Voice of a Poet: The Last Sonatas.

Barbara Nissman (piano). Three Oranges Recordings 3OR-29 TT: 77:56.

BUY NOW FROM Three Oranges Recordings

Schubert playing. The pianist Arthur Schnabel used to say that he played Beethoven for money and Schubert for pleasure. Still, until relatively recently, Schubert's piano music did not often appear in concert. One can speculate on reasons: a lot of it consists of miniatures; almost none of it demands the killer virtuosity that wows an audience; structurally, the longer works become difficult to suss out. For example, the pianist-lecturer Alfred Brendel demonstrated the thematic unities that bind Schubert's last three sonatas to each other. It was news to many, including me. With Beethoven's sonatas, you have the impression of a carefully planned movement, where one section follows another almost inevitably. In Schubert's, you feel as if you watch an iris unfold. His movements are more sectional, less tightly-knit, although he works from a relatively small set of gestures and themes. The Beethoven pianist can often remind you of an explainer of intricate machinery and holds your interest as the connections among parts become clear. You can hear the work of the performer. Schubert resembles Mozart in that the better you play him, the more "natural" you sound, and why should anyone credit you just for playing?

We forget how hard it is to be natural. It's a source of many misunderstandings about Schubert as a composer. The standard rap on him was as a brainless genius who came up with one great melody after another. It is, by the way, the same line about Gershwin and Poulenc, and people also used to say it of Verdi. But, like those others, Schubert worked at his craft. Much of Schubert's mature music is strongly influenced by Beethoven, although his music also differs significantly. Schubert takes a "neat move" of Beethoven and makes it his own. Beethoven tends to highlight his bold strokes. Schubert frequently softens the edges, working to produce song-like structures in his instrumental works. Beethoven seems a tireless inventor of new and exciting keyboard textures (look just at the first movement of the "Waldstein"). Schubert, although capable of this, relies on more usual finger patterns. Perhaps this stems from his lack of Beethoven's virtuoso pianism.

It may surprise many to learn how often Schubert composer turned his hand to the piano sonata. Grove's lists twelve complete and nine fragmentary or unfinished (most of them early attempts). It surprises because they appear on programs so seldom and then, really only the last three. He wrote them in the last few months of his life. Perhaps Brendel's demonstration shouldn't amaze us after all, although it does me. One could argue that these themes and gestures were buzzing around him during this brief period. Still, Schubert often wrote startingly dissimilar works in the same month, but he conceived these sonatas as a set and knowingly used this thematic family. They appear in a wide range of variation and context. The sonatas found a publisher only in 1839, most likely due to Robert Schumann's advocacy of Schubert in general. The publisher even dedicated one of them to Schumann, whom, of course, Schubert never knew. However, they disappointed Schumann, who complained about a lack of novelty and a too-simple style. Brahms, however, was a fan. Schumann's greatness as a musical critic I don't dispute, but Schubert foxed him here. He packed considerable art into these scores. Still, they move with remarkable fluency. Schumann may have missed Beethoven's sense of titanic struggle and mistaken that lack for naiveté.

The Sonata in A's first movement is a case in point. It's a long piece held together mainly by three ideas: a declamatory, syncopated gesture, a curtain-raiser heavy on octaves; delicate arpeggiated triplet runs; and a lyrical idea later on. In addition, one hears a subtle rising chromatic line occurring in the first paragraph which has structural consequences throughout. One can describe the musical journey as sharp oppositions between the first and second idea, with the lyrical theme as a breather. Schumann used the phrase "heavenly lengths" in writing about Schubert's "Great" C major Symphony, and one can apply that concept here. Indeed, the movement runs as long as some entire Beethoven sonatas. The curtain-raiser certainly sets expectations for an expansive movement, and the composer achieves a sense of spaciousness largely through unexpected, unusual modulations to harmonic Nova Zembla (a Schubert specialty).

The second movement is in three-part form, with a quasi-improvisatory middle section sandwiched between a gently syncopated, lyrical theme, tinged with regret. Most of the interest lies in the middle. The harmonies and rhythms become more complex, the textures change more frequently, and the mood darkens considerably. A remarkable recitative section leads to a bit reminiscent of the first idea (but significantly different as well, including the fact that it's in major mode) and makes for an unusual, yet smooth return to the opening, mainly through texture, rather than through conventional harmonic preparation. The movement ends with some dark, broken chords which prefigure the next movement, a scherzo.

"Scherzo" means "joke." This movement playfully comments on ideas in the earlier movements. The large form is again a ternary or "sandwich" form, this time where the sprightly encloses a slower, simpler middle. The opening features the broken chords heard at the end of the second movement, now quick and teasing. Abrupt harmonic changes - to my mind, foreshadowing Prokofiev and Richard Strauss - dominate the movement. The middle hearkens back to the opening measures of the first movement - a syncopated "echo" note, this time dreamy rather than dramatic.

The finale comes in the form of a large sonata-rondo. A rondo takes a primary idea and alternates it with contrasting material: eg, A B A C A B A. A sonata-rondo, as Schubert uses it here, alternates a primary idea and a contrast and then adds a development section of those two ideas: roughly, A B A (development of A and B) A B A. It looks more complicated here than it actually sounds, because Schubert has set the listener on a safe path in brilliant fashion. He begins with a memorably lyrical main idea and a more complicated sharp contrast. Both make it easy for listeners to keep their bearings in a densely packed movement. However, they're not merely utilitarian. The primary idea happens to be one of the most beautiful Schubert ever wrote, and that's saying something. The second idea may not be hummable, but it has a distinct shape and a lot of interest. Schubert's prodigious transformations of both provide a lot of the pleasure of the movement. In addition, triplet arpeggios comprise a large part of the motor that moves the rondo along. Thus, the composer hearkens to the gossamer triplets at the sonata's opening measures, although in the finale the triplets assume various characters as well as light. Schubert ends with a remarkable coda. The primary theme appears in fragments - almost sighs - followed by a strong allusion to the curtain-raiser to the first movement. This not only is yet another cable tying the entire sonata together, but it contributes to the sonata's spaciousness. You feel as if you've traveled a very long way.

I first heard the Sonata in B-flat on my car radio. Usually (as was the case here), I come in at the middle and miss the credits. When that happens and I don't already know the piece, I play Name That Composer. I'm better than average at it, but this piece entirely eluded me - the music so odd that I found it hard even to guess the period. At times, it sounded like Beethoven but it lacked the concentration of Beethoven. Besides, I know the Beethoven sonatas fairly well, and it didn't recall any of them. I guessed Hummel, Voříŝek, even Czerny, as well as Schubert. The last movement struck me as the strangest of the four, with a section that seemed to anticipate Liszt's gypsy mode. At the end, the announcer's repeat of the credits (the pianist: Rudolf Serkin) put me out of my misery.

This final sonata resembles Beethoven the most in the shape of some of its important themes, but the working out belongs solely to Schubert. Some of them start you thinking, "Which Beethoven sonata did he crib this from," but almost immediately it dashes from Beethovenopolis to Schubertland. After many years, I have yet to feel comfortable with its architecture, since it strikes me as the most complex of the set (and the other two late sonatas are no walks in the park).

The first movement works with three major ideas: a noble opening; a falling, melancholy theme; an idea dominated by triplets, sometimes march-like. In addition, a trill in the bass provides a unifying element throughout the entire sonata. The development works with the three main ideas in unusual ways. For example, the first idea appears half-way suspended between B-flat (its home key) and d-minor -- a very unusual harmonic move. The bass trill serves to move the music into surprising tonal areas and signals beginning of the recapitulation. The movement ends with a coda consisting of fragmentary sighs of the first idea and a final growl of the trill.

The slow movement is yet another sandwich; a near-tragic lullaby enclosing a more animated section. It begins with a rocking rhythm with bass "echoes" that recall the opening of the A-major Sonata. The mood resembles Schubert's "Ständchen" ("Leise flehen, meine Lieder"). It moves to a chorale idea over a contradictory rhythmically animated accompaniment. The first section returns, this time mostly in major modes, thus resolving the sighs of its first appearance.

The third movement takes the shape of a scherzo surrounding a trio. Schubert marks the scherzo part "Allegro vivace con delicatezza" - fast, lively, with delicacy. It whirls about, through many key changes, threatening to fly apart, but it never does. In a certain way, the marking contradicts the music, especially the "delicacy" part. The music takes a dramatic turn in the trio, with the bass "echo" in the opening movement of the previous sonata now more like a jab in the ribs.

The finale, another sonata-rondo (but far more integrated than the previous sonata's), this time works from three major ideas: a two-part theme, the first in sputtering minor (very Beethovenian), followed by a major-mode section which reminds me of a polka; an ascending melody in long notes, beneath which the accompaniment percolates; a giddy dotted-rhythm idea, which skips like a spring lamb (my favorite theme of the movement; I smile when I hear it). In addition, an abrupt, heavily accented note in octaves (later also in chords), followed by a slight pause, becomes a unifying gesture throughout the movement. The ascending melody, incidentally, comes from the recap in the slow movement. All three main ideas receive extensive transformation, often outside the formal development section itself. Schubert makes variations on his basic material as easily McDonald's cranks out fries. This makes the movement a little harder to follow than the finale of the previous sonata. I've even made several diagrams of the structure, but I still can't tell for certain where the development ends and the recap begins. Yet it really doesn't matter. The music's vigor carries you through.

Normally, I have no more than two versions of any work in my CD collection. I simply don't have the shelf space. However, I'm hipped on this sonata and broke my rule. Lupu, Clara Haskil, Pollini, Kissin, Rubinstein, Richard Goode, Perahia, and of course Serkin (first recording) have all found room. Some of these performances aren't worth keeping, but my hoarder instincts won't let me get rid of any of them. Rubinstein, Kissin, and, sadly, Serkin need a clue. Pollini and Perahia give you readings that say more about gorgeous pianism than about the sonata. They are "flyover" interpretations. Goode is OK. Haskil struggles, but the struggle yields great interest. Lupu, hands-down my favorite of the bunch, plays as beautifully as anybody and manages to illuminate the architecture. His slow movement might just break your heart. He may be a shade too suave, but that's a quibble.

Amid this company, how does Nissman do? As it turns out, damn well. I put her with Lupu, although her approach differs considerably. Both have long, singing lines, but the nature of their singing differs. Lupu practices a suave pianistic cantabile, not really different from many other players, just better done. Nissman approaches her lines vocally and resembles a great Lieder singer or a Sarah Vaughan. If you listen carefully to just the first few phrases of D. 959, you hear all these micro-crescendi and -decrescendi, slightly behind or ahead of the pulse, shading the line and tempo. Often it sounds as if she gets the piano to breathe. But it isn't some tiptoeing-on-eggshells interpretation. The playing has both muscle and subtlety. For example, Lupu's D. 960 scherzo emphasizes the "delicatezza" marking - as I said, almost too suave. Nissman in general is edgier than Lupu, and the music benefits from it. It reminds you more of the roughness of Beethoven scherzi. You can tell Lupu has rigorously planned his reading. Nissman has done the same, but somehow she manages to sound spontaneous - in other words, "just playing." Yet she manages to elucidate the complex structures of both sonatas, shaping them with assurance. This is definitely one Schubert disc to have.

S.G.S.(June 2022)