Key and diatonic chords: why some songs go together

After playing for a while you start to notice that songs repeat themselves. Not the melody — the chords. The same handful keeps coming back, and when you try to slot in a chord from outside the group it sounds wrong before you can explain why. That has a name: the diatonic chords of a key.

Where a song's chords come from

A song in C uses the notes of the C major scale: C D E F G A B. Build a chord on each of those notes, using only notes from the scale itself, and you get seven chords. Those seven are the family the song is made of.

degree    I     ii    iii   IV    V     vi    vii°
in C      C     Dm    Em    F     G     Am    Bdim

Notice the pattern of major and minor: major, minor, minor, major, major, minor, diminished. That pattern never changes, in any key. If you know the scale, you know the chords.

in G      G     Am    Bm    C     D     Em    F#dim
in D      D     Em    F#m   G     A     Bm    C#dim
in A      A     Bm    C#m   D     E     F#m   G#dim

Degrees are written in roman numerals — upper case for major, lower case for minor. That is why players talk about a "four five one": those are functions, not specific chords, and they hold in any key.

The three that do almost everything

Of the seven, three carry the weight: I, IV and V. In C those are C, F and G. An uncomfortably large amount of popular music lives inside those three, plus the vi (here, Am).

What each one does, in ear language rather than theory language:

• I is home. It is where the song resolves, and where it can end without sounding cut off.
• V is tension. It practically demands a return to the I — the strongest cadence in Western music.
• IV is leaving without urgency. It steps away from home but does not pull back with the same force.
• vi is the relative minor: the same notes as the I, a melancholy colour.

Try it: take a song you play and write the degrees instead of the chords. If I–V–vi–IV shows up in some order, you have just found the progression behind a few hundred hits.

The relative minor

C major and Am minor use exactly the same seven notes. The difference is not the material, it is which note acts as the centre of gravity. That is why every major key has a minor twin, three semitones below:

C ↔ Am      G ↔ Em      D ↔ Bm      A ↔ F#m
F ↔ Dm      Bb ↔ Gm     E ↔ C#m

This is immediately useful on stage. Going from a song in C to one in Am sounds natural with no preparation at all — the ear does not register a key change, only a change of mood.

Neighbouring keys

Two keys sound close when they share almost every note. From C to G a single note changes (F becomes F#). From C to F, also one (B becomes Bb). That one-note kinship is what the circle of fifths describes:

… F  →  C  →  G  →  D  →  A  →  E  →  B …

Immediate neighbours move into each other effortlessly. Keys far apart on the circle — C straight to F#, say — sound like a cut, and sometimes that is exactly the effect you want. But it should be a decision, not an accident.

Where this becomes a setlist decision

Chaining songs in neighbouring keys, or in relative major and minor, makes a show flow without anyone being able to say why. Jumping from E to Eb between two songs creates a sense of stumbling even when both are played correctly.

There is a practical side too, and it may matter more: neighbouring keys tend to ask for similar hand shapes and, with a capo, let you keep the same family of shapes across several songs in a row — fewer changes, fewer mistakes mid-show.

When the chord is not in the family

Sooner or later a song will have a chord that is not on the list. That is neither a mistake nor a rare exception — it is a device. An E7 in a song in C, for example, exists to pull hard towards Am, borrowing the tension of another key's V.

You do not need the full theory to use it. It is enough to recognise that a chord from outside the family is almost always setting up the next one. Mark where it resolves and the logic shows itself.

To put this into show order, see how to build a setlist.