Where is the science in music?
I have taught a course on the physics of music several times. Most physics departments offer a course like this, and typically the teaching is assigned to someone who has some musical background. But musical enthusiasts who take such a course leave vaguely disappointed – for the same reason that the music background of the teacher is irrelevant for teaching the course. That is because these courses are really on the physics of sound: how sound propagates, how it reflects, how it is generated and filtered. And in fact the physics of sound, of musical acoustics, even that of musical instruments themselves has been understood for many years. Here is how you do it: solve linear equations with the right initial and boundary conditions and you’re done. The equations are those for gas flow; they’re a little messy, but they are well known. The key word here is linear; for those with a math background, that translates to “easy-peasy”. I’ve written about this before.
Sensing this vague disconnect between the physics of sound and the science of music, I would give one lecture in my course on musical composition based on a talk I attended in the 1980s. At the time, physicists were curious about developments in the math and computer science community regarding nonlinear fractals. There were joint conferences where physicists and mathematicians would awkwardly interact, like girls and boys at a teenage sock-hop dance. At one of these conferences, I heard a computer science speaker explain that the spectrum of music frequencies has an unusual distribution where the power is inversely proportional to the frequency, called 1/f noise. This is fractal noise with self-similar correlations within correlations within correlations. Then he took some 1/f noise, and programmed it to play notes on a piano.
Up to that point, the physicists were yawning. Then we woke right up: the programmed piano was playing a credible jazz piano cocktail piece. The speaker then used completely random noise to program the piano, which would be like pulling numbers out of a hat (sometimes an avant-garde composition technique), and it sounded like random noise (or, an avant-garde musical composition).
In my physics of music class I would do the same, and – for one lecture anyway – everyone would wake up. These spooky correlations within correlations are at the heart of music.
So, what is the science behind music? As I’ve written before, the scientific origin of the correlations in music is not in physics, but in neuroscience. To understand music, is to understand the spooky correlations in how we think.
This year, our Mini-Science series will address “The Science of Music”, beginning February 26, 2014 – check it out if you like!