You are here

Series II

Lecture 9: Belief ‘In’ and Belief ‘That’

Lecture Note1

Part I

Introduction

Epistemologists have not usually had much to say about believing ‘in’, though ever since Plato's time they have been interested in believing ‘that’. Students of religion, on the other hand, have been greatly concerned with belief ‘in’, and many of them, I think, would maintain that it is something quite different from belief ‘that’. Surely belief ‘in’ is an attitude to a person, whether human or divine, while belief ‘that’ is just an attitude to a proposition? Could any difference be more obvious than this? And if we overlook it, shall we not be led into a quite mistaken analysis of religious belief, at any rate if it is religious belief of the theistic sort? On this view belief ‘in’ is not a propositional attitude at all.

On the face of it, this radical distinction between belief-in and belief-that2 seems plausible to anyone who knows from the inside what religious experience of the theistic sort is like. But to many philosophers it seems to have hardly any plausibility. It seems obvious to them that belief-in is in one way or another reducible to belief-that. This reduction, they would say, is not really very difficult, certainly not difficult enough to be interesting; and that, presumably, is why epistemologists have seldom thought it necessary to discuss belief-in very seriously. Why make such a fuss about this distinction between ‘in’ and ‘that’, when it is little or nothing more than a difference of idiom?

I wish to suggest, however, that the distinction between belief-in and belief-that does at any rate deserve careful discussion. The question whether belief-in is or is not reducible to belief-that is by no means trivial, nor is it at all an easy question to answer.

It is not trivial. Religious belief, whether we like it or not, is quite an important phenomenon. Those who have no religious belief themselves should still try to understand what kind of an attitude it is, and they cannot hope to understand it unless they pay some attention to what is said by those who do have it. Moreover, as we shall see presently, religious belief-in is by no means the only sort. Nearly everyone believes ‘in’ someone or something, whether he believes in God or not.

Nor is our question an easy one to decide. Quite a strong case can be made for each of the two views which have been mentioned, the ‘irreducibility thesis’ on the one hand, and the ‘reducibility thesis’ on the other. The decision between them is made more difficult (though also more interesting) because belief-in, or at least some instances of it, cuts across the boundary sometimes drawn between the cognitive side of human nature, concerned with what is true or false, and the evaluative side, concerned with what is good or evil. Either the boundary vanishes altogether, or we find ourselves on both sides of it at the same time.

There is also a preliminary enquiry, whose importance has not perhaps been fully appreciated by either party in this controversy. Neither perhaps has considered a large enough range of examples. The expression ‘believe in’ is used in a good many different contexts. For all we can tell beforehand, there might be several different sorts of belief-in, and the reducibility thesis might be correct for some of them, but incorrect or highly questionable for others. Let us begin, then, by considering ‘the varieties of believing-in’; and it may be as well to consider some examples of disbelieving-in too.

The Varieties of Believing ‘In’

First we will consider a number of examples which seem prima facie to support the irreducibility thesis. Whatever merits this thesis may have, it is certainly an over-simplification to say that belief-in is always an attitude to a person, human or divine.3

Surely it is perfectly possible to believe in a non-human animal The blind man believes in his guide-dog. A medieval knight or a modern foxhunter might easily believe in his horse. A falconer might believe in this hawk and not believe, or believe less, in that one. And vegetable organisms, as well as animals, can be believed in. A keen gardener might believe in his chrysanthemums, but not in his strawberry plants.

Moreover, it is not only living things which can be believed in. One may believe in a machine. A motorist can believe in his car. Or, if he is more discriminating, he may believe in some parts of its mechanism but not in others, or not much. He may have great confidence in his brakes but less confidence in his battery.

It is even possible to believe in a non-living natural object. Let us consider a remark attributed to the seventeenth-century English statesman, Lord Halifax the Trimmer: ‘The first article, in an Englishman's creed is “I believe in the sea”.’ It is true that for many centuries Englishmen did believe in the sea, though nowadays they would be better advised to believe in the air. Nor need the belief in the sea be confined to inhabitants of islands. It could well be said that the Vikings of the ninth and tenth centuries believed in it too.

Again, one may believe in an event, as opposed to a person or thing. In a war, or at least in its early stages, many people believe in the victory of their country. In some religious beliefs, it is an event which is believed in, and it may be either a past event or a future one. Examples are the Christian belief in the Incarnation, and the Second Coming of Christ.

In all these cases, one is believing in an entity of some sort, whether personal or non-personal, whether a substance or an event. But one may also believe in an institution. An entry in Who's Who many years ago concluded with the words ‘believes in the British Empire’. At the time when they were written, those words were perfectly intelligible, though if someone were to write or utter them now we should be puzzled, because there is no longer a British Empire to be believed in, or disbelieved in either. Again, most people believe in their own university or college or school, and have less belief, or none, in other universities, colleges or schools.

But, further, one can believe not only in an individual entity, but in a class of entities. The falconer may believe in goshawks, in goshawks in general, as a species, and not merely in this particular goshawk of his own. (Indeed, he need not own one himself.) Many people nowadays believe almost to excess in penicillin—not just in this dose of penicillin or that, but in penicillin as such. There are also many who believe almost to excess in computers. In such cases, what is believed in is something very different from an individual person. It belongs to a different logical type: though it is true that the class believed in may happen to be a class of persons (‘I believe in men who have worked their way up to the top, not in those who were born with a silver spoon in their mouths’).

One may also believe in a class of institutions. Some people believe in private preparatory schools, though others disbelieve in them. In a letter to the London Times some years ago, the writer said ‘I believe in railways’.4 More recently a spokesman for a well-known motor company was reported to have said ‘we do not believe in waiting-lists’. The emphasis was on ‘we’. Here we have the converse point. Most motor manufacturers do still believe in waiting-lists. This company's lack of belief in them, or, more probably, disbelief in them, was rather unusual. (Perhaps a waiting-list is not exactly an institution. But it is something like one. It could at any rate be described as ‘a social device’.) Again, there are still some people who do not believe in banks, and prefer to keep their money in a stocking or in a hole under the floorboards. They do not just disbelieve in this particular bank or that. They disbelieve in banks as such. But most people believe in ‘the banking system’ pretty firmly.

But what is believed in may be even more ‘abstract’ than this. One may believe in a procedure or method or policy. Indeed, this is a very common type of belief-in. At one time many Englishmen used to believe in taking a cold bath every morning, and probably some still do. Some people believe in classical education. Many nowadays do not. Instead, they believe in an education which fits one for life in the modem world. Some people believe in abstaining from alcohol when they have to drive a car or pilot an aircraft immediately afterwards. Most of us only go to our dentist when we have toothache. But there are some who believe in going to him regularly once a year, whether they have toothache or not. An interesting example of lack of belief in a method or procedure (or perhaps of disbelief in it) could be noticed in a recent statement by the President of a well known educational body: ‘Of course, we have never believed in measuring the effectiveness of what we do in terms of numbers alone.’ There are many others who would say, more generally, that they have no belief in statistics. But there are some who seem almost to believe in nothing else.

Again, one may believe (or disbelieve) in equal pay for both sexes, in easier divorce, in the abolition of the House of Lords, in aid ‘without strings’ to under-developed countries, in settling all disputes by non-violent methods, and in all sorts of ‘causes’, good, bad or indifferent.

Indeed, it is not easy to set any limits at all to the types of ‘objects’ which may be believed in. (It will be obvious that many of the examples already given take us a very long way from belief in a person, either human or divine.) But since one must stop somewhere, I shall end my list with one more example, belief in a theory.

This is an instructive example, because at first sight belief in a theory might seem so obviously reducible to a set of beliefs that. What is a theory but a logically connected set of propositions? So when someone is said to believe in a theory, surely his attitude is just a rather complicated form of believing ‘that’? He would believe that p, that q, that r, that p entails q, that r is highly probable in relation to q, etc. Now of course such beliefs-that are an essential part of belief-in a theory. But are they the whole of it? If this were a complete account of the believer's attitude, it would be more appropriate to say ‘he accepts the theory’ or ‘he believes that it is correct’ and not ‘he believes in it’. Belief in a theory has some resemblance to belief in penicillin, or belief in an instrument such as the electron microscope. The theory, when you have understood it, gives you power: a power of satisfying intellectual curiosity, of finding things out which were previously unknown, of making verifiable predictions which could not otherwise be made, and of reducing an apparently disconnected mass of brute facts to some sort of intelligible order. When someone believes in a theory, it is this power-conferring aspect of it which he has in mind, and he esteems or values the theory accordingly. It is a fact about human nature that power of this kind is very highly esteemed by some people.

Moreover, a person may still believe in a theory though he is aware that it contains paradoxes which have not yet been resolved. In that case he cannot believe that it is entirely correct. But he may still esteem it highly, and believe in it as an intellectually powerful instrument. He may use it constantly in his own investigations and encourage others to do the same. He relies on the theory, we might even say he trusts it. But in the belief-that sense he does not altogether believe it. If I am not mistaken, this was the attitude which many scientists had to the Quantum Theory in the early days of its development.

Something rather similar applies to metaphysical theories too, or at least to metaphysical theories of the synoptic type, which attempt to provide us with a unified ‘view of the world’ or ‘world-outlook’. Such theories, one may suggest, are believed in rather than just believed in the belief-in sense, and disbelieved in rather than just disbelieved. Indeed, it is doubtful whether words like ‘true’ or ‘false’, ‘correct’ or ‘incorrect’ are the appropriate ones to apply to them. ‘Adequate’, ‘not wholly adequate’, ‘relatively satisfactory’ are expressions which fit the case better. The adherents of a particular metaphysical world view, Schopenhauer's for instance, believe in it somewhat as a plumber believes in his bagful of tools or a housewife in her cookery-book. And like the plumber and the housewife, they may ‘believe in it’ with some reservations. What such a synoptic metaphysician offers us is a systematically ordered set of conceptual instruments, which will enable us (so he claims) to make sense of human experience, to unify apparently disconnected facts and reconcile apparently conflicting ones. To put it negatively, he claims to deliver us from the predicament of having to experience the world as ‘just one thing after another’: a predicament which some human beings dislike intensely, though others do not mind it.

Reducible ‘Beliefs In’

We have now considered a number of examples of belief-in which suggest that it is quite a different attitude from belief-that. We have also seen how very various they are, although belief in a person (human or divine) may well be the most important type of belief-in. Still, despite these differences, they do all support the irreducibility thesis; or at least they seem prima facie to support it.

But it is not very difficult to find examples which point the other way. An obvious one is belief in fairies. Believing in fairies amounts to no more than believing that fairies exist. Again, if someone believes in the Loch Ness monster, he just believes that there is a very large aquatic creature which inhabits Loch Ness. We often hear people say they ‘do not believe in the supernatural’. What they do not believe is that supernatural events occur or that supernatural beings exist.

The same applies sometimes even when one expresses belief in a person. If someone says he believes in King Arthur, he just expresses his belief that there was such a person, or at least a person who had some of the characteristics attributed to Arthur in the earlier versions of the Arthurian legend. For instance, he may believe that in the late fifth and early sixth centuries there was a Romano-British dux bellorum called something like Artorius, who commanded a troop of heavily armed cavalry and defeated the Saxons at Mons Badonicus about the year ad 500. This belief-in is very different indeed from the belief in Artorius which one of his own heavily armed cavalrymen (‘knights’) may have had. There is nothing in it of esteem or trust or loyalty. It is just a case of believing an existential proposition, believing that there was a person to whom a certain complex description applied. It is much the same when a classical scholar believes ‘in’ Homer. He believes that there was one poet and only one who wrote at least the greater part of the Iliad and the Odyssey, and that ‘Homer’ was his name.

Similarly, if someone disbelieves in Arthur or in Homer or in fairies, he just disbelieves an existential proposition. We may contrast this with the disbelief which British Tories had in Mr Gladstone. It was an attitude of disesteem or distrust. We notice, however, that they could not have this disbelief in him unless they believed that there was such a person. To disbelieve in him in one sense they had to believe in him in another. Nor could they disbelieve in his foreign policy unless they believed that he had one.

There does seem to be an attitude which might be called minimal or merely factual belief-in. One might be tempted to call it existential belief-in, since what is believed here is an existential proposition. That indeed is what it is, in the logician's sense of the word ‘existential’, so far as these examples are concerned.5 But Existentialist philosophers have introduced a new and entirely different sense of the word ‘existential’; and in their sense the word would apply to the kind of belief-in illustrated by our previous set of examples, where belief-in seems, on the face of it, to be irreducible to belief-that. It is perhaps one of their merits that they have paid more attention than most other philosophers to beliefs-in of this apparently ‘irreducible’ sort.

Be that as it may, there certainly is a minimal or merely factual sense of ‘believe in’. This is a very common and familiar use of the expression ‘believe in’; and ‘believing in’ in this sense certainly is reducible to ‘belief that’. It is even possible that when a person says ‘I believe in God’ he is expressing no more than a minimal or factual belief ‘in’. He may just believe that there is a God, or that God exists. When a religious person says it, he is almost certainly expressing something more; and this perhaps is the point of Pascal's distinction between Dieu d'Abraham, Dieu d'Isaac, Dieu de Jacob and Le Dieu des philosophes et des savants. It is perfectly possible to believe that God exists without being a religious person at all; and certainly the ordinary use of language allows us to speak of this ‘belief that’ as a ‘belief in’.

Similarly, in 1492 when Columbus set sail, there may well have been geographers who could say, sincerely, that they too believed ‘in’ a westerly sea route from Europe to the Indies. Nevertheless, their belief-in differed very considerably from his. They just accepted an existential proposition, the proposition ‘that there is’ such a sea route. Columbus accepted it too. But he was prepared to risk his life on it. He ‘put his trust in’ this westerly sea route which he believed to exist, and they did not.

Other Examples of ‘Reducible’ Belief-In

There are also examples of a rather different kind where belief-in does seem to be reducible to belief-that. The proposition which the ‘reduction’ yields need not necessarily be an existential proposition. If I believe in the combustibility of nylon and the incombustibility of asbestos, I believe that nylon is combustible and asbestos is not. If I believe in the infrequency of lunar rainbows, I believe that lunar rainbows are infrequent. Again, if a philosopher says he believes in free will, the obvious rendering of this is ‘he believes that all men (or all rational beings) have the power of free choice’. To say instead ‘he believes that free will exists’ or ‘that there is such a thing as free will’ is less explicit, and does not bring out the full force of the belief-in expression which we are trying to analyse.

Sometimes, no doubt, it does not matter very much whether the reduction takes an existential form or not. (One can say ‘he believes that there is such a thing as the longevity of tortoises’.) But there are other cases where it does matter. For instance, someone says ‘I have never quite believed in her blond hair’. If this is to be interpreted in a ‘belief-that’ sense, what exactly was the belief-that which the speaker could never quite hold? It certainly was not the belief that the blond hair existed. He never doubted its existence. But he did doubt whether it was the lady's own. He surmised that she wore a skilfully-made wig; or he surmised that though it was her own hair, its original colour had been very different. And now one or other of these surmises has turned out to be correct.

Finally, it is worth while to notice that the converse rendering of belief-that sentences into belief-in sentences is also possible, at least sometimes; and this gives some support to the doctrine mentioned earlier that the difference between the two is ‘merely one of idiom’. Believing that all whales are mammals could equally be described as believing in the mammality of all whales. Believing that no Englishman plays ice hockey as well as some Canadians does not seem to differ from believing in the inferiority of all Englishman to some Canadians as ice-hockey players. Believing that the Sahara would be habitable if it were irrigated does not seem to differ from believing in [the habitability of the Sahara on condition of its being irrigated]. But here we have to insert brackets to avoid an ambiguity. ‘If the Sahara were irrigated, then I should believe that it is habitable’ and ‘I believe that if the Sahara were irrigated, it would be habitable’ are two different statements. We have to put in the brackets to show that our belief-in statement is the equivalent of the second, not of the first.

Sometimes the change-over from ‘that’ to ‘in’ is not easy to make with our existing terminological resources. The ‘belief-in’ rendering of a ‘belief-that’ sentence may be clumsy, long-winded, and inelegant: for example, ‘I believe in his either coming this morning after breakfast or putting off his visit until lunch-time on the second Sunday of next month’. But certainly there are a good many cases where the difference between ‘in’ and ‘that’ can quite fairly be called a mere difference of idiom, and there are more of them than we might have supposed.

The relevance of this to our main question ‘Is belief-in reducible to belief-that?’ can now be seen. Equivalence is a symmetrical relation. If A is equivalent to B, it follows logically that B is equivalent to A. So if we are willing to use the syntactical expedients which have been illustrated, the number of ‘belief-in’ sentences which have ‘belief-that’ equivalents turns out to be much larger than we thought. Nor does it matter if some of these ‘belief in’ sentences have an exceedingly artificial air, so that no one would in practice be likely to utter them. They are intelligible, however complicated, long-winded and inelegant they may be.

Two Different Senses of ‘Belief-In’

It is only too obvious by now that the question ‘Is belief-in reducible to belief-that?’ is a complicated and difficult one. There is much to be said on both sides. We began by considering a number of examples which suggest rather strongly that no such reduction is possible. Instead, they suggest that belief-in is an attitude quite different from belief-that. We have now considered a number of other examples, which suggest equally strongly that belief-in is reducible to belief-that. What conclusion are we to draw when we consider both sets of examples together?

The obvious conclusion is this: there are two different senses of ‘believe in’. On the one hand, there is an evaluative sense. This is illustrated by believing in one's doctor, or believing in railways, or believing in a procedure such as taking a cold bath every morning. Something like esteeming or trusting is an essential part of belief-in in this sense. (The other part of it would be conceiving or having in mind whatever it is that is esteemed or trusted.) As we have seen, the ‘objects’ of belief-in, in this sense, are enormously various. It is a mistake to suppose that its ‘object’ must always be a person. There is a corresponding sense of ‘disbelief in’, where our attitude is something like disesteem or distrust. This is quite commonly expressed by saying ‘I do not believe in…’, much as dislike is quite commonly expressed by saying ‘I do not like…’. It is illustrated by ‘we do not believe in waiting-lists’, or by the disbelief in Mr Gladstone which most contemporary British Tories had. In this sense of ‘believe in’, believing-in does seem to be a quite different attitude from belief-that and irreducible to it. The same applies to the corresponding sense of ‘disbelieve in’.

On the other hand, there is also a factual sense of ‘believe in’. The most obvious examples of it are the belief in fairies or the belief in King Arthur. Belief-in, in this sense, certainly is reducible to belief-that. In these examples one believes an existential proposition. One believes that there is something to which such and such a description applies. But as we have seen, there are other examples of ‘reducible’ belief-in where the proposition believed is not an existential one. There is also a corresponding and equally reducible sense of ‘disbelieve in’. If someone disbelieves in fairies, he just disbelieves that there are such creatures, or rejects the proposition that there are. And if he disbelieves (in this sense) in free will, he disbelieves or rejects the proposition that human beings have the power of making free choices.

Moreover, just because these two senses of ‘belief in’ are different, the attitude denoted by the one can be combined with the attitude denoted by the other. One may both believe that there is such and such a thing and have esteem for it or trust in it. The writer to The Times who believed in railways is an example. So is Lord Halifax's Englishman, who believed in the sea. Again, one may both believe that there is such a thing and have disesteem for it or distrust in it, like those who say ‘we do not believe in waiting-lists’. Here disbelief-in, in the evaluative sense, is combined with belief-in, in the factual sense: and there is no inconsistency in this combination. In St James' Epistle a similar combination of attitudes is attributed to the devils who ‘believe and tremble’. They believe that God exists, and we may suppose they believe it with full conviction too. At the same time they have an attitude of distrust towards him.

Connections Between the Two Senses

Let us assume that there are these two different senses of ‘believe in’, the evaluative sense and the factual sense. If there are, there is also a close connection between them, when the ‘object’ of evaluative belief-in is an entity of any kind. I cannot trust my doctor unless I at least believe that there is a person to whom the description ‘being my doctor’ applies. But the phrase ‘at least’ is important, as Professor Norman Malcolm has pointed out to me. A person who is believed in, in the evaluative sense, may be known to the believer by personal acquaintance. Malcolm's example is a wife who says of her husband ‘I believe in Tom’. It would not be false to say of her ‘she believes that there is such a person’, but it would be saying too little. Similarly, if you believe in your doctor you probably know him by personal acquaintance (in some degree at any rate) though you do of course believe that he is your doctor, and probably hold other beliefs-that about him as well, for instance that he is about forty years old, and that he lives at No, 50A Tankerville Avenue. Again, the blind man who believes in his guide-dog knows the dog by acquaintance almost as one knows a human friend.

On the other hand, personal acquaintance is certainly not a necessary condition for evaluative belief-in. On the contrary, when one comes to be personally acquainted with a person whom one believes in, one's belief in him may decrease or even vanish altogether. (Fortunately this also applies to disbelief in a person, which quite often decreases or vanishes when one meets him.) To take an example of quite a different kind: the Managing Director may believe in waiting-lists so long as he has no personal experience of being put on a waiting-list himself; but at last this experience befalls him, when he is trying to get his mowing-machine repaired, and then his belief in waiting-lists is considerably shaken and may even be replaced by a disbelief in them.

The opposite ‘conversion’, from disbelief in a method or procedure to belief in it, may occur in a similar way. Many disbelieve in air travel so long as they have not actually tried it. But when they are compelled to try it, because there is no other way of getting to their destination in time, the result of this personal experience is that their disbelief in air travel vanishes and is replaced by a firm belief in it. These examples do, however, show that Professor Malcolm's point about acquaintance is relevant not only to belief in another person, but to other cases of evaluative belief-in as well.

Is Factual Belief-In a Presupposition of Evaluative Belief-In?

Shall we say, then, that anyone who believes in Jin the evaluative sense must also believe in X in the factual sense—this at least, though he may know X by acquaintance too? This is an attractive suggestion. It offers us a neat and tidy way of formulating the relation between the two senses of ‘belief in’; factual belief-in would just be a necessary condition for evaluative belief-in, or a presupposition of it. But unfortunately this is too neat and tidy to be true, if ‘believing in X in the factual sense’ is taken to mean ‘believing that X exists’, which is the natural way to take it. So interpreted, the formula proposed would fit some of the many varieties of evaluative belief-in, but not all of them.

For instance, does it fit belief in a procedure such as taking a cold bath every morning? At first we may be inclined to think it does. For surely anyone who holds this belief-in does also believe ‘that there is’ such a procedure? Some people still take a cold bath every morning and many people did so fifty years ago. Perhaps it will also be said that anyone who believes in taking a cold bath every morning must himself take one every morning (or most mornings); for if he did not, he could not be sincere when he claims to believe in doing so. In that case he not only believes but knows that there is such a procedure, and moreover he has acquired this knowledge-that by personal experience of instances of the procedure.

But these two arguments are inconclusive, A man might believe sincerely and very firmly in taking a cold bath every morning, even though he had been a bed-ridden invalid all his life and had never been able to take a bath at all. Or he might be a Bedouin who lives in the Mesopotamian desert and never has enough water to take a bath. It is conceivable too that neither of these persons has ever heard of anyone elso who took daily cold baths. Either of them might be an original thinker, who has managed to think of this curious procedure for himself; and having thought of it, he values it highly, without ever being able to put it into practice.

The same applies to belief in equal pay for both sexes. This is a policy which a man could believe in (and some presumably did) at a time when it had not been put into practice anywhere. Nor is it true that if such a man was sincere, he must have put it into practice himself, by paying his male butler and his female housemaid equally. He need not have had any employees himself; and if he had, they may all have been of the same sex.

In what sense, then, did he believe ‘that there is such a thing as’ paying the two sexes equally? Only in a pretty tenuous sense of ‘there is’. It did not amount to much more than believing that the concept of paying the two sexes equally for the same work is not self-contradictory, or believing that the proposition ‘the two sexes are paid equally for the same work’, though false at that time, is not necessarily false.

Nevertheless, there was another and not quite so tenuous belief-that which he did have to hold, if he was sincere in believing in this policy. He had to believe that the policy was practicable, or in principle practicable, whatever obstacles might have to be overcome before it was put into practice. And if we like, we can describe this as a belief ‘in its being practicable’, and then we are using ‘belief in’ in its factual sense. Similarly any sincere believer in taking daily cold baths must believe that this procedure is in principle practicable, even if he can never practice it himself and has never heard of anyone else who did.

But what shall we say of belief in an ‘ideal’, such as the ideal of complete unselfishness? A man who believes in this ideal does have to believe that complete unselfishness is logically possible, not self-contradictory as some have alleged that it is. But does he have to believe that it is practicable? Surely a man may sincerely believe in an ideal while admitting that it is quite ‘unrealizable’?

Indeed, it might be said that unrealisability is one of the distinguishing features of an ideal (as opposed, e.g. to a policy). Still, there is a belief-that concerning practicability or realizability which a sincere believer in an ideal has to hold. Whatever his ideal is, he has to believe that approximations to it are practicable, and approximations closer, or much closer, than those which exist at present. It may not be practicable for any human being, still less for most, to be completely unselfish in all his actions, utterances, thoughts, desires and feelings. But it is practicable for almost anyone to be a good deal more unselfish than he has been hitherto. At any rate a person must believe that it is, if he sincerely believes in the ideal of complete unselfishness. And if we like, we can formulate this belief-that in ‘reducible’ or merely factual belief-in terminology. Such a believer may be said to believe in its being empirically possible for these approximations to occur.

‘Belief In’ and ‘Confidence In’

It does seem to be true that factual belief-in is a necessary condition for evaluative belief-in, or a presupposition of it. But we must be careful to add that this factual belief-in may take many different forms, corresponding to the many different varieties of evaluative belief-in. It need not always take the form of believing an existential proposition, believing that X exists or that ‘there is such a thing as X’, when X is what is evaluatively believed in.

Nevertheless, the cases where it does take this form are of considerable importance. They make it very clear that we must distinguish between two different senses of ‘believe in’, and evaluative sense and a factual sense; otherwise we are led into absurd misunderstandings.

For instance, little Belinda says she does not believe in Santa Claus any more. The Christmas presents he brings her now are not nearly so nice as they used to be, and there are not so many of them either. Shall we say ‘Naughty child! Little liar! You certainly do believe in him if you make these complaints about the way he has treated you’? Yet these comments would be justified if the factual sense of ‘believe in’ were the only one.

We may notice, however, that such misunderstandings do not arise if we use the phrase ‘confidence in’ instead of ‘belief in’. ‘Confidence in’ does not have these two difference senses, evaluative and factual; and the distinction between ‘confidence in’ and ‘confidence that’ is a pretty clear-cut one. If I say I have confidence in someone, it is pretty plain that I am expressing an evaluative attitude; and if I have lost confidence in him, because of something he has done or failed to do, what I have lost is pretty clearly an evaluative attitude. Of course, if I do lose confidence in him I must still retain my confidence that he exists. But no one is ever tempted to accuse me of inconsistency on that account. Perhaps Belinda would have been wiser to say ‘I have no confidence in Santa Claus now’. Or she might have said ‘I have no faith in him now’, or ‘I don't trust him any longer’. Then we should have had no temptation to call her a naughty child or a little liar.

Part II

Another Version of the ‘Reducibility’ Thesis

So far it has been argued that there are two senses of ‘believe in’. First there is a factual sense. Here belief-in is clearly reducible to belief-that. It is just the acceptance of a proposition; and the proposition accepted is often, though not always, an existential one. Secondly, there is an evaluative sense of ‘believe in’. Here believing-in amounts to something like esteeming or trusting; and in this second sense, believing-in seems to be quite a different attitude from believing-that.

The conclusion one is inclined to draw is that the reducibility thesis (‘belief-in is reducible to belief-that’) is correct for one of the two senses of ‘believe in’ but incorrect for the other. Unfortunately the question cannot be settled quite so easily. Perhaps it has only been shown that the reduction has to take two different forms, one form for factual belief-in and another for evaluative belief-in.

A reductionist might quite well admit that there are these two different senses of ‘believe in’. Yet he might still claim that evaluative belief-in can itself be reduced to belief-that, if we go the right way about it. All we have to do, he might say, is to introduce suitable value-concepts into the proposition believed. Once we have done this, the difference between factual and evaluative belief-in will turn out to be just a difference in the content of the proposition believed, a difference in the ‘object’ and not in the mental attitude of the believer; and believing that will turn out to be the only sort of believing, a conclusion very welcome to all sensible men.

In this revised version, the reducibility thesis is much more plausible; and whether correct or not, it draws our attention to certain important characteristics of evaluative belief-in which might easily be overlooked. But anyone who wishes to maintain it must indeed be careful to ‘go the right way about it’. He must be careful to choose the appropriate value-concepts, if the proposed reduction is to be plausible. For instance, believing in one's doctor certainly cannot be reduced to believing that he is a morally good man. The value-concept which we must apply to him is not ‘morally good’ but ‘good at…’. Nor will it suffice to believe that he is good at water-colour painting. Of course I may also believe this, and it may be true. But that kind of ‘goodness at…’ is irrelevant if I believe in him as my doctor. I must believe that he is good at curing diseases, or perhaps at curing the diseases to which I myself am particularly liable.

This brings out an important point about evaluative belief-in. When someone expresses a belief in another person, it is always appropriate to ask ‘As what is he believed in by you?’ or ‘What is there about him, in respect of which you believe in him?’ Again, if the falconer believes in goshawks, we may ask ‘What is there about goshawks, in respect of which he believes in them?’ Presumably he believes that hawks of this species are good at catching geese and other large birds in flight. It is true that the phrase ‘good at’ is only appropriate to persons and animals. It hardly makes sense to say that railways are ‘good at’ or ‘bad at’ anything. But other terms closely related to ‘good at…’ may be used, such as ‘efficient’, ‘effective’ ‘good way of…’. The believer in railways believes that railways are an efficient or the most efficient way of transporting large numbers of persons and commodities over long distances by land. The believer in taking daily cold baths believes that this is an effective way of maintaining one's bodily health.

But we need to introduce another value-concept as well. Someone might believe very firmly that railways are a highly efficient way of transporting persons and commodities. But if he were very old-fashioned and eccentric, he might think that this was a reason for not believing in them. He might reject the view, held by nearly everyone else, that mobility is something good for its own sake. He might think it would be better if persons and commodities usually stayed where they are.

Again, we might believe that so-and-so is exceedingly good at extracting information from others by means of torture. But this would be a good reason for not believing in him, or in the policy of employing him in the service of the government or the police. We disapprove of that kind of efficiency, and the greater it is, the worse it is.

We see now what the other value-concept is, which has to be introduced into the proposition believed if this type of reductive analysis is to be plausible. It is colloquially expressed by the phrase ‘good thing that…’. We do not believe it is a good thing that a man is good at extracting information by means of torture. But we do ordinarily believe it is a good thing that our doctor is good at curing diseases. Or the proposition which we believe (according to this analysis) could be formulated thus: ‘My doctor is good at curing diseases, and a good thing too!’ Similarly, the falconer believes it is a good thing that goshawks are good at catching large birds in flight. Some would not agree with him in believing that this is a good thing, even though they do agree that goshawks are good at doing it. Then they do not believe in goshawks, or only in the factual sense of believing that there is such a species of hawks.

Let us now consider Lord Halifax's Englishman who believed in the sea. If this belief-in is to be reduced to belief-that, in the way suggested, what is the proposition which the Englishman believed? It must have been a rather complex one, something like this: ‘It is a good thing that my country, Great Britain, is completely surrounded by sea, since navies are a more efficient and less expensive means of defence than armies.’ (Perhaps he also believed that navies are a more efficient and less expensive means of aggression and conquest.)

The same kind of analysis can be applied to the rather difficult example of ‘believing in oneself’. Most commonly we use this phrase in a negative form. The trouble with Tom is that he does not believe in himself: he will never make a success of his life. What is this belief-in which he lacks? Certainly it is not just a factual belief-in. He does not lack the belief that he exists. He still does not lack it, even if he is a philosopher who accepts a very radical Humian or Buddhist theory of personal identity. According to the analysis we are considering, the proposition which he does not believe would be something like this: ‘I am good at performing most of the tasks I undertake, and a good thing too!’ A man who does believe such a proposition has a general belief in himself, as we might call it. But belief in oneself can be of a more limited or departmental kind. The question ‘As what is so and so believed in by you?’ is still relevant, even when ‘so and so’ is yourself. Thus an undergraduate might believe that he is good at understanding lectures, writing essays, and passing examinations, and that this is a good thing. Then he believes in himself as a student. But he need not believe that he is good at other activities in which he engages, such as football or mending punctures in the tyres of his bicycle.

It seems then that the proposed reduction of evaluative belief-in to belief-that must introduce two value-concepts into the proposition believed: not only ‘good at…’ (‘efficient’, ‘effective’), but also ‘good thing that…’. As we have seen, it need not be at all a good thing that someone should be ‘good at his job’ nor that something is an effective means or method of producing a certain result. And unless we do believe it is a good thing, we shall not believe in him or in it.

The Prospective Character of Belief-In

One merit of this analysis is to draw our attention to the relation between evaluative belief-in and time. If we just say ‘I believe in Mr So-and-so’ or ‘in such and such a policy or procedure’ no temporal predicate is attached to the object of our belief-in. But when we substitute a that-clause for the noun or name-phrase we must use a verb, and verbs have tenses.

We then notice an interesting feature of evaluative belief-in. In all the examples so far given it has a reference to the future, though not necessarily to the future only. It has a prospective character. This is not true of factual belief-in, which can be concerned entirely with the past. Belief in King Arthur is an example. On the other hand, if I believe in my doctor, I believe not only that it is and has been a good thing that he is good at curing my diseases, but that it will continue to be a good thing and that he will continue to be good at curing them.

But does evaluative belief-in always have a prospective character? Surely there can be an evaluative belief in a past event? For instance, Christians believe in the Incarnation. This is not only a belief that it happened, that there was such an event more than nineteen centuries ago. It is an evaluative belief-in as well as a factual one. According to the analysis we are considering, they do of course believe that it was a good thing that this event happened, that the results of it have been highly beneficial to the human race, and that there was no other effective means of producing them. But this is not all they believe. They believe that these results will continue to be beneficial and that there never will be any other effective means of producing them, at least so long as the present world-order continues. This is the sense in which they ‘put their trust in’ the Incarnation; and it is clear that their trust does have a prospective character. To put it another way, there is a connection between evaluative belief-in and hope.

Similarly one may believe in a person who is no longer alive, without having to believe that he is still alive in another world. A student of the Roman Empire may believe in Tacitus and disbelieve in Suetonius, or believe in him much less. (In the factual sense, of course, he believes equally in both of them. He believes that both of them existed and that both were Roman historians.) Is there anything prospective about this belief-in? There is. Some of the writings of both these historians still exist and can still be read. It could be said of each of them that ‘being dead, he yet speaketh’. So the question ‘Do we trust him?’ or ‘How much do we trust him?’ still arises. And there is something prospective about this question. Beliefs about the future are relevant to it. For instance, I believe that archaeological evidence will continue to confirm most of what Tacitus says, but will not confirm so much of what Suetonius says. Moreover, I believe that if the lost parts of Tacitus' writings are discovered some day, they too will be confirmed by archaeological evidence. Trusting Tacitus, then, is not altogether different from trusting one's doctor; and in both cases (according to the analysis we are considering) our trust consists at least partly in a belief-that of a prospective kind. Again, Englishmen have ceased to believe in the sea, though in Lord Halifax's time they did, and so did their successors until about a generation ago. Why is this? The sea is still there and still surrounds our country. But we can no longer rely on it to continue to ‘deliver the goods’ (security, power, etc.) which it did deliver formerly. It never will deliver them again unless there is a complete breakdown in our present technological civilization. If Englishmen still have a creed of this geo-political kind, the first article in it is certainly not ‘I believe in the sea’ but ‘I believe in the air’ or perhaps ‘I believe in inter-continental ballistic missiles’. In that case, according to the analysis we are discussing, we believe not only that aircraft or ballistic missiles have been and still are efficient means of defence, but also that they will continue to be so for some years to come.

Interested and Disinterested Belief-In

This analysis suggests another question which we ought to consider. If I believe that it is a good thing that such and such a state of affairs exists, does the word ‘good’ just mean ‘good for me’, the believer? (Of course, what I believe to be a good thing for me may not in fact be a good thing for me at all. But this is not relevant. We are only concerned to elucidate what it is that I believe.) If ‘good thing’ does always have this sense, we shall have to say that evaluative belief-in is always an interested attitude, never a disinterested one.

This is a plausible suggestion, provided that we are willing to stretch the meaning of ‘for me’. ‘Good for us’ would often be a more appropriate phrase. For instance, the doctor may be the family doctor or the doctor of all the inhabitants of the village in which I live. Then, if I believe in him, I am likely to believe that it is a good thing for us that our doctor is good at curing diseases. Still, in order to believe so, I must in some way ‘identify myself’ with a group (my family or the inhabitants of my village) as the use of the first person plural indicates. It is not enough that he is in fact the family doctor or the village doctor and that I believe him to be so. I must be in some way concerned about the health of the other members of my family or of my fellow-villagers. It must matter to me whether they are well or ill.

Again, I may believe in the Queen's doctor. Then I believe it is a good thing ‘for us’ that he holds this position and is very good at his job. And now ‘us’ has expanded so far that it includes the whole population of Great Britain or even of the entire British Commonwealth. It might even expand so far that it includes the whole of humanity. The believer in penicillin may well believe that it is a good thing for all mankind, for all of us everywhere, that this drug has been invented and is such an efficient means of saving lives and curing diseases. Nevertheless, it seems permissible to say that these beliefs-in are still interested ones, even though it is a matter of ‘our’ interest and not just the interest of ‘me’, the individual believer. There are collective interests as well as individual interests.

There is another question which has a bearing on this one. Indeed, it is perhaps another way of formulating the same question, if ‘interested’ has the wide sense just suggested. We may ask whether ‘good’ (in ‘good thing that…’) always has the sense of ‘instrumentally good’, ‘good as a means’. Clearly this is the sense it has in many of the examples so far considered. It is not an intrinsically good thing that people should take daily cold baths. If anything, it is an intrinsically bad one, since it is often a painful experience to take them. But according to those who believe in this procedure, it is good as a means for maintaining one's health. Again, if someone believes in easier divorce, he certainly need not believe that it is a good thing for its own sake that married couples should be divorced more easily. But he does believe, rightly or wrongly, that if divorce were made easier, this would be an effective means of increasing human happiness or decreasing human misery.

Belief in a Friend

But now let us consider belief in a friend; or rather, let us say belief in someone as a friend, since we might also believe in him ‘as’ something else, for example as a scholar or as a bee-keeper. The analysis of evaluative belief-in which we are discussing makes use of two value-concepts ‘good thing that…’ and ‘good at…’. We find, however, that we have to take a new look at both of them when we consider belief in a friend.

If I believe in someone as a friend, I do believe that it is a good thing for me, advantageous to me, that he is my friend. I believe that he is disposed to be kind to me and to give me what help he can when I need it. So far, my belief in him is an interested one; I believe that my friendship with him is good as a means, a means to my own welfare or happiness, and that it will continue to be so. But is this all I believe? Clearly it is not. I also believe that my friendship with him, and his with me, is something good in itself, and will continue to be so. It is something which I value for its own sake. In this respect, then, my belief in him is disinterested. More than that, I value him for his own sake. According to the type of analysis we are discussing, this would amount to believing that it is just a good thing that he exists, and still would be even of I ‘got nothing out of it’. In this respect again, my belief in a friend is disinterested.

Let us now turn to the other value-concept ‘good at’. Does it make sense to say ‘So and so is a friend of mine, but I do not believe in him at all’? Apparently it must, if ‘good at…’ is an essential part of the analysis of evaluative belief-in. For surely I might be quite convinced that he is ‘no good at anything’, a thoroughly inefficient person; and yet he might be my friend. It might be suggested that he is good at being friendly, and that this, after all, is a pretty important sort of ‘goodness at…’. Very likely he is, but he need not be. I might be his only friend, the only person he gets on well with. Can it be said, then, that at any rate he is good at being friendly with me, though he is no good at being friendly with others?

But there is something odd about this use of ‘good at…’. It is true that my doctor might be good at curing the particular diseases from which I personally suffer, and this might be the reason why I believe in him. It is true too that he need not be good at curing any other diseases. But he would have to be good at curing other people who suffer from the same ones. Similarly, my friend, if he is ‘good at’ being friendly with me, must also be good at being friendly with anyone else who resembles me in the relevant respects, whatever respects they are.

It comes to this: ‘good at…’, in the ordinary sense of the phrase, has a certain generality about it. It refers to a class of some kind. The class in question might in fact have only one member. I might happen to be the only person who has the characteristics required, e.g. the only indolent and red-haired person, educated at Manchester Grammar School, who has sailed twice round Cape Horn. But if ‘good at’ is the appropriate phrase, my friend would have to be good at being friendly with anyone who has this combination of characteristics, even though in fact there is no one else who happens to have them.

Perhaps we can now see what the trouble is. There is a sense in which every person is a unique individual. He is of course a member of many different classes. All the same, there is something unclassifable about him. There is a sense in which he is ‘just himself’. And this is what matters in inter-personal relations such as friendship. A friend likes me just as being myself, and I like him just as being himself. That is why it is inappropriate to say he is ‘good at’ being friendly with me, or that I am ‘good at’ being friendly with him.

Trusting is an essential factor in all evaluative belief-in. But it follows from what has been said that the trust we have in a friend is different from the trust we have in an expert who is ‘good at’ a particular job. The trust we have in an expert has a limited or departmental character. We trust him in so far as he is an engine-driver or an instructor in water-colour painting. But we need not trust him just as a human being. For all we can tell, he may be quite untrustworthy in some of his other activities. Trust in a friend, on the contrary, is non-departmental. We do trust him as a human being. Moreover, we trust him as the individual, unique human being that he is.

It seems then that the proposed analysis in terms of ‘good thing that…’ and ‘good at…’ does not apply very well to belief in a friend. The concept of ‘good at…’ is not relevant to this very important variety of evaluative belief-in. The concept of ‘good thing that…’ is indeed relevant to it. If we believe in a friend we do believe that it is and will continue to be ‘a good thing’ that he exists and is the individual person that he is. But we believe it is a good thing not only as a means, but also for its own sake. Belief in a friend cannot be just an interested belief-in. This is a logical impossibility. If our belief in another person were wholly interested, it would be improper to describe him as our friend.

To make matters more difficult, the expert on whose skill or efficiency we rely, for instance our doctor or teacher, may become our friend as well; and then we believe in him and trust him in two ways at once. Friendship may creep in, unawares, into many of the relations between one person and another. Something like it may also creep in when there is a relation between a person and an animal, for instance between the blind man and his guide-dog, or the falconer and his goshawk. And there is a faint analogue of it when someone gives a proper name to a machine on which he relies (he may call his motor-car ‘Jane’)—After all, the capacity for love, in all its many degrees and forms, is quite an important part of human nature. Philosophical analysts have to put up with it, even though it makes their work more complicated.

The Merits and Defects of this Reduction

We have now considered two different proposals for reducing belief ‘in’ to belief ‘that’. Both are useful. We learn something from each of them when we try to apply it to the many different sorts of examples in which the expression ‘belief in’ is used.

From the first we learn that there are two senses of ‘belief in’; on the one hand, a factual sense where ‘belief in’ is reducible to ‘belief that’, and often though not always consists in believing an existential proposition; on the other hand, an evaluative sense, where ‘believing in’ is equivalent to something like esteeming or trusting.

What do we learn from the second and much more plausible proposal? The aim of it is to show that even though there is an evaluative sense of ‘belief in’, this too can be reduced to ‘belief that’ if suitable value-concepts (‘good thing…’ and ‘good at…’ or ‘efficient’ or ‘effective’) are introduced into the proposition believed.

This second type of analysis is instructive and illuminating. It brings out the prospective character of evaluative belief-in, an important one which we might not otherwise have noticed. It also helps to make clear just what the content of a particular belief-in is. When someone or something is believed in, it is always appropriate to ask ‘as being what is he (or it) believed in?’ The answer often is ‘as being good at such an activity’ or ‘as being a good (efficient) way of achieving such and such a result’. Moreover, the person believing does have to value the activity which he thinks A is ‘good at’ or the state of affairs which he thinks B is an efficient means of achieving; and this is the point of the phrase ‘good thing that…’. There are many beliefs-in whose content cannot be fully explicated unless thsee two concepts ‘good thing that…’ and ‘good at…’ (or ‘efficient’) are brought in somewhere. Finally, once they are brought in, we see that there are two types of evaluative belief-in, interested and disinterested, and also that some evaluative beliefs-in are both at once, when ‘good’ in ‘good thing that…’ includes both ‘good as a means’ and ‘good for its own sake’. This again is an important point, which we might not have noticed otherwise.

But there are defects in this analysis too, however helpful and instructive it is. We shall do well to make all the use of it we can, but it will not take us all the way. As we have seen, it does not fit one very important type of example, belief in a friend. Here the concept ‘good at…’ plays no part, although ‘good thing that…’ does. Our friend's efficiency, or lack of it, is irrelevant to our belief in him, if we do believe in him as a friend.

But further, the proposed reduction does not completely fit any of the examples to which we have tried to apply it. In all of them, it leaves something out. At an earlier stage of the discussion it was suggested that ‘esteeming or trusting’ is an essential feature of evaluative belief-in. We now see, I think, that both esteeming and trusting are essential features of it. This reductive proposal does provide fairly well for the esteeming, by means of the concepts ‘good thing that…’ and ‘good at…’ (or efficient’). But does it provide for the trusting? Can this be done by insisting on the prospective character of evaluative belief-in?

Suppose I believe not only that my doctor has been and is good at curing my diseases, but also that he will continue to be so; and not only that it is and has been a good thing that he is good at this, but also that it will continue to be a good thing. But what if I do believe these two propositions as firmly as you please? Believing them may be a necessary condition for trusting him, but it is not the same as trusting him. Trusting is not a merely cognitive attitude.

To put the same point in another way, the proposed reduction leaves out the ‘warmth’ which is a characteristic feature of evaluative belief-in. Evaluative belief-in is a ‘pro-attitude’. One is ‘for’ the person, thing, policy, etc. in whom or in which one believes. There is something more than assenting or being disposed to assent to a proposition, no matter what concepts the proposition contains. That much-neglected aspect of human nature which used to be called ‘the heart’ enters into evaluative belief-in. Trusting is an affective attitude. We might even say that it is in some degree an affectionate one.

The beliefs-that, to which this reductive analysis draws our attention, are indeed an essential part of our belief-in attitude. When we trust someone or something, these beliefs-that are the ones we must mention in order to answer the question ‘in respect of what do you trust him (or it)?’. And this question is a perfectly proper one, and does require an answer. But when it has been answered, we still have not explained what trusting is, or what it is like to trust or ‘put one's faith in’ someone or something. Perhaps we can only know what it is like by actually being in the mental attitude which the word ‘trusting’ denotes. But fortunately there are few persons, if any, who have never trusted anyone or anything; and if it is disagreeable to be compelled to talk about ‘the heart’, the fact remains that most of us have one, as well as a head.

Application to Belief in God

The most important of all the varieties of evaluative belief-in is belief in God. It is also the most difficult to discuss, if only because so many of us nowadays do not know what it is like to have it. Still, one may ask whether any light is thrown on it by the conclusions we have reached, even though most of the examples discussed have been of a non-religious kind.

Belief in God (in the evaluative sense) clearly does have the ‘warmth’ or ‘heart-felt’ character which we have noticed in other evaluative beliefs-in. It is certainly a pro-attitude, and both esteeming and trusting enter into it. But does the distinction between ‘interested’ and ‘disinterested’ belief-in apply to it, and does it have a prospective character, as non-religious belief-in has?

It looks as if evaluative belief in God were both interested and disinterested at the same time, interested in some respects and disinterested in others. If the phrase ‘good thing that…’ may be used here, then surely it is a good thing for the believer himself (and for all of us) that God is loving, compassionate and merciful, that he answers prayers, that he gives his grace to us, that he is a refuge to us in times of trouble. Nothing could be more advantageous to us than the existence of God, if he is what theists believe him to be. The prospectiveness is there too. We believe not only that all this is and has been ‘a very good thing’ for each of us individually and all of us collectively, but also that it will continue to be so. God has been ‘good to us’ and we trust him to be good to us always, come what may, and even at times when he seems not to be (‘Though he slay me, yet will I trust in him’).

But if we were to stop at that, our belief in God would be an interested belief-in. His existence would only be good as a means, however important and even indispensable that ‘means’ might be; and if we loved him, our love (so far) would only be a kind of ‘cupboard love’. From this point of view he is regarded just as ‘the giver of gifts’. We value his gifts; we are sure that we could not get on without him. But so far, we do not value him for his own sake.

But as soon as we start thanking God for his gifts, being grateful for them with a gratitude which is not just ‘a lively sense of favours to come’, our belief in him ceases to be wholly interested. We are beginning to value him for bis own sake, and to believe that it is a good thing, intrinsically good, that he exists and is what he is; and not just ‘a good thing’, but the fundamental ‘good thing’ without which there would be no others.

At this stage, the nearest analogue in inter-human relationships would be belief in a friend, where there is a similar combination of interested and disinterested believing-in. It is perhaps significant that some theistic mystics have referred to God as ‘The Friend’. But as the definite article indicates, they did not think of him as just one friend among others. Friendship of the ordinary kind is a relation between equals, and in this important respect the analogy breaks down.

After all, if once we make the distinction between interested and disinterested belief-in, it is almost obvious that belief in God is normally a combination of both. It might be perfectly sincere if it were wholly of the interested sort, but we should be inclined to think that there was something incomplete about it. Even so, it would still be quite different from the belief-that of Pascal's philosophes et savants who believe in God only in the factual sense. Even a wholly interested belief in God is still evaluative and not merely factual. It cannot be reduced to the mere acceptance of an existential proposition.

  • 1.

    This lecture, a rather radically revised version of one of the orally-delivered Gifford Lectures, was published in Religious Studies (Vol. I, 1965–6) and is reproduced here by the Editor's kind permission.

  • 2.

    From now on, I shall sometimes write ‘belief-in’ and ‘belief-that’ with hyphens.

  • 3.

    We need not here consider in what sense God may be described as ‘personal’. It is sufficient for our purpose that in theistic religion personal pronouns are

  • 4.

    Sir Egbert Cadbury, The Times, September 9, 1959.

  • 5.

    We shall see presently that there are other example where the propositions believed belong to other logical types (p. 433–4).

From the book: