Drei Klavierstücke D.946 (1828)

Written just five months before Der Hirt auf dem Felsen, the three pieces first published forty years after the composer’s death as Drei Klavierstücke were probably intended for a third set of impromptus. Certainly, the Allegretto in E flat major - a kind of gondola song that meets a storm in C minor and is all but displaced by a more extended second episode - is as spontaneous as anything he ever wrote. The Allegro in C major - a scherzo animated by Hungarian syncopations and held in check by a broader middle section - is comparatively straightforward in construction.

…It was also because of Brahms that Schubert’s Three Piano Pieces finally achieved publication, the first of them with an Andantino episode which is obviously crossed out in the manuscript.

The Three Piano Pieces were written in May, 1928, presumably with a fourth in mind to make up another set of Impromptus. The set was never completed, however, and the three pieces were generally ignored until Brahms (anonymously) edited them and saw them through the press in 1868. As Schubert left it, the first of these pieces, in E flat minor, is an expansive ternary movement. The Allegro assai outer sections and the decorative Andante in the middle - its B major tonality approached and left by way of some remarkable modulations - are themselves ternary structures. It was perhaps because of this symmetry that Schubert decided to leave out the A flat major Andantino. It originally appeared after the reprise of the Allegro assai, which was recalled again at the end. As Brahms published it, including the cancelled but irresistible Andantino, the piece become a kind of rondo.

The pieces published as Drei Klavierstücke (Three Piano Pieces) in 1868, forty years after the composer’s death, were probably conceived as part of a new set of Impromtus – a form which Schubert had made his own and which had already brought him a little commercial encouragement. In the absence of any title on the mansucript, which dates from May 1828, one cannot be certain, but the pieces do have more in common with those of the two complete sets of Impromptus than any of his sonata movements. The fact that he cut out an Andantino section from the first in E flat minor, apparently to contain it in within impromptu proportions and in ternary form, seems to support the idea. He made no such cut, however, in No.2 in C major – which probably explains why some pianists, reluctant to observe Schubert’s cut, restore the irresistible Andantino to its original place near the end of No.1 in E flat minor, extending a ternary construction to a kind of rondo. After all, No.2 in E flat major would be another ternary construction, with gondola-song outer sections and a stormy middle section, if Schubert hadn’t added a whole new episode before the end. There is no problem, however, with the short Allegro in C major, a scherzo animated by Hungarian syncopations and held in check by a broader middle section.

Apart from the dances, Schubert’s piano music had little appeal for the publishers of his day. Only three of the sonatas were published in his lifetime (two more were issued in the year after his death) and the only other piano pieces he saw in print were the “Wanderer” Fantasy, the six Moments Musicals and two of the first set of four Impromptus. It was probably because of the partial success of the first set of Impromptus that he wrote four more in 1827 and completed three of another set in May 1828. While it is true that the manuscript of the 1828 pieces bears no title - which is why they were eventually published as Klavierstücke or Piano Pieces - they are clearly impromptus in all but name.

    The first of them, in E flat minor, is a particularly compelling inspiration. In the standard edition it is presented as a simple, though expansive ternary structure, the Allegro assai outer sections propelled by an impulsive galloping rhythm and a theme obsessively centred on one note. The middle section is a contrastingly serene and decorative Andante in B major which Schubert links to the reprise of the Allegro assai by an even more interesting modulation than the one by which he approached it.

    At one time the composer seriously considered adding an A flat major Andantino at this point and then another reprise of the Allegro assai. Although it is quite definitely crossed out in the manuscript, the Andantino is both irresistible and perfectly legible and is printed in some editions of the piece. It will be included in today’s performance.

…had he lived to see them, Schumann could not have taken any of the Three Pieces for a first movement of a sonata when they were first published in 1868. They were written a few months before Schubert’s death, possibly for a third set of four Impromptus, which he did not have time to complete. Whatever they are, each one is a good example of what a Schubert first movement is not. The first of them, with its persecuted E flat minor beginning, resembles a Chopin scherzo as much as anything - a resemblance which does not end with the complete contrast offered by the B major song in the Andante middle section. Whether the scalic flourishes in this section and the tremolos in right and left hand are orchestral or pianistic is a continuing matter of controversy.

There are some particularly interesting piano sounds in the first trio of the second piece. The main section is a conventionally written Gondollied in E flat major. Then the key changes to C minor, the left hand rumbles at a distance, the right hand shivers in parallel thirds and punches out chords in contradiction of the prevailing 6/8 metre. This stormy episode, which is more colouristic than thematic, develops gradually more dramatic intensity until the tension is relaxed in a modulation to C major - with a gesture which evidently left an impression on Rachmaninov.

In comparison with the second piece, which is complicated by the second trio in A flat minor with its own middle section in B minor, the third piece is simple in construction. It might even pass as a sonata scherzo, with syncopated outer section sin C major and a quieter, rhythmically regular but boldly modulating middle section in D flat major. Certainly, the transition from the middle section back to the main theme is, in contrast to the obvious links of most impromptus, worthy of any sonata.

Apart from the dances, Schubert’s piano music had little appeal for the publishers of his day. Only three of the sonatas were published in his lifetime (two more were issued in the year after his death) and the only other piano pieces he saw in print were the “Wanderer” Fantasy, the six Moments Musicals and two of the first set of four Impromptus. It was probably because of the partial success of the first set of Impromptus that he wrote four more in 1827 and completed three of another set in May 1828. While it is true that the manuscript of the 1828 pieces bears no title – which is why they were eventually published as Klavierstücke or Piano Pieces – they are clearly impromptus in all but name.

The first of them, in E flat minor, is a particularly compelling inspiration. In the standard edition it is presented as a simple, though expansive ternary structure, the Allegro assai outer sections propelled by an impulsive galloping rhythm and a theme obsessively centred on one note. The middle section is a contrastingly serene and decorative Andante in B major which Schubert links to the reprise of the Allegro assai by an even more interesting modulation than the one by which he approached it. At one time the composer seriously considered adding an A flat major Andantino at this point and then another reprise of the Allegro assai. Although it is quite definitely crossed out in the manuscript, the Andantino is both irresistible and perfectly legible and is printed in some editions of the piece. It will/will not be included in today’s performance¹.

The E flat major opening section of the second piece has the gentle flow of a Mendelssohn Gondellied. But then the key changes to C minor, the left hand rumbles at a distance, the right hand shivers in parallel thirds and punches out chords in contradiction of the prevailing 6/8 metre. It is a stormy episode that develops in dramatic intensity until the tension is relaxed in a modulation to C major and, eventually, a return to E flat and the gentle motion of the Gondollied. But whereas Mendelssohn would have ended the piece there, having neatly completed a ternary shape, Schubert goes on to introduce another episode, this one in A flat minor with its own middle section in B minor. If this new departure seems to be leading in the wrong direction, the easily accomplished return to E flat major and the opening material indicates that it did not stray so far off course after all.

The third piece is comparatively simple in construction. It could even pass as a sonata scherzo, with    syncopated Hungarian-style outer sections in C major and a quieter, rhythmically regular but boldly modulating middle section in D flat major. Certainly, the transition from the middle section back to the main theme is, in contrast to the obvious links of most impromptus, worthy of any sonata. So is its brilliant if boisterous coda.

1please delete as appropriate