from the fact, that these words seem to have no appropriate place in so artificial a system as that of the Latin grammar, from which we have derived most of our ideas of the structure of language. They can neither be declined nor conjugated; they are incapable of degrees of comparison; they govern nothing, qualify nothing, connect nothing, and may be left out of the period altogether without affecting the syntactical propriety of its structure. In short, they cannot be parsed. They have no position in the rank and file of the legion, and therefore are at best supernumeraries, if not intruders. In a language so cemented and compacted together as the Latin, not by mortar or pins of independent material and formation, but by organic attachments, natural hooks and eyes, congenital with the words and of one substance with them, this objection to the recognition of constituents so incapable of assimilation is by no means without validity; but in English, and in those other tongues where the relations between important words are determined by mere position or by the aid of distinct and insignificant particles, it strikes us much less forcibly. I shall endeavor to vindicate the elaim of these neglected articulations to rank as legitimate means of vocally expressing human passions, states, affections, and therefore to be called words, though of a rhetorical and dramatic, not of a logical or didactic character.

Considered as a purely natural and spontaneous emission of the voice, we might expect to find similar interjections in all human tongues, but their forms, even when they most resemble each other, are modified by the same obscure influences which diversify the action of the organs of speech in the production of like or analogous sounds among different nations, and consequently they are by no means identical in different languages. The alleged diversity in the cries of the

But

infant, and in the true interjections, which two utterances, psychologically considered, belong to the same general class of expressive sounds, has been urged by some physiologists as a proof of a diversity of origin in the human race. the argument loses something of its weight, when it is shown, as it may be, that in numerous other cases, words common to two or more demonstrably cognate nations, and identical in form and sound, so far as any written notation can express sound, are nevertheless differenced in their pronunciation by those nations as widely as the true interjections are by unrelated races. These distinctions are occasioned by two proximate causes; the one is the employment of different sets of muscles, by different peoples, for the production of the same or similar sounds, the other is the peculiar quality impressed upon articulate sounds by the intonation with which they are uttered.

These two classes of linguistic facts, the production namely of like or analogous sounds in different languages by the em ployment of different organs, or at least muscles, and the fixed character of national intonation in certain languages, have as yet been little investigated by philologists, but they are full of curious interest, and the study of them, however difficult, is essential to the construction of even a tolerably complete system of phonology. Nice distinctions between related sounds depend of course upon the mechanical means employed to produce them, and one reason why an adult so seldom succeeds in mastering the pronunciation of a foreign language, why he is at once recognized as a stranger by his articulation even of words which, according to grammars and dictionaries, are identical with syllables and words of his mother-tongue, is, that to pronounce them like a native, he must call into play muscles not employed, or employed in a

different way, in speaking his own language, and which have become so rigid from disuse, that he cannot acquire the command of them, or, in other words, render them what are called voluntary muscles. Further, the organs of speech act and react upon each other; the frequent play of a given set of muscles modifies the action of neighboring or related mus cles; there is, to use a word, which, if not now English, soon will be, a certain solidarity between them all, and organs accustomed to the deep gutturals of the Arabic, the hissing and lisping sounds of the English, or the nasals of the French and Portuguese, are with great difficulty trained to the pure articulation of languages like the Italian, in which such elements do not exist.

National peculiarities of intonation are still more subtle and obscure, and they are almost equally difficult to seize by the ear, and to reproduce by the lips and tongue. To us, whose intonations belong not to the individual word, but to the whole period, it is difficult to conceive of the tone with which a word is uttered, as a constant, essential, characteristic and expressive ingredient of the word itself. But in monosyllabic languages like the Chinese, where the number of words, differing in the vowel and consonantal elements of which they are composed, must necessarily be very small, other distinctions must be resorted to, and accordingly we find that in such languages a monosyllable, consisting perhaps of one vowel and one or two consonantal elements, and which admits of but one mode of spelling in alphabetic characters, may nevertheless have a great number of meanings, each indicated by a peculiarity of intonation not perhaps appreciable by foreign ears, but nevertheless readily recognizable by a native. These peculiarities are however by no means confined to languages so alien to our own, for they ex

ist in the Danish and the Swedish, both of which are nearly allied to English, and they, no doubt, occur to a considerable, but thus far uninvestigated, extent, in other tongues more familiar to most of us. In such languages, these intonations are constant, and they are also expressive and significant, so far that certain words are under all circumstances pronounced with the same intonation, and thus distinguished from words differing from them in signification, but otherwise identical in sound. Scandinavian phonologists have made these intonations, for which the vocabulary of our language does not even furnish names, a subject of special in quiry; and Rask, one of the most eminent of modern philologists, has subtilized so far upon them, that few of his own countrymen, even, have sufficient acuteness of ear to follow him. But this is not strange, when we learn that the same discriminating phonologist fancied he could detect, what no Englishman or American ever did, a difference between the pronunciation of our two English words pale, pallid, and pail, a water-bucket.*

Yet more etherial than even these subtle shades of difference, is what, to borrow a musical term, may be called the mode in which a given language is spoken. A stranger in Greece or the East is struck at once by a certain sadness of tone, amounting at times almost to wailing, which marks the speech of the people, and especially of the women of the lower classes. Some travellers have ascribed this to the long centuries of humiliation and oppression under which women have groaned in the East; but I think it belongs rather to the races than to the sex; for it is not altogether confined to the women: and, besides, something of the same sort is found

* Rask's Danish Grammar for the use of Englishmen.

among the most primitive and simple tribes, and the fact, if it is a fact, that the music of ancient Greece and Latium, like that of most Oriental countries, was wholly in the minor mode, seems to confirm this view.

The Greek, or to speak more specifically, Alexandrian and other colonial grammarians, carefully investigated the intonation of their language, in both its branches, accentuation, and vocal inflection, and they invented several points, which we call accents, to indicate the particular intonation of the important syllables of the words. What the signification of these points was we do not know; nor does the pronunciation of the modern Greeks afford us any light on the subject. What we call accent, that is, stress of voice, has been generally supposed to have been, among other things, marked by them; but this is disputed. Metrical quantity or prosody, they certainly did not indicate, but left it to general rules, which, in most cases, were sufficiently explicit. The quantity, or relative duration of syllables as it is generally understood, is a quality of sound to which the Greek ear was acutely sensible, and it appears to have been recognized in the earlier Teutonic dialects, but to modern ears, it is, as an element of prosody, much less appreciable. In English verse and more especially by recent poets, rhythm has been made to depend upon and consist in accentuation alone, and those other elements of articulation, which to the ancient classical nations constituted the very essence of poetical melody, are, by the fashion of the day, altogether disregarded. This, I think, is a mistake, but it will be more fitly considered on another occasion.

But, to return from what may be considered a digression, the true interjections, though modified by peculiarities of intonation, have at least a family resemblance, if not an abso