Edmund Campion’s Computer Aided Composition

Cover design for the Quadrivium booklet

At the end of last month, Neuma Records released the album Quadrivium, the title of a four-movement suite by Edmund Campion. The title referred to the four subjects of the medieval liberal arts. In “order of appearance” in the suite, the movements are entitled “Mathematica,” “Geometria,” “Astronomia,” and “Musica.” While I found a Web page for this album on the Amazon.com Web site, the content was somewhat vague; so I feel it appropriate to cite the Web page on the Bandcamp Web site as the most reliable source.

Quadrivium was composed in 1995 when Campion was a Fellow at the American Academy in Rome. He used his fellowship to explore computer-assisted composition, now given the abbreviation “CAC.” This was my first contact with both the label and its abbreviation. Campion elaborated on it as follows:

In these pieces are found my early adoption of CAC techniques including the use of time-varying probability tables and Markov decision processes to create distributed spatial electronics and to manage slowly evolving instrumental harmonies and textures.

My initial impression was that the use of probability as a technique for composition had come a long way since John Cage had applied it to select passages for consultation from the I Ching. More interesting, however, was my second impression, which was that the EUTERPE software tool, which I had first invented in 1967 for a senior thesis under the advice of Professor Marvin MInsky at the Massachusetts Institute of Technology (MIT), had been explicitly designed for “computer-assisted composition.” More specifically, the thesis title was “EUTERPE: A Computer Language for the Expression of Musical Ideas.”

What made this significant was that, at that time, there were very few laboratories in which programming and running software was an interactive process. For example, even an institution as significant as the Bell Telephone Laboratories required “batch processing” for most (if not all) computer usage. The interactive approach at MIT on the East Coast was, at that time, complemented primarily (if not only) by the activities at Stanford University on the West Coast under the supervision of Professor John McCarthy, previously Minsky’s colleague at MIT. Stanford would subsequently apply that technology to music, primarily under the guidance of John Chowning; and the University of California at Berkeley would provide partnership through its Center for New Music and Audio Technologies, initially led by David Wessel.

In retrospect I would now say that EUTERPE was a “polyphonic” programming language. Exploring how polyphony could be evoked and utilized then found its way into my doctoral dissertation, which was given the title “A Parallel Processing Model of Musical Structures.” That model included the sort of probabilistic techniques that Campion would subsequently explore. However, it was based on an infrastructure of “voices,” each of which had its own “program,” which described both its own activity (the notes to be performed) but also the ability to “cue” other voices.

All this is now distant past. (The doctoral thesis was approved in September of 1971.) Now that computers have become “personal,” there are any number of ways in which they have provided “voices” for polyphonic approaches to both composition and performance. (The first Macintosh computer that I purchased was connected to my Yamaha Clavinova.) As had been the case with EUTERPE, probabilistic decision-making is only one of many techniques that can be classified as CAC. Ironically, having now listened to Quadrivium several times, I find myself less aware of a probabilistic infrastructure than I have been in the past when I listened to Cage’s use of the I Ching.

Personally, I think that this speaks well for Campion’s techniques as a composer. To evoke Bell Labs again, the mathematician Richard Hamming began his textbook Numerical Methods for Scientists and Engineers with the motto: “The purpose of computing is insight, not numbers.” All four of the Quadrivium movements prompt insights from the attentive listener. Where polyphony is involved, particularly in the “Musica” movement, the insights emerge from not only the interplay of the four performers (Tod Brody on flute, Peter Josheff on clarinet, Daniel Kennedy on marimba, and Patricia Plude on piano) but also the imaginative approaches to rhythms and sonorities. (Regular readers may recognize those first two names as members of the Earplay chamber ensemble.) In addition, Campion created the electronics for the duo performances in “Mathematica” (with Brody on flutes) and “Astronomia” (with Kennedy on marimba).

I have to say that I find it a bit chilling that Quadrivium has emerged from practices that had taken root half a century ago. However, what I find comforting is that Campion has developed his own techniques to build on those practices. This is the sort of album that can provide a new journey of discovery each time the attentive listener approaches it.