A more important step toward the true sonata style was made by Philipp Emanuel's less romantic brother, Johann Christian, who settled in London, founded the Bach and Abel Concerts, and had a great influence on the boy Mozart. J. C. Bach is the first com poser to lay a dramatic emphasis on the transition between his first and his second group. In crude or deliberately formal examples this has been wittily described as "presenting arms" to the new key. Its point is not that there is any difficulty in appre hending the new key, but that the move into it is dramatic and not decorative. Whether the move be made with intellectual music or with common forms makes no difference. Beethoven preferred, in his most characteristic early works, to disguise it cleverly. In later works he acquired the grand formal breadth of Mozart's chamber-music in this transition.
There is a first group in the tonic, followed by a transition to another key, where there is a second group that usually ends with a neat little cadence-theme. These groups constitute the exposition, which may be repeated. Then follows the development, the function of which is to put the previous materials into new lights, regrouping the figures into new types of phrase, modulating freely and settling, if at all, only in new keys. Eventually a return is made to the tonic, and so to the Recapitulation. This recapitulates the exposition, but it gives the second group in the tonic, and so completes the design. The development and recapitu lation may be repeated; a coda may follow the recapitulations.
This account has required so many words that the illusion is apt to arise that it conveys more information than, say, the statement that the plan of a cathedral is cruciform and that the arms of the cross are called transepts, and so on. It gives us no means of distinguishing an ambling decorative movement by Boccherini from the first movement of a Beethoven symphony; and the description of the development is the only point which would rule out the sequel of our second example as a specimen of sonata-form. Haydn, Mozart and Beethoven differ widely in their handling of every part of the scheme.
The most regular form is to be found in Mozart, whose transi tions are always broad and smooth. The effect of "presenting arms" is evident only in small or perfunctory works; and if it is found at all in larger works it is on such a scale and with such a purpose as Beethoven would give it. The second group contains at least one definite new theme and a number of cadence-phrases in various rhythms. The development is short, consisting of one broad sequential process that leads through a wide range of keys back to the tonic. Sometimes it contains an entirely new theme.
Such an episode, which is generally placed at the beginning, by no means always indicates a lighter style and texture. It may be a relief from unusually concentrated figure-work in the expo sition; and the developments of two of Mozart's most serious works (the C minor serenade for 8 wind instruments, better known as a string quintet, and the G minor pianoforte quartet) are episodic. The return to the tonic always has the effect of being accurately timed after a delightful period of anticipation.
The recapitulation is full and has a deceptive appearance of regularity. In reality it is anything but mechanical. It has just that kind of difference by which stereoscopic pictures produce the effect of binocular vision. In the light of the recapitulation the listener finds that points that were superficial in the exposition have now become solid. The composer instinctively conceives his exposition in relation to the question "How will this sound when it returns?" The minimum change happens automatically with the transition to the second group, for this transition must no longer lead to the complementary key of the exposition. One quaint primitive device in the transition is that of making it not leave the tonic at all but simply come to a pause on (but not in) the dominant. This dominant is then taken literally as a key. In such a case the recapitulation need alter nothing; the second group merely follows in the tonic instead of in the dominant. Even this automatic device makes the recapitulation give a more solid im pression than the exposition; for the pause on the dominant, treated paradoxically in the exposition, is now treated rationally.
We need not deny that formal devices are apt to become mechanical; but we have no right to the a priori opinion that Mozart is writing unimaginatively every time that he decides that the most familiar course is the wittiest. It is much wiser to regard the most exact recapitulation as the extreme case of deli cate balance, and even in the most exact the crucial detail will appear. Here is a case in a difference of a single bar, Mozart's string quartet in E flat (K. 428) has the following clause in the first theme : we ought to say that he invented the most brilliant type of Beet hoven's coda. And these features of his form are not, as has sometimes been alleged, primitive. They are only partially visible in quartets before Op. 5o. Then they appear in full vigour, and Haydn's admiration for Mozart only confirms him in his inde pendence.