Whenever I talked about sound, I stressed the inadequacy of the classical languages that composers had used to describe it. I said that the evolution of the electronic instruments and recording processes had created a situation where the whole question of timbre - the physical quality of sound - had been opened up wide and had become a major focus of compositional attention. Modern composers and producers working in recording studios were experimenting with sound itself and were quite content to use largely traditional "received" forms (such as "the blues") upon which to hang their experiments. It struck me that this had been completely missed by classically trained musicologists, who were always looking for innovation in places where it wasn't happening. They were expecting that any music that deserved the title "new" would be making breakthroughs in harmony, in melody, in compositional structure; but here they were faced with a music that, in those respects, had barely made it past the turn of the century.

Brian Eno, from Scents and Sensibility
(thanks to rb for the link)