that we can not think (that is, combine, associate) without mental objects to associate, and that every mental object is a portion of language. To assert this would not be a "revolution in philosophy," but we might properly call it a revolution in the science of language.
Let us now consider the formation of concepts or general notions upon which the author lays so much stress as supporting his theory. Professor Müller brings forward the doctrine of Locke, Berkeley, and Hume, that "a general is nothing but a particular idea annexed to a general term—that is, to a term from which a customary conjunction has a relation to many other particular ideas and readily recalls them in imagination." It can not be doubted that there is substantial truth in this statement, though it needs qualification, but it does not prove Professor Müller's point. There must, indeed, be a fundamentum in every general notion, a nucleus, a type, a symbol. When we have in our minds the general notion horse, we have some particular horse, either remembered or constructed in imagination from former experiences of horses. With this goes the cognition that there are other objects like this one. To elucidate, I may, perhaps, be permitted to quote from a former analysis of my own: "Whatever association brings up the concept evokes the cognition of one or a small plural number of individuals which are either remembered as wholes or constructed out of remembered parts, and with which is associated the idea that there is a number of objects not definitely recalled which are similar to the individuals before the mind in the particulars characterizing the concept. . . . When we think of man, we remember a particular man, or imagine one; or the mind runs over the representations of several men, after which it rests content with the idea of an indefinite number of men about the same as those ideally presented." Now, the office of a name in such a connection is to furnish a connecting link in thought between a present cognition (or experience generally) and past ones. When I see a moving object in the distance, and as it comes nearer I identify it, I doubtless think by saying to myself, "It is a man." But if I see a strange creature, the likeness of which I had never seen but once before, and which, so far as I am concerned, is nameless, when I observe the second, the first is recalled, and identification takes place. This is just as much thought as if there were the intervention of a name. Suppose I see a third creature, which, by representative association, I class with the other two. Common characters are noticed, and I begin the formation of a general notion. This is completely done by the mere association of any striking resemblance, as a horn, a spotted skin, a peculiar howl, an odor. Any one of these peculiarities may form the nucleus or mark which will recall the creature, and knowledge of it can be communicated to others by gesture, by a picture, or by a word. Thought consists in identification and discrimination in present and past experiences, and between
- "A System of Psychology," Chapter L, Longmans, 1884.