You are here

Chapter XVIII | The Limits of Knowledge

§ 1. The given

Human experience, or human knowledge, has so far been analysed in a rather simple way. More difficult questions have now to be raised, not only about experience itself, but also about the limits of our knowledge and about the status of the world that is known and of the self that knows it. Readers who have found the last chapter difficult enough may be advised to pass straight on to the conclusions outlined in § 8. As human knowledge is still being analysed from the point of view of the knower, the painful practice of speaking in the first person will have to be continued.

If there is any truth in what has been said, my experience (and my world) has to be built up on what is given from moment to moment; and it is by means of the given that I have what may be called direct contact with reality. But what is it that is so given?

The word ‘given’, if it is taken strictly, should be applied only to something of which I am aware without any activity on my part. If sense is supposed to be altogether passive, then the given is what is given to sense alone.

It might be thought easy to describe what is given, yet there are few topics about which philosophers disagree more. The perspectives hitherto taken to be, as it were, the bricks with which I build my experience and my world, cannot possibly be given: like other bricks they have to be made. Many thinkers have held that what is given to sense alone is a ‘sense-datum’; and by this they mean, for example, a bulgy red patch or shape or expanse which I see out there now when I am looking at what I believe to be a tomato. For reasons of brevity tactual and auditory sense-data—not to mention others—will have to be ignored.

The sense-datum theory has taken many forms. In spite of opposition, and almost persecution, when it first migrated from Cambridge to Oxford not so many years ago, it rapidly spread, like a prairie fire, over the English-speaking world, but at present it seems almost to have burnt itself out. One of the best accounts of it is to be found in Professor Price's ‘Perception’—a work which retains its value after twenty years; but he too would probably express his doctrine now in somewhat different language.

When I look at a tomato, I think—though I find it hard to be quite sure—that I can see a definite red bulgy expanse out there now; and if I were an artist like Professor Price, I could no doubt do this much more easily. As it is, I find it most difficult to see a coloured expanse and not to see a tomato. Difficulties of this kind have given rise to modifications in the doctrine, for which the reader may be referred to Professor Price's recent book called ‘Thinking and Experience’. But if subtleties may be ignored, it is reasonable to hold that while I can be mistaken in believing what I see to be a tomato—it may, for example, be an object made of wax or a mirror image or even a hallucination—I cannot be mistaken in saying that I see a red bulgy expanse out there now. It is also reasonable to hold that I build up my experience of a tomato by means of seen red shapes out there. Only by means of seen and felt shapes can I build up my experience of material objects (although it is true that a blind man can build up his different experience of material objects without the aid of colours).

The fundamental difficulty about all this is already suggested by Professor Northrop's account of Eastern ‘nihilism’ as outlined in Chapter VI. There is reason to believe that what is given to sense alone is not anything so definite as a seen shape, but is more like an ‘undifferentiated continuum’—something vague and obscure and indeterminate, perhaps similar to what I feel within my body when I am hardly conscious that I have a body. According to Professor Northrop, it is possible to apprehend such an undifferentiated continuum by means of sense alone without being aware of anything else. Even if few of us are sufficiently practised in ascetic discipline to be capable of such pure apprehension, we may have grounds for accepting his hypothesis about the vagueness of what is given as almost certainly correct.

Many empirical observations might be adduced in support of such a contention. It is borne out, for example, by the experience of blind men who have acquired sight for the first time—they find the utmost difficulty in distinguishing by sight between circles and squares and triangles, although they already know these simple differences by touch. An infant, who lacks this advantage, must find the difficulty still greater. We have to learn how to see coloured shapes out there, and still more we have to learn how to estimate their size and distance. The given may be compared to a letter in invisible ink which can be brought out only by the appropriate procedure.

If we accept this view and assume at the same time that through the given we are in direct contact with a reality which is what it is independently of our experience, it seems impossible to believe that this reality as it is in itself can be identified either with our sense-data or with the material objects we know by their help; for both of these are not merely given, but are also made what they are by activity on our part. On the other hand, if we assume that reality is only what is given, we may be forced to adopt the Eastern doctrine which has been described as maintaining that what is is what is not.

If we look at things ingenuously, what actually seems to be given is not only more definite and precise than an undifferentiated continuum, but also more permanent and orderly than a collection of sense-data. When I look out of my window, what I seem to see is a steady landscape of fields and trees and houses and hills—part of my stable, though changing, world, one of the perspectives of which my experience is made up. But every philosopher would agree that there must be more to this than mere seeing—more than can be given to sense: we are not provided, as it were gratuitously, with slabs of reality which we have only to fit together like a jigsaw puzzle. Even what seems to be ‘given’ must somehow be ‘taken’ as well. It is possible to look for a pipe on the table and yet fail to see it, although it is, as it were, staring us straight in the face the whole time. The word ‘given’ is highly ambiguous. If anything of which I am immediately aware can be described as given, no matter what I may have done to bring it before me, then everything may be so described—even the present argument.

§ 2. Space and time

What I see when I look at a tomato has been described as a coloured expanse. It might equally well be described as an extended colour—we cannot see colours that are not extended. We might also say that it is an enduring colour or—rather more artificially—a coloured duration. There are no such things as mere colours, for whatever is a colour must also be spatial and temporal; and if colours are given out there now, space and time would seem to be so far given too. The same statement applies broadly to all sense-data.

In human experience space and time are all-pervasive. Whatever may be given to me—whether I take this to be an undifferentiated continuum or a red expanse or a tomato or a landscape or even a world—seems to be always a bit of filled time; and part of what fills time is generally, and perhaps always, a bit of filled space. Every bit of time is for me part of a longer time; every bit of space is part of a larger space; and I assume I could go on tacking bit to bit in imagination for ever.

What is given to me can only be given now. But since every ‘now’ is a different ‘now’, what is given is not a fixed point on which I can rest a lever that would move the world: it is at best a sliding point—a perpetually vanishing point—and to have anything before me at all I must continually, as it were, attach a past that is no longer to a present that is always ceasing to be. I know my sense-data, as I know my world, only by a synthesis of times and spaces as well as of sense-data—that is, by a kind of double synthesis which itself goes on in time. Without this I could have no experience, nor could I even begin to speak of my point of view. Without this there would be no world for me.

If this is true, the world I know is not merely given: it is also made or constructed—although not created—by me on what seems to be a most precarious foundation, which is entirely concealed by the building that covers it. Nor should it be forgotten that in constructing my world I construct also my own mental history. The world I know seems to be a kind of between-world—the joint product of my activity and an unknown reality. My knowing can never be a mere passive reception of reality as it is in itself. Reality does not migrate into my mind under its own steam: it appears to me under a form imposed by the limitations of my mind.

This contention is independent of philosophical or scientific theories about the nature of space and time—or of space-time, as we are now expected to call it—but it may be strongly reinforced by such theories. On a simple view we can in thought separate space and time from what is in them, and we assume that there can be only one space and one time. These have the curious characteristic of being composed entirely of relations; and these relations are themselves spaces and times. This does not mean that space and time are things like boxes which can exist apart from what is in them—you cannot make a box out of relations. They are not substances and they are not qualities. They are not even objects, as bodies or minds are objects. They are sometimes spoken of as if they were merely relations between things or even as if they were orders of things; but they seem to be rather the condition of such relations and orders. Sometimes it is said that the word ‘space’—or the word ‘time’—stands only for a set of facts, or even a set of propositions, about objects. It is hard to see any difference between ‘a fact about’ and ‘a true proposition about’; and I must confess that this doctrine is to me incomprehensible: to speak ingenuously—how can a body be located in a set of propositions? But it must at least be recognized that space and time are paradoxical in character—so paradoxical that Kant maintained, and even claimed to demonstrate, that they are contributed by our own minds to the world we know; that they are, as it were, only the spectacles through which we experience as spatial and temporal a reality that in itself is not in space and time. This is a view which has pressed itself for many reasons on thoughtful men and is not unfamiliar in theology. If it could be upheld, it would establish beyond doubt the contention that we can never know reality as it is in itself.

In any case space and time, beyond the ‘here and now’ which may be thought to be given with a sense-datum, have to be constructed by an act of synthesis: they cannot possibly be given either as finite or as infinite wholes, although we may think of them as if they were. Even ‘here’ and ‘now’ have no meaning apart from ‘then’ and ‘there’.

All of this is complicated by doctrines about what is called the ‘specious present’. These attempt to explain the fact that during a relatively short time I may be able to see a body moving and not merely to remember that it was once in a different place from where it is now. For a brief discussion of this problem I can only refer to my book ‘In Defence of Reason’.

§ 3. Imagination

An infant must pass gradually from vague sensation to ordered experience. In spite of having made this journey ourselves we know very little about it, but it should not be confused with what happens in mature experience.

When I see a tomato, it may be said that I must pass from the sign to the thing signified—from the sense-datura to the material object. Yet it would be a mistake to suppose there are two clearly defined stages happening one after another. If this passage occurs, it seems to happen all at once.

But what is it that does happen? The language employed might suggest that we were making an inference or offering an interpretation, but obviously we are doing nothing so sophisticated. When the validity of mystical experience was discussed in Chapter X, I used words like ‘conceiving’, ‘thinking’, ‘judging’; but I had to point out that even in describing our experience of finite objects they were not to be taken in the ordinary sense, and that perhaps we had no recognized word for what is meant. Similar difficulties must arise about words for describing the passage, if there is one, from an undifferentiated to a differentiated continuum and from a material object to an ordered world.

It is all too easy to make hay of this over-explicit language, especially if we interpret it as implying some kind of verbal utterance; but those who concentrate on the inadequacy of the terminology are in danger of missing the problem.

Without entering into these complicated questions, on which Professor Price is probably the best guide, I wish only to call attention to one factor which is too commonly neglected—the factor of imagination.

In our experience we are always doing something more than apprehending what is given here and now; and it is not nearly enough to say that we supplement this by entertaining, or asserting, a series of theoretical propositions. Such a view is possible only to those who believe that all thinking takes place solely by means of words. Yet nothing can be more certain than that men can think also by means of pictures and images, and by gestures as well. Perhaps there must always be something more than these in our thinking; but whatever it may be, this ‘something more’ need not be the utterance, aloud or to ourselves, of verbal propositions.

The very fact that our experience is temporal means that we could have no experience without memory. If we are content to speak in a simple and provisional way, we may say that we must be able to reproduce in imagination what we have sensed and to recognize that the image now produced is like what we sensed before. It should be unnecessary to add that the image need not be visual.

One of the obvious difficulties about this is that such recognition would seem to be impossible unless we already remembered what we had sensed. How otherwise could we recognize that our present image is like it? Some philosophers would be content to say that the image produced feels familiar; but memory must be more than an inference from a felt familiarity. Yet it is true—to take a rather different case—that when we come into a room we may remember that we have been there before without remembering the previous occasion, and presumably without producing any image of what we formerly saw.

In spite of these and other difficulties we may perhaps assume that without images of some kind there could be no memory and no experience. What must be added is that the memory of definite events, which is at least usually accompanied by images, shades off into the other kind of memory which consists in remembering, as we say, how to do something—for example, how to ride a bicycle. So too our activity in imagining seems to shade off from the making of images to knowing how to make them. Between these extreme limits there may be a whole range of activity, including the making of vague or incomplete or wavering images; and this range, I think, is what Kant had in mind when he maintained that imagination produced ‘schemata’ for concepts and not merely definite images. It is essential to recognize that our imaginings may be highly schematic and yet that without them our concepts would be meaningless.

If imagination is an ability to have or produce images of what is not immediately present to sense, its function is not confined to memory images. We expect as well as remember. In our experience of material objects we have not only to reproduce in imagination what we formerly sensed: we also have to fill up gaps in experience and to anticipate the future. This too may be done in a highly schematic way.

What I am trying to suggest is this. In the synthesis necessary to any experience which goes on in time the work of imagination is fundamental. It is by means of imagination, however schematic, that we attach the present to the past (or the past to the present) and construct on the basis of immediate sensation an ordered world in one space and time. In so doing we construct also time and space themselves.

This double process of synthesis is most obvious in the counting and measuring by which we determine the relative sizes and shapes and durations of material objects. This cannot be done without concepts and judgements: imagination by itself would at most give us only a picture, not a part of an ordered spatial and temporal world. Such a world could never be given to mere sense or even to sense plus memory of sense-data. So far as we take it to be given, it is given only to beings who imagine and remember and expect and think.

§ 4. Principles of synthesis

The synthesis which constructs my experience and my world must take place in accordance with principles. I have to exclude chance associations, casual fancies, dream visions—not to mention illusions and hallucinations. My imagining is not merely wild. It is ordered and controlled and—we might even say—intelligent.

When I recognize that what I am looking at is a house, my imaginative synthesis of the sides and back may be guided in accordance with an empirical concept of ‘house’ derived from my past experience of houses. But this suggestion merely postpones the real problem; for I have still to account for the past experience of houses which has enabled me to acquire the empirical concept of ‘house’ by some process of abstraction. Whatever may be the object recognized—and this is the crucial point—my imaginative synthesis is always controlled in accordance with universal principles which are not derived from experience, but are manifested in experience and are its necessary conditions, although they are at first followed unconsciously and are made clear only by later reflexion.

The most obvious principle manifested in the imaginative synthesis of experience is that I must combine in one time and space whatever it is that may be given—we need not here consider afresh whether what is usually supposed to be given must not also be made. Nothing can be an object of experience, or a part of the world I know, except in so far as it is assigned a position, however vaguely, within the framework of time and also—subject to certain qualifications—within the framework of space.

If this view is correct, it may have far-reaching implications.

Must we not, for example, assume that the parts of space (and of time) are homogeneous, and must we not maintain this homogeneity in our imaginative and conceptual constructions? It could never be established by any empirical observations. And since we can know time and space only through what is in them, must we not also in our syntheses assume that there is something permanent which fills space and time—something to which what is given successively may be attached as a state or quality?

This kind of language fits in, not only with ordinary experience, but also with Newtonian physics. We meet with fresh difficulties to-day because modern physics is developing a new language, and it is not clear how far this new language can be used for describing ordinary experience. But the relation between physics and experience is a problem far beyond our present scope; and in any case all physical theories are based on ordinary experience and could not arise without it. It may be that in the future we shall have to modify our language, but for most of us at least this time has not yet arrived.

If in order to be aware of objects in one time and one space we have to combine the given in accordance with universal principles, then all objects as known to us must—as a consequence of our activity—have the characteristic of being so combined. To conceive the principles of our activity is thus to conceive also characteristics which every object must have if it is to be an object for us. That is to say, by making clear to ourselves the principles of the synthesis necessary for experience we can acquire the concept of ‘an object as such’. The concepts of the various characteristics—such as, for example, causality—which every object must have if it is to be an object are properly known as ‘categories’, and these characteristics may be described as ‘categorial’ characteristics. It is a great pity that to-day these words are so often used in a much vaguer and less useful sense.

If this is true, it means that, if we are to have experience of objects, we must make use of certain a priori principles and concepts which are not derived from sense but from our own spontaneous activity.

Such a doctrine, although it is anathema to Logical Positivists, is not unfamiliar to moderate empiricists such as Professor Price and Professor Broad, who are prepared to speak of a priori concepts like ‘material thinghood’ and ‘causality’. It is important to trace—if we can—the origin of such concepts. I have already suggested that one reason, though not the only reason, why we have to assume causes and effects in experience is that in our thinking we must distinguish between grounds and consequents. Similarly it may be thought that one reason, though not the only reason, why we assume something permanent underlying the changes in space and time is that we must in thinking apply predicates to subjects. Concepts such as ‘ground and consequent’, or again ‘subject and predicate’, can never be derived by abstraction from what is given to sense: they arise only because they are manifested in thinking and are subsequently formulated by reflexion—they are not empirical but a priori. The categories, without which we can have no knowledge of objects, may spring partly from our imaginative synthesis in one time and space of what is given to sense and partly from the nature of our thinking as such. But these are highly technical questions and must here be left aside.

To some philosophers the language here employed may be unintelligible, and there may be objections to talk about concepts (or principles) ‘controlling’ or ‘guiding’ our imaginative synthesis and being ‘manifested’ or—still worse—‘operative’ in our thinking. It is therefore with special pleasure that I quote part of a sentence from Professor Price. ‘Concepts or abstract ideas may operate in our minds (and in our behaviour too) without ever being present to our minds’. I should be sorry to ascribe to him my own errors, or even the errors of Immanuel Kant; but this statement puts in a nutshell an essential part of the doctrine I have been trying to expound.

§ 5. The phenomenal world

Whatever be our view of the details, the general doctrine leads to one all-important conclusion.

If the world of objects known to me—and to others—is what it is as the result of a spontaneous synthesis, there can be no reason to believe that this world of ours is the world as it is in itself. We can know the world, not as it is in itself, but only as it must appear to finite minds like our own.

To say this is not to say that we create the world or that it is a product of arbitrary imagination. On the contrary, we are supposing that the synthesis required for experience must conform to principles which are common to all finite human intelligences. Furthermore, we are also supposing that the synthesis is a synthesis of what is given: however indeterminate the given may be, we cannot make it into any sense-data we please, nor can we combine sense-data in a way which ignores their character. And finally we are supposing that what is given is an appearance to us of a reality which is wholly independent of ourselves. This last supposition is admittedly one that can never be demonstrated; but all three suppositions appear to be necessary if men are to be aware of a common world. If these suppositions are adopted, our common world must be described as a ‘phenomenal’ world (a world of appearances) or again as an ‘in-between’ world (one which is the joint product of our minds and an unknown reality). Yet although the world we know is not the world as it is in itself, we must nevertheless take it to be the real world as it must appear to human beings. In spite of the ambiguities of language there are not two worlds, but only one world (1) as it appears to us, and (2) as it is in itself.

To hold this view is not to deny—but, on the contrary, to affirm—that the world as it appears to us is (to speak roughly) a world of bodies occupying a space and time which extend beyond our ken. This is the world explored by science, and we can consistently accept whatever account of it may be given by experts. For purposes of common sense and science we are right to think and act as if it were a world of things as they are in themselves. Our doubts about this are metaphysical doubts—the kind we are expected to dismiss when we serve as members of a jury. To those who hold that the beliefs of common sense and science are hard data exempt from philosophical criticism, all such doubts must seem perverse or even woolly. To those who take a different view of philosophy such an attitude appears to be ingenuous—the symptom of a dogmatic slumber that is still unbroken. This is one reason why philosophers of these two schools often seem to be incapable of understanding one another.

Whatever be the difficulties in the doctrine expounded, it is no accident that scientists, and especially biologists, who reflect on their own procedure often come to a similar conclusion. Thus, for example, Professor J. Z. Young in his Reith Lectures can remark: ‘At best what we are producing is a system of the universe as conceived by man, the talking animal’. He even goes so far as to use the word ‘create’ in this connexion, and says his researches can show ‘that the brain of each one of us does literally create his or her own world’. If this is so, the world we create cannot be reality as it is in itself—unless we assume that we create all the universe there is.

It will be observed that he prefers to speak of our brains where I speak of our minds. The principles of synthesis about which I talk become for him ‘characteristic rules’ which each brain ‘sets up’; and these are identified with ‘patterns of activity’ or ‘models of action’ in the cells of the brain. He combines the language of behaviourism with that of philosophical psychology, though the two go ill together; and if he really identifies electrical activities in the brain with conceiving, it is hard not to think that this is a confusion. Furthermore, we have the very paradoxical result that if the brain creates the world, it must also create itself, for it is only a part, and a very small part, of our world: we have no right, as philosophers, to give it the privileged status of being the sole reality as it is in itself. Professor Young's view, as I see it, does not explain how he can know what he says he knows, and it seems both to recognize and to ignore the limitations of human knowledge in one and the same philosophical theory. What is remarkable is that in spite of using very different languages we seem to arrive at very similar conclusions. At the very least we both agree that ‘we are set about with mysteries’; and this is the beginning of philosophy.

§ 6. Antinomies

The mysteries of the world we know and of the way we know it and of what may lie beyond our knowledge are endless; but some thinkers have found especially mysterious the fact that if we try to think about the world as a whole, we seem to fall into contradictions which are known technically as ‘antinomies’.

Some indication of this has already been given in our account of the cosmological argument for a first cause or a necessary being. It will be enough here to mention a simpler example from Immanuel Kant. We are tempted to say both that the world must have a beginning in time and equally that it cannot have a beginning in time. Such contradictory theories are commonly opposed as thesis and antithesis, and it seems possible to put up an equally good case for either. Whichever side we adopt, our argument may seem unanswerable so long as we confine ourselves to refuting the other. From this it may be inferred that this world can only be an appearance—an appearance of some reality which presumably could be conceived, at least ideally, without contradictions. This conclusion would be a further confirmation of the doctrine we have expounded.

It is only fair to say that this kind of argument is commonly scouted to-day on the ground that it has been superseded by modern advances in physics, mathematics, and mathematical logic. Mr. Bertrand Russell is the leading exponent of this view, and the simplest statement of it is to be found in his book ‘Our Knowledge of the External World’.

It would be foolish for any one who is not an expert to argue with Mr. Russell on his own ground, and he may be right in saying that this kind of argument should now be abandoned. But he is not infallible as an interpreter of Kant; and when he imputes to Kant an ‘elementary blunder’, we may feel inclined to remember that many of the elementary blunders imputed to the sage of Koenigsberg have been blunders on the part of his interpreters.

It is a little surprising—at least at first sight—that Mr. Russell should dispose of the whole argument by refuting the thesis; for on Kant's view the possibility of refuting the thesis (or the antithesis) is precisely the mark of an antinomy. The refutation consists in rejecting Kant's claim that ‘the infinity of a series consists in this—that it can never be completed by a successive synthesis’. Mr. Russell maintains that the notion of infinity is primarily a property of classes and is only derivatively applicable to series. ‘Classes which are infinite are given all at once by the defining property of their members, so that there is no question of “completion” or of “successive synthesis”.’ The italics are mine.

In reply Kant would, I think, say of Mr. Russell, as he said of Leibniz, that he is lacking in transcendental reflexion; or—in simpler language—that he fails to ask whether his conceptions are derived from sense or thought. Kant is not talking about mathematical infinity, but about the world in space and time; and this cannot be given by any definition—unless by a strange revival of the ontological argument. On Kant's view—and surely he is right—the world, whether finite or infinite, can be given only by a successive synthesis; and it is by no means clear that he and Mr. Russell are talking about the same problem.

Curiously enough, Kant's doctrine of the antinomies bears some resemblance to the ‘logical paradoxes’ which have so exercised modern philosophers, and it might seem to have a similar claim for consideration. But these questions are highly technical, and it must be recognized that mathematics and mathematical logic—to say nothing of physics—have gone far beyond anything known in the eighteenth century. The Kantian doctrine may well require reconsideration and restatement. Nevertheless the plain man may still feel himself faced with a mystery when he asks whether the world has or has not a beginning in time. He may even legitimately wonder whether there may be a reality which is not in time at all.

§ 7. The self

If we turn to the self that knows the world, we seem to find further paradoxes, or even contradictions, painfully similar to those already found in Professor Young's account of the brain.

On the view hitherto expounded it is clear enough that not only my body, but also my mental history, can only be part of the phenomenal world I know; for they too, as we have seen, must be a product of a spontaneous a priori synthesis. So far as I know myself as an object, I know myself only as I appear to myself under the limits of finite knowledge. I cannot know what I am in myself any more than I can know what the world is in itself.

The paradox is this. On the one hand, my mental history, so far as I am taken to be a knower, is a history of the way in which I construct my world. On the other hand, my mental history, considered solely as an object known, is only a part of my phenomenal world; and—in spite of many puzzles—it appears to be causally determined by the other parts of my phenomenal world, and in particular by physical bodies (including my own).

The difficulties become even greater if we ask what is the self that is supposed to construct its phenomenal world (including its own mental history) by sensing and imagining and remembering and thinking.

Some would say that I who think and know must be a permanent substance; for only a permanent substance can be aware of a changing point of view, a changing world, and a changing self. It is harder to see what this can mean to-day than it was when the word ‘substance’ had a meaning in physics and could be applied by analogy to the soul.

Others would say that I who know can be only a structure of events. But is it possible that a structure of events could know itself to be such a structure? The events are continually succeeding one another and passing away. They can be a structure to a mind which knows them—It is not so easy to understand how they can be a structure to themselves.

There appears to be a tendency both to affirm and to deny that I am something over and above my past thoughts and my present thinking. If we reject the language of substance, we may be tempted to say that I who think am, not merely a point of view, but a centre of activity, or even the unity of an activity. Such statements serve to call attention to a problem, but it is hard to make them precise. It looks as if I can know myself only as an object—only as a succession of mental and biological events. Yet what can ‘an object’ mean unless it means ‘an object to a subject’? And surely I know—if I know anything—that I am a subject as well as an object, a knower as well as something known, and am for this reason a person and not a thing. Although part of the world of nature, I seem to stand outside of nature and to view it almost as if I were a god. Yet all I seem able to know of myself as a knower are the principles manifested in my cognitive activity—the principles on which my world and my own changing history are held together as objects in one space and time. These principles cannot be causally explained; for they are the conditions of any attempt at causal explanation. They are indifferent to time, and they look almost as if they were an intrusion into our phenomenal world from some other world than this.

All of this is admittedly difficult and obscure. The problem may have been stated in wrong terms; but this is better than saying that there is no problem at all. Part of the difficulty arises from the fact that even to see the problem requires a revolution in our ordinary thinking, and especially in our scientific thinking, which is primarily concerned with known objects and not with our knowing of them. Thus Professor Young, if I may return to him, wishes to abolish the word ‘consciousness’ altogether, because, as he says, if consciousness ‘is a thing in the ordinary sense it could be observed directly like any other object’. Since it cannot be so observed, he concludes that it must be rejected as an ‘occult quality’. But even the simplest of those who use the word do not mean that consciousness is an observed object or ‘a thing in the ordinary sense’. What they mean is that there can be no observed objects unless there is observing; and they use the word ‘consciousness’ for the observing, not for any thing or object observed, and still less for any quality, occult or otherwise, of an object.

In spite of difficulties there seem to be two points of view—or two ways of talking—each equally legitimate; or perhaps we should say that while the common theory describes the objects seen from a point of view, the other attempts to consider what is implied in there being a point of view from which anything can be seen or described. And if ordinary language is any guide, when there is thinking there must be an ‘I’ who think. The assertion that there must be an ‘I’—or even an ‘It’—which thinks and knows presents us at the very least with a question, a puzzle, a problem. To suppose that it can be solved by any talk about objects and qualities is to miss the problem altogether.

§ 8. The limits of knowledge

This has been a long and troublesome discussion, although it has touched only on the fringes of the subject. It is to be hoped that the conclusion will be relatively simple, or at least that it will be easy to see what is its bearing on religious faith.

An attempt has been made to analyse human experience from its own point of view and to determine the conditions without which it could not be what it is. It looks as if we must be given at every moment something utterly obscure and perpetually changing, and on this precarious basis must construct and hold together our ordered world (including our own mental history) in accordance with spontaneous principles of imagination and thought. This process is dominated throughout by the Idea of a whole—we might almost say by the Idea of the whole. Only on the supposition that this idea is operative, whether consciously or unconsciously, in our imagining and thinking can we understand our efforts to construct one world in time and space, and so to construct one time and one space (or one space-time). Such wholes cannot be given, but they seem to be postulated as the necessary conditions of ail our experience.

Unless we assume the miracle of a pre-established harmony, a world so constructed cannot be reality as it is in itself, but only reality as it must appear to finite human minds.

This is the main conclusion of our argument. It rests on the prior assumption that there is a world independent of our experience, and that this independent world, as the source or condition of what is given, determines (in conjunction with our own spontaneous activity) the character of the phenomenal world as we know it.

Although this assumption seems natural and almost inevitable, and although without it it is difficult to see how we can know a common world, it may be rejected by philosophers who otherwise attempt to analyse experience in the same sort of way: they may assume instead that there is no reality independent of our experience; and they will then arrive at a different conclusion.

Some thinkers, for example, may conclude that reality is nothing but a collection of finite experiences. Others—or perhaps we should say simply ‘I’—may conclude that reality is nothing but my own finite experience.

Even the first of these views is hard to accept: it seems to run contrary to a natural human instinct, and even the totality of actual finite experiences would appear to be obviously incomplete. The second is wildly paradoxical: it is almost impossible for me to act and think as if I were the sole reality—a single, solitary, imperfect mind. Both views take human knowledge to be absolute and the world to be all of a piece. Here we need only note that neither of them leaves any room for belief in God.

A more complicated view is possible. Even if we hold that there can be no reality beyond experience, we may recognize that our finite experience, and even the experience of the whole human race, is manifestly incomplete. On this basis some thinkers claim to know that finite experiences are parts of one all-embracing experience. They may not treat human knowledge as absolute, but they do treat it at least as if it could approximate to knowledge of absolute reality or to an absolute experience.

This doctrine, which has been developed with the utmost subtlety, is unpopular at the present time: it seems to claim for man a knowledge beyond his finite powers. Like Platonism, it may afford a basis or background for religion—although perhaps for one that must in the long run be replaced by philosophy. But the modern predicament of religion has arisen because the advance of science, which has enabled men to know so much more about the temporal world, has also—whether rightly or wrongly—made them doubt their ability to know a world of eternal reality or an infinite experience other than their own.

The belief that the ordinary world we know by common sense or science is only the appearance to us of an underlying reality which we can never know has been here supported by a variety of arguments which some may think unnecessarily elaborate. It has often been entertained by thoughtful men without a great deal of argument: to some it may seem that there is no initial probability, or even plausibility, in supposing that limited and imperfect minds can grasp reality—or even a bit of reality—as it is in itself. Indeed it is almost a commonplace to-day that all human knowledge is relative; and it is hard to see what this can mean unless it means that reality as it is in itself can never be known by us. Such a hypothesis is obviously not one that can ever be proved or disproved by science, though it may be suggested by reflexion on scientific procedure; but at least it is a safeguard against the metaphysical assertion that the bounds of science are the bounds of reality and also against idle speculations to which there can be no end.

It should be noted that the arguments here used are entirely different from the cosmological argument: they make no attempt to pass from knowledge of contingent beings in time and space to knowledge of a necessary being not in time and not in space. Nevertheless if they are accepted, we may find ourselves obliged to use at least the negative concepts of the cosmological argument in order to think about reality in itself as the condition of our experience—if, that is, we allow ourselves to think about it at all. We may even be tempted to conceive it as absolute and unconditioned and self-sufficient, however little meaning we can attach to these phrases, and however impossible it must be to find any confirmation in experience.

From the point of view of religion our doctrine does at least recognize the mysteries by which we are beset. Taken in itself it could only lead to a complete suspension of judgement if man were a purely scientific mind; but in abandoning a claim to knowledge it may make room for faith. It accords with much that is said by religious men, when they tell us, for example, that the world is not only a manifestation of God's will, but also a veil behind which He is eternally concealed. We might almost say that for all religion the ordinary world as we know it can only be the appearance of a deeper reality which is beyond human comprehension.

From the book: