The Notes Between the Notes

The standard introduction to Arabic music says it divides the octave into twenty-four quarter tones instead of twelve semitones. This is tidy, it is in a great many textbooks, and it is not what musicians play.

The best evidence for that is an experiment nobody designed as one.

In 1932 Cairo hosted a congress of Arab music, attended by Egyptian and Levantine musicians, European scholars, and Béla Bartók among them. Part of the purpose was to settle theoretical questions, and one of the questions was the tuning. A qanun, the plucked zither with movable bridges that makes retuning practical, was set up to an equal twenty-four-tone scale, exactly as the theory prescribed.

The musicians played it and said it was out of tune. Then they retuned it by ear.

That is as close to a controlled test as this subject gets, and it has been confirmed by measurement since. Maqam intonation is not equal-tempered in twenty-four any more than it is in twelve. Most maqam scales contain a perfect fourth or fifth and all octaves are perfect, and the remaining degrees sit where the tradition puts them, which varies by region and by period.

So what is the twenty-four-tone grid for?

Writing things down. If you want to notate maqam using Western staff notation plus a couple of extra accidentals, you need a finite set of symbols, and rounding every pitch to the nearest quarter tone gives you one. It is a practical compromise that discards detail, in the same way that writing a blues phrase in straight eighth notes is practical and discards detail. The notation is a transcription convention and the music is what players do.

Now the part that the quarter-tone story obscures, which is the more interesting part.

The organising unit of maqam is not a scale at all. It is the jins: a short melodic cell, usually three to five notes, with a characteristic interval pattern and a characteristic behaviour. A maqam is built by joining ajnas, typically one from the tonic upwards and another continuing from where the first ends, and a performance moves between them. Asking which notes are in a maqam is a little like asking which words are in a sentence: true, answerable, and not the structure.

Three ajnas worth knowing, because they show what the twelve-note keyboard can and cannot do.

Jins Bayati spans a perfect fourth and has two neutral seconds above its tonic. A neutral second sits between a minor and a major second, roughly three quarter tones, and two of them stacked land nowhere near any piano key. There is no approximation of Bayati on a keyboard, not even a poor one. This is the fact to reach for when somebody says these are the same twelve notes played expressively.

Jins Hijaz has a large step between its second and third degrees, and it can be faked on a piano. Players have a name for the result, piano hijaz, and it is not a compliment: the big interval comes out at the wrong size, the ones around it adjust, and the cell loses the thing that identifies it.

Jins Saba spans slightly less than a perfect fourth, which means a maqam built on it has no stable fourth degree in the Western sense at all.

One last thing, because it is a useful corrective to the idea that there is a single Arabic tuning to be discovered. Some strict traditionalists reject the twenty-four division entirely and work with a seventeen-tone set, on the grounds that twenty-four was a commercial simplification. The disagreement is live. Which is what you would expect of a living practice and is not what you would expect if a grid had been handed down from theory.