Keys and Key Signatures

A key is the major or minor scale a piece treats as home, named after its tonic — the note of that scale the music keeps returning to and finally settles on.

Being in a key is not a property of a list of notes. It is what a piece does with them. The two phrases above contain exactly the same five pitches; the first one is in C because it leans on C, leaves it, and comes back to it, and the second one leaves that job unfinished. Play enough phrases that behave like the first and a listener stops hearing seven equal notes and starts hearing one note with six others arranged around it.

That arrangement is a hierarchy, not a set. The tonic is home; the fifth degree above it is the note that sets up the return; the seventh degree sits a half step below home and leans on it. A key is a scale plus that hierarchy, which is why two pieces using identical notes can be in two different keys.

A key also comes with a modality. The same seven notes will make a major key if the third degree above the tonic is the wide one and a minor key if it is the narrow one, and the whole colour of the music changes with it. Naming a key therefore takes two words: a note and a modality — C major, A minor, E flat major.

A key signature is the group of sharps or flats printed immediately after the clef on every line of music, which alters those letters in every octave for the rest of the piece unless a later sign cancels them.

It is worked out, not looked up. Build the major scale from the note you want as home and see which letters the pattern forced you to alter: from D, the formula puts a half step between the third degree and the fourth and another between the seventh and the octave, which can only be done by writing F sharp and C sharp. Those two sharps are D major’s key signature. The same procedure gives one sharp for G, one flat for F, and so on for all fifteen of them.

Two things follow from printing it once rather than repeatedly. The reader learns the whole alteration before the first note, so the page can be read as plain note heads at speed. And an accidental printed inside a bar recovers its meaning: it is no longer routine bookkeeping but a signal that this note comes from outside the key, which is exactly the kind of thing a reader should notice.

A key signature identifies the notes, not the home. Two sharps means D major or B minor, and the signature cannot tell you which — for that you look at where the music starts, where the bass falls at the end, and which of the two notes the phrases keep returning to.

The circle of fifths is the twelve keys arranged in order of rising fifths, so that every key stands between the two whose notes it shares all but one of.

Nothing about it has to be memorised, because it is a consequence rather than a diagram. Take any major scale, go up a perfect fifth, apply the formula from there, and six of the seven notes come with you: the fifth degree of the old scale becomes the tonic of the new one, and only the note a half step below that new tonic has to be raised. One step, one new sharp, every time. Going down a fifth instead does the mirror image and adds one flat at a time. Do it twelve times and you have visited all twelve notes and returned to the start, which is what makes it a circle and not a list.

The two sides meet opposite C, at the point where six sharps and six flats describe the same sounds. That key is written F sharp major or G flat major depending on which direction you arrived from, and the choice is a matter of which is easier to read rather than anything the ear can settle. The same seam is why there are fifteen key signatures for twelve keys.

What the arrangement is good for is distance. Keys that sit next to each other share six of their seven notes, so a piece can slip between them almost without a seam; keys on opposite sides share only two, and moving between them is an event. That single measurement is behind most of what the circle gets used for: counting a signature, choosing a key to modulate to, and explaining why a progression that keeps falling by fifths sounds like it is going somewhere.

A relative key is the key that shares another key’s seven notes and its key signature while treating a different one of them as home: every major key has a relative minor whose tonic is three half steps below its own, and every minor key a relative major three half steps above.

A parallel key is the key that shares another key’s tonic and changes its notes instead. C major and C minor are parallel: both call C home, and the minor lowers the third, sixth and seventh degrees a half step, so a piece that moves between them keeps its centre of gravity and changes its colour.

The two relationships are worth keeping apart because they do opposite things. A move to the relative minor changes where home is without altering a single note or the signature, so it can happen inside a phrase and often goes unnoticed as a key change at all — it is why the same eight bars can sound major at the start of a verse and minor by the end of it. A move to the parallel minor keeps home exactly where it was and repaints everything around it, which is impossible to miss.

Finding either takes no memorisation. For the relative minor, count three half steps down from the major tonic — a minor third — and start the same notes there; for the relative major, count three up. For the parallel, keep the note and change the word, then lower the third, sixth and seventh: the signature moves three places anticlockwise around the circle, which is what turning C major into C minor and its empty signature into three flats amounts to.

Choosing a key is deciding which pitch a piece’s home note will sit at, and the decision is made on practical grounds: the range of the voices and instruments that will play it, and which notes those instruments produce most easily.

Range comes first. A melody has a highest note and a lowest note, and the key is what sets where those land. A song that strains a singer at the top is moved down until it does not, and the key it ends up in is whatever that move produced — which is why the same song appears in different keys on different recordings by different singers.

After range comes what an instrument gives away free. A guitar and a violin both have open strings, which sound louder, need no finger and are perfectly in tune, so music written for them gravitates towards the keys that contain those notes: E, A, D and G for guitar; G, D, A and E for violin. Brass and wind instruments pull the other way. A trumpet, a clarinet and a tenor saxophone are each built around a fundamental note that is not C, and their simplest fingerings and cleanest tone fall in the flat keys, which is why band and jazz repertoire lives in B flat, E flat and F.

None of which is the same as a key having a character. The idea that D major is triumphant or E flat major heroic is a real tradition with a real origin, but that origin is the tuning systems in use before the nineteenth century, not the keys themselves — and the difference it names has not existed on a modern piano for two hundred years.

Transposition is moving every note of a passage by the same interval, so that the distances inside it are untouched and only the pitch it sounds at changes.

The method is two steps. Measure the interval from the old tonic to the new one — C up to E flat is a minor third, C up to G a perfect fifth — and then move every note in the passage by that exact interval, accidentals included. Nothing is decided note by note: if one note moves a minor third, all of them do, or the tune is no longer the same tune.

Spelling is where transposing goes wrong on paper. Move the letter names by the same number of letters as the pitch moved, then take the accidentals from the new key’s signature: the third note of the tune above is the third letter of each key, E in C, G in E flat, B in G. Choosing spellings by sound alone — writing F sharp where the new key calls for G flat — gives a passage that plays correctly and reads wrongly, and a reader will stumble over it.

This is why a key is worth thinking of as a position rather than a set of notes. A singer asking for a song "a bit lower", a guitarist clamping a capo on the second fret, an arranger writing out a trumpet part, a producer dragging a MIDI clip up three semitones: all four are doing the same operation, and all four rely on the fact that the music itself survives it intact.