You are here

Series II

Lecture 6: Self-Verifying Beliefs

In this lecture I shall discuss a paradoxical class of beliefs to which William James drew attention in his essay The Will to Believe. He claims that there are situations ‘where faith in a fact can help create the fact’.1 But one of them had been noticed some nineteen centuries earlier by Vergil. He says of one of the crews competing in a boat-race possunt quia posse videntur, ‘they can because they think they can’.2 If a lecture required a text, like a sermon, these words of the illustrious poet would serve the purpose very well. Obviously there is something paradoxical about them. Can it ever be true that ‘thinking makes it so’? We ordinarily suppose that facts are independent of our beliefs. Facts are what they are, we say, no matter what anyone believes. Someone's belief that it is raining has no tendency whatever to make it true that it is raining (nor, of course, to make it false either).

But perhaps this is not quite the right way to formulate our ordinary view of the matter. For example, it is a fact that at 5.30 p.m. I open my front door and walk down the road to the post-box, and the state of affairs comes about because of a belief which I have, my belief that the post-box will be cleared at 5.45 p.m. In this kind of way, innumerable facts clearly are dependent, or partly dependent, upon people's beliefs, since our beliefs affect our actions and our actions cause changes in the world. Our ordinary assumption is not so much that facts in general are independent of beliefs in general, but rather that the fact which verifies a belief is independent of that particular belief. That the post-box is cleared at 5.45 p.m. is something completely independent of my belief that it is, though not independent of other beliefs, for instance those held by post office administrators about the time it takes to transport letters from here to the sorting office and from there to the railway station.

Again, I believe that there are several motor cars now passing down the High Street in Oxford in a westerly direction. Let us suppose that there are in fact fifteen motor cars doing so, which makes my belief correct. This fact is no doubt dependent on other people's beliefs, for example the belief (correct or not) held by one of the drivers that this is the quickest way to the railway station, and the belief (correct or not) held by another that he will enjoy the musical comedy which he believes is going to be performed at the theatre this evening. But the fact that the motor cars are moving in this way is completely independent of my belief.

In short, we ordinarily suppose that the state of affairs which makes a particular belief correct or incorrect (verifies it or falsifies it) is entirely independent of that belief, though it may sometimes be dependent in some degree upon other beliefs-namely, when the state of affairs in question is one which human actions can influence. ‘Nothing is made true’, we say, ‘simply by the belief that it is so. Believing a proposition p has not even a tendency to make p true, nor to make it false either.’ This is what we ordinarily assume; and this assumption does seem at first sight to be an essential ingredient in our notion of ‘objectivity’ or ‘objective reality’. The objective is what is so, no matter what beliefs anyone holds about it. This is part, at least, of what we mean when we speak of ‘hard’ facts. I once heard a philosopher say ‘the objective is the objectionable’.3 It is objectionable because it so often falsifies our beliefs, and also in another way, because it so often prevents our wishes from being fulfilled. It is a fact that it is a cold day, whether you like it or not. It is a fact that I am still sitting in the train at 11.30 p.m., though I should much prefer to be in bed.

But these assumptions about ‘objectivity’ seem to be contradicted by examples like the one quoted from Vergil ‘They can do it because they think they can’. Here the very fact which makes the belief correct seems to be dependent, or partly dependent, on the belief itself. Or, as James puts it, ‘faith in a fact’ has apparently ‘helped to create that fact’.

Other Examples

Let us now consider some other examples. The paradoxical phenomenon which we are discussing is not at all uncommon.

1. Some years ago a paragraph appeared on the sports page of a Sunday newspaper. Its title was ‘Failure in the foursomes. Beaten because they expected it’. This resembles Vergil's case. Believing that you will fail can cause you to fail, just as believing that you will succeed can cause you to succeed.

2. If you believe that you are going to get hurt (for example in a game of mixed hockey) the chances are that you will get hurt. So just don't think about it at all.

3. If the patient believes that he will be well again a fortnight from now, this makes it more likely that he will be well again by then.

4. Suppose you have to walk across a narrow plank over a small river. If you believe that you are going to fall off into the water, you probably will fall off. On the other hand, if you believe you are going to get across safely, the chances are that you will in fact get across safely.

5. If a large number of business men believe that there is going to be a slump or a devaluation of the currency, this makes it likely that there will in fact be a slump or a devaluation of the currency.

The same phenomenon can be noticed when the proposition believed is negative; or, if one likes to put it so, disbelief can be as effective as belief. When the competitors in the golf match expected to be beaten, we could equally say that they expected not to win or draw; and that the business men just mentioned believe that economic prosperity will not continue, or that the currency cannot be maintained at its present value for much longer.

Or let us consider another patient who believes that he will not get well for a very long time. This belief does seem to retard his recovery. He remains ill for more than a month, whereas most people with his physical symptoms are completely restored to health in three weeks.

Again, the man confronted by the plank across the river may be in a state which is the exact converse of the one described by Vergil. He is quite sure that he cannot possibly walk across it, even though he has just seen his companions do it, apparently without any difficulty. He cannot do it, because he believes he cannot.

This kind of belief-induced inability may take a more general form. Robert seldom succeeds in any of his undertakings because he has no belief in his own capacities. Whenever any task confronts him, or at least any important task, he believes that he cannot manage it, and because he believes this, he cannot in fact manage it.

Beliefs of this paradoxical kind need a special name. Let us call them self-verifying beliefs. Let us say that a belief is self-verifying if the belief that p either makes p true or at any rate increases the probability of p. (Perhaps you do not actually fail in your undertaking, but the belief that you are going to fail makes it appreciably more likely that you will.)

It is important to include this second case, when the probability of a proposition is increased in some degree by someone's believing the proposition, because even this appears at first sight to contradict our ordinary assumptions about the relation of beliefs to facts, and to be inconsistent with our ordinary notion of ‘objectivity’. If believing a proposition p has even a tendency to make p come true, the truth or falsity of p cannot be wholly independent of the belief that p.

In the example about crossing the plank, perhaps the man who believed that he could do it fell off into the water all the same. But he did get more than half way across, whereas the men who believed they could not do it just remained standing on the bank.

Perhaps the patient does not actually recover from his illness. But the paradox we are discussing still arises, if his belief that he was going to recover made his recovery more likely, or less unlikely, than it would have been otherwise. And we may have evidence that it did have this effect. The patient eventually died of his disease; but he lived longer than a patient with this type of illness would normally be expected to live.


Before going further, there is a difficulty, or complication, which we must consider. Are these not cases where a belief, so far from being self-verifying, is actually self-falsifying? Surely this is the characteristic feature of the state of mind called over-confidence? For instance, it might be said, an undergraduate who believes very firmly that he will get a First Class in his final examinations is likely, for that very reason, not to get one. It might even happen that because of this belief of his he fails to get a degree at all.

Again, I once heard a doctor remark ‘One can make out a good case for saying that belief that you will recover makes it more likely that you will not’. Why is this? Presumably because a very firm belief that you will recover may lead you to neglect reasonable precautions. An over-confident influenza patient may get out of bed and do a full day's work as soon as his temperature goes down to normal, and the result is that his temperature goes up again.

In the same kind of way, the over-confident undergraduate is so sure he will get a First Class that he neglects to study several of the books prescribed, and does not bother to go to any lectures in the last three weeks of term.

What shall we say about such examples? It might be argued that self-falsifying beliefs are just as interesting, from an epistemological point of view, as self-verifying ones. On the face of it, the existence of self-falsifying beliefs would be equally incompatible with our ordinary assumption that facts are independent of beliefs. What is puzzling is that believing a proposition p should have any effect at all upon the facts, one way or the other; either in the way of making p true or in the way of making p false. Our ordinary assumption is that believing a proposition has not even a tendency to make the proposition false (any more than a tendency to make it true) and neither reduces nor increases the probability of the proposition.

But perhaps we have not yet grasped the point of this objection about over-confidence. Perhaps it is this—if someone wishes to maintain that there are self-verifying beliefs, will he not have to admit that the same belief is self-verifying in some cases and self-falsifying in others? And this in turn may suggest that the whole conception of a belief which ‘alters the facts’ is a mistaken or confused one. If the same belief is followed in some cases by a situation which verifies it, and in others by a situation which falsifies it, surely the actual result, in either case, must be independent of the belief itself? Whether the patient recovers or not depends on his bodily state, and it makes no difference at all what his beliefs about it may be. Again, it might be said, your belief that you were going to get across the plank safely did not make it any more likely that you would get across safely (though in fact you did). For here is another man who holds the same belief about himself as you hold about yourself, and holds it equally firmly; and yet he does not in fact get across safely. He loses his balance when he is half way across and falls into the water.

Now if anyone were to maintain that believing a proposition is ever a sufficient condition for that proposition's being true, this objection would show, I think, that he was mistaken. But the suggestion we are discussing is not nearly so extravagant as that. The suggestion was only that in certain special cases believing a proposition p is a cause-factor or a part-cause which contributes to bringing about a state of affairs which makes the proposition true. At the most, it would be a necessary condition or conditio sine qua non; something whose absence would prevent p from being true. (For instance, there may be some things a person cannot do, unless he believes he can.)

But many other conditions must obviously be fulfilled as well. No matter what my beliefs are, I shall not succeed in getting across the river safely if the plank is struck by lightning when I am half way across, or if my companions who have crossed already (large, heavy men) have strained the timber to breaking-point. More important still, some of these other conditions which must be fulfilled are psychological ones. I may believe firmly that I am going to get across safely, but I shall not succeed unless I pay careful attention to what I am doing. Suppose a peregrine falcon flies past while I am on my way (I happen to be interested in rare birds) and I turn my head to have a better look at it. Then I shall almost certainly lose my balance and fall into the water.

This is just the kind of thing which the over-confident person is liable to do. He not only believes that he will be successful, but also believes, quite mistakenly, that this belief—the belief ‘I am going to succeed’—is sufficient by itself to ensure that he will in fact succeed.

It is not quite true, therefore, that the same belief is self-verifying in some cases and self-falsifying in others. The over-confident man does, of course, believe that he will succeed; and the sensible man, who is neither over-confident nor under-confident, likewise believes that he will succeed. So far, we can say rather loosely that both of them hold ‘the same’ belief, in the sense that A believes that A will succeed and B believes that B will. But the over-confident man also holds another belief which the sensible man rejects. It is what philosophers call a ‘second-order’ belief: the belief, or perhaps the unconscious taking for granted, that a belief in one's own success is sufficient by itself to ensure that one will in fact succeed; which it certainly is not, though it may be a necessary condition for success (a sine qua non) in certain cases.

I would suggest that the disposition to hold such a second-order belief is what we mean by ‘over-confidence’. At any rate, it is clear that the total belief-states of the over-confident man and the sensible man are not the same. Both believe that they are going to succeed in crossing the river, but the one man believes something which the other disbelieves—namely that believing one is going to succeed is sufficient by itself to ensure success.

We have now considered a number of examples which may help to illustrate the meaning of the term ‘self-verifying belief’. But so far we have done nothing to solve the problem which was mentioned at the beginning. The mere suggestion that there might be self-verifying beliefs, still more the assertion that there are, does seem to conflict with our ordinary assumption that the real world is independent of our beliefs about it.

Let us try to imagine a world in which every proposition you believe is made true merely by your believing it, a world in which thinking always ‘makes it so’. Would it be a world at all? Surely it would not, in any ordinary sense of the term ‘world’. Or, if one prefers to put it so, it would be a purely subjective world. According to some speculative persons, the next world is of that sort. But surely this present world is not.

Now of course it has not been suggested that all our beliefs are self-verifying. Far from it. As we shall see presently, the class of self-verifying beliefs is quite a narrow one, even though we include in it cases where believing a proposition has no more than a tendency to make the proposition come true. But the problem still arises if there are any self-verifying beliefs at all, even in this rather wide sense of the term ‘self-verifying’. If any beliefs at all have this character (and we can hardly avoid admitting that some do have it) it would seem, at least at first sight, that our ordinary notion of objectivity will have to be revised.

What Kinds of Belief Are Self-Verifying?

Before we attempt to solve this problem, it will be useful to consider just what this paradoxical class of beliefs consists of, and ask whether they have any distinctive feature in common. We notice at once that they are concerned with a very restricted subject matter.

In the first place, they are all beliefs about human beings, or at any rate about conscious beings. It might perhaps be unwise to say that only human beings can hold them. Perhaps non-human animals have some self-verifying beliefs, in so far as they can be said to have beliefs at all. Our cat at home is a very enterprising jumper. Perhaps he can only jump from the floor to the top of a cupboard because he believes he can (potest quia posse videtur). It is also logically possible that there are superhuman conscious beings who have beliefs, and that some of these beliefs are self-verifying. Indeed, if there are such conscious beings, we might perhaps expect that their self-verifying beliefs would be more numerous and more varied than our own.

Secondly, self-verifying beliefs always refer to the present or the future and never to the past: that is, they always refer to what is present or future at the time when the belief exists. The past cannot be altered, whatever our beliefs about it may be. On the other hand, it would be a mistake, though a tempting one, to suppose that self-verifying beliefs are concerned only with the future. When a man believes in this self-verifying way that he is able to do something, he has the ability now, though the exercise of it may be future.

Could we say, thirdly, that all self-verifying beliefs are beliefs which some conscious being holds about himself—that they are all ‘self-regarding’ beliefs? If we do say this, we shall have to include the believer's body as part of himself, in order to make room for such examples as ‘he will get well if he believes he will get well’. If his belief alters the facts, or tends to alter them, in such a way that the proposition believed comes true, the facts in question are at least partly facts about the state of the believer's body. Indeed, the same applies, very often, to ‘he can do it because he thinks he can’. The ‘doing’ is very often a psychophysical process. Walking across the plank over the river is an activity which obviously contains bodily events as well as mental ones; and in the example quoted from Vergil the thing which the competitors could do ‘because they thought they could’ was winning a boat-race.

It is true that there are some cases, and important ones too, where the ‘doing’ is something purely mental, or as near to being so as any activity of an embodied conscious being can be. There is scope for what is called ‘faith’ in intellectual or imaginatively-creative activities. When we say that someone never succeeds in any of his undertakings because he has no belief in his own powers, the undertakings we have in mind may include constructing philosophical theories or composing sonatas. Again, one of the things we may succeed in doing if we believe we shall succeed, and fail in doing if this belief is lacking, is to alter our own emotional attitudes, for example to forgive someone towards whom we have hitherto felt resentful, or to cultivate the habit of seeing the other man's point of view.

Nevertheless, it is clear that in many of the cases where we say ‘he can do it because he thinks he can’ or ‘he succeeded in doing it because he believed he would succeed’, the doing is a psycho-physical activity. So if we wish to say that self-verifying beliefs are invariably beliefs which some person holds about himself, we shall certainly have to include his body as part of himself. (We need not, of course, suppose that his body is the whole of him.)

Are All Self-Verifying Beliefs Self-Regarding?

But is it true that all self-verifying beliefs have this self-regarding character? It may seem so from the examples so far given. The question however, is a little more complicated than it looks. In the first example of all, possunt quia posse videntur could quite well be translated ‘they can do it because other people think they can’; and so translated, the statement might still be true. Again, it might be true that the patient's chances of recovery are increased if other people (his doctor, for instance, or his family) believe he will recover; and conversely that his chances of recovery are decreased if these other persons do not believe he will recover, and still more if they believe that he will not. It might also be said that the man who has to walk across the plank over the river is more likely to succeed if his companions believe he will succeed, and more likely to fail if they believe he will fail.

On the other hand, it might be argued that this effect of other people's beliefs is secondary and indirect. Thus if the patient is told by his family or his friends or his doctor that they believe he is going to get better, it is quite likely that he himself will begin to believe so, even though previously he did not; and then what is self-verifying is not their belief about him, but his own belief about himself, which is caused or partly caused by their communication of their belief to him. Nor is it enough to say that they talk to him in this way ‘just to encourage him’ and leave it at that. For what is encouragement, or being encouraged? It consists at least partly in being induced to believe that one's prospects are better than one had previously believed them to be.

Indeed, it is not even necessary that these other people should tell the patient what their belief is. Their behaviour and demeanour, the emotions they express, the topics they discuss in his presence (for instance, plans for taking him with them on holiday in Denmark next summer)—all this may convey to him what their belief is, even if they do not actually tell him what it is. In the converse case, where they believe that he is not going to recover, they are very unlikely to tell him explicitly what their belief is. Nevertheless, from their behaviour and demeanour he may be able to ‘gather’ that they do believe this. And that might be just as effective in inducing him to hold the same belief about himself.

It is also conceivable that even though these beliefs which other people hold about him are not communicated to the patient explicitly, nor conveyed to him in the indirect ways which have been mentioned, they may still affect him telepathically. At any rate, it would be unwise to exclude this possibility. But then again the beliefs of other people would only affect the situation indirectly. The self-verifying belief would still be the one which the patient holds about himself, even though it is the telepathic influence of other people's beliefs which caused him to hold it.

But this kind of explanation (including the telepathic variant of it) does not cover all the facts. It can happen that A's belief about B has a ‘self-verifying’ character even though B does not himself share that belief. For instance, John's Headmaster believes that John is capable of getting a scholarship at the University. John himself does not believe this, and still does not believe it in spite of all the Headmaster says to him. All the same, he is persuaded to sit for the scholarship examination, which he would never have thought of doing otherwise; and then, to his own intense surprise, he does get a scholarship. Here, of course, the Headmaster's belief is only verified because of the effect it has on John's own mind. But the effect it has is not at all to alter the boy's own belief about himself. When he was persuaded, reluctantly perhaps, to take the examination, he did not in the least believe that he was capable of getting a scholarship, as is shown by his intense surprise when he does get one. The Headmaster's belief verifies itself not by altering the boy's beliefs, but by altering his volitions and his actions.

This example seems to show that a self-verifying belief about a person need not necessarily be a ‘self-regarding’ belief, one which this person holds about himself, though very often it is.

Something similar might occur in several of the other examples we have considered, for instance the one about crossing the plank. Suppose that my companions all believe that I can cross safely. Some of them tell me so, others convey their belief to me by their gestures or their facial expressions. One slaps me kindly on the back, another smiles in an encouraging way from the other side and beckons to me to come over. I still do not believe myself that I can manage it. All the same, because they have conveyed to me that they believe it, I am induced to make the attempt; and then I find, to my own surprise, that I do in fact get across safely. Here again a person is able to do something, not because he believes he can (until he actually did it, he had no such belief) but because other people believe that he can; it is sufficient if these other people's beliefs, when they are conveyed to him, affect his volitions and his actions.

It is worth while to add that self-verifying beliefs do play a considerable part in personal relations between one person and another, as William James points out in his essay The Will to Believe.4 Nor need these relations be very intimate or continuous. Perhaps my companions who help me to cross the plank by expressing or conveying to me their belief that I can do it, are people whom I do not know at all well. Perhaps I had never even met them until this morning, when we set out on our cross-country walk together. Perhaps the reluctant scholarship candidate had never even spoken to the Headmaster before (it may be a very large school) and just has one interview with him, lasting only twenty minutes. Indeed, the personal relations need not be of the ‘face-to-face’ kind at all. They may be conducted entirely by correspondence.

All that is required is that one person should express or convey to another some kind of ‘faith’ which he has in this other person. Because A believes that B is an honest man, and makes it clear to B that he does hold this belief, B is more likely to be honest in fact, or become so or remain so, than he would have been otherwise. It is not necessary that A should actually tell B, in speech or writing, what he believes. It is sufficient if his belief is conveyed to B in some indirect manner. (Indeed, this is likely, perhaps, to be more effective than direct ‘telling’). For instance, they are talking about some plan of action which B is considering, and A discusses it on the assumption that B will ask for, and obtain, a day's leave of absence from his place of employment and will be able to earn the extra money he will need. By conducting the conversation on that assumption, he conveys to B his belief that B is going to behave honourably, without actually saying that he holds it.

Of course A's faith may be falsified after all. Perhaps B just absents himself without permission and steals the money from the till in the shop the night before. But still, we ordinarily suppose that A's faith makes it appreciably more likely that B will ‘act up to’ the belief which A has about him. The faith which one person has in another is an attitude which has a tendency, at least, to justify itself, if he can somehow convey it to the other person.

Finally, to complete our survey of the kinds of belief which have a self-verifying character, we may notice that they do not have to be reflective beliefs, arrived at after a careful scrutiny of the evidence for and against the proposition believed. They may equally well be unreflective takings-for-granted. As we have seen, the man who believes he is going to get across safely does have to take reasonable precautions while he is on his way across. But he need not necessarily consider the evidence for and against the proposition ‘I can do it’ or ‘I am going to get across safely’. It is quite possible that he has evidence for this proposition: for instance, he has often tried to do the same sort of thing before, and has been successful rather more often than not. But he need not consciously recall these facts about his past, or estimate what degree of probability they confer on the proposition which he now accepts and acts upon. He may well accept it without question, just take for granted that he will arrive safely on the other bank of the river: and this is enough to make it likely that he will in fact get across safely.

So far, then, our conclusions are these:—

1. Self-verifying beliefs are always about the present or the future, never about the past.

2. They are always beliefs about conscious beings.

3. They are often, but not always, self-regarding beliefs, beliefs which some person holds about himself (his body being counted as part of himself).

4. They need not be reflective beliefs. Often they are unreflective takings-for-granted.

We see now that this paradoxical self-verifying character is confined to quite a small class of beliefs, though quite an important one, since many of the beliefs contained in it have a direct relevance to human well-being. But however small the class of self-verifying beliefs may be, this does nothing to remove the paradox we are discussing. It still arises if any beliefs at all are self-verifying or have a self-verifying tendency. The idea that any proposition can be made true, or even likely, merely by being believed does seem to conflict with our ordinary assumptions about objectivity. Is there any way of solving this paradox? I shall now suggest that there is.

The ‘Act-Object’ Distinction

Let us begin by stating the proposed solution in an over-simplified form, more or less in the way an old-fashioned Realist philosopher of the 1920s would have put it, if he had happened to consider our problem.

It might be said, the word ‘belief’ is ambiguous: we must distinguish carefully between the believing and what is believed. Believing is a mental state or attitude. But what is believed is not a mental state or attitude at all; it is what logicians call a proposition, something which is either true or false.

Now it may be claimed that once we distinguish clearly between the believing and what is believed, the paradox which has been troubling us disappears. For surely there is nothing paradoxical in the suggestion that believings may ‘alter the facts’? Mental states certainly do have effects. They cause changes in the person in whom they occur or exist, changes in his mind and often (perhaps always) in his body, and thereby they may indirectly cause changes in his physical environment. And since believings are mental states, they may perfectly well cause changes both in the mind and in the body of the believer. Indeed, it is perfectly obvious that believings do have effects, and often quite important ones too: whereas the suggestion that a proposition could cause changes is not so much false as nonsensical, though of course the uttering of it, or the writing of it down on paper, might very well do so.

Just what effects a believing might have is a purely empirical question; and there is no a priori reason why some believings should not cause changes of such a kind that the proposition believed becomes true, or more probable than it would have been otherwise. If we find empirical evidence (as we do) that some believings have effects of this kind, we may be a little surprised, since it is clear that most believings do not have them. But there is no monstrous paradox here. It is just one more illustration of the familiar fact that mental states have effects.

Thus, it would be argued, the statement that facts are independent of our beliefs is true in one sense of the word ‘belief’ but false in another. It is true that facts are always independent of the proposition believed; but they need not always be independent of the believing. Suppose the proposition is ‘Timothy will succeed in walking across this plank’. It does not follow in the least that because there is this proposition, there will also be a state of affairs which makes it true, nor even that such a state of affairs is in the least degree probable. But suppose that someone, for instance, Timothy himself, believes this proposition; then his believing may quite well bring about a state of affairs which makes the proposition true, or at least more likely than it would have been otherwise. It may even be that in some cases believing a proposition p is a necessary condition for p's being true (p would not have been true, unless it had been believed).

Defects in this Formulation

There are two defects in this proposal for solving our problem. It is over-simplified in two different ways. First, it treats believings as if they were events or occurrences. It uses the term ‘mental states’ in an occurrent sense. Acquiring a belief is indeed a mental occurrence, and so is losing it. But having a belief (or ‘holding it’ as we say) is a disposition rather than an occurrence. So long as we have it, it is liable to manifest itself by occurrences of various kinds, both mental and psycho-physical. Instead of talking about ‘believings’ as if they are themselves events, we should talk about the events which are manifestations of a belief-disposition. One of them is the using of the proposition believed as a premiss in our practical deliberation, when we are considering what to do, and the giving of attention to the proposition and to the inferences which may be drawn from it. These are mental events or occurrences and they certainly have effects of one sort or another; but it is misleading to describe them as ‘believings’.

The second defect, or over-simplification, is a rather strange view about propositions. They are spoken of as if they were entities, which are somehow ‘there’ and are quite independent of our thinking and speaking. You may have noticed the curious statement ‘it does not follow that because there is such and such a proposition there will be a state of affairs which makes it true’.5 But this difficulty can be avoided by reformulating the proposed solution in a slightly more complicated way.

Let us first translate the rather alarming statement ‘There is the proposition p’ into the harmless one ‘The indicative sentence s makes sense’, which is in turn equivalent to ‘The indicative sentence s is either true or false’ (or ‘would be either true or false if someone were to assert it’). Let us suppose also that s is an empirical sentence, not an a priori one such as ‘3 + 2 = 5’, Then we proceed as follows: From the fact that the sentence s is either true or false, it does not in the least follow that s is true rather than false, nor of course that it is false rather than true. Nor does it in the least follow that the sentence is likely to be true rather than false, nor yet that it is likely to be false rather than true. What makes such an indicative sentence true or false is the state of the world, the actual state of affairs. (Human minds, and of course human bodies too, are part of the world.)

This is the cash-value of the statement that ‘facts are independent of propositions’; and if the statement is translated in the way just suggested (as a statement about indicative sentences), it does not commit us to the view that propositions are subsistent entities. It merely amounts to saying that if an empirical indicative sentence s makes sense, nothing whatever follows from this as to the existence or non-existence of an actual state of affairs which the sentence would describe. And this principle is an essential part of our ordinary conception of ‘objectivity’.

On the other hand, if some person understands and believes an indicative sentence s, this understanding of his and this belief which he holds will certainly have effects of one sort or another; and one of these effects, in suitable cases, might quite well be that a state of affairs comes into existence which makes the sentence true, or that the coming into existence of such a state of affairs becomes more probable than it would otherwise have been. Whether believing ever does have such effects is a question of empirical fact, and the answer to this empirical question seems to be ‘Yes, in certain special cases’.

If so, there is after all no paradox in the notion of a self-verifying belief, provided we distinguish carefully between the believing and what is believed.

Self-Verifying Sentences

There would however be a paradox in the notion of a self-verifying indicative sentence, a sentence which merely by making sense would thereby bring into existence the state of affairs described by it, or make that state of affairs more likely to exist.

Even magicians have not gone so far as to embrace such a paradox as this. They have only claimed that the utterance of certain sorts of sentence which make sense, or the writing of them (with suitable ceremonial precautions) tends to bring about a state of affairs such as the sentence would describe. They have not supposed that the mere fact that the sentence makes sense has any such consequences.

It is true that logicians, a very different class of persons, might claim that there are self-verifying sentences. An example would be ‘This sentence contains five words’. There certainly is a sense in which it is self-verifying. This sentence is true just because it is the sentence that it is.

But there are two points to notice here. First, it is a very peculiar sentence, in that it is about itself. But the sentences with which we have been concerned in this discussion are not about themselves, though they may be, and often are, about the person who believes them or asserts them (‘I am going to get across the plank safely’ ‘I shall be out of bed again a week from now’). Secondly, the sense in which ‘This sentence contains five words’ is self-verifying is not a causal sense. The sentence does not alter the facts in such a way as to make itself true. It does not alter anything. What makes it true (if the word ‘make’ is here appropriate) is just a fact about the sentence itself, the fact that there are indeed exactly five words in it. Although we can say it is true ‘because of’ the number of words it contains, this is not the causal sense of the word ‘because’.

But when it is claimed that some propositons are made true, or tend to be made true, because they are believed, the word ‘because’ is being used in a causal sense. The contention is that believing does sometimes cause changes of such a sort that what is believed becomes true, or more likely to be true than it would otherwise have been. Provided we use the term ‘self-verifying’ in the causal sense which it has had throughout this discussion, it can be said, I think, that the notions of a self-verifying sentence—a sentence which merely by making sense altered the facts in such a way as to make itself true—would be an excessively paradoxical one, whereas the notion of a self-verifying belief-attitude would not.

Other Propositional Attitudes

Hitherto, we have been wholly concerned with self-verifying beliefs. But we may notice that it is not only belief which has this self-verifying tendency in suitable cases. Other propositional attitudes may have it too. Merely entertaining a proposition, barely thinking of it without either belief or disbelief, sometimes has a self-verifying tendency, especially when the entertaining is attentive and prolonged. Suppose the man confronted by the narrow plank across the river can be induced by others (or can induce himself) just to fix his attention on the proposition ‘I am going to get across safely’, the chances are that he will in fact get across safely. He need not necessarily believe the proposition. Indeed, we may have pretty good evidence afterwards that he did not believe it; we may observe that he is greatly surprised to find himself standing safely on the other side. It is sufficient if he entertains the proposition attentively (‘fixes his mind on it’) and excludes from his attention the contradictory of it ‘I am not going to get across safely’.

Presumably this is what happens in what is called ‘suggestion’, including hypnotic suggestion, and we must remember that there is such a thing as self-suggestion and even auto-hypnosis. It is an empirical fact that suggestion, including self-suggestion, is sometimes an effective method of enabling people to do things which they could not otherwise have done. Nor need these ‘doings’ be merely physical ones. By means of self-suggestion, a man may be enabled to solve an intellectual problem which has hitherto baffled him, or the kind of stylistic problem which confronts nearly all writers from time to time, when we ‘know roughly what we want to say but cannot see how to put it’, (You say to yourself before going to sleep ‘To-morrow morning at 9.30 a.m. some appropriate words will come to me’; and more often than not they do.) In a similar way, suggestion, including self-suggestion, can enable one to alter one's emotional attitude to another person, for instance to change it from resentment to forgiveness: a result which one has been quite unable to bring about by a mere effort of will, or even by repeated and persevering efforts of will. It is worth while to mention another example, trivial though it is, because anyone can easily test it for himself. By merely entertaining the proposition ‘My left forefinger is tingling’ one can actually make it tingle (or at any rate some people can) provided one goes on entertaining the proposition attentively for half a minute or so.

In these examples (unlike the ones discussed earlier) we have been assuming that the proposition is entertained quite neutrally. One does not believe it or disbelieve it or doubt it or wonder about it. It is just bare or pure entertaining, just thinking of the proposition in an attentive manner and continuing to do so for some time.

Perhaps the word ‘neutrally’ may mislead. I have heard it objected that if we do entertain a proposition p in a completely neutral way, we are ipso facto entertaining its contradictory not-p at the same time.6 The reply is that there are at least two kinds of neutrality. Our attitude may be neutral as between acceptance and rejection. We neither believe p nor disbelieve it. This is the sort of neutrality I have in mind. But our attitude may be neutral in another way also, in that we give equal attention to both p and not-p. In this respect the person I am discussing is not neutral. His whole attention is given to one of these two alternatives—and that is the one which ‘comes true’ or tends to come true.

But it would seem that self-verification can still occur when an attitude other than belief is added to this ‘bare entertaining’, an attitude of questioning or wondering, for instance. I not only entertain the proposition p; I wonder whether it is true, or ask myself how probable it is. Perhaps I also consider some of the logical relations which this proposition has to other propositions. I notice the consequences which follow from it, either with certainty or with probability. This is sometimes called ‘supposing’ (If p, then what?’) and then I may find, disconcertingly, that the situation alters in such a way that the proposition p becomes true—especially if it is a proposition about myself or my own actions.

This may be illustrated by the unfortunate predicament of the beginner in bicycling, who sees an old lady crossing the street very, very slowly in front of him. He keeps on asking himself, ‘Am I going to run into her? Am I going to run into her? What will happen if I do?’ The result is that, sure enough, he does run into her. Here again, the important point is that his mind dwells upon the propositon, even though his attitude towards it is merely one of questioning and supposing, and not of belief.

Doubt and Disbelief

Not only so: the additional attitude we have, over and above the more entertaining may be an attitude of doubt, or even an attitude of disbelieving or rejection. Of course, we sometimes cease to entertain a proposition once we have rejected it. Having rejected it, we often dismiss it from our minds altogether and no longer think of it at all. After a careful look at the clouds and the weathercock, I reject the proposition ‘It will clear up in half and hour’ and decide to spend the whole afternoon in the library. And then I pay no more attention to this conceivable but (as I am persuaded) very improbable change in the weather.

But this ‘dismissal’ of a rejected proposition does not always happen. The proposition may be of intense interest to us, as the proposition ‘I am never going to get better’ is to a person who is seriously ill. He cannot dismiss this proposition from his mind. He is likely to entertain it attentively and repeatedly, even though he always entertains it disbelievingly, with an attitude of rejection. I would suggest that this makes the proposition appreciably more likely to be true. It alters the situation for the worse, or at least tends to prevent it from altering for the better. The same thing happens in wartime, if people repeatedly entertain the proposition ‘We are going to be defeated’, even though they always reject it.

In such examples as these, it almost looks as if disbelieving a proposition had a tendency to make the proposition true; and if one wishes a proposition p to be false, it is perhaps a good piece of psychological hygiene to give up disbelieving p and believe its contrary instead (not just its contradictory, not-p, but its contrary). Instead of disbelieving that he will remain an invalid for the rest of his life, the patient should believe that he is going to get well. Instead of disbelieving that your Country will lose the war, try to believe that it will win.7

It seems, then, that with propositions of the sort we have been discussing, it is the entertaining of a proposition (rather than the believing which sometimes accompanies it), which has a self-verifying tendency; and especially, prolonged and attentive entertaining. Likewise, perhaps, the most important point about faith is not so much that a man believes the propositions with regard to which he has faith, though he does of course believe them, but rather that his thoughts dwell upon these propositions, he meditates upon them, considers the consequences which follow from them, illustrates them by means of examples, real or imaginary. In short, he entertains them more frequently and more attentively than other men who do not share his faith, and questions of the form ‘p, so what?’ are often on his mind.

If there is anything in these considerations, the problem we have been discussing was incorrectly formulated at the beginning. It should have been called the problem of self-verifying thoughts or entertainings; and ‘Thinking makes it so’ might have been a more appropriate title for this lecture than ‘Self-verifying beliefs’.

‘Thinking Makes It So’

Let us now see what can be done to solve the problem which was stated at the beginning of this lecture. If there are self-verifying beliefs or self-verifying thoughts, are we compelled to alter our ordinary assumption that facts are independent of our beliefs, or more generally of our thoughts, entertainings of propositions whether believed or not? Is there something wrong with our ordinary conception of ‘objectivity’?

If we use the phrase ‘Thinking makes it so’ as a title for the class of phenomena we have been discussing, we notice at once that there are innumerable cases in which thinking does at least contribute to ‘making it so’, and yet we are not puzzled by this at all. There is a perfectly ordinary and familiar way in which a person's thoughts ‘come true’, namely whenever he succeeds in doing something which he intended to do. In any such action, the agent must have the thought of the state of affairs which he sets himself to bring about.

It is certainly not true, then, that facts are wholly independent of thoughts, and if our ordinary assumptions about objectivity imply that they are, there is something wrong with these assumptions. We have only to look at all these entities by which we are surrounded—chairs, tables, houses, tea-pots, motor-buses, clocks, gardens, streets. These objects came to be what they are, and where they are, because the human body, especially its central nervous system, its muscles and its limbs, is an apparatus whereby thoughts can ‘come true’. And that is why the world around us, or at least the surface of this planet, is full of entities which the old Idealist philosophers might have called ‘objectified thoughts’.

Again, we speak of ‘hard facts’. But if all facts were equally hard, there would not be much point in this expression. Perhaps we only call them ‘hard’ when there is nothing whatever that we can do about them. But if so, there are at any rate degrees of ‘hardness’. We can quite often do something to alter some state of affairs which displeases us or fails to satisfy us. And then some thought which we have is made to come true. We tend to forget that thinking—the entertaining of propositions—is a necessary constituent of doing, or at any rate of intentional doing. There are no doubt puzzles about doing. Indeed one of my own teachers, H. A. Prichard, was greatly puzzled by the question ‘How is it possible to do anything?’ But nobody supposes, nor should he suppose, that doing involves any epistemological puzzles about ‘objectivity’ or the relation of thoughts to facts.

Intermediate Causal Links.8

When we say that facts are independent of our beliefs (or, more generally, independent of our thoughts) perhaps we mean that the entertaining of a proposition is never a sufficient condition for the existence of a state of affairs which makes it true. It may be a necessary condition, a sine qua non, and in any successful voluntary action it is; but pure and simple thinking never ‘makes it so’. This is our ordinary assumption. So if we encounter cases where pure and simple thinking does appear to ‘make it so, we are greatly puzzled.

To put it in another way; what would puzzle us very much would be a thought which directly caused its own verification, directly brought about a state of affairs whereby the thought ‘comes true’, without any intermediate causal links at all. This is very different indeed from anything that happens in ordinary voluntary action, where a long and complicated causal chain, with many intermediate links, intervenes between the entertaining of a proposition and the state of affairs which makes the proposition true (for instance the proposition ‘I shall be home by half past twelve’). Even in a very simple action, such as voluntarily blowing one's nose, or writing one's name in the Visitors' Book, there are many intermediate links between the thought and the event which fulfils it or makes it ‘come true’.

Now in the examples we have been discussing (unlike these familiar examples of ordinary voluntary action) it seems as if a thought does directly make itself come true, without any intervening causal links. It seems that pure and simple thinking does sometimes ‘make it so’. But perhaps if we look more closely, we may be able to find at least some intermediate links between the thought and the event which verifies it or fulfils it. Or if ‘find’ is too strong a word to use, at least we may be able to form some more or less plausible hypothesis about the kind of intermediate links there might be.

Let us consider the Vergilian type of example (possunt qua posse videntur) from this point of view. Why is the thought ‘that I can do it’ so important? Partly because it makes you free to try. But perhaps it is better to begin with the negative case (non possunt quia non posse videntur). This was illustrated by the example of the man who cannot walk across the plank because he thinks he cannot, and also by the golfers who were ‘beaten because they expected it’, if we take this to mean ‘They could not win or draw because they thought they could not’. Such thoughts as ‘I cannot do it’, ‘It is quite impossible that I should succeed’, plainly have an inhibiting effect. They prevent one from even trying to do the thing in question. The converse thought ‘I can do it’ enlarges the field of choice for you. It makes you free to try. And if you do try with this thought in mind, there is at least a chance that you will succeed; whereas there is no chance of success at all if you do not even try. When the negative thought ‘I cannot do it’ verifies itself, it does so by preventing you from trying, and this is an intermediate link between the thought and the state of affairs which verifies it. And when the positive thought ‘I can do it’ verifies itself, the intermediate link is the trying which this thought makes possible.

Thoughts and Emotions

Another relevant point is that the entertaining of a proposition, especially a proposition about oneself, is liable to have emotional effects: not only when it is believed, but also when it just occupies one's attention. The thought ‘I cannot possibly get across’ is very likely to have such effects when one wishes to get across, and other people (who have crossed already) are impatiently awaiting one on the other side. In such a situation, one is likely to feel both frightened and ashamed; and fear and shame are incapacitating emotions. The same applies to the golfers who expect to be beaten in the golf match (assuming that they do not wish to be beaten).

It is a very familiar fact that emotions can have a profound effect upon a person's bodily state. Here, then, is another intermediate link, which helps us to understand how the entertaining of a proposition about oneself can bring about a state of affairs which makes the proposition true. The entertaining of the proposition arouses emotions, and these in turn have a disabling effect upon one's body, so that it really is impossible for one to do the thing which one thinks it impossible to do. And perhaps the ‘positive’ thought ‘I can do it’ or ‘We can win some holes at any rate’ makes itself come true, or tends to make itself come true, by preventing the person from having such upsetting and disabling emotions; it enables him to ‘keep cool’. This perhaps is the way in which this positive thought makes one free to try, and to make full use of whatever capacities one has.

The same intermediate link—the emotional state which the entertaining of a proposition is liable to induce—is also relevant to the medical examples: the patient who gets better, at at least gets better more quickly, because he thinks ‘I am going to get better’, and the converse case where ‘I am never going to get better’ retards his recovery, or is a contributory factor in preventing it. Whichever of these thoughts he has, whether the positive thought or the negative one, it is likely to arouse quite strong emotions in him: and it is not very surprising that these in their turn should affect his bodily state. No doubt there are complications here (unconscious or subconscious wishes may be relevant as well). But at any rate there is no reason to suppose that ‘pure and simple thinking’ is sufficient by itself to bring about a change in his state of health without any intermediate causal links at all.

Effects of Self-Suggestion

Finally, let us consider the writer who gets stuck in his work, and manages to solve his problem by making a suggestion to himself before he goes to sleep. Perhaps it will be said that the intermediate link here is a purely physiological one. When he got stuck in his work yesterday evening he was tired. At 9.30 this morning, after a good night's sleep, he is fresh again; so it is not very surprising that the problem which seemed so difficult yesterday is now solved quite easily.

No doubt this is a relevant point, but it can hardly be the complete explanation. If it were, there would be no difference between ‘sleeping on’ some problem and merely sleeping. This rather odd phrase ‘sleeping on’ implies that the problem was in your mind just before you went to sleep. (It is helpful to make a brief mental survey of the main points in it just before going to sleep). Something else was in your mind as well: the thought, one might say the ‘quietly confident’ thought, that the solution is going to occur to you at 9.30 tomorrow morning.

Moreover, anyone who has used this method must have noticed that when the solution does come, it very often comes ‘ready made’. The word ‘come’ is important. It is not that one makes a fresh start with the process of groping or searching, and succeeds this time whereas one failed before. The answer ‘just comes’, almost as if it were being told to you by someone else. All you have to do is to seize upon it and write it down at once; before it fades out of your mind again. After that, you will, no doubt, have to grope and search once more in order to work out its implications; and this will no doubt be easier because you have had a good night's sleep.

Here it is more difficult to find the intermediate link between the thought ‘the solution will come’ and the event which verifies it. The Ancients get round this difficulty by personifying the ‘missing link’. They called it ‘the Muse’ (or ‘a Muse’, for they supposed that there were nine of them). Sometimes they would begin a poem by invoking the Muse: a very proper procedure, so far as it goes, and it may well have been a psychologically effective one, but somewhat too speculative if interpreted literally.

We need not go quite as far as this. Instead, we can use the hypothesis of subconscious or unconscious mental activity. We can suppose that there are thought-processes which go on in us, even though we are not conscious of the processes themselves, but only of the results which they eventually produce. There is no a priori reason for thinking that a person must be consciously aware of every mental event which occurs in him; and there are, of course, other phenomena, not only in our intellectual lives but in our emotional and conative lives as well, which cannot easily be explained unless we are willing to suppose that mental events (important ones too) can occur in a person without his being aware of them.

We notice also—and there is some consolation in it—that yesterday's work was by no means wasted, even though it seemed to end in failure. It turns out to have been a necessary condition for this morning's success. Without these apparently unavailing efforts, there would have been no material, so to speak, for our subconscious or unconscious mental processes to work upon. The necessary material was provided by the memories, or memory-traces, which yesterday's efforts left behind them.

For our purposes, however, it is sufficient to point out that here too we can at any rate suggest an intermediate link between the thought and the event which verifies it or fulfils it. We do not have to suppose that the thought ‘the solution will come to me to-morrow morning’ is sufficient by itself to bring about the desired result, by a kind of miraculous action-at-a-distance in time.

Nevertheless, it is important to consider the fairly numerous and various types of case in which ‘thinking it will be so’ does indeed contribute to ‘making it so’; and in some of them, this thinking does seem to be a necessary condition for the coming about of the event which ‘makes it so’, even though it is never a sufficient one.

Are there still some cases left over, where ‘thinking makes it so’ without any intermediate causal links at all? Perhaps there are, but for that very reason we describe them as ‘paranormal’ or ‘supernormal’. What puzzles us about both telepathy and telekinesis is precisely the fact that we have not so far succeeded in finding any intermediate causal links in either case. It does not follow that we never shall. But hitherto we have not found any intermediate links between the agent's experience and the percipient's experience in telepathy. In telekinesis too, where a person's thoughts or wishes appear to cause movements in objects outside his body without any kind of physical contact, direct or indirect, we have not yet succeeded in finding any series of intermediate causal links between the thoughts or wishes on the one hand, and the physical movements on the other.

  • 1.

    William James: Selected Papers on Philosophy (Everymans Library), p. 119.

  • 2.
    Aeneid V, line 231.
  • 3.

    This remark, I think, was made by Professor J. A. Smith of Oxford in a lecture which I attended as an undergraduate.

  • 4.
    Selected Papers on Philosophy (Everyman's Library), p. 118.
  • 5.

    P. 363, above.

  • 6.

    I am much indebted to an American colleague for calling my attention to this difficulty, and greatly regret that I cannot recall his name.

  • 7.

    I have discussed the puzzling concept of ‘trying to believe’ in an essay called Belief and Will (Aristotelian Soc. Supp. Vol. 1954, pp. 1–26, reprinted in Philosophy of Mind., edited by Prof. S. N. Hampshire, pp. 91–116).

  • 8.

    In this section, I am much indebted to some critical comments and suggestions made by Miss F. Anstey of the University of Reading.

From the book: