Sound: not as simple as it sounds: an interview with Joshua Fineberg: a child of psychoacoustics and the computer revolution, the "spectral music" movement is turning Western art music on it ear

Joshua Fineberg is the John L. Loeb Associate Professor of the Humanities at Harvard University and a composer whose works are widely performed in the United States and Europe. He collaborates with computer scientists and music psychologists to help develop tools for computer-assisted composition, electronic sound manipulation, and in music perception research. In 2004 Fineberg became the U.S. editor of The Contemporary Music Review. In 2006 his book Classical Music, Why Bother? Hearing the World of Contemporary Culture through a Composer's Ears was published by Routledge. Fineberg is associated with the movement known as "spectral music," which draws on acoustics and computer technology to explore the fundamental nature of sound (a spectrum is a representation of a sound in terms of the amount of vibration at each of the individual frequencies that make it up). In spectral composition, timbre often eclipses melody as the primary musical element.

What is spectral music?

All such labels are kind of awkward, but the common thread is that rather than taking for granted certain sonic categories as the most musically relevant way to divide up the soundstream--notes, for example, that are played for particular durations at a particular volume--you start from the assumption that what you have is the soundstream itself. Though sound can be parsed in the traditional way, it can be parsed lots of other ways. By understanding the physical and psychophysical principles of sound, you can gain an understanding into the possibilities and methods best adapted to modifying sound over time.

In what ways has spectral composition been influenced by science and technology?

This is a kind of music that couldn't have happened without the progress in acoustics and psychoacoustics in the late 1960s and 1970s when people started having access to the first analog sonograms and then electronic software sonograms that let you see the interior composition of sound. The personal computer enabled people to analyze sounds more easily with less demanding equipment.

Spectral music has been called a post-electronic approach to music. Sometimes it uses electronic synthesis; often it doesn't. But the knowledge acquired in order to make (synthesize) sounds from scratch is essential to writing this kind of music. To really control sound as we want to, we must understand enough to be able to make it.

Is the movement French?

Initially it was centered around an ensemble called L'Itinerarie in early 1970s Paris, a very experimentally oriented group that was trying out these ideas. But once you had the basic concepts it became very clear that you needed computer tools. Because in France music in general is not at the universities, except for musicology or music history, the place where a lot of this happened was IRCAM (Institut de Recherche et Coordination Acoustique/ Musique). It's a research institute started in the mid-1970s and built on the idea of providing tools to musicians at the interface between science and art. In the early 1980s, the computer-assisted composition techniques that IRCAM was developing prompted them to bring spectral composers into the fold.

Spectral music became one of the first real implementations of computer-assisted composition that went beyond what I would call algorithmically produced music. What spectral composers wanted was much more like what happens in computer-assisted design in architecture.

In music there has been a long tradition of laborious hand-calculation, but the reality is that if you spent four weeks calculating something, you're going to use it whether it turned out to be what you wanted or not! Whereas if you spent twenty minutes on it with a computer, you might have the courage to go back and try it seven or eight times until you really find the thing you were looking for.

Some see spectral music as a reaction to the artificiality of serialism and 12-tone-row music. Is it somehow more natural?

We are creatures that are tremendously sensitive to timbre because the vowels of language depend on timbral perception, as does our auditory scene analysis. The fact that we are relatively less good at identifying things like pitches and intervals is part of why for a long time they were interesting. But when you start thinking you can do anything that is mathematically possible with musical symbols, you get a kind of speculative music that at a certain time loses all contact with perceptual reality. Spectral music certainly strove to reground musical discourse in human perception and cognition.

Your book asks, Classical Music, Why Bother? What's your answer?

In order for subsidized art to survive it must be seen as having importance and intrinsic value independent of how much entertainment value it has. You're not just going out and buying it. You're supporting it because you believe that the world is a richer place with this art in it. That is a much harder sell than it used to be. Now, most people believe that in most domains, there isn't better or worse. You also see this in the debate over evolution, in the idea that we should teach all the alternatives.