and Dionysius of Halicarnassus even fixes the difference between a grave and an acute accent, by comparing it to a quint, which, in the circumflex, may have occurred twice on the same vowel. Cicero also says: "Est in dicendo etiam quidam cantus." The accent then was literally what its name implies, a musical element of the language, although modern languages have long since forgotten what was meant by the Latin "accentus” (ad cantus), or the Greek "prosody" (πpòs won): two features of those ancient idioms, so highly valued and carefully observed that we need not be astonished at the energetic and rather contemptuous manner in which Julian compares the songs of Germans on the banks of the Rhine, to the cawing of hoarse birds, and the sounds heard further north to the neighing of horses and barking of dogs.

Our poets and orators have comparatively little to do with the extreme subtlety and complexity of distinction with which the Greeks distinguished between accent and quantity. The former, moreover, we know, was constantly changing both character and place in the same word, accompanying each slight change of form, quantity or value, by flexion, composition or derivation. Now, accent and quantity are almost always alike, and but here and there a faint trace of former laws survives. The Italian occasionally recalls the precision with which the Romans avoided accenting the last syllable, and thus earned from the fastidious Greeks the name of "grandiloquentes," by placing an accent upon the last part of a word only when it is needed to mark the loss of a final syllable. In a similar way, French words that have been naturalized in English, show the effect of the Anglo-Saxon preference for an accent on the first

ACCENT IN LANGUAGE.

66

221

syllable: "jugement" has become "judgment;" gouvernement," "government;" "capitaine," "captain ;" and under such influence most of them lost their last syllable entirely, until Voltaire could say, with a certain air of truth, that in speaking, the English gained two hours a day over the French, as they "swallowed half of their words."

How important, nevertheless, this element may become in establishing the proper relation between the accent of a language and the character of a nation, can be seen from the simple but suggestive fact, that the Romance languages, exhibiting their inherited sense of material beauty, accent almost invariably the last syllable but one, while the German 's sense of spiritual beauty accents whatever syllable may be most important, without regard to euphony or established rule. The delicacy of the French ear, and the peculiarities of the musical element of that language are well known; although in the course of time much of this original delicacy must have been lost, enough is left to mark it with a peculiar and indelible stamp. The delicate distinction between hard and soft consonants, the extremely gentle aspiration of the "h aspiré,” and especially the graceful sound of nasal vowels and the liquid 7, form transitions from vowels to consonants, and a gentle connection between them, which few other languages can rival. A French critic, Dupuis, called these nasal sounds, from the analogy between the diatonic scale of French vowels and the notes of their music, the true "bémols" of the language.

Bunsen extended already the effects of this musical element in language to whatever belongs to peculiarities of intonation and the greater or less prevalence of certain classes of sounds in

each idiom. But there are even still fainter features in language, more difficult to perceive and yet characteristic and essential almost in proportion to their spiritual nature. Such are the "click" of the Hottentot, or the mere lengthening of a syllable which alters its meaning. Mexican nouns, for instance, when they terminate in a vowel, form their plural merely by lengthening the last sound in the form of a long aspiration, which, at first, makes no other impression upon the stranger than a pause, during which the vowel seems to lose itself into air. In the southern dialect of the Guarana language, the suffix of the perfect "yma" is pronounced more or less slowly, as the past, which it indicates, is more or less recent.

Another success obtained by the comparative study of languages, according to the principles just referred to, is the establishment of a standard by which languages may at once be assigned to one or the other class of idioms. This standard is not based upon mere external circumstances, the present location, the outward form, or the historical fate of a language. It rests upon that only characteristic which can never fail or mislead-the degree of perfection with which any given idiom fulfils the great purposes of language.

MONOSYLLABIC CLASS.

223

CHAPTER XXXIX.

THE THREE GREAT CLASSES OF LANGUAGES.

The Monosyllabic-The Agglutinizing-The Inflected.

It has already been said that language has to reflect the picture or idea produced on our mind by means of sound, so as to communicate it to others in a tangible shape. It has therefore to find utterance for the whole activity of the mind. But the mind never ceases for a moment to reproduce outward impressions, and proceeds with the rapidity of lightning. Language, on the other hand, has to translate, as it were, these impressions one by one into words, to arrest the fleeting thought in its progress, and to clothe the ethereal idea in a bodily frame. Moreover, the mind of man has an almost unlimited sphere: it ranges from world to world, and perceives alike the well-defined, positive impression and the vaguest sensation of an instant. Language has, to reproduce all these perceptions, only a few sounds, and a poor, decaying idiom. It cannot follow the mind, therefore, in its inconceivably rapid perception of imponderable thought: it must now and then lag behind; it must be satisfied with a partial and fragmentary reproduction; and can convey only the outline or the essence of ideas. The proportion, now, in which this is accomplished by different idioms forms the stand

ard by which they are judged, and suggests the three classes in which they are commonly divided.

In examining this mutual relation of mind and language, it is found that the latter has to perform two duties, which cause the principal difference between the various idioms. In our mind, namely, no idea, no thought rises in the abstract—that is, standing clearly alone and disconnected from all other thoughts. We know, on the contrary, that the mind conceives at once, and necessarily together with the idea-the relation, also, in which it stands to other ideas--its mental connection. To perceive an abstract idea, it is well known, is one of the most difficult processes of mind, and there is good reason to doubt whe-. ther it can ever be fully accomplished. Our thoughts consist, therefore, of two distinct parts: the idea itself, or what we might call the material of the thought, and the relation of that idea to others. The substantial part finds its equivalent easily enough in language; it surely has a name for each object; but it is not so easy to represent the relation also instantly and fully. This relation is so unsubstantial, so delicate a perception of the mind, that, as was said, the material word cannot at once follow, nor always adequately express it. Generally, therefore, the idea itself only is rendered in language by a genuine word, a rooti the relation, on the contrary, is but indicated or suggested by some addition to this root, or by a slight change in its form. For language has no separate means, no special material by which to represent this ideal relation, which connects the abstract idea of our mind with actual existence and object with object. It can only employ the same material which already served to render the idea itself, once more and in a peculiar