that of the first inventor of the machine. How infinitely less the intelligence of a brute who may happen to use a machine of the kind! Is the intelligence of a squirrel or white mouse which turns in its wheel-cage greater than that either of the child who purposely gets the wheel-cage to put its pet in, or that of the man who made the cage? Mr. Romanes must somehow see this, for he says, "In order that he should assign names, primitive man must first have had occasion to make his preconceptual statements about the objects, qualities, etc., the names of which afterwards grew out of these statements, or sentence-words." That is to say, he must have been an essentially intellectual person.

Mr. Romanes next considers † the value of these supposed earlier sentence-words. After stating his hypothesis about the genesis of such early words with the help of gesture-the sound having no meaning apart from the gesture--he says, "From these now wellestablished facts, [!] we may gain some additional light on . . . the extent to which primitive words were 'abstract' or 'concrete,' 'particular' or 'general,' and therefore, 'receptual' or 'conceptual.'" Here he censures Prof. Max Müller for proclaiming the truth that language proceeded from the abstract to the concrete, or, as Mr. Romanes phrases it, that human thought "sprang into being Minerva-like, already equipped with the divine inheritance of conceptual wisdom."

He blames the Professor for adopting, as he says, "the assumption that there can be no order of words which do not, by the mere fact of their existence, p. 335. § p. 336.

* p. 332.

† p. 334.

imply concepts." He tells us that the Professor "does not sufficiently recognize that there may be a power of bestowing names as signs, without the power of thinking these signs as names." Mr. Romanes thus implies that a name cannot denote a concept unless he who employs it adverts to the fact of its being a name. But a name signifies a concept, without any advertence on the part of the utterer of it to its conceptual nature, or to the fact that it is a name; nor is it less conceptual in essence because the utterer of it is at the time of his utterance and for some time afterwards unable from circumstances to advert to and recognize the fact that it is a name. Mr. Romanes gives,* as his case in point, the instance of a child of his who "on first beginning to speak had a generalized idea of similarity between all kinds of brightly shining objects, and therefore called them all by the one denotative name of 'star.' The astronomer has a general idea answering to his denominative name of 'star;' but this has been arrived at after a prolonged course of mental evolution, wherein conceptual analysis has been engaged in conceptual classification in many and various directions: it therefore represents the psychological antithesis of the generalized idea, which was due to the merely sensuous associations of preconceptual thought. Ideas, then, as general and generic severally occupy the very antipodes of Mind." This is really nonsense. The child's term "star," was in its way as good and true a "universal" as the term "star" of the greatest astronomer who ever lived or shall live. But the p. 336.

two terms, though identical in sound and appearance, denote two very different concepts or universals-as truly as the term "trumpeter" respectively stands for the two very distinct concepts-a man and a pigeon.

"No one," he says, "will maintain that the sentencewords of young children exhibit the highest elaborations of conceptual thought, on the ground that they present the highest degree of 'generality,' which it is possible for articulate sounds to express.' Indeed! we reply. We ourselves will maintain it, and stoutly, too, if Mr. Romanes considers the word "thing," as used by young children, to be a “sentence-word." Naturally he denies to early man what he thus denies to the child. Just as naturally we affirm that primitive man in a sentenceword, even if thought out only by the aid of gesture, may, nay, must have, attained to concepts of the very highest generality, though, of course, neither the child, the ancient man, nor the modern peasant, recognizes its nature and generality by a reflex mental act. We altogether, then, deny the distinction which Mr. Romanes seeks to establish between generic and general ideas, other than the distinction (which is profound indeed) between (1) general ideas and (2) psychical states which are no ideas at all, but the mere unconscious, consentient energies named by us "Sensuous Universals."

The next point urged by Mr. Romanes is the resemblance which he affirms to exist between the syntax of gesture-language, that of baby-talk, and what he therefore assumes to have been the mode of speech of primitive man. This we do not in the least care to p. 338.

contest. It shows how perfectly logical gesture-language may be, and therefore, we may infer, always was as soon as it existed at all.

He then endeavours to show that language was at first essentially sensuous (what he calls receptual), and not intellectual. Here we must distinguish: As we have said again and again, being rational animals, we must use bodily signs to denote our thoughts, and require to have our conceptions first aroused by the incidence of sense-impressions in groups and groups of groups. Every highest conception of ours depends on the recognition of preceding acts of conception, and these on the imagination of the sense-impressions which called them forth. Thus there is, and must be, a sensuous element accompanying every concept.* But this sensuous element is not the concept itself, since it exists beside, or rather, underlies the concept. Our earliest perceptions, though, of course, truly conceptual, contain concepts of a lowly order, called forth by sense cognitions. Nevertheless, the very highest universals, even that of "being," are latent in every one of them. Now, Mr. Romanes, believing as he does that the lower concepts are but sense cognitions with names to them, naturally declares † that the evolutionist would clearly "expect to find more or less well-marked traces, in the fundamental constitution of all languages, of what has been called 'fundamental metaphor '-by which is meant an intellectual extension of terms that originally were of no more than sensuous signification. And this," he adds, "is precisely what we † p. 343.

* See "On Truth," p. 88.

do find." But "what we do find" is exactly what our combined intellectual and corporeal nature would lead us to expect, and is absolutely fatal to the doctrine of the common nature of man and brute. As we have before said, the very existence of " metaphor" is proof positive of the intellectual nature and activity of the human mind. Had not the intellect the power of apprehending through sense, and expressing by sensible signs, things which are beyond sense, metaphor could not exist. Neither could it exist if thought arose from language and followed it, instead of the opposite.

It is precisely because speech is too narrow for thought, and because words are too few to convey the ideas of the mind, that metaphor exists. It is interesting to note that figurative, metaphorical language is natural to, and especially abundant amongst, various uncultured tribes. Mr. Romanes says, "The whole history of language, down to our own day, is full of examples of the reduction of physical terms and phrases to the expression of non-physical conceptions and relations." We say, not the "reduction," but the "elevation" of such terms; and how could such elevations take place if "names" preceded "thoughts"?

With truth does Mr. Romanes say that metaphor is universal, and he quotes Carlyle as making the just remark, “An unmetaphorical style you shall seek in vain, for is not your very attention a stretching to?" The sensuous element in language does not show that the earliest ideas were themselves sensuous, but rather the wonderful spontaneity of the human intellect, † pp. 343, 344.

See above, p. 233.