Beethoven, Piano Sonata No. 23 in F minor, Op. 57 (“Appassionata”)

Beethoven wrote the sonata in 1804–1806 and dedicated it to Count Franz von Brunsvik. It was published in February 1807 by the Bureau des Arts et d’Industrie, as his Op. 57 — the “Appassionata” name wasn’t his: the Hamburg publisher Cranz attached it to a four-hand arrangement issued after Beethoven’s death, and it stuck. Beethoven himself just called it the sonata in F minor.

A personal note

Frederic Lamond’s recording is the fastest one on the list below, and it’s not close. He’s about four full minutes ahead of everyone else, and this is a guy with a direct line to the piece if anyone has one: a pupil of Liszt, who studied with Czerny, who studied this very sonata with Beethoven himself. If anybody should have known how the finale goes, it was him.

So why does his third movement clock in at 4:40, when everyone else runs seven to eight and a half minutes? The marking is Allegro ma non troppo. Fast, but not too much. At 4:39 I want to ask him what part of “non troppo” he didn’t understand.

I don’t think Lamond forgot how the piece goes. I think it’s the record business. He recorded in the 78 era, when a single side only held about four or five minutes before the needle ran out. The movement gets squeezed to fit the disc, not to fit the tempo marking. Schnabel made the first complete recorded cycle of the Beethoven sonatas in the early 1930s, under that same limit, though his own Appassionata here doesn’t run away from the pack like Lamond’s does. Maybe he found a way around it, or just had more sides to work with.

I used to think Igor Levit’s recording was quick. Then I saw Lamond’s time.

One more thing worth noting: the Serkins, father and son, land on very different answers for the same movement. Rudolf takes the third movement at 8:10. His son Peter takes 9:24, over a minute slower, a generation later, in the same family. Hard to say if that’s just two different pianists or something closer to a son answering his father.

(Room for more as the list grows — and when I’m ready, this is where the link to my own recording will go.)

Structure

The first movement moves quickly through startling changes in tone and dynamics, built from an economical handful of ideas. The main theme, stated quietly in octaves, is a down-and-up arpeggio in dotted rhythm that cadences on the dominant — then is repeated immediately a semitone higher, in G-flat. That move to the Neapolitan key (the flatted supertonic) is one of the sonata’s real structural hinges: it isn’t a passing color, it’s the basis of the main theme of the finale too. The rhythm of that opening theme is sometimes traced to the Scottish folk song “On the Banks of Allen Water” — British folk tunes were well known in Vienna at the time, and Beethoven, like Haydn, arranged a number of them for British publishers, though the direct link to this theme is disputed rather than settled fact.

Czerny, who studied it with Beethoven himself

Carl Czerny is mostly remembered today for the études every pianist survives as a student, but he was also Beethoven’s own pupil, and the Appassionata was one of the specific sonatas he studied directly with the composer — alongside Op. 13, Op. 14 No. 1, Op. 14 No. 2, Op. 31 No. 2, Op. 101, and the Andante from Op. 28, plus all the piano concertos except Op. 19. Czerny later became one of the sonata’s first serious editors, and his own eleventh piano sonata — written in F minor, “Op. 57” in his own catalog of works — is generally read as a deliberate homage to his teacher’s, sharing not just the key and opus number but something of the first movement’s energy and power.

Czerny’s five editions of Op. 57 (Haslinger I, Haslinger II, Cocks, Wessel, and Simrock, spanning the 1820s to the 1850s) don’t agree with each other on tempo:

	Allegro assai (i)	Andante con moto (ii)	Allegro ma non troppo (iii)
Haslinger I	♩. = 120	♪ = 120	♩ = 138
Haslinger II	♩. = 108	♪ = 108	♩ = 138; Presto 𝅗𝅥 = 92
Cocks	♩. = 120	♪ = 120	♩ = 138
Wessel	♩. = 138; Più allegro = 160	♪ = 92	♩ = 152; Presto 𝅗𝅥 = 100
Simrock	♩. = 120	♪ = 112	♩ = 138; Presto 𝅗𝅥 = 96

Haslinger I and Cocks agree exactly, and across the earlier three editions the pulse of the second movement is consistently the same as the first — a relationship that breaks down entirely in the 1850s editions, where Wessel’s markings swing from very fast in the first movement to noticeably slower in the second, then very fast again in the third. Czerny also quietly standardizes one of Beethoven’s own markings: the autograph’s rinforzando in bar 226 of the third movement appears in all five editions as ritard. or ritardando instead — an error that goes back to the first printed edition and seems to have simply been carried forward rather than corrected.

On the finale, Czerny’s own written advice (in his supplement to the Piano Forte School, On the Proper Performance of all Beethoven’s Works for the Piano) is specific:

The present Finale must not be played too fast. The passages are to be performed with distinct equality and lightness, only slightly legato, and but seldom impetuously. The movement and power first continually increase on the repetition of the second part and towards the conclusion, and the Presto winds up the Sonata with all the power which can be elicited from the Pianoforte, by employing all its means.

By Czerny’s own admission, that volume doesn’t claim to simply record how Beethoven played — it’s his update of that style for the pianists of his own later era.

Recordings compared

Forty-one recordings, timed movement by movement and rated out of five — fastest total time to slowest. A few have comments worth keeping alongside the numbers.

Performer	i	ii	iii	Total	Rating	Notes
Edith Fischer	10:31	—	—	10:31*	★★★★	*first movement only
Larry Clapp	8:23	5:51	—	14:14		April 3, 1987 (no third movement)
Larry Clapp	8:38	5:51	—	14:29		1987 Robert Casadesus Int’l Piano Competition
Larry Clapp	9:06	5:55	—	15:01	★★★★★
Frederic Lamond	8:03	4:37	4:40	17:20	★	Considered the leading authority on Beethoven’s piano music before Schnabel; Breitkopf & Härtel published his edition of the sonatas. Plays too fast.
Wilhelm Backhaus	9:27	6:00	5:12	20:39	★★★
Larry Clapp	8:40	6:05	7:02	21:47		Tape 7 (ASU, I think, based on audience response)
Artur Schnabel	9:00	6:01	7:13	22:14	★
Maurizio Pollini	9:03	5:51	7:27	22:21	★★★
Yundi Li	9:20	6:21	7:29	23:10
Stewart Goodyear	9:51	6:04	7:19	23:14	★
Igor Levit	9:23	6:03	7:52	23:18	★★★
Emanuel Ax	9:30	6:05	7:56	23:31	★★★
Vladimir Horowitz	9:57	5:31	8:15	23:43
Lazar Berman	9:33	6:21	7:52	23:46	★★★★
Rudolph Buchbinder	9:00	6:44	8:03	23:47	★★
Arthur Rubinstein	9:29	6:32	7:57	23:58	★
Sviatoslav Richter	10:56	6:03	7:02	24:01	★★★
Russell Sherman	9:29	6:40	7:55	24:04
Boris Giltburg	9:53	6:24	7:57	24:14
Gary Graffman	9:24	6:48	8:03	24:15	★★★★
Murray Perahia	10:07	6:18	7:55	24:20	★★★★★
Dinis Schemann	10:02	5:38	8:43	24:23	★★★★★	Extremely accurate control, clear notes everywhere.
Stephen Kovacevich	10:14	5:57	8:13	24:24	★★★★★
Rudolph Serkin	9:23	7:06	8:10	24:39	★★★
Alfred Brendel	9:56	6:38	8:08	24:42
Van Cliburn	10:11	6:08	8:25	24:44
Wilhelm Kempff	9:53	6:01	8:52	24:46	★★★★	Recorded all the sonatas; worked with John O’Conor; played for Sibelius’s Hammerklavier.
Ingrid Fliter	9:56	6:33	8:20	24:49
Nikolai Lugansky	10:30	6:28	7:53	24:51	★★
Valentina Lisitsa	10:27	6:24	8:11	25:02	★★★★
Justus Frantz	10:02	6:54	8:09	25:05	★★★★★
Adolph Drescher	10:19	6:27	8:28	25:14	★★★★
Yuja Wang	10:14	6:50	8:11	25:15
Mikhail Pletnev	11:01	6:57	7:32	25:30
Emil Gilels	11:07	6:28	7:58	25:33	★★★★
Claudio Arrau	10:23	7:02	8:19	25:44	★★★★★
Evgeny Kissin	11:04	7:16	8:09	26:29	★★★★
Daniel Barenboim	10:40	7:39	8:13	26:32	★★★★★	Uses an octave on the low C in the coda.
Lang Lang	10:49	7:18	8:29	26:36	★★
Peter Serkin	10:03	7:20	9:24	26:47	★★
Glenn Gould	15:02	11:08	5:25	31:35	★

What Lenin heard in it

Maxim Gorky recorded Lenin’s remarks on the piece after hearing it played at a private gathering in 1920:

I know of nothing better than the Appassionata and could listen to it every day. What astonishing, superhuman music! It always makes me proud, perhaps with a childish naiveté, to think that people can work such miracles! … But I can’t listen to music very often, it affects my nerves. I want to say sweet, silly things, and pat the little heads of people who, living in a filthy hell, can create such beauty. These days, one can’t pat anyone on the head nowadays, they might bite your hand off. Hence, you have to beat people’s little heads, beat mercilessly, although ideally we are against doing any violence to people. Hm — what a devilishly difficult job!

It’s one of the odder footnotes in the piece’s afterlife: a head of state, mid-revolution, naming this sonata as music he had to ration for himself because of what it did to him.

Further reading

• Czerny’s own advice on the sonata, On the Proper Performance of all Beethoven’s Works for the Piano (supplement to the Piano Forte School)
• a modern study of Czerny’s five editions of Op. 57 against Beethoven’s autograph and against Czerny’s own written advice — the source for the tempo table and the rinforzando/ritardando note above

The score is in the public domain and available at IMSLP.