'It may be shown,' says Mr Herbert Spencer,[4] 'that Music is but an idealisation of the natural language of emotion, and that, consequently, Music must be good or bad according as it conforms to the laws of this natural language. The various inflections of voice which accompany feelings of different kinds and intensities, are the germs out of which Music is developed. It is demonstrable that these inflections and cadences are not accidental or arbitrary: but that they are determined by certain general principles of vital action; and that their expressiveness depends on this. Whence it follows that musical phrases, and the melodies built on them, can be effective only when they are in harmony with these general principles. It is difficult here properly to illustrate this position. But perhaps it will suffice to instance the swarms of worthless ballads that infest drawing-rooms, as compositions which science would forbid. They sin against science by setting to music ideas that are not emotional enough to prompt musical expression: and they also sin against science by using musical phrases that have no relation to the ideas expressed, even when these are emotional. They are bad because they are untrue. And to say they are untrue is to say they are unscientific.'
In these words we may find a starting-point for sound criticism. If a musical composition is to make any bid for the rank of classic it must, as a primary essential, be genuine in feeling: by which we mean, that it must not only be original, though originality is implied and included, but that, in Wordsworth's fine phrase, it must be inevitable. To recognise a melody as perfect is to feel, when we come to know it, that it could not possibly have been written in any other way: that its phraseology, whether simple or complex, whether obvious or recondite, is the necessary outgrowth of the thought which it embodies. Of course this law does not preclude the element of surprise, which is one legitimate factor of musical effect. The hearer, like the composer, may sometimes be 'stung with the splendour of a sudden thought' and roused into a moment of exquisite consciousness by an unexpected cadence or a new modulation. But if the surprise be more than temporary, it is inartistic. Before we reach the conclusion of the work, we must be convinced that the effect in question bears some vital and organic part in the total structure: that it is, in short, a prediction which is justified by a future fulfilment. And, in that case, we end by acknowledging that it was not an isolated and deliberate attempt to stir our wonder, but part of an established plan which only astonished us at the moment because we were unable to foresee its issue.