Scientists swing into action to explain jazz mystery
In 1939, Louis Armstrong asked audiences a question that has tormented jazz scholars ever since: ‘‘What is this thing called swing?’’
Scientists now believe that they have finally provided an answer, with a study suggesting that ‘‘swing’’, the Holy Grail of many jazz musicians, is the result of tiny musical nuances so subtle that they have eluded detection for a century.
The term ‘‘swing’’ was used by early jazz artists to describe what they felt was a specific playing style. ‘‘Generally, jazz is supposed to swing – it’s supposed to have this feel,’’ said Professor Theo Geisel of the Max Planck Institute, in Munich, a physicist who usually works on chaos theory but also plays saxophone.
There are exceptions: free jazz, characterised by disorderly improvisation, doesn’t have to swing. ‘‘But nearly everything else does,’’ Geisel said.
‘‘From the old days and Louis Armstrong, to bebop and hard bop – when jazz musicians come together, they expect to produce this thing called ‘swing’.’’
One contributory factor, known as the ‘‘swing ratio’’, had already been identified. It refers to pairs of notes and how, in jazz, the first will often be held slightly longer than the second, even though the sheet music may call for them to be the same length.
‘‘Everybody can hear this,’’ Geisel said. ‘‘It’s an important ingredient – but it’s . . . not enough for swing’’.
In search of the missing element, his team performed a series of experiments.
In one, they used a computer to manipulate the timing of various musicians’ parts in recordings by jazz ensembles. Two features were changed – notes played on downbeats, which would correspond to the downstrokes of a conductor’s baton, and notes on offbeats, which would occur between two baton strokes.
In one version of the music, only the piano soloist’s downbeats were delayed, by 30 milliseconds each. In another version, the researchers delayed both the soloist’s downbeats and the offbeats.
The scientists then asked 37 jazz musicians to rate how much ‘‘swing’’ they could hear in the manipulated music.
Delaying both downbeats and the offbeats did not yield swing.
However, swing did emerge when only the soloist’s downbeats were delayed, while their offbeats remained in sync with the rhythm section.
More precisely, when the downbeats were delayed but the offbeats stayed the same, the music was 7.48 times more likely to be rated by the jazz musicians as swinging.
‘‘Professional jazz musicians could perceive a pleasant ‘friction’ between soloist and rhythm section,’’ Geisel said. However, the musicians could not identify the cause of the effect.
Next, Geisel and his colleagues analysed more than 450 solos by well-known jazz musicians, including Armstrong, Miles Davis, Duke Ellington and John Coltrane. They found downbeats with a delay between 10 and 50 milliseconds in nearly all of them.
‘‘This subtle method of creating the swing feel is used only unconsciously by jazz musicians – they were not aware of it,’’ Geisel said.
He said the swing code would never entirely be cracked, however.
‘‘There are other ingredients. When you listen, let’s say, to [pianist] Oscar Peterson, you can hear that he is putting accents on different notes. This is part of swing. But extracting a rule on how he does it? It’s probably impossible.’’
The findings are reported in the journal Communications Physics.