Skip to main content Accessibility help
Hostname: page-component-559fc8cf4f-xbbwl Total loading time: 0.31 Render date: 2021-03-06T12:37:03.101Z 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:  04 June 2019

E-mail address:


Atheist critiques of religious belief tend to assume that this belief is of broadly the same kind as that treated by philosophers: a disposition to take as true the content of some propositional knowledge claim. I argue here that it is not. What is primarily meant by ‘belief’ in a religious context refers to something quite different, and while this kind of belief sometimes implies belief in the philosophers’ sense it doesn't always do so, and even when it does it is with significant qualifications. Similar considerations apply to other words, such as ‘truth’, ‘knowledge’ and ‘meaning’ itself. In each case atheists tend to impose on religious discourse a particular philosophical meaning rather different from that typically intended.

Research Article
Copyright © The Royal Institute of Philosophy 2019 

In his recent book, The Meaning of Belief: Religion from an Atheist's Point of View (2017), Tim Crane sets out to provide a richer and more realistic view of religion than he finds in the New Atheist critiques of Dawkins, Dennett, Grayling, Harris and Hitchens. Religion, he suggests, is not just about belief but also about rituals and routines, moral practices, a sense of belonging and identification with a community. Given this context, believers are unlikely to be swayed by the kind of arguments advanced by the New Atheists, powerful as he thinks they are. Indeed, while most religious beliefs are, in his view, false, the religious package as a whole has evident benefits and seen in this broader context the holders of these false beliefs may not necessarily be acting irrationally.

Anthropologists and sociologists have long treated religion as a form of social practice, and the recognition of this dimension in a philosophical context is welcome. Religion is surely not just a matter of belief. But it begs a question. To what extent is it a matter of belief at all, in Crane's sense of the word? As he goes to some pains to make clear, religious belief on his account is exactly the kind of belief in which philosophers normally trade. It is a disposition to take as true the content of some propositional knowledge claim, which to be meaningful at all must be either true or false: in this case a claim such as ‘God exists’ or ‘There is some unseen order’ or ‘There is some transcendent being’, together with a range of claims specific to different religions.

From the perspective of a philosopher or of an atheist, this assumption about the meaning of belief is unremarkable, and religious writers too have often seemed to go along with it. But it is not what people have historically meant by belief, either in the context of religious profession or more generally. Nor is it quite what people tend to mean by belief in everyday conversation today. Religious belief as understood by religious believers is, I shall suggest, generally something quite different from the kind of epistemic, propositional belief assumed by their critics. And while it may sometimes entail or imply such epistemic belief, that is not something we can take for granted. The relationship is, to say the least, problematic.

Although I shall focus in this article on the meanings of religious belief, as an important topic in itself, it is worth noting that there is a much broader underlying issue. Analytical philosophy depends on rigorous, logical argument, and that is only possible if the terms of the argument are clearly and tightly defined. Crane himself notes at one point, in reference to a quote from Quine, that debates often flounder because opponents use the same words with quite different meanings, and we are not going to be able to build any kind of logical argument if our definitions shift from one step to the next. We get round this problem in the sciences by specifying measurements, so that a word like ‘force’ has a technical meaning in physics quite distinct from its meanings in everyday conversation. Similarly in philosophy, we use words like ‘belief’ with well-defined conventional meanings, and this enables us to make rigorous arguments. But they are arguments about belief as we define it and may be of only limited relevance to belief as experienced and understood in the wider world.

To illustrate this point, consider the title of Crane's book: ‘The meaning [definite article, singular] of belief’. There is an implicit assumption here that the word ‘belief’ in this context has only one meaning, or only one correct meaning. If someone ‘believes’ in a different sense from the philosophers’ sense that just doesn't count. As we shall see, however, religious belief is not generally the same as epistemic belief. There is also an underlying assumption about another word, ‘meaning’, namely that this applies to words (like ‘belief’) and statements.

In philosophy the meaning of a word is said to be determined either by reference to some objective reality or, on the more dominant view, by reference to some shared consensus on how the word may be used, in what Wittgenstein called a language game. There is also an assumption that a ‘meaningful statement’ is one that has a single, unambiguous meaning, and that is either true or false. We use the word in a similar way in everyday life when we refer to a dictionary meaning. But look up ‘meaning’ in a dictionary and you will find, first, that it has many different meanings; and second, that almost all of these are quite different from the one assumed by philosophers. In particular, the meaning associated with a word or statement refers primarily to the intent of the speaker, and not to the word or statement itself: to what I mean or you mean. It is personal and subjective. Even the objective-sounding ‘meaning of life’ refers to what is intended by God: no God, no intent, no meaning.

In many contexts it doesn't matter much if what I mean by a word and what you mean are different. For everyday purposes rough agreement is generally sufficient and where it isn't we can resolve differences by pointing to examples. But when the reference is to concepts rather than objects, as it often is in philosophy (to knowledge, for example, or belief or truth or justice or goodness, or indeed God), things get much more difficult. Concepts exist in our minds and they can only be defined in terms of other concepts. They can be illustrated and analysed, and if you and I differ on what we mean by justice, as we might well, we can certainly debate the merits of alternative conceptions. But there are no grounds on which I can say of your conception ‘That's not “justice”.’ The most I can say is ‘That's not what we normally mean by “justice” within a particular discourse’ – with the suggestion, perhaps, that you don't properly understand the discourse. But you might reasonably respond ‘So much the worse for the discourse. I understand your conventions, but justice is not just a matter of conventional definitions, it's more important than that.’

We can reasonably talk of ‘the meaning of belief’, but only (in line with Wittgenstein) as an approximately shared intent within a particular context and a particular community – a community of philosophers, say, or a community of like-minded religious. What we cannot reasonably do is assume that the same intention is shared across contexts or across communities. And it turns out not only that the words ‘belief’ and ‘believe’ can have many different meanings, but that as with ‘meaning’ itself most of these are quite sharply at odds with that assumed by philosophers.

Until relatively recently, in historical terms, the word was never applied to propositions, and people never spoke of ‘believing that’. One believed a person, or believed in a person, and what was typically meant by this was that one put one's trust or confidence in them, committed oneself to them, or held them dear. (The root of the word is the same as that of ‘love’ and the word ‘beloved’ still captures something of the old meaning of ‘believed’.) The verb ‘believe’ was used in a religious context as a translation of the Latin verb credo, which carried very much the same range of meanings and was also never used of propositions. If you wanted to say in Latin that you believed that such and such was the case, the word you would use was opinio. Credo itself was used to translate the Greek pisteō, which had a slightly different range of meanings including elements of loyalty and obedience, but again carried no connotations of propositional belief. An appropriate translation of the Nicene Creed around the time of the Reformation would begin something like ‘I commit myself to one God …’ or ‘I pledge allegiance to one God …’. The elements of loyalty and obedience were also captured in the Latin use of the word fides as the noun corresponding to credo, which was at first translated into English as ‘belief’ but later as a new word ‘faith’.

The meanings of words vary from time to time as well as from person to person and context to context, so what people mean by ‘believe’ today is not generally what they meant by it five hundred years ago. The Reformation brought some changes, the Scientific Revolution brought others, and by the late nineteenth century, in the wake of Darwin's evolutionary theory and early attempts to formalize scientific method, scientists and philosophers were beginning to use the word in much the way philosophers do today, as indicating some kind of knowledge claim. The older meanings didn't disappear, however, and in a religious context, especially, the word has retained many of its earlier connotations.

When people use the word ‘believe’ in ordinary conversation today it is usually in connection with religion. They may use it in other contexts in which values come to the fore, as in moral or political beliefs, but even here there's usually some religious or pseudo-religious connotation. We don't normally assert that we ‘believe that X’, where X is something that can be settled as a matter of simple fact or through scientific reasoning, even if we're philosophers. We either just state the fact or, if we wish to express doubt or certainty, use words like ‘I think that X’ or ‘I know that X’. The one context in which we might say that someone ‘believes that X’ is where we think they're wrong, and this points to another feature of our everyday use of the language of belief, that it typically carries connotations of error or doubt. To say that ‘I believe that X’ is less likely to be an objective truth claim than a polite way of pushing one's point while refraining from such a claim. To say that ‘He believes that X’ almost always carries the suggestion that he might well be wrong, and often implies that he is wrong.

These connotations are even more apparent when we use the phrase ‘believe in’. If I say that I believe in something, it is typically a statement of trust and commitment to some ideal, and always carries the recognition, however faintly, that my trust may be misplaced. If there were no such risk, there would be no need for trust. Similarly, to say that ‘She believes in something’ is almost always to cast doubt on her judgement, to assert that her trust is misplaced. We might say that Jack believes in astrology. We are unlikely to say that he believes in chemistry. If we say that he believes in fairness or democracy we are probably suggesting that he is a bit of an idealist. And if we use the phrase ‘believe in X’ as synonymous with ‘believe that X exists’, it is with the clear implication that X does not exist. We might say that Jill believes in ghosts; we are unlikely to say that she believes in black holes. And if someone were to say that Jill ‘believed' in climate change, you would probably, and quite reasonably, conclude that they didn't!

Given the way our language has evolved, it is tempting not only to translate any profession of religious faith (‘I believe in God’) into an objective knowledge claim (‘I believe that God exists’), but also to presume that claim is false, or at least unfounded. Before going down that road, however, we really should stop and ask what it is, or rather how it is, that religious believers believe. If they say that they believe in God, or in Jesus Christ, for example, what do they mean? What is the intention behind the statement? And to what extent does this or does it not entail epistemic belief of the kind assumed by philosophers?

The first of these questions cannot be answered definitively. Any religion embraces a wide range of personal beliefs. If we are to make any progress at all, however, we need to give some kind of answer, at least to give a rough indication of the range of possible meanings. To keep things relatively simple, I shall begin by restricting the discussion to Christian belief, and shall attempt to illustrate the range of attitudes covered through four very loosely characterized approaches.

It seems clear, first, that there are many Christian believers whose beliefs do take more or less the form assumed by philosophers and by their atheist critics. They believe that there is a God; that Jesus Christ was both a really existing biological human being and that he is also, uniquely among humans, the Son of God; that he died and was resurrected; that we too will be resurrected, in body; that the Bible is a work of divine revelation, and so on. Moreover they understand these claims, and those of the Bible generally, as being true in a straightforward factual and indeed literal way – in much the same way, in other words, that philosophers understand propositions to be true or false.

From an atheist perspective, these believers (we might call them fundamentalists, but without the connotations of extremism we tend to associate with that word) are a soft target. Taken on their own, their beliefs cannot be justified on any rational grounds, and attempts to provide justifications within the philosophical literature are probably tolerated only out of courtesy. But that is often not the point. Not only are the beliefs supported by a context of belonging, identification and emotional experience, they are typically a product of that context. The function they serve is not so much to make knowledge claims, though they do, as to make sense of powerful emotional experiences, bind people in a community, and demarcate that community from others. The community might be focused on a particular ‘truth’ that others seem to neglect, or on an authority figure, or on the rejection of ‘false gods’. But whatever the focus the element of demarcation appears to be central, and that may well be served better by beliefs that people can't accept on purely rational grounds, beliefs that require commitment rather than mere acquiescence.

At the other end of the spectrum is another large group of Christians, whom we might assume to be believers but for whom belief of any kind is beside the point. For many Christians, including many churchgoers, religion is about a set of ritual practices, identification with a community, and a shared set of moral values, in which they choose to participate but without any particular sense of trust or commitment, let alone any epistemic beliefs. They may well join in the liturgy and recite the Creed, but they do so as part of a ritual performance, the significance of which is quite independent of its linguistic content. Should we call them Christians? Well, they call themselves Christians, and the churches are happy to treat them as such.

So far, so relatively straightforward. But a great many Christians, including a great many clergy and, I would suggest, the great majority of theologians, fit into neither of these groups. My third group of believers might be characterized as holding epistemic beliefs of the kind rejected by atheists, but in a more modest way than the first group, and with an acute awareness of the limitations of language and human cognition. This is the province of much Christian theology as well as much lay belief, and the approach differs from the norms of analytical philosophy in important respects.

First, it assumes that both language and thought are always imperfect and imprecise, that any statement is open to multiple interpretations, and that all knowledge is approximate and provisional. This assumption is not at all unreasonable. Indeed, it's shared by many working scientists and it is a standard tactic for atheists to charge religious believers with claiming a degree of certainty that science, which is also open to empirical critique, doesn't. That charge can only be sustained, however, if at all, against the first of our groups. The believers in this third group may profess subjective conviction, but they don't generally profess epistemic certainty. And whereas philosophers typically respond to the limitations of our language and comprehension by restricting themselves to that which can be treated more or less logically, and treating everything else as meaningless, Christian thinkers quite reasonably prefer to forego that precision in favour of a more comprehensive view.

They also use the words ‘truth’ and ‘knowledge’ rather differently from analytical philosophers. As Bernard Williams has observed (Truth and Truthfulness, 2002), our ordinary concept of truth includes two elements, accuracy and sincerity. When philosophers talk of a proposition being true or false, the reference is to its accuracy only. However, the historical meanings of the word are associated more with sincerity. ‘Truth’ is allied to ‘troth’ and like ‘belief’ it was historically applied to people, not propositions, a true person being one who was constant and straight and could be trusted and relied upon. When we talk of a person being true or truthful, that is still where the emphasis lies. In the Christian context the ultimate accurate truth is considered to be beyond our comprehension. Something is considered ‘true’ if it is thought to contain some element of that truth, inevitably partial and poetically expressed, and if it is to be trusted, and is not misleading.

Similarly with knowledge. To ‘know’ in English originally meant to be acquainted with, as in knowing a person, rather than to know about, and this is the primary sense in which it is still used in a Christian context. (Other languages use different words for the two conceptions, but in English they are conflated.) To know something is not to know it as an epistemic fact, but to be acquainted with it. The reference may be to the truth, but with the sense just noted, and with the explicit qualification that any human knowledge is partial: to be acquainted with someone is not to know everything about them, even if they're merely human, let alone if they're God.

Christians in this group do accept that in believing in God, in the old sense of the phrase, they are also admitting the epistemic claim that God is. But they also accept that God, the transcendent or ultimate reality, is ineffable, and that to the extent that we try to say anything about him at all we inevitably wind up casting him in our own human image. We have no way of seeing beyond the boundary of the immanent world, so what we describe is effectively what we see of our own human condition, reflected back from that boundary. Similarly with the characteristic claims of Christian dogma. If you take the claim that Jesus is the only son of God literally, it is not wrong: it is nonsensical. But if you take it as a necessarily inadequate attempt to capture poetically some very partial and obscure element of the human condition that we can never hope, given our limited comprehension, to pin down precisely, you might well feel justified in asserting its truth, in the broader sense of that word, and claiming to know it, in the sense of acquaintance.

There is, finally, a fourth significant group of Christians whose sympathies are probably with the third group, but who are not in any sense theologians and haven't thought through the kinds of questions I have been discussing. They believe in God and in Jesus, and in the gospels, in the old sense of commitment and trust, but simply struggle to go any further. If you push them as to whether they believe the claims of Christian dogma in an epistemic sense, they will give all sorts of answers, ranging from ‘not really’ to ‘I suppose so, sort of’, via ‘yes, in my own way’, ‘I don't not believe’, or just ‘I don't know’. You ask a question, they feel obliged to give some sort of answer, but the underlying intent is to the effect that epistemic belief is not really what it's about for them, and not something on which they want to commit themselves. The essence of their religion is a powerful story, and an associated set of values and ideals, in which they trust to interpret their emotional experiences and guide their lives.

The characterizations I have given are very crude, and no doubt many believers won't fit into any of them. But they serve to draw attention to the possibility that what Christians mean by belief may be very different both from what philosophers mean and from what atheists assume. They may not claim anything by way of propositional beliefs at all, and even when they do these may be heavily qualified.

Of course, intention is not everything, and the atheist may well argue that even though Christians do not claim to hold the propositional beliefs attributed to them, these are rationally implied by the beliefs they do hold. Hence, perhaps, the discomfort of my fourth group when pushed to go there. Philosophers have generally held that ‘believing in’ something necessarily implies believing that it exists, though it may have additional affective content too.

Where what is believed in is some kind of natural or supposedly supernatural entity, this seems uncontroversial. You can't believe in ghosts without believing that they exist. But where the object of belief is a thought or a concept, matters are not so straightforward. A belief in fairness, say, surely entails a belief in the existence of a concept of fairness, but that is quite compatible with a world that is wholly unfair. Moreover, the concept referred to need have no objective existence, outside our thoughts. If you believe in fairness, I ask you for examples, and say ‘no, that's not fairness’, I might have a basis for criticizing your use of language, but it is not a basis for criticizing your belief.

The question then arises, whether the Christian God is an entity or a concept – a thing or a thought – and the answer is surely that in so far as He is either of these, God must be a concept. People sometimes say that God exists, which would suggest an entity, but in the monotheist religions God is not an existent but the transcendent, unknowable ground of all existence – as one theologian has put it, God is the question, not the answer. So to believe in God entails having some concept of God that is meaningful to the believer, but that's probably as far as we can go.

Similarly with Jesus. To believe in Jesus qua human being would certainly imply his existence, but that is not what matters for Christian belief. What Christians believe in is Jesus qua Son of God, and that is surely a concept: a concept that exists, in meaningful if varying ways, in the minds of Christian believers, if not in those of their critics.

Underlying the debate between atheists and religious believers is a choice between two grounding assumptions: that what we are in principle able to know about the world, in the epistemic sense, is all there is to it; or that it is not. That choice is probably a matter of temperament. The question is: what can we reasonably build on the latter assumption? Like Crane and other atheists, I personally find the specifics of the Christian gospel unbelievable, epistemically or otherwise. But the grounding assumption is not at all unreasonable and if as philosophers we are going to criticize the beliefs people build on it, we should at least ask what and how they believe and not just assume that their intentions, in using the language of belief, are the same as ours. In this case they are surely not.

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: 11
Total number of PDF views: 76 *
View data table for this chart

* Views captured on Cambridge Core between 04th June 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? *