Originally published in Dialogue, Issue 41 November 2013.
“A wise man”, declared David Hume, “proportions his belief to the evidence”. Although the relationship between evidence and belief is sometimes complicated, this principle – which has become known as evidentialism – offers a straightforward rule-of-thumb when evaluating beliefs. In short, a belief is considered unreasonable, or unjustified, without sufficient evidence.
For Hume, it was a matter of wisdom. For later writers, however, the stakes were higher. William Clifford, writing about 130 years after Hume, restated the evidentialist position with greater force:
To sum up: it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence.
Clifford’s claim is normative – that is, he describes unjustified belief as wrong and insists we have a moral obligation to believe only with sufficient evidence. He makes this point with the compelling example of a shipowner who has reason to believe that his vessel is not seaworthy. Despite these doubts, the shipowner convinces himself of the vessel’s safety and trusts to Providence. When the ship sinks mid-ocean, with considerable loss of life, the owner collects the insurance money and tells no one of his doubts. Clifford rightly asks, What shall we say of him? Few would disagree that the shipowner is morally at fault, and Clifford makes clear the reason:
... because he had no right to believe on such evidence as was before him.
In this way, Clifford makes it a moral obligation to challenge and reflect on our beliefs until they are justified by the available evidence. Like the shipowner, we are at fault if we abuse the ‘sacred faculty’ of belief and to do so constitutes ‘one long sin against mankind’.
It seems reasonable to insist on a certain standard of justification for our claims. Ordinary beliefs should require evidence to support them; and extraordinary beliefs must therefore require extraordinary evidence. It is in the matter of extraordinary beliefs – that is, religious and supernatural claims – that the principle of evidentialism becomes controversial (it is no accident that Hume’s remark appears in a discussion of miracles). Evidentialism represents a significant challenge to religious faith: it is only reasonable to believe in God if there is sufficient evidence to justify that belief. If the evidence is, at best, merely probable then full belief in God is unwarranted – or, in Clifford’s less compromising language, morally wrong.
To answer this challenge, medieval philosophers attempted to provide exactly the evidence required: the traditional arguments for the existence of God, accounts of miracles and religious experiences – this attempt to ground religious belief in reason and evidence is known as natural theology and most philosophers nowadays agree that the project failed – certainly, few are convinced by the ontological argument or theories of intelligent design. For the purposes of this discussion, therefore, I will assume that there is not sufficient evidence for God’s existence to meet the evidentialist’s challenge.
But religious belief continues, despite the absence of evidence. As Clifford would have it, faith is the error made by the shipowner. Does that make faith irrational? There seem to be two possible approaches to defend faith against this charge. Since we have assumed that there is not sufficient evidence to compel belief in God, then for faith to be rational it must either be justified by something other than evidence, or be a special sort of belief that does not require evidence at all.
The first way to defend faith against the evidentialist is to claim that a belief can be reasonable even without sufficient evidence. Philosophers of faith often make this point, arguing that the value of religion does not depend upon its supernatural claims but its capacity to motivate people to acts of kindness and compassion. We can call this the pragmatic argument for faith. Pragmatists insist that a belief is reasonable if adopting it makes a practical difference, perhaps by making life better or leading to a more moral outlook. On this view, the truth of a belief is independent of its justification. Whether or not God exists is beside the point: what matters is the effect of believing. If faith leads to consolation or hope, a better quality of life or a more settled conscience, then it is rational to believe.
Perhaps the most famous pragmatic argument for faith was made by William James around 20 years after Clifford’s evidentialist challenge. The Will to Believe is:
... an essay in justification of faith, a defense of our right to adopt a believing attitude in religious matters, in spite of the fact that our merely logical intellect may not have been coerced.
James rejects the evidentialist’s standard of justification and argues that under certain conditions we are faced with a stark choice: remain sceptical or take a leap of faith. While scepticism protects us from error, a willingness to believe sometimes leads to rewards that the sceptic will never receive. This seems to be the case with close personal relationships: trust in someone’s good nature may be necessary for that person’s kindness to reveal itself. In other words, the evidence comes after the belief. The same may be true of explicitly religious beliefs: to have faith is to open up the possibility of a relationship with God, to meet the divine half-way. If so, then to follow the evidentialist’s advice – to suspend belief until the evidence arrives – is to pay a very high price. As James says:
... a rule of thinking which would absolutely prevent me from acknowledging certain kinds of truth if those kinds of truth were really there, would be an irrational rule.
In cases like these, James argues, it is rational to believe even without sufficient evidence. The rewards of faith are sufficient to justify the commitment, and the demands of evidentialism are misplaced.
This is an attractive argument: all it takes is a little courage to win access to hidden knowledge. Unfortunately, the same willingness to believe may also lead to deception and delusion, and without proper evidence it is not possible to tell the difference. James finds Clifford’s rule of sufficient evidence to be too limiting, but surely any rule is by nature restrictive: the rules exist precisely to help us distinguish between true belief and delusion. If we are too sceptical, argues James, we may miss out on valuable knowledge: but believing anything and everything, just in case, is simply foolish. To be fair to James, he advocates faith only under narrow conditions, but even so he is less interested in truth than in the advantages of belief. Clifford pre-empted this argument by insisting:
The fact that believers have found joy and peace in believing gives us the right to say that the doctrine is a comfortable doctrine, and pleasant to the soul; but it does not give us the right to say that it is true. And the question which our conscience is always asking about that which we are tempted to believe is not, “Is it comfortable and pleasant?” but, “Is it true?”
For Clifford, it is a matter of conscience, not comfort. The truth has value over and above its effect on one philosopher’s state of mind, and therefore such wish-fulfilment is neither rational nor morally acceptable. This is an austere but compelling position. Philosophers generally seek to escape ungrounded commitments, trusting instead to the value of rational enquiry and empirical evidence. James’ desire to escape these restrictions is understandable, human even, but hardly constitutes a rational defence of faith.
Interestingly, it seems that atheism can be defended by the same criterion. If a belief is considered reasonable solely on the grounds of the advantages it brings, then we might look to the experience of those who have abandoned the religious life and found more freedom and peace in their non-belief. Contemporary advocates of atheism (the ‘new atheists’) have repeatedly invoked this argument: religion is a harmful phenomenon and the sceptical life brings its own rewards. If this is the case, and there is evidence to support it, then we should be wary of the pragmatic argument: any criterion of rationality that justifies both belief and non-belief alike does not seem to be doing its job.
The argument is sometimes made that ‘faith’ in science also rests on pragmatic foundations. We are justified in taking certain basic propositions on faith (for example, the uniformity of nature) because to do so allows us to measure and describe the world more effectively. We can believe in science because it works. A pragmatist might then point to faith and say the same holds true: belief in God leads to peace and contentment. It just works. However, there is an important distinction to be made here. We say that science ‘works’ because of its truthfulness: predictions are made and met; practical problems are solved; diseases are cured; and so on. Truth matters in science. By the pragmatist’s justification of faith, however, a belief ‘works’ not because it is truthful but because it leads to satisfaction or improved behaviour. Truth is not the issue: in fact, the pragmatist claims that it is rational to believe whether or not the belief is true. Such a position reveals a profound disregard for the value of truth. It is instructive to consider the history of scientific discovery which has been consistently unsettling: from Galileo to Darwin and beyond, our comfortable beliefs have been undermined by the scientific quest for truth – not a pragmatic quest for comfort.
It seems, then, that the pragmatic justification for faith, in its failure to distinguish between true and false belief, is inadequate to guarantee rationality. Therefore the only remaining way to defend faith against the evidentialist’s challenge is to understand it as a special type of belief that does not depend upon evidence.
As the evidentialist notes, most of our beliefs emerge because of evidence presented to us: the rational response to sufficient evidence is to form a proportional belief. But not all of our beliefs arise in this way. When I see a laptop before me, or I remember recharging the laptop last night, I am forming beliefs that are non-inferential: that is, they occur immediately and are not inferred from other beliefs. Perceptions and memories seem to be self-evident and the beliefs that arise from them are called basic. Without these basic beliefs, we would not be able to make much progress with our thinking.
The important point here is that philosophers believe it is rational to form these beliefs, even though there is no evidence to support them. The special class of basic beliefs is immune from the evidentialist’s challenge. This is for two reasons. Firstly, there are criteria to determine which of our beliefs are legitimately basic: traditionally, those criteria require such a belief to be self-evident, evident to the senses or incorrigible (meaning I cannot be mistaken about the belief, such as ‘I am in pain’). Even so, some basic beliefs can be wrong: if I am prone to hallucination or false memories then of course my beliefs about the laptop can be called into question. So a ‘properly’ basic belief must also be warranted or supported by other considerations. In this case, since I have good reasons to trust my perceptions and memory at this moment, my beliefs about the laptop are warranted. It is this combination of immediacy and theoretical support that makes it rational to hold properly basic beliefs.
Alvin Plantinga has argued that belief in God satisfies these criteria. That is, faith is a properly basic belief and therefore does not require evidence to justify it. If Plantinga is correct, then the evidentialist’s challenge is irrelevant and we can acknowledge that faith is rational after all. In one sense, this might sound unsurprising: belief in God is of course fundamental to religious people (although not necessarily in the epistemological sense we are describing). But to many thinkers, the claim that belief in God is as basic as a perception of pain or a memory, sounds unlikely.
We can present here three main objections which seem to undermine Plantinga’s claim that faith is properly basic. Although a theist may form a basic belief about God’s existence, the value of that belief depends largely on the background theory by which it is warranted. If I perceive a horse where you perceive a unicorn, my basic belief is warranted by the likelihood of horses; your belief in the existence of unicorns, of course, is less of a warrant. These background theories that justify our basic beliefs are not themselves basic, and therefore they are contestable and open to argument. If belief in the existence of God is properly basic, then the hypothesis of theism must itself be reasonable – and several hundred years of philosophy have shown that it is, at the very least, arguable. Plantinga’s claim seems to imply that if God exists, belief in him may well be basic. But most thinkers are aware of the philosophical problems behind that ‘if’ which leave any basic belief in God largely unwarranted.
The second objection is acknowledged by Plantinga himself. If a religious person’s commitments are enough to warrant properly basic belief, then surely anything goes. The example Plantinga uses is the Great Pumpkin from the Peanuts cartoon: every Halloween Linus waits in vain for the Great Pumpkin to appear. Is such a belief properly basic? If so, then the criteria have surely been loosened too far. Plantinga denies this and claims that there is an obvious and relevant difference between belief in the Great Pumpkin and belief in God – but to the sceptic, any supernatural claim is as unlikely as any other.
Following on from this, Plantinga implies that the criteria for properly basic beliefs will differ from individual to individual. Followers of the Prophet, of Buddha, of L Ron Hubbard – each will arrive at different basic beliefs determined by the other theoretical commitments they bring with them. One of Plantinga’s own examples involves revelation through scripture:
Upon reading the Bible, one may be impressed with a deep sense that God is speaking to him.
And another, we might say, may be impressed with a deep sense of the unreliability of ancient documents, or the power of superstition over the pre-scientific mind. Plantinga accepts that the atheist will not form basic beliefs in the same way as the theist, defending a relativism about belief that is deeply unsettling. It seems that any belief can indeed be properly basic for you – and not for me – if it is compatible with your theoretical commitments. It seems implausible that a candidate for properly basic belief could be warranted for a Christian but not for an atheist. Plantinga here is rejecting a shared and universal rationality in favour of an epistemological relativism. What is right for you isn’t right for me. To reject this common framework of reason may be necessary for his argument, but it is too high a price to pay in order to rationalise a belief in God.
This discussion has assumed a strict criterion for the rationality of belief. The evidentialist challenge should be taken seriously because it is philosophically and ethically important to distinguish between justified and unjustified beliefs. It seems that faith does not meet that challenge; nor can it be justified pragmatically; nor can it claim to be a properly basic belief. There are good reasons to think that each of these criticisms is true. By the assumed standards, therefore, faith fails the test of rational belief.
Of course, for people of faith these epistemological questions may seem irrelevant. The desire to believe is, perhaps, a very human thing and there is no doubt that faith brings hope and consolation to many. However, the fact remains that many thinkers find faith problematic. Whatever reasons people may have for their religious beliefs, those reasons do not appear philosophically robust. In a world where religious claims spill over into matters of public policy, or affect the freedom of individuals, we should perhaps insist on higher standards for belief. The truth matters. And if Clifford is right, to trust one’s faith uncritically is to risk being irresponsible.