Religious experience

It is familiar by now that one might hope to explain the rationality of belief in the existence of God by appealing to the basic epistemic warrant conferred on that belief by experiences of God, or experiences which indicate or suggest the presence of God. Thus, Alston has appealed to the warrant bestowed by perceptual contact with the world to indicate the sort of epistemic upshot which religious experiences might have, and aside from the perceptual model, philosophers have explored analogies with experiences of personal relationships.Footnote ³

It is not my aim in this article to discuss these suggestions. It is a premise of the argument I am presenting that religious experience can indeed confer warrant on a person's belief in the existence of God (if the experience is their own, at least), but I shall not defend that premise. One version of the argument I am presenting – the version which I will in fact be setting out – requires that religious experience confers immediate or basic warrant. But another version would not require this. The crucial premise for arguments of the sort I am interested in here is that religious experience is warrant-conferring in a way that testimony is not (perhaps, but not necessarily, because testimonial warrant is subject to defeaters which experiential warrant is not subject to). In this article, that premise is defended by appealing to the immediate or unconditional nature of experiential warrant, contrasted with the mediated or conditional nature of testimonial warrant. But the crucial premise might be defended otherwise: all that is required is to show that experiential warrant is, in fact, more secure than testimonial warrant, and this might be achieved by showing that whilst experiential warrant can defeat its defeaters, testimonial warrant cannot.Footnote ⁴

Some philosophers who emphasize the warrant-conferring power of religious experience might well think that the testimony of those who have enjoyed a warrant-conferring religious experience can be warrant-conferring for an audience, at least under some conditions. You might already think, for your own reasons, that this is not correct, and that religious testimony fails to transmit warrant. If so, all well and good for the argument. In the next section, though, I will say something about why we might be entitled – perhaps required, even – to accept that the warrant-conferring powers of experience and of testimony differ.

Scepticism about testimonial warrant

The argument I am setting out relies on an epistemic asymmetry between a person who enjoys an experience which rationally grounds their belief in the existence of God and those who enjoy no such experience and must rely on testimony. Since Hume famously expressed scepticism about the capacity of testimony to transmit warrant in the religious domain, I have taken the liberty of calling the argument ‘Humean’.Footnote ⁵ As will emerge as this section unfolds, though, the scepticism upon which the argument relies does not depend upon Hume's own reasons for such scepticism, for there are reasons for taking it seriously which were not noted by Hume.

In section 10 of the Enquiry Concerning Human Understanding, Hume presents two quite different sorts of reasons for doubting that either scriptural or more recent reports of miraculous events could be sufficient to epistemically ground our belief that they happened. The first I will call methodological and the second practical.

The methodological worry Hume presses against relying on testimony concerning the miraculous is that testimony, qua evidence, must be weighed against any and all countervailing evidence (here the ‘must’ expresses what is necessary in order for someone to be a responsible or reasonable thinker). But it is in the very nature of the miraculous that it runs counter to the law-governed order of nature: that is what it is for something to be truly (and not merely figuratively) miraculous. So a report of some miraculous event (or, more accurately, a report of some event as miraculous) is immediately in tension with everything we take ourselves to know about the workings of nature, in so far as we take ourselves to know that nature is law-governed. Hume's point is that what we take ourselves to know about the law-governed order of nature is so much better confirmed and more familiar than the idea of a non-rule-governed natural event that it would be rationally indefensible to give any serious epistemic weight to anyone's testimony that a miracle has occurred – it would amount to ignoring the fact that the countervailing evidence overwhelmingly outweighs the testimonial evidence.

The practical concerns Hume raises are of a familiar sort: particularly in matters of religion, people are eager to convince others to share their views (perhaps from the best of motives, such as hoping to save their souls), and they are naturally enthusiastic (in the eighteenth-century sense of that term in which it pertains, according to the OED, to ‘ill-regulated or misdirected religious emotion’ and ‘extravagance of religious speculation’), so the confidence with which they present their own views often exceeds the quality and/or quantity of their evidence.Footnote ⁶ All in all, there is more reason to doubt the word of those who are apparently convinced of their miraculous experiences, or at least to withhold judgement, than to trust them, and the problem only deepens when it is not the direct testimony of those who purport to have had the experience which one is presented with, but rather second-hand or even more mediated reports of such experiences.Footnote ⁷

As I said, I do not intend to motivate testimony scepticism by simply endorsing Hume's claims in the Enquiry. But there are, I think, plenty of philosophers who would (and indeed do) say that Hume's methodological point is well taken because it reflects an independently plausible holism about justification. Similarly, you do not have to be Hume to think that the importance of someone's religious commitments to them and the way in which some people – perhaps the people who are most likely to proselytize their faith and testify as to its grounds – want to believe give grounds for unusual caution with respect to the epistemic status of their testimony about religious matters.

Perhaps most philosophers who would be willing to take Hume's methodological and practical concerns about religious testimony seriously are inclined to believe that miracle reports and other theological claims are false. But, crucially, they needn't be. So what I want to do, now, to motivate testimony scepticism beyond simply appealing to Hume's authority is to explore some of the reasons for being sceptical about religious testimony even if you are a religious realist, and to do that by reflecting on a debate between Alvin Plantinga and Philip Quinn.

Plantinga's ‘reformed epistemology’ is ‘reformed’ in the same sense that Protestant theology amounted to a reformation: in each case, immediacy enjoys a prominent role. Theologically, the Protestant Reformation involved (amongst other things) the insistence that contact with God is available via direct engagement with the word of the Bible, and that one might enjoy a personal relationship with God without that being mediated by the Church with its rituals and works. The reformed epistemology argued that belief in God might be epistemologically basic, requiring none of the intellectual sophistication of rigorous theology to confer warrant on it, meaning that some of our religious beliefs might be immediately warranted (warranted not by inferences, but, perhaps, by experiences directly).

Quinn objected to the reformed epistemology movement on the grounds that in order to be warranted, the religious beliefs not grounded in inference to which Plantinga referred would have to be undefeated by other beliefs which a reflective, well-informed person reasonably held, and that in fact reflective, well-informed adults typically do have the sorts of defeating beliefs which would suffice to undermine warrant. Such defeating beliefs include knowledge of the theories of Freud and Marx, for example, which make salient the possibility that our religious beliefs are not, as they might appear to be, immediate responses to the divine but rather attitudes which are a product of psychological or social forces which are nothing to do with the divine at all. Even if such theories do not succeed in establishing that our religious beliefs actually are the products of non-truth-tracking mechanisms, the salient possibility that they are seems to undercut the basicness of their epistemic status: if those religious beliefs are warranted after all, it is surely because of some further reasons we have to be confident in them (reasons which Quinn suggests are the product of natural theology, thus restoring the arguments of natural theology to the place from which reformed epistemology sought to banished them). Other defeaters for our religious beliefs include our knowledge of and reflection upon the existence of evil, which (Quinn thinks) is sufficiently problematic to require a theodicy (i.e. a theoretical account of why a benevolent God allows evil to happen) if we are to maintain warranted religious beliefs. Quinn is confident that the demand that our religious beliefs be supported by a good dose of theory and argument in order for them to be warranted can be met, but he thinks it is a genuine demand and that appealing to brute epistemic basicness fails to take the significance of rational defeaters seriously.Footnote ⁸

Plantinga, of course, disagrees about the significance of Quinn's defeaters. It is not the purpose of this article to adjudicate the debate between Plantinga (or any reformed epistemologist) and Quinn (or any critic of reformed epistemology), though. The point I want to make is that it is at least prima facie plausible that doubts about the epistemic status of religious testimony might be raised on grounds similar to those upon which doubts about the epistemic status of supposedly immediately warranted experiential beliefs are raised, and that they are more powerful with respect to testimony than with respect to experientially warranted beliefs. Even if you are a reformed epistemologist who, perhaps by appealing to the perceptual model, is satisfied that religious experience can confer basic warrant on religious beliefs, you might still think that the warrant conferred by testimony is subject to the sorts of defeaters to which Quinn (or indeed Hume) adverts. It is one thing to be warranted in one's own belief on the basis of perceptual evidence; but it is quite another to be warranted in one's belief on the basis of another person's testimony of their purported perceptual evidence.

This might strike some as counterintuitive. Surely, one might say, if I am in fact warranted in believing that p on the basis of my perceptual evidence, you are in fact warranted in believing that p on the basis of my truthful testimony of that evidence, though you might not know that you are thus warranted, because you might not know that my testimony is truthful. The point is that you can be warranted in believing that p without knowing that you are warranted in believing that p.

Let us grant that you can indeed be warranted in believing that p without knowing that you are warranted in believing that p. What follows from this? All that immediately follows is that it might be the case that you are warranted in believing that p even though as it happens you do not know that that is your epistemic situation. As far as whether you are warranted in believing that p despite not knowing that you are is concerned, though, everything hangs on why you do not know that you are warranted in believing that p. If you have simply never got around to forming a belief about the epistemic status of your belief that p, then not knowing that you are warranted in believing that p does not undermine your being, in fact, warranted in believing that p. On the other hand, if you fail to believe (hence know) that you are warranted in believing that p because you take seriously some considerations which would seem to undermine the epistemic status of your belief that p, then the epistemic status of your belief that p might be defeated and not merely unknown.

It ought to go without saying – but is worth saying anyway, to be absolutely clear – that nothing I have said relies on the obviously false thought that defeater beliefs which you might happen to have suffice to make the belief that p untrue. It is not the truth of your belief which is in question here, but rather your being rationally entitled to believe it, given your psychological and epistemic situation. That, indeed, is why the argument I am exploring is open to the religious realist, who thinks that the beliefs in question are (at least in some cases) true. The point is that what is rationally acceptable – and hence epistemically respectable – as a belief for one person is not guaranteed to be rationally acceptable for another. The epistemic status of a belief is a function of things other than merely the truth value of the proposition believed.

Suppose, then, that you are the sort of sophisticated modern person whom Quinn has in mind, with a stock of knowledge – or at least reasonably well informed opinion – about the world and about various ideas about the world which amount to rational defeaters for religious beliefs. Now suppose that the grounds for your religious beliefs are (or would be – you might be considering whether to adopt some religious belief rather than reflecting upon a belief you already have) testimonial. Testimony, as Hume observed, does not confer basic epistemic warrant on a belief. So even if you are willing to say, with the reformed epistemologists, that the beliefs Quinn adverts to are insufficient to defeat basic epistemic warrant for religious beliefs, you might still think that those beliefs defeat the testimonial warrant a person would otherwise have for their religious beliefs, since testimonial warrant (unlike immediate perceptual warrant, for example) depends upon the balance of evidence, including the evidence of one's beliefs about the origins of the experience on which a person's testimony (even if truthful) is based.

Let us take a straightforward perceptual case as a model. Suppose I see a tomato in front of me. This fact – the fact that I see the tomato – warrants my belief that there is a tomato in front of me, and warrants it immediately. I do not need to have a high degree of confidence in an anti-sceptical argument in order to enjoy the warrant which my perception affords me. Now suppose that I come across a clever sceptical argument to the effect that perception is systemically unreliable and can confer no warrant on my beliefs about the external world, and suppose that I become convinced by this argument. My predicament will be that of a person who has warrant for their belief that there is a tomato in front of them (supposing that I cannot shake the belief, despite my scepticism), but refuses to acknowledge that they have such warrant.

On the other hand, the epistemic predicament of another person who has not seen the tomato might be rather different. Suppose your only reason for believing that there is a tomato in front of me is that I tell you that there is (I am on the phone to you, perhaps). Plausibly, the fact that I embrace scepticism about my perceptual access to the external world is not a defeater for your belief (if you form such a belief) that there is a tomato in front of me, even if your only reason for believing that there is a tomato in front of me is that I tell you that there is, or say things which, stripped of their sceptic's qualifications (‘if I didn't know better’), strongly suggest that I see it. But if you are the sceptic, the epistemic situation is plausibly rather different when you receive my testimony. It now looks like your warrant for your belief that there is a tomato in front of me is defeated, though I retain the (immediate, perceptual) warrant I have and testify truthfully. That is because your warrant is not immediate or basic, but merely testimonial, and testimonial warrant, unlike perceptual warrant, is sensitive to an audience's other beliefs.

There is a reason why testimonial, but not perceptual, warrant is sensitive to one's other beliefs. It is that testimonial warrant depends upon the degree to which it is rational to trust the source of testimony. Were testimony a way of coming to know things in the way perception or pure reason are, then trust would not come into it. But testimony, as Hume pointed out, relies upon people, not quasi-mechanical sub-personal perceptual systems or abstract relations of ideas, and people are complex in respect of their motives for telling us things and in respect of their belief-forming mechanisms. Perception and reasoning either work or they go wrong. People also either succeed or fail in their aims, but what calls for trust as a mediating epistemic condition in the case of testimony is that we can never quite take it for granted what those aims are. Perception is never setting out to convince us of what it would be good for us to believe, even though it is not true, for example. Perception is necessarily prima facie trustworthy in the sense that genuine perception is essentially a matter of presenting or representing how the world actually is (or a matter of producing experiences by means of a generally reliable process, perhaps), so all we have to decide is whether a putatively perceptual experience really is perceptual.Footnote ⁹ That is why stipulating that an experience is genuinely perceptual suffices, in the example above, to establish that the person whose experience it is has prima facie warrant, at least, for the belief they form on the basis of it, regardless of their attitude towards their own epistemic situation. Stipulating that a putative piece of testimony really is testimony in such an example, however, settles no questions about the epistemic status of one's beliefs formed on the basis of that testimony, for it remains to be established whether that testimony is plausibly truth-revealing: testimony can be genuine but non-truth-revealing, and not just because of untruthfulness – it might be that the testifier simply has the wrong end of the stick which they are then, truthfully (in the sense of trying to tell the truth), passing on.Footnote ¹⁰

The fictionalist argument

If the upshot of all this is right – if, that is, testimonial warrant is subject to defeaters which immediate experiential warrant is not – then there is room for the thought that whilst A's warrant for their religious beliefs may be excellent, because A has enjoyed the sort of religious experience which confers immediate warrant, B may be doomed to never enjoy any meaningful warrant for their religious beliefs if they, B, are reliant on A's testimony. B may simply have too many of the (perfectly reasonable) beliefs which defeat testimonial warrant for those religious beliefs to be warranted to any significant degree.

If you agree with Quinn about the role of natural theology and theodicy in undermining the defeating effect of our beliefs about the origins of religious ideas and the problem of evil, then you are liable to say that in fact the possibility described in the previous paragraph is not the way things are: true, B's religious beliefs will not be immediately warranted, and they would lack warrant entirely if there were to be no rational considerations which could make it reasonable to believe in the face of their putatively warrant-defeating beliefs, but in fact B has only to avail themselves of some natural theology and theodicy in order for them to overcome the warrant-defeating effects of those putatively warrant-defeating beliefs.

But what I want to point out is that one might agree that the possibility described in the first paragraph of this section is a real possibility, but also doubt the power of natural theology, theodicy, and indeed anything other than immediate experience to overcome the warrant-defeating powers of some of our reasonably held beliefs. One might, for example, think that natural theology and theodicy are hopeless attempts to know or speculate about what cannot be known or even intelligibly speculated about. One might be a ‘negative theologian’ of the sort Maimonides is said to have been. And one might think the view that God is only to be conceived of by means of saying what he is not to entail that one cannot begin to say what could constitute evidence for his deeds, or what considerations he would be moved by in allowing evil. Nothing about accepting the view that the warrant for non-basic beliefs is subject to being defeated by our other beliefs entails or rationally commits one to thinking that that defeat is bound to be itself overcome by other considerations.

Now one might run the following argument. It is important – morally, spiritually, or otherwise – for a person to be committed to the existence of God, and to have some commitments about what God requires of us, or what we might do to honour him. In short, it is important that the thought that there is a God plays some positive role in a person's life, and a meaningful role at that.Footnote ¹¹ Remember that for all that's been said in the forgoing discussion, one might be a religious believer, who might have had the sort of experience which, according to their epistemology at least, confers immediate warrant on their religious beliefs. So, from their point of view at least, they have no reason to be embarrassed about making these claims about the importance of the idea of God in people's lives, nor about interpreting them in a full-bloodedly realist way: there is a God, and it is morally and spiritually important (and not just pragmatically or psychologically important) to let the existence of God be a force in one's life.

But it is one thing for a person to have a reason (whether they are aware of it or not) to live a life in which religious commitments play an important part, and quite another for them to be in a good epistemic position with respect to religious beliefs. If it is true that we all have a reason (the best sort of reason, a reason of a moral and/or spiritual sort) to have some sort of faith which is an active force in guiding our deliberations about what to do and framing the way we think about the most important things in life, it does not follow that we are warranted in holding religious beliefs. Equivalently, it does not follow from our lacking the sort of epistemic warrant which would put our religious beliefs in good standing that we have no reason to have some sort of faith.

Now it might be that what we have so far amounts to a reason for thinking that it can be morally or spiritually good for us to adopt certain beliefs regardless of the fact that we lack epistemic warrant for them. But that is not the only response one might have if one accepts the views expressed in the previous two paragraphs. Another option is to think that the faith which is morally and/or spiritually required of us need not amount to doxastic faith, i.e. belief, at all. One might then think that we can satisfy the relevant moral and/or spiritual demands without sacrificing rationality or epistemic virtue. (This might be appealing even if rationality and epistemic virtue are less worth having than the relevant moral and/or spiritual goods.) It might be good enough, morally and/or spiritually, if our faith is non-doxastic.

Various proposals about the nature of non-doxastic faith have been made in the literature, sometimes suggesting that the essence of faith is hope rather than belief, sometimes that it is trust, and sometimes that it is some more general sort of ‘acceptance’.Footnote ¹² This is not the place to discuss which of these proposals, if any, are promising as charitable models of what faith is typically like. The argument I am presenting assumes that there is some form of non-doxastic faith which it is possible to have, and argues that in at least one respect it would be good to have that sort of faith instead of doxastic faith. In particular, it would be better to have non-doxastic faith than either (i) no faith at all or (ii) doxastic faith if you are one of those for whom testimony is the only source of evidence for your religious beliefs. The idea that it might be good to have some kind of non-doxastic faith is one which many religious realists would be willing to assent to anyway: non-doxastic faith is often recognized to be important even if warranted belief is also available.

I take it that it is not implausible that, if there is a God of something like the Judaeo-Christian sort, at least, then a life in which the idea that God exists (and has some particular traits) plays some significant role is preferable to one in which the idea of such a God plays no significant role. This might be so for a number of different reasons. Perhaps a life of faith makes available the sort of personal relationship with God which is unsurpassably good. It might also be that, whether or not a personal relationship with God is the greatest of goods, there are other goods which faith makes available or involves. These might be intrinsic or instrumental goods.Footnote ¹³

One might doubt whether a merely non-doxastic faith could sustain a personal relationship, at least, in the right sort of way, for one might think that it is crucial for such a relationship that one believes that the object of one's attention exists. I will not settle that issue here, but suffice it to say that things might not be so straightforward. Think, for example, about a person who receives a letter one day purporting to be from their long-lost (and hitherto believed dead) mother. There is enough in the letter to make it reasonable for them to hope that it is what it purports to be, so they reply and subsequently receive another letter. This exchange of letters goes on for some time, and all the time this person hopes that they are writing to their mother, and receiving news, reminiscences, and words of kindness from her, and what they write back is just what one would write to one's mother (in respect of, for example, affection, concern for the other's well-being, intimacy, etc.); but they are never confident enough that the letters are genuine to bring themselves to believe that their correspondent is who she purports to be. Indeed, they are not entirely confident that they are corresponding with a particular person at all, and not a criminal organization, various members of which at different times write a letter as part of an elaborate long-term con trick. Their hope is not unreasonable, however, and it would seem off key to describe them as open minded as to the true nature of the correspondence – they invest a great deal of not only hope but also effort and trust in their relationship with this correspondent. (Notice that the situation is not one in which they go as far as to believe that the letters are not from their mother.) Now suppose that the person writing the letters is, in fact, their mother. Are we sure that we could not reasonably describe their exchange of letters, then, as a meaningful personal relationship? (Remember that from the point of view of the religious realist, this could be a precise analogue, for they take it that the God with whom others might have a ‘non-doxastic personal relationship’ does exist.)Footnote ¹⁴

Note also that it seems to make sense to think that a person could take inspiration or guidance from someone whom they fail to believe exists (and even from someone whom they believe not to exist). So, those who believe or know that God exists might perfectly sensibly think that those who lack warrant for belief in God could nonetheless benefit from being guided by a conception of God which though not bolstered by any doxastic assent to his existence could still amount to, in fact, being guided by the best possible exemplar of love, or justice or whatever. Think, by way of an apt analogy, of the way a person might be guided by their understanding of what Sherlock Holmes would do, were he to be faced with a problem to solve. They would not need to believe that he ever existed in order to be convinced that he represents a paradigm of rationality which one might aspire to emulate. And it need not be that one simply recognizes in Holmes a manifestation of those aspects of rationality which one already grasps: thinking about what Holmes would do might be a way of coming to understand new aspects of thinking rationally. This is all true whether or not Sherlock Holmes existed, and whether or not he is believed to have existed. When it comes to God, it might be that aspiring to emulate his nature is only morally or spiritually worthwhile if he exists. But the realist thinks that he does exist, so that is no problem. What matters, here, is that given that God exists, and given that we all therefore have a reason to emulate his nature, it is not unreasonable to think that anything which puts us in a position to do that is a good thing, and existential beliefs are not the only attitudes which put us in that position.

Of course, full-blown religious realists are not bound to think that non-doxastic faith is just as good as faith which involves proper belief, or even knowledge of the existence of God. The point is rather that if you are the sort of religious realist who thinks that some faith is always better than no faith at all (better, that is, for anyone), and you think that one ought not to believe without sufficient warrant, then you might reasonably recommend non-doxastic faith to others as a non-ideal way of avoiding the worst (which is to have no faith of any sort). In this, you might be like the doctor who knows that drug A is the most beneficial to those for whom it is suitable, but that their patient ought not to have drug A because it is harmful to them in some other way, and so gives them drug B which is not so comprehensively effective (it cures the most serious symptoms, but leaves plenty of less serious ones untouched). Indeed the analogy can be extended, for just as our doctor might hope that their patient would be strong enough to take drug A (and thus be cured of their residual symptoms) once they have been restored to some extent by drug B (though they could not take drug A in their current weakened state), the religious realist might hope that a person who has non-doxastic faith might be in a better position than someone who has no sort of faith to enjoy (perhaps because they are more open to) the sorts of experiences which would rationally ground full-blown belief in, and perhaps even knowledge of, the existence of God. (Perhaps this is what Pascal had in mind.) Indeed, if having non-doxastic faith puts a person in a particularly advantageous position with respect to enjoying those sorts of experiences, then that fact might be sufficient reason for a religious realist to recommend that sort of faith.Footnote ¹⁵