applications, such as impetus, onslaught, appetitus, etc.

A number of English words, such as petition, petulance, competition, repetition, pen, pinnacle, feather, and many more, can all be traced back, step by step, and letter by letter, to this one root PAT, which is supposed by a kind of natural selection to have survived out of a large number of sympathic sounds, and to have become in the end a phonetic type expressive of the general concept of rapid motion.

All this is no doubt again hypothetical, and how could it be otherwise? But I do not see that we can entirely reject a theory which derives certain roots, expressive of the movements of objects, in just the same manner which we have followed ourselves in explaining the formation of roots, expressive of our own subjective movements. In one of my lectures on the Science of Language I examined in great detail the immense progeny of the root MR, to grind, to crush. This root has been traced back to the sound which men utter while engaged together in the act of grinding or crushing. Now even in this case it is an open question whether these sounds are not to a certain extent imitations of the noises produced by our own acts, rather than of the sounds which we utter while occupied in grinding. At all events it would be difficult to say what exact share should be assigned to one or the other source. These imitations or sympathic sounds must at first have been very numerous, expressing such various shades of meaning as crashing, crunching, crushing, thrashing, smashing, cracking, creaking, rattling, clattering, mawling, marring, etc., till at last, after dropping all

The true

Speech.

that seemed too special, there remained the smooth and manageable Aryan root MAR, with the general meaning of grinding or reducing to small particles. It should be clearly understood, however, that men who while grinding or platting utter the almost involuntary sounds of Mar-mar, nativity of or Ve-ve, are not yet speaking, as little as a child or an animal which screams on being pinched can be said to be conversing. All that I contend for is that our own acts are the first and the only direct objects of knowledge. They are what we will and what we know, and while of objects we can know one side only, we know all that there is to be known of our acts. It is when we remind ourselves or others of these acts by means of the sounds which used to accompany them, the clamor concomitans, that we make our first step towards real language. That first step manifested itself most naturally in the mood which we call the Imperative, i. e. the reminder, whether addressed to ourselves or to others. In all this we do not postulate anything which the most pedantic psychologist could wish to disallow. We do not suppose that the primitive speakers, on wishing to convey a command, looked about to find a sign, which sign they discovered in the sounds uttered involuntarily by themselves and their fellow-labourers in the performance of the commonest acts of their daily life. No, there was nothing premeditated in the process which changed those sounds into signs. man wished to grind, the sound Mar returned naturally to his lips, and when he wished that others should help him in grinding, the same sound was uttered, only in a voice so loud that it should

When a

reach those for whom it was intended.

It was the effect produced by these shouts which left on the mind of the man who shouted and of his clansmen who obeyed, the new impression that these shouts were useful signs, i. e. a means of making others think as we think, know as we know, will as we will, do as we do. This is the true nativity of language, a very humble manger, it may seem, and yet perhaps the most momentous event in the whole history of mankind.

of Roots.

The difference between the theory which ascribes The growth the origin of roots to sounds uttered by men while engaged in common acts, and the other which sees in some at least of these sounds more or less conscious imitations of noises produced by these acts themselves, is really not so great as the upholders of the one or the other opinion imagine. What is important in both is the parallelism between the process of generalisation of single sensations and the production of abstract concepts on one side, and the process of mutual friction and smoothing down of the sounds accompanying these sensations, on the other. This double process might either go on till the highest generalisations are reached, or it might be stopped at certain points, so that some roots retained something of their sharper outline and became popular on that very account, while others, more general in form and meaning, were used most frequently, till their real origin was completely forgotten.

In this struggle between generalisation and specialisation many roots must have crossed each other, and the summum genus of going, moving, doing, being, etc. must have been reached from many

different starting-points. This would explain how even during what may be called the Radical Period small communities, after a very short separation, became mutually unintelligible. We could thus perfectly understand the natural growth, not only, as in later times, of different dialects of languages sharing the same capital in common, but likewise of different families of languages, such as the Aryan and Semitic, possessing each its own peculiar roots, and yet, it may be, proceeding in the beginning from a common centre. I say we should be able to understand it, but I feel bound to add, if we clearly apprehend the process of the generalisation and specialisation of the radical elements of human speech, we should likewise understand that, from the nature of the case, it would be impossible ever to prove it.

Roots which occur in Pânini's Dhâtupâtha, and which I should be quite willing to sur- Onomatopoie render as onomatopoeic, are: KAS, to Roots. cough, KÛG, to hum, KUŇG, to rustle, KRAKSH, to crash, KHARG, to creak, KSHU, to sneeze, KSHVID and KSHVID, to hum, GUNG, to hum, KUMB, to kiss, KULUMP, to suck, PRUTH and PROTH, to snort, MA, to bleat, RAT, to howl, RÂ, to bark, SHTHIV, to spue, HIKK, to sob, HESH, to whinny, HRESH, to neigh, etc. Some of them might, no doubt, be disputed, but not one of them is of any importance as helping us to account for real words in Sanskrit. Most of them have had no offspring at all, others have had a few descendants, mostly sterile. Their history shows us clearly, how far the influence of onomatopoeia may go, and if we once know its legitimate sphere, we

shall be less likely to wish to extend it beyond its proper limits.

Further

of Roots.

Having made these two admissions, not so much on account of their practical importance, for Modification they concern the most insignificant portion of language only, as because it seems right, whenever there is an opportunity, to yield frankly and openly to arguments which have some reason on their side, we have now to consider whether, with Noire's theory, we have really all we want for understanding the nature of roots and the growth of language. He has taught us to recognise in the conscious and creative social acts of men, as accompanied by various natural sounds, the true germs of concepts embodied in language. We have learnt how that consciousness of repeated self-willed acts becomes, to all intents and purposes, what we mean by an original concept, and how the phonetic sign inseparable from it is what we mean by a root-word. Even Hume could not maintain that the consciousness of the continuously repeated act of digging was but a singular impression, nor could Berkeley object that the mind had no idea of such continued acts. And as certain sounds naturally accompany every one of these acts and serve to recall it, we are not driven to the admission that, after the idea or the concept was formed, its phonetic body had to be fetched from elsewhere to give it security and shelter.

But though roots expressive of acts are no doubt the most primitive, the most numerous and important elements in the growth of human language and human thought, we must remember that in speaking we have to do more than to express our acts and what is connected with them. We