Schubert: the last piano sonatas, part 3

Today I am going to have a look at the first movement of the second of the three last piano sonatas by Schubert, the one in A major, D. 959. I've mentioned before the quite good Wikipedia article on these three sonatas, so I recommend reading it again. It contains an interesting discussion of how the first movement is structured. What sticks in my mind is how Schubert has a great deal of modulation in the exposition, but confines the development itself to the key of C mostly with occasional departures to B and C minor before settling into A minor. This is the tonic minor, A minor, which has the same key signature as C major, so all he has to do to begin the recapitulation is change to the major. One of Schubert's typical harmonic gestures is the juxtaposition of major and minor.

So what's special about this? The traditional approach to the development section is to begin by destabilizing the tonality and putting one or more themes from the exposition through some harmonic changes and fragmentation. But Schubert does more of this in the exposition! Schubert's themes tend to be fairly complex, often consisting of ternary structures. The first theme begins with this phrase:

Click to enlarge

The first phrase is the first three lines and it dovetails with a restatement of the opening, with a different continuation. These triplets, in a chromatic version, will make up the second part of this first theme. Notice that the phrase is fifteen measures long--the sixteenth measure begins a new phrase. Theorists tend to be a bit fetishistic about irregular phrase lengths: eleven measures long in Haydn or five in Mozart, and this is certainly interesting. But the reason it is interesting is because so many composers tend to keep cranking out eight or sixteen measure phrases without much variation. I tend to think that the real magic here is not that the great ones wrote irregular phrases, but that they made them sound so normal!

As Wikipedia notes, the second theme is more lyrical and in four-part harmony. The theme begins on the third measure of this excerpt, where it says pp:

This theme is ternary, meaning that it has three parts, the first one shown above, a middle part:

Click to enlarge

which is a further development of the chromatic triplets from theme one, and then a return to the first part. This excerpt I just quoted contains the kind of fragmentation and harmonic velocity that is usually reserved for the development section, but which, as I said, is not found in the development of this movement.

What Schubert has constructed is a kind of reversal of some of the basic characteristics of the first movement sonata form. In Haydn, Mozart and even Beethoven, a harmonically stable theme is presented, often followed by another in the dominant and this exposition is followed by a development in which the harmony is unstable and wide-ranging and the theme or themes are broken up and sequenced. While Schubert does indeed have two themes, the second in the dominant, each theme contains a development section with fragmentation and harmonic instability. But the development itself has little actual development. Instead, it is a more calm contrast with the exposition and recapitulation. The development sounds rather Mozartean, in fact.

Another unusual quality is the extent to which the subdominant is stressed, not just in the coda, which is normal, but in the first theme itself. Finally, Schubert ends the movement with an unusual cadence. Instead of the nearly mandatory V7 to I (in A major an E G# B D chord followed by A C# E) he has, in place of the dominant seventh, this chord: B flat, D G#, which is an Italian augmented sixth chord, normally used to prepare the dominant. The effect here is to again, bend things toward the subdominant, D.

But enough talk, let's listen to the first movement of the sonata. This is Rudolf Serkin whom I prefer because, unlike most of the other clips available, he repeats the exposition: