When we ask what the word "justified" can mean in this connection, it is not easy to find an answer. Within the sphere of the immediately known the meaning of the word is plain enough. When I have constructed in my imagination a certain image or complex of images embodying a belief as to matter of fact, I say the mental operation is justified when I can substitute for the idea the percept which it is supposed to represent, or can know indirectly that this might be done according to known laws of the appearance and disappearance of percepts. Thus I perceive the outside of a tree-trunk and form an idea of what lies under the bark. I have reason to know that by stripping off the bark I can substitute for the image I have formed the corresponding percept. And if I see at a distance a similar tree growing upon an inaccessible cliff, and form an image of what lies under its bark, I may still regard this as justified by the possibility of referring to cases in which a similar image, arising out of a similar experience, has been found to be justified. It is a legitimate inference that, if circumstances were somewhat different, the proper percept might take the place of this image too. It is evident, however, that the word "justified" cannot be used in this or any analogous sense in speaking of the relation not of an image to a percept, but of an image or a percept to a something that, by hypothesis, cannot itself enter into experience at all. What then can the word mean? It at first interests us to know that "some Snarks are Boojums," but our interest lapses when we discover that we have absolutely no mark by which we may know a Boojum from anything else.
But I must not be drawn into digressions. The points with which I am concerned are these:—First: When a man says he sees this tree or that house, he ordinarily speaks as if there were but the one object in his thought. If he distinguishes at all between an immediate and a mediate object, the language that he uses would not indicate that he does so. And even after men have entered into lengthy arguments for the purpose of marking this distinction, and insisting that things are not single as they seem to the unreflective, they still indulge in this peculiar use of language, which would imply either that they have forgotten for the time being their own distinction between the immediate and the mediate, or that they regard the two as the same in Sense I, and to be treated as one. Certainly, in their reasonings upon this subject, men who hold to the two kinds of objects do confound them with one another, and strengthen their faith in the two by this misconception, as I shall show later. We have here, then, what we may call a kind of sameness, or pseudo-sameness, which deserves investigation, and which one should be careful not to confound with sameness of other kinds. Whether the word same is commonly applied in the premises is indifferent to my purpose. In the remainder of the present section I will consider the relation between the mental representative and its assumed correlatives.
Second: If we are to accept not merely the world of objects immediately known, but also a world or two worlds corresponding to this, and yet distinct from it, we cannot be sure our list of samenesses is complete unless we traverse in our search for the different kinds all the spheres of being in which we believe, and of which we think we can have some knowledge. In the section following this one I will try to discover the kinds of sameness which a believer in "external" things may reasonably attribute to them within their own sphere. In this there is no question of the relation of something in one sphere to a correlative something in another.
For the first point. What is in a man's mind when he is thinking of his percept as having a "real" object corresponding to it, I have shown to be as follows: He has in mind an immediate object and a duplicate of this, not necessarily altogether like it, imagined as standing in a causal relation to it and represented by it. When it occurs to him that this imagined duplicate is itself an immediate object and not the "real" one, he does as much for it, and provides it with a similar duplicate. In every case, when he tries to think of an object immediately perceived as having a "real" correlate, he simply furnishes it with an imaginary double in this way. What else is he to do? He is trying to think of two objects; the "real" object cannot, it is said, be in the mind; he must then imagine it. If he is a Lockian he will have in mind the immediate object and two imaginary ones, one signifying the "real" object as a bundle of qualities, and the other, a highly vague one, picturing the "substance."
Now, since this is all that can be before the man's mind, any kind of sameness which concerns the percept and the "real" thing must mean to him some relation between the immediate object and the image or images of which I have spoken. When this is realized it is seen that we have here not a new kind of sameness, a distinct experience, but a kind already discussed. The relation between the immediate object and the images described is simply that between representative and thing represented. This I have already examined within the field of what is recognized as immediately known. Here, too, it would seem that we are in the field of the immediately known, since we have to do with percepts and ideas, but though these images are in this field, they are here, so to speak, under protest, and their framing is supposed to be justified[18] only by something assumed to be not in this field. When this something is thought of at all it is thought of in just the way I have described. This is what thinking it means. Nevertheless, this duplicate world is assumed to be a world apart, and for this reason I have considered the sameness of percepts and their corresponding "real" objects by itself. It gives us sameness in sense seventh.
In writing the foregoing I have had in mind chiefly the position of the Lockian. I need not consider at length that taken by the Kantian, for what I have said will, with little change, apply to it also. If I hold to a "noumenon" as corresponding to my "phenomenon," and yet deny to it all qualities whatsoever, I must, to retain any appearance of consistency, represent it in my mind in the very vaguest possible way. Nevertheless, I must represent it, or I am not thinking of it at all, and I must relate the phenomenon and this vague representation in the way described. A true consistency would, of course, make impossible the whole process, for it would make impossible the giving of any quality at all to the representation, and the putting any relation between it and the phenomenon. In so dark a night cats do not merely turn grey, they disappear. On the other hand, if one is too liberal with this "noumenon," it palpably ceases to be a "noumenon," and degenerates into something very like a "phenomenon." The illusion must not be lost. Both these conflicting tendencies may be well illustrated in Mr. Spencer's "Unknowable." If we really refuse to allow to the consciousness of it "any qualitative or quantitative expression whatever,"[19] our vague image wholly disappears and there is nothing left in our consciousness but the "phenomenon." While if we follow the "First Principles" in coaxing it back into existence by allowing it reality,[20] and causality,[21] and a freedom from limits,[22] and printing its name with a capital letter,[23] as though it were even better than other things—if we do all this there is danger of the convalescent's becoming too robust altogether. The problem has its parallel in the practical problem of paying wages:—one must not pay too little, or he loses his laborer; nor too much, or he loses his money. The thing is to find the happy mean which will keep an object of thought before a man's mind and yet not make him lose all appearance of consistency.
But in which ever of the ways mentioned a man thinks of "real" things, he does what I have described. And when he implies that the immediately known and the "real" are in some sense the same, as he does when he talks as if there were but the one object, or asserts that we do not know things as they are in themselves but only as they appear to us; he really uses the word same in the fifth sense that I have given. The fact that he is using it to indicate the relation between percepts and a certain class of ideas which he has come to regard as duplicates of his percepts does not make the use of the word a new one. Whatever may be the state of affairs outside of his consciousness, this is all that takes place within it; and the word same, used in this connection, can mean no more to him than I have said.
Sec. 9. To avoid needless prolixity I will class together and very briefly treat of the kinds of sameness which one may attribute to "external" things. It is not necessary to go at length into the discussion of these, for since the "external" world, as it is assumed by those who have faith in it, is, to the man thinking it, simply a more or less modified duplicate of the world of things immediately perceived; and since all ground for attributing to it any determinations at all must be found in that which is immediately perceived; we may naturally look to find in it nothing that we have not already found in this immediate world. How, indeed, could anything else get into it? We cannot have in mind what is by hypothesis out of mind. The "real" world is then, in the mind of the man who thinks it, a world of imagined objects, and the world of imagination depends for its material upon the world of sense. A little reflection will show that the kinds of sameness of these "realities" are only the kinds of sameness already discussed duplicated, and assumed to belong to a new world.
1. An "external" quality or group of qualities may be said to be the same with itself at any one instant. Here we have Sense I carried over into the field of imagined duplicates.
2. An "external" quality existing at one time may be said to be the same with an "external" quality existing at another time, to indicate that the two are similar. The same thing may be true of any group of qualities. Here we have Sense II.