laid on the principal syllable (verbal accent).1 Next, by emphasizing the principal one of a series of words (i.e. the principal syllable of this word, by means therefore of the verbal accent) it unites the series into a single sentence. In like manner several sentences, by the prominence given to the principal sentence, are made to form a period. And even beyond this limit, accent operates in still greater divisions, according as the mind by means of it is able to master the quantity. This depends, on the one hand, on the mental clearness and vivacity of the speaker or reader; on the other, on his power over his voice on his elocution. That which constitutes the principle of unity in these divisions of speech is also the principle of their separation; and it is accordingly accent which effects the division of the sense the separation of the words and the division into sentences and periods which is designated in writing partly by interspaces, partly by punctuation.2 But the influence of accent is by no means limited to giving prominence to those parts of spoken language which (in the way just described) receive the intonation, and form the exponents of the contents of a whole series, the illuminated peaks, as it were, which tower up out of the obscure mass of words. It embraces within its

sphere the whole mass of words, its intonations being graduated and classified according to their logical relations. And only by this fact can be explained the secret of its power,

1 Cf. Wm. von Humboldt: Ueber Entstehung der grammatischen Formen, in den Abhandlungen der Berliner Akademie 1824, p. 423. According to its principle the accent ought strictly always to fall upon that syllable which is for the notion the chief syllable, which constitutes the logical centre and the kernel of the word, therefore the stem syllable. But this is the case actually only in the German language, where the terminations have been by degrees subdued and suppressed. In other languages, where this has not been accomplished, the accent must depend on the strength of the final syllables, according to a phonetic (rhythmical) law. See below.

2 But only the Hebrew writings of the Old Test. in their present form designate the divisions of the sense within the period (the so-called verse) simply by its system of accents, and so by that which constitutes the principle of the division. Respecting the ancient designation of the larger sections and their historical development into the present accentuation, see the above-mentioned Essay in the Stud. und Krit. 1837, p. 836 sq., and Heb. Gram. §§ 18-22.

viz. that it is able to blend a plurality of sounds into unity, and to mark single sounds as exponents of the contents of the whole series. This is done simply by making prominent one of a series of sounds, and thus giving to the others a point around which they may group themselves (in ascending and descending gradation), forming, therefore, a centre and kernel, and thus producing a unity of sound, which represents to the ear the unity of idea. In the case of the verbal accent, as related to the other syllables of the word, this is at once obvious. Inasmuch, now, as the accent of the sentence is nothing else than the verbal accent of the principal word, and therefore, in order to be the accent of the sentence must make itself more prominent than the verbal accent of the other words of the sentence, and inasmuch as the same is true in a still higher degree of the accent of the leading sentence in a period, there results at once a gradation, a rank, among the accents, according to the logical importance of the sphere of each in relation to the whole. At the same time, however, it is, as a matter of course, to be inferred from this that the gradation is not confined to the tones of the principal words of the sentence, but extends to that of all the other words; in short, that the tone of every word accords with its logical relation to the whole. And this is fully confirmed by a closer consideration and comparison of the accents with which the separate words are spoken.1 Furthermore, a similar difference will show itself in respect to the duration of the tone, or of the rapidity of the movement of the voice among the different parts of the discourse. As in every word the unaccented (earlier) syllables hasten towards the accented syllable, so in every sentence the accent hastens from one word quickly to another more closely connected with it in sense, or dwells longer on another, and separates it from the rest, in accordance with the notion it is aiming to express; here pressing towards the chief word of the sentence as its highest point; there calmly passing it by, and gradually sinking down. Thus, out of the rough image

1 Of this more below, in treating of the rhythmical principle of speech.,

of articulate sound there rises up a finer, more spiritual, as it were a rectified image, which, with its infinitely fine gradations and shades, presents exactly the order of the notions (the logical relations), as well as the relations of the feelings, involved in the discourse. These gradations of the accent are primarily gradations of the force or strength of the tone of the voice; but, since every increase of force is also connected with a slight raising of the scale, there results at the same time a certain melody in speech.

Thus far the principle of accent, and of the melody which it introduces into speech, seems to be a purely mental one, concerning merely the understanding and (in so far as the emotions participate in it) the feelings. Accordingly one might think that accent is something voluntary, arbitrary, which may be used or omitted at pleasure; an ornament, or an accomplishment, without which, indeed, speech does not fully express what is in the mind (therefore essential to the perfection of speech), but which one may in many cases forego. But this would be a complete error. Accent is rather a physical necessity, which in speaking we cannot avoid, even if we would. Even when one takes pains to speak without accent (e.g. for the sake of affecting gentility, or of concealing his feelings), he can do no more than to diminish the gradations of it as much as possible and make them unnoticeable, but cannot entirely suppress them. Accent must therefore grow out of a law of nature, to which the voice in its progress is bound. And this law is in physics well known as the law of motion in all fluids. It is the law of undulation, of fluctuation. That the voice also moves according to this law, that its course is "undulatory," i.e.

This, by the way, is etymologically the fundamental notion of the words for motion: In German bewegen, related to wogen, wage, wiege, wagen, etc.; Latin, ago (properly vago, Aeol. Báyw, from which vagus, vacillo, wackeln, cognate veho, oxos, wagon, vectis, scales); from it agina, scales, aris, axle, arilla, shoulder (contracted, ala), especially of things which move around a fixed centre; again,

os (i.e. equiponderant); oculus, Sanscr. akshas, eye (from its rolling, axle-like motion), etc. Cf. my Essay in A. Kuhn's Zeitschrift für vergleichende Sprachforschung, VIII. 370 sq.

constantly rises and falls, is elevated and depressed, was long ago observed,' and the law in this application known by the name of rise and fall.2 But how does the voice come to be subject to this law? The more immediate cause lies in the so-called beating of the pulse or heart, i.e. in the undulatory strokes of the blood, and of the breathing which stands in reciprocal relation to it. For speaking is an action connected with expiration, and is produced by sounds of various kinds being elicited from it, as the air pssses through its canal at two principal points (the head of the windpipe and the mouth) through the co-operation of the organs there situated. The tones produced at the first place are clear tones (vowels); the others are sounds, more or less perceptible, which serve as accompaniments to the former (hence called consonants), and together with them form a single sound or syllable. It is consequently clear that, when one speaks, the supply of breath (i.e. of the air thrown out after each inhalation) which is expended in the production of articulate sounds, is divided into as many parts or single expulsions of the breath as there are separate members or single sounds in the discourse. Since, however, the breath expelled proceeds from a source characterized by undulation, i.e. from the beating of the heart, its separate expulsions cannot flow out in a uniform, smooth stream, but must constantly rise and fall in waves, like those of the blood in the beating of the pulse, i.e. exhibit a constant alternation of strength and weakness, elevation and depression. And this alternation expresses itself, of course, in the tones produced by it, primarily therefore in an alter

1 E.g. by DeWette, Introd. to Commentary on the Psalms, No. VII., p. 52 of ed. 3. (It is wanting in the 4th ed.)

2 Probably borrowed from the term used by the Greek grammarians, &pois and és which is applied by Priscianus and modern writers on metre (especially by Bentley, in the Schediasma de metris Terentianis, I.) to the voice, but properly refers to the movement of the foot or hand in beating time, and has, therefore, just the opposite meaning, viz. Oéos is the stroke of the foot or hand (Latin ictus) which accompanies the accented part of the measure; pois is the unaccented part. So sublatio and positio in Quinctilian.

& Vid. my Essay "Von der Natur und den Arten der Sprachlaute" in Jahn's Jahrbücher der Philologie, 1829, Vol. I. No. 4, p. 451 sq.

nation of strong and weak syllables, and thus manifests itself as a law of the movement of the voice. This is the natural law from which the so-called accent or tone proceeds, which, in this aspect of it, is nothing else than those elevations (summits) of the waves of the breath and voice, or of the stronger expulsions of the breath, which, alternating with weaker expulsions or depressions of the voice, produce in speech the antithesis of tone and tonelessness, of accented and unaccented parts of speech, like the antithesis of light and shade. Now, this antithesis, and its regular, constant return, is in speech, strictly speaking, what is designated by a Greek word, much used but little understood, rhythm 2 (Lat. numerus or numeri), and is the same thing as measure. Accent, as the climax of this, appears accordingly to be of rhythmical origin and nature; i.e. the origin of it, as well as the law of its movement, is not chiefly logical, but physical, i.e. traceable to the rhythm or undulation of the blood and breath, and hence of the voice.

-

That this is the origin and character of accent is shown by observing in all languages at least in all which have long and short syllables and any definite accent at all the rules respecting the position of the accent, or the determination of the location and quality of the stress. This whatever influence etymology and composition may in particular cases have upon it is everywhere subject to, and conditioned by, the higher law of rhythm or statics. Inasmuch as this law of accentuation has, so far as I know, not been sufficiently

1 Voice is, properly speaking, the clear tone (so vox, pwvh) produced in the glottis, contained in the vowel (hence vocalis, perheis) and constituting the lond, sounding element of the language; then in a wider sense, instrument of speech in general. The breath in its modifications just mentioned works primarily only on the tone of the glottis, or a vowel which it produces; but, since this is the soul of the syllable, and of language in general, it works by means of it on the whole language.

2 Etymologically it denotes (from ¿éw, to flow) a stream = ῥεῦμα. Applied to the flow of speech, it must designate either, as Buttmann thinks, an easy, flowing motion, or, since that is too vague for the figure and the thing meant, the wavelike, rocking, uniform rise and fall of the motion, and so just the essence of the so-called rhythm. Cf. above, note 4.