In "Reason and Belief in God," I suggested that such propositions as:

1. God is speaking to me.
2. God disapproves of what I have done, and
3. God forgives me for what I have done.

are properly basic for at least some believers in God; there
are widely realized sets of conditions, I suggested, in which such
propositions are indeed properly basic. And when I said that these
beliefs are properly basic, I had in mind what Quinn calls the narrow
conception of the basing relation.[1] I was taking it that a person S
accepts a belief A on the basis of a belief B only if (roughly) S
believes both A and B and could correctly claim (on reflection) that B
is part of his evidence for A. S’s belief that there is an error in
some argument against p will not typically be a belief on the basis of
which he accepts p and will not be a part of his evidence for p.[2]

This is important for the following reason. In arguing that belief in
God is properly basic, I meant to rebut the claim made by the
evidentialist objector: the claim that the theist who has no evidence
for theism is in some way irrational. What the evidentialist objector
objects to, however, is not just believing in God without having a
response to such objections to theism as the argument from evil. He
concedes that the theist may perfectly well have an answer to that
objection and to others; but as long as she has no evidence for the
existence of God, he says, she can’t rationally believe. As the
evidentialist objector thinks of evidence, then, you don’t have evidence
for a belief just by virtue of refuting objections against it; you must
also have something like an argument for the belief, or some positive
reason to think that the belief is true. I think this conception of
evidence is an appropriate conception; but in any event it is the
relevant conception, since it is this conception of evidence that the
evidentialist objector has in mind in claiming that the theist without
evidence is irrational.

As I see it, then propositions like (1) – (3) are properly basic for
many persons, including even such intellectually sophisticated adults as
you and I. Quinn disagrees: " . . . I conclude that many, perhaps most,
intellectually sophisticated adult theists in our culture are seldom if
ever, in conditions which are right for propositions like those
expressed by (1) – (3) to be properly basic for them."[3] Why so? I
think Quinn is inclined to agree, first, that there are conditions in
which such beliefs are properly basic for a person; such conditions
might be those of a child brought up by believing parents, or perhaps of
an adult in a culture in which skeptics had not produced the sorts of
alleged reasons for rejecting theistic belief that are at present
fashionable. The problem for intellectually sophisticated adults in our
culture, he says, is that many potential defeaters of theistic belief
are available; and we have substantial reason to think them true. One
kind of defeater for a belief (the kind Quinn is concerned with here) is
a proposition incompatible with the belief; Quinn cites

4. God does not exist

as a potential defeater of theism. And the problem for the
intellectually sophisticated adult theist in our culture, says Quinn, is
that many substantial reasons for believing (4) have been produced.

There are defeaters for theistic belief, then; and in the presence of
defeaters, an otherwise properly basic belief may no longer be properly
basic. More exactly, according to Quinn 

it seems plausible to suppose that conditions
are right for propositions like those expressed by (1) – (3)
to be . . . properly basic for me only if (i) either I have
no sufficiently substantial reason to think that any of
their potential defeaters is true, or I do have some such
reason, but for each such reason I have, I have an even
better reason for thinking the potential defeater in
question is false, and (ii) in either case my situation
involves no epistemic negligence on my part.[4]

Quinn goes on to say that he is not in this fortunate condition with
respect to theistic belief; he knows of substantial reasons, he says, to
think that (4) is true, and it is not the case that for each such reason
he has, he has an even better reason for thinking (4) false. So (by Q*)
belief in God is not properly basic for him; and he suspects the same
goes for most of the rest of us.

Now here I find myself in solid disagreement. We must first ask what
are these "very substantial reasons" for thinking that what (4)
expresses is true.[5] What would be some examples of such substantial
reasons for atheism? Quinn’s answer: "After all, nontrivial
atheological reasons, ranging from various problems of evil to
naturalist theories according to which theistic belief is illusory or
merely projective, are a pervasive, if not obtrusive, component of the
rational portion of our intellectual heritage."[6 ] So these substantial
reasons for thinking theism false would be the atheological argument
from evil together with theories according to which theistic belief is
illusory or merely projective; here perhaps Quinn has in mind Marxist
and Freudian theories of religious belief.

I should remark immediately that the Marxist and Freudian theories he
alludes to don’t seem to be even reasonably cogent if taken as reasons
for believing (4), or as evidence for the nonexistence of God, or as
reasons for rejecting belief in God. Freud’s jejune speculations as to
the psychological origin of religion and Marx’s careless claims about
its social role can’t sensibly be taken as providing argument or reason
for (4) , i.e., for the nonexistence of God; so taken they present
textbook cases (which in fact are pretty rare) of the genetic fallacy.
If such speculations and claims have a respectable role to play, it is
instead perhaps that of providing a naturalistic explanation for the
wide currency of religious belief, or perhaps that of attempting to
discredit religious belief by tracing it to a disreputable source. But
of course that doesn’t constitute anything like evidence for (4) or a
reason to think theism false. One might as well cite as evidence for
the existence of God St. Paul’s claim (Romans 1) that failure to believe
in God is a result of sin and rebellion against God. None of the
naturalistic theories according to which theism is illusory or merely
projective seem to me to have any strength at all as arguments or
evidence for the nonexistence of God-although they may be of interest in
other ways.

This leaves us with the atheological argument from evil as the sole
substantial reason for thinking (4) true. And initially this argument
seems much stronger as a reason for rejecting theistic belief. But is
it really? Until recently, most atheologians who urged an atheological
argument from evil held that

5. God exists and is omniscient, omnipotent, and wholly good

is logically incompatible with the proposition

6. there are 10[13] turps of evil

(where (6) is just a way of referring to all the evil our world
in fact displays). At present, I think atheologians have given up the
claim that (5) and (6) are incompatible, and quite properly so.[7 ] What
they now say is that (5) is unlikely or improbable
with respect to (6); and Quinn (himself, of course, no atheologian)
says, "What I know, partly from experience and partly from testimony,
about the amount and variety of non-moral evil in the universe confirms
highly for me the proposition expressed by (4)."[8 ] But is this really
true? Does what Quinn and the rest of us know about the amount and
variety of non-moral evil in the world confirm highly the nonexistence
of God? This is not the place to enter a discussion of that difficult
and knotty problem (difficult and knotty at least in part because of the
difficult and confusing character of the notion of confirmation); for
what it is worth, however, I can’t see that it does so at all. So far
as I can see, no atheologian has given a successful or cogent way of
working out or developing a probabilistic atheological argument from
evil; and I believe there are good reasons for thinking that it can’t be
done.[9] I am therefore very much inclined to doubt that (6) "highly
disconfirms" (5) for Quinn. At the least what we need here is some
explanation to show just how (or even approximately how) this
disconfirmation is supposed to go.

So first, these alleged substantial reasons for rejecting theism warrant
a good deal of skepticism. But secondly, even if we concede that there
are such reasons, Quinn’s conclusion won’t follow; this is because (Q*),
as it stands, is pretty clearly false. The suggestion is that if I have
a substantial reason for thinking some defeater of a proposition (for
example, it’s denial) is true, then I can’t properly take the
proposition as basic unless I have an even stronger reason for thinking
the defeater in question false. But surely this is to require too much.
Suppose an atheologian gives me an initially convincing argument for
thinking that (5) is in fact extremely unlikely or improbable on (6).
Upon grasping this argument, perhaps I have a substantial reason for
accepting a defeater of theistic belief, namely that (5) is improbable
on (6). But in order to defeat this potential defeater, I need not know
or have very good reason to think that it is false that (5) is
improbable on (6); it would suffice to show that the atheologian’s
argument (for the claim that (5) is improbable on (6)) is unsuccessful.
To defeat this potential defeater, all I need to do is refute this
argument; I am not obliged to go further and produce an argument for the
denial of its conclusion. Quinn takes

(4) God does not exist

to be a potential defeater for the propositions (1) – (3); but to defeat
the potential defeater offered by an argument for (4) I need not
necessarily have some argument for the existence of God. There
are undercutting defeaters as well as rebutting defeaters.[10]

There is another and more subtle point here. Quinn seems to be thinking
along the following lines: suppose I take some proposition as basic, but
have substantial evidence from other things I believe for some defeater
of this proposition-a proposition incompatible with it, let’s say. Then
(according to Q*) I am irrational if I continue to accept the
proposition in question, unless I also have good evidence for the
falsehood of the defeater. So if I accept a proposition p, but believe
or know other things that constitute strong evidence of some defeater q
of p, then, says Q*, if I am not to be irrational in continuing to
accept p as basic, I must have a reason for thinking q false, a reason
that is stronger than the reasons I have for thinking q true.

Now my question is this: could p itself be my reason for
thinking q false? Or must that reason be some proposition distinct from
p? Consider an example. I am applying to the National Endowment for
the Humanities for a fellowship; I write a letter to a colleague, trying
to bribe him to write the Endowment a glowing letter on my behalf; he
indignantly refuses and sends the letter to my chairman. The letter
disappears from the chairman’s office under mysterious circumstances. I
have a motive for stealing it; I have the opportunity to do so; and I
have been known to do such things in the past. Furthermore an extremely
reliable member of the department claims to have seen me furtively
entering the chairman’s office at about the time when the letter must
have been stolen. The evidence against me is very strong; my colleagues
reproach me for such underhanded behavior and treat me with evident
distaste. The facts of the matter, however, are that I didn’t steal the
letter and in fact spent the entire afternoon in question on a solitary
walk in the woods; furthermore I clearly remember spending that
afternoon walking in the woods. Hence I believe in the basic way

7. I was alone in the woods all that afternoon, and I did not
steal the letter.

But I do have strong evidence for the denial of (7). For I
have the same evidence as everyone else that I was in the chairman’s
office and took the letter; and this evidence is sufficient to convince
my colleagues (who are eminently fairminded and initially well disposed
towards me) of my guilt. They are convinced on the basis of what they
know that I took the letter; and I know everything they know.

So I take (7) as basic; but I have a substantial reason to believe a
defeater of (7). According to Q*, if I am to be rational in this
situation, I must have even better reason to believe that this
potential defeater is false. Do I? Well, the only reason I have for
thinking this potential defeater false is just (7) itself; I don’t have
any independent reason to think the defeater false. (The
warrant I have for (7) is nonpropositional warrant; it is not
conferred upon (7) by virtue of my believing that proposition on the
basis of some other proposition, for I don’t believe (7) on the basis of
any other proposition.)

In this situation it is obvious, I take it, that I am perfectly rational
in continuing to believe (7) in this basic way. The reason is that in
this situation the positive epistemic status or warrant that (7) has for
me (by virtue of memory) is greater than that conferred upon its
potential defeater by the evidence I share with my colleagues. We might
say that (7) itself defeats the potential defeater; no further
reason for the denial of this defeater is needed for me to be rational.
Suppose we say that in this sort of situation a proposition like (7) is
an intrinsic defeater of its potential defeater. When a basic
belief p has more by way of warrant than a potential defeater q of p,
then p is an intrinsic defeater of q-an intrinsic defeater-defeater, we
might say. (A belief r is an extrinsic defeater-defeater if it defeats
a defeater q of a belief p distinct from r.)

So my question here is this: how is Quinn thinking of these reasons for
thinking the defeating proposition false? I am inclined to
believe that he intends Q* to be read in such a way that these reasons
have to be extrinsic defeater-defeaters; but if so, then his
principle, I think, is clearly false. On the other hand, perhaps it is
to be understood as saying something like

Q** If you believe p in the basic way and you have reason to believe a
defeater q of p, then if you are to be rational in continuing to believe
p in this way, p must have more warrant for you then q does.

I am not certain this principle is correct, but I am also not inclined
to dispute it. The central point to see, however, is that if a belief p
is properly basic in certain circumstances, then it has warrant or
positive epistemic status in those circumstances in which it is properly
basic-warrant it does not get by virtue of being believed on the
evidential basis of other propositions. (By hypothesis it is not
believed on the evidential basis of other propositions.) To be
successful, a potential defeater for p must have as much or more warrant
as p does. And p can withstand the challenge offered by a given
defeater even if there is not independent evidence that serves either to
rebut or undercut the defeater in question; perhaps the nonpropositional
warrant that p enjoys is itself sufficient (as in the above case of the
missing letter) to withstand the challenge.

But how does all this apply in the case in question, the case of belief
in God and the alleged defeaters Quinn mentions? As follows. If there
are circumstances in which belief in God is properly basic, then in
those circumstances such belief has a certain degree of warrant or
positive epistemic status. Now suppose a potential defeater arises:
someone claims that the existence of 10[13] turps of evil makes theism
improbable, or he claims that theistic belief arises out of nothing more
reputable than a kind of widespread human neurosis. Two questions then
arise. First, how does the degree of nonpropositional warrant enjoyed
by your belief in God compare with the warrant possessed by the alleged
potential defeater? It could be that your belief, even though accepted
as basic, has more warrant than the proposed defeater and thus
constitutes an intrinsic defeater-defeater. When God spoke to Moses out
of the burning bush, the belief that God was speaking to him, I daresay,
had more by way of warrant for him than would have been provided for its
denial by an early Freudian who strolled by and proposed the thesis that
belief in God is merely a matter of neurotic wish-fulfillment. And
secondly, are there any extrinsic defeaters for these defeaters?
Someone argues that the existence to 10[13] turps of evil is
inconsistent with the existence of God; I may then have an extrinsic
defeater for this potential defeater. This defeater-defeater need not
take the form of a proof that these propositions are indeed consistent;
if I see that the argument is unsound, then I also have a defeater for
it. But I needn’t do even that much to have a defeater. Perhaps I am
no expert in these matters but learn from reliable sources that someone
else has shown the argument unsound; or perhaps I learn that the experts
think it is unsound, or that the experts are evenly divided as to its
soundness. Then too I have or may have a defeater for the potential
defeater in question, and can continue to accept theistic belief in the
basic way without irrationality.

By way of conclusion then: Quinn claims that intellectually
sophisticated adult theists in our culture are seldom in epistemic
circumstances in which belief in God is properly basic; for they have
substantial reason to think that some potential defeater of theism is
true, and do not have, for each such defeater, even stronger reason to
think it is false. But first, it isn’t necessary that they have reasons
independent of their belief in God for the falsehood of the
alleged defeaters. Perhaps the nonpropositional warrant enjoyed by your
belief in God is itself sufficient to turn back the challenge offered by
the alleged defeaters, so that your theistic belief is an intrinsic
defeater-defeater. And second, extrinsic defeaters of the alleged
defeaters need not be evidence for the falsehood of those defeaters;
they may instead undercut the alleged defeaters; they may be, for
example, refutations of atheological arguments. (And here Christian
philosophers can clearly be of service to the rest of the Christian
community.) My opinion (for what it is worth) is that for many theists,
the nonpropositional warrant belief in God has for them is indeed
greater than that of the alleged potential defeaters of theistic belief-
for example, Freudian or Marxist theories of religion. Furthermore,
there are powerful extrinsic defeaters for the sort of potential
defeaters of theism Quinn suggests. The atheological argument from
evil, for example, is formidable; but there are equally formidable
defeaters for this potential defeater. I am therefore inclined to
believe that belief in God is properly basic for most theists-even
intellectually sophisticated adult theists.

Notes

1. Philip Quinn, "In Search of the Foundations of Theism," Faith and
   Philosophy 2 (October 1985): 20-1.
2. Faith and Rationality, ed. A. Plantinga and N.
   Wolterstorff (South Bend: The University of Notre Dame Press, 1983), pp.
   84-5.
3. Quinn, "Search," p. 481.
4. Ibid., p. 483.
5. Ibid., p. 481.
6. Ibid.
7. See, for example, Chapter IX of my The Nature of the
   Necessity (Oxford: The Clarendon Press, 1974).
8. Quinn, "Search," p. 481.
9. See my paper "The Probabilistic Argument from Evil,"
   Philosophical Studies (1980): 1-53.
10. I owe these terms to John Pollock. The distinction between
    undercutting and rebutting defeaters is of central importance to
    apologetics. If the propriety of basic belief in God is threatened by
    defeaters, there are two ways to respond. First, there is negative
    apologetics: the attempt to refute the arguments brought against theism
    (the atheological argument from evil, the claim that the conception of
    God is incoherent, and so on). Second, there is positive apologetics:
    the attempt to develop arguments for the existence of God. These are
    both important disciplines; but it is only the first, clearly enough,
    that is required to defeat those defeaters.