I. Warrant and the Argument from Religious Experience

The first objection is really less an objection, so it seems to me, than a confusion, a failure to make an important distinction. Anthony O’Hear considers the idea that theistic belief might be justified or receive warrant (it is hard to tell which he is thinking of) in a direct way, not by way of argument or inference. (In my terms, the question is whether theistic belief might be properly basic, either with respect to warrant or with respect to justification.) Referring to William James, John Baillie, and others, he notes that one suggestion as to how this might go would be by way of religious experience, broadly conceived. He then goes on to say:

It is the idea of direct personal contact with a non-sensory reality that non-believers will find hard to grasp. In order to bring out the nature of the difficulty, I will consider the extent to which religious experience can provide evidence for the existence of a reality beyond the experience itself. Presumably people who are convinced that they are in personal contact with some super-reality will not often attempt to argue or prove their conviction at all, nor will their conviction be arrived at inferentially, any more than we naturally infer from statements about our sensations to statements about physical objects. Nevertheless the question of the extent to which the conviction can be justified by the experience naturally arises.⁴²²⁴²²   Experience, Explanation and Faith (London: Routledge and Kegan Paul, 1984), p. 27. Page references to O’Hear are to this work.

Here there are several questions. First, note that this quotation illustrates that initial difficulty I mentioned above: is O’Hear talking about justification, or rationality, or warrant, or what? This isn’t clear 327from what he says here or elsewhere. Despite the occurrence of ‘justified’ in the last line of the quotation, I don’t think he’s really speaking of justification—and in any event, as we have already seen (above, pp. 99ff.), the question of justification is too easily answered to be interesting. Although O’Hear speaks of “personal contact,” perhaps his question is best construed in the present context as the question whether religious experience could put us in epistemic contact with a nonsensory reality (i.e., one that can’t, ordinarily, be seen, heard, touched, etc.) such as God; and that question, I take it, is the question whether beliefs about such a nonsensory reality could acquire warrant by way of religious experience.

Now his initial suggestion is that there is something problematic about the very idea of a human person’s being in cognitive contact (the kind required by warrant) with a nonsensory reality such as God. Why is this problematic? O’Hear doesn’t directly answer that question, but proposes to “bring out the difficulty” by turning to the question whether “religious experience can provide evidence for the existence of a reality beyond the experience itself.” This sounds like he thinks the way to answer the question

Does religious experience make it possible for us to have the right sort of cognitive connection with God?

is to ask

Is there a good argument from the existence of the experience in question to the existence of God: an argument whose premises report the experience in question and whose conclusion is that there is such a person as God?

That this is what he has in mind is confirmed by what he says a bit later:

Christians, for example, tend to explain this unpredictability [of religious experience of the relevant sort] by saying that these experiences are a gift of God. This may be so, but saying it certainly weakens attempts to argue from the experience to the reality. (p. 44)

This clearly suggests that what is at issue, here, with respect to the question whether theistic belief can have warrant by virtue of religious experience, is whether there is a good argument from premises reporting that experience to the existence of God. O’Hear goes on to say that what we are really asking for is

grounds on which the religious experiences we have . . . could be regarded as experiences of an objective sort. (p. 45)

the answer, he says,

will have to be in terms of the explanatory power of the hypothesis that religious experiences are due at least in part to the existence 328and operation of an objective religious reality, rather than due to merely worldly factors, such as features of a person’s psychology, chemistry or upbringing. (p. 45)

His thought, then, as far as I can make it out, is twofold:

(a) theistic belief can have warrant by virtue of religious experience only if there is a good (noncircular) argument from premises reporting the occurrence of such experiences to the existence of God

and

(b) such an argument will have to involve as a premise the proposition that the existence of God is the best explanation of religious experiences.

(Of course such an argument would also have to provide reasons for thinking that premise true.)

I say (a) is part of O’Hear’s thought; perhaps ‘assumption’ would be a better term because he doesn’t explicitly make this claim but rather just takes it for granted. Another way to put the assumption: theistic belief can have warrant by way of religious experience only if some theistic argument from religious experience is successful. This assumption is widely shared and seldom argued; as I shall maintain, however, it has the substantial disadvantage of being false. In fact, one of the main points to see here is that the question whether theistic belief can receive warrant by way of religious experience (and thus in the basic way) is a wholly different question from the question whether there is a good argument from the existence of religious experience to the existence of God. (Not only are these different questions: an affirmative answer to the first does not require an affirmative answer to the second.) I shall argue that (a) is false. (a)’s being false doesn’t distinguish religious experience and theistic or Christian belief from other kinds of experience and belief: perceptual experience and belief, memorial experience and belief, a priori experience and belief, and the like, all resemble Christian belief in this respect. In each of these cases, it is entirely possible that the beliefs in question have warrant even if there is no good argument from the existence of the experience in question to the truth of those beliefs.⁴²³⁴²³   See Warrant and Proper Function (hereafter WPF), pp. 61ff. and 93ff.

This is one of the most important things to see here; before arguing this claim, however, I want to note another writer who also simply assumes that (a) is true, without so much as raising the question whether it is. According to the late J. L. Mackie,

an experience may have a real object: we ordinarily suppose our normal perceptual experience to be or to include awareness of independently existing material spatio-temporal things. The question then is whether specifically religious experiences should be taken to have real objects, to give us genuine information about independently existing supernatural entities or spiritual beings.⁴²⁴⁴²⁴   The Miracle of Theism (Oxford: Clarendon Press, 1982), p. 178. Page references to Mackie are to this work.

So far so good: this is the question whether religious experience can or does provide warrant for belief in “independently existing supernatural entities or spiritual beings” such as God. But Mackie goes on:

Whether their content [i.e., the content of religious experiences] has any objective truth is the crucial further question. . . . The issue is whether the hypothesis that there objectively is a something more gives a better explanation of the whole range of phenomena than can be given without it. (p. 183)

Mackie concludes his examination of the possible warrant conferred by religious experience with these words:

if the religious experiences do not yield any argument for a further supernatural reality, and if, as we have seen in previous chapters, there is no other good argument for such a conclusion, then these experiences include in their content beliefs that are probably false and in any case unjustified. [I take it ‘unjustified’, here, means ‘without warrant’.] (p. 186)

Here we see the very same assumption at work as in O’Hear. Like the latter, Mackie assumes that theistic (or other religious) belief could get warrant by way of religious experience only if there is a good argument from the existence and character of that experience to the existence of God (or “something more”). (And like O’Hear, he also seems to endorse (b).) Neither Mackie nor O’Hear argues for this claim, simply taking it utterly for granted that the only way a belief (or at any rate a religious or theistic belief) could possibly receive warrant from experience would be by way of an implicit argument from the existence and properties of that experience to the truth of the belief in question. But why think a thing like that? It certainly isn’t self-evident. In fact, once we explicitly raise the question whether it is true, (a) looks extremely problematic. Presumably one wouldn’t want to say that perceptual beliefs get warrant from experience only if there is a good (noncircular) argument from the existence of perceptual experience to the truth of perceptual beliefs; if not, however, what is the reason for saying it in the case of theistic or Christian belief?

Mackie makes this assumption, I believe, because he makes another: that theistic and Christian belief is or is relevantly like a scientific 330hypothesis—something like special relativity, for example, or quantum mechanics, or the theory of evolution. Still speaking of whether theistic belief can receive warrant by way of religious experience, he (characteristically) remarks: “Here, as elsewhere, the supernaturalist hypothesis fails because there is an adequate and much more economical naturalistic alternative” (p. 198). This remark is relevant only if we think of belief in God as or as like a sort of scientific hypothesis, a theory designed to explain some body of evidence, and acceptable or warranted to the degree that it explains that evidence. On this way of looking at the matter, there is a relevant body of evidence shared by believer and unbeliever alike; theism is one hypothesis designed to explain that body of evidence, and naturalism is another; and theism has warrant only to the extent that it is a good explanation thereof, or at any rate a better explanation than naturalism.

But why should we think of theism like this? Why should we think of it as a kind of hypothesis, a sort of incipient science? Consider the extended A/C model of chapters 8 and 9. On that model, it is not that one notes the experiences, whatever exactly they are, connected with the operation of the sensus divinitatis, and then makes a quick inference to the existence of God. One doesn’t argue thus: I am aware of the beauty and majesty of the heavens (or of my own guilt, or that I am in danger, or of the glorious beauty of the morning, or of my good circumstances): therefore there is such a person as God. The Christian doesn’t argue: “I find myself loving and delighting in the great things of the gospel and inclined to believe them; therefore they are true.” Those would be silly arguments; fortunately they are neither invoked nor needed. The experiences and beliefs involved in the operation of the sensus divinitatis and IIHS serve as occasions for theistic belief, not premises for an argument to it.

The same holds for, say, memory beliefs. Obviously one could take a Mackie-like view here as well. One could hold that our beliefs about the past are really like scientific hypotheses, designed to explain such present phenomena as (among other things) apparent memories, and if there were a more “economical” explanation of these phenomena that did not postulate past facts, then our usual beliefs in the past would have no warrant. But of course this is merely fantastic; we don’t in fact accept memory beliefs as hypotheses to explain present experience at all. Everyone, even small children and others with no interest in explaining anything, accepts memory beliefs. We all remember such things as what we had for breakfast, and we never or almost never propose such beliefs as good explanations of present experience and phenomena. And the same holds for theism and Christian belief in the suggested model.

So Mackie apparently believes that

(c) theistic belief is or is relevantly like a quasi-scientific hypothesis, designed to explain religious experience (perhaps among other things).

This explains why he believes (a), that is, that theistic belief can get no warrant from religious experience unless there is a good argument from premises reporting the experiences to the existence of God. As we have seen, however, (c) is false.

Well, perhaps Mackie would insist on (a) even if it is clear that Christians do not take belief in God or Christian belief generally as hypotheses; perhaps he would nonetheless insist that the only way in which such belief could possibly get warrant would be by being successful quasi-scientific hypotheses. But precisely this is what is refuted by the A/C model of chapter 6 and the extended A/C model of chapters 8 and 9. These models show that it is clearly possible that theistic and Christian belief have warrant, but not by way of being hypotheses that nicely explain a certain range of data. For if Christian belief is, in fact, true, then obviously there could be such cognitive processes as the sensus divinitatis and IIHS or faith. As we saw, beliefs produced by these processes would meet the conditions necessary and sufficient for having warrant: they would be the result of cognitive faculties functioning properly in a congenial epistemic environment according to a design plan successfully aimed at truth. Hence (a) is plainly false. It is plainly false that Christian belief has warrant (and could constitute knowledge) only if there is also a good argument from the existence of the experiences involved in the operation of IIHS to the truth of Christian belief; and the same point holds for theistic belief and the sensus divinitatis. Why suppose that if God proposes to enable us to have knowledge of a certain sort, he must arrange things in such a way that we can see an argumentative connection between the experiences involved in the cognitive processes he selects and the truth of the beliefs these processes produce? That requirement is both entirely gratuitous and also false, since it doesn’t hold for such splendid examples of sources of knowledge as perception, memory, and a priori intuition.