The Rare Part of Perfect Pitch

Absolute pitch is the ability to hear a note on its own and name it, or to sing a named note without hearing anything first. It is usually called perfect pitch, it is usually described as rare, and the figure most often quoted is about one person in ten thousand. That figure is old and rough, but nobody disputes the order of magnitude: in the general population, the ability is very unusual.

The usual explanation is that it is a kind of memory, an unusually precise recall of what each note sounds like. That explanation is pointing at the wrong part.

In 1994 Daniel Levitin asked forty-six people to sing a familiar pop song from memory, and compared what they sang with the recording. Two in five hit the exact pitch of the record at least once in two attempts, and nearly half came within two semitones both times. These were not people who could name notes. They were remembering the pitch of something they had heard many times, always on the same recording, always in the same key.

A later replication across six European laboratories found a smaller effect, about a quarter singing the exact pitch, and it varied from lab to lab. A quarter is still a very long way from one in ten thousand.

Glenn Schellenberg and Sandra Trehub came at it from the other side in 2003. They played college students with little or no musical training five-second excerpts of television theme tunes, some at the original pitch and some shifted up or down by one or two semitones. The students picked out the originals at better than chance. With unfamiliar recordings the same task was at chance, so they were not hearing some artefact of the shifting. They were hearing a difference from a memory.

So the memory is widespread. What almost nobody has is a name for it: a fixed, instant link between a pitch heard on its own and a label such as F sharp. That is the rare part, and it behaves like something learned rather than something given.

The clearest sign is age. Almost everybody with absolute pitch began musical training in early childhood, usually before about six. People who start later very rarely acquire it, however talented and however hard they practise, which is the pattern of a critical period: a window in which the brain makes an association that it later finds very hard to make.

The second sign is language. In 2006 Diana Deutsch and colleagues compared first-year students at the Central Conservatory of Music in Beijing, all Mandarin speakers, with first-years at the Eastman School of Music in Rochester, New York, who spoke no tone language. Absolute pitch was far more common in Beijing, and the difference held when students were compared by the age at which they had begun lessons. In Mandarin the pitch contour of a syllable changes which word it is, so those children had been attaching meanings to pitches since infancy. Deutsch's reading is that absolute pitch sits closer to speech than to memory: a habit of labelling, practised early.

It also runs in families, which is why early training cannot be the whole story. The careful position is that early experience appears necessary and is not sufficient.

Once you see it as a label rather than a memory, some odd facts stop being odd.

The labels are tied to a convention. Concert pitch was fixed at A = 440 hertz at a conference in London in 1939, and later adopted as an international standard. Before that, the pitch of A drifted between cities, decades and orchestras. A possessor trained on a modern piano carries a set of names anchored to that agreement, and when a baroque ensemble tunes to A = 415, about a semitone lower, every note arrives under the wrong name. Some possessors find that close to unbearable. The rest of the audience notices nothing, because the rest of the audience was never attaching names.

And it is not the same thing as musicianship. Relative pitch, hearing the distance between notes, is what lets a musician transpose, harmonise and play in tune with others, and it can be trained at any age. Plenty of fine musicians do not have absolute pitch and do not miss it. Some who do have it report that it gets in the way when they have to transpose, because the names keep insisting.

None of this makes absolute pitch less remarkable. It makes it more specific. It is not a better ear. It is an early association between a sound and a name, of the kind nearly all of us made with words and almost none of us made with notes.