FIGURE XII. EXAMPLES OF PHRASES. Gavotte from BACH’S Fifth French Suite.
Bourrée, from BACH’S Third Suite for 'Cello.
It will be seen at once that in each case the second phrase answers or supplements the first. Like it in length and in general contour, it is at the same time more positive and final, so that the combined effect of the two is much like that of a couplet of verses. The first phrase, in fact, arouses in our minds an expectation, which only the second can satisfy; so that we have here a new and larger application of the now familiar device for binding together successive impressions. So characteristic is the supplementation of one phrase by another that theorists have adopted a set terminology suggested by it, calling the first phrase in all such cases the “antecedent phrase,” and the second the “consequent phrase.” It will also be noted, however, that the pair of phrases, once heard, becomes itself a unit in the mind, and arouses a new expectation of further matter to establish a still larger balance; and a reference to the pieces of Bach cited will show that Bach in each case follows up his pair of two-measure phrases by a four-measure phrase which supplements them as they supplemented each other. And so the process goes on, the piece growing ever larger and more complex by a regular accretion, until at last a phrase of definite and entire finality is reached, and the movement stands complete. All short songs and dances illustrate this progressive accretion of phrases into larger and larger units, by means of a constant unfolding of new expectations and fulfilments. To trace it out, to analyse what the composer has so ingeniously built up, is one of the most fascinating of studies; for it shows us how the simplest song is organic like a crystal, a flower, or an animal.
It is neither possible nor desirable to lay down here any rigid rules as to the metrical or harmonic relationships between the phrases. Generally, the metrical balance is fairly simple; a two-measure phrase is usually answered by another of the same length; two such phrases are often answered by a single four-measure phrase. But sometimes four measures are answered by two; and not infrequently three- or five-measure phrases appear unexpectedly but with quite satisfactory effect. The sense of balance must be given—that is all we can say: just how it shall be given will depend, as Mr. Weller would say, “on the taste and fancy of the composer.” As for the harmonic relationships, endless variety is possible. Yet we may here point out certain general principles. Every phrase, as we have seen, ends with some sort of a cadence, strong or weak according to the harshness of the dissonance it contains and the nearness of its final chord to the tonal centre, or key-note, of the piece. Now, as the salient tones of any key are its tonic and its dominant, the most obvious and natural course for the composer is to embody these in the successive phrases; and as the tonic conveys the impression of finality it is natural to use that last. A glance at Figure XII will show that Bach makes his antecedent phrase, in the first instance, end with a tonic chord, but a weak one; in the second instance, with a dominant. In both cases the consequent phrase ends with a strong tonic. Thus the harmonic as well as the metrical relations produce the effect of expectation and fulfilment, of antithesis between a transitive and a final impression. This is the general principle of all harmonic structure. The final impression is given by a strong tonic chord; the mediate impression, arousing the sense of anticipation, is given by some weaker and contrasting harmony, in the vast majority of cases the dominant chord. A full sense of the inexhaustible capabilities of this sort of harmonic structure can be gained only by a careful analysis of many pieces such as the movements of Bach’s suites. To this the reader is recommended.
When once composers had grasped the possibilities of structure by means of harmony, they quickly proceeded to work them out in the large, as applied to a complete musical form. They began to organize whole pieces by means of a grouping or ordered antithesis of different harmonic centres. Working without models and in the dark, they made many false starts and wrong moves, they tried many hybrid and unstable forms; but eventually, in the course of years of experiment, they developed two great types of structure, based on fundamental principles, and embodied, with unimportant minor modifications, in almost all the suite-movements of the seventeenth and of later centuries. The first of these two great general types of structure, called Binary Form, contained two distinct members or sections; the second, called Ternary Form, contained three sections.