Why people develop particular musical tastes is surely determined at least as much by culture and environment as by the music itself. But it's remarkable that we are so good at distinguishing one style from another. We can sometimes assign a genre after hearing less than a second of music, and well informed listeners may be able to identify specific composers from just a few bars of their music, even if they've never heard it before. Genres become dividing and subdivided as if with a surgical scalpel.

Nearly every composer or performer makes music within a tradition: the art is to build on what is already familiar, but to create from it something original. Too much originality, and the audience are confused and alienated. Too little, and we are bored. There's good evidence to support what we might intuit about musical audiences: that they are inherently conservative, preferring the familiar to the novel. Even the Beatles sold fewer records as their music became more experimental, the sequences of notes departing ever more from the conventional structures of pop. Sadly, our instinctive determination to fit new music into the patterns and formats we have learnt already tends to limit our sonic horizons; sometimes we must be ready to abandon our expectations and search for new listening strategies.

Perhaps digital music media will help to reveal what those strategies are and how they might be altered. Music companies are keen to discover new ways of classifying the vast libraries of music that people typically store on their MP3 players, which reach beyond traditional distinctions of genre and draw more on non-traditional groupings according to factors such as mood or timbre. We're not there yet, but these efforts suggest that there may be very different ways to draw the maps of the musical universe.