§ 1. Religious thinking

If religious life is bound up with belief in God's existence and attributes, it must contain an element of thinking. Religious men may reasonably claim that their thinking is neither inferential nor discursive—that it is more like intuition or feeling than like reasoning; and they may be right in asserting that if it is explicitly formulated as a philosophical argument, it is likely to be emptied or distorted. We are now, it is to be hoped, in a better position to appreciate these real difficulties and dangers; but the conceiving or thinking or judging that must be present—under whatever name—in man's living relation to God, is manifestly in need of closer study.

There are many ways in which this almost unmanageable task might be attempted, but a brief discussion may perhaps be made a little easier if we consider some of the traditional arguments for the existence of God.

It may be thought that to do this is to abandon the basis of religious experience and to take refuge in the arid region of scholastic discussions as remote from ordinary thinking as they are alien to modern philosophical taste. Such an objection cannot be accepted without further enquiry. The traditional proofs may be unconvincing as abstract arguments and may be inadequate to the intellectual factor in religious experience, but they have grown out of such experience: they are ways in which men of great acuteness have attempted to justify or defend their religious beliefs, and it is in relation to religious experience that they should be examined.

We must proceed at first along a well-beaten track—it has been trodden by many pilgrims before us and may be called the Hill of Difficulty. The arguments in question have been elaborated with the utmost subtlety and complexity, perhaps in order to meet the objections of critics. The subtle and the complex must here be passed over if there is to be any hope of making natural theology intelligible to those who are not professional philosophers. There may even be advantages in this avoidance of the subtle; for although an invalid argument may be made valid by further premises, the very complexity of an argument may conceal its fallacies. To simplify these proofs, so far as one can, may mean a failure to do them justice; but at least it may open up a possible line of enquiry, and any errors can be checked and corrected by more elaborate expositions, of which there is no lack.

§ 2. Arguments for the existence of God

The theoretical arguments traditionally supposed to prove the existence of God may be grouped under three main heads.

First of all, we may argue—according to some philosophers—from certain thoughts or conceptions of our own; for example, we may argue that a perfect being must exist since existence is contained in the very concept of a perfect being. Secondly, we may argue from our general experience of the world; for example, we may maintain that since anything in the world must have a cause, there must therefore exist a first cause, to which we give the name of ‘God’. Thirdly, we may argue from particular experiences of special features in the world; we may hold, for example, that the organs of living creatures are so marvellously adapted for purposes of life that they must be the product of a supremely intelligent designer or creator.

Put thus nakedly, these possible lines of argument must seem unconvincing, however much they may have carried conviction to men in the past—or at least may have reinforced convictions already held on other grounds. In actual thinking they are not so sharply distinguished from one another as the classification might suggest. Rather they run into one another and may afford each other mutual support or corroboration. On the other hand, so far as they provide separate links in a chain of argument, the whole chain will have the strength only of its weakest link. Nor can it be denied that they rest on assumptions which are not readily made to-day.

Even at first sight it may seem that all these types of purely intellectual argument have a common defect: they seek to pass from non-religious premises to a religious conclusion. Admittedly, even from the religious point of view, they are concerned with intellectual abstractions; but we must be careful not to take the word ‘religious’ too narrowly. It seems to be a conviction of religious consciousness itself that God is revealed (or reveals Himself) in two ways—firstly, in the world that He has made, and, secondly, in the heart of the worshipper. We have been concerned with the second way, the way of the heart, in attempting to examine the numinous feeling and the mystic union and the living relation to the Absolute Thou. So far as we are now to be concerned with God's revelation in the world, we are bound to start from our ordinary experience of that world. And so far as we are to be concerned with the way of the head, we are bound to consider even the most abstract thinking about God. We are not entitled to set all this aside at the outset on the ground that it can in no sense be a factor in religious experience and so cannot lead to a religious, or even to a theological, conclusion.

There is one further general point. It is impossible to establish the existence of anything without some conception of its character. Hence the alleged proofs of God's existence must be in some way concerned also with His nature and attributes.

§ 3. The ontological argument

This double concern is obvious in the argument from the concept of a perfect being to His existence: it is an inference from God's nature to His reality.

This argument, which was formulated first by Anselm and elaborated by such thinkers as Descartes, Leibniz, and Hegel, is commonly known as the ontological proof. It has many variations but may be put most simply as follows.

God is by definition a perfect being and indeed—if we may speak of degrees in perfection—a supremely perfect being. But it is self-contradictory to regard a supremely perfect being as non-existent; for to lack existence must be an imperfection. Hence a perfect being must exist, and so God must exist. This conclusion is said to be as certain as the mathematical theorem that a triangle must have the sum of its interior angles equal to two right angles.

Such an argument may be in need of further premises: perhaps we ought to show that a perfect being can be conceived without contradiction and also that there can only be one perfect being. Into these complexities it is not necessary to enter. The crux of the argument is the contention that we are entitled, and indeed obliged, to pass from the thought or notion of God's perfection to knowledge of His existence—to knowledge that He must necessarily exist.

There is something very peculiar about this attempt at a proof. Arguments to the existence of anything usually start from knowledge that something exists and go on to maintain that if this exists, something else must exist—if there is smoke, there must be fire. Such arguments are based on experience, if we take experience to be sensuous experience of something existent in time and space. All theoretical proofs of the existence of God except the ontological one start from experience of an existent world and argue to the existence of its first cause or its creator. The ontological argument is unique in not being based on experience of an existent world. It is not based on any experience of any existent at all—not even on what men sometimes claim to be religious experience of an existent God. It is based on the bare concept of perfection—on pure thinking unsullied by thought or feeling. It professes to pass from the pure thought of perfection to knowledge of God's existence, or to enjoy such knowledge by means of this pure thought alone.

From the religious point of view the concept of perfection, in spite of its vagueness, is the most satisfactory basis on which to rest any attempt at an ontological proof; the religious man can worship only a God who is and is perfect. But the concept from which the argument starts can be expressed in other words. Anselm, for example, took the concept of God to be the concept of that than which nothing greater can be conceived. Others have taken it to be the concept of the most real being, the supreme being, the being of beings, and so on. These concepts have the advantage of appearing less anthropomorphic, and they are metaphysically preferable to the concept of perfection, except when its human associations have been removed. But unless they cover the notion of perfection, they will not, from the religious point of view, offer a satisfactory basis for any proof of the existence of God. From a logical point of view they may lead only to the empty tautology that whatever is most real must be most real.

In whatever form the argument is stated it will be instinctively rejected by common sense. Some may see in it a pitiable exhibition of human weakness. How is it possible that men of outstanding intellectual ability should have been taken in by so manifest a sophism? We can never be entitled to infer from the mere thought of an object that the object itself must exist. No purely verbal definition can give us knowledge of reality.

The attitude of common sense in this matter seems to be fundamentally sound. Nevertheless it is a mistake to suppose that we can take an argument apart from its context and dismiss it summarily without regard to the whole philosophical system of which it forms a part. This particular argument merits a little more consideration because of the importance attached to it by thinkers whose eminence in other fields is beyond dispute.

§ 4. The philosophical background

For Anselm the ontological argument had a background of Platonism imperfectly understood. In such a context the argument becomes more plausible. Common sense ascribes existence and being, at least primarily, to the world we perceive—to the ordinary world of stars and planets, rivers and mountains, tables and chairs, plants, animals, and human beings. Whatever else may exist or be, these things exist and are. But Plato did not agree with common sense. In his language the ordinary world is in a state of flux and cannot be said to be but only to become: it tumbles about between being and not-being. Things can be said to be only if they are eternally what they are and so are outside the flux of time. Such things he found in numbers and mathematical figures, but above all in what we call universals and he called Forms or Ideas—for example, in beauty itself, goodness itself, and so on. Though beautiful things come into being and pass away, beauty itself neither comes into being nor passes away, but remains eternally what it is. Only of the Forms (and such other things as may be grasped by thought and not by sense) can it be truly asserted that they are.

Thus it can be said that for Plato the objects of thought, not the objects of perception, are real; but by this he did not mean that they existed merely as objects of thought. On the contrary, they are for him eternal or timeless realities in their own right, whether we think about them or not. No doubt we come to know beauty itself (or beauty in itself) by a process of thinking stimulated by the sight of beautiful and perishable objects in the changing world; but even this process is to be described, at least mythologically, as the ‘recollection’ of a reality known in some former life.

Such a doctrine about our knowledge of universals cannot legitimately be extended without further argument to our knowledge of individual souls and of God, even although the intellectual element in the soul is said to be akin to the Forms; but at least the claim to know such Forms in virtue of this kinship helps to make knowledge of ultimate reality by means of pure thinking less suspect and eases the obvious difficulty of the ontological argument.

The philosophy of Plato with its claim to know by pure intelligence an eternal world more real than our perceptible world of change will always have a special attraction for religious men; and no one who has not been at one time a Platonist can know fully what it is to enjoy philosophy. But perhaps we must now say that Plato exaggerated the importance of universals as some modern philosophers exaggerate the importance of words. Men who make a philosophical discovery generally make too much of it.

In view of the Platonic doctrine it may be suggested that the ontological proof is misdescribed when regarded as an inference from our thinking, or even from our concepts, to the independent reality which is thought or conceived. What we conceive, it may be said, is a real essence, a timeless and independent universal; and the inference to existence—if it is an inference—is based on this real essence, and not on our thinking and conceiving.

Few philosophers would now hold that by pure thinking we are able to know real essences; and many would maintain that such a doctrine is unintelligible. But even if we set aside these grave objections, it is still hard to see how we can infer from a real essence to the existence of an individual who partakes of this real essence or in whom the essence is embodied. From our knowledge of beauty itself (if we have such knowledge) we cannot infer that beautiful things must exist. The argument must turn on the special character of the essence in question: it must be an essence such that only one individual can be its embodiment; and the individual must be such that it embodies only the essence and nothing more. In short, the ordinary distinctions of individual and universal—or of matter and form—must be overcome. In accordance with such a view God has been described as pure form, and the distinction between His essence and His existence has been denied. We are told, for example, that it is a mistake to say ‘God is good’. If we speak strictly, we must say ‘God is goodness’.

All this raises questions as to the sense in which God is said to exist. The doctrine does not mean that God exists as a universal exists, or even as a real essence exists. It means that He exists both as an individual and as a universal; and we have here a real essence which is supposed to determine one individual, and only one individual, through and through. This may seem merely to repeat the ontological argument in different language. But, in any case, if God's essence and His existence are indistinguishable, there can be no inference from one to the other. The ontological argument has become an assertion, intelligible, if at all, in the light of a whole philosophy.

A similar development may be found in philosophies of a very different kind. An idealist, for example, may reject the common-sense distinction between our thinking and what we think, and may tell us that to be real is to be thought—no doubt in some rather special sense of the word ‘thought’. As is sometimes said—not without risk of ambiguity—‘The real is the rational, and the rational is the real’. In such a setting the ontological argument may again find its place, although it will no longer be a simple inference, but rather the assertion of a principle which is worked out in a whole philosophy and must stand or fall with that philosophy as a whole. Doctrines of this type are found in Hegel and his followers, and the philosophy of Hegel is not to be disposed of in a couple of paragraphs. It is enough to say that it makes claims to a knowledge of ultimate reality which are not easy to accept.

Even if we are willing to adopt such theories, it is hard to see how the ontological argument can be valid for us unless we can have knowledge of God's perfection, and indeed of His essence, as He is in Himself. So bold an assumption is not to be lightly made. Philosophers and saints alike may be shocked at Hegel's claim for his own logic, when he says that ‘The content of logic is the representation of God as He is in His eternal essence and before the creation of nature and any mortal mind’. If St. Thomas rejected the ontological argument on the ground that no human being can know God as He is in His eternal essence, perhaps he is to be congratulated on his modesty and good sense.

§ 5. The concept of perfection

So far nothing has been said about the origin of our concept of perfection. How do we arrive at it? Is it merely a concept that we happen to have?

The simplest view would be that we derive it from our acquaintance with degrees of excellence. We find, particularly in human beings, a gradation from the less powerful to the more powerful, the less wise to the more wise, the less good to the more good. We suppose this gradation to be completed ideally in a being that is most powerful, most wise, most good—or even all-powerful, all-wise, all-good. We conceive certain human qualities as raised, so to speak, to a higher power—indeed to the highest possible power—and we apply them by analogy to God. Thus we speak of God as possessing absolute might, absolute wisdom, absolute goodness, and it is such characteristics that are conceived in our concept of perfection. If we believe in degrees of reality, we may also suppose the divine perfection to cover the possession of supreme or absolute reality. This is the basic supposition of the present argument.

Some thinkers would hold that by this means we get no clear concept at all—we use the word ‘perfection’, but we have no corresponding concept. Even if we have such a concept, it must be inadequate to God's nature since it is derived mainly from human qualities known to us. In any case we have now adulterated the ontological argument with elements drawn from experience—precisely what we are claiming to avoid.

It may be replied to this that so far from our concept of perfection being drawn from experience, it is, on the contrary, pre-supposed by experience—if the word ‘experience’ may be used to cover awareness of degrees of excellence. We can say that one man is more wise or more good than another only because we already possess the notion of absolute wisdom and absolute goodness—even if this notion is first evoked by experience of lesser degrees of wisdom and goodness. How could we know that anything is imperfect unless we already had some notion of perfection?

This kind of argument is unfashionable at the present time. It does not follow that it is invalid. But even if we accept it, there is at least a risk that we are still contaminating the unsullied purity of the ontological argument. There is a danger that we may now be arguing, not first from knowledge that the imperfect exists to our concept of perfection and then from our concept of perfection to the existence of the perfect, but directly from the existence of the imperfect to the existence of the perfect. This is quite a different kind of argument: it is no longer an ontological argument from pure thought to existence. The direct argument from existent imperfections to an existent perfection, as we shall see later, comes much more naturally to ordinary men than the ontological argument; and it is probably true that unless the argument from existent imperfections had already been used (and perhaps found incomplete), the ontological argument would never have been invented. But this does not justify us in confusing two arguments that are fundamentally different.

In Kant's Critique of Pure Reason we can find a much more elaborate account of the concept of the most real being (ens realissimum) as the basis of the ontological proof. He arrives at it by considering such abstract logical questions as the synthesis of all possible predicates and the relation of negation to affirmation. Such complexities He outside our present scope. We must try to look at the ontological argument free from all possible accretions; and in order to do this most effectively it is safer for the present to regard the concept of perfection as like the high priest Melchizedek—without father, without mother, and without descent.

§ 6. Existence

Let us turn to the ‘being’ or existence' we are supposed to infer legitimately from our concept of perfection.

There is much ambiguity in the word ‘exists’ and still more in the word ‘is’. We say readily enough that the conspiracy existed only in his imagination or even that centaurs existed only in Greek mythology. This usage is frowned on by sticklers for precision, but at least we can be sure that such imaginary existence is the last thing the supporters of the ontological argument wish to ascribe to God. Similarly we might perhaps say that some things exist as objects of thought—for example, propositions or even mathematical triangles. Some philosophers would prefer to say that these ‘subsist’ rather than ‘exist’, while others would repudiate such language altogether. But here again the ontological argument manifestly does not seek to prove that God exists or subsists as an object of thought. On the contrary, it seeks to pass from the premise that God exists as an object of thought (an object thought under the concept of perfection) to the conclusion that He must exist in Himself, independently of all our thinking. It is precisely this transition that is so difficult to follow.

We do not help ourselves very much when we say that if anything exists in any strict sense, it exists in itself independently of our thinking and imagining; for we are only using the word ‘exists’ over again. Perhaps it is impossible to say more than that existence is existence, and that if we do not know what existence is, then nobody can tell us. But in the case of finite things, like tables and chairs, we can indicate the conditions under which we are prepared to say that they exist. It is not enough to say that we must be able to see or touch them; for in certain pathological states (perhaps even when artificially induced) men are credibly said to see, and even to touch, things that do not exist at all. Material objects exist only if they are open to public inspection and have a determinate position in public time and space. This means that they must be subject to causal laws—or at least to laws accepted by physics and in some degree by common sense. If we believe that such things as electrons exist, we do so on the ground that although they cannot be observed themselves, they are connected by physical laws with objects open to public inspection. It is more difficult to speak of finite minds even in this over-simplified way, because they appear to have a determinate position in time but not in space—except in the sense that they are connected somehow with a body which is in space. But these refinements we can here neglect.

When we say that God exists, we do not mean to assert that He occupies a determinate position in public time and space or that He is subject to physical laws. The conditions under which we assert His existence are not the conditions under which we assert the existence of material objects or even of finite minds. Hence if we choose to define existence by reference to the conditions under which finite objects are said to exist, we cannot say that God exists. The word ‘exists’ like any other word can be applied to God only by analogy with our ordinary usage. We have to think away the conditions under which the word is applied to finite beings; and this is why it is hard to be certain what we are thinking when we say that God exists.

There are some who infer from such considerations that it is meaningless to speak of God's existence. But there are others who infer that the concept of God's existence is supremely intelligible because it can be grasped by pure thought without any reference to the conditions under which we have sensuous experience of finite objects. These are the two extremes, and the second of them is the one which finds expression in the ontological argument.

It is wise not to be too dogmatic on either side. If a thinker proposes to apply the word ‘exists’ only to finite things in time and space, then—for him—a statement that something exists outside time and space is by definition meaningless. He has made it impossible for himself to say significantly that God exists; he has perhaps closed his own mind to problems hitherto considered important; but he has not even contradicted, much less disproved, the theological assertion that God exists; for no reputable theologian would affirm that God exists as a finite object in space and time. On the other hand, he has challenged the theologian to explain what can be meant by the existence of an infinite being, not in time and not in space.

The theologian, in his turn, can afford even less to be dogmatic; for he has to admit that he ascribes existence to God only by analogy. The thought that something not finite may exist, though not in space and time, may be a necessary thought (if we allow our thinking to stray beyond the problems soluble by science); but it is purely negative and tells us nothing about the positive character of such a possible existence. To find any positive character the theologian may have recourse, perhaps he must have recourse, to a consideration of the existence of finite minds; and he may say that to exist is to be active in thinking and willing. But here again he has to ascribe such activities, and consequently such existence, to God only by analogy. Furthermore, if to be temporal is to be finite, he has to say that God's thinking and willing is not in time, and to face the objection that in that case we cannot understand it as thinking and willing at all.

The theologian has to admit that he sees God in a mirror—in a riddle. He is even forced into paradoxes. For him God is everywhere, and yet nowhere. If the expression may be used, God is everywhen, and yet nowhen. God must be conceived as thinking, but not as we think; as willing, but not as we will; as existing, but not as we exist. It may be possible to show that these paradoxes necessarily arise from the limitations of our own finite thinking; but to say we know that God must exist because we understand so clearly the nature of His existence is to arrogate to ourselves an insight wholly beyond our finite human powers.

§ 7. The inference from perfection to existence

With these considerations in mind we can now try to look at the ontological argument as a bare logical inference which professes to be valid in itself.

Unless perfection, when ascribed to God, were understood in a very special sense, the inference to God's existence would not even be plausible. We do not argue that a perfect man must exist, because if he did not exist, he would not be perfect. Such an argument would be received only with derision. But parallels of this kind always arouse the ire of those who champion the ontological argument. They tell us—not always too politely—we must be stupid not to see that the argument holds for one special concept and for that only. This, however, is precisely what they are being asked to prove. Those who seek to refute the argument by the use of such workaday parallels hold it to be a characteristic of concepts as such that we cannot pass from them to the reality of what is conceived; and they deny flatly that any concept can be so special as to justify an exception to this rule. All you can do with any concept by a purely logical operation is to analyse it and so to make clearer what you are already thinking. You cannot argue from a concept to an existent object. You can only argue from one existent object to another.

If we take an ordinary empirical concept, like the concept of elephant, we should never dream of saying that we could prove the existence of elephants merely by analysing the concept. In order to show that elephants exist we must, as it were, go beyond our concept and appeal to observation or experience. In technical language ‘Elephants exist’ is a synthetic, and not an analytic, proposition.

From a logical point of view the proposition ‘A perfect being exists’ is on exactly the same footing. If we could go beyond our concept of perfect being and appeal to experience of God—whether by means of a unique feeling or a special sense or a faculty of divination—we might say that we had shown the existence of a perfect being adequate to our concept. The extreme difficulty of any such claim we have already seen; for no feeling, however unique, and no sense, however special, can by itself be experience of a perfect being. But this difficulty need not trouble us here. The whole point of the ontological argument is that no such experience is necessary: we can show that God exists merely by analysing the concept of a perfect being.

This is surely a quite incredible claim. Even if we admit that the concept of a perfect being includes within itself, not only the concept of absolute goodness, absolute wisdom, and so on, but also the concept of absolute reality, we have still to get somehow from the concept of absolute reality to absolute reality itself. If these are different, how can we make the transition? If they are identical, how can there be a transition at all? Are we perhaps combining two contradictory views and supposing ourselves to pass from one to the other on the hidden assumption that there is no difference between them?

It is no use telling us in this connexion that all our concepts must refer to reality and that we are making a false separation between conception and reality. If our concepts can refer to reality, the separation has already been made. And although all our thinking may in some sense refer to reality, this is the reason why our judgements can be false as well as true. Our concept of chimaera refers in some way to reality; for it is the concept of an animal with a lion's head, a goat's body, and a serpent's tail. But, in spite of this, when we judge that chimaeras exist, our judgement is false. Similarly we all, let us hope, have some experience of wisdom, of goodness, and even of existence; but if it be granted that these can be conceived as absolute and as all combined in the concept of a perfect being, it does not follow that we make a true judgement when we say ‘A perfect being must exist’.

It would be sheer sophistry to tell us that even if we can infer the existent only from the existent, we can nevertheless infer the existent from a concept since a concept is itself an existent. This is a mere abuse of language; for concepts do not exist in the same sense as things. If the statement that concepts exist is taken to mean that a particular thought occurs at a particular time, we are no farther on. The ontological argument is based on what I am thinking, not on the fact that I am having a particular thought now.

§ 8. Existence is not a quality

It will be replied, with some acerbity, that the concept of perfect being is still being treated as if it were an ordinary concept: it is a very special concept with the special property of guaranteeing the existence of its object. To say this, however, is only to repeat once more what we are asking to be proved and not merely asserted. And even those who use this language assimilate the concept of a perfect being to ordinary concepts in certain respects. Some of them tell us we can know that a perfect being must exist as certainly as we can know that a triangle, defined as a three-sided figure, must have its three interior angles equal to two right angles.

Here I must confess that I am in the unhappy position of not even being able to infer from a mere analysis of the concept of a three-sided figure that a three-sided figure must have three angles. In order to see this I have first of all to consider how a three-sided figure can be constructed either on paper or in imagination; but this disability may be due to a lack of acquaintance with the higher mathematics. I am, however, able to infer from analysis of the concept of a three-sided figure that a three-sided figure must have three sides; and I am willing to suppose that all mathematical thinking can be reduced to similar analytic operations.

What are we to make of this parallel between mathematical thinking and the ontological argument? It appears to imply that existence is an extra quality which is thought in the concept of perfect being along with other qualities like absolute wisdom and absolute power—just as three-angledness is an extra quality which is thought in the concept of a three-sided figure. But it is sheer error to regard existence as one quality among others. Existence is not a quality at all. A thing with all its qualities either exists or it does not. It receives no additional quality by existing. The dodo as an animal species is unfortunately now extinct; but the concept of dodo, even if it could be a concept of all the qualities in the dodo, is precisely the same concept whether the dodo exists or not.

The objection to the ontological argument may also be put in linguistic terms. If we decide to say that a being is perfect only when we are able to say that it exists, we are entitled to say that a perfect being must exist. But this alleged necessity is a logical necessity based on the use of language, and it should not be mistaken for an insight into the real necessity of things. If, in spite of protests, I may again resort to ordinary concepts, we might also decide to say that a thing is an elephant only if it exists; and after all how could it be an elephant unless it did exist? But if we inferred from this that an elephant must exist, meaning anything more than that we should refuse to call anything an elephant unless it did exist, we should merely be deceiving ourselves. The existence of a thing does not depend on the usage of a word.

Such a linguistic interpretation would—with complete consistency—be repudiated by those who claim to know real essences and their interconnexions by means of pure thinking. Even if we are unable to agree with them, we may find at least plausible the doctrine that human beings can ‘see’ a necessary connexion between being good and being intelligent, or even between being supremely good and being supremely intelligent—a necessary connexion based on something more ultimate than an arbitrary use of words. But when we are told that they can also ‘see’ a necessary connexion between being perfect and actually existing, the position is far otherwise. The relation between perfection and real existence cannot be assimilated, except by the grossest confusion, to a relation between interconnected qualities.

If this contention is sound, it is fatal to any ontological proof of God's existence; and, in spite of a recent attempt to prove otherwise, it is equally fatal to any ontological disproof of God's existence. If the ontological argument is ever to become valid by being incorporated into the structure of some philosophical system, it will have to suffer a sea-change in the process. Yet the fallacy, even in its simplest form, is not so easy to detect that we need despise men who accepted the argument in circumstances very different from our own; and not many of us would be able to say where the argument had gone wrong if we were meeting it for the first time.

§ 9. Theoretical argument and religious conviction

It may seem that a great deal of energy and time has been spent on arriving at a purely negative conclusion, which most of us could reach by inspection—namely, that the ontological argument is not valid. But we may also have gained something more positive—perhaps a slightly better understanding of some of the concepts men have used in trying to think about God. At least we may have seen that concepts applicable to finite beings cannot be used literally in thinking about God, but only by some sort of analogy. This may warn us against accepting a too crude theology, which inevitably leads to an equally crude rejection of theology.

We embarked on this enquiry, not for its own sake, but in the hope of getting some light on the intellectual factor in religious experience. Perhaps we have not been too successful even here. The plain religious man, and even the saint, may make little use of the concepts we have considered in abstraction, but they do make some use of them. Furthermore, it is true to say that what is nowadays—sometimes rather confusedly—called an ‘attitude’ presumes a certain character in its object, even when there is little or no attempt to set forth this character in conceptual terms. The religious attitude is an attitude of worship, and we may say that it presumes an object of worship which is and is perfect. We may even perhaps say that it presumes an object of worship which not only exists but exists necessarily and unconditionally—which has no source or ground in anything other than itself.

Some of us may think it would be better for religion if men worshipped perfection without what has been called its ‘fallacious existential trimmings’; but this fails to satisfy a religious need which has its roots very deep in human nature. Men can worship only a God who is. Perhaps those who elaborated the ontological argument were working out unconsciously, on a high level of abstraction, some more obscure, and perhaps more fundamental, process of thought and feeling deeply involved in religious experience. This does not make the argument valid and may even be taken to throw doubt on the deeper process which it is alleged to express. We may be told that the ontological argument is a fallacious rationalization of an irrational process which is equally fallacious—a mere projection of our own desires, a kind of wishful thinking now made only too clear.

Whatever we may think of this, it seems that pure thinking, detached from experience and from life, is unable to give us knowledge of God's existence—or of His non-existence. In this respect it resembles pure feeling detached from thought. But pure feeling permeated by thought may be at least a source of personal religious conviction, even if this fails to pass the test of philosophy. If religious experience is of the whole man, it is unlikely to fit neatly into our tidy pigeon-holes; and in it feeling and thought appear so to interpenetrate that to some minds the very thought of a perfect or most real being whose existence springs from, or is identical with, his essence may be a source of numinous feeling and so of religious conviction. Perhaps after all we have learned something of the intellectual factor in religious experience. Even although the ontological argument is the one that comes least naturally to us, it does seem to express in its own arid intellectual way a demand of religious consciousness itself.