At first I didn’t find cellular automata very interesting. As a simple way to generate complex patterns, yes, they were neat, but the black-and-white pixelated pyramids they tended to generate had a samey quality that didn’t seem super artistically interesting to me. It wasn’t until I saw Gwen Fisher’s “Pixel Paintings” that I changed my mind. Gwen, a mathematician and visual artist, had incorporated colors and imperfections into her use of multistate cellular automata. The paintings hovered somewhere precariously between abstract and representational art: you could see, or imagine, lava flows, rivers, hanging gardens.
Immediately I thought this had musical implications, and I set about mapping colors to everything I could think of: pitches, samples, timbres, more esoteric parameters. The results were immediately compelling, and fun to play with. For example, the initial conditions (or seed) of an automaton can have a great impact on the automaton’s evolution. Seeding a particular automaton with random values causes it to evolve into other random-seeding patterns (Figure 1), while seeding it with a a row of a single value causes it to settle into a periodic fluctuation of more rows containing only a single value (Figure 2). Initializing it with the seed 1111211111112111 yields a more intricate and coherent pattern that displays many symmetries and near-symmetries (Figure 3). This seed can be thought of in musical terms as a backbeat, with accents on beats 2 and 4 of a 4/4 measure. In other words, seeding an automaton with a musically cogent patterns can yield other musically cogent patterns. Another interesting feature of this pattern is that it is a long-term stable structure; after 41 rows (or measures), it repeats.
The patterns also respond satisfyingly to interventions, like putting your hand underneath a faucet and watching it change the flow of the water below. They also present an interesting challenge to the listener/observer, who has to decide what information is most musically relevant. This perception can change depending on how the automata are turned into sound, from noise to percussion to melody and harmony. These shifting representations seem to interrogate how we as listeners derive musical meaning from patterns, and in larger sense, how we decide what information is relevant to us. Often in looking for patterns we see something that’s not there. It turns out there is a lovely word for this: apophenia. The abbreviated form of this, Apop, became the piece’s title.