Lesson 1 · AMEB List B · Baroque
No. 6 of The Well-Tempered Clavier, Book II — a 61-bar prelude of unbroken motion, and a 27-bar fugue that is one of the tightest pieces of counterpoint Bach ever wrote.
Johann Sebastian Bach never left Germany, had twenty children, and worked his whole life as a church and court musician. After he died, people thought he was old-fashioned. Today he is often called the greatest composer who ever lived.
1685 born in Eisenach → Weimar (organist) → Köthen 1717–23 (court music for a prince — that's where Book I was written) → Leipzig 1723–50 (Thomaskantor: music director of the town churches and the famous choir school).
① He is the end of the Baroque — the era is dated to his death in 1750.
② He wrote both books of the Well-Tempered Clavier (1722 and c. 1740).
③ His works are numbered BWV = Bach-Werke-Verzeichnis (Bach Works Catalogue), compiled by Wolfgang Schmieder in 1950 — ordered by type, not by date.
Handel and Domenico Scarlatti (born the same year, 1685), plus Telemann, Vivaldi, Rameau, Couperin. Telemann was godfather to C. P. E. Bach.
Born four weeks and about 130 km apart — and they never once met. Bach tried twice. He rode to Halle and arrived to find Handel had already left. The second time Bach was too ill to go, so he sent his son Wilhelm Friedemann with an invitation; Handel declined. Bach did get Handel's music, though: he owned, copied and performed Handel's Brockes-Passion in Leipzig on Good Friday 1746.
Almost everyone assumes AMEB's four lists mean Baroque / Classical / Romantic / Modern. At AMusA they don't. Your Bach sits in List B — the counterpoint list — a list that runs from Bach's Forty-Eight and Handel's suites through Mendelssohn's and Reger's preludes and fugues to Shostakovich's Op. 87. Meanwhile Haydn's sonata is your List A work, because List A is the large multi-movement repertoire. AMEB is sorting by what kind of thing a piece is, not by when it was written — and the fact that a 1742 fugue and a 1951 fugue live in the same list is itself an argument about what a fugue is. Full list breakdown →
The BWV catalogue puts all the cantatas first (BWV 1–224) and the keyboard works much later (BWV 846–893 are the Well-Tempered Clavier). So BWV 85 is a church cantata, but BWV 875 is your prelude & fugue. If an examiner asks how the catalogue works, that contrast is the perfect answer.
It's Bach's giant project: 48 preludes and fugues — two books of 24 — proving that you can write beautiful music in every single key.
Each book has 24 preludes & fugues, rising chromatically: C major, C minor, C♯ major, C♯ minor, D major, D minor… all the way to B minor. D minor is therefore No. 6.
Book I — finished 1722 at Köthen. Book II — compiled c. 1739–42 in Leipzig, about twenty years later. Your piece is from Book II, Bach's mature "modern" collection.
In Bach's Germany, Clavier meant any keyboard instrument with strings — harpsichord or clavichord — not the piano, which was brand new and almost unknown. So he wrote for a harpsichord-like sound.
Well-tempered means the instrument is tuned so all 24 keys work — but each key still sounds slightly different in colour. That is NOT the same as equal temperament, where every key sounds identical. (Equal temperament only became standard after Bach died.) Nobody knows exactly which tuning Bach used — but the whole point of the collection is showing off all 24 different key-colours. Say this and you sound like a real musician.
Put the music on while you read the next two sections. This recording follows a scrolling score, so you can watch the notes fly past.
The prelude is a moto perpetuo — "perpetual motion" — an unbroken stream of semiquavers that never stops, shared between the hands (watch for the hand-crossings).
Two halves, but no repeat signs (so it's a continuous binary, not a dance binary). Part I: D minor → A minor. Part II: A minor → back to D minor. Then a tiny coda (bars 57–61) over a held tonic note, ending in D major.
The exchange is real but it is three bars, not four: bars 2–4 in the right hand come back as bars 6–8 in the left, note for note an octave down, while bars 2–3 left hand go up two octaves into bars 6–7 right hand. (Bars 1 and 5 do not correspond — a nice detail to have checked yourself.) Later the whole opening returns on the dominant: bars 26–29 = bars 5–8 up a perfect fifth, then bars 30–33 = bars 1–4 up a perfect fifth — the two four-bar units coming back in reverse order.
The oscillating broken chords (bar 2: B♭–G–D–G) are the keyboard's version of a violinist rocking across two strings, and that is why the hands have to cross. Bach spent his Weimar years transcribing Italian concertos — seven of the sixteen in BWV 972–987 are Vivaldi — so the idiom was in his hands. Say "violinistic figuration", not "bariolage": true bariolage depends on the colour of an open string, and a keyboard hasn't got one.
There is an earlier, much shorter version of this prelude, catalogued BWV 875a — and "shorter" means 43 bars against your 61. Its oldest witness is a copy made by Bach's pupil J. C. Vogler in 1729, ten years before Book II was assembled. Scholars trace three stages: 43 bars → 53 bars → 61 bars, the final growth coming from Bach's revision of about 1742–44. So this prelude grew by nearly half while you were not looking.
Your pocket giveaway: the demisemiquaver flourishes in bars 22 and 24. Count them — there are exactly eight demisemiquavers in the entire prelude and all eight are in those two bars. Nowhere else, no exceptions. Point at one and say "this run is a later addition — Bach revised this prelude twice", and you have shown you really know the score. (There is an early version of the fugue too — a D minor Fughetta in Agricola's copy — so don't say "only the prelude was revised".)
A fugue is a musical conversation: one voice (or "part") starts a tune called the subject, then other voices copy it, one by one, while the music chatters on. Bach's fugue is only 27 bars — but it is one of the cleverest in the whole book.
The subject is exactly 2 bars and has two halves. First half: twelve triplet semiquavers — D–E–F–G–F–E–F–G–A–B♭–A–G. Note that it zig-zags rather than simply climbing; don't say "a rising scale". Second half: a leap up to high D, then a slow chromatic descent in quavers — D–C♯–C♮–B♮–B♭–A — which then keeps going G–F–E and leaps back up. That falling chromatic fourth is the passus duriusculus, the Baroque "lament" figure: the same shape Bach used in the cantata Weinen, Klagen (BWV 12) and then reused for the Crucifixus of the B minor Mass. Two opposite affects packed into one subject is a very late-Bach idea.
The comparison that impresses: that shape — a turning figure followed by a chromatic descent — is the same architecture as the royal theme of The Musical Offering, written five years later. Same fingerprint, one of Bach's late obsessions.
⚠️ Do not call this "the chromatic fugue." No standard reference gives this piece a nickname — no work in Book II has one. The nickname belongs to a different D minor piece with a chromatic three-voice fugue: the Chromatic Fantasia and Fugue, BWV 903. Knowing that the two get confused, and saying so, is worth more than the nickname ever was.
Alto (bar 1, alone) → Soprano (bar 3, the answer) → bar 5 → Bass (bar 6). So: middle, top, bottom. Most notes call bar 5 a "codetta", but it is not neutral filler: it carries two statements of the subject's head upside-down (alto on G, soprano on A). The inversion arrives before the exposition is even finished. Say that and you have said something most candidates miss. (Exposition ends bar 8 on the usual reading; some analysts run it to bar 10.)
The subject starts D–E–F–G; the answer starts A–B♮–C–D. That B natural is the proof: it's an exact transposition up a fifth, with no adjustment. (A "tonal" answer would have tweaked a note to stay in the key.)
A second line in plain (duple) semiquavers against the subject's triplets. It appears exactly four times, each an exact transposition: alto b3 (against the answer), soprano b6 (against the bass entry), then it vanishes after the exposition and returns only twice more — bar 14 and bar 26, both times in the bass. "It disappears after the exposition except at bars 14 and 26, both in the bass" is a sentence worth having ready.
The fugue's engine is stretto (a voice starting before the previous one has finished) plus inversion (the subject upside-down), and it runs them together. The climax is bar 25, in the tonic, where all three voices state the subject's head inside a single bar, one crotchet apart:
beat 1 — alto, right way up (on D) · beat 2 — bass, upside-down (on A) · beat 3 — soprano, right way up (on D).
So the two entries that are half a bar apart are the two direct ones, with the inverted form wedged between them a crotchet later — rectus, inversus, rectus. Direct and inverted really are sounding at the same time; just don't say they are the pair half a bar apart.
How many strettos? Say "it depends how you count" and you will be right: published analyses give three or four, and a note-by-note scan of every statement of the head finds six points of overlap (bars 5–6, 10, 14, 17–18, 21, 25). The number is a matter of counting criterion, not of fact — which is itself the better answer.
The key plan is deliberately tiny — D minor (i), F major (III), A minor (v) and G minor (iv), and nothing else. The excitement comes from the counterpoint, not from wandering keys, which is exactly why 27 bars is enough. (Don't forget the F major: in a one-flat key signature it needs no accidentals, so it is easy to miss when you're scanning the page for sharps and flats — but the episodes go there twice.)
Bach wrote no dynamics. Shape the music by how many voices are sounding and how tense the harmony is — think steps (terraces), not long romantic waves.
The harpsichord has no sustaining pedal, so Bach's counterpoint is written to be heard without it. On a modern piano, use only tiny "touch" pedals to connect what the hand can't reach — never enough to blur the voices.
The Baroque rule: a trill begins on the note above, on the beat. (Bach's own chart — the Explication of 1720 — and his son C. P. E. Bach's book of 1753 both say so.) Very few ornaments are written in this pair — and the prelude's mordents are largely a Version A reading, so which ones you play is partly an edition decision. Know which.
The fugue's subject moves in triplet semiquavers and its countersubject in plain semiquavers, so the two grids collide. Pencil-mark bars 6, 9, 11 and 21: those are the places where a note in one voice lands off the triplet grid of another. Practise them slowly and decide — mathematically exact, or slightly assimilated? — because that is a question you may well be asked.
Differentiate the subject's two halves with touch, not dynamics: play the rising sextuplets smooth and even, then let the chromatic quavers be slightly heavier and more "spoken". Then use exactly the same articulation every single time the subject appears — that consistency is what keeps the tune audible in every register.
With this piece, "which edition?" is not small talk. BWV 875 survives in two different authentic traditions, and your copy has silently picked one.
British Library Add. MS 35021 — 22 loose bifolia, 21 of the 24 pairs (Nos. 873, 874 and 881 are missing), and no title page at all.
The B autograph is lost. Its best witness is the copy made by J. C. Altnickol — Bach's pupil and later his son-in-law — which carries corrections in Bach's own hand.
The Bärenreiter/NBA prints BWV 875 twice, keeping the two traditions strictly apart. Henle does the opposite: one text representing "Bach's latest thoughts". Neither is wrong — but you should know which you have.
The prelude's mordents are largely a Version A feature, and so is the figuration in the bars around the demisemiquaver flourishes. So "why do you play that ornament?" has a source answer, not just a taste answer.
The London manuscript is called an autograph, but it isn't all Bach's hand. Several pairs were copied by his second wife, Anna Magdalena Bach — and BWV 875 is one of them (along with 871 and 878). She was a professional singer before she married him, and she copied a great deal of his music; her hand is so like his that scholars have argued for decades about which is which.
And the manuscript's journey is a story in itself: J. S. Bach → his son Wilhelm Friedemann → the composer and piano-builder Muzio Clementi → an English family → the British Museum in 1896 → the British Library. The man who wrote the sonatinas every young pianist plays once owned the paper your prelude is written on.
The Well-Tempered Clavier was never published in Bach's lifetime. It circulated hand to hand, in copies, for fifty years — and it made every composer who touched it.
When Simrock finally put the Forty-Eight into print in 1801, they published the second book first — Book I followed in 1802. You can even see it in the plate numbers: Book II is plate 138, Book I is plate 166. So the collection the world met first is the one your piece comes from. If an examiner asks when the Well-Tempered Clavier was published, "1801 — and curiously, Book II came out before Book I" is a far better answer than a bare date.
You will hear that the Forty-Eight and Beethoven's sonatas are the pianist's Old and New Testaments, usually credited to the conductor Hans von Bülow. It is a lovely line and everyone repeats it — but no one can point to where he actually wrote it. Say "traditionally attributed to von Bülow" and you get the credit for the quotation and for the caution.
Book I's 1722 title page has a decorative squiggle of loops drawn above the words. In 2005 a scholar, Bradley Lehman, argued the loops are a code for the exact tuning Bach wanted, and the idea went round the world. Then two rebuttals appeared in the very same journal, one calling the chain of reasoning "full of weak links", and several other people decoded the same squiggle into different tunings — which is usually a sign that a code has no single solution.
And here is the part that matters for you: it is an argument about Book I's title page — and Book II has no autograph title page at all. So even if the code were real, it wouldn't attach to your piece. Knowing why the famous theory doesn't apply is a more grown-up answer than repeating it.
In the exam, a strong answer has four moves: Name it → Place it → Contextualise it → Land it on a decision. Here's the shape, using the fugue subject.
"The subject is exactly two bars and it has two characters. The first half is twelve triplet semiquavers, zig-zagging upward — light and dancing — and the second half leaps to high D and descends chromatically through a fourth, which is the passus duriusculus, the Baroque lament figure; Bach uses the same shape for the Crucifixus. Packing two opposite affects into one subject is a late-Bach trait — the royal theme of The Musical Offering is built the same way. So I play the triplets smooth and even and let the chromatic quavers be more weighted and speaking — and I keep that identical for every entry, including the inverted ones, so the subject stays audible wherever it appears."
How is The Well-Tempered Clavier organised?
Two books of 24 preludes & fugues — one pair in every major and minor key, rising chromatically (C major, C minor, C♯ major…). So D minor is No. 6.
Is the fugue's answer real or tonal — and which note proves it?
Real. The answer starts A–B♮–C–D; the B natural shows it's an exact transposition up a fifth, not adjusted to stay in the key.
Where is the fugue's climax, and exactly what happens there?
Bar 25, in the tonic: all three voices state the subject's head within one bar, a crotchet apart — alto right way up on D, bass inverted on A, soprano right way up on D. Direct and inverted forms sound together. (Careful: the two entries half a bar apart are the two direct ones.)
How does the fugue end?
On a bare D — all three voices converge to an open octave and unison with no third at all, so it is neither major nor minor. That is not an editorial variant: both surviving traditions print it — and no nineteenth-century performing edition added a third either; Czerny, Busoni, Tovey and Bischoff all print the bare D. Compare the prelude, which does end on a bright D major Picardy third.
Where does the countersubject go after the exposition?
It disappears. It sounds four times in all — alto bar 3, soprano bar 6, then only bar 14 and bar 26, both times in the bass.