There is, therefore, a fundamental difference in this respect between single word-sounds and single colors or tones; they are not sufficiently impressive in themselves, not sufficiently separable from their meanings, to have anything except the slightest value as mere sounds. In collocation, however, and quite apart from rhythm and alliteration, this minute expressiveness may add up to a considerable amount. In Matthew Arnold's lines,
Swept by confused alarms of struggle and flight
Where ignorant armies crash by night,
the hardness and difficulty of the consonants in their cumulative force become an independent element of expressiveness, strengthening that of the meaning of the words. Or in Tennyson's oft-quoted line, "the murmuring of innumerable bees," the sounds taken together have a genuine imitative effect, in which something of the drowsy feeling of the hive is present.
Following the general law of harmony between form and content, the beauty of sound should be functional; that is, it should never be developed for its own sake alone, but also to intensify, through re-expression, the mood of the thoughts. The sound-values are too lacking in independence to be purely ornamental. Poetry does indeed permit of embellishment—the pleasurable elaboration of sensation—yet should never degenerate into a mere tintinnabulation of sounds. The rimes in binding words should bind thoughts also; the tonalities or contrasts of vowel and consonant should echo harmonies or strains in pervasive moods.
It is by rhythm, however, that the chief expressiveness of the mere medium is imparted to verse. But here again we shall find sound and meaning intertwined—a rhythm in thought governing a rhythm in sound.
Only as a result of recent investigations can a satisfactory theory of modern verse be constructed. The making of this theory has been largely hampered, on the one hand, by the application of the quantitative principles of classical verse to our poetry; and, on the other hand, by forcing the analogy between music and verse. The insufficiency of the quantitative scheme for English verse is not difficult to perceive. Such a scheme presupposes that syllables have a fixed quantity of duration, as either long or short, and that rhythm consists in the regularity of their distribution. But, although there are differences in the duration of syllables, some being longer than others, there are no fixed rules to determine whether a syllable is short or long; and, what is a more serious objection, it is impossible to find any regularity in the occurrence of shorts and longs in normal English verse,—in all verse that has not been written with the explicit purpose of imitating the Greek or Latin. An examination of any line of verse will verify these statements. Take, for example, the first three lines of Shakespeare's song,
Blow, blow, thou winter wind,
Thou art not so unkind
As man's ingratitude.
Here the quantitative scansion is perhaps as follows:—
- - - - U -
- - U x U -
U - U - U -
I have given the word "so" a double scansion because I conceive it impossible to determine whether it is really long or short. At any rate, there is certainly no regularity in the distribution of shorts and longs, except in the last of the three lines, and no correspondence, except in that line, between the quantitative scansion and the rhythmical movement of the verses. And whenever such a correspondence exists, it is due either to the fact that the incidence of stress tends to lengthen a syllable or to the fact that, oftentimes, in polysyllabic words, mere length will produce a stress. This is the modicum of truth in the quantitative view. But obviously stress governs, quantity obeys.