Piano Sonata No. 1 in D minor, Op. 28

I grew up in unquestioning ignorance of Rachmaninov’s First Piano Sonata. Like Tchaikovsky’s Piano Concerto (there are three!) and Dvořák’s Piano Quintet (there are two!), Rachmaninov often seems to have written only one piano sonata – that in B flat minor. He composed the Sonata No. 1 in Dresden, alongside his great Second Symphony and the unfinished opera Monna Vanna, and his letters from that time had already expressed doubts about the work. Thus, in a letter to his friend Nikita Morozov in 1907: ‘I’m finishing the piano Sonata now… The other day, I played the Sonata to [Oskar von] Riesemann, and he didn’t seem to like it. In general, I’m beginning to notice that whatever I’ve been writing lately – nobody seems to like. And I myself often wonder if it isn’t all rubbish.’

From these general worries, he moves on to specific ones, about the sonata’s scope: ‘The sonata is undoubtedly wild and endlessly long. About 45 minutes, I think. What drew me into such dimensions was the programme, or rather one guiding idea. It is three contrasting characters from a certain work of world literature. Of course, no programme will be given [to the audience], though it is beginning to occur to me that if I had revealed the programme, the Sonata would be clearer.’ And the prognosis: ‘No one will ever play this work because of its difficulty and length, and perhaps most importantly, because of its dubious musical merits.’

Such harsh self-assessment seemed to accompany Rachmaninov for much of his life – but was often disconnected from reality. For instance, Rachmaninov expressed very similar sentiments in letters about his Second Symphony – but the symphony was a triumph at its premiere, and has remained one of Rachmaninov’s most beloved works ever since. However, in the case of the First Piano Sonata, his prediction unfortunately landed closer to reality, and the sonata has remained – very unjustly, in my opinion! – among his less-loved and less-performed works since its composition.

The original version of the sonata was about ten minutes longer. Rachmaninov performed that version to a circle of contemporaries in Moscow, including Nikolay Medtner and Konstantin Igumnov, and on their advice shortened the first and last movement by some 110 bars. (Rachmaninov also acknowledged the symphonic scale of the work, saying that it would have been a symphony, if not for the purely pianistic language of the music.) It was Igumnov who gave the sonata’s premiere in Moscow in October 1908, to a lukewarm reception.

Rachmaninov later revealed to Igumnov that the ‘three characters from a certain work of world literature’ were Faust, Gretchen and Mephistopheles from Goethe’s drama, and that each movement was conceived to be a musical portrait of one of them, similar to Liszt’s ‘Faust’ Symphony. But at the same time, Rachmaninov noted that he had discarded this idea during the composition process, and none of it remained in the final work.

This, and similar pronouncements by composers (e.g. Ravel about La Valse being nothing but pure music) are hard for us to assess – do we take the composers at their word, or do we look at the music as if they had not spoken? Without attempting to answer the question conclusively for all cases, I would like to share my personal interpretation of the sonata – a work which I believe still contains multiple links to Rachmaninov’s original idea.

As an aside: from a purely compositional point of view, the sonata is a masterclass in achieving a wide range of expression through very sparse motivic means. As with Liszt’s Sonata in B minor, Rachmaninov develops the sonata’s entire melodic language from a handful of motifs that recur throughout, often transformed. Even if we treat these motifs as pure music, the web they create helps unify the sonata’s sprawling canvas.

The first movement, for me, is about Faust’s quest for knowledge, and the quest’s perceived futility. As such, much of the movement is spent in agitation, turbulence and discontent. The very opening sets the mood: an austere rising and falling fifth, like an ascetic pronouncement scrutinized for the truth, interrupted right away by an outburst (angry? defiant? tragic?). A second motif at 1 0:34, weary and sighing, could symbolise Faust’s resignation – but if so, it is short lived. The music (and Faust) keep winding themselves up, reaching a first anguished peak at 1:45.

Once this turbulence calms down, a new motif emerges [2:44] – a plainchant in parallel octaves, impersonal yet benevolent, surrounded by calm figurations. Not only is this in strong contrast to the preceding turmoil, but the link with church music is unmistakable. I take this motif to represent Faust’s yet-uncorrupted soul, or faith, or God. (Though I wonder if, on the most literal level, this motif’s initial appearance might have been inspired by a stage instruction in the first scene of Faust, where a chime of bells is called for, followed by a chorus of angels singing.)

The next moment to highlight is the second half of the development, from 6:37. Rachmaninov constructs a multi-step build-up, all through the use of the opening motif. It reaches new heights of turmoil [7:59], with the opening motif played in the left hand in fortissimo octaves and chords, but then, at 8:25, as we move into a major key, a sudden hope emerges. This sense of hope grows, becomes more and more certain, and finally reaches a peak at which the music, to me, seems to be nearly delirious with triumph [8:34]. I read this as Faust’s believing, momentarily, that his quest for knowledge has finally been successful, and the intensity of his victory is as extreme as the depth of his despair has been before.

But… this is not to last. After a very confident restatement of the faith/soul motif at 8:53, Rachmaninov subtly modulates from major to minor [9:10], and the effect is staggering: to me it is as if all certainty drains from Faust, much quicker than it has come:

FAUST: And here, poor fool! with all my lore I stand, no wiser than before; (Goethe: Faust, Part I, Scene 1, trans. B. Taylor)

The last moment to point out in the first movement is perhaps the most remarkable – towards the end of the reprise, yet another painfully turbulent scene builds up (from 11:30 onwards). Once again there is a sense of despair to the music, of an anguished rage, this time soothed by a new motif [12:25], which we will later recognise as the main motif of the second movement – that of Gretchen. We could read this as love being the only thing that can save Faust at this point; but at the same time, high above Gretchen’s motif appear shimmering, slightly malicious figurations, which could also hint at Faust’s temptation; Gretchen’s love for him is pure, whereas his for her is not necessarily so. Whatever the case, the movement ends with a mostly peaceful restatement of the soul/faith motif, and the very last notes are a falling fifth; the reverse of the opening, thus closing Faust’s quest for now.

The second movement is an oasis of lyricism and tranquillity in what is otherwise a rather dark work. The rocking fifths in the bass, no longer searching or discontent, seem to soothe like a lullaby, and above them, Gretchen’s motif is restated again and again [ 2 from 00:32]. The middle section [2:40 – 4:50], also developed from the same motif, does introduce some anxiety, but the emotional colour is very different – there is tenderness throughout, which was almost completely absent from the first movement. (I read this as stemming from the fundamental difference between the two characters – Faust’s self-centredness and Gretchen’s purity.) The high point of the movement arrives at 7:32, where the music bursts into a ravishing cascade of trills surrounding Gretchen’s motif. This is such a rare texture for Rachmaninov (I cannot easily think of another such abundance of trills in his piano music), and its passion stands out so much in this movement, that the dramatic impact is very strong – an ecstasy of love. The movement ends with a duet of the lovers [8:19], repeating Gretchen’s motif at each other.

The finale, Mephistopheles’ domain, starts with an avalanche of sound, a demonic ride (towards the Walpurgisnacht revels?), with shreds of motifs appearing and disappearing through the whirlwind of figurations. Soon, the Dies irae motif makes its entrance [ 3 1:32] – a medieval plainchant describing the Day of Judgement, which was an obsession of Rachmaninov’s throughout his creative life, and which fits here naturally through its connotations of hellfire and damnation. Rachmaninov introduces several sections in the exposition [2:32, 3:10, 3:49], which, despite the illusion of presenting new material, all derive from motifs introduced in the first movement, transformed and recombined. In this, Rachmaninov again follows in Liszt’s footsteps, who treated his Mephistopheles movement in the same way – I read this as referencing Mephistopheles’ inability to create, only to distort.

A shock comes at 4:35, as the music emerges out of a tumble down the keyboard into a restatement of Gretchen’s motif [4:42], followed by a full section based on an inversion of this motif, pale and melancholy [5:34]. Here, I feel, the music follows Goethe’s text quite literally, when in the middle of the revels, Faust sees an apparition of Gretchen –

FAUST: ….Mephisto, seest thou there, Alone and far, a girl most pale and fair? … I must confess, it seems to me As if my kindly Margaret were she.

Mephistopheles tries to snatch Faust’s attention away [6:07]:

MEPH: Let the thing be! All thence have evil drawn: It is a magic shape, a lifeless eidolon.

But the image reappears [6:20]:

FAUST: Forsooth, the eyes they are of one whom, dying, No hand with loving pressure closed; That is the breast whereon I once was lying,— The body sweet, beside which I reposed! (Goethe: Faust, Part I, Scene 21, trans. B. Taylor)

But Mephistopheles intervenes more forcefully [6:53], and the music builds up towards a fateful restatement of the Dies irae motif in the bass [7:30], which launches us into the reprise, the frantic figurations crowned each time with a short snippet of the Dies irae motif, as if doom has drawn nearer.

A similar scene is repeated after the reprise [11:24], which leads into the coda [13:05], and the final build-up of the work, when at the moment of greatest anguish and collapse, the soul/faith motif is hammered out in triple forte [14:28] before the sonata forcefully concludes in a minor key. Rachmaninov followed in Berlioz’s footsteps this time, ending the work with Faust’s damnation (as in the original version of the story), and not his salvation as per Goethe.

The above is a highly personal reading of the work. It is impossible for me to say what part of it really lies in the music and what comes from imagination. But it is equally impossible for me to ignore the totality of the threads linking the sonata to Faust’s story. Whatever the case, the sonata is a work of great dramatic power, supported by Rachmaninov’s incredible pianistic writing and motivic imagination. I am very happy to see it coming out of the shadow of its younger sibling and gradually putting to rest Rachmaninov’s gloomy predictions about its future.

Piano Sonata No. 2 in B flat minor, Op. 36 (1931 version)

Rachmaninov’s Second Piano Sonata appeared five years after the first, in 1913, the same year that saw him write the choral symphony The Bells. Bell-like sonorities can also be frequently encountered in the Second Sonata, from the silvery tinkling bells strewn through the lyrical sections to the great tumult of bells large and small at the climaxes of the first and second movements. However, I do not think we should make too much of this proximity, as the sound image of bells is so inherent to Rachmaninov’s musical language as to be found through all his works: as early as his First Piano Concerto (Op. 1) and as late as his last work, the Symphonic Dances – and indeed throughout both other works on this album!

Eighteen years later, in 1931, Rachmaninov revised the sonata extensively, cutting away more than a fifth of its size, and rewriting significant parts of the remaining material. He seems to have felt the textures of the original version were too dense, and the overall plot line not taut enough, saying to his friend Alfred Swann: ‘so many voices are moving simultaneously and [the sonata] is too long. Chopin’s Sonata lasts 19 minutes, and all has been said.’ (He was referring to Chopin’s Second Piano Sonata, a staple piece in his repertoire, which he had recorded in 1930).

The issue of cuts and shortenings in Rachmaninov’s works is a well-known one, as he himself made or authorised cuts in several works: not only in the First Sonata above, but also in the Second Symphony and the Third Concerto (his own recording of it uses the shortened version). Yet that issue becomes somewhat problematic (if not outright painful) once we consider that the reason behind those cuts might well have had to do with Rachmaninov’s lack of self-confidence and his doubt about audiences’ reception, rather than with any desire to improve the works musically. For instance, in 1931 he wrote to his friend and fellow composer Nikolay Medtner, describing a series of performances of another work of his, the Variations on a Theme of Corelli:

Not once have I played them in full. I was guided by the coughing of the audience. Whenever the coughing increased I would skip the next variation… In one concert… the coughing was such that I played only ten variations (out of twenty). My record was in New York, where I played eighteen. However, I hope that you will play all of them and that you will not cough.

This is hard reading… The Second Sonata, however, is a somewhat different case. Whereas the cuts mentioned above are just that – cuts, i.e. clean jumps from point ‘A’ to point ‘B’ in the score – the revised version of 1931 goes far beyond that. Besides reworking much of the piano writing (always with the aim of achieving leaner, more transparent textures), Rachmaninov inserted new sections almost every time he made a cut, and this new material alters the face of the sonata to such an extent as to make the revised version a separate, independent work.

Several analyses of the two versions agree that the sonata’s form and structure are better served by the first, 1913, version. As they point out, many of the sections Rachmaninov removed contained important thematic links and development, and at times whole chunks of the sonata form have been done away with, such as the recapitulation of the main subject in the finale. I fully agree with this on paper, but my feeling is that Rachmaninov was once again concerned with how the sonata would be perceived by the listeners – though in a good way this time.

Looking at the sections that were removed, I think we can find a common thread – they are all, for want of a better word, meandering. They often interrupt whatever line Rachmaninov was following at that time, and embark on a new micro-journey of their own. These might contain some really effective and highly pianistic music, yet on balance I feel they detract rather than add to the sonata’s unity. The revised version seems to me much more directional and single-minded, and the resulting structure more organic and easier to follow. (Putting directionality above expansiveness is of course completely a matter of taste. The opposite argument could be made, saying the original version is that much richer due to the way Rachmaninov explores adjacent ground rather than sticking to a single path.)

Through the newly inserted sections Rachmaninov also reinforces the melodic links between the sonata’s movements. The entire first movement is based on a short motif – four descending chromatic notes, followed by a dotted falling interval – which is first heard at the beginning of the work, right after the initial plunging passage (which already contains all of the motif’s notes but one). This motif occurs frequently in full throughout the movement, and its chromatic notes form the basis for much of the passage work, and for the lyrical second subject [ 4 1:58].

In both versions, this motif appears once more, in a reminiscing episode in the middle of the second movement, after the melancholic main theme of the movement has been explored in several variations, and a wonderful, heart-aching climax has been reached [ 5 3:23]. However, in the revised version this episode is fully reworked: it incorporates a longer quote from the first movement, bringing with it the agitated mood of the sonata’s opening, and gaining so much drive, drama and tension as to make the original episode (which is very beautiful on its own) almost pale in comparison.

Later on, in the coda of the second movement, the haze that follows the overpowering bell-frenzy dissolves, and there emerges in the revised version the lyrical second subject from the first movement [4:50]. This intertwines the two movements even more. And finally, to link in the third movement as well, Rachmaninov subtly alters the opening descending passage to include the notes of the first movement’s main motif.

As you will have probably guessed by now, I’m a full-throated advocate of the revised version. But in either version the sonata is a powerful, passionate, exhilarating work of blinding brilliance and great emotional intensity. It contains much darkness and turbulence, but also places of purity and light, and it is these, as is often the case with Rachmaninov, and in direct contrast to his First Sonata, that reign triumphant in the end.

The Isle of the Dead – Symphonic Poem after the Painting by Arnold Böcklin, Op. 29 (1909)

(arr. Georgy Kirkor and Boris Giltburg for piano, 1957/2023)

The inspiration behind one of Rachmaninov’s most sombre works, the symphonic poem The Isle of the Dead, lay in the eponymous painting by the Swiss Symbolist painter Arnold Böcklin. The painting is singularly evocative: a small rowboat bears a white-veiled figure to a rocky island in a mirror-still sea; it approaches the water gate, behind which stands a foreboding copse of cypresses – the shade so deep that nothing can be discerned behind the first line of trees. Sheer cliffs guard all other approaches. There is a deep sense of awe and stillness in the painting; Böcklin himself, while not explaining the painting’s meaning, wrote that the viewer should be able to dream themselves into the world of shadows, until they would fear disturbing the solemn silence with a loud word. The painting was an enormous success, leading Böcklin to paint six different versions of it. Reproductions abounded, owned among others by Lenin, Freud and Hitler.

Rachmaninov saw a black-and-white print of the painting in Paris in early 1907, and it was this print that inspired him two years later to compose a symphonic poem. (When he later saw one of the colour originals, he was disappointed: ‘If I had seen first the original, I, probably, would have not written my Isle of the Dead. I like it in black and white.’) Whatever Böcklin’s attitude to death was, Rachmaninov must have feared it; at least in this piece, the key emotions for me are not so much stillness, awe, or solemnity, but rather a sense of stark dread and hopelessness, which continues almost unrelieved throughout much of the piece. (I can’t help but think of Rachmaninov’s setting of Nunc dimittis from his Vespers, Op. 37 – again filled with heavy, inescapable grief, with none of the expectant hope of life after death that we can find in other composers’ settings of the same text.)

The music starts out of nothingness; an open fifth appears in the bass; and then Rachmaninov’s first masterstroke: the pulse which he had found for the piece, an irregular 5/8 meter, instantly evoking the pattern of a rowing stroke. The notes on the rising pattern are perfectly chosen as well, containing the minor third, to give us a sense of a minor key, and then two fourths in a row, radiating emptiness. Much of the music of the piece is simply a build-up of multiple voices playing this pattern, with slight harmonic changes – a journey without a foreseeable end, its perpetuity holding much of its terror.

When a melodic motif does briefly appear at 7 1:00, we are perhaps not surprised to recognise it as the first notes of the Dies irae motif – Rachmaninov’s constant companion, and probably nowhere as organically incorporated into music as in the Isle. Together, the two motifs construct our journey for the next minutes, gliding through some lighter terrain (e.g. 4:23 and 5:39), but also building up in waves towards increasingly powerful climaxes (7:34, 8:12).

At 9:57 new music appears, bringing a sense of life, and love, and possibility – all the things that no longer belong in death, yet there they are, against all hope. (Juxtaposed with the First Sonata, I think there is a parallel here with the recurrences of Gretchen’s motif in the finale. Both are lights in the dark; pale and distant in the Sonata, much more present and vibrant in the Isle). The yearning that suffuses the music is heart-breaking, as we know in our heart of hearts that this – hope, reunion, memory, whatever it is – cannot last. And indeed, the music turns troubled at 11:18, gradually building up towards the next climax [12:03], and then, after a tremendous polyphonic and polyrhythmic build-up of the Dies irae motif at 12:36, to the apex of the piece [13:41], when all collapses.

Out of this crash there first emerges a truly spooky section built from the Dies irae motif played in various voices over each other [13:58]; the ticking of the clock constituting a visceral evocation of the dwindling time. And then a final, acutely personal, recall of the music of hope [14:59], which eventually slides down into the depths, bringing back the opening music [16:24]. In this coda, the Dies irae motif floats high above the rest, a solitary mourning voice [16:53]. And then, at 17:27, Rachmaninov grants us the small miracle of a major-key harmony. It is akin to a final light in the dark, its gleam extinguished a few seconds later, leaving us with a modicum of, if not solace, then acceptance, as the boat continues its journey into nothingness.

A note on the transcription: as a basis, I have used a two-hand piano reduction by the Soviet musicologist Georgy Kirkor. Kirkor’s reduction, I believe, was not meant for concert performance; he reshuffled much of Rachmaninov’s complex three-layer writing so that the end result would be comfortably playable with two hands. I tried to reinstate Rachmaninov’s layering, admittedly making the piece less playable, but hopefully bringing it a step closer to the original. As for the audacity of approaching this inherently orchestral work on a piano, my reasons were twofold. On the one hand, it is one of my favourite pieces of music by any composer, and I dearly wished to engage with it as a performer. And on the other, I saw the piano version as a black-and-white version of the colourful orchestral original: a kind of a parallel in reverse with the black-and-white print that inspired Rachmaninov.