signs used by deaf and dumb people are by no means identical all over the globe 1.

Children again, long before they are able to speak, can imitate the acts of eating, drinking, riding on their father's knees, and thus express their little wishes; but a wish is not a thought, as little as fear and horror. If some philosophers like to call these states of feeling thought, they may do so at their own. peril, but they ought at all events to let us know, in order that others may be able to discount such license.

Images.

Some more serious philosophers put in a claim for images. Images, they say, such as our senses leave in our memory, may surely be called thought. They may, no doubt, if only we let others know that in our own philosophical dialect we use thinking in that extended sense. But it is surely better to distinguish and to keep the term imagination for signifying the play of our images. I myself hold it impossible that human beings should have real images without first having framed them in names; and among physiologists, Helmholtz denies the possibility of our having perceptions without names. But, of course, if careful observers, such as Mr. Galton, assure us that they have images without knowing what they are images of, and without remembering what they are called, we are bound to believe them, even though we cannot follow them. What they are anxious for is evidently to show that animals, though they have no language, have images, that they combine these images, and that their acts, their sensible, or, as they like to call

1 Mallery, Sign Language among the North-American Indians.

them, their rational acts, are determined by them. Let that be so, at least for argument's sake. But even then, is not this imagination or even this reasoning without language utterly different from imagination and reasoning with language? Suppose a dog, instead of coming to me, as one of my dogs did, expressing his uneasiness and then dragging me on to his rug which was red, and showing me that it was occupied by my other dog, who ought to have been on his own rug which was blue, looking at me reproachfully till I had ordered the other dog away, and then taking possession with all the pride of an injured innocent of his own red rug-suppose that dog, instead of wheedling and barking were suddenly to stand up on his hind legs and say to me, 'The other dog has taken my rug; please, Sir, order him away,' should we not almost go out of our mind?

Or let us place an infant and a grown-up man side by side, the one struggling and crying for a cup of milk, the other saying plainly, 'I should like that cup of milk.' Is not the distance between these two acts immeasurable, the one being merely the result of the direct or reflex action of our senses, the other the result of a growth that has gone on for thousands of years? The grown-up man also, if he were dying of thirst, might no doubt rush towards the cup and swallow it without saying a word, and we might call the expression of his impetuous features language, and his rushing movements reason. But we should gain nothing by the use of this metaphorical language. There are philosophers who tell us that an infant could not stretch out its arms without going through a silent syllogism: By stretching out our arms we

obtain what we wish for; I wish for this cup of milk, therefore I stretch out my arms.' It may It may be so, but we know nothing about it and never shall, till the infant is able to speak, that is to say, ceases to be an infant, and then tells us what it thinks.

Between the infant, however, and the man who is able to speak, there is not a distance of ten or twenty years only. The language which he has accepted is the result of intellectual labour carried on for thousands of years. The original framing of our words and thoughts is a process which no one but the geologist of language has even the most remote idea of, and to suppose that one human being could, in the space of ten or twenty years, have accumulated the wealth of his grammar and dictionary is like believing that the earth with its mountains and rivers could have been made in six days. It is extraordinary that the same argument, which has been answered ad nauseam, is brought forward again and again. It is quite true that the infant and the parrot are for a time without language, and that both learn to say after a time, 'How do you do?' But the child learns to speak human language, while the parrot never speaks Parrotese.

Involuntary and voluntary sounds.

The next step after what has been called the language of gestures, leads us on to involuntary and voluntary sounds. I call involuntary sounds interjections which have a direct natural origin, which express joy, fear, anger, admiration, assent or dissent. To us, accustomed to our own interjections, there seems a natural appropriateness in their sound, but here too a comparative study teaches us that it is not

so. No, for instance, does not always mean no; in Syrianian it means yes. Even in Irish we find for No, not only naicc, but also aice 1. Voluntary interjections I call such imitative sounds as bow wow for dog, and moo for cow. Here, too, we find that what seems to us perfectly natural and intelligible, is not always so. Whereas to our ears the dog says bow wow, he says kliff klaff to a German ear.

It is extremely difficult to render inarticulate sounds by our alphabet. Many attempts have been made to write down the sounds uttered by birds, but hitherto with small success. A great phonetician, well acquainted with the latest theories of physiological phonetics, has spent many days and nights in watching the notes of the nightingale; and what do you think his rendering has come to? The real note of the nightingale, as reduced to alphabetical writing, is:

Dailidurei faledirannurei lidundei faledaritturei. You know that before languages were studied scientifically, it was a very general idea that all human speech arose in that way, and that the ultimate elements of our words were imitations of natural sounds or involuntary interjections. I called these theories the Bow-wow and Pooh-pooh theories. Some philosophers have lately added a third theory, which they call the Yo-heho theory, but which is really a subdivision only of the Pooh-pooh theory. By a most extraordinary mistake this theory has been ascribed to Noiré, who was really one of its most determined opponents. According to this theory language would have been derived directly from the

1 Zeuss, Grammatica Celtica. Yes in Old Irish is iss ed, est hoc,' or simply ed, Goth. ita.-Whitley Stokes,

=

cries uttered by people while engaged in pulling, rubbing, digging, rowing, and similar primitive occupations.

In this the supporters of this Yo-heho theory have, no doubt, touched on a very important phase in the growth of language and thought, as we shall see presently; but if they look upon sounds such as Yo-heho as mere interjections, they are still in the bitterness of the Pooh-pooh theory, that is to say, they have not even perceived the difficulty of the problem which they wish to solve.

The names Bow-wow, Pooh-pooh, and Yo-heho theories have sometimes been objected to as too homely, and as possibly offensive. But as these theories in their crude form are no longer held by any scholar, these names are really quite harmless, and they are certainly useful, because they tell their own tale. If we are afraid of them, we must use the cumbersome names of Mimetic, Onomatopoetic or Interjectional theory, every one of them requiring an elaborate commentary.

The Bow-wow, Pooh-pooh, and Yo-heho theories.

These three theories, however, were by no means so illogical as they seem to us now. They were no doubt a priori theories, but they had certain facts to support them. There are interjections in every language, and, by the general analogy of language, some of them have been raised into verbs and adjectives and substantives. Hush, for instance, the German husch, is an interjection which in German is used to drive away birds, to express any quick movement, to attract attention, while in English it is now chiefly