Bach on the Piano
Playing the Chromatic Fantasia and Fugue
The odd beginning of Bach’s Chromatic Fantasia and Fugue in D minor, BWV 903: scales up and down, brusque and brisk. Many pianists emphasize the brusqueness and briskness. Alternatively one may bring out the harmonic implications, the chords or arpeggio’s hidden in the scales. Every scale is an arpeggio, with passing tones.
J.S. Bach, Chromatische Fantasie und Fuge BWV 903, Lodewijk Muns, October 2022
Scales and arpeggio’s are typically the keyboardist’s practice material. As such, they are ready to hand for improvisation; insubstantial, but effective. Almost as if the hands were thinking for themselves.
These thinking hands keep moving closely together, in passages that could essentially be played by one hand, until they come to linger around the middle of the keyboard. Still restless, the fingers are playing chords up and down, as if waiting for clear directions from the mind: this is the way to go.
The manuscript sources of BWV 903 here use an abbreviated notation. The chords are written as simple block chords, the word arpeggio, ‘harping’, tells the player to go on breaking them up and down. This is a feature of baroque harpsichord style that can be easily and effectively adapted to the piano. Profiting from the pedal (the right, damper lifting pedal), one may increase sonority, double the bass in octaves, apply dynamics (forte, piano), and emphasize a certain ‘voice’, usually the top, sometimes the middle or the bass.
The Chromatic has made its successful transition from the harpsichord to the piano early on. When Felix Mendelssohn played the piece in February 1840, the audience “in a frenzy” demanded a repeat (“ganz rasend und stürmisch”, he wrote to his mother on March 7). Later that year, in a letter to his sister Fanny, he enthusiastically describes his success with playing the arpeggio’s in a pianistic manner. “… in this way the single harmonic sequences sound splendidly on the stout modern grands” (“auf den dicken neuern Flügeln”, letter of November 14, 1840). He takes the trouble of writing out a few arpeggio’s in full. Notable is his practice of initially arpeggiating each chord twice; an example that has found no following, to my knowledge.
His contemporary Franz Liszt too played the Chromatic in his concerts. According to one critic, the music theorist A.B. Marx, he “stormed through fantasia and fugue as if in a bacchanalian frenzy (the fugue twice as fast as one is used to – and capable of – hearing it) …” Marx was irritated by the arpeggio’s as they were usually played. For him, the seriousness and coherence of the work demanded that the chords be played as briefly arpeggiated block chords, “rock masses” in the wayward passagework.
Marx was contradicted in the same musical journal by the scholar and musician F.K. Griepenkerl, who had published his own edition of BWV 903 in 1819. Continuous arpeggiation is, according to Griepenkerl, the historically correct execution. A curious detail is that Griepenkerl makes the duration of each arpeggiated chord dependent on the varying number of notes: this destroys the metre, but creates a “higher kind of rhythm”. Most performers have chosen the more obvious solution of maintaining the beat by playing faster and slower groups of notes, depending on the number of voices in the chords.
…And a Speaking Voice
The hands find their way, with a few strange and surprising (‘chromatic’) detours, to a temporary resolution on the dominant chord. But the top note lingers thinly in the air. From this emerges a true voice, a voice that hesitatingly seems to try to say something, wordlessly (recitativo). Seeking affirmation, doubting again, impatiently protesting, finally sinking away in deep melancholy.
Particularly striking is how seamlessly these speech-song-like inflections are fused with the unsingable passage work of the keyboard idiom. In this, and in its imitation of operatic manners, the recitative is almost humorous. Maybe its melancholy and humour are not mutually exclusive. Although the harpsichord player can achieve much with timing, for the pianist there are regions to explore beyond what could have been originally conceived.
Practices in Print
Some clues for the way BWV 903 was performed in the pre-recording era are contained in the annotated editions, which were mainly produced for the amateur market. (Mendelssohn refused to do so: he was an early champion of the Urtext). For BWV 903 the editions are numerous. Liszt was generally reluctant to add interpretative detail to his editions or arrangements of Bach’s works, but his edition of the Chromatic (Schlesinger, Berlin, 184?) contains a lot of dynamic detail, particularly in the recitative. Interesting are the abrupt tempo changes, which suggest a very flexible and dramatic interpretation of each phrase. Surprisingly, the arpeggiations are not written out. Octave doubling is applied only at the most obvious places in the fugue.
It cannot generally be assumed that pianists’ editions faithfully reflect their own playing. In the case of Liszt, there seems to have been no typical ‘Lisztian’ interpretation anyway. A famous anecdote relates that he played Bach’s A minor Prelude and Fugue for organ, BWV 543, three times in a row privately. First in “the very perfection of the classical style,” then with “a slightly more picturesque movement, with the effects demanded by an improved instrument,” and finally, “as I would play it for the public … to astonish, as a charlatan!” (Rosen 1995: 510-11). This would be a good exercise for any pianist.
Hans von Bülow, pianist and conductor, a student of Liszt, published his edition around 1865 (Bote & Bock). It raises the practice of annotation to a new level and reflects the increased power of the period’s piano’s (and pianists’ muscles). Doublings in octaves and sixths, even in the recitative, add to the work’s pathos.
The extraordinary status of BWV 903 in the repertoire is evidenced by the fact that in the early 20th century Universal-Edition in Vienna produced no less than four different editions, by Julius Röntgen (1901), Albert Schweitzer (1912), Ferruccio Busoni (1910?, originally published by Simrock, cop. 1902), and by the music theorist Heinrich Schenker (1910), who supplemented it with thirty pages of commentary.
A dislike for what Schenker calls the “childish” arpeggiation up and down, with the top note always steady on the beat (the standard practice), causes him to propose a curiously zigzagging realization of the arpeggio’s. The recitative has a crude alternation of p (for each phrase) and f (chords and doubled phrases), which Schenker copied from the Bach-Gesellschaft-Ausgabe of 1890, which in turn took it from an early 19th-century manuscript. As far as I know, Schenker’s edition had no impact on performance history.
Busoni, on the other hand, has been influential through his teaching, recording, and Bach arrangements (which even include an arrangement of BWV 903 for cello and piano). His edition was marketed as an ‘interpretation’, and the detail of his annotations surpasses Bülow’s. However, Busoni is remarkably sparing with octave doublings. Most notably, he has left the fugue completely untouched; there are not even the most obvious octave doublings. Presumably, Busoni saw a strict division between a subjective-expressive fantasia and the objective, ‘pure’ counterpoint of the fugue.
Busoni writes out the arpeggio’s (dolce, quasi arpa) on two extra staves, each time preceded softly by a block chord (soft, muffled, quasi organo, somewhat stately) – as if he wanted to appease both adversaries, Marx and Griepenkerl, sixty years later. With Busoni we have entered the recording era, but he seems not to have recorded this particular work. His arpeggiation effect can be heard in later recordings, such as that of Edwin Fischer (1931), and György Sándor’s majestic interpretation (1950; adding a lower octave).
‘Majestic’ is a word that could characterize one of the three styles which can conveniently be identified in recorded performances of the 20th century. The other two I would could ‘dry’, and ‘expressive’ (or ‘sentimental’, if that didn’t have negative connotations). Maestoso, secco, espressivo.
The division does not quite coincide with the three styles demonstrated by Liszt, since the dry style is an early 20th-century, anti-romantic, neobaroque invention. Glenn Gould, the most famous exponent of the dry style, took a decidedly anti-romantic stance, but may still be considered an arch-romantic in his veneration for the ‘purity’ of counterpoint. The improvisatory unruliness of BWV 903 seems to have kindled his intense antipathy (“That’s Bach for people who don’t like Bach”). Alexis Weissenberg’s 1972 recording may exemplify the style.
The majestic style is based on a romantic, 19th-century image of Bach – one of Teutonic seriousness and masterful complexity. Whereas ‘dry’ performances make the piano sound as if it were a kind of upgraded, super-heavy harpsichord, the majestic style adopts the organ as its sound model. The actual sound of the 18th-century harpsichord may have been virtually unknown to 19th-century musicians. According to Bülow…
Doubling of passages and strengthening of chords seemed necessary to increase the colouring; calling to mind the sound of the organ seemed more in Bach’s spirit than constricting one’s imagination through associations with the spinet [i.e. harpsichord] or clavichord. (My transl.)
A.B. Marx’s essay on the Chromatic has contributed to this image of Bach-the-Majestic. However, the work poses a problem for this conception, and Marx saw it as his task to solve it. The fantasia is a rhapsodic, loosely structured work that seems to lack a strong connection with the fugue, which in turn is rather playful and in its contrapuntal devices not particularly sophisticated. However, Bach, is Marx’s assumption (and this was a dogma still prevalent in 20th-century musicology), or any composer of genius, is too serious to have written anything flimsy or not quite unified. If the work appears to lack consistency, it must be us who are not listening properly. Accordingly, he set out to describe how the work can be seen as “a unified outpouring … sustained by a single idea or mood”. This idea he finds in a narrative or dramatic structure: a progression from quest through crisis (recitative) towards increasing affirmation (the fugue).
Very much in the same spirit, Bülow has attempted…
… to clarify the psychological inner unity of the fantasy and the fugue, by presenting both pieces as the two-part monologue of one and the same person. The previous contrast between them gave the impression of a dreaming poet who is replaced by a grumbling schoolmaster. (My transl.)
One of his ways to achieve this is by starting the fugue pp and building a very long climax (bar 60, poco a poco animando il tempo sin’ al fine).
The third style, the expressive one, comes close to Liszt’s ‘picturesque’ style. It seeks out the expressive content of the work in all its details, and handles the instrument accordingly. In this perspective, octave doublings, rather than a violation of the text, might be said to be a matter of execution.
An expressive style is, according to Griepenkerl, the ‘true’ Bachian way of playing. Griepenkerl could claim to have inherited an unbroken tradition of Bach performance (he was a student of J.N. Forkel, who was a student of Bach’s eldest son, Wilhelm Friedemann). According to him, Bach & Sons played all fugal voices “with such abundance of expression … that each became a complete, true, and powerful picture of a state of mind”. That is surely exaggerated, and authenticity of tradition across generations is a dubious concept – musical taste changed radically around 1750. We might translate Griepenkerl’s claim into more modest terms: that the player must give each voice its proper dynamic shape. This could be a timeless adage; about the proper dynamic shape there will always be debate.
It is, anyhow, the style with which I am most in sympathy. Among the more interesting recordings in this style, I would mention that of Edwin Fischer, who also produced an fairly sparsely annotated edition in 1928 (Ullstein). In his edition, he warns the player to avoid “extreme pp and ff“. In his recording, after the dry and extremely fast passage work, he plunges into precisely those extremes, and with great effect. For a synthesis of the expressive and majestic, I would recommend Samuel Feinberg’s 1948 recording.
Recordings of a few favourite pieces by Liszt, Chopin and Debussy are found on my piano website, aandepiano.art
Dirst, M. 2012. Engaging Bach: The Keyboard Legacy from Marpurg to Mendelssohn (New York: Cambridge University Press).
Griepenkerl, F.K. 1848. “Noch einmal: J.S. Bach’s chromatische Phantasie”. Allgemeine Musikalische Zeitung Nr. 7, 16-2-1848, Sp. 97-100
Marx, A.B. 1848. “Seb. Bach’s chromatische Fantasie: Einige Bemerkungen”. Allgemeine Musikalische Zeitung Nr. 3, 19-1-1848, Sp. 33-41.
Mendelssohn Bartholdy, Felix. 2013. Sämtliche Briefe. Bd. 7 (Kassel: Bärenreiter).
Rosen, Charles. 1995. The Romantic Generation (Cambridge MA: Harvard University Press).