Composing a perfect pitch

Can musical taste be analysed scientifically? Steve Farrar reports on how record companies are banking on science to seduce the public.

A few haunting, liquid chimes of Gamelan music filled the laboratory. The four Balinese musicians listened intently to the first notes of the recording. Although they were accomplished performers of this traditional Indonesian music, performing to packed houses worldwide, the obscure songs they listened to in David Huron's laboratory had been selected because they were well outside their experience. Before it had really got going, the melody was abruptly cut short. Then, like a cross between a casino roulette table and the 1970s television game show Name that Tune , the musicians were asked to predict the next note.

Each placed gambling chips on a schematic array of differently pitched gongs that are the heart of the distinct Gamelan sound. Guess right and the stake came back tenfold. Guess wrong and the gambled chips were forfeited. By the end of the session, the worst player had notional winnings of $300,000 - best cleaned up with almost $4 million.

When Huron, professor of music and head of the cognitive and systematic musicology laboratory at Ohio State University, persuaded 25 seasoned American musicians to attempt the same test, they did not fare so well. While one came close to the worst Balinese score, several lost everything. Huron believes the results indicate just how reliant our appreciation of melody is on our cultural background. It is just one of a host of projects he has initiated to explore the phenomenon of music, from why some tunes stick in our heads to what makes a sound musical in the first place.

The challenge is daunting, but Huron is increasingly optimistic that many of the questions he is posing are, for the first time, coming within the grasp of human intellectual endeavour. The study of music is about to become as much a scientific discipline as linguistics or sociology and Huron is in the vanguard of this revolution. "Music has been a more traditional humanities discipline," he says. "It hasn't been in the hypothesis-testing business and the reason is lack of data. The essence of transforming something into a science is being drowned in data."

Traditional musicologists -and Huron is at pains not to criticise them - might spend decades pursuing a single line of inquiry. Huron wants to use new data-mining techniques to unlock answers to many of his questions from vast databases of music. And he wants to do this in a matter of days and weeks so that poor hypotheses can be readily discarded in favour of more promising alternatives. A host of different problems could be tackled in a systematic fashion using statistical analysis techniques. The team has already fed 6,000 European folk songs, gathered by collectors over the centuries, into their database. Different questions can be answered by using specially devised programs. For example, a contour map of ballad popularity across Europe can be easily drawn up, revealing the regions that like them slow and romantic.

Huron is soon to repeat the exercise with a study of Native American music. The approach also allows other aspects of music to be explored, for example testing accepted notions such as the "golden arch" of melody, the rise and fall of tone in initial musical phrases (such as Twinkle, Twinkle Little Star ). It turns out that the adage does indeed hold true when averaged over thousands of different tunes.

"New technology is opening up music, turning it from a humanity into a science," says Huron. "Music is an incredibly rich experience with many ways to interpret it and we're not in any danger of being able to explain it. But I think we are on the cusp of a new age of music scholarship."

He is not the only one to note the change. A few years ago, Huron had dinner with three executives from MoodLogic, an internet music company based in San Francisco. The businessmen were desperate to recruit music experts, especially those who also understood information technology and could apply rigorous scientific standards to their work. While sitting in the restaurant, the executives called some of those Huron suggested to make immediate job offers, such was their enthusiasm. The music industry has a global turnover of $210 billion a year. Yet research spending has always been desultory.

Identifying the raw material that feeds this vast undertaking -the musicians and songwriters who have the ability to achieve stardom -relies on the whim of an elite band of A&R people, gurus who have the "knack" of picking out the sensations of the future. The industry has done very well out of this arrangement but is now facing a revolution that threatens the whole "gut instinct" approach. Its executives are increasingly convinced that compact discs and cassette tapes are destined to go the way of the vinyl LP, and that some form of internet, possibly wireless, music distribution will come to dominate, even as they strive to throttle web-based music-exchange service Napster. They envisage software agents that can learn about an individual's preferences, such as listening to soft jazz late at night and hip-hop in the gym, and then seek out music to fit. This means selecting favourite tunes as well as identifying pieces that it is confident the person will enjoy, even dropping in unexpected oddities that it feels might engage that particular listener.

Questions such as what constitutes personal taste, can musical hooks be identified and how can you summarise a piece of music to create a musical "thumbnail", are taxing the minds of an industry that has previously been largely indifferent to such matters. To find the answers, major corporations such as Sony and Warner, as well as the thrusting dotcoms, are turning to experts like Huron.

And because he finds these questions fundamentally interesting, he is happy to work with them and, in the process, play a role in changing the way people experience music. "Until three years ago, I had interacted with business people about as frequently as any other music professor - which is to say, virtually never," he says. "Now the music industry has discovered music research. As academics, we can choose to stand on the sidelines and lob our usual smug invective, or we can roll up our sleeves and try to shape a musical culture worth having."

Deep in the Amazon, the interior of Papua New Guinea and the northern mountains of China, there are people who have not heard western music. But their number is few and dwindling. Huron recalls a report filed by an anthropologist working in the isolated, jungle-cloaked Indonesian province of Irian Jaya. The scholar reported encountering a family group of hunter-gatherers deep in the jungle and persuading them to let him compile an inventory of their belongings. There were steel blades and other tools. There was also a Hard Rock Café T-shirt, a working tape player, batteries and a handful of cassettes. If musicologists such as Huron want to test cultural influences on musical appreciation, they know they have to act quickly before the last "uncontaminated" few get exposed to the Spice Girls, Madonna or U2. A sobering prospect.

"It may well be that 100 years from now, musicologists will be cursing our generation for not grasping this opportunity before there is no culture left on earth that has not heard western music," he acknowledges. But such studies are expensive and ambitious. Getting academics out into remote communities far from the beaten track is a difficult task. Typically, Huron is thinking big. He proposes sending experimenters into the field to identify and gain the cooperation of subjects. Then a series of double-blind tests, free from the interfering bias of the experimenter, could be downloaded from Huron's university via a satellite phone to probe the musical appreciation of different cultures.

Can people with diverse cultural backgrounds really understand each other's music? Results could be beamed back to the lab for analysis and any follow-up problems and questions ironed out and resolved before the experimenter left the site. "We have the technology and resources to pull it off," Huron enthuses, "though not the funding - yet." The expense of such an exercise is far beyond the means of modern musicology. Perhaps Huron's suitors in the music industry will help chronicle the wisdom of cultures that they are playing such a major role in altering forever.