IN this chapter, I wish to consider whatever would ordinarily pass for knowledge of particular matters of fact, in so far as this is not obtained by a process of deliberate scientific inference. I want to consider this as far as possible independently of the scientific laws based upon it, though not completely without reference to the primitive beliefs by which common sense draws inferences from perceptions. In particular, I wish to abstain from introducing the causal theory of perception, unless, on investigation, this should prove impossible. It will be understood that my purpose is epistemological: I am considering perception because it is involved in the premisses of empirical sciences, not because it is interesting as a mental process. It is of course necessary to consider its intrinsic character, but we do not do this for its own sake, we do it for the sake of the light that it may throw upon the character and extent of our knowledge.

We are met at the outset by a difficulty due to the fact that philosophical terminology is inappropriate when the views to be expressed are in any way unusual. "Knowledge" and "belief" both have connotations which are inconvenient for the purpose I have in view. They are both commonly applied in orthodox psychology to something conscious and explicit, such as is, or may be, already expressed in words. For our purposes, it is desirable to include more primitive occurrences, such as may be supposed to exist in animals. Obviously a bird can see an approaching man, and fly away in consequence. I wish to include under "perception" what happens in the bird, and also to say that the bird "knows" something when it sees a man, though I shall not venture to say what it knows.

But at this point a good deal of caution is necessary. My knowledge of the bird is part of my knowledge of the external world, and is partly, if not wholly, physical knowledge. Therefore when I am asking: how do I know about the physical world? I have no right to begin by comparing my knowledge with that of a bird. I must start from myself and my own cognitions, and use the bird only to suggest hypotheses. This caution applies also to what was said in Chapter XV.

Again, there is always a danger, in epistemology, of putting the less certain before the more certain. My knowledge of the process of perceiving is less certain, and less primitive, than my knowledge of percepts. When I say, "I know that I have just heard a clap of thunder," I am saying something not so indubitable as when I say, "There has just been a clap of thunder." It is facts of this latter kind that are required as premisses in physics. A man might be completely competent as a physicist if he knew such propositions as "There has just been a clap of thunder" even if he knew no propositions such as "I know that there has just been a clap of thunder." The consideration of our knowing, as opposed to what we know, is forced on us by the fact that what we think we know sometimes turns out to be false; if this were not the case, an analysis of matter need not consider our knowing at all. As it is the case, we are compelled to examine our knowing, as well as what we know, with a view to discovering, if possible, how to minimize the risk involved in taking as knowledge what, on reflection, we still believe to be knowledge.

We are often urged to adopt an artificial naivete in investigating problems concerning what we know; if we do not do so, we are accused of the "psychologist's fallacy." Now in certain problems this caution is quite proper, but in others it is not. My problem is: What do I, here and now, know about the external world, and how do I know it? It is obvious that my knowledge of the external world cannot be dependent upon (say) how long it takes a fish to learn to recognize the man who feeds it, since this supposes that I know all about the fish and the man and the feeding. Facts about the perceptions of babies, such as we considered in Chapter XV., come under the same head. Long before I can know that there are babies, I must know many other things about the external world. I want to start from what comes epistemologically first in my existing knowledge now; and in this problem, obviously, I cannot assume that I already know all about the experiences of animals and babies. There must therefore be no artificial naivete, but a straightforward investigation of my knowledge as I find it.

The position may be illustrated by Chuang-Tze's story of the two philosophers on the bridge. The first says: "See how the little fishes are darting about. Therein consists the pleasure of fishes." The second replies: "How do you, not being a fish, know wherein consists the pleasure of fishes?" To which the first retorts: "How do you, not being I, know that I do not know wherein consists the pleasure of fishes?" My position is that of the second philosopher. If other philosophers know "wherein consists the pleasure of fishes," I congratulate them; but I am not thus gifted.

When I try to disentangle the primitive from the inferred elements in what I take to be my knowledge, I find that the task is not really very difficult, except in certain niceties. The primitive part seems something like this: There are coloured shapes which move, there are noises, smells, bodily sensations, the experiences which we describe as those of touch, and so on. There are relations among these items: time-relations (earlier and later) among all of them, and space-relations (up-and-down, right-and-left, and the relations by which localization in the body is effected) among many of them. There are recollections of some of these things; this seems indubitable, although it is not easy to say in what a recollection consists, or how it is related to what it recollects. There are also expectations; by this I mean something just as immediate as memory. Everyone knows the story of the Orangeman who fell off a scaffolding and murmured as he fell: "To Hell with the Pope, and now for the—bump." He was experiencing expectation in the sense in which I mean it. Of thoughts other than memories and expectations, it is not necessary to take account when our sole purpose is to reach the primitive basis of our knowledge of matter.

In the above account, I have omitted many things which I formerly "knew," and which, apparently, most other people "know." I have omitted "objects." In former days, my apparatus of non-inferential knowledge included tables and chairs and books and persons and the sun and moon and stars. I have come to regard these things as inferences. I do not mean that I inferred them formerly, or that other people do so now. I fully concede that I did not infer them. But now, as the result of an argument, I have become unable to accept the knowledge of them as valid knowledge, except in so far as it can be inferred from such knowledge as I still consider epistemologically primitive.

The argument in question would naturally, but not validly, express itself in terms of the causal theory of perception. What I see—so it might be urged—is causally dependent upon the light waves that reach my eye, and these waves might be reflected or refracted in such a way as to deceive me concerning their source. This way of stating the argument is invalid because it assumes more knowledge of the physical world than we have any right to assume at our present level. But the facts upon which it relies can be easily made available, without any undue assumption of knowledge, for the purpose of proving our conclusion. In certain cases in which we seem to have immediate knowledge of objects, we find ourselves surprised by something totally unexpected. The dog listening to "his master's voice" on the gramophone may serve as an illustration. He thinks he perceives his master, but in fact he only perceives a noise. In restaurants which wish to look larger than they are, one whole wall sometimes consists of looking-glass, and it is easy to suppose that one perceives diners at tables, when in fact they are mere reflections. Perspective can be made to deceive. When I say "deceive," in this connection, I mean "rouse expectations which are not fulfilled." It is useless to multiply examples. The upshot is that what seems like perception of an object is really perception of certain sensible qualities together with expectations of other sensible qualities—the commonest case being something visual which rouses tactual expectations. It is found that the occasional deceptive experiences are not, in themselves, distinguishable from those that are not deceptive. Hence we conclude that we have to do with a correlation which is usual but not invariable, and that, if we wish to construct an exact science, we must be sceptical of the associations which experience has led us to form, connecting sensible qualities with others with which they are often but not always combined.

The above argument is based upon principles which common sense can be brought to accept, and has a conclusion which physics has accepted, though perhaps without fully realizing its scope. The argument is not "philosophical," in the sense of coming from a region quite different from that of science and ordinary knowledge. It proceeds merely on the usual principle of trying to substitute something more accurate for a belief which has been found to lead to error on occasion. It has as a consequence that "matter," in physics and in philosophy, if legitimate at all, cannot be altogether identified with the common-sense notion of a material object, though it will have a certain connection with this notion, since the common-sense belief in material objects does not usually lead to false expectations.

Some misunderstandings must be guarded against as regards expectation and error. Neither of these is primarily intellectual; I should be inclined to say that both are primarily muscular—or, we may say, nervous, in order not to seem paradoxical. Suppose you set to work to lift a watering-can: you may adjust your muscles in the way appropriate if the can is full, or in the way appropriate if it is empty. If they are adjusted to a full can when the can is empty, you receive a shock of surprise on experiencing the lightness of the can. You would describe your experience by saying, "I thought the can was full of water." But as a rule, in such situations, there has not been anything that could be called "thought"; there has been physiological adjustment as a result of a stimulus. Of course there may have been "thought"; and whatever "thought" may be, it certainly can produce the kind of muscular effects which we are considering. But these effects can be produced more directly, and usually are. There is so little essential difference between a process involving "thought" and one not involving it that it seems a mistake to confine the notions of truth and error to intellectual processes; they ought rather, it seems to me, to be applied to the complete reaction of a person to a situation, in which "thought" is only one element. But it will not do, at our present level, to introduce physiology, since we are considering how we know about matter, and must not therefore assume that we already know about the matter in our own body. However, the phenomena are easily described in the way which our problem demands. In the case of the watering-can, the vivid part of the experience is the surprise. But by means of attention a number of other elements can be observed. We can observe the feelings which are interpreted as meaning muscular adjustment to a heavy load; we can observe the visual appearance described as the can coming up with a jerk; we can observe the sudden change in what, for short, we may call muscular feelings. It is impossible to describe all this without circumlocution, since the natural words to use presuppose physiology; but it is clear that there is a great deal that can be directly observed, without invoking any theory. In such a process, what comes earlier may be described as "error" because of the emotion of surprise which follows. Where the activity which has been begun runs its course without leading to this emotion, we shall say that there is not error. I hesitate to ascribe "truth" to something pre-intellectual, but at any rate we may say that there is "correctness," or that what has succeeded to the sensation (or perception) which came at the beginning of the process has been "correct." We may shorten this by saying that the response to a stimulus may be "correct" or "erroneous." But the longer phrase has the merit of not assuming so much knowledge of causal relations.

In the situations to which the above analysis applies, we have the advantage of a perfectly definite criterion of correctness or error. The feeling of surprise marks error, and the absence of this feeling marks correctness. It must not be supposed that we have normally an explicit prevision, still less an explicit inference; all that can be said is that we are in such a condition that one sort of event will cause surprise while another sort will not. Consider the experience we have all had, of "thinking" we were at the bottom of a staircase when in fact there was another step to go down. In such a case, when we "think" we are at the bottom, we do not think at all, for if we did we should not make such a silly mistake. Indeed, we might say (or an Irishman might): "I thought I was at the bottom because I wasn't thinking."

It is fairly clear that all our elementary intellectual processes have pre-intellectual analogues. The analogue of a general causal belief is a reflex or a habit. A dog goes to the dining-room when he hears the dinner-bell, and so do we. In the case of the dog, it is easy to suppose that he has merely acquired a habit, without having formulated the induction: "Dinner-bells are a cause, or an effect, or an indispensable part of the cause, of dinner." We, however, can formulate this induction, and we shall then suppose that it is because we have done so that we go into the dining-room when we hear the bell. In fact, however, we may be just as merely habitual as the dog. The elementary inductions of common sense are first habits, and only subsequently beliefs. We may say that if, in our experience, is accompanied by either often or in some emotionally important manner, this fact causes first a habit which would be rational if were always accompanied by , and then a belief that is always accompanied by —the latter being a rationalization of the pre-existing habit.

General propositions may thus form part of our thinking from the start. Such general propositions are merely the verbal expression of habits. The hand-eye co-ordination becomes firmly fixed as a motor habit, and then, when we think, we conclude that what can be seen can often be touched—in fact, that it can be touched in circumstances which we know in practice, though we might have difficulty in formulating them exactly. Such general propositions are synthetic, and are in a certain sense a priori; for, though experience has caused them, they are not obtained by inference from other propositions, but by rationalizing and verbalizing our habits; that is to say, their antecedents are pre-intellectual. The trouble with them is that they are never quite right. Common sense, do what it will, cannot avoid being surprised occasionally. The object of science is to spare it this emotion, and create mental habits which shall be in such close accord with the habits of the world as to secure that nothing shall be unexpected. Science has, of course, not yet achieved its ideal: the Great War and the earthquake of Tokyo took people by surprise. But it is hoped that in time such events will no longer disturb us, because we shall have expected them. However, I do not wish at this stage to consider our knowledge of general propositions; it is particular matters of fact that concern us at present.

Although, in our less intellectual moods, we act as the result of a sensation without stopping to think (e.g. when we blink because we see something approaching the eye), yet we can, when we choose, react to a stimulus in the way which is called "knowing" it, and we often react involuntarily in this way. It is not necessary, in an analysis of matter, to decide what "knowing" is; it is only necessary to decide what is known, in so far as this is relevant to our knowledge of physics. The list which I gave earlier in the present chapter was designed to be such as would exclude the risk of error, using "error" in the sense which I have been defining. Common sense is liable to err—of this we have already given instances. We cannot therefore include the common-sense notion of an "object" or "thing" as part of what we know. But the sensible qualities which can be analyzed out of the "thing" can be admitted without ever leading us into error. These, therefore, are to be accepted as genuinely known.

It is a remarkable fact that all such knowledge, when not inferential, arises at about the same time as what is known, though it may survive for an indefinite time in the form of memory. This is the essential peculiarity, which we mentioned earlier, that distinguishes the empirical premisses of empirical knowledge. These consist of facts which become known spontaneously at about the time when they occur, and cannot be known sooner except by elaborate and more or less doubtful inferences from other such facts. The process of getting to know such facts without inference is called "perception," and knowledge derived wholly or partly from perception is said to be based on experience. A Greek could know the multiplication table as well as we do, but he could not know the biography of Napoleon.