Why the clefs look like that

Look at a treble clef for a moment without reading it. It is an absurd shape: a loop, a spiral, a tail that hooks under the staff and comes back up. Nobody would design that. Nobody did.

It is the letter G, written quickly, by hand, for six hundred years.

The problem the clef solves

A staff on its own says nothing. Five lines and four spaces give nine positions, and pitch rises as you go up — but from where? The lines are a ruler with no zero marked. Without a zero the same written note is any note at all.

Early Western notation had exactly that problem. Before the eleventh century, melodies were written as neumes drifting above the text: shapes that told a singer who already knew the tune where it rose and fell, and told a singer who did not know it almost nothing.

The fix is credited to Guido of Arezzo, in the early eleventh century, and it is disarmingly simple. Draw the lines, then label one of them. Guido drew a line for F and a line for C and marked them — the story has it in red and yellow ink — and the moment one line has a name, every other line and space has one too, because the intervals between them are fixed.

That label is the clef. The word is French for key, from the Latin clavis: the sign that unlocks the staff.

Three letters

The labels were letters, and scribes wrote them the way scribes write anything — fast, and then faster. Three of them survived into print, and each one still points at the line it names.

The G clef. The spiral curls around one line, and that line is the G above middle C:

G4

Everything else on the staff follows from it. Put a note on the line just below the curl and you have F4; two lines up from the curl and you have D5. The shape has lost every trace of a letter G, but the crossing point has not moved in centuries.

The F clef. Two dots, and the line between them is the F below middle C:

F3

The dots are what is left of the two horizontal strokes of a written F. The curve above them is the stem.

The C clef. Two facing curves, and the point where they meet is middle C:

C4

Three letters, three pitches, one each side of middle C and middle C itself. That is the whole system.

Why more than one?

Here is the part that a modern reader, used to seeing only two clefs, tends to find backwards. The multiplicity is not decoration and it is not tradition for its own sake. Five lines are not enough.

Nine staff positions cover a tenth. A trained voice covers closer to two octaves. So any single fixed clef leaves some music hanging off the top or the bottom of the staff, on ledger lines — and ledger lines are slow to read, because you have to count them.

A movable clef solves that by sliding the zero. The C clef in particular was written on whichever line kept a given voice comfortably inside the staff, which is why it has five historical names for five positions: soprano on the first line, mezzo-soprano on the second, alto on the third, tenor on the fourth, baritone on the fifth.

Two of those are still in daily use, and for exactly the original reason:

• Alto clef, C on the middle line — the viola's clef. A viola part in treble clef would sit mostly below the staff; in bass clef, mostly above it. Alto clef centres it.
• Tenor clef, C on the fourth line — used for the upper register of the cello, the bassoon and the trombone, where bass clef would mean a forest of ledger lines.

Seen that way, the clef is not a historical accident that survived. It is a piece of engineering: a way of moving the page to fit the instrument instead of moving the instrument to fit the page.

What did not survive

Plenty. Clefs have been written on more lines than we now use, staves have had four lines and six and ten, and the same piece has been copied with different clefs by different scribes for different singers. What settled the question was not theory but printing: movable type rewards a small fixed set of signs, and a small fixed set is what we got.

The residue of that is the one clef that names no letter at all — the neutral clef, two thick vertical strokes, used for untuned percussion. It labels nothing because there is nothing to label: a snare drum has no pitch, so the staff is only a rhythm grid, and the sign is there to say do not read this as pitch.

The thing worth taking away

A clef is not a difficulty added to notation. It is the answer to the only hard question notation had: relative to what?

Every line on a staff is a lie until something declares a zero, and for a thousand years the way we have declared it is by writing a letter, badly, in the left margin.