Bosanquet describes it as follows:

If I say, pointing to a particular house, "That is my home," it is clear that in this act of judgment the reference conveyed by the demonstrative is indispensable. The significant idea "my home" is affirmed, not of any other general significant idea in my mind, but of something which is rendered unique by being present to me in perception. In making the judgment, "That is my home," I extend the present sense-perception of a house in a certain landscape by attaching to it the ideal content or meaning of "home;" and moreover, in doing this, I p.r.o.nounce the ideal content to be, so to speak, of one and the same tissue with what I have before me in my actual perception. That is to say, I affirm the meaning of the idea, or the idea considered as a meaning, to be a real quality of that which I perceive in my perception.

The same account holds good of every perceptive judgment; when I see a white substance on a plate and judge that "it is bread" I affirm the reference, or general meaning which const.i.tutes the symbolic idea "bread" in my mind, to be a real quality of the spot or point in present perception which I attempt to designate by the demonstrative "this." The act defines the given but indefinite real by affirmation of a quality, and affirms reality of the definite quality by attaching it to the previously undefined real.

Reality is given for me _in_ present sensuous perception, and _in_ the immediate feeling of my own sentient existence that goes with it. (Pp. 76, 77.)

Again, he says that the general features of the judgment of perception are as follows:

There is a presence of a something in contact with our sensitive self, which, as being so in contact, has the character of reality; and there is the qualification of this reality by the reference to it of some meaning _such as can_ be symbolized by a name (p. 77).

Our point of contact with reality, the place where reality gets into the thought-process, is, according to this view, to be found in the simplest, most indefinite type of judgment of perception. We meet with reality in the mere undefined "this" of primitive experience. But each such elementary judgment about an undefined "this" is an isolated bit of experience. Each "this" could give us only a detached bit of reality at best, and the further problem now confronts us of how we ever succeed in piecing our detached bits of reality together to form a real world.

Bosanquet's explanation is, in his words, this:

The real world, as a definite organized system, is _for me_ an extension of this present sensation and self-feeling by means of judgment, and it is the essence of judgment to effect and sustain such an extension (p. 77).

Again he says:

The subject in every judgment of Perception is some given spot or point in sensuous contact with the percipient self. But, as all reality is continuous, the subject is not _merely_ this given spot or point. It is impossible to confine the real world within this or that presentation. Every definition or qualification of a point in present perception is affirmed of the real world which is continuous with present perception. The ultimate subject of the perceptive judgment is the real world as a whole, and it is of this that, in judging, we affirm the qualities or characteristics.

(P. 78.)

The problem is the same as that with which Bradley struggles in his treatment of the subject of the judgment, and the solution is also the same. Bradley's treatment of the point is perhaps somewhat more explicit. Like Bosanquet, he starts with the proposition that the subject of the judgment must be reality itself and not an idea, because, if it were the latter, judgment could never give us anything but a union of ideas, and a union of ideas remains forever universal and hypothetical. It can never acquire the uniqueness, the singularity, which is necessary to make it refer to the real. Uniqueness can be found only in our contact with the real. But just where does our contact with the real occur? Bradley recognizes the fact that it cannot be the _content_--even in the case of a simple sensation--which gives us reality. The content of a sensation is a thing which is in my consciousness, and which has the form which it presents because it is in my consciousness. Reality is precisely something which is not itself sensation, and cannot be in my consciousness. If I say, "This is white,"

the "this" has a content which is a sensation of whiteness. But the sensation of whiteness is not reality. The experience brings with it an a.s.surance of reality, not because its content is the real, but because it is "my direct encounter in sensible presentation with the real world."[46] To make the matter clearer, Bradley draws a distinction between the _this_ and the _thisness_. In every experience, however simple, there is a content--a "thisness"--which is not itself unique.

Considered merely as content, it is applicable to an indefinite number of existences; in other words, it is an idea. But there is also in every experience a "this" which is unique, but which is not a content. It is a mere sign of existence which gives the experience uniqueness, but nothing else. The "thisness" falls on the side of the content, and the "this" on the side of existence. It is exactly the distinction which Bosanquet has in mind in the pa.s.sages quoted in which he tells us that "reality is given for me _in_ present sensuous perception, and _in_ the immediate feeling of my own sentient existence which goes with it;" and again when he says: "There is a presence of a something in contact with our sensitive self, which, as being so in contact, has the character of reality." The same point is made somewhat more explicitly in his introduction when he says that the individual's present perception is not, indeed, reality as such, but is his present point of contact with reality as such (p. 3).

But has this distinction between the content of an experience and its existence solved the problem of how we _know_ reality? When Bosanquet talks of knowing reality, he means possessing ideas which are an accurate reproduction of reality. It is still far from clear how, according to his own account, we could ever have any a.s.surance that our ideas do represent reality accurately, if we can nowhere find a point at which the content of an experience can be held to give us reality.

The case is still worse when we go beyond the problem of how any particular bit of reality can be known, and ask ourselves how reality as a whole can be known. The explanation offered by both Bradley and Bosanquet is that by means of judgment we extend the bit of reality of whose existence we get a glimpse through a peep-hole in the curtain of sensuous perception, and thus build up the organized system of reality.

In a pa.s.sage previously quoted, Bosanquet tells us that all reality is continuous, and therefore the real subject of a judgment cannot be the mere spot or point which is given in sensuous perception, but must be the real world as a whole. But how does he know that reality is continuous, and that the real world is an organized system? Our only knowledge of reality comes through judgment, and judgment brings us into contact with reality only at isolated points. When he tells us that reality is a continuous whole, he does so on the basis of a metaphysical presupposition which is not justifiable by his theory of the judgment.

The only statement about reality which could be maintained on the basis of his theory is that some sort of a reality exists, but the theory furnishes equal justification for the a.s.surance that this reality is of such a nature that we can never know anything more about it than the bare fact of its existence. Moreover, the bare fact of the existence of reality comes to us merely in the form of a feeling of our own sentient existence which goes with sense-perception. But the mere a.s.surance that somewhere behind the curtain of sensuous perception reality exists (even if this could go unchallenged), accompanied by the certainty that we can never by any possibility know anything more about it, is practically equivalent to the denial of the possibility of knowledge.[47]

Although the denial of the possibility of knowledge seems to be the logical outcome of the premises, it is not the conclusion reached by Bosanquet. At the outset of his treatise, Bosanquet propounds the fundamental question we have been considering in these words: "How does the a.n.a.lysis of knowledge as a systematic function, or system of functions, explain that relationship in which truth appears to consist, between the human intelligence on the one hand, and fact or reality on the other?" His answer is: "To this difficulty there is only one reply.

If the object-matter of reality lay genuinely outside the system of thought, not only our a.n.a.lysis, but thought itself, would be unable to lay hold of reality." (Pp. 2, 3.) The statement is an explicit recognition of the impossibility of bridging the chasm between a reality outside the content of knowledge and a known real world. It brings before us the dilemma contained in Bosanquet's treatment of the subject of the judgment. On the one hand the subject of the judgment must be outside the realm of my thoughts. If it were not, judgment would merely establish a relation between my ideas and would give me no knowledge of the real world. On the other hand, the subject of the judgment must be within the realm of my thoughts. If it were not, I could never a.s.sert anything of it; could never judge, or know it. The stress he lays on the first horn of the dilemma has been shown. It remains to show his recognition of the second horn, and to find out whether or not he discovers any real reconciliation between the two.

Bosanquet sums up the section of the introduction on knowledge and its content, truth, with the following paragraph:

The real world for every individual is thus emphatically _his_ world; an extension and determination of his present perception, which perception is to him not indeed reality as such, but his point of contact with reality as such. Thus in the enquiry which will have to be undertaken as to the logical subject of the judgment, we shall find that the subject, however it may shift, contract, and expand, is always in the last resort some greater or smaller element of this determinate reality, which the individual has constructed by identifying significant ideas with that world of which he has a.s.surance through his own perceptive experience.

In a.n.a.lyzing common judgment it is ultimately one to say that _I judge_ and that _the real world for me, my real world, extends itself_, or maintains its organized extension. This is the ultimate connection by which the distinction of subject and predication is involved in the act of affirmation or enunciation which is the differentia of judgment. (Pp. 3, 4).

Here the subject of the judgment appears as an element of a reality _which the individual has constructed_ by identifying significant ideas with that world of which he has a.s.surance through his own perceptive experience. But the very point with reference to the subject of the judgment previously emphasized is that it is not and cannot be something which the individual has constructed. The subject of the judgment must be reality, and reality does not consist of ideas, even if it be determined by them. It does not mend matters to explain that the individual has constructed his real world by identifying significant ideas with that world of which he has a.s.surance through his own perceptive experiences, because, as we have seen, "the individual's perceptive experiences" either turn out to be merely similar mental constructions made at a prior time, so that nothing is gained by attaching to them, or else they mean once more the mere shock of contact which is supposed to give a.s.surance that some sort of reality exists, but which gives no a.s.surance of what it is. That and what, this and thisness still remain detached. When he talks of _the real world for any individual_ we are left entirely in the dark as to what the relation between _the real world as it is for any individual_ and _the real world as it is for itself_ may be, or how the individual is to gain any a.s.surance that _the real world as it is for him_ represents _the real world as it is for itself_.

Another attempt at a reconciliation of these opposing views leaves us no better satisfied. The pa.s.sage is as follows:

The real world, as a definite organized system, is _for me_ an extension of this present sensation and self-feeling by means of judgment, and it is the essence of judgment to effect and sustain such an extension. It makes no essential difference whether the ideas whose content is p.r.o.nounced to be an attribute of reality appear to fall within what is given in perception, or not. We shall find hereafter that it is vain to attempt to lay down boundaries between the given and its extension. The moment we try to do this we are on the wrong track. The given and its extension differ not absolutely but relatively; they are continuous with each other, and the metaphor by which we speak of an extension conceals from us that the so-called "given" is no less artificial than that by which it is extended. It is the character and quality of being directly in contact with sense-perception, not any fixed datum of content, that forms the constantly shifting center of the individual's real world, and spreads from that center over every extension which the system of reality receives from judgment. (P.

77.)

In this pa.s.sage by the "given" he evidently means the content of sensory experience, the thisness, the what. It is, as he says, of the same stuff as that by which it is extended. Both the given and that by which it is extended are artificial in the sense of not being _real_ according to Bosanquet's interpretation of reality; they are ideas. But if all this is admitted, what becomes of the possibility of knowledge? Bosanquet undertakes to rescue it by a.s.suring us again that it is the character and quality of being directly in contact with sense-perception, not any fixed datum of content, that forms the center of the individual's real world and gives the stamp of reality to his otherwise ideal extension of this center. Here again we find ourselves with no evidence that the _content_ of our knowledge bears any relation to reality. We have merely the feeling of vividness attached to sensory experience which seems to bring us the certainty that there is some sort of a reality behind it, but this is not to give a.s.surance that our ideal content even belongs rightfully to that against which we have b.u.mped, much less of _how_ it belongs--and only this deserves the t.i.tle "knowledge."

In the chapter on "Quality and Comparison," in which he takes up the more detailed treatment of the simplest types of judgment of perception, he comes back to the same contradiction, and again attempts to explain how both horns of his dilemma must be true. The pa.s.sage is this:

The Reality to which we ascribe the predicate is undoubtedly self-existent; it is not _merely_ in my mind or in my act of judgment; if it were, the judgment would only be a game with my ideas. It is well to make this clear in the case before us, for in the later forms of the judgment it will be much disguised. Still the reality which attracts my concentrated attention is also within my act of judgment; it is not even the whole reality present to my perception; still less of course the whole self-existent Reality which I dimly presuppose. The immediate subject of the judgment is a mere aspect, too indefinite to be described by explicit ideas except in as far as the qualitative predication imposes a first specification upon it. _This_ Reality _is_ in my judgment; it is the point at which the actual world impinges upon my consciousness as real, and it is only by judging with reference to this point that I can refer the ideal content before my mind to the whole of reality which I at once believe to exist, and am attempting to construct. The Subject is both in and out of the Judgment, as Reality is both in and out of my consciousness. (Pp. 113, 114.)

The conclusion he reaches is a mere restatement of the difficulty. The problem he is trying to solve is how the subject _can_ be both in and out of the judgment, and how the subject without is related to the subject within. The mere a.s.sertion that it is so does not help us to understand it. His procedure seems like taking advantage of two meanings of sense-perception, its conscious quality and its brute abrupt immediacy, and then utilizing this ambiguity to solve a problem which grows out of the conception of judgment as a reference of idea to reality.

Turning from his treatment of the world of fact to his discussion of the world of idea, from the subject to the predicate, as it appears in his theory of the judgment we find again a paradox which must be recognized and cannot be obviated. An idea is essentially a meaning. It is not a particular existence whose essence is uniqueness as is the case with the subject of the judgment, but is a meaning whose importance is that it may apply to an indefinite number of unique existences. Its characteristic is universality. And yet an idea regarded as a psychical existence, an idea as a content in my mind, is just as particular and unique as any other existence. How, then, does it obtain its characteristic of universality? Bosanquet's answer is that it must be universal by means of a reference to something other than itself. Its meaning resides, not in its existence as a psychical image, but in its reference to something beyond itself. Now, any idea that is affirmed is referred to reality, but do ideas exist which are not being affirmed? If so, their reference cannot be to reality. Bosanquet discusses the question in the second section of his introduction as follows:

It is not easy to deny that there is a world of ideas or of meanings, which simply consists in that identical reference of symbols by which mutual understanding between rational beings is made possible. A _mere_ suggestion, a _mere_ question, a _mere_ negation, seem all of them to imply that we sometimes _entertain_ ideas without affirming them of reality, and therefore without affirming their reference to be a reference to something real or their meaning to be fact. We may be puzzled indeed to say what an idea can mean, or to what it can refer, if it does not mean or refer to something real--to some element in the fabric continuously sustained by the judgment which is our consciousness.

On the other hand, it would be shirking a difficulty to neglect the consideration that an idea, while denied of reality, may nevertheless, or even must, possess an identical and so intelligible reference--a symbolic value--for the rational beings who deny it. A reference, it may be argued, must be a reference to something. But it seems as if in this case the something were the fact of reference itself, the rational convention between intelligent beings, or rather the world which has existence, whether for one rational being or for many, merely as contained in and sustained by such intellectual reference.

I only adduce these considerations in order to explain that transitional conception of an objective world or world of meanings, distinct from the real world or world of facts, with which it is impossible wholly to dispense in an account of thought starting from the individual subject. The paradox is that the real world or world of fact thus seems for us to fall within and be included in the objective world or world of meanings, as if all that is fact were meaning, but not every meaning were fact. This results in the contradiction that something is objective, which is not real. (Pp. 4, 5.)

In the seventh section of the introduction Bosanquet explains his meaning further by what the reader is privileged to regard as a flight of the imagination--a mere simile--which he thinks may, nevertheless, make the matter clearer.

We might try to think that the world, _as known to each of us_, is constructed and sustained by his individual consciousness; and that every other individual also frames for himself, and sustains by the action of his intelligence, the world in which he in particular lives and moves. Of course such a construction is to be taken as a reconstruction, a construction by way of knowledge only; but for our present purpose this is indifferent. Thus we might think of the ideas and objects of our private world rather as corresponding to than as from the beginning identical with those which our fellow-men are occupied in constructing each within his own sphere of consciousness. And the same would be true even of the objects and contents within our own world, in as far as an act or effort would be required to maintain them, of the same kind with that which was originally required to construct them.... Thus the paradox of reference would become clearer. We should understand that we refer to a correspondence by means of a content. We should soften down the contradiction of saying that a name to meet which we have and can get nothing but an idea, nevertheless does not stand for that idea but for something else.

We should be able to say that the name stands for those elements in the idea which correspond in all our separate worlds, and in our own world of yesterday and of today, considered as so corresponding. (Pp. 45, 46.)

According to this view, the idea obtains the universality which const.i.tutes it an idea by a sort of process of elimination. It is like a composite photograph. It selects only the common elements in a large number of particular existences, and thus succeeds in representing, or referring to, all the particular existences which have gone to make it up. But when we come to consider the bearing which this view of universality, or generalized significance, has on our estimate of the knowledge-process, we feel that it has not solved the problem for us. In the first place, the idea _in its existence_ is just as particular when regarded as made up of the common elements of many ideas as is any of the ideas whose elements are taken. A composite photograph is just as much a single photograph as any one of the photographs which are taken to compose it. The chasm between the particularity of the psychical image and the universality of its meaning is not bridged by regarding the content of the image as made up by eliminating unlike elements in a number of images. The stuff with which thought has to work is still nothing more than a particular psychical image, and the problem of what gives it its logical value as a general significance is still unsolved.

Nor does it seem possible to find anything in the _existence_ of the image which could account for its reference to something outside of itself. The _fact_ of reference itself becomes an ultimate mystery.[48]

But even waiving this difficulty, the judgment must still appear truncated, if it really totally disregard a part of its content--_i. e._, the particular existence of the image as part of the judging consciousness. The theory holds that the particular existence of the image has no logical value. It is only its meaning, or general reference, which has logical value. But the image _qua_ image is just as real as that to which it is supposed to refer. If the judgment really does ignore its existence, then it ignores a portion of the reality it attempts to represent, and stands self-confessed as a failure.[49] At still another point, ideas, as Bosanquet represents them, prove to be unsatisfactory tools to use in the work of building up reality. In Bosanquet's words: "The meaning tyrannizes over the psychical image in another respect. Besides crushing out of sight its particular and exclusive existence, it also crushes out part of its content" (p. 74).

The idea, as we use it, is not, as to content, a complete or accurate representation of anything real. To take Bosanquet's ill.u.s.tration:

Some one speaks to me of the aegean sea, which I have never seen.

He tells me that it is a deep blue sea under a cloudless sky, studded with rocky islands. The meanings of these words are a problem set to my thought. I have to meet him in the world of objective references, which as intelligent beings we have in common. How I do this is my own affair, and the precise images at my command will vary from day to day, and from minute to minute.

It sounds simple to say that I combine my recollections of sea and sky at Torbay with those of the island-studded waters of Orkney or the Hebrides. Even so, there is much to adjust and to neglect; the red cliffs of Torbay, and the cloudy skies of the north. But then again, my recollections are already themselves symbolic ideas; the reference to Torbay or the Hebrides is itself a problem set to thought, and puts me upon the selection of index-elements in fugitive images that are never twice the same. I have _first_ to symbolize the color of Torbay, using for the purpose any blue that I can call to mind, and fixing, correcting, subtracting from, the color so recalled, till I reduce it to a mere index quality; and _then_ I have to deal in the same way with the meaning or significant idea so obtained, clipping and adjusting the qualities of Torbay till it seems to serve as a symbol of the aegean. (Pp.

74, 75.)

And by the time all this is performed what sort of a representation of reality is the idea? Evidently a very poor and meager and fragmentary one.

It is so poor and fragmentary, that it cannot itself be that which is affirmed of reality. It must be some other fuller existence to be found in the world of meanings which is affirmed. And yet how the meager content of the idea succeeds in referring to the world of meanings, and acting as the instrument for referring a meaning to reality, is not at all clear. It seems impossible to explain reference intelligibly by the concept of a _correspondence_ of contents.

The fundamental difficulty in the interpretation of the predicate is the same one that we encountered in the interpretation of the subject. If the predicate is to be affirmed of reality (and if it be not, it has no logical value), then it must, when affirmed, be in some sense an accurate representation of reality. But the predicate is an idea, and, moreover, an idea which is, both in its existence and in its meaning palpably the outcome of transformations wrought upon given sensory contents by the individual consciousness. Since the one point of contact with reality is in sensory experience, the more simple sensory experiences are reacted upon and worked over, the farther they recede from reality. The idea seems, therefore, in its very essence, a thing which never can be affirmed of reality. As image it is itself a reality, but not affirmed; as meaning it is that reality (the image) manipulated for individual ends. Why suppose that by distorting reality we get it in shape to affirm _of_ reality? Moreover, the farther an idea is removed from immediate sensory experience--in other words, the more abstract it becomes--the less is the possibility of affirming it of reality. The final outcome of this point of view, if we adhere rigorously to its logic is that the more thinking we do, the less we know about the real world. Bosanquet avoids this conclusion by a pure act of faith. If knowledge is to be rescued, we must believe that the work done by consciousness upon the bits of reality given in sensory experience really does succeed in building up a knowledge of reality for us. As Bosanquet puts it: "The presentation of Reality, qualified by an ideal content, is one aspect of Subject and Predication; and my individual percipient consciousness determining itself by a symbolic idea is the other. That the latter is identified with the former follows from the claim of conscious thought that its nature is to know."[50] (P. 83.)

To sum up the situation, Bosanquet starts out with the a.s.sumption that by knowledge we must mean knowledge of a world entirely independent of our ideas. If we fail to make this a.s.sumption, knowledge becomes merely a relation between ideas. But its whole importance seems to us to rest on the conviction that it does give us knowledge of a world which is what it is quite independently of our ideas about it, and cannot in any sense be modified by what we think about it. What knowledge does is to give us a copy or representation of the real world, whose value depends on the accuracy of the representation. And yet when we examine any individual knowing consciousness, the subject which appears within the judgment is never some portion of the world which exists outside of the knowing consciousness, but always some portion of the world which exists within the knowing consciousness, and which is const.i.tuted by the knowledge process. The predicate which is affirmed of reality is constantly found to derive its meaning, its generalized significance, not from its correspondence with, or reference to, the real world outside of the knowing consciousness, but from reference to a world of meanings, which consists in a sort of convention among rational beings--a world whose existence is distinctly within the knowing consciousness and not outside of it.[51] Between the real world, as Bosanquet conceives it, and the world of knowledge, we find inserted on the side of the subject, the world _as known to each of us_, and on the side of the predicate, the _objective world of meanings_. Neither of these is the real world. Both of them are ideal, _i. e._, are constructions of the individual consciousness. We nowhere find any satisfactory explanation of how these ideal worlds are related to the real world. There is merely the a.s.sertion that we must believe that they represent the real world in order that we may believe that knowledge exists. But the fact remains that whenever we try to a.n.a.lyze and explain any particular judgment, what we find ourselves dealing with is always the world as it exists to us as subject, and the objective world of meanings as predicate. If we stop here, then knowledge turns out to be just what Bosanquet a.s.serted at the outset that it was _not, i. e._, a relation between ideas. When we demand a justification for going farther than this, we find none except the claim of conscious thought that its nature is to know--a claim whose justice we have no possible means of testing, and which would not, even if admitted, be of the slightest value in deciding which _particular_ judgment is true and which false.

Bosanquet's development of his subject has proved to be throughout the necessary logical outcome of the presuppositions with reference to reality from which he starts. The fundamental difficulty of erecting a theory of the knowledge-process upon such a basis is recognized by him at the start in a pa.s.sage already quoted: "If the object-matter of reality lay genuinely outside the system of thought, not only our a.n.a.lysis, but thought itself, would be unable to lay hold of reality"

(p. 2). But, in spite of this a.s.sertion, his fundamental conception of reality remains that of a system which does lie outside the thought-process. His theory is an attempt to reconcile the essentially irreconcilable views that reality is outside of the thought-process, and that it is inside of the thought-process, and he succeeds only by calling upon our faith that so it is.

If it be true, as it seems to him to be, that we are compelled to adhere to both of these views of reality, then surely there is no other outcome. It means, however, that we finally resign all hope of _knowing_ reality. We may _have faith_ in its existence, but we have no way of deciding what particular judgment has reality in it as it should have it, and what as it should not. All stand (and fall) on the same basis.

But does not Bosanquet himself point out a pathway which, if followed farther, would reach a more satisfactory view of the realm of knowledge?

He has shown us that the only sort of reality _we know_, or can know, is the reality which appears within our judgment-process--the reality as known to us. Would it not be possible to drop the presupposed reality outside of the judgment-process (with which judgment is endeavoring to make connections) and content ourselves with the sort of reality which appears within the judgment-process? In other words, may there not be a satisfactory view of reality which frankly recognizes its organic relation to the knowledge-process, without at the same time destroying its value as reality? Is it possible to admit that reality is in a sense const.i.tuted in the judgment without making it at the same time the figment of the individual imagination--"a game with ideas"?

Let us a.s.sume for the moment that the real difficulty with Mr.

Bosanquet's conception, the error that keeps him traveling in his hopeless circles, is the notion that truth is a matter of reference of ideas as such to reality as such, leading us to oscillate between the alternatives that either all ideas have such reference, and so are true, const.i.tute knowledge; or else none have such reference, and so are false; or else are mere ideas to which neither truth nor falsity can be attributed. Let us ask if truth is not rather some _specific_ relation within experience, something which characterizes one idea rather than another, so that our problem is not how an idea can refer to a reality beyond itself, but what are the marks by which we discriminate a true reference from a false one. Then let us ask for the criterion used in daily life and in science by which to test reality.

If we ask the philosophically unsophisticated individual why he believes that his house still exists when he is away from it and has no immediate evidence of the fact, he will tell you it is because he has found that he can go back to it time and again and see it and walk into it. It never fails him when he acts upon the a.s.sumption that it is there. He would never tell you that he believed in its existence when he was not experiencing it because his mental picture of his house stood for and represented accurately an object in the real world which was nevertheless of a different order of existence from his mental picture.

When you ask the physicist why he believes that the laws of motion are true, he will tell you that it is because he finds that bodies always do behave according to them. He can predict just what a body will do under given circ.u.mstances. He is never disappointed however long he takes it for granted that the laws of motion are true and that bodies behave according to them. The only thing that could make him question their truth would be to find some body which did not prove to behave in accordance with them. The criterion is the same in both cases. It is the practical criterion of what as a matter of fact will work. That which can safely be taken for granted as a basis for further action is regarded as real and true. It remains real so long, and only so long, as it continues to fulfil this condition. As soon as it ceases to do so, it ceases to be regarded as real. When a man finds that he can no longer obtain the accustomed experience of seeing and entering his house, he ceases to regard it as real. It has burned down, or been pulled down.

When a physicist finds that a body does not, as a matter of fact, behave as a given law leads him to expect it would behave, he ceases to regard the law as _true_.

The contrast between the nave view of the criterion of reality and the one we have just been discussing may be brought out by considering how we should have to interpret from each standpoint the constant succession of facts in the history of science which have ceased to be facts. For ill.u.s.tration take the former fact that the earth is flat. It ceased to be a fact, says the theory we have been reviewing, because further thought-constructions of the real world convinced us that there is no reality which the idea "flat-world" represents. The idea "round-world"