as to prevent them from responding to it
appropriately. No reason is suggested for withholding evidence
from those who do not fall into this category—from those, for
example, who have felt their corruption and the emptiness of life
without God and who have begun to search for God with proper
motives. No reason is suggested, in short, for supposing that God
would permit the occurrence of reasonable nonbelief.
12. For a recent discussion of Pascal that seems to follow him here, see Thomas
V. Morris, “The Hidden God,” Philosophical Topics, 16 (1988), 5-21.
13. And this is of course how we must consider the Presumption Argument,
having rejected the Just Deserts Argument not only because of the force of Penel¬
hum’s criticism but also because of the implications of Christian charity.
Now it may be said that I am burdening Pascal with a problem

which he did not consider to exist, that he considered it to be the
case that all nonbelief is culpable and that those who recognize
their wretchedness and respond with humility receive immediate
satisfaction (i.e., find that God “appears openly to them”). This
claim is mistaken, however. Pascal recognized, for example, that
nonbelievers may search diligently but without success, as the fol¬
lowing passages from the Pensees show:
Amongst those who are not convinced, I make an absolute distinc¬

tion between those who strive with all their might to learn and
those who live without troubling themselves or thinking about it.
I can feel nothing but compassion for those who sincerely lament
their doubt, who regard it as the ultimate misfortune, and who,
sparing no effort to escape from it, make their search their principal
and most serious business. [Fragment 427, p. 156]
There are only two classes of people who can be called reasonable:
those who serve God with all their heart because they know him
and those who seek him with all their heart because they do not
know him. [Fragment 427, p. 160]
This is what I see and what troubles me. I look around in every
direction and all I see is darkness. Nature has nothing to offer me
that does not give rise to doubt and anxiety. If I saw no sign there of
a Divinity I should decide on a negative solution: if I saw signs of a
Creator everywhere I should peacefully settle down in the faith.
But, seeing too much to deny and not enough to affirm, I am in a
pitiful state, where I have wished a hundred times over that, if there
is a God supporting nature, she should unequivocally proclaim him,
and that, if the signs in nature are deceptive, they should be com¬
pletely erased; that nature should say all or nothing so that I could
see what course I ought to follow. Instead of that, in the state in
which I am, not knowing what I am or what I ought to do, I know
neither my condition nor my duty. My whole heart strains to know
what the true good is in order to pursue it: no price would be too
high to pay for eternity. [Fragment 429, pp. 162-163]14
14. It is hard to know what to make of this passage. If it is to be taken as

autobiographical, it certainly contrasts sharply with other fragments in which Pas¬
cal seems quite confident of God’s existence. Perhaps it represents a prior state of
his self. Perhaps, on the other hand, it is meant to represent the proper occupation of
Because of such passages as these, it seems we must say that Pascal

accepted the existence of the problem I have mentioned, however
implicitly, and that he either had no solution to offer or had some
solution other than the ones we have hitherto considered or held
one or another or both of the arguments just distinguished—
namely, the Stimulus and Presumption Arguments—to be capa¬
ble, suitably revised and extended, of accommodating it. These
would appear to be the interpretive options. The first option may
well be correct, but it is the least interesting, and so I will not
consider it further. The second may be right, but I do not think so:
I can find no independent, supplementary explanation in Pascal.
We are therefore left with the third option. In pursuing it, I will
indulge in maneuvers that may seem at times to find little warrant
in the text. But it must be remembered that I am, in the final
analysis, more interested in possible arguments (in what Pascal could
have said) than in determining exactly what the historical Pascal
meant to say.
That Pascal considered a period of diligent searching to be a con¬
dition of revelation is indicated by a phrase that occurs several
times in the Pensees. Here is one example: “God has appointed visi¬
ble signs in the Church . . . [and] has . . . hidden them in such a
way that he will only be perceived by those who seek him with all
their heart” (fragment 427, p. 155, my emphasis; cf. fragment 149,
p. 80). What may, however, strike us about Pascal’s use of the
phrase “with all their heart’’ is that given what he says about rea¬
sonable doubters in the passages quoted above, namely, that they
have sought with all their heart, even if his claim were to be ac¬
cepted, he would be faced with essentially the same problem as
that referred to earlier—namely, the problem of explaining why
individuals who have met the standard set by God may nonetheless
fail to meet God. The claim that God will only reveal himself to
those who seek with all their heart may seem, therefore, to be
unhelpful (assuming Pascal is right about reasonable doubters),
even if justified.
the nonbeliever, namely, seeking, and so to contrast with indifference, which Pas¬
cal calls “monstrous” (fragment 427, p. 157). Whatever the case, it strongly sug¬
gests that Pascal considered “seeking without finding” to be part of the human
condition (although note the absence of any reference to the Christian revelation).
It seems to me, however, that Pascal need not be faced with this
problem. For the phrase “with all their heart” is ambiguous. The
purity of desire, of motive, of intention (and perhaps depth of convic¬
tion on matters such as one’s wretchedness) to which it seems to
refer can be seen as instantiated in one who, upon recognizing her
unworthiness, is filled with remorse and, perhaps in prayer, ex¬
presses a desire to be properly related to God, even if she is subse¬
quently deterred from pursuing her search by pride or selfish de¬
sires; or as instantiated only in one who, through long, persistent,
unremitting searching, has a deeply ingrained attitude of humility
and desire for the life of faith, one who is no longer at all likely to
be deterred by lesser motives from pursuing a (proper) relationship
with God. If the phrase “with all their heart” is indeed ambiguous
in this way, Pascal can claim that the reasonable doubters to whom
he refers do not necessarily seek for God “with all their heart” in
the second, stronger sense of that phrase, and that it is this second
sense that is intended when it is said that God will reveal himself
clearly to those who seek with all their heart.
Whether Pascal saw this distinction or not, it is clearly one he
could employ. That he would be inclined to employ it (more specif¬
ically, that when he says God will be revealed to those who seek
with all their heart, he has in mind an attitude developed and ex¬
pressed in the course of a long and toilsome search) is suggested by
the following passage from the Pensees, in which it is said that ob¬
scurity can only provide the basis for an objection to Christian
theism if the one who complains of it has engaged in a very long
and very thorough search:
The obscurity in which they [i.e., religious skeptics] find them¬

selves, and which they use as an objection against the Church, sim¬
ply establishes one of the things the Church maintains [viz., that
God has hidden himself so only those who seek with all their heart
will find him] without affecting the other [viz., that those who do
genuinely seek will be satisfied], and far from proving his teaching
false, confirms it.
In order really to attack the truth they would have to protest that
they had made every effort to seek it everywhere, even in what the
Church offers by way of instruction, but without any satisfaction. IJ
they talked like that they would indeed be attacking one of Chris¬
tianity’s claims. [Fragment 427, p. 155, my emphases]
But even if we assume that Pascal does not necessarily face the
problem in question, and that he considers a long and thorough
search and a deeply ingrained attitude of humility to be necessary
for a clear revelation of God, we must still ask how this latter claim is
to be justified. It is only if we can find an argument for it that we will
have located the Pascalian answer to the problem of reasonable
nonbelief.
As I suggested earlier, such an argument, if it exists, is likely to
involve an extension of the Stimulus Argument or the Presump¬
tion Argument, or both. Do either of these arguments admit of
extension? It seems to me that, quite obviously, both do. For Pas¬
cal can claim the following: “The knowledge of God is a gift that
must be carefully given. Humans who have begun to search for
God (and, indeed, those who have honestly searched for some
time) may still be tempted, by virtue of their sinful nature, to fall
back into older and less worthy patterns of behavior; in particular,
they might very well relate externally and in a relationship-inhibit¬
ing way to any good evidence they were given. It is therefore im¬
portant that God remain hidden until they have persevered over a
long period and have developed a deeply ingrained attitude of humil¬
ity and a desire for a proper relationship with God strong enough
to overwhelm selfish desires. The hiddenness of God itself helps to
bring this development about by continually bringing home to hu¬
mans the poverty of a life without God and by continually
prompting them to seek God humbly. God’s hiddenness can there¬
fore be understood in terms of patience. Although he wishes to be
more deeply and intimately related to human beings (and, indeed,
because he wishes this), God is willing to wait until good evidence
will be religiously efficacious. It is only if he is patient that God can
ensure that those who come to be aware of his presence are those
willing to submit in the right way to what they see.”15
This, then, is the explanation of reasonable nonbelief to which
the arguments in the Pensees most naturally lead. Having got it out
15. One of the virtues of this interpretation is that it is able to make good sense
of Pascal’s repeated claim that in order to acquire a proper belief in God humans
need first to diminish their passions, and that to diminish passion we must resort to
habit (see fragments 119, p. 60; 125, p. 61; 418, p. 152; and 821, p. 274).
into the open, we are now in a position to see whether it stands up

to critical scrutiny. This will be my concern in the next section.
The Pascalian Solution: An Assessment
Before we move to a critical discussion of Pascal, two points

should be noted. First, in the context of our discussion, the notions
of Divine hiddenness and disclosure must be understood in terms
of the notion of evidence sufficient for belief: God is disclosed if he
has made such evidence available and hidden if he has not. Hence,
for the Pascalian explanation to succeed in this context, it must
show that it is plausible to suppose that a propensity for presump¬
tion is the reason why evidence sufficient for belief is not generally
available. This requirement makes the Pascalian’s task more diffi¬
cult than it would otherwise be. For claims about the dangers of
Divine disclosure (as we saw in the previous chapter) have much
more force when the notion of disclosure is given a very strong
interpretation (when, e.g., it is understood to involve reference to
an overwhelming manifestation of glory); and the interpretation to
which we are committed by the terms of the problem of reason¬
able nonbelief, as I have developed it, is clearly not of this sort.
Second, although it may seem that I have in describing the Pas¬
calian’s task just now neglected the role of the extended Stimulus
Argument, this argument is quite obviously a weak argument and
so may be dispatched here at the outset. In order for it to succeed
as an independent reason for supposing that God would remain
hidden (in the relevant sense), it must be plausible to suppose that
Divine hiddenness is necessary for humans to be stimulated to a
continuing awareness of their wretchedness. But this is not the
case. Indeed, a Divine revelation might do the job quite well. Now
of course, if the extended Presumption Argument is correct, such
direct methods might well be ruled out; and so we might be led to
conclude that the extended Stimulus Argument is at worst depen¬
dent on the extended Presumption Argument. But even this would
be too strong a conclusion. For it does not follow from the claim
that the extended Presumption Argument is correct that God must
hide in order to stimulate the awareness in question. Even if that
argument is correct, it might be sufficient for such an awareness
that, for example, much evil exist—in other words, there remain
indirect methods of stimulating such an awareness which do not
require Divine hiddenness.16 I conclude therefore that the extended
Stimulus Argument does not succeed and that the success of the
Pascalian solution depends on the extended Presumption Argu¬
ment. For this reason, I will be focusing on the latter.
Pascal’s claim about the likelihood of a presumptuous response
to clear evidence of God’s existence may seem stronger than it oth¬
erwise would if we assume, as he does, that a historical “Fall” actu¬
ally took place. If we assume this, we are in effect assuming that
there is inductive evidence for the claim that, were God’s existence
to be revealed today, a presumptuous response would follow. But
in light of the findings of disciplines like evolutionary biology and
biblical criticism, it is hard to see how such an assumption could be
successfully defended. Thus it would seem that we do not have the
inductive evidence in question and, as a result, that Pascal’s argu¬
ment is deprived of (possibly) vital support from the very begin¬
ning.
But there is more that can be said on this score. For the pre-Fall
state of affairs—a state of absolute intellectual clarity and without
misery or death—seems to be, for Pascal, the paradigm of a situa¬
tion in which God is not hidden.17 Consequently (we may sur¬
mise), when he refers to Divine disclosure, he is inclined to think of,
inter alia, a world without evil, and the disastrous response o f the first
humans to life in such a world. This interpretation, whatever its other
merits, certainly helps explain why Pascal considered a presump¬
tuous response to Divine disclosure to be so likely. For it makes
some sense to suppose that if we knew of God’s existence and were
untroubled by evils which in the actual world bring home to us
our limitations, we might develop an inflated self-conception and
begin to respond to God in inappropriate ways. However, espe-
16. As I suggest immediately below, Pascal may assume that there is a necessary
connection between evil and hiddenness. But I will argue that this view is incor¬
rect.
17. This is perhaps to be expected since, as we have seen, God’s hiddenness is,
on his view, bound up with the corruption of nature and the introduction of mis¬
ery and death. It is only natural to think of the removal of hiddenness as involving
the removal of those phenomena that brought it about in the first place (cf. Pensees,
fragment 149, pp. 79-80; fragment 449, p. 170).
dally without the classical doctrine of the Fall, Divine hiddenness

and disclosure need not be thought of in terms of glory lost and
glory regained. On our definition, for example, hiddenness and
disclosure involve considerably less than this. In particular, on our
definition it is not necessary that the evils of the world be removed
in order for God to be disclosed. Therefore, if the notion of a pre¬
sumptuous response to Divine disclosure is, for Pascal, bound up
with the classical picture of the Fall, as I have indicated it might be,
this can only lessen the force of any responses to the argument of
this book that are based upon it.
It seems, however, that on Pascal’s view, there is other inductive
evidence as well, namely, the arrogance of philosophers who take
themselves to have found proofs of God’s existence. And so the
Pascalian might argue in response as follows: “The notion of prov¬
ing God’s existence seems perfectly compatible with living in a
world filled with evil and suffering. Therefore Pascal’s notion of a
presumptuous response is capable of being detached from the notion
of a restored pre-Fail glory. Even if we reject the notion of a Fall,
and think of hiddenness and disclosure in terms not connected with
it, the problem of presumption remains.”
But even if we ignore the fact that to disclose himself (in the
relevant sense) God need only provide evidence sufficient for be¬
lief, and set aside the possibility that he could provide nontheoretical
evidence, we must say that this claim is only weakly supported by
the consideration it cites. As Penelhum writes, referring to the cor¬
rupt motives Pascal ascribes to philosophers, “one must surely say
that sometimes the attempt to prove the existence of God has not
been undertaken from such motives as these. When Pascal thinks
about dogmatic philosophers and their attempts at proof, he thinks
too readily of Descartes.”18 If now we recall the fact that God need
only provide evidence sufficient for belief, and reintroduce the possi¬
bility of nontheoretical evidence—if, for example, we think in
terms of experiential evidence of the sort described in Chapter 2—
then the support provided for the Pascalian’s claim by the arro¬
gance of philosophers becomes negligible.
Four points may be noted here. First, the notion of proof carries
18. Penelhum, God and Skepticism, p. 95.

48I The Force of the Argument
with it an air of finality, whereas that of evidence sufficient for

belief does not (evidence of the latter sort may fall well short of
proof). Those who construct a proof therefore always have more
reason to boast than those who find themselves with evidence suf¬
ficient for belief, and hence are more likely to become arrogant.
Second, religious experiences of the sort in question, unlike theo¬
retical proofs, are commonly viewed as coming from “outside”;
the subject of a theistic religious experience does not perceive her¬
self to be in control of her situation. Hence experiential evidence
would be less likely to inspire a sense of achievement in the one to
whom it was made available. Since the sense of significant achieve¬
ment and control is often involved in arrogance, this suggests that
recipients of experiential evidence would not be as likely as others
to respond arrogantly (especially if those near the experients were
clearly also receiving experiential evidence). Third, religious expe¬
rience has its own distinctive psychological effects, and arrogance
is not very naturally construed as one of them. Feelings of grati¬
tude, joy, reassurance, astonishment, guilt, or dismay seem more
likely. This third point is of course closely related to the second in
that the feelings mentioned are ones we would naturally associate
with a communication from a higher, moral source, whereas arro¬
gance is more likely where there is no sense that one is the recipient
of such a communication. Finally, and most important, part of
what God might communicate to us through religious experience
is the very message of wretchedness and corruption that Pascal suggests a
Divine disclosure would inhibit. Religious experiences, it can be
argued, are not all likely to provoke an arrogant response, inas¬
much as they would awaken in us a sense of our wretchedness and
corruption (a state incompatible with arrogance). This fourth point
is made all the more interesting by the fact that Pascal, writing his
Pensees, seems at times to have recognized it:
The God of the Christians is a God of love and consolation: he is a

God who fills the soul and heart of those whom he possesses: he is a
God who makes them inwardly aware of their wretchedness and his
infinite mercy: who unites himself with them in the depths of their
soul: who fills it with humility, joy, confidence and love: who
makes them incapable of having any other end but him. [Fragment
449, p. 169]
The Christians’ God is a God who makes the soul aware that he is
its sole good: that in him alone can it find peace; that only in loving
him can it find joy; and who at the same time fills it with loathing
for the obstacles which hold it back and prevent it from loving God
with all its might. Self-love and concupiscence, which hold it back,
are intolerable. This God makes the soul aware of this underlying
self-love which is destroying it and which he alone can cure. [Frag¬
ment 460, p. 178]
What should we make of this? Why did Pascal not see these
claims as providing a counterexample to the Presumption Argu¬
ment? Why were they not seen as suggesting an alternative to theo¬
retical proof (and, for that matter, to the restoration of primeval
glory) and, hence, as rendering otiose the appeal to philosophical
arrogance? Whatever the reason may actually have been, it seems
clear that if religious experience can have the effects Pascal de¬
scribes, it could, if given, rule out a presumptuous response; and,
therefore, that God’s disclosure (in the relevant sense) need not at
all provoke such a response. That it can have these effects is
strongly suggested not only by Pascal’s own reports but also by
the testimony of many believers to experiences apparently of God
which brought home to them their corruption—experiences which
were unexpected and unsought, but which changed the direction
of their life and (on their interpretation) led to a deep and fulfilling
relationship with God. The critic of Pascal therefore has strong
inductive evidence of his own, evidence which seems to render
irrelevant the evidence adduced by the Pascalian in support of his
claim and, indeed, to positively establish the contrary claim.
I would suggest, therefore, that in focusing on the effects of
proofs on philosophers, Pascal has failed to take into account other
possible sources of evidence, in particular, the evidence of religious
experience, which, so far from provoking an arrogant response,
seems likely to produce its opposite. We could also vary the
emphasis and say that he is wrong to focus on the effects of any
evidence whatever on philosophers. If God were going to make
evidence sufficient for belief generally available, he would not
likely provide evidence of the sort that (sometimes) makes philoso¬
phers arrogant, namely, theoretical proofs, but rather evidence as¬
similable by anyone, no matter how sophisticated or lacking in
sophistication. If religious experience fits this description, as it

seems to do, and if God could through religious experience make
us aware of our wretchedness and move us in the direction of hu¬
mility, then, again, the Pascalian solution seems to lack plau¬
sibility.19
How might the Pascalian respond to these claims? Could he ar¬
gue that even religious experience would, at least in some cases,
provoke a presumptuous response if given too early, and that it
must therefore (in these cases at least) be postponed until individ¬
uals have come a long way on their own? Can he say that even
when the evidence is experiential, the threat of presumption is such
that God must be patient, waiting until evidence will be religiously
efficacious?
When I think about these possible rejoinders, my first inclination
is to reiterate points already made and to suggest that evidence
could always be provided in such a way as to avoid a presumptuous
response. Surely an experience apparently of God—a loving, holy
God—would not inspire such a response in anyone. But let us sup¬
pose for the sake of argument that the Pascalian’s claim is plausible.
Does it follow that the extended Presumption Argument provides
the required rebuttal? As it seem to me, it does not. For we may
still reject a claim which must be plausible if it is to succeed as
such, a claim which I have hitherto assumed to be correct, namely,
that a loving God would indeed withdraw to the extent in question
if this were necessary to prevent a presumptuous response. Even if
we accept the former claim, we may show the unacceptability of
the Pascalian explanation as a rebuttal by providing reasons for re¬
jecting the latter, and this I propose to do.
But first we must be clear about what it is we are rejecting. If it
could be shown that, given evidence sufficient for belief, at least
some human beings would be forced to respond presumptuously,
with no hope of ever being able to do otherwise (unless God again
became hidden), then I would agree that a loving God, wishing to
establish a proper relationship with human beings, would remain
19. Pascal himself offers support for the antecedent of this claim: “Do not be
astonished to see simple people believing without argument. God makes them love
him and hate themselves. He inclines their hearts to believe” (Petisees, fragment
380, p. 138).
The Importance of Inwardness [ i5 i
withdrawn. But I hope it will be obvious that this is not the case.2'1
Given all the considerations I have adduced, it seems unreasonable
to suppose that in the circumstances in question, anyone would
ever have more than an inclination to respond arrogantly or pre¬
sumptuously—an inclination, moreover, capable of being overrid¬
den. In other words, while some might be tempted to respond
inappropriately, everyone would be free not to. The most, there¬
fore, that a defender of Pascal could ever plausibly claim is that,
given good evidence, some might be inclined to respond presump¬
tuously, and it is the claim that God would remain withdrawn to
prevent humans from yielding to such an inclination that I reject.
My reasons for rejecting this claim are drawn from the discus¬
sion in Chapter 5. There I assumed, with Swinburne, that a loving
God would wish to give us a choice with respect to our own des¬
tiny, and that to do this, he must, inter alia, give us a range of
natural inclinations to act badly which it is up to us to yield to or
to suppress. He must leave us morally free. Now if God remained
withdrawn in the face of a possible presumptuous response to
theistic evidence, he would in effect be preventing us from choos¬
ing how to handle the gift of Divine disclosure; he would be ren¬
dering us morally unfree in this respect. And in preventing us from
choosing how to respond to Divine disclosure, he would be re¬
moving the possibility of an explicit choice with respect to our
relationship with himself—a choice which many theologians have
considered to be essential to a choice of destiny.
Perhaps it will be replied to this that there are other possible
choices one might make with respect to a relationship with God—
a choice to seek God, for example—and that it is not at all a bad
thing for God to be patient and to allow us to mature before per¬
mitting choices as significant as the one in question. But for this to
be plausible, the danger of a presumptuous response—and what is
more, a response with continuing relationship-inhibiting effects—
must be quite great indeed, and we have already noted the defects
of this view. Given the nature of human beings and the revelatory
20. Strangely, in order to suppose that it is the case, one must accept a claim
about human beings that is directly opposed to the claim advanced by Hick and
Swinburne (see Chapter 5): one must claim that humans would, in circumstances
of the sort in question, be morally unfree in respect of certain bad actions.
options open to God, even in those few cases where a presump¬

tuous response might seem possible (cases of nonbelief that might
turn into presumptuous belief if evidence were given), the choices
of the individuals in question could not be viewed as more likely to
be relationship-inhibiting than relationship-building, and so I sug¬
gest that a God who values both our friendship and our freedom
would not withdraw in order to prevent them from being made.
(If he withdrew, although a presumptuous response would be pre¬
vented, friendship would be rendered impossible and freedom
would be diminished, and so it seems clear that he would produce
neither an outweighing nor an offsetting good by doing so.) To
withdraw under such circumstances would betray not so much pa¬
tience as an un-Godly wariness and mistrust.21
My assessment of Pascal’s contribution to our discussion is
therefore a negative one. When we consider that only evidence suf¬
ficient for belief is required, that this could be provided without
removing the evil of the world, through religious experience, that
religious experience could communicate to us our corruption and
need for humility, and finally, that the danger of human presump¬
tion, even were it to exist, would only provide opportunity for a
significant exercise of moral freedom—when we consider all these
things, I say, our judgment must be that the Pascalian solution
does not succeed.22
Arguments Possibly or Actually Attributable to

Kierkegaard
In the case of Kierkegaard, even more than in the case of Pascal,

a certain tentativeness in interpretation is required. Although here
there is no shortage of completed works, there is a shortage of
21. Interestingly, if some human beings might respond presumptuously to good

evidence of God’s existence, and if so doing (or not so doing) would constitute an
important exercise of moral freedom, we are provided with an additional reply to
the arguments of Hick and Swinburne. These writers hold that good evidence
would remove moral freedom. Here, however, we have an exercise of moral free¬
dom that requires for its existence that the evidence be good.
22. This conclusion is further supported by the discussion in the penultimate
section of Chapter 7, where I argue that there is room for a sort of Divine with¬
drawal within the context of a Divine-human relationship.
works in which the author’s ideas are clearly stated. This is because
Kierkegaard preferred an indirect style. He wrote under pseudo¬
nyms and filled his books with parable, metaphor, irony, humor,
and poetry, hoping that by forcing his readers to toil for results, he
would cause them to become subjectively involved with the matters
discussed. Such involvement, he held, was necessary for the appre¬
hension of ethical and religious truths.23
I have more than one reason for noting here these features of
Kierkegaard’s writing style. For Kierkegaard’s most general inten¬
tions as a writer seem also to have been intentions he ascribed to
God, by reference to which he thought the hiddenness of God
could be explained. His (apparent) view was that God communi¬
cates with us indirectly in order to stimulate the proper sort of
subjective involvement in our existence as human beings, and in
order to prevent situations and states incompatible with such in¬
volvement. I will take it as my task to clarify and develop this view
and the reasons for Divine hiddenness it suggests, and to assess the
force of these reasons against the argument of Part i.24
23. For a discussion of Kierkegaard and indirect communication, see Louis P.

Pojman, The Logic of Subjectivity (University, Ala.: University of Alabama Press,
1984), pp. 148-154.
24. It should be noted that another, apparently conflicting, view on Divine hid¬
denness is also suggested by Kierkegaard, namely, that each of us knows there is a
God and that if God seems hidden, this is because we have stifled or suppressed
such knowledge: “There [has] never been an atheist, even though there certainly
have been many who have been unwilling to let what they know (that the god
exists) get control of their minds” (Philosophical Fragments, Howard V. Hong and
Edna H. Hong, eds. and trans. [Princeton: Princeton University Press, 1985], “Se¬
lected Entries from Kierkegaard’s Journals and Papers,” pp. 191-192). On this
view, acceptance of God’s existence must be understood in terms of knowledge,
not of faith. As Kierkegaard writes, “I do not believe [i.e., have faith] that God
exists . . . , but know it; whereas I believe that God has existed” (ibid., p. 222).
But this distinction is not upheld in other of his writings, where, as we will see,
God’s existence is said to be objectively uncertain and hence, like the paradox of
the Incarnation, a suitable object for faith; and the reasons for Divine hiddenness he
offers elsewhere emphasize not sin, but a Divine withdrawal which causes diffi¬
culties even for those who are not “unwilling” to become aware of God. Clearly
there is a problem of interpretation here. I will not, however, attempt to solve it,
but will assume for the sake of argument that Kierkegaard did not espouse a Cal¬
vinist position on this matter (if he did, he is adequately answered by considera¬
tions adduced in Chapter 3), and develop the other reasons for Divine hiddenness
suggested by his writings.
As suggested in the preceding paragraph, the basic structure of

the explanation I am (tentatively) attributing to Kierkegaard
closely resembles the basic structure of Pascal’s account. Like Pas¬
cal, Kierkegaard seems to have assigned to the hiddenness of God
both a positive and a negative function: Divine hiddenness, on his
view, both intensifies subjectivity and prevents certain situations
and states that would inhibit it.
In a moment I will be adding detail to this basic structure, but
before I do so, it will be useful to reflect for a moment on how
“hiddenness” and “subjectivity” are to be understood in this con¬
text. As I indicated earlier, the latter term, as used by Kierkegaard,
implies, minimally, involvement in one’s existence. But how can
this involvement be further characterized? An attempt at exhaus¬
tive characterization is of course out of the question, but perhaps
the following points, drawn from Kierkegaard’s Concluding Un¬
scientific Postscript, will help make this notion a little clearer.25
The subjective individual, as portrayed in the Postscript, focuses
his interests on some idea and, through his own decisions, brings
his life into conformity with it. As Louis Pojman explains, “the
important thing in [Kierkegaardian] subjectivity is appropriation,
the resolution and integration of an idea in one’s life.”2*’ The objec¬
tive individual, by contrast, pursues learned inquiries into ideas, to
see, for example, whether they are really instantiated. But as he
does so, Kierkegaard writes, the “infinite, personal, passionate in¬
terest of the subject . . . vanishes more and more, because the
decision is postponed, and postponed as following directly upon
the results of the learned inquiry” (p. 28). “All decisiveness, all
essential decisiveness, is rooted in subjectivity. A contemplative
spirit, and this is what the objective subject is, feels nowhere any
infinite need of a decision, and sees no decision anywhere. This is
the falsum that is inherent in all objectivity” (p. 33). Subjectivity,
unlike objectivity, involves an “internal decision in which an indi¬
vidual puts an end to the mere possibility and identifies himself
with the content of his thought in order to exist in it” (p. 302).
25. S0ren Kierkegaard, Concluding Unscientific Postscript, David F. Swenson and

Walter Lowrie, trans. (Princeton: Princeton University Press, 1941). Further refer¬
ences to this work, in both text and notes, will be made parenthetically.
26. Pojman, Logic of Subjectivity, p. 66.
The Importance of Inwardness [i55
The decisions made by the subjective individual involve risk and

cost; the ideal is that of persistent exertion and striving (p. no).
This costly exertion and striving in the face of great risk Kierke¬
gaard calls “passion” (p. 381).27 In passion the individual takes con¬
trol of his life; so far from letting his thought and action be deter¬
mined from without, he is se//^determined and self-directed. Instead
of drifting along, responding automatically to inclinations, he
“transforms his entire existence in relation [to his idea], and this
transformation is a process of dying away from the immediate” (p.
432).
The highest passion of all, according to Kierkegaard, is the pas¬
sion of faith (p. 118). Here we have the greatest risk, the absolute
venture. The one who has faith stakes his whole life on an objec¬
tive uncertainty (indeed, in the case of the idea of the Incarnation,
on an absurdity). Faith’s venture is to believe against all odds that
there is a God and that, although eternal, he has entered time in the
Incarnation. Since these are the most difficult of ideas to appropri¬
ate, they require the highest degree of decisiveness. As Kierkegaard
puts it: “Christianity . . . requires that the individual risk his
thought, venturing to believe against the understanding . . . this is
the absolute venture and the absolute risk” (p. 384). “The absurd
[must] stand out in all its clarity—in order that the individual may
believe if he wills it; I merely say that it must be strenuous in the
highest degree so to believe” (p. 190). “There can be no stronger
expression for inwardness than . . . when . . . facing the tremen¬
dous risk of the objective insecurity, the individual believes” (p.
188).
Although it is easy to miss this point, it seems that for Kierke¬
gaard, subjectivity has (what we may call) a definite direction. The
individual who exercises it will realize her potential for the infinite,
that is, her potential for a relationship with God. God is the infinite
Subject, and so it is only to be expected that those who come to
know him will do so by transcending the finite in subjectivity.
This occurs in the moment of highest passion—when one believes
27. For a good discussion of this notion, see Robert M. Adams, “Kierkegaard’s
Arguments against Objective Reasoning in Religion,” in Steven M. Cahn and
David Shatz, eds., Contemporary Philosophy of Religion (New York: Oxford Uni¬
versity Press, 1982), pp. 221-227.
in God against all reason. In that moment one is closest to God: “It
is only momentarily that the particular individual is able to realize
existentially a unity of the infinite and the finite which transcends
existence. This unity is realized in the moment of passion” (p.
176). “God is a subject, and therefore exists only for subjectivity in
inwardness” (p. 178). The subjective individual, to realize her po¬
tential, must believe “against the understanding,” and feel “the
peril of lying upon the deep, the seventy thousand fathoms, in order
there to find God” (p. 208, my emphasis).
Kierkegaard’s notion of subjectivity is notoriously difficult, and
this account of it is doubtless incomplete in some respects. But it
conveys, I think, some of his most important emphases and will
suffice for our purposes. What now of the notion of God’s “hid¬
denness”? How is it understood?
Kierkegaard himself hardly ever uses the term, but it is clear that
what it is commonly taken to mean is expressed by him in other
ways. He speaks in the Postscript of the “objective uncertainty” of
God’s existence (by which he means, at the very least, the improb¬
ability of God’s existence) and of God’s “elusiveness” (pp. 218-
219). As the term “elusive” suggests, Kierkegaard apparently holds
that many of the epistemic difficulties we may face vis-a-vis God’s
existence are a result of God’s intentional withdrawal. It is sug¬
gested that God could (in some sense) reveal himself more clearly
but chooses not to.28 Since it is not just human blindness but God’s
intentional withdrawal that lies behind Divine hiddenness, the anx¬
iety and uncertainty humans may feel with respect to God’s exist¬
ence are not (at least not in the first instance) necessarily signs of
culpability.29 And so we may say that Kierkegaard’s understanding
28. This view is suggested most clearly in the famous parable of the king and
the humble maiden. In his application of this parable to theology, Kierkegaard
writes of God’s dilemma: “Who grasps the contradiction of this sorrow: not to
disclose itself is the death of love; to disclose itself is the death of the beloved. . . .
How grievous it is to have to deny the learner that to which he aspires with his
whole soul and to have to deny it precisely because he is the beloved” (Fragments,
p. 30). Here again we encounter the “patient God.”
29. I say “at least not in the first instance” because Kierkegaard suggests that a
failure to develop subjectively, and so a failure to come to believe that there is a
God, may be culpable (see Postscript, pp. 363, 379).
of Divine hiddenness entails our own, that is, entails that God has
permitted the occurrence of reasonable nonbelief.30
Keeping these comments about “subjectivity” and “hiddenness”
in mind, we may now return to the sketch of Kierkegaard’s expla¬
nation given earlier and begin to fill in some of the detail. It was
suggested that he assigns to Divine hiddenness both a positive and
a negative function. I will begin with the former.
An argument for the view that divine hiddenness fulfills a posi¬
tive function is implicit in the following passage from the Postscript:
An objective uncertainty held fast in an appropriation-process of the most

passionate inwardness is the truth, the highest truth attainable for an
existing individual. . . . The truth is precisely the venture which
chooses an objective uncertainty with the passion of the infinite. I
contemplate the order of nature in the hope of finding God, and I
see omnipotence and wisdom; but I also see much else that disturbs
my mind and excites anxiety. The sum of all this is an objective
uncertainty. But it is for this very reason that the inwardness be¬
comes as intense as it is, for it embraces this objective uncertainty
with the entire passion of the infinite. . . .
But the above definition of truth is an equivalent expression for
faith. Without risk there is no faith. Faith is precisely the contradic¬
tion between the infinite passion of the individual’s inwardness and
the objective uncertainty. If I am capable of grasping God objec¬
tively, I do not believe, but precisely because I cannot do this I must
believe. [P. 182]31
30. Of course, Kierkegaard would not say, as I would, that it follows that evi¬
dence sufficient for belief has not been provided, for this statement, as I have
defined it, presupposes that belief is involuntary, and Kierkegaard denies that this
is so.
31. Alastair Hannay has some helpful comments on this passage which relate
also to our emphasis: “Faith is being certain in spite of the objective uncertainty.
. . . In describing faith as the ‘contradiction’ . . . between the ‘infinite passion of
inwardness’ and the ‘objective uncertainty,’ Kierkegaard is referring to the conflict
between the subjective, or personal, certainty one seeks, and has found, on the one
hand, and the objective uncertainty, on the other, which has made the finding of
the former depend on a strenuous personal choice. The passion of inwardness is
the active passion of someone who has decided upon a risky course of action and
knows that the determination with which he pursues it would give way to despair
were he to fall back on the impersonal authority of reason. The objective uncer-
It is clear from this passage that on Kierkegaard’s view, the ob¬

jective uncertainty of God’s existence is necessary for intense in¬
wardness in relation to the idea of God’s existence and hence for
faith. But the claim that God’s existence is objectively uncertain
entails the claim that God is hidden. Therefore, Kierkegaard can
conclude, God’s hiddenness is necessary for faith and accordingly
fulfills a positive function in relation to it.
This argument is a Kierkegaardian version of Pascal’s ill-fated
Stimulus Argument. But we can see (without committing our¬
selves to its success) that it is a much stronger argument, in the
sense that it can claim more. Given his understanding of faith,
Kierkegaard can claim that faith, unlike an awareness of wretched¬
ness, logically requires Divine hiddenness: if we accept his concept
of faith at all, we ipso facto accept the necessity of Divine hidden¬
ness for its instantiation. (If Pascal could have claimed this much,
no doubt his argument would not have fared so badly.)
I move now to Kierkegaard’s apparent claim that Divine hidden¬
ness has an important negative function—that it prevents certain
situations and states that would inhibit subjectivity. This claim is
more complex than the former claim and will take a little longer to
develop. My procedure will be as follows. I will first let Kierke¬
gaard speak for himself by quoting at some length from the Post¬
script. Then I will present what I take to be the important points
(implicit and explicit) in a somewhat more systematic fashion, con¬
cluding with a brief comparison of Kierkegaard’s thoughts on this
subject in the Postscript with those expressed in the Fragments.
Precisely because he himself is constantly in process of becoming

inwardly . . . , the religious individual can never use direct commu¬
nication, the movement in him being the precise opposite of that
presupposed in direct communication. Direct communication pre¬
supposes certainty; but certainty is impossible for anyone in process
of becoming, and the semblance of certainty constitutes for such an
individual a deception. [P. 68, n.]
tainty is neither replaced nor obscured by the subjective certainty, rather it is a

necessary condition of the latter” (Kierkegaard [London: Routledge and Kegan Paul,
1982], p. 127, my emphasis).
No anonymous author can more cunningly conceal himself. . . than

God. He is in the creation, and present everywhere in it, but directly
he is not there; and only when the individual turns to his inner self
. . . does he have his attention aroused, and is enabled to see God.
Nature is, indeed, the work of God, but only the handiwork is
directly present, not God. Is not this to behave, in His relation to
the individual, like an elusive author who nowhere sets down his
result in large type, or gives it to the reader beforehand in a preface?
And why is God elusive? Precisely because He is the truth, and by
being elusive desires to keep men from error. . . .
If God were to reveal himself in human form and grant a direct
relationship, by giving Himself, for example, the figure of a man six
yards tall, then [we] would doubtless have [our] attention aroused.
But the spiritual relationship to God in truth, when God refuses to
deceive, requires precisely that there be nothing remarkable about
the figure. . . . When God has nothing obviously remarkable about
him, [we are] perhaps deceived by not having [our] attention
aroused. But this is not God’s fault, and the actuality of such a de¬
ception is at the same time the constant possibility of the truth. But
if God has anything obviously remarkable about him, he deceives
men because they have their attention called to what is untrue, and
this direction of attention is at the same time the impossibility of the
truth. [Pp. 218-220]
The argument suggested by these passages is the following: “If

strong objective evidence of God’s existence were made generally
available, we would be deceived in two related ways.32 First, we
would be deceived into thinking that we had a proper understand¬
ing of religious truth—that the answers were clear, and striving
unnecessary. Second, we would be deceived into thinking that
God could be fully understood in terms of objective categories,
and that an external, prudential response to God was sufficient to
establish a proper relationship with him. The results of such decep¬
tion would be disastrous for beings disposed (as we are) to avoid
the strenuosity of subjectivity.33 For we would become complacent,
32. Kierkegaard means “deceived by God”: God, fully aware of our disposition
to avoid subjective strain, would, in providing objective evidence, be knowingly
causing the formation of false beliefs.
33. Kierkegaard writes, “In general it is quite inconceivable how ingenious and
inventive human beings can be in evading an ultimate decision” (Postscript, p. 379).
thinking that we had arrived when really there was still a lot to be
done, failing to recognize that the religious life, rightly pursued, is
subjectively demanding, and hence failing to pursue it rightly. Our
knowledge of God would be superficial, since the most that can be
known of God objectively is still radically incomplete: God is
Spirit and can only really be known through the activation of our
own spirit in inwardness. Our relationship with God would be
shallow, not deep, personal, and strenuous as befits a relationship
with the infinite Subject. We would remain happily within our fi¬
nite parameters, relating to God transactionally, not inwardly, and
hence not really relating to him at all. ”34 It is only a short step from
this argument (the Deception Argument, as I will call it) to the
claim that if God is hidden, the deception in question and its sub¬
jectivity-inhibiting corollaries will be prevented. Therefore, Kier¬
kegaard can claim that the hiddenness of God fulfills an important
negative function in relation to the life of faith.
It may be useful to note, before concluding this part of the dis¬
cussion, that the main emphasis of the Deception Argument seems
also to be traceable in the Philosophical Fragments.35 In chapter 2 of
that work, Kierkegaard presents his famous parable of the king and
the humble maiden. The main point of this parable is often taken
to be that God must come among us incognito if our freedom is to
be preserved.36 Kierkegaard, in other words, is often viewed as a
precursor of Hick. But while I do not wish to deny that such an
emphasis may be found in the Fragments, it is important to note
that it is not the only emphasis. In particular, the claim that, were
God to be openly revealed, we would be deceived and consequently
fail to reach a proper understanding of God, is also to be found:
The learner is in untruth . . . and yet he is the object of the god’s

love . . . the god wants to be his teacher, and the god’s concern is to
bring about equality. If this cannot be brought about, the love be-
34. On relating to God transactionally, see ibid., pp. 378-379.

35. This finding, of course, is to be expected if my interpretation of Kierkegaard
so far has been on the right track (for the Postscript is a “postscript” to the Frag¬
ments). If a similar emphasis can be found in the Fragments, this provides confirma¬
tion for my interpretation here.
36. See, for example, Penelhum, God and Skepticism, p. 79.
The Importance of Inwardness [ i6i
comes unhappy and the instruction meaningless, for they are unable
to understand each other. . . .
The poet’s task is to find a solution, a point of unity where there
is in truth love’s understanding . . . for this is the unfathomable love
that is not satisfied with what the object of love might foolishly
consider himself blissfully happy to have. [Pp. 28-29]
What wonderful self-denial to ask in concern, even though the

learner is the lowliest of all persons: Do you really love me? For he
himself knows where the danger threatens, and yet he knows that
for him any easier way would be a deception, even though the
learner would not understand it.
For love, any other revelation would be deception, because either
it would first have had to accomplish a change in the learner . . . and
conceal from him that this was needed, or in superficiality it would
have had to remain ignorant that the whole understanding between
them was a delusion. . . . For the god’s love, any other revelation
would be a deception—if I pleaded with him to change his resolu¬
tion, to manifest himself in some other way . . . then he would look
at me and say: Man, what have you to do with me. . . . Or, if he
just once stretched out his hand to bid it happen ... I would then
very likely see him weep also for me and hear him say: To think
that you could become so unfaithful to me and grieve love in this
way; so you love only the omnipotent one who performs miracles,
not him who humbled himself in equality with you. . . .
The situation of understanding—how terrifying, for it is indeed
less terrifying to fall upon one’s face while the mountains tremble at
the god’s voice than to sit with him as his equal, and yet the god’s
concern is precisely to sit this way. [Pp. 33-35]
It is natural to interpret these passages as suggesting that, were

God to be clearly revealed, we would be caused to acquire false
beliefs about him—beliefs which made the inward awareness of
God all the more difficult to realize; and that God, in view of this,
remains patiently withdrawn so that a proper relationship and a
proper understanding in inwardness may be achieved.
In bringing this expositon of Kierkegaard to a close, we may
note that his arguments do indeed provide possible rebuttals for
the prima facie case we are considering, for each suggests that God
has reason not to put his existence beyond reasonable nonbelief for
all human beings at all times. Whether they provide at least one
actual rebuttal as well is the question to which I now turn.
A Critique of Kierkegaard’s Arguments
The Kierkegaardian Stimulus Argument, stated informally in the

last section, can be stated more formally as follows:
(1) Intense inwardness in relation to the idea of God is essen¬

tial to faith.
(2) Such inwardness requires that one choose to believe that

there is a God in the face of objective uncertainty (im¬
probability).
(3) One can only choose to believe that there is a God in the
face of objective uncertainty if God is hidden.
(4) Divine hiddenness is necessary for intense inwardness in

relation to the idea of God. (From (2) and (3))
(5) Divine hiddenness is necessary for faith. (From (1) and

(4))
This argument turns into a possible rebuttal of the argument

from the reasonableness of nonbelief if we add the assumption that
faith (in Kierkegaard’s sense) is plausibly viewed as an outweighing
good or is clearly an offsetting good (in the senses of Chapter 4).
For then we may plausibly draw the further conclusion that God
has reason to hide himself. But as it seems to me, the argument
ultimately fares no better than its Pascalian counterpart. For it is
not at all clear that faith in Kierkegaard’s sense is a good of the sort
required. It is, for example, hard to see why such an intense form of
inwardness should be idealized. Indeed, it would seem that a prop¬
erly balanced view of the religious life requires that intensity of the
sort valued by Kierkegaard not be idealized. Such intensity seems
too narrow, excluding as it does many other good things in life
which a loving God might wish us to experience and enjoy. As
Robert Adams puts it:
Certainly much religious thought and feeling places a very high

value on sacrifice and on passionate intensity. But the doctrine that
it is desirable to increase without limit, or to the highest possible

degree (if there is one) the cost and risk of a religious life is less
plausible (to say the least) than the view that some degree of cost and
risk may add to the value of a religious life. The former doctrine
would set the religious interest at enmity with all other interests, or
at least with the best of them. ... In a tolerable religious ethics
some way must be found to conceive of the religious interest as
inclusive rather than exclusive of the best of other interests—includ¬
ing, I think, the interest in having well-grounded beliefs. [My em¬
phasis]37
Kierkegaard seems to have an “all-or-nothing” view—either we

focus entirely on risk and sacrifice, striving subjectively with an
infinite passion and, as part of this, deciding to believe religious
propositions against all reason, or else we live a life that is reli¬
giously worthless. But surely this view is mistaken. A life of grad¬
ual development and transformation, involving risks and sacrifices
but other goods as well, including the good of evidence for one’s
beliefs, seems to more nearly conform to the Christian ideal. But if
so, then intense inwardness in relation to the idea of God is not
essential to faith (i.e., the first premise of Kierkegaard’s argument
is false) and, consequently, Kierkegaard’s argument does not suc¬
ceed.
It seems that the only way Kierkegaard can escape this objection
is if he can show that all subjective endeavor would be ruled out by
God’s self-disclosure. If no subjective advance of any kind would
be possible given strong evidence of God’s existence, then the
more moderate view I have advocated (which presupposes that
some subjectivity is compatible with well-grounded belief in God’s
existence) would not represent a genuine alternative. But this is to
say that his argument is, at best, dependent on the Deception Ar¬
gument, yet to be considered. We will discuss this argument in due
course, but before doing so we must consider one other objection
to the Kierkegaardian Stimulus Argument—the most serious ob¬
jection, in my opinion.
Kierkegaard’s understanding of faith seems clearly to presuppose
that belief is voluntary: we are told that to have intense inward-
37. Adams, “Kierkegaard’s Arguments against Reasoning,” p. 227.

ness, we must choose to believe propositions viewed as improb¬

able. But this presupposition is not one we are in a position to
accept. For as I argued in the Introduction, there is very good rea¬
son indeed to suppose that belief is (logically) involuntary. But if
so, Kierkegaard’s argument must be rejected as containing an inco¬
herence: it makes no sense to suppose that what he recommends
can be done, or that God would wish it to be done.
The most that Kierkegaard can affirm, in light of this objection,
is a revised understanding of intense inwardness, according to
which we may choose to accept the claim that there is a God in the
face of objective uncertainty, that is, act as if it were true. Accep¬
tance in this sense does not entail belief and is indeed possible. But
such a modification would not significantly improve the quality of
Kierkegaard’s argument. It is true that acceptance of God’s exis¬
tence in the face of objective uncertainty requires that God be hid¬
den, but now we must ask why God would consider a state of
affairs in which such acceptance was possible to have a value as
great as the value of a situation in which belief was possible. As I
argued in Chapter i, belief is a logical precondition for certain theo¬
logical benefits and certain valuable religious states commonly as¬
sociated with faith. (Kierkegaard seems to agree, for he places
great emphasis on the notion of subjective conviction.) Hence,
other things being equal, we must suppose that the faith God
would value entails belief.
Because of these considerations, I would suggest that the revised
understanding of intense inwardness is little better than the unre¬
vised. Indeed, it seems that the modified version of Kierkegaard’s
argument is also ultimately dependent for its success (in this con¬
text) on the Deception Argument. For only if the provision of evi¬
dence sufficient for belief would greatly inhibit progress toward a
deep and personal relationship with God could acting-as-if be seen
as an attractive alternative. It is therefore to the Deception Argu¬
ment that we must direct our attention. For the purposes of discus¬
sion, let us state it a little more formally:
(6) If strong, objective evidence of God’s existence were

made available to them, human beings would form (false)
beliefs entailing that subjectivity is of no great impor-

tance.
(7) Human beings are disposed to avoid the strenuosity of

subjectivity.
(8) Individuals who are disposed to avoid the strenuosity of

subjectivity and who believe that subjectivity is of no
great importance will not become subjective.
(9) Therefore, if strong, objective evidence of God’s exis¬

tence were made available to them, human beings would
not become subjective.
I will assume here that God would indeed wish to facilitate the
exercise of subjectivity, even, if necessary, at the cost of permitting
reasonable nonbelief, meaning by this that some degree of risk and
sacrifice and commitment to the pursuit of genuinely religious
goals (including the goal of a relationship with God) are essential to
the most valuable kind of religious life. (I do not mean that God
would wish us to attempt to believe at will what we take to be
improbable propositions; we have already seen that the notion of
such “believing” is incoherent.) I will also assume that premises (7)
and (8) of the Deception Argument, as stated above, are true, for
their claims seem clearly correct. This leaves premise (6). What can
be said about it?
I would suggest that, whereas premise (6) must for our purposes
be taken as referring to evidence sufficient for belief, it is only
plausible when given the stronger interpretation that Kierkegaard
himself seems inclined to give it. If we think in terms of theoretical
proofs and striking displays of Divine power, as he seems to do,
we can begin to see how someone could come to assert it. For if
we were allowed access to such phenomena, we might indeed tend
to form false beliefs of the sort Kierkegaard mentions. If we wit¬
nessed an exhibition of Divine power, perhaps we would feel that
there was nothing left for us to do, that all that was required of us
was to stand back and watch God work, applauding at appropriate
moments, like observers at some celestial fireworks display. Per¬
haps if we were granted access to a successful theoretical proof, we
would be drawn away from the inner life and into a maze of argu¬
ments, thinking that this was the Divinely appointed way of ac-
i 66 ] The Force of the Argument
quiring further knowledge of God. In these ways our attention

might indeed be called to what is “untrue.” But it cries out for
demonstration that these are the only ways evidence sufficient for
belief could be provided. In particular, we must once again stress
that religious experience could provide the necessary evidence, and
that, so far from leading to the formation of beliefs entailing that
subjectivity is of no importance, such experience could inspire sub¬
jectivity. The Kierkegaardian claims that evidence sufficient for be¬
lief would hinder us from turning to our inner self, but if God
could meet us there in a way that motivated us to respond to him in
love and trust but also provided evidence sufficient for belief, then
it seems that his claim must be rejected.
Therefore, my response to the Deception Argument is as fol¬
lows: Evidence sufficient for correct belief need not lead to the
acquisition of false beliefs about God and so God need not deceive
us in providing it. Such evidence could, indeed, produce true be¬
liefs, beliefs entailing that subjectivity is of great importance, by
revealing God to us as personal, holy, and loving. Hence Divine
hiddenness is not necessary in order for humans to have the oppor¬
tunity to become subjective. Anyone who remained complacent or
inappropriately objective upon experiencing Divine self-disclosure
would not be deceived by God but would be, at most, self-de¬
ceived.
I suggested earlier in this section that well-grounded belief, if
compatible with subjectivity, is to be preferred to acting-as-if
which—although it makes possible a distinctive sort of inward¬
ness—cannot rival the value made possible by belief. In other
words, God might choose not to facilitate well-grounded belief if
this entailed a loss of subjectivity, but if it did not, he would
greatly prefer it.
We are now in a position to claim that well-grounded belief is
not incompatible with subjectivity and, therefore, to draw the rele¬
vant conclusion. But no doubt those who favor Kierkegaard’s ar¬
guments will wish to argue that although there is an inwardness of
belief, the inwardness of uncertainty is much deeper and more pro¬
found—that it is only if I construe the two kinds of inwardness as
unequal in value that I get my conclusion, and that this is not the
case. In drawing the chapter to a close, I attempt to articulate this
complaint and give my final response.
Experiential evidence, the Kierkegaardian will say, is just like

any other evidence in that it would, if given, diminish the value of
the religious life by vastly reducing the cost and risk entailed by it.
It is only when one seeks for God without any assurance that he is
there that one is truly in a position to venture, to risk all. If we were
never required to seek, we would never be required to strive in¬
wardly to achieve the knowledge of God; and surely it is a valuable
thing for persons in process of becoming to strive in this way.
To this I respond by suggesting that if the God met in experi¬
ence is the infinite Subject of Kierkegaard’s writings, there will be
no end to the process of “coming to know God” even for the one
who has believed from the start. There will always be new facets
of the Divine nature to discover and appreciate. No matter where
one begins, therefore, a lifetime of striving is only a start upon the
way. Thus, surely, there is no need for God to remain hidden so
that we can strive to know him!
As for cost and risk, these also exist for the believer, although,
again, in a different form. The one who believes that there is a God
has open to her a lifetime of worship and costly service. If she fully
commits herself, she as surely “exists in” the idea of God as the one
who, yet uncertain, undertakes to act-as-if. And since she also has
belief, and so has access to the valuable states it alone facilitates, we
must conclude that her state is, all things considered, much to be
preferred to that of the one who is (only) in a position to act-as-if.
[ 7 ]
Investigation, Diversity,
and Responsibility
In the foregoing, I have closely examined certain influential dis¬

cussions of Divine hiddenness in an attempt to find a plausible re¬
sponse to the problem of reasonable nonbelief. But in each case my
conclusion has been negative: the best-known attempts at an expla¬
nation of our situation in terms of the hiddenness of God fail to
provide a rebuttal for the argument of Part i. In this chapter I
consider a motley set of less well-known arguments. Some of
these, although clearly important in their own right, are related to
arguments already discussed; others suggest entirely new lines of
thought. Only the first—Joseph Butler’s Probation Argument—
exists in developed form. The others have not heretofore been de¬
veloped by anyone and are at most hinted at in some passage or
other. Nevertheless, my aim with respect to each argument is the
same: to give to it a clear shape and assess its force.
Butler on Intellectual Probation
Joseph Butler devotes the better part of a chapter (part 2, chapter

6) of his greatest work, The Analogy of Religion,' to a problem in
1. The full title is The Analogy of Religion, Natural and Revealed, to the Constitu¬
tion and Course of Nature. I have used thej. H. Bernard edition of Butler’s works
(The Works of Bishop Butler, 2 vols. [London: Macmillan, 1900]). Volume 2 con-
Investigation, Diversity, and Responsibility [
many ways similar to our own, the problem posed for Christians
by the following skeptical claim: “If the evidence of revelation ap¬
pears doubtful, this itself turns into a positive argument against it,
because it cannot be supposed that, if it were true, it would be left
to subsist upon doubtful evidence.”* 2 Now, as his formulation of
this claim suggests, the problem to which Butler addresses himself
and the problem of reasonable nonbelief do differ in at least one
respect: the evidence Butler refers to as appearing “doubtful”
(whose apparent doubtfulness he attempts to explain) is evidence
adduced in support of Christian revelatory claims and not the evi¬
dence of theism (with which we have been concerned).3 Hence to
see what force Butler’s arguments have in the context of our dis¬
cussion, it will be necessary to adapt them for our purposes—to set
on one side the question of doubtfulness attaching to revelatory
claims, and focus instead on whether the explanations he offers
succeed as explanations of apparent doubtfulness in theistic evi¬
dence. I will accordingly assume hereafter that the evidence to
which Butler refers is evidence of this latter sort.4
A central claim of Butler’s Analogy is that we are subjected to
temptation in this life so that we may become fitted, through the
development of good moral character, for the life to come. Ours is
a “state of probation . . . intended for moral discipline and im¬
provement.” We have been placed in it so that we might “qualify
ourselves, by the practice of virtue, for another state, . . a future
state of security and happiness.”5
It will be noticed that this view in some ways parallels the claim
of Swinburne, discussed in Chapter 5, with respect to the relation
between temptation and a genuine choice of destiny. This may in-
tains the Analogy, and Butler’s fifteenth Rolls Sermon, to which I will also refer, is
to be found in volume 1. The numbers appearing in my references to the Analogy
are for the part, the chapter, and the paragraph, respectively. In references to Ser¬
mon 15, only the paragraph number appears.
2. Butler, Analogy, 2, 6, 1.
3. The reason for this difference is that the evidence for God’s existence did not
appear doubtful to either Butler or his opponents, the deists.
4. I will, however, return to the distinction between evidence for theism and
evidence for Christian revelatory claims at a later stage of the discussion. As we
will see, it is a distinction that the critic of (the adapted) Butler can exploit.
5. Butler, Analogy, 1, 5, 1.
cline us to ask whether the two views of Divine hiddenness also

have something in common. Is it Butler’s opinion that we would
not be subject to temptation and, thus, not in a state of probation if
our epistemic situation were improved? Some passages suggest
that this is so. For example, in the fifteenth of Butler’s Rolls Ser¬
mons, “Upon the Ignorance of Man,” we read the following:
Now if the greatest pleasures and fains of the present life may be
overcome and suspended, as they manifestly may, by hope and fear,
and other passions and affections; then the evidence of religion, and
the sense of the consequences of virtue and vice, might have been
such, as entirely in all cases to prevail over . . . afflictions, diffi¬
culties and temptation; prevail over them so, as to render them ab¬
solutely none at all. But the very notion ... of a state of discipline
and improvement, necessarily excludes such sensible evidence and
conviction of religion, and of the consequences of virtue and vice.6
To this version of the Probation Argument, it seems to me, the

responses to Swinburne urged in Chapter 5 apply equally well, and
so I will consider it no further. But there is a second version of the
argument in Butler’s writings, which makes an independent claim,
and to which he gives much more attention. In Sermon 15, its
point is made briefly:
Religion consists in submission and resignation to the divine will.

Our condition in this world is a school of exercise for this temper:
and our ignorance, the shallowness of our reason, the temptations,
difficulties, afflictions, which we are exposed to, all equally contrib¬
ute to make it so. . . . Therefore difficulties in speculation as much
come into the notion of the state of discipline, as difficulties in prac¬
tice: and so the same reason or account is to be given of both.7
In the Analogy, it is made at greater length:
The evidence of religion not appearing obvious, may constitute one

particular part of some men’s trial in the religious sense; as it gives
scope for a virtuous exercise, or vicious neglect of their understand-
6. Butler, Sermon 15, 9.

7. Ibid.
Investigation, Diversity, and Responsibility [ 171
ing, in examining or not examining into that evidence. There seems

no possible reason to be given, why we may not be in a state of
moral probation, with regard to the exercise of our understanding
upon the subject of religion, as we are with regard to our behaviour
in common affairs. The former is as much a thing within our power
and choice as the latter. . . . Thus, that religion is not intuitively
true, but a matter of deduction and inference; that a conviction of its
truth is not forced upon everyone; this as much constitutes religious
probation, as much affords sphere, scope, opportunity for right and
wrong behaviour, as anything whatever does.8
Temptations render our state a more improving state of discipline,

than it would be otherwise: as they give occasion for a more atten¬
tive exercise of the virtuous principle, which confirms and
strengthens it, more than easier or less attentive exercise of it could.
Now speculative difficulties are, in this respect, of the very same
nature with these external temptations. For the evidence of religion
not appearing obvious is to some persons a temptation to reject it,
without any consideration at all; and therefore requires such an al¬
ternative exercise of the virtuous principle, seriously to consider that
evidence, as there would be no occasion for, but for such tempta¬
tion. And the supposed doubtfulness of its evidence, after it has
been in some sort considered, affords opportunity to an unfair mind
of explaining away, and deceitfully hiding from itself, that evidence
which it might see. . . . [Such temptation], as it calls forth some
virtuous efforts, additional to what would otherwise have been
wanting, cannot but be an additional discipline and improvement of
virtue.9
The claim of the second version of the Probation Argument, as

indicated by these passages, concerns not moral temptation and
moral probation simpliciter, but one particular sort of moral tempta¬
tion and one particular sort of moral probation. These we may, fol¬
lowing Penelhum, call intellectual temptation and intellectual pro¬
bation. If God’s existence were made more obvious, (the adapted)
Butler argues, the choices with respect to theistic evidence of (in
Penelhum’s words) “frivolous inattention, easy rejection and heed¬
less negativity” and, on the good side, “serious scrutiny, lengthy
8. Buder, Analogy, 2, 6, 8.
9. Ibid., 2, 6, 12.
reflection and wrestling with unwanted doubts” would not be

open to us.10 But, it is implied, it is a good thing that humans are
subject to these additional choices and temptations, since they pro¬
vide further opportunity for the development of moral character.
Therefore (the conclusion follows), God cannot be expected to
make his existence any more obvious than it is.11
Butler might seem to be on firmer ground here than with the
first argument. For does not intellectual probation, as here defined,
by its very nature require that God’s existence be less than obvious?
How could we be in doubt about something that has been made
clear to us? Hence it might seem that even if moral probation of
other sorts is compatible with a clear indication of God’s existence,
intellectual probation is not, and thus that even if Swinburne’s argu¬
ment and the first version of the Probation Argument do not suc¬
ceed, this argument may.12
10. Terence Penelhum, Butler (London: Routledge and Kegan Paul, 1985), pp.
195, 196. Penelhum is one of very few recent writers who give attention to But¬
ler’s philosophy of religion, and I have been greatly helped by his careful study.
11. Whether Butler considered this argument to be an extension of the other or a
narrower version of it depends on whether or not, in the final analysis, he endorsed
the latter. The first passage from Sermon 15 quoted above suggests that he did.
But it is important to note that the Analogy, which treats these matters at much
greater length, seems to favor a contrary view. There it is clearly implied that “real
immoral depravity and dissoluteness” is perfectly compatible with a “distinct con¬
viction” of the truth of religion (Analogy, 2, 6, 8), and from this it follows that
clear evidence would not necessarily remove moral temptation. For this reason, it
seems to me preferable to suppose that Butler viewed the second Probation Argu¬
ment as a narrower, more acceptable, version of the first.
12. Swinburne appears at times to recognize the possibility of extending his
“choice of destiny” argument to choices with respect to investigation of theistic
evidence (see, e.g., Faith and Reason [Oxford: Clarendon Press, 1981], p. 86). He
also suggests the possibility of construing as valuable the cooperative investigation
of theistic evidence (The Existence of God [Oxford: Clarendon Press, 1979], pp.
188-189). The arguments I deploy against Butler can, however, be applied to these
arguments with little adaptation. One possible response to the latter argument is
suggested by Swinburne himself when he notes that “men seem only to be begin¬
ning to take the opportunities which exist for co-operation for long-term practical
ends” (ibid., p. 189). There would appear to be endless opportunities for human
cooperation on important matters even if the question of God’s existence does not
need to be cooperatively investigated, and so—especially given that doing so
would preclude, in many cases, the great good of personal relationship with God
in this life—it seems unlikely that God would withdraw his presence on this ac¬
count.
I will have occasion to question this distinction later in the chap¬

ter, but for now let us suppose that apparent doubtfulness in theis-
tic evidence is necessary for intellectual probation and see how
Butler’s argument fares, given this assumption. It seems to me that
even if we accept (as I am prepared to do) that a loving God might
well subject human beings to a period of moral probation, we need
not accept the further claim that such a God would impose intellec¬
tual probation. If we assume, as we must, given the arguments in
Chapter 5, that a wide range of temptations to do bad actions
would remain even for individuals provided with evidence suffi¬
cient for belief (and thus that the provision of such evidence is
compatible with moral probation), and if we recognize, as we
should, given the arguments of Chapter 1, that apparent doubtful¬
ness in theistic evidence must (by virtue of its inhibition of belief)
remove, for as long as it persists, the possibility of an explicit per¬
sonal relationship with God, and all that entails, we should regard
as at any rate initially suspicious the claim that God would wish to
add to the aforementioned temptations the intellectual temptations
attendant upon apparently doubtful evidence. Because of its nega¬
tive effects and apparent superfluity, intellectual probation surely
requires more in the way of support than Butler’s claim—namely,
that the temptations attendant upon it give further occasion for
“discipline and improvement of virtue”—seems capable of provid¬
ing. Given this justification alone, intellectual probation seems
clearly the lesser of the two goods and so cannot provide the basis
for a rebuttal of the sort we require.
In light of these points, it seems that if his claims are to go
through, Butler must show that we have special reasons for suppos¬
ing that God would wish to impose on at any rate some individuals
a period of intellectual probation, reasons additional to the reasons
we have just by virtue of the fact that such probation is a form of
moral probation and so provides further occasion for “improve¬
ment of virtue.” And toward the end of his discussion, Butler does
indeed seem to suggest such reasons:
[There does not] appear any absurdity in supposing, that the spec¬
ulative difficulties, in which the evidence of religion is involved,
may make even the principal part of some persons’ trial. For as the
chief temptations of the generality of the world are the ordinary

motives to injustice or unrestrained pleasure; or to live in the neglect
of religion from the frame of mind, which renders many persons
without feeling as to anything distant, or which is not the object of
their senses: so there are other persons without this shallowness of
temper, persons of a deeper sense as to what is invisible and future;
who not only see, but have a general practical feeling, that what is
to come will be present, and that things are not less real for their not
being objects of sense; and who, from their natural constitution of
body and of temper, and from their external condition, may have
small temptations to behave ill, small difficulty in behaving well, in
the common course of life. Now when these latter persons have a
distinct full conviction of the truth of religion, without any possible
doubts or difficulties, the practice of it is to them unavoidable, un¬
less they will do constant violence to their own minds; and religion
is scarce any more a discipline to them, than it is to creatures in a
state of perfection. Yet these persons may possibly stand in need of
moral discipline and exercise in a higher degree, than they would
have by such an easy practice of religion. Or it may be requisite, for
reasons unknown to us, that they should give some further mani¬
festation what is their moral character, to the creation of God, than
such a practice of it would be. Thus . . . what constitutes . . . the
probation ... of some persons may be the difficulties in which the
evidence of religion is involved; and their principal and distin¬
guished trial may be, how they will behave under and with respect
to these difficulties.13
The argument here is that for some people, intellectual probation

is not, as we have assumed, just an added form of probation but
the only form; without it, they would not be subjected to serious
temptation at all. These are people of an intellectual bent who are
not troubled by the “ordinary” motives to injustice, unrestricted
pleasure, and the neglect of religion in favor of immediate (sensual)
13. Buder, Analogy, 2, 6, 13. It is interesting to note that this passage also sug¬
gests a reason for denying the (apparent) claim of Sermon 15 (viz., that a much
improved epistemic situation would rule out moral temptation and probation).
Butler suggests that many people, if provided with a proof of the truth of religion,
would still neglect it “from that frame of mind which renders many persons with¬
out feeling as to anything distant, or which is not the object of their senses.” This
is essentially the same point I made toward the end of Chapter 5, namely, that it is
often quite difficult to give up short-term pleasure for the sake of long-term goals.
fulfillment, people who would not have any difficulty with the
practice of religion should a demonstration of its truth become
known to them. Therefore, if they also require a period of moral
probation (and Butler assumes they do), such individuals should
not be provided with a demonstration of religion, but should be
required to face the temptations attendant upon difficulties in the
evidence.14 In this way they too may be disciplined and provided
with an opportunity for the improvement of virtue.
What should we say in response to this argument? Does it make
plausible the view that a loving God would require at any rate
some individuals to investigate the question of his existence before
arriving at evidence sufficient for belief? As it seems to me, it does
not, for the following reasons.
First, individuals of the sort Butler describes must be exceed¬
ingly few in number, if they exist at all. No doubt there are per¬
sons of an intellectual bent who have a “deeper sense as to what is
invisible or future” and who recognize that things are “not less real
for their not being objects of sense.” But how many are also free
from the temptations to which objects of sense give rise? Suppose,
however, that there exist individuals who combine these qualities.
Now we must ask why Butler considers them to be in need of
discipline and improvement of virtue. If their virtue is apparent,
why should they be subjected to additional tests? Surely God
would only sacrifice personal relationships for the sake of intellec¬
tual probation, even for a short while, if there were individuals
who stood in need of it. Let us, however, also grant that the indi¬
viduals in question are in need of discipline. The question that then
arises is how intellectual probation could ever provide it if other
14. Note that Butler’s point here is not just a narrower version of Swinburne’s
prudential claim, that is, his point is not that this particular class of individuals
should not be provided with strong evidence because of their special sensitivity to
its prudential implications. His point is rather that clear evidence should not be
provided because the individuals in question stand in need of intellectual probation,
and this (logically) necessitates difficulties in the evidence. According to Butler,
these individuals would find religious practice easy in the circumstances described,
not for any reason directly associated with the availability of a demonstration of
religion, but because of their general rationality and freedom from sensual desire.
They are inclined to do what is right even if it goes against short-term interests and
so, seeing religious practice to be morally right, would not be tempted to do any¬
thing else.
forms of moral probation do not. Butler attempts to move us on

to intellectual probation by demonstrating that these individuals
are not subject to other temptations, but it is unclear how apparent
doubtfulness in the evidence could tempt anyone to be frivolous or
dismissive or to avoid serious scrutiny who was not in the “com¬
mon course of life” inclined to reject long-term goals in favor of
short-term pleasures or to give up in the face of difficulty. Those
for whom intellectual temptations can arise must, it seems, be indi¬
viduals who would be tempted in other contexts of life as well.
Butler seems therefore to be faced with a dilemma: either the indi¬
viduals in question cannot be subjected to temptation at all, finding
no “difficulties,” intellectual or other, to be a source of temptation,
or they can be subjected not only to intellectual but also to other
forms of probation. Whichever alternative is chosen, the special
importance of intellectual probation falls away. Let us now go one
step farther, however, and suppose that even this point can be an¬
swered, and that (despite what has been said) intellectual probation
is a genuine possibility for the persons in question. It seems to me,
finally, that Butler is wrong in thinking that the extraordinary in¬
dividuals he describes could be put to the test in no other way, and
thus that intellectual probation is in the end unnecessary even for
them. In particular, it seems he underestimates the difficulties
likely to be encountered by those who give themselves wholehear¬
tedly to the life of faith. One who was undeterred by the “ordi¬
nary” temptation to neglect religion would surely, upon being ap¬
prised of its truth, seek to follow closely in the footsteps of its
most exemplary practitioners, who continually perform super¬
erogatory actions and sometimes give up life itself for the sake of
others; and such a form of life could not be “easy” for anyone.
It seems, therefore, that Butler’s argument does not go through.
Given that intellectual probation would have side effects such as a
loving God would seek, other things being equal, to avoid, and
given that its possible good effects would be superfluous (i.e.,
given that other forms of probation, suitable for everyone,
abound), God would surely not view its existence as having a
value as great as the value of a strong epistemic situation in relation
to theism, and so would not impose it on anyone.15
15. Perhaps it will be claimed that even if the instrumental value of intellectual
probation is not comparatively very great, the choices it makes possible have great
There is another route to this conclusion, which, before moving

on, I would like briefly to indicate. We have so far assumed that
apparent doubtfulness in theistic evidence is necessary for intellec¬
tual probation—that, unlike moral probation of other sorts, intel¬
lectual probation is not compatible with a clear indication of God’s
existence. But this assumption may be challenged, and in two
ways. The first retains the connection between intellectual proba¬
tion and theistic evidence, and suggests that temptations with re¬
spect to this evidence would persist even if it supported a theistic
conclusion. The second proposes a broadening of the notion of
intellectual probation to allow for temptation arising with respect
to evidence adduced in support of other important religious propo¬
sitions.
Beginning with the first of these, we may note that its central
claim is one we have already encountered several times in Part 2.
This is Penelhum’s point, that self-deception is possible even when
good evidence for theism presents itself and that temptations to
self-deception may therefore arise not only when the evidence is
obscure but also when it is clear. In Penelhum’s book on Butler,
the point is made in the following ways:
If we do have the freedom to respond or not to respond to the signs

of God, this freedom exists if the signs are clear, as well as if they
are not. It is very well known (and Butler, who says such wise
things about self-deceit, certainly knows it) that people can ignore
obvious facts, can fail to attend to what stares them in the face, and
can invent, or eagerly grasp on to, difficulties that would not im¬
press them at all if they were not biased against learning unwelcome
truths. The very facts about us that make intellectual probation a
intrinsic value. There is value in free choices to pursue the good quite apart from
the value of the improved character to which such choices may lead. But to this we
must reply that the intrinsic value of such additional free choices surely does not
exceed the intrinsic value of personal relationship with God. And since, as we have
seen, God would value such relationship not only for the good—instrumental or
noninstrumental—it would produce in the life of the believer, but also for its own
sake, we must conclude that, all things considered, intellectual probation would
not be viewed by God as facilitating a good as great as the good that would (appar¬
ently) need to be sacrificed in order to make it possible.
real possibility show that it could exist quite well, in all the forms
Butler suggests for it, if the signs of God’s presence were clear.16
It is just mistaken to suppose that there are no moral tests involved

in acknowledging a fact which we know for certain, as opposed to
one which we judge to be merely probable. If the acknowledgement
is a painful one, showing our failures and defects to the world, there
are no limits to the ingenuity with which we will persuade ourselves
that we do not know it after all, or will distract ourselves from
paying attention to it.17
It seems, therefore, that even if God’s existence were made more

obvious—if, say, evidence sufficient for belief were made avail¬
able—many of the choices with respect to theistic evidence listed
earlier, namely, frivolous inattention, easy rejection, heedless nega¬
tivity, serious scrutiny, lengthy reflection, and wrestling with
doubt, would remain open to us. Theistic evidence would continue
to be a source of intellectual temptation, and thus intellectual pro¬
bation would still be possible: the bad choices mentioned would
remain tempting for many and the good choices difficult.
Penelhum’s argument succeeds, I think, in showing that many
individuals would find theistic evidence to be a source of moral
temptation even if its verdict were much clearer than it is. But
what of those extraordinary individuals, described by Butler, who
would happily begin the practice of religion if convinced of its
truth? Surely they would be immune to the temptations outlined
by Penelhum. Surely they would require evidence that could be
reasonably doubted. Now it seems to me that the criticisms already
given of Butler’s claims with respect to these individuals suffice to
show that, were they to exist, they would not require intellectual
probation at all. But suppose that these objections can be an¬
swered. We may yet deploy the second criticism, mentioned
above, of the view that doubtfulness in theistic evidence is neces¬
sary for intellectual probation. Returning to the distinction be¬
tween the evidence of theism and the evidence of revelation, we
can argue that apparent doubtfulness in theistic evidence is not re¬
quired for intellectual probation because the evidence of revelation
16. Penelhum, Butler, p. 195.

17. Ibid., p. 197.
may appear doubtful, and hence may—if the notion of intellectual

probation is suitably broadened—provide the basis for a fully satis¬
factory form of intellectual probation. There is no reason to sup¬
pose that if the question of God’s existence has been settled, all
related questions—such as whether God has in some way revealed
himself and his purposes for humankind more fully—must also
be settled: good evidence for God’s existence may coexist with a
paucity of evidence for revelatory claims.18 But if the evidence rele¬
vant to such claims (e.g., the claim that God is revealed in Christ)
may remain open to reasonable doubt, then even if the evidence
for God’s existence no longer is, we may still be put in a state of
intellectual probation with respect to the truth of religion.19 There¬
fore, even if there are individuals who require a stronger form of
intellectual probation, this does not at all imply the need for appar¬
ent doubtfulness in theistic evidence.
Now perhaps one objection to this argument will be that if the
individuals in question were provided with good evidence for
God’s existence, their motivation to pursue questions about a Di¬
vine revelation would be much greater—they would wish to know
God’s will in order to be in a position to conform their lives to it—
and so the tests represented by such questions would be much less
severe than I have suggested. But in claiming that these individuals
would be tempted were the doubtful evidence evidence concerning
God’s existence and that they require such doubtfulness for proba-
tional purposes, the critic has in effect already conceded that they
would not automatically do what they saw reason to do. (They
would see that serious investigation of theistic evidence was called
for but be tempted not to pursue it.) And so he cannot claim here
that finding good reason to investigate the evidence of revelation,
the individuals in question would have no desire to do otherwise.
Since the only more general justification for this claim would seem
18. Some may wonder whether we are not committed by the terms of the argu¬
ment developed in Part 1 to the claim that any evidence of revelation would also be
clear. They are referred to the end of Chapter 1, where a response to this sort of
objection is provided.
19. I assume that the evidence of revelation may remain open to reasonable
doubt. If, as I have argued, God’s existence may be reasonably doubted, this would
seem to be true a fortiori with respect to more specific claims about God’s action in
the world.
to require the acceptance of arguments discussed in Chapter 5,

which I have rejected, I suggest that it is an unsuccessful claim.20
In consequence of the arguments just given, we are, it seems, in
a position to conclude once more that given its negative implica¬
tions and the superfluity of its good effects, a loving God would
not impose intellectual probation vis-a-vis theistic evidence on any¬
one.21 Indeed, now we may claim that various forms of probation
including intellectual probation would be preserved even if God’s
existence were beyond reasonable nonbelief. Butler’s arguments
therefore do not provide the rebuttal we require.22
20. In defense of this assessment, we may also call as witness the (unadapted)
Butler, for he and his opponents presupposed the existence of God, and yet the
difficulties in which the evidence of revelation is involved concerned them—in
particular, were taken by Butler as providing intellectual probation. The fact that
God’s existence was already (in his opinion) well-evidenced did not in his view in
any way lessen the value of the intellectual probation provided by doubtfulness in
the evidence of revelation.
21. It may be claimed that questions about God’s existence—about whether
there is a religious reality at all—are far deeper and more serious than questions
that presuppose the truth of the theistic claim, and so ones God would wish to be
among those we are given the opportunity of investigating. For this reason, it may
be said, it is a mistake to put both forms of investigation into the same category, as
I have done. Investigation of God’s existence would be viewed by God as the more
valuable of the two. But I suggest it is not at all obvious that the investigations
into religious matters we would be in a position to pursue were we in possession
of strong evidence for God’s existence and (perhaps) growing in relationship with
God would be any less interesting, deep, serious, and so on, than the investigation
of questions concerning God’s existence. If we suppose there to be a God who is
infinitely rich and inconceivably great, we may surmise that there is no limit to the
number of interesting and fruitful religious investigations it is possible to pursue
even after having satisfied ourselves that there is a God, and so that there is no way
to legitimately judge, from our perspective, that one such investigation is signifi¬
cantly “deeper,” more “serious,” and so on, than another. Even if there is such a
difference, however, I would claim that investigation of revelatory claims provides
a reasonable substitute for theistic investigation, and that since the former investiga¬
tion is compatible with theistic belief and the goods for which it is necessary,
whereas the latter is not, we must view a loving God as preferring it, all things
considered.
22. For an argument similar to Butler’s, which defends an understanding of
religious doubt as trial, seej. R. Lucas, “Doubt: A Sermon,” in Freedom and Grace
(London: SPCK, 1976), pp. 120-122. But Lucas also suggests an answer to this
argument, construed as an argument against the prima facie case I have developed,
namely, that the doubt in question may occur and have the desired effects within
the context of a relationship with God, and so need not occur prior to its develop¬
ment. I look more closely at suggestions of this sort in the penultimate section of
this chapter.
Investigation, Diversity, and Responsibility [ i8i
The Diversity Argument
I turn now to a very different response to our problem, which

we may call the Diversity Argument.23 It can be formulated as fol¬
lows:
(1) Religious diversity is at least as great a good as the good

God would sacrifice by allowing reasonable nonbelief to
occur.
(2) Diversity in religion would be vastly reduced if our epis-
temic situation in relation to theism were a strong one.
(3) Therefore, it is plausible to suppose that God would not
allow such a situation to obtain.
The term “religious diversity” which appears in this argument

admits of at least two interpretations. It can be seen as referring to
the actual concrete religious traditions (Judaism, Christianity, Is¬
lam, Buddhism, Hinduism, and so on) or to the free expression of
religious imagination, creativity, and devotion, which (in their bet¬
ter moments) these traditions represent. We have, that is, the fol¬
lowing alternative constructions of (1) and (2):
(T) The existence of Judaism, Christianity, Islam, Buddhism,

Hinduism, and other actual religious traditions is at least
as great a good as the good God would sacrifice by al¬
lowing reasonable nonbelief to occur.
(2') The map of the actually existing religious traditions
would be greatly changed if our epistemic situation in re¬
lation to theism were a strong one.
(1") The free expression of religious imagination, creativity,

and devotion is at least as great a good as the good God
would sacrifice by allowing reasonable nonbelief to occur.
(2") The expression of creative religiousness would be greatly
inhibited if our epistemic situation in relation to theism
were a strong one.
We may begin by assessing the force of the second set of prem¬

ises. As it seems to me, the most vulnerable member of this set is
23. I am grateful to Robert Adams for pointing out to me the possibility of such
a response. The formulation of it that appears here is, however, my own.
(2").24 It is clear that religious diversity in its sense is exemplified at

a (perhaps infinite) number of possible worlds. In particular, it is
clearly exemplified at those possible worlds in which the epistemic
situation of humans in relation to theism is a strong one. But if so,
then (2") is false, and this version of the Diversity Argument is
unsuccessful.
Let us consider how this response can be developed. The claim
to which it is opposed can be expressed as follows: “If a strong
epistemic situation in relation to theism were to obtain, human
religiousness would be reduced to a narrow and stifling unifor¬
mity.” Unless some such claim can be shown to be true or plausi¬
bly viewed as true, we have no reason to be deterred by (2"). But it
seems to me that all such claims are false. We may note, first of all,
that what is likely to follow from God presenting himself to the
experience of all individuals capable of recognizing him in the
manner described in Chapter 2 is not a uniform pattern of reli¬
giousness, but rather patterns of religious life that are (at one level
at least) compatible, united under a common acceptance of God as
personal and loving. Even this may be saying too much. For if
humans would remain free to reject God, as I have argued they
would, there would presumably remain the possibility of religious
beliefs and practices incompatible at all levels with traditions built
up on the experience of God, resulting directly or indirectly from
the rejection of that experience by some individual(s) at some point
in time. Suppose, however, that the “compatibility” claim is one to
which the critic of the Diversity Argument is committed. (2") is
still false; for compatibility of the sort in question does not rule out
diversity. Even within a particular tradition such as Christianity,
where the various denominations are united not only under a com¬
mon acceptance of God as personal and loving, but also under a
common acceptance of Jesus of Nazareth as in some sense revela¬
tory of God, there is room for diversity—for the expression of
creative religiousness. To take but the most obvious example: even
though all denominations make some connection between “Jesus”
and “God,” the connection is not by any means the same in all
24. I will assume here that (1") is true. Religious diversity in its sense can be
viewed as intrinsically valuable or as having value for the sort of reason discussed in
Chapter 5.
cases. Indeed, disputes rage. This suggests that, were a strong epis-
temic situation in relation to theism to obtain, religious diversity
(in the sense in question) would not at all be ruled out. Quite the
contrary. If the object of the experiences in question were indeed
the inexhaustibly rich source of existence, we would expect tradi¬
tions that arose on the basis of such experiences to (potentially)
reflect this richness.25 While a certain direction might be given to
religious creativity, creativity could hardly be restricted in any im¬
portant sense. (And if such direction would not prevent diversity,
its presence should be seen as an advantage of the situation in ques¬
tion, rendering it preferable, all things considered, to the actual
situation; for creative chaos is surely not the religious ideal.) When
all of this is taken into account, it seems clear that where “religious
diversity” is understood as in (1") and (2"), the Diversity Argument
fails.
Can a more favorable judgment be rendered if the premises of
the Diversity Argument are (1') and (2') instead of (1") and (2")? It
seems to me that the answer must again be a negative one. Here
the problem is not the second premise but the first. It is not the
second because it is likely that any world in which a strong epis-
temic situation in relation to theism obtained would be one in
which the map of the religious traditions was different. If God had
always followed a policy of presenting himself to the experience of
individuals capable of recognizing him, then (it is likely) various
traditions now existing would not have arisen and/or various tradi¬
tions not now existing would have arisen (and/or various traditions
now existing would have existed in different form). Such an im¬
portant input into the religious life of humankind would surely
have had some effect.
Let us assume on the basis of this consideration that (2') is true.
What is to prevent us from accepting (C) as well and declaring the
argument sound? In response to this it can be argued, first of all,
that (1') would appear to derive any plausibility it has from (1"). If
the actual religious traditions have value, it is because they in some
way instantiate the free expression of religious imagination, cre-
25. In line with this point, we should recall what was mentioned in Chapter 2,
that in the situation in question there would be nothing preventing individuals
from having experiences apparently of God of other sorts.
ativity, and so forth. But then, given that there are various possible
traditions that might have existed instead and that might also have
instantiated religious creativity, including religious traditions com¬
patible with a strong epistemic situation in relation to theism, and
given the extra value we must associate with traditions of the latter
sort, we must surely conclude that the existence of the actual reli¬
gious traditions is not as great a good as the good their existence
precludes, and so that (T) is false.
Perhaps it will be replied that Judaism, Christianity, and other
religious traditions are valuable not just in the sense of the preced¬
ing paragraph but also in an instrumental sense, constituting (to ap¬
propriate a Christian term) important “means of grace” for their
adherents. But surely we must answer, again, that many other
possible traditions, had they existed instead, would have per¬
formed the same function. In particular, if God exists and all expe¬
riences of grace are ultimately traceable to his benevolence, then, as
I have argued, possible situations in which the religious life of hu¬
mankind and explicit belief in God’s existence are more closely
linked than in ours must be ones in which the grace of God is more
readily appropriated. For these reasons and the reasons given
above, I conclude that (i'), which implies that God might wish the
religious life of humankind to be just as it is, is false. The argument
from (T) and (2') is therefore no more successful than the argu¬
ment from (1") and (2"), and so the Diversity Argument fails.
Reasonable Nonbelief and Actual Existents
An argument in some ways related to the second version of the

Diversity Argument but sufficiently independent to warrant sepa¬
rate treatment can be derived from Robert Adams’s essay “Exis¬
tence, Self-Interest, and the Problem of Evil.”26 In this essay,
Adams argues, following Leibniz, that we owe our existence to
past evils:
I could not have existed without past evils that have profoundly
affected the course of human history, and especially the “combina-
26. Robert Adams, “Existence, Self-Interest, and the Problem of Evil,” chap. 5
in The Virtue of Faith (New York: Oxford University Press, 1987).
tions of. . . people and marriages.” ... A multiplicity of interacting

chances, including evils great and small, affect which people mate,
which gametes find each other, and which children come into being.
The farther back we go in history, the larger the proportion of evils
to which we owe our being; for the causal nexus relevant to our
individual genesis widens as we go back in time. We almost cer¬
tainly would never have existed had there not been just about the
same evils as actually occurred in a large part of human history.27
Adams claims that it follows from this that “God has not wronged
us in causing or permitting those evils, if he is going to see to it
that we will have lives that are worth living on the whole. What
right could I have against satisfying the necessary conditions of my
coming to be, and how could I be injured by satisfying them, if
my life will be worth living?”28 On Adams’s view, this argument
makes a significant contribution to theodicy, for if all those who
claim they have been wronged by God in fact owe their existence to
the evils God has allowed (and live a reasonably good life), the
possibility of justified complaint would seem largely ruled out.
Critics of the case developed in Part 1 who are impressed by this
argument will wish to argue by analogy as follows: “If it were not
for the fact that God has not always followed a policy of presenting
himself to the experience of all human beings capable of recogniz¬
ing him, we—that is, the actually existing human beings, includ¬
ing those constructing and evaluating this argument—would very
likely not have come into existence. The absence of such an impor¬
tant input into the life of humankind, though it may seem regret¬
table, has in fact affected ‘the course of human history’ (including
especially the religious life of humankind) and (consequently) the
‘combinations of. . . people and marriages’; and we would almost
certainly not have existed had things been otherwise.” Since we are
referring here to an alleged evil of omission rather than an evil of
commission, this point is perhaps more naturally put the other way
around: “Had God always followed the policy in question, we
would likely not have come into existence. For such an important
input would have affected the ‘combinations of . . . people and
27. Ibid., p. 66.

28. Ibid., p. 67.
marriages,’ resulting in the existence of human beings other than

ourselves.” Either way, the conclusion which may seem to follow
is that since the evil of which we complain is in fact a necessary
condition of our existence, God cannot be said to have wronged us
in preventing a strong epistemic situation in relation to theism
from obtaining.
I find this argument unsatisfactory for more than one reason. I
have argued (in Chapter i) that what a perfectly loving God would
do cannot in all cases be determined by considering what are his
obligations. Thus, even if nothing which it would be wrong of
God to allow happens to me, difficulties may remain. But let us set
these on one side. It seems to me that the argument is in much
deeper trouble than this. As we have seen, Adams states that the
problem of evil must be understood in terms of God having (al¬
legedly) wronged us. This suggests that the alternative which the
one who argues from evil has in mind is a situation in which she
(or some other individual) is not wronged. On this view, the arguer
is claiming, in effect: “If God existed, I would exist and suffer no
evil of this kind.” But this claim entails that a situation in which
God exists and does no wrong is one in which the existence of the
complaining individual is preserved. It is for this reason that Adams’s
argument seems so effective. For if, as he claims, these evils must
exist if the arguer is to exist, then the arguer is not in a position to
claim that God would bring about a situation in which she exists
and suffers no evil. The situation she is saying God might be ex¬
pected to bring about is in fact impossible.
Adams’s (apparent) view of the position of one who argues from
evil is mistaken, however. It is mistaken, first of all, in supposing
the following claim to be coherent: “God exists and has wronged
me.”29 If, as Adams assumes, God is essentially perfectly good and
so cannot wrong anyone, this claim is in fact /^coherent. Further¬
more, why would anyone seeking to argue from evil to the nonex¬
istence of God make this claim? It is, after all, a claim entailing that
God exists. Perhaps Adams would reply that what he assumes to
be coherent and attributes to his arguer is not this, but rather
29. Adams suggests his acceptance of the coherence of this claim by arguing
against it at length.
something like: “An evil has befallen me which it would be wrong

of God to allow.” But if this is the arguer’s claim, it does not entail
that a situation in which God exists and does no wrong is one in
which the arguer’s existence is preserved. On this interpretation of
her claim, the arguer may, indeed, allow that if there were a God,
other human beings would exist instead of her.
My point here can perhaps be made clearer if we bring to bear
the distinction between de re and de dicto readings of propositions.
Consider the following proposition:
(4) If there were a God, all human beings would live a life
unblemished by evil e.
Adams’s arguer appears to accept a de re reading of this proposi¬

tion, expressible as follows:
(5) If there were a God, all human beings (that is, all of the
human beings who actually exist) would live a life un¬
blemished by e.
I have suggested, however, that the arguer from evil need not
claim this much. What she may accept instead is the following de
dicto reading of the proposition in question:
(6) If there were a God, all human beings (that is, all existing
human beings, whoever they might be) would live a life
unblemished by e.
If (5) is true, then, if God exists, I exist. But the same need not be
said about (6), and so (6) is not subject to Adams’s criticism. Hav¬
ing noted this, we can now consider:
(7) Josie Bloggs (a human being who actually exists) lives a

life blemished by e.
This proposition brings us back once more within the ambit of the
de re. But it is important to note that while it clearly follows from
(7) and (5) that there is no God, the same conclusion follows from
(7) and (6). That is, actual suffering can count against the existence
of a perfectly good God even if in a world created by such a God,

the sufferer would not have existed. Josie Bloggs, as suggested
above, need not exist for God to exist, and so she need not be
among the human beings mentioned by (6) who would come into
existence and live a life unblemished by e if there were a God. But
if Josie Bloggs does exist, it follows from (6) that unless she lives a
life unblemished by e, God does not exist. To put it another way:
if (6) is true, God does not coexist with human beings who live
lives blemished by e, and so if Josie Bloggs’s life is blemished by e,
God does not coexist with Josie Bloggs. We can therefore recom¬
mend to the arguer that she choose a proposition like (6) to express
her claim; she clearly is not limited to propositions like (5).
The application of this reasoning to the problem of inculpable
nonbelief should be apparent. The one who argues from the rea¬
sonableness of nonbelief claims that
(8) If there were a God, the epistemic situation of human be¬

ings in relation to theism would be a strong one.
He means by this, not the de re proposition,
(9) If there were a God, the epistemic situation in relation to

theism of actually existing human beings would be a
strong one,
but rather the de dicto proposition,
(10) If there were a God, the epistemic situation in relation to

theism of existing human beings, whoever they might be,
would be a strong one.
This, however, does not prevent him from adducing as a further

premise the de re proposition,
(11) The epistemic situation in relation to theism of the human

beings who actually exist is not a strong one,
Investigation, Diversity, and Responsibility [
and deriving
(12) God does not exist.
For although the human beings who actually exist need not exist
for God to exist, and so need not be among the human beings
mentioned by (10) who would come into being and benefit from a
strong epistemic situation if there were a God, if they do exist, it
follows from (10) that unless their epistemic situation is a strong
one, God does not exist. As (10) states, God and human beings in a
weak epistemic situation do not coexist, and so if actual human
beings exist in such a situation, God and actual human beings do not
coexist.
If this is correct, the Adams-type objection can be easily circum¬
vented. For then anyone can consistently claim both that the situa¬
tion of reasonable nonbelief in which she finds herself poses a
problem of evil, and that, had things been otherwise, she would
not have existed.
Perhaps it will be replied to my argument that I have missed the
point of the Adams article. Adams’s point, it may be said, is that if
we consider our existence to be on the whole a good thing, the fact
that conditions necessary for our existence have been satisfied must
also be viewed by us as a good thing, even if the satisfying of those
conditions can be seen by us to have involved the occurrence of
what is in itself evil. But to this it seems right to respond by saying
that the existence of individuals who would have come into being
had these evils not occurred and the necessary conditions of our
existence not been satisfied might also have been a great good. In¬
deed, where the evil in question is the existence of reasonable non¬
belief, the well-being achievable by them would have been much
greater than the well-being that actually existing individuals are in
a position to achieve. Would not God then have preferred their
existence to ours?
But a response to Adams need not rest on this point alone. (If it
did, it might be weakened by his claim that “God could be per¬
fectly good and . . . cause or permit evils that are necessary for
good ends that he loves, even if those goods are not the best states
of affairs obtainable by him.”30) We can point out, in addition, the

value God (if he exists) places on personal relationship for its own
sake. Adams speaks of the creatures God creates as creatures he
loves, but as the discussion in Part 1 indicates, if God loves, then
the creatures he creates will (for reasons of well-being and the in¬
trinsic value of personal relationship) exist in the presence of evi¬
dence sufficient for belief, other things being equal. And as an at¬
tempt to fill out this ceteris paribus clause, Adams’s argument
seems inadequate. It provides a good to compete with the good of
personal relationship with God only if our existence would have
some special value that the existence of individuals in a better posi¬
tion, epistemically, to relate personally to God would not have.
But it seems that this condition is not satisfied: there seems no
reason for God, in advance of our existing, to decide in favor of us
and attempt to steer things our way. 31 Indeed, Adams himself veers
from his chosen course at the last moment. After having suggested
that God might well have wished to bring us into being, and so
permitted the evils in question, he writes: “I am suggesting, in
effect, that the existence of creatures such as we are, with the charac¬
teristic, subtle, and sometimes bittersweet values and beauties of
human life, may also be a good of the relevant sort that is loved by
God” (my emphasis).32 But if in the end Adams wishes to say only
“creatures such as we are,” as apparently he does, and as it seems
he ought to do, then his whole argument collapses. For no reasons
have been given or could be provided for supposing that exactly
those evils that have occurred are necessary for the existence of
creatures such as we are; only for supposing they are necessary for
our existence. And if creatures other than ourselves but such as we
are could be as valuable as ourselves—as surely we must admit—
then given all the points made above, we must conclude that there
is here no rebuttal for the argument of Part 1.
30. Adams, Virtue of Faith, p. 72.

31. Indeed, the view suggested here seems slightly absurd. While it is one thing
to speak of the various evils with which history is littered as necessary conditions
of our existence, it is another altogether to claim that God has allowed these evils
with a view to bringing us—the humans beings who actually exist—into being.
32. Adams, Virtue of Faith, p. 72.
The Responsibility Argument
An explanation of reasonable nonbelief in terms of the value of

human responsibility has not been explicitly developed by anyone.
But the possibility of such an explanation is suggested by certain
passages in the literature and so must be considered. Swinburne,
for example, writes that “a God has reasons for creating a world in
which humanly free agents have deep responsibility for each
other—provided by the opportunity to harm each other in various
ways, e.g. by curtailing each other’s knowledge, freedom and
power, and the opportunity to benefit each other in converse
ways.”33 Swinburne gives special emphasis to the view that it is
good that humans have the power not just to benefit but also to
harm:
A world in which agents can benefit each other but not do each
other harm is one where they have only very limited responsibility
for each other. ... [If] I cannot harm you, you will be moderately
all right whatever I do. Your well-being will not then depend
greatly on me. God has reason for going beyond that. A God who
gave agents only such limited responsibility for their fellows would
not have given much.34
Taking these points into account, the critic may once again pro¬
ceed by analogy and suggest that there may well be opportunities
to harm or benefit which require for their existence that some indi¬
viduals for some period of time be (or possibly come to be) in a
state of reasonable nonbelief.35 We must now look at how such a
suggestion could be filled out.
(i) The most obvious development of this suggestion would in-
33. Swinburne, Existence of God, pp. 190-191.

34. Ibid., p. 189. There is an obvious connection between Swinburne’s points
about responsibility and his “choice of destiny” arguments: emphasized here are
deep and significant choices affecting other human beings. In more recent writing
Swinburne more explicitly connects the two discussions, referring to the value of a
“free and responsible choice of destiny.” See his “Knowledge from Experience,
and the Problem of Evil,” in William J. Abraham and Steven W. Holtzer, eds.,
The Rationality of Religious Belief (Oxford: Clarendon Press, 1987), p. 154.
35. The nature of the alternative state of affairs suggested here will be clarified
when we come to the second form of the Responsibility Argument.
volve the claim that it is a good thing that some individuals be in a

state of reasonable nonbelief so that others, more favored in this
respect, are given the opportunity to benefit or harm them by pro¬
viding them with evidence sufficient for belief, or else failing to do
so. In providing some human beings with this opportunity, God
gives to them a deep responsibility for the well-being of their fel¬
lows. He gives to them the power to assist them toward a personal
relationship with God, and all that entails, or to leave them in ig¬
norance of God’s existence; and it is a good thing that such respon¬
sibility be given.
But this explanation has an air of unreality about it. For it to
succeed, the opportunity in question must be one that some human
beings in our world in fact have, and if they are to have it, they
must be possessed of evidence that clearly supports God’s existence
and that honest inquirers will see to support God’s existence.36 But
this does not seem to be the case. The evidence that various indi¬
viduals have claimed to provide support for God’s existence is of¬
ten evidence that inculpable nonbelievers who fail to believe even
after long investigation and soul-searching have considered. Thus,
even if there is a God, the evidence of those who believe in him is
not clearly such as to render plausible the view that they have the
responsibility in question.
Despite this problem, let us suppose that there exists the possi¬
bility of responsibility of a sort. Perhaps some inculpable non¬
believers would have evidence sufficient for belief if they heard
what believers could tell them, even if others would not. The ques¬
tion we must then face is whether it is plausible to suppose that a
loving God would consider the value of the responsibility given to
convinced theists by the presence of inculpable nonbelief to be
great enough to allow it to occur.
One point which suggests itself is that since our world is one in
which human beings clearly already have great responsibility for
each other’s well-being, and since giving this spiritual form of re-
36. For surely, if God deprives A of evidence so that B can have the opportunity
of benefiting or harming A by bringing evidence to him or failing to do so, it is to
be expected that the evidence B is actually in a position to give A will be strong;
otherwise there is no sense in the supposition that B is in a position to benefit or
harm A.
sponsibility to some would prevent others from gaining access to

God, a loving God might be expected to refrain from giving this
additional responsibility. The claims of Swinburne and others
about the value of responsibility are large claims even when (as is
usually the case) claims about the value of spiritual responsibility
are not seen as entailed by them. Because this is so, and because of
the value that would go unrealized if God gave to some this further
responsibility, special reasons for supposing God to value it seem
clearly required.
We can go farther as well. For, having noted the many forms of
deep responsibility we have independently of the responsibility in
question, it may seem to us that a loving God would not wish to
withdraw his presence to facilitate additional responsibilities, but
would in fact wish to do the opposite: because the responsibilities
we have are so great, God, like a good parent, would wish to
make it possible for us to draw on the resources of a personal rela¬
tionship with him, in order to better enable us to fulfill our respon¬
sibilities.
To these arguments it may be replied that according to the ex¬
planation in question, the human beings to whom this extra re¬
sponsibility is given are ones who have become to some extent
aware of God’s presence and who therefore are in a position to
relate personally to God and to draw on the resources provided by
such relationship: those who are asked to share their evidence of
God’s presence with others must, logically, first have it them¬
selves. But it is hard to see how this point mitigates the difficulties
faced by the proponent of the Responsibility Argument. Let us
look once more at the responses outlined above. I have suggested
(in the second response) that a personal relationship with God
would help us successfully manage our responsibilities. I have not
suggested that it would make things easy, unless the arguments of
Chapter 5 can be successfully refuted, such a suggestion cannot be
upheld. Thus, even the individuals who according to the Respon¬
sibility Argument are given access to God cannot be viewed as
finding life unchallenging; far from it. And of course, either way,
those who inculpably fail to believe and so do not have such access
must be viewed as lacking an important resource. We can therefore
argue as follows. Since, even in the absence of the extra respon-
sibility suggested as appropriate for them by the Responsibility Ar¬

gument, very considerable responsibility-related challenges would
remain for humans given access to God, and since, were this re¬
sponsibility given to them, others would be prevented from expe¬
riencing such relationship and would suffer the lack of an impor¬
tant resource for the fulfilling of their responsibilities, a loving God
would not consider the value of the responsibility given to con¬
vinced theists by the presence of inculpable nonbelief to be great
enough to warrant allowing it to occur.
Other points can be adduced in support of this conclusion. First
of all, as was suggested in the discussion of Butler, even if all hu¬
man beings were to arrive at belief in a loving God, their under¬
standing of God and of his relation to the world would remain
imperfect, and so some of what they had yet to learn might well be
specially revealed by God. If God were to reveal himself further in
this way, then even if he wished to give to some human beings
additional spiritual responsibilities, he would not need to deprive
anyone of evidence for belief in his existence but could instead
leave it up to certain individuals historically and geographically
well-placed (and evidentially well-fortified!) whether the special rev¬
elation became generally known or not.37
It seems to me that there are reasons for supposing that, given a
choice, a loving God would in fact prefer to tie the spiritual respon¬
sibilities of human beings to the dissemination of a special revela¬
tion. For one thing, as I noted in Chapter 2, perhaps the most
religiously efficacious way of coming to believe in the existence of
a loving God is through one’s own experience; and so if there is a
God, we have some reason to suppose that he would provide hu¬
man beings with experiential evidence of his existence. Of course,
once the experiential foundation, so to speak, was laid, a variety of
ways of growing in spiritual understanding, of building on that
foundation, might be appropriate and religiously efficacious, in¬
cluding ways associated with the communications of other human
37. And given that God’s existence was well-evidenced, no one would be in a
position to claim that God should, for reasons of personal relationship, have made
this special information available to everyone. For as I have argued, personal rela¬
tionship with God is to be developmentally understood, and belief in God’s existence
is sufficient to enter upon such relationship.
beings in contact with God. To put this point another way: there is
some reason to suppose that a loving God would wish to provide
evidence for belief in his existence directly, not through some hu¬
man intermediary; for otherwise the acceptance of evidence would
perhaps not be as likely to evoke a personal response, and further
information about God, however provided, would not be as
readily assimilated and spiritually appropriated. A second reason to
suppose that God would prefer to tie the spiritual responsibilities of
human beings to the dissemination of a special revelation is that it
is possible (some would say likely) that many human beings would
fail in their spiritual responsibilities, or that the fulfillment of these
responsibilities would be long delayed. Taking this into account, a
loving God would surely wish those for whom others had been
made spiritually responsible to be provided with some basis for
communion with himself in the interim—a more general knowl¬
edge of God’s existence and loving nature which would allow
them to experience at any rate some of the benefits associated with
a personal relationship with God.
Much of what I have said about general and special revelation
and human responsibility for dissemination of the latter matches
traditional Christian thinking about God’s self-revelation and the
nature of the missionary task. It will be noted, however, that on
the traditional view, the all-important message of God’s love and
forgiveness is not a part of the general revelation but is intimately
connected to the story of Christ. And so it may seem that my
distinction between the two notions is incorrectly drawn. Indeed,
it may seem that since the evidence for God’s existence I have been
mentioning is evidence for the existence of a perfectly loving God,
there is really no distinction left at all: what I am asking for includes
the special revelation of Christianity (or at least its major theme),
and so there is no “further revelation” for me to appeal to as an
alternative source of spiritual responsibility.
To this I would reply by saying that I have been describing what
seems to be the policy a perfectly loving God would pursue, and if
this conflicts with traditional Christian thinking, so be it: tradi¬
tional Christian thinking on these matters may be internally incon¬
sistent. The message of a God of love Christians have (in the actual
world) tried to convey has sometimes not taken root in the experi-
ence of those to whom it was delivered because other concepts had

taken root instead. If there were a loving God, things would, it
seems, be otherwise; then, surely, this message would precede the
missionaries. Does this mean that there would be nothing left for
missionaries to convey? Would there then be no filling out of the
general revelation to which various individuals could contribute? It
seems not. In such a situation we might still, for example, be able
to speak of the responsibility of some to bring to others the news
of one who wonderfully exemplified the agapeistic character of
God, who provided the basis for a community of love in the body
of his teachings. And even leaving aside the traditional Christian
connection between a fuller knowledge of God and the person of
Christ, we could speak of the responsibility of those who had ex¬
perienced God’s presence and grown in interpersonal relationship
and in relationship with God to seek to enter into dialogue with
members of other communities and share their experiences and the
fruits of their spiritual labors. In this way all human beings could
(potentially) have spiritual responsibilities. We could go farther still
and suggest that if, as has been argued, some humans might very
well reject the evidence for God’s existence with which they were
provided, inculpable nonbelief need not exist in order for humans
to have “missionary” responsibilities. If belief could be rejected,
there would be opportunity for others who did not reject it to
admonish, exhort, awaken, or else fail to do so. There would be
nonbelief for believers to address, even if not inculpable nonbelief.
Given these points, we may add to the previous argument
(which is, I think, sufficient in itself to defeat this form of the Re¬
sponsibility Argument) the claim that various forms of spiritual
responsibility would in fact be facilitated by a strong epistemic situ¬
ation in relation to theism. Not only do humans have many forms
of deep responsibility. If God were self-disclosed, certain others
would be added to the list. There is therefore no reason to suppose
that a loving God would for reasons of human responsibility deprive
anyone of awareness of his presence. The value he would realize in
so doing would be far less great than the value he would give up.
(ii) A second version of the Responsibility Argument can also be
developed. It claims that humans have been given responsibility
for the spiritual well-being of future generations, and that some or all
present-day cases of inculpable nonbelief are due to past abuses of

this responsibility. Bad human choices over time—in particular,
the choices of humans to rebel against God and to reject the good
evidence given them by God—have gradually reduced the level of
religious awareness, resulting ultimately in a secular society in
which exist human beings who have never seen any need to attend
to the possibility of a religious dimension, beings who find the idea
of God alien, foreign, even ridiculous and so (in many cases at
least) inculpably fail to believe that there is a God. In theological
terms, this is the view that some of those who fail through no fault
of their own to believe in God are nonetheless inhibited by sin in
the (indirect) sense that they have inherited (biologically and/or
culturally) and now express involuntarily dispositions and values
inimical to belief.
In responding to this form of the argument, we may once again
wish to point out the difficulties faced by the one defending the
empirical claim on which it depends, namely (in this case), that at
some time in the past, humans had in their possession good evi¬
dence of the sort in question. It is unclear how we are to be led to
the conclusion that this claim is plausible. If it is plausible, it must
be plausible to suppose that there was a time when inculpable non¬
belief was impossible. But there seems no non-question-begging
way to defend this view; that is, it seems that any such view de¬
pends on the prior assumption of (at least) the plausibility of sup¬
posing God to exist, an assumption the argument of this book (to
which the Responsibility Argument is a response!) casts into ques¬
tion. For the empirical evidence seems to provide no reason to sup¬
pose that there have not always been individuals who have inculpa¬
bly denied the existence of God or remained agnostic or
unreflectively failed to believe, and so seems to provide no reason
to suppose that evidence of the required sort has ever been gener¬
ally available. Furthermore, in this case it will not suffice to say
that at any rate some individuals at some time knew of considera¬
tions that, although not obviously good evidence, would have
constituted evidence sufficient for belief for someone or other in a
future generation. If the evidence human beings had in their pos¬
session was not such as no one could inculpably reject, there is the
clear possibility of arguing that the present decline in belief is due
to the judgments upon careful investigation of certain individuals at

certain times that the evidence available to them was not support¬
ive of belief. And from this it follows that it is clearly possible that
individuals have always passed on the considerations that seemed to
them good evidence, but that reasonable nonbelief arose despite
this, which is to say that humans never had the responsibility in
question, for it was not in their power to ensure that individuals in
the next generation would not reasonably fail to believe: despite
their best efforts, reasonable nonbelief might come to exist. I con¬
clude, therefore, that there are serious difficulties in the way of
showing that the empirical claim in question is plausible. But if it is
not plausible, then the argument to which it belongs does not con¬
stitute a rebuttal.
We can argue not only that the claims of the second version of
the Responsibility Argument seem not to square with the facts, but
also that there is reason to suppose that the facts on which it de¬
pends would not be permitted to obtain by a loving God. Humans
already have deep responsibilities for the well-being of future gen¬
erations and regularly abuse them.38 If future generations exist, we
will have wronged them in many ways, for example, in our obses¬
sive concern for material goods, leading to poor management of
the fragile environment in which they will live. Given the deep
responsibilities we already have and our propensity to abuse them,
a loving God, concerned to enter into personal relationship with
human beings of all generations, recognizing the difficulties hu¬
mans of all generations must face, would, we might expect, not
give us the further responsibility in question. It is important to
note that nothing in this argument implies that the human beings
to whom God reveals himself in each generation should not be in a
position to reject God and a personal relationship with him for
themselves; by granting humans freedom, God makes it possible for
38. This fact is the one we must emphasize in response to those who would
claim that the opportunity to prevent others from coming to believe that there is a
God constitutes a much deeper and therefore much more valuable form of respon¬
sibility than those here mentioned. It would be a great responsibility. But the harm
that might well—and I should say, would probably—result, would be commen-
surately great, and so I suggest a perfectly loving God, concerned not only to give
human beings responsibility, but also to enter into personal relationship with
them, would rest content with giving other responsibilities.
them to reject him. But by the same token, it seems that a loving
God would not give anyone the opportunity to put others in a
position where neither explicit acceptance nor explicit rejection is
possible. God, if he exists, is concerned to make it possible for each
human being, at all times at which she or he exists and is capable,
to be personally related to himself, and will, other things being
equal, see to it that this is the case, unless that individual chooses
otherwise. Given the deep responsibilities humans could have even
given such a strong epistemic situation in relation to theism, the
second version of the Responsibility Argument does not seem to
me to be capable of adducing considerations strong enough to off¬
set this concern.
We may conclude this section with the following observation.
There is a tendency among some writers to value the giving of
freedom and responsibility almost limitlessly: their view seems to
be that if having some of these commodities is a good thing, then
more must always be better. But this is to forget that the context
for all theistic talk about these matters must be the love of God,
and that love not only grants freedom and responsibility, but de¬
sires personal relationship. A balanced view, it seems, must allow
for the realization of the relational aspirations of love, if their realiz¬
ation is compatible (as I believe it clearly is) with the giving of
much freedom and responsibility. Because of the nature of Divine
love, we may expect to find ourselves with freedom and respon¬
sibility, but by the same token, not in unlimited measure.
The Cognitive Benefits of Doubt
Most of the counterarguments we have considered suggest that

the cognitive deprivation represented by the reasonableness of non¬
belief is tolerated by God for the sake of some noncognitive good,
such as moral freedom or deep responsibility.39 The argument
taken up in this section chooses a different tack. It focuses on the
reasonableness of doubt, suggesting that the deprivation it repre¬
sents facilitates certain cognitive benefits: inculpable doubt, although
39. I say “most” because Kierkegaard’s Deception Argument, inasmuch as it

suggests that reasonable nonbelief would ward off false beliefs about God, can be
construed as referring to cognitive benefits.
it can be seen as depriving one of proper access to God, can also be

seen as creative of a fuller, deeper, more perfect understanding of
God and of God’s relation to the world.
This argument can be filled out as follows: “Inculpable doubt
involves a careful probing of issues surrounding the proposition(s)
over which one is in doubt. In the case of theism, doubt—if incul¬
pable—will result in a deeper understanding of the claim that there
is a God, and of the nature and proper conduct of the religious life.
The reasonable doubter, through long thinking combined with
deep religious concern, will clarify to herself (and perhaps also to
others) the meaning of such propositions as ‘God is good,’ ‘God is
loving,’ and so forth, and as a result of her investigations, will
grow to appreciate more fully the character and potential depth of
the life of faith. There is consequently a sense in which, even in the
midst of doubt, one may be coming to know God better. Though
as doubters we may complain that our knowledge of God is re¬
stricted, were God to reveal himself to us, we would have to say
that we knew him better than we otherwise could have; the per¬
sonal relationship with God into which we would then enter
would not have been hindered by our (previous) doubt but would
be the fuller and the richer for it. We have reason, therefore, to
suppose that God would wish inculpable doubt to obtain, and so
have reason to suppose that if there is a God, a strong epistemic
situation in relation to theism will not obtain.”
This is an interesting argument. Unlike many of the others I
have considered, it addresses directly the central difficulty repre¬
sented by the reasonableness of nonbelief, namely, its relationship-
inhibiting effects, and seeks to show that these can be mitigated.
Although it cannot be argued that while in a state of doubt the
doubter enters fully into personal relationship with God, the results
of her sojourn in the wilderness can be seen as contributing directly
to such a relationship by providing her with the‘deeper under¬
standing of spiritual things required for its flourishing. She may
arrive at a more adequate conception of the religious life just because
she is barred from fully entering into it herself. It may seem, there¬
fore, that we have here a good that God would wish to facilitate
even at the cost of the goods for which belief is required. Indeed, it
may seem that no real sacrifice would be incurred; for until a
deeper grasp of spiritual things is acquired, there can be no signifi¬

cant progress in the Divine-human relationship and so no real
movement toward the benefits of such relationship outlined in
Chapter i.
Must we say then that the argument from the cognitive benefits
of doubt is successful? Does it provide a rebuttal for the argument
of Part i? It seems to me that despite what has been said, the an¬
swer must once again be a negative one; although of considerable
interest, this argument ultimately fails.
Let me begin my defense of this claim by pointing out that in
order for the argument to succeed, reasonable doubt must be plau¬
sibly viewed as not just sufficient, but necessary, for the benefits to
which it refers. Even if God would value the deeper understanding
in question, if it could be achieved in other ways, he would not
(for reasons the reader should by now be able to supply) for its
sake subject anyone to doubt. But this condition does not seem to
be satisfied; reasonable doubt is apparently unnecessary for such a
deeper understanding. Although it involves a careful probing of
relevant issues, such investigation may also be the result of, for
example, experiences apparently of God. Surely many of those
who would experience God if our epistemic situation were in rele¬
vant respects improved would be motivated by their experience to
seek to know God more fully and to explore the various possi¬
bilities of the religious life. They would begin in a relationship
with God and grow in it. And so it seems that although there will
be progress in relationship with God only if one is developing a
deeper grasp of spiritual things, and although reasonable doubt
may facilitate this, the further claim that without doubt the relation¬
ship must stagnate cannot be upheld. Many would be properly
motivated by an experience of God, and so a sacrifice would in¬
deed be incurred if belief were postponed.
But it is possible for the proponent of the argument under con¬
sideration to make his claim more clear and precise, and so (appar¬
ently) to avoid this difficulty. It is indeed “many,” he may say,
who would be prompted by their experience of God to embark
upon a deeper quest, and not “all.” There are also individuals who
would take God and their experience of God largely or completely
for granted and settle down into an undemanding and relatively
shallow religious life; and it is only to such individuals that the

argument from the cognitive benefits of doubt applies. In their case a
period of investigation motivated by doubt is plausibly viewed as
essential to the achievement of the benefits in question, and in their
case no real sacrifice would be incurred if belief were to be post¬
poned.
This qualified version of the argument faces a serious problem,
however. For it would seem that the individuals in question, if
likely to respond inappropriately—in particular, without religious
zeal—to experiences apparently of God, could hardly be relied on
to conscientiously investigate the question of God’s existence in the
absence of such experience. Inculpable doubt, which requires careful
investigation, clearly implies just the sort of attentiveness and con¬
cern for the truth which the individuals in question are said to lack.
And if it is replied that doubt about God, unlike experience of God,
motivates to inquiry, we may point out that such doubt com¬
monly already involves a measure of religious concern. If the indi¬
viduals in question really would settle into spiritual indifference
upon experiencing God, we would expect (in the absence of clear
evidence for God’s existence) to find them not among the doubters
but in the company of unreflective nonbelievers, and so would not
expect them to be reaping any cognitive benefits of the relevant
sort at all. The proponent of the (clarified) argument therefore
faces a dilemma from which his argument apparently cannot es¬
cape unscathed: either the individuals in question would not in fact
benefit from the absence of clear evidence, because inclined to be
apathetic and religiously unconcerned, or (if they are alert and con¬
cerned) they do not require an interval of doubt, because likely to
be sufficiently motivated to inquiry by religious experience.
But perhaps the proponent of the (clarified) argument will reply
by suggesting that it is presumption that is at issue—presumption of
the sort likely sometimes to be inspired by religious experience,
even in individuals who would otherwise (in particular, if in
doubt) pursue careful investigation. But if this is the sort of reason¬
ing on which the argument depends, we are back to considerations
dealt with in Chapter 6. In effect, a version of Pascal’s Presumption
Argument is then being deployed to defend the view that doubt
may be necessary in some cases: growth in spiritual understanding,
Investigation, Diversity, and Responsibility f 203
it is suggested, is incompatible with the sort of presumptuousness

sometimes exhibited by those who have been granted proximity to
God. If, however, the Pascalian’s moves fail in Chapter 6, they
cannot be revived here. Humility is at least as likely to be the hu¬
man response to religious experience as presumption, and religious
humility is a close ally of investigative zeal. In any case, God
would not seek to predict our likely response and prevent all situa¬
tions in which inappropriate responses might occur, but would al¬
low us an explicit choice with respect to relationship with himself
and respect our decision, working within the parameters it estab¬
lished.
This response is greatly strengthened by a fourth, very power¬
ful, consideration available to the critic of the argument from the
cognitive benefits of doubt as clarified above. Suppose that what is
in fact quite unlikely is shown to be plausible, namely, that indi¬
viduals exist who would be properly motivated to religious inquiry
in the absence of clear evidence, but who would settle into pre¬
sumption if granted proximity to God. We may yet point out that
there is a sort of Divine withdrawal that is compatible with a
strong epistemic situation in relation to theism. What I have in
mind here is analogous to a state often referred to as “the dark
night of the soul”—a sense of God’s absence which severely tests
the believer’s faith and, in particular, works against the sort of easy
assurance and presumption in question.40 God may for a time leave
the believer in relative darkness, intending this to result in doubts,
even if not in doubt, that is, intending the believer to be troubled
by questions that shake her confidence and motivate her to exam¬
ine more closely the content of her belief.41 And surely such a pe-
40. Such an interruption in the experience of God’s presence need not result in
inculpable doubt because it is not necessary to lose evidence sufficient for belief in
order to be so tested. Although it may seem that if the primary evidence of God’s
existence is experiential, I must always sense God’s presence in order always to
have evidence sufficient for belief, this is not the case. I may be forced to depend
on past experience, to remain true to it; and this past experience, together with the
past and present experience of others and nonexperiential evidence, may be per¬
fectly sufficient for some degree of belief.
41. Now perhaps the believer will come to doubt, unreasonably. Does this
mean that, contrary to the argument of Chapter 1, doubt is compatible with a
personal relationship with God? I would say no, for the believer, since she has
riod of withdrawal would facilitate the same cognitive benefits as

are facilitated by doubt. Indeed, there is some reason to suppose
that God would prefer the former approach to the latter. The be¬
liever, having had experiences apparently of God, may have more
in the way of motivation because she senses a loss, and so may be
inspired to explore matters of faith even more deeply than the one
who is in doubt.42
We are therefore in a position to argue as follows: Given that, as
the previous argument shows, God would not second-guess hu¬
man responses to evidence of his presence but leave humans free to
explore more deeply the nature of faith or to sink into spiritual
nonchalance, we must expect that God would not subject to doubt
those who antecedently seem likely to fall into the latter category.
But this does not mean that there is no way for God to prompt
those who in fact do respond inappropriately to a deeper inquiry:
something analogous to “the dark night of the soul” after belief
may readily substitute for doubt prior to belief, and may indeed
have a stronger effect. Since this is the case, there is no reason to
suppose that God would, for the sake of its cognitive benefits,
leave anyone in a state of inculpable doubt, and so the argument
from the cognitive benefits of doubt fails.
responded presumptuously to her belief, may not have been in a personal relation¬
ship with God in the first place. But even if she was, if she doubts unreasonably
and so fails to believe or to experience the benefits of belief, the individual has
withdrawn herself from the relationship. (I assume here for the sake of discussion
that God exists.) It may seem that we should say she is still in the relationship, but
this, I think, is because the individual, while in doubt, is likely to be influenced in
various ways by her prior belief. Her life is still strongly affected by her having
been personally related to God, and indeed, she may once again become personally
related to God. We need not say that she still is so related to account for this.
42. The view I have been defending is well expressed, from the perspective of
theological commitment, by John Macquarrie: “As happens also in some of our
deepest human relationships, the lover reveals himself enough to awaken the love
of the beloved, yet veils himself enough to draw the beloved into an even deeper
exploration of that love. In the love affair with God (if we may so speak) there is
an alternation of consolation and desolation and it is in this way that the finite
being is constantly drawn beyond self into the depths of the divine” (In Search of
Deity [London: SCM Press, 1984], p. 198).
A Cumulative Argument
We have seen that available counterarguments fail to show the

plausibility of supposing that a perfectly loving God would allow
the occurrence of reasonable nonbelief. There is still one last re¬
sponse we must consider—a cumulative argument that can only be
clearly formulated now that the claims of each of the other argu¬
ments have been brought out into the open. It suggests that if we
view the good to which the good of belief is opposed as the con¬
junction of the various states of affairs for which the occurrence of
reasonable nonbelief is necessary, it will become quite obvious to
us that we have a good greater than or equal to the good sacrificed
by allowing reasonable nonbelief to occur, and (given that reason¬
able nonbelief is logically necessary for it) that the argument from
the reasonableness of nonbelief can therefore be rebutted. Which
are these states of affairs? Given the conclusions reached in the rele¬
vant chapters, they would seem to be the following: (i) the possi¬
bility (for many) of a Kierkegaardian subjectivity of uncertainty; of
venturing, risking all, in the search for God; (ii) the possibility (in
many cases) of choosing to investigate God’s existence (or failing
to do so) in the absence of any clear reason to believe that there is a
religious reality at all; (iii) the existence of the actual religious tradi¬
tions, that is, Hinduism, Buddhism, and so forth; (iv) the existence
of those human beings who actually exist; and (v) the respon¬
sibility (for some) of transmitting evidence sufficient for belief to
individuals who would otherwise lack it and so lack the benefits
made possible thereby.43 Is the conjunctive state of affairs repre¬
sented by (i)-(v) plausibly viewed as having a value greater than
(or alternately, does it have a value clearly equal to) the value that
is sacrificed in allowing reasonable nonbelief to occur?
The correct answer to this question, it seems, is no. We may
point out, first of all, what various arguments in Part 2 have al¬
ready suggested, namely, that quite apart from facilitating personal
relationship with himself, in providing evidence for all, God would
ipso facto bring about the existence of various states of affairs that
43. As I suggested earlier in this chapter, it seems less than plausible to suppose
that this state of affairs obtains, but for the sake of argument, I include it.
would otherwise not have obtained, many of which are broadly of

the same type as those the objector is (correctly) claiming would
have to be given up in order for such evidence to be provided. Let
us briefly compare these states of affairs and the value they would
bring into the world with those mentioned by the objector, re¬
membering that the latter are incompatible with the former—that
the former states of affairs and the value they represent are in¬
cluded in what is sacrificed in allowing reasonable nonbelief to oc¬
cur. The alternative list looks like this: (vi) the possibility for all
(and so also for those who would otherwise have been in a position
to develop the subjectivity of uncertainty) of cultivating the in¬
wardness of belief; of striving to know God more fully; (vii) the
possibility for all who did not resist the evidence with which they
were provided of pursuing questions concerning the nature and
activity of a God already believed to exist; of pursuing (to the ex¬
tent that circumstances and abilities allowed) the way of “faith
seeking understanding”; (viii) the existence of religious traditions
instantiating religious creativity, imagination, and the like, no
doubt differing in many respects from those that actually exist, but
perhaps in many cases more closely attuned to the truth; (ix) the
existence of creatures other than ourselves but “such as we are”;
and (x) the opportunity for those who would have failed to believe,
had circumstances been otherwise, to grow in relationship with
God, to share, or fail to share, the fruits of their spiritual labors
with others, and to awaken, or fail to awaken, others to a realiza¬
tion of the importance of faith.
What can we say about the comparative values of the conjunc¬
tive states of affairs represented by (i)-(v) and (vi)-(x)? Obviously
such judgments are difficult to make. But this, I would argue, is
just the point we should emphasize: it seems that there is little to
choose between the two alternatives; that their value is approx¬
imately equal.44 And I would go on to point out that we have
reached this conclusion without taking account of (xi) the possibility
for all at all times of personal relationship with God and the great
instrumental and noninstrumental value this represents—which
44. Further support for this conclusion is to be found in the detailed arguments
of Chapter 6 and previous sections of Chapter 7.
surely tips the balance decisively in favor of the second list. That is,
the conjunctive state of affairs represented by (vi)-(xf) (for the ex¬
istence of which the absence of reasonable nonbelief is necessary)
seems quite clearly the state of affairs we should (and a loving God
would) prefer. A perfectly loving God would not be deterred by
considerations of the sort the objector can marshal from bringing
about, as he would naturally be inclined to do, the state of affairs
represented by (xi), realizing that what would be sacrificed in so
doing could in large part be made up in other ways.
So it seems that a cumulative argument must also fail. Indeed, in
putting it forward, the objector helps bring the larger picture more
sharply into focus, enabling us to see more clearly the force of the
original claim.
Conclusion
Our investigations into the question of reasonable nonbelief have

come to an end. As we have seen, in the case of each of the argu¬
ments discussed in the previous three chapters, it is possible to
show that there is no logical connection of the required sort be¬
tween the permission of reasonable nonbelief and the good offered
for our consideration, and/or that the state of affairs said to be of
great value does not in fact constitute an outweighing or offsetting
good, and/or (in one or two cases) that the state of affairs is not
plausibly viewed as obtaining. Since, as we saw in Chapter 4, all
the conditions here mentioned (the actual existence of the state of
affairs, the logical necessity of reasonable nonbelief for its exis¬
tence, and its status as an outweighing or offsetting good) must be
plausibly viewed as satisfied (and in the case of the existence of an
offsetting good, clearly satisfied) if there is to be a rebuttal, we may
conclude that there is no rebuttal among the available counterargu¬
ments. That is, premise (2) of the argument from the reasonable¬
ness of nonbelief, as stated in Chapter 4, can apparently withstand
all available challenges, and so there seem to be good grounds for
supposing it to be true, and hence for concluding that the argu¬
ment is sound.
How ought believer and nonbeliever to regard this conclusion?
The answer, it seems to me, is likely to be different for different
Conclusion [ 209
cases, depending on the extent to which the various claims I have

made are accepted and on whether the fifth assumption of the In¬
troduction, namely, that the relevant evidence (exclusive of evi¬
dence adduced in this book) does not clearly favor either theism or
atheism, is accepted or denied. Let us suppose, however, that my
claims are fully accepted by some individual S. What should be the
effect of this on the belief of S concerning God’s existence? Con¬
sider first the situation in which 5 takes the relevant (independent)
evidence to strongly favor the claim that there is a God (perhaps
she has had powerful religious experiences or considers herself to
be in possession of a simple deductive proof of God’s existence).
Must S, to be rational, deny or come to doubt the existence of
God because of the undefeated argument from the reasonableness
of nonbelief? It may seem not. For S’s situation would appear to be
analogous to that of Jim, who believes that he has just been run¬
ning with an acquaintance Joe in the city park and then is told by
very reliable witnesses that Joe died in a plane crash the previous
day. Surely the testimony must be denied in this case. Jim, it
seems, must believe that the apparently reliable witnesses are in
fact deceitful or deceived. And something similar, it may seem,
must be said about S. S must believe that there are countervailing
considerations even though she can find none, because of the force
of her other evidence for God’s existence (i.e., for the denial of the
conclusion of the argument from the reasonableness of nonbelief).
But consider the following expansion of this analogy. Suppose
that Jim, out of curiosity, travels to the site of the plane crash,
reads the local newspapers, which mention Joe’s name in connec¬
tion with the crash, talks to more witnesses, including members of
Joe’s family, who tell him Joe has indeed died, and finally goes to
the funeral chapel and sees what is to all appearances the body of
Joe. At some point the conflicting considerations must put him (at
least!) in a state of doubt as to whether it was really Joe he was
with in the park, and we would surely be inclined to say that any
other response to the evidence would be unreasonable. What does
this suggest concerning the rationality of S continuing to believe
that there is a God? It seems to me to suggest that if S’s other
evidence remains the same, only up to a certain point will it be
21 o ] Conclusion
reasonable for S to say, because of the force of her other evidence,

that there must be a theologically acceptable explanation of the oc¬
currence of reasonable nonbelief even though she can find none.
Here too there may come a time when it suddenly becomes clear
that the evidence provided by the occurrence of reasonable non¬
belief is at least as weighty as contrary evidence. This would be the
case, for example, if S came to believe very strongly that if there
were a good served by reasonable nonbelief, this information
would be available to her. Such a belief might result from addi¬
tional evidence suggesting that any good served by reasonable
nonbelief would be logically possible to know; that God, if he ex¬
ists, would reveal to earnest inquirers any good served by reason¬
able nonbelief; that the relevant moral facts were known to S, and
so forth. Beliefs of this sort, if formed by S, might very well cause
her to view her other evidence with some suspicion, or simply lead
to perplexity about which evidence was weightier. Either way, just
as in the case of Jim and Joe, doubt could at some stage be the
reasonable result. Now of course, both in the case of 5 and in the
case of Jim and Joe, the contrary evidence may not remain con¬
stant. Jim may, a few days after the crash, bump into a living,
breathing Joe while running in the park, and come (reasonably) to
believe it was all a hoax, and S may happen on yet another appar¬
ent proof of God’s existence! The point remains, however, that
even for one who considers herself to be in possession of extremely
good contrary evidence, the argument from the reasonableness of
nonbelief, if undefeated, must continue to exert a certain evidential
pressure and cannot be ignored.
This conclusion holds a fortiori for anyone who is less strongly
convinced on independent grounds that there is a God. In particu¬
lar, the many who believe that the relevant (independent) evidence
does not clearly favor either theism or its denial must, if my argu¬
ment seems to them to be undefeated by available counterargu¬
ments, come to believe that there is no God. In their case, there is
little for the new evidence for atheism to overcome. The situation
of an individual 5 who is a member of this group is comparable to
that of Kim, who catches a glimpse of a person in the corner of the
park who appears to be Flo, but might very well not be, and then
is told by reliable witnesses that Flo died in a plane crash the pre-
Conclusion [ 2 i i
vious day.1 Kim, if reasonable, will surely judge on the basis of her
total evidence that it was probably not Flo whom she saw. In this
case, there is no significant presumption for the new evidence to
defeat. And if Kim goes to the park again the next day and has the
same experience—that is, again has a fleeting glimpse of an indi¬
vidual resembling Flo—she, unlike Jim, who bumps into Joe and
sees him face to face, will judge, if reasonable (and if no other
evidence has come to light), that the individual she briefly sees is
probably not Flo. Although here too there are possible alternative
explanations of the witnesses’ testimony, given that her evidence
for believing that it was Flo she saw is not strong in the first place,
Kim has no reason to appeal to such possibilities. So too in the case
of S, if she finds herself without other reasons that clearly suggest
the existence of God. Without strong independent evidence for the
denial of my argument’s conclusion, S has no reason to appeal to
the possibility of an explanation unknown to her, perhaps beyond
her grasp. Without the indirect support for an appeal to the possi¬
bility of unknown explanations afforded by strong independent ev¬
idence of God’s existence, S must, if she agrees with this argu¬
ment, come to believe that there is no God.
My conclusion has important implications for a number of posi¬
tions and arguments in contemporary philosophy of religion. I will
mention two here—the positions of agnosticism and noneviden-
tialism.2 Nonevidentialism was originally and most famously en¬
dorsed by Pascal.3 On this view, one is pragmatically (as opposed
1. The analogy might be a closer one if we said that Kim is informed that had it
been Flo, she would have caught more than just a glimpse—that Flo likes to walk
out in the open where everyone can see her! But it is the point about independent
evidence I am emphasizing here, and for that, the present analogy will suffice.
2. The second of these positions, albeit familiar, does not seem to go by any one
name. I call it “nonevidentialism” to contrast it with a view popularly known as
“evidentialism”—the view that belief is rational only if based on adequate evi¬
dence.
3. I am of course referring to Pascal’s famous “wager” argument. See his Pen-
sees, A. J. Krailsheimer, trans. (Harmondsworth: Penguin, 1966), fragment 418.
William James, while disagreeing with Pascal on specifics, reaches a similar conclu¬
sion. See William James, The Will to Believe and Other Essays in Popular Philosophy
(New York: Dover Publications, 1956). For more recent restatements of this view,
see Stephen T. Davis, Faith, Skepticism, and Evidence (Cranbury, N.J.: Associated
University Presses, 1978); Robert M. Adams, “Moral Arguments for Theistic Be-
212] Conclusion
to evidentially) justified in believing that there is a God if one has

legitimate ends the pursuit of which is facilitated by such belief,
and if the question of God’s existence cannot be settled on eviden¬
tial grounds. Now it will be apparent that if the argument we have
considered succeeds, nonevidentialism is, if not false, irrelevant.
For if that argument is correct, “the question of God’s existence
cannot be settled on evidential grounds” represents an impossible
state of affairs, and so what is according to nonevidentialism a nec¬
essary condition of justifiably believing in God on pragmatic
grounds can never be realized. Any apparent inconclusiveness in
the evidence must, if that argument succeeds, itself be taken as a
consideration (evidentially) justifying the conclusion that God does
not exist.
Agnosticism would seem to be in the same boat. The agnostic
says that when the evidence appears inconclusive, one must sus¬
pend judgment—that in such circumstances, neither theism nor
atheism is a rational belief. But again, if the argument we have
considered succeeds, this position must be rejected as, at best, irrel¬
evant. For if that argument is correct, individuals who find the
evidence inconclusive (and hold the argument to be correct) must
ipso facto become atheists: the weakness of theistic evidence—a
fact they will take to be confirmed in their own experience—must
in their case be viewed as itself a consideration that tips the balance
in favor of atheism. Hence neither the pragmatic leap nor the ag¬
nostic slump is justified for anyone who accepts the argument
from the reasonableness of nonbelief as sound.
Of the various points that strengthen the argument of this book,
one stands out, and it may be useful, in drawing the volume to a
close, to bring it more clearly into focus. This point is the follow¬
ing: the reasons for Divine self-disclosure suggested by reflection
on the nature of love are not reasons for God to provide us with
some incontrovertible proof or overwhelm us with a display of
Divine glory. Rather, what a loving God has reason to do is pro-
lief,” in C. F. Delaney, ed., Rationality and Religious Belief (Notre Dame: Univer¬
sity of Notre Dame Press, 1979); and Nicholas Rescher, Pascal's Wager (Notre
Dame: University of Notre Dame Press, 1985).
Conclusion [213
vide us with evidence sufficient for belief One of the consequences

of this is that moral freedom, perhaps the most important good we
have discussed, need not be infringed in order for God to be dis¬
closed in the relevant sense. If this were not so, we could add to
the list of goods for which the occurrence of reasonable nonbelief is
necessary, “the possibility of morally free choice,” and in that case,
the argument from the reasonableness of nonbelief would be de¬
feated: if God’s hiddenness were required for the preservation of
moral freedom, we might indeed expect a hidden God. But as it is,
the kind of revelation that would remove moral freedom need not
at all be associated with an argument of the sort presented here.
That various writers have been little exercised by the relative
weakness of theistic evidence is, it seems to me, largely due to a
failure to recognize this point. To put it another way: the relevant
(epistemic) alternatives to our actual situation have not been prop¬
erly delineated. By specifying, as I have, what we might expect
our epistemic situation to be like if a loving God exists, I hope to
have set up a standard by which various attempts to solve the
problem of weak theistic evidence can be evaluated. Many such
attempts seem only to show that some overwhelming manifesta¬
tion is not to be expected, and so provide no reason at all to sup¬
pose that God should not be self-disclosed when the latter concept
is more carefully and adequately defined.
I stated in the Introduction that it was my aim in this book to
show the importance of the argument from the reasonableness of
nonbelief for the philosophy of religion. This, I think, I may fairly
claim to have done. As we have seen, it is an implication of our
discussion that individuals who doubt or weakly believe must, if
they accept this argument, come to believe that there is no God,
and that even those who consider themselves to be in possession of
strong independent grounds for belief in the existence of God
ought to take it seriously, and seek to answer it or to acquire addi¬
tional grounds for belief. I myself have been unable to find reason
to suppose that it is so much as plausible that a perfectly loving
God would be hidden, and so the prospects for a future counter¬
argument that would remove this threat to theism and revive the
possibility of belief must appear dim (however much one may feel
214 ] Conclusion
that the existence of such an argument, given the profundity of the

notion of Divine love, would be an immensely good thing). But it
may be that new evidence will turn up. Perhaps stronger counter¬
arguments can be devised. I hope that because of my efforts, others
will be inspired to show that this is indeed the case.
Index
Adams, Marilyn M., 25-26 Cognitive benefits of doubt, argument

Adams, Robert M., 17, 20, 22, 30 n., from the, 199-204
63-64, 155 n., 162-163, 184-190, Cognitive freedom, 97-107, iio-m
211 n. Cohen, L. J., 30 n.
Agnosticism, 67, 212
Alston, William P., 1 n., 35, 48 n., Dark night of the soul, 203-204
51 n., 55-56, 57 n., 72 Davis, Stephen, 72, 211 n.
Dilley, Frank B., 1 n., 98
Barth, Karl, 4, 24 n. Divine hiddenness, 1,3, 28-29, 96,
Belief: categorical, 31 n., 32; defined, 105, 168, 170; and evidence suffi¬
30; and evidence, 33, 60-62; graded, cient for belief, 145; and the Fall,
31-32; involuntary, 9-10; and prob¬ 132—133, 137-138, 146-147; as pa¬
ability, 35-38 tience, 144, 151, 156 n.; problem of,
Belief in God: and belief in other reli¬ 5-6; types of, 4
gious propositions, 40-41; entailed Divine hiddenness, reasons for: in But¬
by personal relationship with God, ler, 168-180; in Hick, 96-115; in
30, 203 n.; and evidence, 33-40; im¬ Kierkegaard, 153-167; in Pascal,
plicit, 41-43; and religious experi¬ 132-152; in Swinburne, 116-130.
ence, 101-102, 104, 106-107, 109, See also Cognitive benefits of doubt;
112-115. See also Religious experi¬ Reasonable nonbelief: and actual
ence existents; Religious diversity; Re¬
Berkhof, Hendrikus, 22 n. sponsibility
“Beyond reasonable nonbelief.” See Divine love: and the concept of God,
Strong epistemic situation in relation 10-11; as entailing benevolence, 17;
to theism and the fulfillment of obligation,
Butler, Joseph, 168-180, 194 26-27; and human freedom, 27-28,
38; as involving concern to facilitate
Calvin, John, 74-75 belief in Divine existence, 30-43;
Calvinian view of nonbelief, 74—82 and the seeking of personal relation-
216] Index
Divine love (cent.) gious experience, 109; as require¬

ship, 17-29, 199; as universal, 23- ment of reasonable nonbelief, 61-64;
24. See also Belief in God; God; Per¬ as in some cases not knowingly ne¬
sonal relationship with God glected, 65-69. See also Cognitive
Divine self-disclosure, 145-147, 212- benefits of doubt
213
Doubt: defined, 60; as (sometimes) in¬ James, William, 211 n.
culpable, 64-69; inculpable, criterion Jantzen, Grace, 29
of, 60-64; and sin, 74-82. See also Jiingel, Eberhart, 69
Cognitive benefits of doubt; Reason¬
able nonbelief Kant, Immanuel, 118 n.
Kellenberger, J., 76 n.
Epistemic parity between propositions, Kenny, Anthony, 67 n., no
59-60, 64-65 Kierkegaard, Soren, 152-167, 199 n.
Epistemic rationality, 60-64 Kuhn, Thomas, 70
Evidence: defined, 33; public and pri¬
vate, 71-73. See also Belief; Belief in Lindbeck, George, 70
God Lucas, J. R., 180 n.
Evil, problem of, 6-9, 47-48, 86, 184- Luther, Martin, 5
189
McClendon, James W., 70
Feenstra, Ronald J., 76 n. Mackie, J. L., n n., 12 n.
McKim, Robert, 1 n., 98
Gale, Richard, 56 n., 73 n. Macquarrie, John, 27, 45, 70, 204 n.
Gilkey, Langdon, 71 n. Mavrodes, George, 51 n., 72-73
God: idea of, 10-n; as seeking per¬ Mesle, C. Robert, 1 n., 98
sonal relationship for its own sake, Miracles, 2, 47-48
21-23, l77 n., 190. See also Divine Mitchell, Basil, 20
hiddenness; Divine love; Personal Moral freedom, 134-135, 151-152,
relationship with God; Religious 213; arguments from the value of,
experience 96-130; and cognitive freedom,
Goldman, Alvin, 31 n. 105-107, 110-112, 134-135; and de¬
Gutting, Gary, 5m., 55 sire for Divine approval, 126-129;
and expectation of punishment for
Hannay, Alistair, 157 n. bad actions, 117-118, 122-125; as
Hebblethwaite, Brian, 19-21 genuine choice of destiny, 116; lim¬
Hepburn, Ronald, 1 n., 67 n., 73 n. ited, 116, 117 n.; and supererogatory
Hick, John, 1 n., 27, 70 n., 72, 96- action, 130; and temptation, 116,
117, 134-135, 151 n., 152 n.; rela¬ 118-119. See also Self-deception
tion to Kierkegaard, 160; relation to Morris, Thomas V., 1 n., 70-71,
Pascal, 104 n., 134-135, 151 n., 140 n.
152 n.
Hume, David, 76, 78 Nonbelief: defined, 58; types of, 58-59
Nonevidentialism, 211-212
Inscrutability of God-purposed goods, Nygren, Anders, 17 n.
88-91
Intellectual probation, 168-180 Offsetting good, defined, 86
Investigation, 172 n., 180 n.; and O’Hear, Anthony, 1 n., 98
moral freedom, 104, 108; and reli¬ Outweighing good, defined, 86
Index [217
Pannenberg, Wolfhart, 24 n. Rescher, Nicholas, 212 n.

Pascal, Blaise, 68, 132-152, 154, 158, Responsibility, argument from the
202-203, 211-212 value of, 191-199
Paul, St., 129 Revelation, evidence of, 178-179, 194-
Pelikan, Jaroslav, 5 196
Penelhum, Terence, 1 n., 6-7, 110- Rowe, William L., 86 n., 90
iii, 113, 136, 140, 147, 160 n., Russell, Bruce, 86 n., 88 n., 89 n.
171-172, 177-178
Personal relationship with God: as be¬ Schlesinger, George, 1 n., 4 n., 76 n.
ginning in this life, 25-29; benefits Self-deception, 38-39, 53, 56-57, 75~
of, 19-21; as developmental, 28; and 76, 78-79; and human freedom, 27-
moral freedom, 27-28, 105-108, 28, 103-104, 110-112, 115, 117,
198-199; and salvation, 29. See also 123-124, 128; and indirectly volun¬
Belief in God; Divine love; Reli¬ tary belief, 10 n.; and investigation,
gious experience 62, 65-68
Plantinga, Alvin, 8, 9 n., 33 n., 53, Subjectivity, 153-156, 158-161, 164-
63-64, 72, 74-75 167; and faith, 155, 157-158, 162-
“Plausible,” stipulatively defined, 84 164. See also Religious experience
Pojman, Louis, 153 n., 154 Strong epistemic situation in relation
Price, H. H., 11 to theism: arguments for impos¬
Prima facie case, defined, 84 sibility of, 45-46; defined, 39; possi¬
bility of, 44-57. See also Religious
experience
Rahner, Karl, 41-43, 69 Swinburne, Richard, 10, n n., 12 n.,
Reasonable nonbelief: and actual exis- 36-38, 46, 49, 51 n., 53-55, 60-63,
tents, 184-190; actual occurrence of, 64 n., 72 n., 96-97, 116-130, 151,
58-82; conditions of God’s permit¬ 152 n., 169-170, 172, 175 n., 191-
ting, 85-86; defined, 3 n.; problem 193; relation to Butler, 169-170,
of, 2-3, 83; and problem of evil, 6- 172 n., 175 n.; relation to Pascal,
9. See also Doubt; Weak epistemic 151 n., 152 n.
situation in relation to theism; Weak
theistic evidence Talbot, Mark R., 1 n., 75-82
Rebuttal of prima facie case, conditions
of, 84-88 Vanstone, W. H., 18
Religious ambiguity, 70-71, 97, 107-
108 Ward, Keith, 24 n.
Religious diversity, arguments from Weak epistemic situation in relation to
value of, 181-184 theism: defined, 39-40. See also Rea¬
Religious experience: as evidence for sonable nonbelief; Weak theistic evi¬
belief, 33“34, 48-57, H7-I50, 194- dence
195; and the experience of evil, 52, Weak theistic evidence, 1-3; and Di¬
56-57, 113; and God’s action, 50; as vine justice, 3 n. See also Reasonable
not inordinately intrusive, m-112; nonbelief; Weak epistemic situation
and a presumptuous response, 150- in relation to theism
152, 202-203; and progress in rela¬ Wiles, Maurice, 50 n.
tionship with God, 48, 201-204; and Williams, Bernard, 10 n.
subjectivity, 166-167. See also Belief Wolterstorff, Nicholas, 37-38
in God Wykstra, Stephen, 88-91
Cornell Studies in the Philosophy of Religion
Edited by William P. Alston
God, Time, and Knowledge

by William Hasker
On a Complex Theory of a Simple God

by Christopher Hughes
Time and Eternity

by Brian Leftow
Theology in the Age of Scientific Reasoning

by Nancey Murphy
Divine Hiddenness and Human Reason

by J. L. Schellenberg
The Nature of God

by Edward R. Wierenga
Library of Congress Cataloging-in-Publication Data
Schellenberg, J. L., 1959-

Divine hiddenness and human reason / J. L. Schellenberg.
p. cm. — (Cornell studies in the philosophy of religion)
Includes index.
ISBN 0-8014-2792-4 (alk. paper)
1. Theism. 2. Hidden God. 3. God—Knowableness. 4. Knowledge,
Theory of (Religion) I. Title. II. Series.
BL200.S34 1993
212'. 6—dc20 92-32633

(Cornell Studies in The Philosophy of Religion) ) J.L. Schellenberg - Divine Hiddenness and Human Reason-Cornell University Press (2006)

Uploaded by

Copyright:

Available Formats

(Cornell Studies in The Philosophy of Religion) ) J.L. Schellenberg - Divine Hiddenness and Human Reason-Cornell University Press (2006)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Cornell Studies in The Philosophy of Religion) ) J.L. Schellenberg - Divine Hiddenness and Human Reason-Cornell University Press (2006)

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What arguments does the book make?

What arguments does the book make?

Divine Hiddenness

and Human Reason

Cornell Studies in the Philosophy of Religion

Cornell University Press, Ithaca and London

Copyright © 1993 by Cornell University

First published 1993 by Cornell University Press.

International Standard Book Number 0-8014-2792-4

© The paper in this book meets the minimum requirements

Part i Framing the Argument

Part 2 The Force o f the Argument

I am indebted to a number of individuals and institutions for the

ities Research Council of Canada. Finally, I thank my wife,

Many religious writers, sensitive to the difficulties in which our

i. Several contemporary writers have touched on this problem. See Terence

offered suggestions as to how the weakness of theistic evidence might yield an

occurs.2 Hence we can argue from the weakness of theistic evi¬

2. Nonbelief is reasonable, I will stipulate, if and only if it is not the result of

and reformation theologians affirmed that God’s existence could be

(1) God exists and has intentionally withheld (or permitted to

But (1) entails that God exists, and so it would be contradictory,

to interpret talk of God’s hiddenness as implying a commitment to

8. Of course, the theological problem is closely connected to ours inasmuch as

(2) If God exists and is perfectly loving, reasonable nonbelief

is necessarily true, or equivalently, that

(3) God exists and is perfectly loving

is logically incompatible with

(4) Reasonable nonbelief occurs.

9. Terence Penelhum, “Divine Goodness and the Problem of Evil,” Religious

And in reply to such a claim, it would surely be tempting to argue,

human beings. Accordingly, the problem of reasonable nonbelief,

(2') a includes the state of affairs consisting in God’s being perfectly

what we “believed” was true and so would not really believe.12

argument defeats the case that I present, my claim should be con¬

Framing the Argument

We saw in the Introduction that the concept of God is the con¬

1. The classic treatment of the concept of “agape” is to be found in Anders

connection and to develop certain of its implications.3 In particular,

A Neglected Feature of Divine Love

“God seeks to be personally related to us.” In claiming that this

Forgiveness, reconciliation, peace and justice sound much the same

Basil Mitchell and Robert Adams argue similarly:

That a personal relationship with God would contribute ethically

5. Brian Hebblethwaite, The Ocean of Truth (Cambridge: Cambridge Univer¬

suggests, it is arguable that even the experiential aspects of our

8. Hebblethwaite, Ocean of Truth, p. 15.

personal relationships in question not merely because they would

The contrast between Agape and Eros is popularly seen as a special

God’s love for us is surely seen as involving a desire for certain

If Adams’s claims are correct—and it seems to me that they are—

10. Adams, Virtue of Faith, pp. 187-188.

One thing is sure, that there is no theological justification for setting

Since God is the one God of all beings, he can be no respecter of

The Divine love is limitless. . . . [It] extends to all creatures impar¬

The view defended by the theologians cited seems clearly cor¬

reason, environmental or genetic, seems at some time utterly inca¬