With this, the companion piece for our popular Integral Edition Series post, we continue and conclude our survey of Franz Schubert's piano sonatas. This post concerns all the largely unfinished or posthumously published works, none of which appear in the Integral Edition Series. Schubert wrote a lot of unfinished works that were fortunately not burned, as was often the fate of lesser or unfinished works by others. It's difficult to properly number these works, though one can always rely on the Deutsch numbers. We'll continue numbering them after the first set, which ended with Sonata #15 being the B Flat Major D. 960, which Schubert completed in September, 1828, just weeks before his death. You'll find in this set all those works that you've perhaps heard before, yet they were missing from the first set.
Sonata #16 in E Major (1816-18) D. 549, D. 549AConsidered a fragment, this piece became a five movement work, the missing movements apparently found later and the first publication occurred in 1843. It is available, complete and beautifully laid out, in the third volume of Henle's urtext series edited by Paul Badura-Skoda. This E Major Sonata would be contemporaneous with the 1817 sonatas, so it makes many allusions to Mozart, Beethoven and Haydn. But this is the fourth voice of the Viennese classical style, the voice of the singer. One finds all kinds of textures and moods in Schubert. Think Mozart when you're playing these, because if you over play any of it, much of the music's uniqueness will be ruined. Just because there are a lot of chords doesn't mean that we want to understand by this the instruction to bang a lot on the piano. Nothing like that is ever intended. This music is a distillation of the styles of the other classical masters and even those who came long before them. Subtlety abounds. I chose Wilhelm Kempff's performance, as it superbly exemplifies all these characteristics.
1. Allegro moderato
2. Scherzo: Allegro
4. Scherzo: Allegro - Trio: Più tardo
5. Allegro patetico
Sonata #17 in f sharp minor (1817) D. 571, D. 604, D 570
This sonata is concocted of fragments, the first and last movements happen to lack recapitulations, which have been supplied by various editors. It's apparently not that hard to complete Schubert sonata-allegro forms, since in his finished works, he rarely deviates from his method, so that's what's frequently done. Some pianists, actually stop playing after the verifiable music actually penned by Schubert runs out. Richter and Brendel did this. It can be done as acceptable performance practise, but I've always thought it somewhat odd; if you're going to play a fragment, might as well make the best of finishing it if you can. Besides which, and perhaps this was even more true in earlier more romanticized periods, when one hears the music suddenly stop and knows its because the composer didn't get around to finishing it, one can then afford oneself the lament that only if the poor fellow had managed to live longer, that he'd have finished it himself. Believe me, as it is, plenty goes unfinished. One always hopes to accomplish what one can.
Victor Stanislavsky plays the first movement, including a recapitulation that is similar to that found in the Badura-Skoda Henle edition. The Andante is complete and is played here very well by an unidentified pianist. I don't usually like including unauthenticated performances, but we're dealing with the rare and difficult here. We want to give our readers an idea of the composition, so that perhaps they might make something of it themselves.
For the third movement, I chose a period instrument rendition by Trudelies Leonhardt. It gives you the sense of the times in which Schubert worked, with the instruments he had available, a transition period where you can really see how he was dealing with how to mould and shape the old dance forms with new harmonies and textures. The fourth movement is played by Alwin Bär who has the recording fade out as the development section nears its end. This piece is completed in the Henle edition, so I suppose you'd have to get a copy to find out how it ends.
1. Allegro moderato [D 571]
2. Andante [D. 604]
3. Scherzo [D 570]
4. Allegro [D 570]
Sonata #18 in C Major (1818) D. 613, D. 612
This sonata is also a concoction of fragments. The first movement is again offered here played quite competently by an unauthenticated pianist. It lacks a recapitulation, and one is supplied here. The one in the Henle edition will be somewhat different only near the end where Badura-Skoda provides a more conservative rendering. The second movement, a flowery period Adagio, is complete and here played by Peter Frankl. I regret not finding a recording of the last movement. That also can be found in the Henle edition, its recapitulation completed by Paul Badura-Skoda.
1. Moderato [D. 613]
2. Adagio [D. 612]
3. Allegretto [D3 613 said to be a Siciliana]
Sonata #19 in f minor (1818) D. 625
This is one of the more outstanding sonatas among the unfinished ones and one of the most recognizable. Richter played it, at least the parts actually written by Schubert. Paul Badura-Skoda played the version of it he completed in the Henle edition and his version is presented here. The middle movements are sometimes switched. Either way, the punchy scherzo theme offers a kind of shock to whatever preceded it. The Scherzo is complete, as is the Adagio. The fourth movement is missing the left hand part after a while, so Badura-Skoda completed it.
2. Scherzo: Allegretto – Trio
3. Adagio [D. 505]
Sonata #20 in C Major “Relique” (1825) D. 840
This is another amazing discovery, the latest to be found unfinished work in this genre by Schubert. It was given the name “relique” in the belief that it was his last written piano sonata, but it was in fact written three years earlier in 1825, and left to gather dust somewhere until it was discovered, put together and first published as a fragment in 1881, the same year Brahms' second piano concerto was premièred and Béla Bartók was born. It's first two movements were complete, and that's how much of it most people ever get to hear. But there were enough of the last two movements to nearly complete them. We are pleased to have Sviatoslav Richter's nearly complete performance of this recreated work. He quits playing when the last of the real Schubert ends and where editors like Badura-Skoda have completed the work.
And believe it or not, there's actually one more; a completed version of Sonata #7 in E Flat Major D. 568 except it's in D Flat Major, D. 567, and there are some other changes to some of the music in the middle movements and so on. It's considered an earlier version of the more popular E Flat sonata (our #7). I wasn't able to find any recordings of it and as far as I know, no one ever plays it. It's included in the third volume of the Henle Badura-Skoda edition, should you wish to become the first to claim it as your own.
Thank-you once again for coming along with us to investigate the unfinished works of Franz Schubert. We're lucky to have them. We'd be even more lucky to hear newer and better performances and recordings of all of them.