Skip to main content Accessibility help
Hostname: page-component-559fc8cf4f-s5ss2 Total loading time: 0.593 Render date: 2021-03-06T12:12:31.192Z Has data issue: true Feature Flags: { "shouldUseShareProductTool": true, "shouldUseHypothesis": true, "isUnsiloEnabled": true, "metricsAbstractViews": false, "figures": false, "newCiteModal": false, "newCitedByModal": true }

Article contents


Published online by Cambridge University Press:  11 December 2019

E-mail address:


Religious beliefs are often criticized as lacking the rational justification we expect of factual knowledge claims. In this article I suggest that while religious believers do often claim ‘knowledge’ of the ‘truth’ they typically use these words in traditional, and indeed still current, senses that are quite different from the senses assumed both by their atheist critics and by standard theories of knowledge. The claims are not primarily claims of factual accuracy, subject to the norms of what philosophers call theoretical reasoning, but claims of acquaintance with what can be trusted in the making of practical judgements, and subject to the rather different norms of practical reasoning. This does not exempt them from rational critique, but it does call for a different kind of critique from that usually offered.

Research Article
Copyright © The Royal Institute of Philosophy 2019

A common atheist criticism of religion is that holding religious beliefs is irrational. What is meant by this is that when you treat religious beliefs as knowledge claims and apply to them the rational criteria we have developed to assess knowledge claims in all other fields, they comprehensively fail. In an earlier article I argued that what religious believers typically mean by ‘belief’ is in fact very different from what their atheist critics, and philosophers generally, mean by the word. Outside religious fundamentalism, I argued, belief in the religious context is often not concerned with knowledge claims at all, and even when it is these may be heavily qualified. I also commented in passing that similar reservations apply to religious uses of the words ‘truth’ and ‘knowledge’. In this article I want to expand on those comments. Focusing on those cases in which religious belief can be seen as making or implying knowledge claims, I shall suggest that what distinguishes religious from atheist approaches to such claims is not so much different standards of evidence, as is generally assumed in the literature, as different conceptions of truth and knowledge. When believers claim the kinds of truth and knowledge that are taken for granted by their critics, they are indeed being irrational. But that is not generally what they are claiming and the criticisms are consequently, to some extent at least, misplaced.

According to standard philosophical theory, knowledge is true justified belief or, in philosophers’ shorthand, TJB+, where the ‘+’ allows for some modifications, the details of which are much debated, needed to rule out cases in which a person's beliefs are justified, and happen to be true, but only accidentally and for reasons unconnected with the justification. A key set of assumptions underlying this formulation concerns the concept of truth: that the word ‘true’ applies to propositions; that it refers to their accuracy or correctness; that it is in some sense objective – there is some absolute truth of the matter; and that it is effectively binary. Any meaningful proposition must be either true or false, albeit sometimes with probabilities attached (so we might say that a proposition has a probability x of being true and a probability of 1 − x of being false, but it can't be both true and false, or neither true nor false, or a bit of both).

There is a practical problem with this formulation, in that the only way in which we can determine what is true is by means of a process of justification. It is consequently this process of justification, rather than any objective truth, that is crucial in assessing knowledge claims. We conventionally say that a proposition is true if its content corresponds to how things really are in the world, objectively, but in all but the simplest cases this is not self-evident. We have to work in practice with some kind of reasoned collective judgement: the judgement, for example, of an expert scientific community. Recognizing this difficulty, some philosophers suggest that we treat a proposition as true if it forms part of a coherent and self-consistent set of propositions that are consistent with what reality throws at us by way of evidence. Others suggest that we treat as true whatever works, pragmatically, in our encounters with reality, where reality is defined as whatever is outside the control of our will. On either theory knowledge in the sense of a shared state of knowledge (what we collectively know) comes down to justified belief. And knowledge in the sense of an individual's knowledge becomes belief for which both the individual and the wider community have reasoned justifications. So, for example, we may know as a community that elementary particles follow the laws of quantum mechanics, because rigorous scientific experiment and reasoning have shown this to be the case, while Jack may personally know the same fact because he has read it in a textbook, the authority of which he has very good reasons to accept.

Most working scientists take the view that there is in fact some objective truth about the way things are, and establishing this truth is the aim of their research. They also recognize, however, that what we take to be the truth at any particular time is both approximate and provisional. Newton's laws of motion, for example, were taken for two hundred years to be true. We now know them to be only approximately true, under certain conditions. The same can be said of much that is counted as scientific knowledge, an important characteristic of which is that the process of justification is always open to challenge and revision, both through the gathering of new evidence and through the critical assessment and analysis of existing evidence.

This is achieved by imposing some basic criteria on what can be admitted as hypotheses or knowledge claims and on the processes by which these are tested. A knowledge claim, to be taken seriously, must lead to predictions that can be tested empirically and must, in particular, be open to refutation. Where alternative hypotheses are in play, they should be tested not only against the observable evidence but also against each other, to see which better fits the evidence. And the observations drawn on must be both open to scrutiny and replicable by others.

These criteria apply not only to scientific but also to everyday knowledge claims. If you claim that there are flowerpots at the bottom of the garden, anybody can in principle come in, take a look, make sure they are there and make sure that they are indeed flowerpots, and not some things that just happen to look like flowerpots from a distance or in a certain light. The criteria are met. If you claim that there are fairies as the bottom of the garden, which while visible to you are undetectable by other human beings or by any kind of recording devices, the criteria are not met. We might reasonably conclude either that you are making it up, or that the fairies you (genuinely) see are a figment of your imagination.

Atheists charge that the evidence offered in support of religious belief is of the fairies rather than the flowerpots kind. Typically it is evidence from divine revelation in one form or another – through supposedly sacred texts, through personal testimony, and above all through personal experience. Christians, for example, tend to rest their knowledge claims on the authority of the Bible as a revealed text, but if you press them on why they trust in that authority they will generally refer back to their personal experience, or to the personal experiences of people they trust. This kind of evidence is inherently subjective, untestable and immune to challenge. The observations might in some sense be shared across a community. There might be a great many people who, when they go to the bottom of the garden, see fairies. But there are also a great many people, adherents of other religions, who see elves. And there is no procedure you can specify by which anyone sceptical of these claims could in principle see either, and no way in which the rival claims could either be tested against each other or refuted. The evidence is good evidence of a shared psychological state of belief, but this is readily explicable in terms of well-understood psychological and socio-psychological factors. Additional hypotheses, in terms of religious knowledge claims, lead to no testable predictions and are, by our normal epistemological standards, redundant.

In some cases, Christian writers have been tempted to respond by accepting (wittingly or otherwise) the underlying assumption of what constitutes a valid knowledge claim and attempting to rebuff the charge that their evidence is insufficient to support such a claim. Some argue that rational enquiry is not limited to science and that there is no reason why the rules of reasoning applicable to religious matters should conform to those applicable to matters of natural fact: religious reasoning, they suggest, has its own rules, within which evidence from personal experience is perfectly valid. Others argue that the evidence on which they draw, though it can't be directly tested, is well validated by processes of critical scrutiny and cross-examination. These may not provide the standards of rigour expected in the natural sciences but they are sufficient to eliminate intentional deceit or evident madness and are not significantly different from those used in other areas, such as the interpretative discourses of history and the humanities.

Neither of these responses meets the objection. As we have noted, the epistemological norms on which the atheists rely are not peculiar to science, and according to any of our normal standards of reason the evidence cited in support of religious claims, while good evidence of a propensity to believe, is not good evidence of what seems to be claimed. Those norms are, however, tied to particular notions of knowledge and truth, and one question we might reasonably ask is how far religious knowledge claims share those notions. To what extent are they in fact claiming what atheists, and some of their Christian antagonists, assume?

In the early 2000s the broadcaster Joan Bakewell conducted a series of radio interviews on the subject of belief. In one of them she asked Timothy Radcliffe, former Master General of the Dominican Order of friars, about literal interpretations of the Bible. ‘I think’, he said, ‘the Gospel writers had a much richer sense of truth than we tend to.’ This is the sort of comment that immediately arouses atheist suspicion, but there is an important point behind it. Bernard Williams, in his Truth and Truthfulness (2002), identifies two components of our conception of truth: accuracy, or correctness, and sincerity, or trustworthiness. Theories of knowledge are concerned almost exclusively with accuracy, and when atheists critique religious belief it is the accuracy of its claims they reject. The origins of the English word ‘truth’, however, lie in the concepts of sincerity and trustworthiness, and in a religious context this is still the dominant aspect.

Until relatively recently, in historical terms, ‘truth’ denoted a personal quality. It was a person who was true or false, not a proposition, and in judging someone true or false it was their character you were judging – their sincerity, integrity, loyalty, trustworthiness, steadfastness, constancy – not their statements. And we still talk today of a true friend, or of true love. In the biblical context, the word was used to translate Hebrew, Greek and Latin terms (èmèt, alētheia, veritas) that had closely related connotations. In each case the contrast was with what was misleading, inconstant, deceptive, insincere or mendacious, and not with what was inaccurate or incorrect.

As Williams insists, the aspects of accuracy and sincerity can never be completely separated. If a person was considered sincere, one had good reason to trust in the accuracy of their statements. Indeed, in a pre-scientific context, the accuracy of a statement or proposition was generally treated as a function of the ‘truth’ of its author, rather than of any objective test of its accuracy, and to describe something as true was to say both more and less than to say it just was. More, in the sense that it came on trustworthy authority, and would therefore not mislead. Less, in the sense that it remained provisional: only God could know the ultimate truth.

In the course of the seventeenth century, the terms ‘true’ and ‘truth’ gradually lost some of this anchoring in personal trust and trustworthiness and began to take on more modern propositional meanings, especially in a scientific or philosophical context. And by the latter part of the nineteenth century, these meanings were well established in popular use. Truth came to denote primarily the accuracy of a statement or proposition, without referencing the sincerity – or even the accuracy, seen as a personal virtue – of the person making the statement.

We still retain today, however, the opposition between telling the truth and lying, so that to describe someone as telling the truth or not telling the truth references their behaviour at least as much as the content of their assertion. And in a religious context the older sense of the word remains dominant. Propositions such as the claims of the gospels or the Church's articles of faith are considered by Christians to be true in the sense that they come on reliable authority (the divine authority of revelation) and can be trusted (‘believed in’ in the original sense of that phrase) not to mislead. But only within Protestant fundamentalism are they considered to be true in the sense of being accurate statements of the facts. Indeed most theologians and a great many practising Christians would say that as statements of literal fact they just don't make sense. They are at most attempts to describe symbolically, in human language, things that can't be described in human language, and they can only ever constitute partial and incomplete glimpses of what might be called the ultimate truth. So a Christian might, for example, state that the bread of the eucharist is ‘truly' the body of Christ, but without for a moment thinking of this as a natural fact.

Similar considerations apply to the language of ‘knowledge’. To ‘know’, in English, originally meant something like to be acquainted with, and referred primarily to people. You knew a person, you were acquainted with them, you had experience of them, you recognized them. This knowledge might range from the highly intimate, in the Old Testament sense of knowing someone sexually or in the sense of understanding someone particularly well (as, for example, in the Delphic injunction to ‘Know thyself’), to the remote, in the sense of knowing of someone, without necessarily having met them. By extension you might ask somebody whether they would know somebody else, i.e. recognize them, if they met them, or you might ‘know’ something other than a person – you might know pain, for example, or grief. All these shades of meaning corresponded roughly to the German kennen or erkennen, the French conna ître, the classical Greek gign ōskein (from which gnosis) and the Latin novisse and cognoscere, all of which share the same root. And it was this kind of knowing that was referred to in a Christian context: the knowledge of God, say, or of the Christian gospel. To know something, in a religious context, was to be acquainted with something, the reliability of which you trusted.

Of course, English people also knew about things, and whereas other languages had other words (German wissen, French savoir, Greek epistasthai, Latin scire, etc.) English used the same words, ‘know' and ‘knowledge'. So knowing might also refer to something learnt: how to do something, how to behave properly, a trade or profession, a text learnt by rote, or a simple fact. Inevitably the meanings intermingled, so that knowing in a factual or propositional sense carried shades of the root meaning of personal knowing and vice versa. And over the centuries the propositional meaning has become much more dominant, and the personal meaning much less so. The claimed knowledge of a person, or indeed of God, came to be met with demands for factual evidence: you say you know him, but what evidence do you have for that? So while we still invest the idea of knowing a person with an expectation of knowing something about them, we are much less prone to invest the idea of knowing a fact with an expectation of personal acquaintance and authority.

In a religious context, however, the original meaning is still dominant. When people talk about knowledge in a religious context, it is primarily some kind of personal acquaintance to which they refer. Sometimes this is explicit. Christians will often say, for example, that they have come to know Jesus, by which they mean that they feel the presence of some being or some experience they identify with Jesus, not that they know facts about him (though they might, of course, claim that as well). Sometimes it is implicit. They might say, for example, that they know the truth that Jesus is the Son of God, and this looks, on the face of it, like a standard knowledge claim – except that, taken literally, it is incomprehensible. What they mean is that they hold in their mind some notion of ‘Jesus as the Son of God’. They are intimately acquainted with the notion and they put their trust in the sincerity of the texts and experiences that have brought them to this acquaintance.

This use of language puts their claims on a different footing from that assumed by their critics. To the extent that religious believers claim knowledge of truths in the sense in which both philosophers and atheists use those terms, they are engaging in what we sometimes call ‘theoretical’ reasoning, and can quite reasonably be criticized for falling far short of the standards of reasoning we call for in such cases. But to the extent that their claims are about acquaintance with the sincerity and reliability of religious ideas, they are engaging in what we call ‘practical’ reasoning: the kind of reasoning we use to make practical judgements as to how to proceed in any particular circumstances. And the standards we apply here are different.

Practical reasoning is about making judgements and decisions consistent with our experienced sensations, feelings, desires and beliefs, and explaining these judgements and decisions by giving appropriate reasons. It is the kind of reasoning we use when we decide to take off a coat because we're too hot, or choose not to take it off because although we're too hot we'd be embarrassed by what we're wearing underneath. It is also, more substantially, the kind of reasoning we use when deciding where to live, what jobs to apply for, or how to live our lives. To be considered sound, this reasoning has to make a certain kind of sense, but it lacks the precision and certainty of sound theoretical reasoning. The premises and assumptions on which we draw are typically complex, multi-faceted and open-ended. And in the case of the feelings and desires that motivate our actions they are not strictly replicable. Moreover, we usually have multiple aims and desires, many of which conflict, and few of which can be formulated clearly, let alone measured. Some reasons weigh more heavily on us at some moments, others at others. Some weigh more heavily on some people than they do on others. We make a decision, but we could just as reasonably have made a different decision.

In this context, the demands of rationality are inevitably moderated. We expect people to be broadly consistent in their reasoning, to take account of known facts, and to avoid blatant contradictions or factually erroneous assumptions. But we don't expect logically rigorous justifications, and we accept that what count as reasons for Jack may not do so for Jill and vice versa. And while we might well not share someone's aims and objectives we don't condemn them as irrational for having those aims and objectives, providing that they are broadly self-consistent.

Religious beliefs arise and function in a practical context. They go with being a member of a community, identifying with a community, holding to a moral code, taking part in ritual practices. And they acquire their salience from their application to practical problems: how to behave in a certain situation, how to cope in an emotional crisis, how to make and build friendships, how to help other people. In this context they are almost always, in the first instance, the products of practical rather than theoretical reasoning. This may be simply a matter of trusting in the authority of parents and teachers, or finding that the beliefs you inherit or come across in your friends seem to meet your needs. It may be a matter of finding in religious beliefs something that helps you make sense of your emotional experiences in a way that nothing else does. Whatever the particular route, the beliefs are the products of reasoning about what it makes sense in practice for you to trust in, and not of reasoning about what is factually true.

Some believers go on to argue that what is true for them, in the sense of being a practically reliable guide, must be true more generally, as a matter of fact. They shift their ground from the judgements of practical reasoning to the justifications of theoretical reasoning, and in doing so they open themselves up to the charge of irrationality. But many don't. They hold to the ‘truth’ of their religion and to the ‘knowledge’ that gives them access to that truth, but they don't confuse, and don't pretend to confuse, that knowledge and truth, and the practical reasoning on which it's based, with the knowledge, truth and theoretical reasoning appropriate to natural facts.

Atheists tend to see this as a fudge. From their perspective, knowledge is knowledge, truth is truth, and religious beliefs are factual knowledge claims like any other. To use words like ‘knowledge’ and ‘truth’ in other ways is simply to muddy the waters and dodge the issue. But this response seems to me both lazy and itself unreasonable. It's not, after all, as if believers were using words in a Humpty Dumpty way, or with intention to deceive. They are using them in a way in which they have always been used, and still are used in other contexts in which practical judgements are called for. Nor is it as if practical reasoning is itself immune to criticism. An atheist might well find much to question in the practical reasoning of religious belief, especially in the notion of revelation and the uses made of this. All practical reasoning relies on some assumptions of fact, and while it can be quite reasonable to make such assumptions without proper justification – if, for example, these correspond to the particular situation with which you're most concerned in reaching your judgement – you will need very strong reasons indeed to make assumptions that are demonstrably false. It may be reasonable, under certain circumstances, to live in a fantasy world, but unless you're seriously ill it's unlikely to be wise.

A strong commitment to a practical judgement also runs the risk of objectification. Much as atheists tend to assume that their meanings of truth, knowledge and belief are universal, Christians often assume that the practical reasoning that works so well for them must work for everyone else too, and so slide into making objective knowledge claims of just the kind atheists find irrational. Not everyone falls into this trap, however, and questioning the assumptions people pragmatically make in living their lives is a very different matter from questioning their theoretical reasoning. It is typically about questioning their wisdom rather than their rationality, and it needs to take careful account of their personal circumstances, motives and desires.

Full text views

Full text views reflects PDF downloads, PDFs sent to Google Drive, Dropbox and Kindle and HTML full text views.

Total number of HTML views: 15
Total number of PDF views: 100 *
View data table for this chart

* Views captured on Cambridge Core between 11th December 2019 - 6th March 2021. This data will be updated every 24 hours.


Send article to Kindle

To send this article to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about sending to your Kindle. Find out more about sending to your Kindle.

Note you can select to send to either the or variations. ‘’ emails are free but can only be sent to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

Available formats

Send article to Dropbox

To send this article to your Dropbox account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Dropbox.

Available formats

Send article to Google Drive

To send this article to your Google Drive account, please select one or more formats and confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your <service> account. Find out more about sending content to Google Drive.

Available formats

Reply to: Submit a response

Your details

Conflicting interests

Do you have any conflicting interests? *