Primitive Culture/Chapter 5

by Edward Burnett Tylor

 

CHAPTER V.

EMOTIONAL AND IMITATIVE LANGUAGE.

Element of directly expressive Sound in Language — Test by independent correspondence in distinct languages — Constituent processes of Language — Gesture — Expression of feature, &c. — Emotional Tone — Articulate sounds, vowels determined by musical quality and pitch, consonants — Emphasis and Accent — Phrase-melody, Recitative — Sound-Words — Interjections — Calls to Animals — Emotional Cries — Sense-Words formed from Interjections — Affirmative and Negative particles, &c.

In carrying on the enquiry into the development of culture, evidence of some weight is to be gained from an examination of Language. Comparing the grammars and dictionaries of races at various grades of civilization, it appears that, in the great art of speech, the educated man at this day substantially uses the method of the savage, only expanded and improved in the working out of details. It is true that the languages of the Tasmanian and the Chinese, of the Greenlander and the Greek, differ variously in structure; but this is a secondary difference, underlaid by a primary similarity in method, namely, the expression of ideas by articulate sounds habitually allotted to them. Now all languages are found on inspection to contain some articulate sounds of a directly natural and directly intelligible kind. These are sounds of interjectional or imitative character, which have their meaning not by inheritance from parents of adoption from foreigners, but by being taken up directly from the world of sound into the world of sense. Like pantomimic gestures, they are capable of conveying their meaning of themselves, without reference to the particular language they are used in connexion with. From the observation of these, there have arisen speculations as to the origin of language, treating such expressive sounds as the fundamental constituents of language in general, and considering those of them which are still plainly recognizable as having remained more or less in their original state, long courses of adaptation and variation having produced from such the great mass of words in all languages, in which no connexion between idea and sound can any longer be certainly made out. Thus grew up doctrines of a 'natural' origin of language, which, dating from classic times, were developed in the eighteenth century into a system by that powerful thinker, the President Charles de Brosses, and in our own time have been expanded and solidified by a school of philologers, among whom Mr. Hensleigh Wedgwood is the most prominent.^[1] These theories have no doubt been incautiously and fancifully worked. No wonder that students who found in nature real and direct sources of articulate speech, in interjectional sounds like ah! ugh! h'm! sh! and in imitative sounds like purr, whiz, tomtom, cuckoo, should have thought that the whole secret of language lay within their grasp, and that they had only to fit the keys thus found into one hold after another to open every lock. When a philosopher has a truth in his hands, he is apt to stretch it farther than it will bear. The magic umbrella must spread and spread till it becomes a tent wide enough to shelter the king's army. But it must be borne in mind that what criticism touches in these opinions is their exaggeration, not their reality. That interjections and imitative words are really taken up to some extent, be it small or large, into the very body and structure of language, no one denies. Such a denial, if anyone offered it, the advocates of the disputed theories might dispose of in the single phrase, that they would neither be pooh-poohed nor hooted down. It may be shown within the limits of the most strict and sober argument, that the theory of the origin of language in natural and directly expressive sounds does account for a considerable fraction of the existing copia verborum, while it raises a presumption that, could we trace the history of words more fully, it would account for far more.

In here examining interjectional and imitative sounds with their derivative words, as well as certain other parts of language of a more or less cognate character, I purpose to bring forward as far as possible new evidence derived from the languages of savage and barbarous races. By so doing it becomes practicable to use a check which in great measure stops the main source of uncertainty and error in such enquiries, the habit of etymologizing words off-hand from expressive sounds, by the unaided and often flighty fancy of a philologer. By simply enlarging the survey of language, the province of the imagination is brought within narrower limits. If several languages, which cannot be classed as distinctly of the same family, unite in expressing some notion by a particular sound which may fairly claim to be interjectional or imitative, their combined authority will go far to prove the claim a just one. For if it be objected that such words may have passed into the different languages from a common source, of which the trace is for the most part lost, this may be answered by the question, Why is there not a proportionate agreement between the languages in question throughout the far larger mass of words which cannot pretend to be direct sound-words? If several languages have independently chosen like words to express like meanings, then we may reasonably suppose that we are not deluding ourselves in thinking such words highly appropriate to their purpose. They are words which answered the conditions of original language, conforming as they do to the saying of Thomas Aquinas, that the names of things ought to agree with their natures, 'nomina debent naturis rerum congruere.' Applied in such comparison, the languages of the lower races contribute evidence of excellent quality to the problem. It will at the same time and by the same proofs appear, that savages possess in a high degree the faculty of uttering their minds directly in emotional tones and interjections, of going straight to nature to furnish themselves with imitative sounds, including reproductions of their own direct emotional utterances, as means of expression of ideas, and of introducing into their formal language words so produced. They have clearly thus far the means and power of producing language. In so far as the theories under consideration account for the original formation of language, they countenance the view that this formation took place among mankind in a savage state, and even, for anything appearing to the contrary, in a still lower stage of culture than has survived to our day.^[2]

The first step in such investigation is to gain a clear idea of the various elements of which spoken language is made up. These may be enumerated as gesture, expression of feature, emotional tone, emphasis, force, speed, &c. of utterance, musical rhythm and intonation, and the formation of the vowels and consonants which are the skeleton of articulate speech.

In the common intercourse of men, speech is habitually accompanied by gesture, the hands, head, and body aiding and illustrating the spoken phrase. So far as we can judge, the visible gesture and the audible word have been thus used in combination since times of most remote antiquity

1 Among the principal savage and barbaric languages here used for evidence, are as follows: — Africa: Galla (Tutschek, Gr. and Dic.), Yoruba (Bowen, Gr. and Dic.), Zulu (Döhne, Dic.). Polynesia, &c.: Maori (Kendall, Vocab., Williams, Dic.), Tonga (Mariner, Vocab.), Fiji (Hazlewood, Dic.), Melanesia (Gabelentz, Melan. Spr.). Australia (Grey, Moore, Schürmann, Oldfield, Vocabs.). N. America: Pima, Yakama, Clallam, Lummi, Chinuk, Mohawk, Micmac (Smithson. Contr. vol. iii.), Chinook Jargon (Gibbs, Dic.), Quiché (Brasseur, Gr. and Dic.). S. America: Tupi (Diaz, Dic.), Carib (Rochefort, Vocab.), Quichua (Markham, Gr. and Dic.), Chilian (Febres, Dic.), Brazilian tribes (Martius, ' Glossaria linguarum Brasiliensium'). Many details in Pott, 'Doppelung,' &c. in the history of our race. It seems, however, that in the daily intercourse of the lower races, gesture holds a much more important place than we are accustomed to see it fill, a position even encroaching on that which articulate speech holds among ourselves. Mr. Bonwick confirms by his experience Dr. Milligan's account of the Tasmanians as using 'signs to eke out the meaning of monosyllabic expressions, and to give force, precision, and character to vocal sounds.' Captain Wilson remarks on the use of gesticulation in modifying words in the Chinook Jargon. There is confirmation to Spix and Martius' description of low Brazilian tribes completing by signs the meaning of their scanty sentences, thus making the words 'wood-go' serve to say 'I will go into the wood,' by pointing the mouth like a snout in the direction meant. The Rev. J. L. Wilson, describing the Grebo language of West Africa, remarks that they have personal pronouns, but seldom use them in conversation, leaving it to gesture to determine whether a verb is to be taken in the first or second person; thus the words 'ni ne' will mean 'I do it,' or 'you do it,' according to the significant gestures of the speaker.^[3] Beside such instances, it will hereafter be noticed that the lower races, in counting, habitually use gesture-language for a purpose to which higher races apply word-language. To this prominent condition of gesture as a means of expression among rude tribes, and to the development of pantomime in public show and private intercourse among such peoples as the Neapolitans of our own day, the most extreme contrast may be found in England, where, whether for good or ill, suggestive pantomime is now reduced to so small a compass in social talk, and even in public oratory.

Changes of the bodily attitude, corresponding in their fine gradations with changes of the feelings, comprise conditions of the surface of the body, postures of the limbs, and also especially those expressive attitudes of the face to which our attention is particularly directed when we notice one another. The visible expression of the features is a symptom which displays the speaker's state of mind, his feelings of pleasure or disgust, of pride or humility, of faith or doubt, and so forth. Not that there is between the emotion and its bodily expression any originally intentional connexion. It is merely that a certain action of our physical machinery shows symptoms which we have learnt by experience to refer to a mental cause, as we judge by seeing a man sweat or limp that he is hot or footsore. Blushing is caused by certain emotions, and among Europeans it is a visible expression or symptom of them; not so among South American Indians, whose blushes, as Mr. David Forbes points out, may be detected by the hand or a thermometer, but being concealed by the dark skin cannot serve as a visible sign of feeling.^[4] By turning these natural processes to account, men contrive to a certain extent to put on particular physical expressions, frowning or smiling for instance, in order to simulate the emotions which would naturally produce such expressions, or merely to convey the thought of such emotions to others. Now it is well known to every one that physical expression by feature, &c., forming a part of the universal gesture-language, thus serves as an important adjunct to spoken language. It is not so obvious, but on examination will prove to be true, that such expression by feature itself acts as a formative power in vocal language. Expression of countenance has an action beyond that of mere visible gesture. The bodily attitude brought on by a particular state of mind affects the position of the organs of speech, both the internal larynx, &c., and the external features whose change can be watched by the mere looker-on. Even though the expression of the speaker's face may not be seen by the hearer, the effect of the whole bodily attitude of which it forms part is not thereby done away with. For on the position thus taken by the various organs concerned in speech, depends what I have here called 'emotional tone,' whereby the voice carries direct expression of the speaker's feeling.

The ascertaining of the precise physical mode in which certain attitudes of the internal and external face come to correspond to certain moods of mind, is a physiological problem as yet little understood; but the fact that particular expressions of face are accompanied by corresponding and dependent expressions of emotional tone, only requires an observer or a looking-glass to prove it. The laugh made with a solemn, contemptuous, or sarcastic face, is quite different from that which comes from a joyous one; the ah! oh! ho! hey! and so on, change their modulations to match the expression of countenance. The effect of the emotional tone does not even require fitness in the meaning of the spoken words, for nonsense or an unknown tongue may be made to convey, when spoken with expressive intonation, the feelings which are displayed upon the speaker's face. This expression may even be recognized in the dark by noticing the tone it gives forth, while the forced character given by the attempt to bring out a sound not matching even the outward play of the features can hardly be hidden by the most expert ventriloquist, and in such forcing, the sound perceptibly drags the face into the attitude that fits with it. The nature of communication by emotional tone seems to me to be somewhat on this wise. It does not appear that particular tones at all belong directly and of themselves to particular emotions, but that their action depends on the vocal organs of the speaker and hearer. Other animals, having vocal organs different from man's, have accordingly, as we know, a different code of emotional tones. An alteration in man's vocal organs would bring a corresponding alteration in the effect of tone in expressing feeling; the tone which to us expresses surprise or anger might come to express pleasure, and so forth. As it is, children learn by early experience that such and such a tone indicates such and such an emotion, and this they make out partly by finding themselves uttering such tones when their feelings have brought their faces to the appropriate attitudes, and partly by observing the expression of voice in others. At three or four years old they are to be seen in the act of acquiring this knowledge, turning round to look at the speaker's face and gesture to make sure of the meaning of the tone. But in later years this knowledge becomes so familiar that it is supposed to have been intuitive. Then, when men talk together, the hearer receives from each emotional tone an indication, a signal, of the speaker's attitude of body, and through this of his state of mind. These he can recognize, and even reproduce in himself, as the operator at one end of a telegraphic wire can follow, by noticing his needles, the action of his colleague at the other. In watching the process which thus enables one man to take a copy of another's emotions through their physical effects on his vocal tone, we may admire the perfection with which a means so simple answers an end so complex, and apparently so remote.

By eliminating from speech all effects of gesture, of expression of face, and of emotional tone, we go far toward reducing it to that system of conventional articulate sounds which the grammarian and the comparative philologist habitually consider as language. These articulate sounds are capable of being roughly set down in signs standing for vowels and consonants, with the aid of accents and other significant marks; and they may then again be read aloud from these written signs, by any one who has learnt to give its proper sound to each letter.

What vowels are, is a matter which has been for some years well understood.^[5] They are compound musical tones such as, in the vox humana stop of the organ, are sounded by reeds (vibrating tongues) fitted to organ-pipes of particular construction. The manner of formation of vowels by the voice is shortly this. There are situated in the larynx a pair of vibrating membranes called the vocal chords, which may be rudely imitated by stretching a piece of sheet india-rubber over the open end of a tube, so as to form two half-covers to it, 'like the parchment of a drum split across the middle;' when the tube is blown through, the india-rubber flaps will vibrate as the vocal chords do in the larynx, and give out a sound. In the human voice, the musical effect of the vibrating chords is increased by the cavity of the mouth, which acts as a resonator or sounding-box, and which also, by its shape at any moment, modifies the musical 'quality' or 'timbre' of the sound produced. This, not the less felt because its effects are not registered in musical notation, depends on the harmonic overtones accompanying the fundamental tone which alone musical notation takes account of. It makes the difference between the same note on two instruments, flute and piano for instance, while some instruments, as the violin, can give to one note a wide variation of quality. To such quality the formation of vowels is due. This is perfectly shown by the common Jew's harp, which when struck can be made to utter the vowels a, e, i, o, u, &c., by simply putting the mouth in the proper position for speaking these vowels. In this experiment the player's voice emits no sound, but the vibrating tongue of the Jew's harp placed in front of the mouth acts as a substitute for the vocal chords, and the vowel-sounds are produced by the various positions of the cavity of the mouth, modifying the quality of the note, by bringing out with different degrees of strength the series of harmonic tones of which it is composed. As to musical theory, emotional tone and vowel-tone are connected. In fact, an emotional tone may be defined as a vowel, whose particular musical quality is that produced by the human vocal organs, when adjusted to a particular state of feeling.

Europeans, while using modulation of musical pitch as affecting the force of words in a sentence, know nothing of making it alter the dictionary-meaning of a word. But this device is known elsewhere, especially in South-East Asia, where rises and falls of tone, to some extent like those which serve us in conveying emphasis, question and answer, &c., actually give different signification. Thus in Siamese, há=to seek, hã=pestilence, hà=five. The consequence of this elaborate system of tone-accentuation is the necessity of an accumulation of expletive particles, to supply the place of the oratorical or emphatic intonation, which being thus given over to the dictionary is lost for the grammar. Another consequence is, that the system of setting poetry to music becomes radically different from ours; to sing a Siamese song to a European tune makes the meaning of the syllables alter according to their rise and fall in pitch, and turns their sense into the wildest nonsense.^[6] In West Africa, again, the same device appears: thus in Dahoman so=stick, só=horse, sò=thunder; Yoruba, bá=with, bà=bend.^[7] For practical purposes, this linguistic music is hardly to be commended, but theoretically it is interesting, as showing that man does not servilely follow an intuitive or inherited scheme of language, but works out in various ways the resources of sound as a means of expression.

The theory of consonants is much more obscure than that of vowels. They are not musical vibrations as vowels are, but noises accompanying them. To the musician such noises as the rushing of the wind from the organ-pipe, the scraping of the violin, the sputtering of the flute, are simply troublesome as interfering with his musical tones, and he takes pains to diminish them as much as may be. But in the art of language noises of this kind, far from being avoided, are turned to immense account by being used as consonants, in combination with the musical vowels. As to the positions and movements of the vocal organs in producing consonants, an excellent account with anatomical diagrams is given in Professor Max Müller's second series of Lectures. For the present purpose of passing in review the various devices by which the language-maker has contrived to make sound a means of expressing thought, perhaps no better illustration of their nature can be mentioned than Sir Charles Wheatstone's account of his speaking machine;^[8] for one of the best ways of studying difficult phenomena is to see them artificially imitated. The instrument in question pronounced Latin, French, and Italian words well: it could say, 'Je vous aime de tout mon cœur,' 'Leopoldus Secundus Romanorum Imperator,' and so forth, but it was not so successful with German. As to the vowels, they were of course simply sounded by suitable reeds and pipes. To affect them with consonants, contrivances were arranged to act like the human organs. Thus p was made by suddenly removing the operator's hand from the mouth of the figure, and b in the same way, except that the mouth was not quite covered, while an outlet like the nostrils was used in forming m; f and v were rendered by modifying the shape of the mouth by a hand; air was made to rush through small tubes to produce the sibilants s and sh; and the liquids r and l were sounded by the action of tremulous reeds. As Wheatstone remarks, the most important use of such ingenious mechanical imitations of speech may be to fix and preserve an accurate register of the pronunciation of different languages. A perfectly arranged speaking machine would in fact represent for us that framework of language which consists of mere vowels and consonants, though without most of those expressive adjuncts which go to make up the conversation of speaking men.

Of vowels and consonants capable of being employed in language, man is able to pronounce and distinguish an enormous variety. But this great stock of possible sounds is nowhere brought into use altogether. Each language or dialect of the world is found in practice to select a limited series of definite vowels and consonants, keeping with tolerable exactness to each, and thus choosing what we may call its phonetic alphabet. Neglecting such minor differences as occur in the speech of individuals or small communities, each dialect of the world may be said to have its own phonetic system, and these phonetic systems vary widely. Our vowels, for instance, differ much from those of French and Dutch. French knows nothing of either of the sounds which we write as th in thin and that, while the Castilian lisped c, the so-called ceceo, is a third consonant which we must again make shift to write as th, though it is quite distinct in sound from both our own. It is quite a usual thing for us to find foreign languages wanting letters even near in sound to some of ours, while possessing others unfamiliar to ourselves. Among such cases are the Chinese difficulty in pronouncing r, and the want of s and f in Australian dialects. When foreigners tried to teach the Mohawks, who have no labials in their language, to pronounce words with p and b in them, they protested that it was too ridiculous to expect people to shut their mouths to speak; and the Portuguese discoverers of Brazil, remarking that the natives had neither f, l, nor r in their language, neatly described them as a people with neither fé, ley, nor rey, neither faith, law, nor king. It may happen, too, that sounds only used by some nations as interjectional noises, unwritten and unwriteable, shall be turned to account by others in their articulate language. Something of this kind occurs with the noises called 'clicks.' Such sounds are familiar to us as interjections; thus the lateral click made in the cheek (and usually in the left cheek) is continually used in driving horses, while varieties of the dental and palatal click made with the tongue against the teeth and the roof of the mouth, are common in the nursery as expressions of surprise, reproof, or satisfaction. Thus, too, the natives of Tierra del Fuego express 'no' by a peculiar cluck, as do also the Turks, who accompany it with the gesture of throwing back the head; and it appears from the accounts of travellers that the clicks of surprise and admiration among the natives of Australia are much like those we hear at home. But though here these clicking noises are only used interjectionally, it is well known that South African races have taken such sounds up into their articulate speech and have made, as we may say, letters of them. The very name of Hottentots, applied to the Namaquas and other kindred tribes, appears to be not a native name (as Peter Kolb thought) but a rude imitative word coined by the Dutch to express the clicking 'hot en tot' and the term Hottentotism has been thence adopted as a medical description of one of the varieties of stammering. North-West America is another district of the world distinguished for the production of strange clucking, gurgling, and grunting letters, difficult or impossible to European voices. Moreover, there are many sounds capable of being used in articulate speech, varieties of chirping, whistling, blowing, and sucking noises, of which some are familiar to our own use as calls to animals, or interjectional noises of contempt or surprise, but which no tribe is known to have brought into their alphabet. With all the vast phonetic variety of known languages, the limits of possible utterance are far from being reached.

Up to a certain point we can understand the reasons which have guided the various tribes of mankind in the selection of their various alphabets; ease of utterance to the speaker, combined with distinctness of effect to the hearer, have been undoubtedly among the principal of the selecting causes. We may fairly connect with the close uniformity of men's organs of speech all over the world, the general similarity which prevails in the phonetic systems of the most different languages, and which gives us the power of roughly writing down so large a proportion of any one language by means of an alphabet intended for any other. But while we thus account by physical similarity for the existence of a kind of natural alphabet common to mankind, we must look to other causes to determine the selection of sounds used in different languages, and to account for those remarkable courses of change which go on in languages of a common stock, producing in Europe such variations of one original word as pater, father, vater, or in the islands of Polynesia offering us the numeral 5 under the strangely-varied forms of lima, rima, dima, nima, and hima. Changes of this sort have acted so widely and regularly, that since the enunciation of Grimm's law their study has become a main part of philology. Though their causes are as yet so obscure, we may at least argue that such wide and definite operations cannot be due to chance or arbitrary fancy, but must be the result of laws as wide and definite as themselves.

Let us now suppose a book to be written with a tolerably correct alphabet, for instance an ordinary Italian book, or an English one in some good system of phonetic letters. To suppose English written in the makeshift alphabet which we still keep in use, would be of course to complicate the matter in hand with a new and needless difficulty. If, then, the book be written in a sufficient alphabet, and handed to a reader, his office will by no means stop short at rendering back into articulate sounds the vowels and consonants before him, as though he were reading over proofs for the press. For the emotional tone just spoken of has dropped out in writing down the words in letters, and it will be the reader's duty to guess from the meaning of the words what this tone should be, and to put it in again accordingly. He has moreover to introduce emphasis, whether by accent or stress, on certain syllables or words, thereby altering their effect in the sentence; if he says, for example, 'I never sold you that horse,' an emphasis on any one of these six words will alter the import of the whole phrase. Now, in emphatic pronunciation two distinct processes are to be remarked. The effect produced by changes in loudness and duration of words is directly imitative; it is a mere gesture made with the voice, as we may notice by the way in which any one will speak of 'a short sharp answer,' 'a long weary year,' 'a loud burst of music,' 'a gentle gliding motion,' as compared with the like manner in which the gesture-language would adapt its force and speed to the kind of action to be represented. Written language can hardly convey but by the context the striking effects which our imitative faculty adds to spoken language, in our continual endeavour to make the sound of each word we speak a sort of echo to its sense. We see this in the difference between writing and telling the little story of the man who was worried by being talked to about 'good books.' 'Do you mean,' he asked, speaking shortly with a face of strong firm approval, 'good books?' 'or,' with a drawl and a fatuous-benevolent simper, 'goo-d books?' Musical accent (accentus,^[9] musical tone) is turned to account as a means of emphasis, as when we give prominence to a particular syllable or word in a sentence by raising or depressing it a semi-tone or more. The reader has to divide his sentences with pauses, being guided in this to some extent by stops; the rhythmic measure in which he will utter prose as well as poetry is not without its effect; and he has again to introduce music by speaking each sentence to a kind of imperfect melody. Professor Helmholtz endeavours to write down in musical notes how a German with a bass voice, speaking on B flat, might say, 'Ich bin spatzieren gegangen. — Bist du spatzieren gegangen?' falling a fourth (to F) at the end of the affirmative sentence, and rising a fifth (to f) in asking the question, thus ranging through an octave.^[10] When an English speaker tries to illustrate in his own language the rising and falling tones of Siamese vowels, he compares them with the English ones of question and answer, as in 'Will you go? Yes.'^[11] The rules of this imperfect musical intonation in ordinary conversation have been as yet but little studied. But as a means of giving solemnity and pathos to language, it has been more fully developed and even systematized under exact rules of melody, and we thus have on the one hand ecclesiastical intoning and the less conventional half-singing so often to be heard in religious meetings, and on the other the ancient and modern theatrical recitative. By such intermediate stages we may cross the wide interval from spoken prose, with the musical pitch of its vowels so carelessly kept, and so obscured by consonants as to be difficult even to determine, to full song, in which the consonants are as much as possible suppressed, that they may not interfere with the precise and expressive music of the vowels.

Proceeding now to survey such parts of the vocabulary of mankind as appear to have an intelligible origin in the direct expression of sense by sound, let us first examine Interjections. When Horne Tooke spoke, in words often repeated since, of 'the brutish inarticulate Interjection,' he certainly meant to express his contempt for a mode of expression which lay outside his own too narrow view of language. But the epithets are in themselves justifiable enough. Interjections are undoubtedly to a certain extent 'brutish' in their analogy to the cries of animals; and the fact gives them an especial interest to modern observers, who are thus enabled to trace phenomena belonging to the mental state of the lower animals up into the midst of the most highly cultivated human language. It is also true that they are 'inarticulate,' so far at least that the systems of consonants and vowels recognized by grammarians break down more hopelessly than elsewhere in the attempt to write down interjections. Alphabetic writing is far too incomplete and clumsy an instrument to render their peculiar and variously-modulated sounds, for which a few conventionally-written words do duty poorly enough. In reading aloud, and sometimes even in the talk of those who have learnt rather from books than from the living world, we may hear these awkward imitations, ahem! hein! tush! tut! pshaw! now carrying the unquestioned authority of words printed in a book, and reproduced letter for letter with a most amusing accuracy. But when Horne Tooke fastens upon an unfortunate Italian grammarian and describes him as 'The industrious and exact Cinonio, who does not appear ever to have had a single glimpse of reason,' it is not easy to see what the pioneer of English philology could find to object to in Cinonio's obviously true assertion, that a single interjection, ah! or ahi! is capable of expressing more than twenty different emotions or intentions, such as pain, entreaty, threatening, sighing, disdain, according to the tone in which it is uttered.^[12] The fact that interjections do thus utter feelings is quite beyond dispute, and the philologist's concern with them is on the one hand to study their action in expressing emotion, and on the other to trace their passage into more fully-formed words, such as have their place in connected syntax and form part of logical propositions.

In the first place, however, it is necessary to separate from proper interjections the many sense-words which, often kept up in a mutilated or old-fashioned guise, come so close to them both in appearance and in use. Among classic examples are φέρε! δεῦτε! age! macte! Such a word is hail! which as the Gothic Bible shows, was originally an adjective, 'whole, hale, prosperous,' used vocatively, just as the Italians cry bravo! brava! bravi! brave! When the African negro cries out in fear or wonder mámá! mámá!^[13] he might be thought to be uttering a real interjection, 'a word used to express some passion or emotion of the mind,' as Lindley Murray has it, but in fact he is simply calling, grown-up baby as he is, for his mother; and the very same thing has been noticed among Indians of Upper California, who as an expression of pain cry, aná! that is 'mother.'^[14] Other exclamations consist of a pure interjection combined with a pronoun, as οἴμοι! oimè! ah me! or with an adjective, as alas! hélas! (ah weary!) With what care interjections should be sifted, to avoid the risk of treating as original elementary sounds of language what are really nothing but sense-words, we may judge from the way in which the common English exclamation well! well! approaches the genuine interjectional sound in the Coptic expression 'to make ouelouele,' which signifies to wail, Latin ululare. Still better, we may find a learned traveller in the 18th century quite seriously remarking, apropos of the old Greek battle-shout, ἀλαλά! ἀλαλά! that the Turks to this day call out Allah! Allah! Allah! upon the like occasion.^[15]

The calls to animals customary in different countries^[16] are to a great extent interjectional in their use, but to attempt to explain them as a whole is to step upon as slippery ground as lies within the range of philology. Sometimes they may be in fact pure interjections, like the schû schû! mentioned as an old German cry to scare birds, as we should say sh sh!, or the aá! with which the Indians of Brazil call their dogs. Or they may be set down as simple imitations of the animal's own cries, as the clucking to call fowls in our own farm-yards, or the Austrian calls of pi pi! or tiet tiet! to chickens, or the Swabian kauter kaut! to turkeys, or the shepherd's baaing to call sheep in India. In other cases, however, they may be sense-words more or less broken down, as when the creature is spoken to by a sound which seems merely taken from its own common name. If an English countryman meets a stray sheep-dog, he will simply call to him ship! ship! So schäp schäp! is an Austrian call to sheep, and köss kuhel köss! to cows. In German districts gus gus! gusch gusch! gös gös! are set down as calls to geese; and when we notice that the Bohemian peasant calls husy! to them, we remember that the name for goose in his language is husa, a word familiar to English ears in the name of John Huss. The Bohemian, again, will call to his dog ps ps! but then pes means 'dog.' Other sense-words addressed to animals break down by long repetition into mutilated forms. When we are told that the to to! with which a Portuguese calls a dog is short for toma toma! (i.e., 'take take!') which tells him to come and take his food, we admit the explanation as plausible; and the coop coop! which a cockney might so easily mistake for a pure interjection, is only 'Come up! come up!'

'Come uppe, Whitefoot, come uppe, Lightfoot, Come uppe, Jetty, rise and follow, Jetty, to the milking shed.'

But I cannot offer a plausible guess at the origin of such calls as hüf hüf! to horses, hühl hühl! to geese, deckel deckel! to sheep. It is fortunate for etymologists that such trivial little words have not an importance proportioned to the difficulty of clearing up their origin. The word puss! raises an interesting philological problem. An English child calling puss puss! is very likely keeping up the trace of the old Keltic name for the cat, Irish pus, Erse pusag, Gaelic puis. Similar calls are known elsewhere in Europe (as in Saxony, pûs pûs!), and there is some reason to think that the cat, which came to us from the East, brought with it one of its names, which is still current there, Tamil pûsei! Afghan pusha, Persian pushak, &c. Mr. Wedgwood finds an origin for the call in an imitation of the cat's spitting, and remarks that the Servians cry pis! to drive a cat away, while the Albanians use a similar sound to call it. The way in which the cry of puss! has furnished a name for the cat itself, comes out curiously in countries where the animal has been lately introduced by Englishmen. Thus boosi is the recognized word for cat in the Tonga Islands, no doubt from Captain Cook's time. Among Indian tribes of North- West America, pwsh, pish-pish, appear in native languages with the meaning of cat; and not only is the European cat called a puss puss in the Chinook Jargon, but in the same curious dialect the word is applied to a native beast, the cougar, now called ' hyas puss-puss,' i.e., 'great cat.'^[17]

The derivation of names of animals in this manner from calls to them, may perhaps not have been unfrequent. It appears that huss! is a cry used in Switzerland to set dogs on to fight, as s—s! might be in England, and that the Swiss call a dog huss or hauss, possibly from this. We know the cry of dill! dilly! as a recognized call to ducks in England, and it is difficult to think it a corruption of any English word or phrase, for the Bohemians also call dlidli! to their ducks. Now, though dill or dilly may not be found in our dictionaries as the name for a duck, yet the way in which Hood can use it as such in one of his best-known comic poems, shows perfectly the easy and natural step by which such transitions can be made: —

'For Death among the water-lilies, Cried "Duc ad me" to all her dillies.'

In just the same way, because gee! is a usual call of the English waggoner to his horses, the word gee-gee has become a familiar nursery noun meaning a horse. And neither in such nursey words, nor in words coined in jest,

1 See Pictet, 'Origines Indo-Europ.' part i. p. 382; Caldwell, 'Gr. of Dravidian Langs.' p. 465; Wedgwood, Dic. s.v. 'puss,' &c. ; Mariner, 'Tonga Is. (Vocab.)'; Gibbs, 'Dic. of Chinook Jargon,' Smithsonian Coll. No. 161; Pandosy, 'Gr. and Dic. of Yakama,' Smithson. Contr. vol. iii.; compare J. L. Wilson, 'Mpongwe Gr.' p. 57. The Hindu child's call to the cat mun mun! may be from Hindust. mâno=cat. It. micio, Fr. mite, minon, Ger. mieze, &c.='cat,' and Sp. miz! Ger. minz! &c.='puss!' are from imitations of a mew. is the evidence bearing on the origin of language to be set aside as worthless; for it may be taken as a maxim of ethnology, that what is done among civilized men in jest, or among civilized children in the nursery, is apt to find its analogue in the serious mental effort of savage, and therefore of primæval tribes.

Drivers' calls to their beasts, such as this gee! gee-ho! to urge on horses, and weh! woh! to stop them, form part of the vernacular of particular districts. The geho! perhaps came to England in the Norman-French, for it is known in France, and appears in the Italian dictionary as gio! The traveller who has been hearing the drivers in the Grisons stop their horses with a long br-r-r! may cross a pass and hear on the other side a hü-ü-ü! instead. The ploughman's calls to turn the leaders of the team to right and left have passed into proverb. In France they say of a stupid clown 'Il n'entend ni à dia! ni a hurhaut!' and the corresponding Platt-Deutsch phrase is 'He weet nich hutt! noch hoh!' So there is a regular language to camels, as Captain Burton remarks on his journey to Mekka: ikh ikh! makes them kneel, yáhh yáhh! urges them on, hai hai! induces caution, and so forth. In the formation of these quaint expressions, two causes have been at work. The sounds seem sometimes thoroughly interjectional, as the Arab hai! of caution, or the French hue! North German jö! Whatever their origin, they may be made to carry their sense by imitative tones expressive to the ear of both horse and man, as any one will say who hears the contrast between the short and sharp high-pitched hüp! which tells the Swiss horse to go faster, and the long-drawn hü-ü-ü-ü! which brings him to a stand. Also, the way in which common sense-words are taken up into calls like gee-up! woh-back! shows that we may expect to find various old broken fragments of formal language in the list, and such on inspection we find accordingly. The following lines are quoted by Halliwell from the Micro-Cynicon (1599): —

'A base borne issue of a baser syer, Bred in a cottage, wandering in the myer, With nailed shooes and whipstaffe in his hand, Who with a hey and ree the beasts command.'

This ree! is equivalent to 'right' (riddle-me-ree=riddle me right), and tells the leader of the team to bear to the right hand. The hey! may correspond with heit! or camether! which call him to bear 'hither,' i.e., to the left. In Germany har! här! har-üh! are likewise the same as 'her,' 'hither, to the left.' So swude! schwude! zwuder! 'to the left,' are of course simply 'zuwider,' 'on the contrary way.' Pairs of calls for 'right' and 'left' in German-speaking countries are hot! — har! and hott! — wist! This wist! is an interesting example of the keeping up of ancient words in such popular tradition. It is evidently a mutilated form of an old German word for the left hand, winistrâ, Anglo-Saxon winstre, a name long since forgotten by modern High German, as by our own modern English.^[18]

As quaint a mixture of words and interjectional cries as I have met with, is in the great French Encyclopædia,^[19] which gives a minute description of the hunter's craft, and prescribes exactly what is to be cried to the hounds under all possible contingencies of the chase. If the creatures understood grammar and syntax, the language could not be more accurately arranged for their ears. Sometimes we have what seem pure interjectional cries. Thus, to encourage the hounds to work, the huntsman is to call to them hà halle halle halle! while to bring them up before they are uncoupled it is prescribed that he shall call hau hau! or hau tahaut! and when they are uncoupled he is to change his cry to hau la y la la y la tayau! a call which suggests the Norman original of the English tally-ho! With cries of this kind plain French words are intermixed, hà bellement là ila, là ila, hau valet! — hau l'ami, tau tau après après, à route à route! and so on. And sometimes words have broken down into calls whose sense is not quite gone, like the 'vois le ci' and the 'vois le ce l'est' which are still to be distinguished in the shout which is to tell the hunters that the stag they have been chasing has made a return, vauleci revari vaulecelez! But the drollest thing in the treatise is the grave set of English words (in very Gallic shape) with which English dogs are to be spoken to, because, as the author says, 'there are many English hounds in France, and it is difficult to get them to work when you speak to them in an unknown tongue, that is, in other terms than they have been trained to.' Therefore, to call them, the huntsman is to cry here do-do ho ho! to get them back to the right track he is to say houpe boy, houpe boy! when there are several on ahead of the rest of the pack, he is to ride up to them and cry saf me boy! saf me boy! and lastly, if they are obstinate and will not stop, he is to make them go back with a shout of cobat, cobat!

How far the lower animals may attach any inherent meaning to interjectional sounds is a question not easy to answer. But it is plain that in most of the cases mentioned here they only understand them as recognized signals which have a meaning by regular association, as when they remember that they are fed with one noise and driven away with another, and they also pay attention to the gestures which accompany the cries. Thus the well-known Spanish way of calling the cat is miz miz! while zape zape! is used to drive it away; and the writer of an old dictionary maintains that there can be no real difference between these words except by custom, for, he declares, he has heard that in a certain monastery where they kept very handsome cats, the brother in charge of the refectory hit upon the device of calling zape zape! to them when he gave them their food, and then he drove them away with a stick, crying angrily miz miz; and this of course prevented any stranger from calling and stealing them, for only he and the cats knew the secret!^[20] To philologists, the manner in which such calls to animals become customary in particular districts illustrates the concensus by which the use of words is settled. Each case of the kind indicates that a word has prevailed by selection among a certain society of men, and the main reasons of words holding their ground within particular limits, though it is so difficult to assign them exactly in each case, are probably inherent fitness in the first place, and traditional inheritance in the second.

When the ground has been cleared of obscure or mutilated sense-words, there remains behind a residue of real sound-words, or pure interjections. It has long and reasonably been considered that the place in history of these expressions is a very primitive one. Thus De Brosses describes them as necessary and natural words, common to all mankind, and produced by the combination of man's conformation with the interior affections of his mind. One of the best means of judging the relation between interjectional utterances and the feelings they express, is to compare the voices of the lower animals with our own. To a considerable extent there is a similarity. As their bodily and mental structure has an analogy with our own, so they express their minds by sounds which have to our ears a certain fitness for what they appear to mean. It is so with the bark, the howl, and the whine of the dog, the hissing of geese, the purring of cats, the crowing and clucking of cocks and hens. But in other cases, as with the hooting of owls and the shrieks of parrots and many other birds, we cannot suppose that these sounds are intended to utter anything like the melancholy or pain which such cries from a human being would be taken to convey. There are many animals that never utter any cry but what, according to our notions of the meaning of sounds, would express rage or discomfort; how far are the roars and howls of wild beasts to be thus interpreted? We might as well imagine the tuning violin to be in pain, or the moaning wind to express sorrow. The connexion between interjection and emotion depending on the physical structure of the animal which utters or hears the sound, it follows that the general similarity of interjectional utterance among all the varieties of the human race is an important manifestation of their close physical and intellectual unity.

Interjectional sounds uttered by man for the expression of his own feelings serve also as signs indicating these feelings to another. A long list of such interjections, common to races speaking the most widely various languages, might be set down in a rough way as representing the sighs, groans, moans, cries, shrieks, and growls by which man gives utterance to various of his feelings. Such for instance, are some of the many sounds for which ah! oh! ahi! aie! are the inexpressive written representatives; such is the sigh which is written down in the Wolof language of Africa as hhihhe! in English as heigho! in Greek and Latin as (Greek characters) heu! cheu! Thus the open-mouthed wah wah! of astonishment, so common in the East, reappears in America in the hwah! hwah-wa! of the Chinook Jargon; and the kind of groan which is represented in European languages by weh! ouais! (Greek characters) vae! is given in Coptic by ouae! in Galla by wayo! in the Ossetic of the Caucasus by voy! among the Indians of British Columbia by woï! Where the interjections taken down in the vocabularies of other languages differ from those recognized in our own, we at any rate appreciate them and see how they carry their meaning. Thus with the Malagasy u-u! of pleasure, the North-American Indian's often-described guttural ugh! the kwish! of contempt in the Chinook Jargon, the Tunguz yo yo! of pain, the Irish wb wb! of distress, the native Brazilian's teh teh! of wonder and reverence, the hai-yah! so well known in the Pigeon-English of the Chinese ports, and even, to take an extreme case, the interjections of surprise among the Algonquin Indians, where men say tiau! and women nyau! It is much the same with expressions which are not uttered for the speaker's satisfaction, but are calls addressed to another. Thus the Siamese call of hē! the Hebrew he! ha! for 'lo! behold!' the hói! of the Clallam Indians for 'stop!' the Lummi hái! for 'hold, enough!' — these and others like them belong just as much to English. Another class of interjections are such as any one conversant with the gesture-signs of savages and deaf-mutes would recognize as being themselves gesture signs, made with vocal sound, in short, voice-gestures. The sound m'm, m'n, made with the lips closed, is the obvious expression of the man who tries to speak, but cannot. Even the deaf-and-dumb child, though he cannot hear the sound of his own voice, makes this noise to show that he is dumb, that he is mu mu, as the Vei negroes of West Africa would say. To the speaking man, the sound which we write as mum! says plainly enough 'hold your tongue!' 'mum's the word!' and in accordance with this meaning has served to form various imitative words, of which a type is Tahitian mamu, to be silent. Often made with a slight effort which aspirates it, and with more or less continuance, this sound becomes what may be indicated as 'm, 'n, h'm, h'n, &c., interjections which are conventionally written down as words, hem! ahem! hein! Their primary sense seems in any case that of hesitation to speak, of 'humming and hawing,' but this serves with a varied intonation to express such hesitation or refraining from articulate words as belongs either to surprise, doubt or enquiry, approbation or contempt. In the vocabulary of the Yorubas of West Africa, the nasal interjection huñ is rendered, just as it might be in English, as 'fudge!' Rochefort describes the Caribs listening in reverent silence to their chief's discourse, and testifying their approval with a hun-hun! just as in his time (17th century) an English congregation would have saluted a popular preacher.^[21] The gesture of blowing, again, is a familiar expression of contempt and disgust, and when vocalized gives the labial interjections which are written pah! bah! pugh! pooh! in Welsh pw! in Low Latin puppup! and set down by travellers among the savages in Australia as pooh! These interjections correspond with the mass of imitative words which express blowing, such as Malay puput, to blow. The labial gestures of blowing pass into those of spitting, of which one kind gives the dental interjection t' t' t'! which is written in English or Dutch tut tut! and that this is no mere fancy, a number of imitative verbs of various countries will serve to show, Tahitian tutua, to spit, being a typical instance.

The place of interjectional utterance in savage intercourse is well shown in Cranz's description. The Greenlanders, he says, especially the women, accompany many words with mien and glances, and he who does not well apprehend this may easily miss the sense. Thus when they affirm anything with pleasure they suck down air by the throat with a certain sound, and when they deny anything with contempt or horror, they turn up the nose and give a slight sound through it. And when they are out of humour, one must understand more from their gestures than their words.^[22] Interjection and gesture combine to form a tolerable practical means of intercourse, as where the communication between French and English troops in the Crimea is described as 'consisting largely of such interjectional utterances, reiterated with expressive emphasis and considerable gesticulation.'^[23] This description well brings before us in actual life a system of effective human intercourse, in which there has not yet arisen the use of those articulate sounds carrying their meaning by tradition, which are the inherited words of the dictionary.

When, however, we look closely into these inherited sense-words themselves, we find that interjectional sounds have actually had more or less share in their formation. Not stopping short at the function ascribed to them by grammarians, of standing here and there outside a logical sentence, the interjections have also served as radical sounds out of which verbs, substantives, and other parts of speech have been shaped. In tracing the progress of interjections upward into fully developed language, we begin with sounds merely expressing the speaker's actual feelings. When, however, expressive sounds like ah! ugh! pooh! are uttered not to exhibit the speaker's actual feelings at the moment, but only in order to suggest to another the thought of admiration or disgust, then such interjections have little or nothing to distinguish them from fully formed words. The next step is to trace the taking up of such sounds into the regular forms of ordinary grammar. Familiar instances of such formations may be found among ourselves in nursery language, where to woh is found in use with the meaning of to stop, or in that real though hardly acknowledged part of the English language to which belong such verbs as to boo-hoo. Among the most obvious of such words are those which denote the actual utterance of an interjection, or pass thence into some closely allied meaning. Thus the Fijian women's cry of lamentation oile! becomes the verb oile 'to bewail,' oile-taka 'to lament for' (the men cry ule!); now this is in perfect analogy with such words as ululare, to wail. With different grammatical terminations, another sound produces the Zulu verb gigiteka and its English equivalent to giggle. The Galla iya, 'to cry, scream, give the battle-cry' has its analogues in Greek (Greek characters), 'a cry,' (Greek characters) 'wailing, mournful,' &c. Good cases may be taken from a curious modern dialect with a strong propensity to the use of obvious sound-words, the Chinook Jargon of North-West America. Here we find adopted from an Indian dialect the verb to kish-kish, that is, 'to drive cattle or horses'; humm stands for the word 'stink,' verb or noun; and the laugh, heehee, becomes a recognized term meaning fun or amusement, as in mamook heehee, 'to amuse' (i.e., 'to make heehee') and heehee house, 'a tavern.' In Hawaii, aa is 'to insult;' in the Tonga Islands, úi! is at once the exclamation 'fie!' and the verb 'to cry out against.' In New Zealand, hé! is an interjection denoting surprise at a mistake, hé as a noun or verb meaning 'error, mistake, to err, to go astray.' In the Quiché language of Guate- mala, the verbs ay, oy, boy, express the idea of 'to call' in different ways. In the Carajas language of Brazil, we may guess an interjectional origin in the adjective ei, 'sorrowful,' and can scarcely fail to see a derivation from expressive sound in the verb hai-hai 'to run away' (the word aie-aie, used to mean 'an omnibus' in modern French slang, is said to be a comic allusion to the cries of the passengers whose toes are trodden on). The Camacan Indians, when they wish to express the notion of 'much' or 'many,' hold out their fingers and say hi. As this is an ordinary savage gesture expressing multitude, it seems likely that the hi is a mere interjection, requiring the visible sign to convey the full meaning.^[24] In the Quichua language of Peru, alalau! is an interjection of complaint at cold, whence the verb alalauñini, 'to complain of the cold.' At the end of each strophe of the Peruvian hymns to the Sun was sung the triumphant exclamation haylli! and with this sound are connected the verbs hayllini 'to sing,' hayllicuni, 'to celebrate a victory.' The Zulu

1 Compare, in the same district, Camé ii, Cotoxó hiehie, euhiähiä, multus, -a, -um. halala! of exultation, which becomes also a verb 'to shout for joy,' has its analogues in the Tibetan alala! of joy, and the Greek (Greek characters), which is used as a noun meaning the battle-cry and even the onset itself, (Greek characters), 'to raise the war-cry,' as well as Hebrew hillel, 'to sing praise,' whence hallelujah! a word which the believers in the theory that the Red Indians were the Lost Tribes naturally recognized in the native medicine-man's chant of hi-le-li-lah! The Zulu makes his panting ha! do duty as an expression of heat, when he says that the hot weather 'says ha ha'; his way of pitching a song by a ha! ha! is apparently represented in the verb haya, 'to lead a song,' hayo, 'a starting song, a fee given to the singing-leader for the haya'; and his interjectional expression bà bà! 'as when one smacks his lips from a bitter taste,' becomes a verb-root meaning 'to be bitter or sharp to the taste, to prick, to smart.' The Galla language gives some good examples of interjections passing into words, as where the verbs birr-djeda (to say brr!) and birēfada (to make brr!) have the meaning 'to be afraid.' Thus o! being the usual answer to a call, and also a cry to drive cattle, there are formed from it by the addition of verbal terminations, the verbs oada, 'to answer,' and ofa, 'to drive.'

If the magnific and honorific o of Japanese grammar can be assigned to an interjectional origin, its capabilities in modifying signification become instructive.^[25] It is used before substantives as a prefix of honour; couni, 'country,' thus becoming ocouni. When a man is talking to his superiors, he puts o before the names of all objects belonging to them, while these superiors drop the o in speaking of anything of their own, or an inferior's; among the higher classes, persons of equal rank put o before the names of each other's things, but not before their own; it is polite to say o before the names of all women, and well-bred children are distinguished from little peasants by the way in which they are careful to put it even before the nursery names of father and mother, o toto, o caca, which correspond to the papa and mama of Europe. A distinction is made in written language between o, which is put to anything royal, and oo which means great, as may be instanced in the use of the word mets'ké or 'spy' (literally 'eye-fixer'); o mets'ké is a princely or imperial spy, while oo mets'ké is the spy in chief. This interjectional adjective oo, great, is usually prefixed to the name of the capital city, which it is customary to call oo Yedo in speaking to one of its inhabitants, or when officials talk of it among themselves. And lastly, the o of honour is prefixed to verbs in all their forms of conjugation, and it is polite to say ominahai matse, 'please to see,' instead of the mere plebeian minahai matse. Now an English child of six years old would at once understand these formations if taken as interjectional; and if we do not incorporate in our grammar the o! of admiration and reverential embarrassment, it is because we have not chosen to take advantage of this rudimentary means of expression. Another exclamation, the cry of io! has taken a place in etymology. When added by the German to his cry of 'Fire!' 'Murder!' Feuerio! Mordio! it remains indeed as mere an interjection as the o! in our street cries of 'Pease-o!' 'Dust-o!' or the â! in old German wafenâ! 'to arms!' hilfâ! 'help!' But the Iroquois of North America makes a fuller use of his materials, and carries his io! of admiration into the very formation of compound words, adding it to a noun to say that it is beautiful or good; thus, in Mohawk, garonta means a tree, garontio a beautiful tree; in like manner, Ohio means 'river-beautiful:' and Ontario, 'hill-rock-beautiful,' is derived in the same way. When, in the old times of the French occupation of Canada, there was sent over a Governor-General of New France, Monsieur de Montmagny, the Iroquois rendered his name from their word ononte, 'mountain,' translating him into Onontio, or 'Great Mountain,' and thus it came to pass that the name of Onontio was handed down long after, like that of Cæsar, as the title of each succeeding governor, while for the King of France was reserved the yet higher style of 'the great Onontio.'^[26]

The quest of interjectional derivations for sense-words is apt to lead the etymologist into very rash speculations. One of his best safeguards is to test forms supposed to be interjectional, by ascertaining whether anything similar has come into use in decidedly distinct languages. For instance, among the familiar sounds which fall on the traveller's ear in Spain is the muleteer's cry to his beasts, arre! arre! From this interjection, a family of Spanish words are reasonably supposed to be derived; the verb arrear, 'to drive mules,' arriero, the name for the 'muleteer' himself, and so forth.^[27] Now is this arre! itself a genuine interjectional sound? It seems likely to be so, for Captain Wilson found it in use in the Pelew Islands, where the paddlers in the canoes were kept up to their work by crying to them arree! arree! Similar interjections are noticed elsewhere with a sense of mere affirmation, as in an Australian dialect where a-ree! is set down as meaning 'indeed,' and in the Quichua language where ari! means 'yes!' whence the verb ariñi, 'to affirm.' Two other cautions are desirable in such enquiries. These are, not to travel too far from the absolute meaning expressed by the interjection, unless there is strong corroborative evidence, and not to override ordinary etymology by treating derivative words as though they were radical. Without these checks, even sound principle breaks down in application, as the following two examples may show. It is quite true that h'm! is a common interjectional call, and that the Dutch have made a verb of it, hemmen, 'to hem after a person.' We may notice a similar call in West Africa, in the mma! which is translated 'hallo! stop!' in the language of Fernando Po. But to apply this as a derivation for German hemmen, 'to stop, check, restrain,' to hem in, and even to the hem of a garment, as Mr. Wedgwood does without even a perhaps,^[28] is travelling too far beyond the record. Again, it is quite true that sounds of clicking and smacking of the lips are common expressions of satisfaction all over the world, and words may be derived from these sounds, as where a vocabulary of the Chinook language of North-West America expresses 'good' as t'k-tok-te, or e-tok-te, sounds which we cannot doubt to be derived from such clicking noises, if the words are not in fact attempts to write down the very clicks themselves. But it does not follow that we may take such words as deliciæ, delicatus, out of a highly organized language like Latin, and refer them, as the same etymologist does, to an interjectional utterance of satisfaction, dlick!^[29] To do this, is to ignore altogether the composition of words; we might as well explain Latin dilectus or English delight as direct formations from expressive sound. In concluding these remarks on interjections, two or three groups of words may be brought forward as examples of the application of collected evidence from a number of languages, mostly of the lower races.

The affirmative and negative particles, which bear in language such meanings as 'yes!' 'indeed!' and 'no!' 'not,' may have their derivations from many different sources. It is thought that the Australian dialects all belong to a single stock, but so unlike are the sounds they use for 'no!' and 'yes!' that tribes are actually named from these words as a convenient means of distinction. Thus the tribes known as Gureang, Kamilaroi, Kogai, Wolaroi, Wailwun, Wiratheroi, have their names from the words they use for 'no,' these being gure, kamil, ko, wol, wail, wira, respectively; and on the other hand the Pikambul are said to be so called from their word pika, 'yes.' The device of naming tribes, thus invented by the savages of Australia, and which perhaps recurs in Brazil in the name of the Cocatapuya tribe (coca 'no,' tapuya 'man') is very curious in its similarity to the mediæval division of Langue d'oc and Langue d'oïl, according to the words for 'yes!' which prevailed in Southern and Northern France: oc! is Latin hoc, as we might say 'that's it!' while the longer form hoc illud was reduced to oïl! and thence to oui! Many other of the words for 'yes!' and 'no!' may be sense-words, as, again, the French and Italian si! is Latin sic. But on the other hand there is reason to think that many of these particles in use in various languages are not sense-words, but sound-words of a purely interjectional kind; or, what comes nearly to the same thing, a feeling of fitness of the sound to the meaning may have affected the choice and shaping of sense-words — a remark of large application in such enquiries as the present. It is an old suggestion that the primitive sound of such words as non is a nasal interjection of doubt or dissent.^[30] It corresponds in sound with the visible gesture of closing the lips, while a vowel-interjection, with or without aspiration, belongs rather to open-mouthed utterance. Whether from this or some other cause, there is a remarkable tendency among most distant and various languages of the world, on the one hand to use vowel-sounds, with soft or hard breathing, to express 'yes!' and on the other hand to use nasal consonants to express 'no!' The affirmative form is much the commoner. The guttural i-i! of the West Australian, the ēē! of the Darien, the a-ah! of the Clallam, the é! of the Yakama Indians, the e! of the Basuto, and the ai! of the Kanuri, are some examples of a wide group of forms, of which the following are only part of those noted down in Polynesian and South American districts—ii! é! ia! aio! io! ya! ey! &c., h'! heh! he-e! hü! hoehah! ah-ha! &c. The idea has most weight where pairs of words for 'yes!' and 'no!' are found both conforming. Thus in the very suggestive description by Dobrizhoffer among the Abipones of South America, for 'yes!' the men and youths say héé! the women say háá! and the old men give a grunt; while for 'no' they all say yna! and make the loudness of the sound indicate the strength of the negation. Dr. Martius's collection of vocabularies of Brazilian tribes, philologically very distinct, contains several such pairs of affirmatives and negatives, the equivalents of yes!'—'no!' being in Tupi ayé—aan! aani!; in Guato ii!—mau!; in Jumana, aeae!—mäiu!; in Miranha ha ú!—nani! The Quichua of Peru affirms by y! hu! and expresses ' no,' 'not,' 'not at all,' by ama! manan! &c., making from the latter the verb manamñi, 'to deny.' The Quiché of Guatemala has e or ve for the affirmative, ma, man, mana, for the negative. In Africa, again, the Galla language has ee! for 'yes!' and hn, hin, hm, for 'not!'; the Fernandian ee! for 'yes!' and 'nt for 'not;' while the Coptic dictionary gives the affirmative (Latin 'sane') as eie, ie, and the negative by a long list of nasal sounds such as an, emmen, en, mmn, &c. The Sanskrit particles hi! 'indeed, certainly,' na, 'not,' exemplify similar forms in Indo-European languages, down to our own aye! and no!^[31] There must be some meaning in all this, for otherwise I could hardly have noted down incidentally, without making any attempt at a general search, so many cases from such different languages, only finding a comparatively small number of contradictory cases.^[32]

De Brosses maintained that the Latin stare, to stand, might be traced to an origin in expressive sound. He fancied he could hear in it an organic radical sign designating fixity, and could thus explain why st! should be used as a call to make a man stand still. Its connexion with these sounds is often spoken of in more modern books, and one imaginative German philologer describes their origin among primæval men as vividly as though he had been there to see. A man stands beckoning in vain to a companion who does not see him, till at last his effort relieves itself by the help of the vocal nerves, and involuntarily there breaks from him the sound st! Now the other hears the sound, turns toward it, sees the beckoning gesture, knows that he is called to stop; and when this has happened again and again, the action comes to be described in common talk by uttering the now familiar st! and thus sta becomes a root, the symbol of the abstract idea to stand!^[33] This is a most ingenious conjecture, but unfortunately nothing more. It would be at any rate strengthened, though not established, if its supporters could prove that the st! used to call people in Germany, pst! in Spain, is itself a pure interjectional sound. Even this, however, has never been made out. The call has not yet been shown to be in use outside our own Indo-European family of languages; and so long as it is only found in use within these limits, an opponent might even plausibly claim it as an abbreviation of the very sta! ('stay! stop!') for which the theory proposes it as an origin.^[34]

That it is not unfair to ask for fuller evidence of a sound being purely interjectional than its appearance in a single family of languages, may be shown by examining another group of interjections, which are found among the remotest tribes, and thus have really considerable claims to rank among the primary sounds of language. These are the simple sibilants, s! sh! h'sh! used especially to scare birds, and among men to express aversion or call for silence. Catlin describes a party of Sioux Indians, when they came to the portrait of a dead chief, each putting his hand over his mouth with a hush-sh; and when he himself wished to approach the sacred 'medicine' in a Mandan lodge, he was called to refrain by the same hush-sh! Among ourselves the sibilant interjection passes into two exactly opposite senses, according as it is meant to put the speaker himself to silence, or to command silence for him to be heard ; and thus we find the sibilant used elsewhere, sometimes in the one way and sometimes in the other. Among the wild Veddas of Ceylon, iss! is an exclamation of disapproval, as in ancient or modern Europe; and the verb shârak, to hiss, is used in Hebrew with a like sense, 'they shall hiss him out of his place.' But in Japan reverence is expressed by a hiss, commanding silence. Captain Cook remarked that the natives of the New Hebrides expressed their admiration by hissing like geese. Casalis says of the Basutos, 'Hisses are the most unequivocal marks of applause, and are as much courted in the African parliaments as they are dreaded by our candidates for popular favour.^[35] Among other sibilant interjections, are Turkish sûsâ! Ossetic ss! sos! 'silence!'

Fernandian sia! 'listen!' 'tush!' Yoruba sió! 'pshaw!' Thus it appears that these sounds, far from being special to one linguistic family, are very widespread elements of human speech. Nor is there any question as to their passage into fully-formed words, as in our verb to hush, which has passed into the sense of 'to quiet, put to sleep' (adjectively, 'as hush as death'), metaphorically to hush up a matter, or Greek (Greek characters) 'to hush, say hush! command silence.' Even Latin silere and Gothic silan, 'to be silent,' may with some plausibility be explained as derived from the interjectional s! of silence.

Sanskrit dictionaries recognize several words which explicitly state their own interjectional derivation; such are hûñkâra (hûm-making), 'the utterance of the mystic religious exclamation hum!' and çiççabda (çiç-sound), 'a hiss.' Besides these obvious formations, the interjectional element is present to some greater or less degree in the list of Sanskrit radicals,which represent probably better than those of any other language the verb-roots of the ancient Aryan stock. In ru, 'to roar, cry, wail,' and in kakh, 'to laugh,' we have the simpler kind of interjectional derivation, that which merely describes a sound. As to the more difficult kind, which carry the sense into a new stage, Mr. Wedgwood makes out a strong case for the connexion of interjections of loathing and aversion, such as pooh! fie! &c., with that large group of words which are represented in English by foul and fiend, in Sanskrit by the verbs pûy, 'to become foul, to stink,' and piy, pîy, 'to revile, to hate.'^[36] Further

evidence may be here adduced in support of this theory. The languages of the lower races use the sound pu to express an evil smell; the Zulu remarks that 'the meat says pu' (inyama iti pu), meaning that it stinks; the Timorese has poöp 'putrid;' the Quiché language has puh, poh 'corruption, pus,' pohir 'to turn bad, rot,' puz 'rottenness, what stinks;' the Tupi word for nasty, puxi, may be compared with the Latin putidus, and the Columbia River name for the 'skunk,' o-pun-pun, with similar names of stinking animals, Sanskrit pûtikâ 'civet-cat,' and French putois 'pole-cat.' From the French interjection fi! words have long been formed belonging to the language, if not authenticated by the Academy; in mediæval French 'maistre fi-fi' was a recognized term for a scavenger, and fi-fi books are not yet extinct.

There has been as yet, unfortunately, too much separation between what may be called generative philology, which examines into the ultimate origins of words, and historical philology, which traces their transmission and change. It will be a great gain to the science of language to bring these two branches of enquiry into closer union, even as the processes they relate to have been going on together since the earliest days of speech. At present the historical philologists of the school of Grimm and Bopp, whose great work has been the tracing of our Indo-European dialects to an early Aryan form of language, have had much the advantage in fulness of evidence and strictness of treatment. At the same time it is evident that the views of the generative philologists, from De Brosses onward, embody a sound principle, and that much of the evidence collected as to emotional and other directly expressive words, is of the highest value in the argument. But in working out the details of such word-formation, it must be remembered that no department of philology lies more open to Augustine's caustic remark on the etymologists of his time, that like the interpretation of dreams, the derivation of words is set down by each man according to his own fancy. (Ut somniorum interpretatio ita verborum origo pro cujusque ingenio prædicatur.)

1. C. de Brosses, 'Traité de la Formation Mécanique des Langues,' &c. (1st ed. 1765); Wedgwood, 'Origin of Language' (1866); 'Dic. of English Etymology' (1859, 2nd ed. 1872); Farrar, 'Chapters on Language' (1865).
2. 1
3. Bonwick, 'Daily Life of Tasmanians,' p. 140; Capt. Wilson, in 'Tr. Eth. Soc.,' vol iv. p. 322, &c.; J. L. Wilson, in 'Journ. Amer. Oriental Soc.,' vol. i. 1849, No 4; also Cranz., 'Grönland,' p. 279 (cited below, p. 186). For other accounts, see 'Early Hist. of Mankind,' p. 77.
4. Forbes, 'Aymara Indians,' in Journ. Eth. Soc. 1870, vol. ii. p. 208.
5. See Helmholtz, 'Tonempfindungen,' 2nd ed. p. 163; McKendrick, Text Book of Physiology, p. 681, &c., 720, &c.; Max Müller, 'Lectures,' 2nd series, p. 95, &c.
6. See Pallegoix, 'Gramm. Ling. Thai.'; Bastian, in 'Monatsb. Berlin. Akad.' June 6, 1867, and 'Roy. Asiatic Soc.,' June, 1867.
7. Burton, in 'Mem. Anthrop. Soc.,' vol. i. p. 313; Bowen, 'Yoruba Gr. and Dic.' p. 5; see J. L. Wilson, 'W. Afr.,' p. 461.
8. C. W., in 'London and Westminster Review,' Oct. 1837.
9. 'Accentus est etiam in dicendo cantus obscurior.' — Cic. de Orat.
10. Helmholtz, p. 364.
11. Caswell, in Bastian, 'Berlin. Akad.' l.c.
12. Horne Tooke, 'Diversions of Purley,' 2nd ed. London, 1798, pt. i. pp. б0-3.
13. R. F. Burton, 'Lake Regions of Central Africa,' vol. ii. p. 333; Livingstone, 'Missionary Tr. in S. Africa,' p. 298; 'Gr. of Mpongwe lang,' A. B. C. F. Missions, Rev. J. L. Wilson, p. 27. See Callaway, 'Zulu Tales,' vol. i. p. 59.
14. Arroyo de la Cuesta, 'Gr. of Mutsun Lang.' p. 39, in Smithsonian Contr., vol. iii.; Neapolitan mamma mia! exclamation of wonder, &c., Liebrecht in Götting. Gel. Anz 1872, p. 1287.
15. Shaw, 'Travels in Barbary,' in Pinkerton, vol. xv. p. 669.
16. Some of the examples here cited, will be found in Grimm, 'Deutsche Gr.' vol. iii. p. 308; Pott, 'Doppelung.' p. 27; Wedgwood, 'Origin of Language.'
17. 1
18. For lists of drivers' words, see Grimm, l.c.; Pott, 'Zählmethode,' p. 261; Halliwell, 'Dic. of Archaic and Provincial English,' s.v. 'ree;' Brand, vol. ii. p. 15; Pictet, part ii. p. 489.
19. 'Encyclopédie, ou Dictionnaire Raisonné des Sciences, &c.' Recueil de Planches, Paris, 1763, art. 'Chasses.' The traditional cries are still more or less in use. See 'A Week in a French Country-house.'
20. Aldrete, 'Lengua Castellana,' Madrid, 1673, s.vv. harre, exe.
21. 'There prevailed in those days an indecent custom; when the preacher touched any favourite topick in a manner that delighted his audience, their approbation was expressed by a loud hum, continued in proportion to their zeal or pleasure. When Burnet preached, part of his congregation hummed so loudly and so long, that he sat down to enjoy it, and rubbed his face with his handkerchief. When Sprat preached, he likewise was honoured with the like animating hum, but he stretched out his hand to the congregation, and cried, "Peace, peace; I pray you, peace."' Johnson, 'Life of Sprat.'
22. Cranz, 'Grönland,' p. 279.
23. D. Wilson, 'Prehistoric Man,' p. 65.
24. 1
25. J. H. Donker Curtius, 'Essai de Grammaire Japonaise,' p. 34, &c. 199. In former editions of the present work, the directly interjectional character of the o is held in an unqualified manner. Reference to the grammars of Prof. B. H. Chamberlain and others, where this particle (on, o) is connected with other forms implying a common root, leaves the argument to depend wholly or partly on the supposition of an interjectional source for this root. [Note to 3rd ed.]
26. Bruyas, 'Mohawk Lang.,' p. 16, in Smithson. Contr. vol. iii. Schoolcraft, 'Indian Tribes,' Part iii. p. 328, 502, 507. Charlevoix, 'Nouv. France,' vol. i. p. 350.
27. The arre! may have been introduced into Europe by the Moors, as it is used in Arabic, and its use in Europe corresponds nearly with the limits of the Moorish conquest, in Spain arre! in Provence arri!
28. Wedgwood, 'Origin of Language,' p. 92.
29. Ibid., p. 72.
30. De Brosses, vol. i. p. 203. See Wedgwood.
31. Also Oraon hae—ambo; Micmac é—mw.
32. A double contradiction in Carib anhan! = 'yes!' oua! = 'no!' Single contradictions in Catoquina hang! Tupi eém! Botocudo hemhem! Yoruba eñ! for 'yes!' Culino aiy! Australian yo! for 'no!' &c. How much these sounds depend on peculiar intonation, we, who habitually use h'm! either for 'yes!' or 'no!' can well understand.
33. (Charles de Brosses) 'Traité de la Formation Mécanique des Langues, &c.' Paris, An. ix., vol. i. p. 238; vol. ii. p. 313. Lazarus and Steinthal, 'Zeitschrift für Völkerpsychologie,' &c., vol. i. p. 421. Heyse, System der Sprachwissenschaft,' p. 73. Farrar, 'Chapters on Language,' p. 202.
34. Similar sounds are used to command silence, to stop speaking as well as to stop going. English husht!! whist! hist! Welsh ust! French chut! Italian zitto! Swedish tyst! Russian st'! and the Latin st! so well described in the curious old line quoted by Mr. Farrar, which compares it with the gesture of the finger on the lips: —
    'Isis, et Harpocrates digito qui significat st!'

    This group of interjections, again, has not been proved to be in use outside Aryan limits.

35. Catlin, 'North American Indians,' vol. i. pp. 221, 39, 151, 162. Bailey in 'Tr. Eth. Soc.,' vol. ii. p. 318. Job xxvii. 23. (The verb shârak also signifies to call by a hiss, 'and he will hiss unto them from the end of the earth, and behold, they shall come with speed,' Is. v. 26; Jer. xix. 8.) Alcock, 'The Capital of the Tycoon,' vol. i. p. 394. Cook, '2nd Voy.' vol. ii. p. 36. Casalis, 'Basutos,' p. 234.
36. Wedgwood, 'Origin of Language,' p. 83, 'Dictionary,' Introd. p. xlix. and s.v. 'foul.' Prof. Max Müller, 'Lectures,' 2nd series, p. 92, protests against the indiscriminate derivation of words directly from such cries and interjections, without the intervention of determinate roots. As to the present topic, he points out that Latin pus, putridus, Gothic fuls, English foul, follow Grimm's law as if words derived from a single root. Admitting this, however, the question has to be raised, how far pure interjections and their direct derivatives, being self-expressive and so to speak living sounds, are affected by phonetic changes such as that of Grimm's law, which act on articulate sounds no longer fully expressive in themselves, but handed down by mere tradition. Thus p and f occur in one and the same dialect in interjections of disgust and aversion, pub! fi! being used in Venice or Paris, just as similar sounds would be in London. In tracing this group of words from early Aryan forms, it must also be noticed that Sanskrit is a very imperfect guide, for its alphabet has no f, and it can hardly give the rule in this matter to languages possessing both p and f, and thus capable of nicer appreciation of this class of interjections.