Here it is as if, after placing on his canvas the two main chords of the cadence, dominant and tonic, he took, while the colors were still wet, a brush, and with the softest imaginable touch drew it across the entire face of the picture. The grace-notes, most of which, it will be noted, are dissonant to the main harmony, are no more meant to be heard individually than the spots of paint in a Monet are meant to be seen individually; they are a running of the colors, blurring the otherwise too bald outline. Chopin's scores are full of these delicate veilings and obscurations. In a majority of cases they are produced, as in this instance, by the right hand, above a clear harmony in the lower register. But sometimes, more daringly,[23] he assigns the web of dissonance to the left hand, in the middle register or even in the bass, thus gaining an extraordinary lurid gorgeousness of coloring. The passage in the third ballade, beginning at the change of signature to four sharps (Figure XVIII), is an instance.
Or again, as in the "Meno mosso" of the Scherzo, opus 39, both hands first deliver bold, clear chords, and then weave a shimmer of light above them. In all such cases, it is obvious that the dissonances in question do not belong to the essential melodic and harmonic lines of the composition; they are, as Mr. Hadow says, "effects of superficies, not effects of substance," and may be compared to those local blurs made by a draughtsman's stump in a charcoal sketch, or, as before suggested, to those surprisingly rich mixed tints produced in impressionistic paintings by a multitude of minute brush-strokes.
The at first sight very elaborate modulations of Chopin which have provoked so much discussion are but a further application of the same principle. They are really not modulations at all, in the classic sense of transitions from one key to another having a structural value, but rather amplifications of the groups of grace-notes that constantly embroider the tunes. Their function is sensuous rather than structural, and we might describe them by coining the word "grace-chords." Of the twelfth measure of the second nocturne, for example, Mr. Hadow well says that "when we see it on paper it seems to consist of a rapid series of remote and recondite modulations, but when we hear it played ... we feel that there is only one real modulation, and that the rest of the passage is an iridescent play of color." Another striking instance is the following measure in the "Polonaise-Fantaisie," a composition in which effects of this sort abound.