CHAPTER II.

THOUGHT AND LANGUAGE.

IF our analysis of the human mind is right, if all that we call thought finds its last consummation in language, the next question, namely how the growth of the human mind can be studied, is easily answered: it must be studied in the history of language.

Words the

signs of

concepts.

This conclusion, which after the discoveries of the Science of Language seems inevitable, might have been arrived at long ago. It has always been considered as one of the glories of Locke's philosophy that he established the fact that names are not the signs of things, but in their origin always the signs of concepts. It is true that Hobbes1 had already enunciated the same important truth, namely that words are signs of concepts and not of things. But that would in no way detract from Locke's merit, for truth is common property, and it is chiefly the use which a philosopher makes of any given truth which secures him his position in the history of philosophy. I know quite well that Mill considered this distinction between words as the signs of concepts,

1

Computation or Logic, chap. ii. See Mill's Logic, book i chap. ii.

and words as the signs of things as of little consequence, but this must depend altogether on the use which can be made of it. To my mind Locke's insistance on words being the signs of concepts, and not of things, is of the greatest importance for everything that is called philosophy. And even Mill, though he argues against this theory, frequently adopts it unawares. When Mill says1 that a word ought to be considered as the name of that which we intend to be understood,' this is clearly our concept, and not the thing apart from our concept. Mill admits that we know nothing of the inmost nature of fire and water, that heat is not like the steam of boiling water nor the feeling of cold like the east wind. But it is of these subjective sensations that our concepts are made up. Why then should Mill call it a capital error 'that the investigation of truth consists in contemplating and handling our ideas' or conceptions of things, instead of the things themselves? he who in the same chapter declares that a previous mental conception of facts is an indispensable condition' of all thought and belief? In another place Mill says, 'that in using a proper name we put a mark, not indeed upon the object itself, but, so to speak, upon the idea of the object.' But why 'so to speak,' considering that we can do nothing else? If we use a general name, if we say Dog, do we mean the thing, or our concept of it? Is there anything corresponding to Dog? Is not Dog, like every

other name, the name of a thing that cannot possibly Who ever saw saw We may see a

exist?

a dog?

spaniel, or a greyhound, or a dachs-hund, we may see a black or a white or a brown dog, but a dog no, human eye has ever seen. Therefore when we say dog, we can only mean our concept of a dog, that is, our concept of many or all dogs, and it is the name of that concept which we use to denote any single dog. The same with a tree. No one ever saw a tree, but only this or that fir tree, or oak tree, or apple tree; but then again, no one ever saw an apple tree, but only a few parts of it, a little of the bark, a few leaves, an apple here and there; and of all this again one side only. Tree, therefore, is a concept, and, as such, can never be seen or perceived by the senses, can never acquire any phenomenal or intuitional form. We live in two worlds, the world of sight, and the world of thought; and strange as it may sound, nothing that we think, nothing that we name, nothing that we find in our dictionary, can ever be heard, or seen, or felt. We can even have names for things which never existed, such as hobgoblins, also for things which exist no more or which exist not yet, such as the grapes of th last harvest and the grapes of the next. These can hardly be called things, as separate from our concepts of them, and our names in this case clearly refer either to what we have never seen, or what we see no longer or not yet, or at all events to what we have never seen in that form in which we conceive it. The mere fact that I call a thing past or future ought to be sufficient to show that it is my concept I am speaking of, and not the thing as independent of me.

If this is so with the names of concrete things, it must be so all the more with the names of attributes.

Attributes

always abstract.

All attributes, according to the schoolmen, are abstract terms, though it does not follow that all abstract terms are attributes. All attributes are abstract terms, because in forming them we must drop that of which the attributes are attributes. We see the white snow, the white milk, the white horse, but it is only by dropping all except the white colour (I speak here of artificial or scholastic abstraction) that we get white as an attribute. Even when we speak of a white thing, we are speaking of a concept we have made for ourselves, for our experience never offers us anything that is nothing but a white thing. So whether we use white as an attribute, or whether we speak of a white thing, we speak of concepts which we have made, and the words which we use of them are names of our concepts, not names of things.

As a matter of principle, therefore, the distinction between words as signs of concepts and words as signs of things cannot be said to be of little consequence. It forms the watershed between the two great streams of philosophical thought, the nominalistic and realistic, and if we adopted the view which Mill professes to embrace, it might seem to follow, without much difficulty, that idealistic philosophers have no right to use language at all.

But this does not concern us at present. What concerns us in this place is the practical importance which the view attacked by Mill has for our own special inquiry. For if we are right in holding the opposite view, if words are indeed the signs and outward embodiments of our concepts, the origin

and growth of concepts, that is, of all human thought, need no longer be studied as a mere theory, but becomes an historical study resting on facts, namely on the facts of language. Only while Locke still looked on words as arbitrary signs, the Science of Language has taught us this new lesson, that words are neither arbitrary nor necessary, but always reasonable and intelligible signs of concepts.

If we take, for instance, such a word as name, Locke would have said that it was a Name. sound arbitrarily chosen to signify what we call a thing.' Nor does the word name convey more to an ordinary speaker than this its traditional meaning, which he learns either from his parents or from a dictionary. The student might go a step further, and, by comparing other languages, he might learn that Latin, Greek, and Sanskrit contain very similar words to express the same idea, namely nomen in Latin, ovoua in Greek, nâman in Sanskrit. This would teach him that name in English and Name in German could at all events not have been chosen by his own Saxon ancestors, but must have existed a long time ago, when the Teutonic language had not yet branched off from Sanskrit, Greek, and Latin. The question then arises whether that word was chosen arbitrarily by the primitive framers of Aryan speech, or with an intelligent purpose. Placed under the microscope of the comparative grammarian, name, the Sanskrit nâman, is seen to consist of a root NÂ, originally GNA, to know, and of a suffix which generally expresses an instrument or an act. We thus perceive that name meant originally a great deal more than what we call a thing. It

G