Published online 10 December 2007 | Nature | doi:10.1038/news.2007.359

News

Music is in our genes

African cultures that sing alike tend to be genetically similar.

Feel the rhythm: music is tightly tied to biology.Punchstock

A study of 39 African cultures has shown that their genetics are closely linked to the songs they sing. Music, it seems, could reveal deeper biological connections between people than characteristics, such as language, that change rapidly when one culture meets another, says Floyd Reed, a population geneticist at the University of Maryland in College Park, who led the study.

"Other aspects of these populations’ cultures have undergone tremendous change, but the music seems to persist," he says. "In a way music is very resilient to cultural change."

The work, presented late November at the American Anthropological Association’s annual meeting in Washington DC, compares modern genetic data to a catalogue of traditional songs gathered in the 1950s and 1960s by ethnographer Alan Lomax. Lomax, best known for his recordings of American folk music and his popularization of singers such as Woodie Guthrie and Lead Belly, collected some 5,500 songs from 857 cultures.

To reveal connections between musical styles, Lomax, who died in 2002, and ethnomusicologist Victor Grauer created a system called cantometrics. This classifies vocal songs based on a sliding scale for 37 traits, such as yodelling and tempo.

Victor Grauer, now retired and living in Pittsburgh, Pennsylvania, wondered whether their cantometric database could be fused with modern genetics. Grauer contacted Sarah Tishkoff, a geneticist at the University of Maryland who studies human diversity. She offered the project to her post-doc, Floyd.

“At first I just dismissed it, I thought 'this is kind of nuts',” Floyd says. But after a night’s sleep he decided it had potential.

Sing along

In the 1970s, Luigi Luca Cavalli-Sforza, a geneticist at Stanford University in California, argued that combining genetics with fields such as linguistics and demography could help trace human ancestry. But Cavalli-Sforza hadn't looked at how genes and songs travelled together across cultures.

To make this comparison, Floyd crunched through Lomax and Grauer’s catalogue to convert the cantometric database into a two-dimensional scatter plot, with each dot representing a culture. The closer two dots are, the more similar their music.

For example, the music of African bushmen cultures, such as the Juhoansi (listen to Juhoansi audio clip) , fell close to that of pygmy tribes such as the Aka (listen to Aka audio clip). Both include yodelling and a vocal trick called an interlock, in which people sing in tight succession. Further away are styles by groups such as the Hutu (listen to Hutu audio clip), who live in central Africa and sing in unison.

When Floyd compared the graph with a database of genotypes from more than 3,000 people in Africa, he found a correlation between genes and songs. In other words, cultures that had grouped together musically tended to share genetic markers.

The link was stronger than the correlation between songs and geography: cultures next door to each other weren't as likely to sing the same tunes as were cultures with similar genotypes.

ADVERTISEMENT

Measure that yodel

“This is very cool stuff,” says Erez Lieberman, a mathematical biologist at Harvard University in Cambridge who has analysed the evolution of language using darwinian principles. He cautions that the method for scoring songs is quite subjective, making it difficult to use quantitatively. “How do you figure out the metrics of what's yodelling and what's not yodelling?” he says.

Floyd agrees that it is hard to quantify culture. But he says he thinks that Lomax's cantometrics must tap into something fundamental, since the statistical significance of the link to genetics is so high.

Taking an evolutionary perspective on song and language has the power to uncover how cultural traits travel the world, Lieberman adds. 

Commenting is now closed.