I’ve continued to work on the Schubert piano sonata I described in one of the least-read AUFS posts in history. I feel pretty confident on the first movement, though there are still places that need work. Now most of my efforts are directed at the second movement, which begins on pg. 12 of this PDF score.
Its most striking feature is a repeated background pattern in the left hand (modelled by the four C-sharps in the first measure). Being familiar with the piece from recordings, I didn’t find it difficult to execute, though it might have been hard to figure out how it all fit together without the recording. Yet it does present a question of interpretation: should the top note, which crosses over into right-hand territory, sound like a part of the melody? To use the first measure as an example: should the top C-sharp sound like it’s leading into the chords the follow, or should there be a marked distinction between the left-hand pattern and the right-hand melody? Alfred Brendel’s recording (available on Spotify) opts to incorporate the left-hand part into the melody, but I prefer to highlight the contrast — playing the final left-hand note more softly, as though it’s a distant background chime.
One could make this decision solely on the basis of personal inclination, but I prefer to ground my interpretation in my understanding of the first movement. One of its strangest aspects is a seemingly “random” G-flat trill in the left hand, which halts the action twice during the piece (on the introduction of the main theme). Only in its last appearance (also on pg. 12 of the PDF) does the right hand “acknowledge” it by playing a chord in sync with the F on which the trill resolves. It’s almost as though a fly that has been buzzing around the piece has been swatted — although as I note in my previous post, the odd G-flat trill does seem to introduce a kind of derangement into the piece, which includes unexpected dissonances and which undergoes a sudden shift into a more somber mood toward the middle, in a section that seems to anticipate the second movement.
It’s appropriate that the second movement should have its own “alien body” in the form of the left-hand pattern — and also appropriate that it should be more tightly integrated (through fitting with the dominant key signature, for instance, and shifting to fit with the harmonies). Yet have we really arrived at a point where that foreign element is fully incorporated? I don’t think so, at least not until the end of the movement, when the rough edges have been smoothed off (in the form of a rhythmic shift that hits each beat in the measure, as opposed to the jerkier early version). Making that top note into a part of the melody seems to me to be forcing matters prematurely.
An und für sich 
Sonata, not concerto.
Fixed, thanks.
I would also highlight the contrast between the arpeggiated octaves in the left hand and the chords in right — but for the opposite reason. In the first section, notice the keys that predominate in these left-hand octaves: C# and G# at first, followed by E and B. These are obviously the tonic and dominant scale degrees for C# minor and E major, respectively. This wouldn’t be a big deal to point out unless we compare this to the right hand. In the right hand, there doesn’t seem to be any resolution towards a nice, satisfying C# minor chord (the key signature of the movement). Actually, the note of C# itself only occasionally appears in these right-hand chords — when it does, it takes on a subsidiary role in, say, an inverted A major chord. What’s happening in the right hand is a downward stepwise progression that is willing to repeat the motion after 2-3 steps in order to avoid a C# in the melody (the upper voice of the chords). Another way of seeing this is the behavior of B# in these chord progressions. Being the leading tone in a C# minor scale, a B# would “naturally” want to resolve up to the C#. But in this movement, Schubert suspends that potential resolution before repeating the stepwise motion yet again.
I guess this is a long way of saying that rather than viewing the left hand as a foreign body, you should see it as structurally necessary in giving this second movement an identity. Try playing the right hand by itself: without those crucial C# octaves (or G#, E, and B, depending on where you’re looking), the key signature becomes ambiguous because of the right hand’s self-sabotaging nature. If you jump to the third section of this movement, the left hand at this point almost seems to be mocking the impotence of the right hand, as it now acquires a 3-note sixteenth note pattern that includes two C# followed by the leading tone B# that subsequently resolves to yet another C#. Contrast all of this with the middle section, where notes are repeated to the point of redundancy in both hands while the right hand is able to resolve to an A (the key signature of this section): conflict is absent.
If I were playing, I would probably play the outer notes of the 4-note pattern (with their staccato markings) a little more softly than the syncopated inner two notes because of their proximity in time to the playing of the right hand.
If you’re interested, you should listen to the second movement of Ravel’s Gaspard de la nuit, where this separation of the tonic key from the surrounding chordal harmonies becomes the primary aural effect one listens for (an external Bb octave is sustained throughout while the surrounding chords have to accommodate for the corresponding lack of a Bb, forcing the harmonies into unusual directions). I probably overanalyzed here, but I hope this helps.
The great Schubert B-flat Sonata is the rightful envy of all non-pianists like me, and the equal of any in the repertoire. You must be enjoying the heck out of this experience, MAKARIE!
Now Alfred Brendel is of course a very successful pianist and much lauded abroad, but I have never found his interpretations arresting or satisfying. Even among relatively pedal-heavy pianists, I prefer the sound and style of his fellow Austrian and coeval, Ingrid Haebler. For this particular sonata, I would rather recommend the much superior (and more interesting) readings of Annie Fischer, Lili Kraus, Tatiana Nikolayeva, Ingrid Haebler (with minor reservations), and Andreas Staier (fortepiano). All these are on YouTube, and all agree with you, though each in a personalized way.
You are right to point to the analogy of the ectopic G-flat trill in the first movement, which forms a greater exo-consistency of the “unassimilable” from ALL the movements: the limping figure in the left hand in amidst the hemiolas in the Trio of the third movement, and the disruptive midnight gongs of the fourth. These uncanny gestures are indispensable to the compellingly asymmetrical aesthetic and the architecture of the piece. But the internal reason for your interpretive choice is clear: The c# (and its successors in the passage) in question is part of an ostinato figure grounded in the ponte (a static harmonic bridge subtending a whole passage) used throughout the A section of the movement, and therefore belongs with the dotted, octave gestures of the left hand – in a crossover technique hearkening back to the toccata tradition of the Baroque, when it would also involve the use of a second (higher) keyboard.
Question remains though: Why all these “extraneous” touches? One could well approach this as a Benjaminian Stillstellung, the creation of a Jetztzeit disrupting ordinary chronos. In Schubert’s context, it can be seen as both an aesthetic and a political gesture, the former breaking free of the classical molds he inherited, and the latter puncturing the superficial smoothness of life in the Biedermeier cultural-political milieu. Of course, Schubert wrote this sonata as he was approaching his own final Jetztzeit, and with it diverted strikingly from a gray march to the syphilitic grave.
Thank you both for your detailed responses — you’ve given me a lot to ponder, as well as some recordings to track down. As a practical matter, I do wind up doing what calvinpv suggests — playing the top note of the arpeggiated octaves (thank you for the term, I wasn’t sure if it was proper to call it an arpeggio when it’s just octaves) softer to avoid conflict.
I probably should have clarified above that by tonic, dominant, and leading tone, I just mean the 1st, 5th, and 7th steps in a major/minor scale, respectively. The tonic and dominant scale steps in particular are critical to grounding tonality in the Classical and Romantic periods — if they have no prominence in a musical work, then you’re likely listening to an Impressionist work, where the tonality is wandering aimlessly trying to find its proper home.
As for the trill in the first movement, I think David’s remark about hearkening back to the Baroque can also apply here. As far as I know, musicians in the Baroque period had the freedom to add ornamental flourishes not formally marked on the score. In the Schubert sonata, he clearly wants to extend the duration of that F major chord in the right hand for as long as it’s tolerable. If the left hand were to instead play a trill-less Gb, the result would be underwhelming, to say the least. What I find puzzling about the trill is not the Gb or the F on which it lands, but the Ab. The F should be expected; in fact, its ubiquity in the opening theme is almost nauseating. The Gb, meanwhile, is often a useful chromatic device that provides a smooth transition from passages in Bb major (where F is predominant) to passages in Eb major (with its G natural). But the trill misfires and includes an Ab, bringing the momentum to a halt, as you put it.
Anyways, glad to help.
The Gibet movement (II) of Ravel’s Gaspard de la nuit (where the image is presented of a man hanged on the gallows against the noonday sun) does have the same kind of sustained ponte and ostinato figure, and recordings of note include those of Robert Casadesus (under Ravel’s guidance),Walter Gieseking, François Samson, Werner Haas, Arturo Benedetti Michelangeli, Ivo Pogorelich and Pierre-Laurent Aimard. Of these, the first (1922) is the boldest in hammering out that persistent b-flat – if the piano roll be any index of the original balance. Most, however, keep that b-flat fairly subtle.
The low G-flat trill in the Schubert is certainly jarring to the ear, but hardly resistant to harmonic analysis. It is clearly a “Neapolitan” (a harmony based on the flatted second) embellishment to the dominant (F), whence it departs and to which it resolves. This effect is usually perpetrated on the flatted 2nd scale degree in the main key of a piece, as at the very end of the great C-major Quintet, composed in the same last year of Schubert’s life. It’s a way of queering the conventional harmonic expectations, which he did by so many other compositional means – not least in abrogating (almost) the distinction between the major and minor modes. In this trill, the G-flat also gestures in that direction.