Parse Trees is an approach to live improvised electronic music that I was exploring for a few years. My interests lied in attempting to flatten the syntactical space that seems to naturally impose itself between words and "abstract" sounds. I wanted to see if I could create a music where words and sounds seemed to link up together into grammatical structures; where a word could "point" to a sound as another unit in its syntax tree, and vice-versa. What I found is that the brain resists this quite strongly. It seems to always channel these two types of stimuli: connecting words together into one channel and sounds into another, separate channel running in parallel. In this sense I consider the project a failure, but one that nevertheless landed somewhere else that is interesting for its own sake.

The project began with an invitation to perform at a festival of improvised music. I wanted to play a set that would be entirely improvised but would not sound like "Improvised Music" (as a genre). I also had a common feeling of being bored with my approach to live improvisation at the time. Specifically I felt frustrated by how difficult it was for me when performing live with analog hardware to make very sudden changes in form or texture on the drop of a dime. I always had to re-patch things which takes time and has a high chance that the sound you arrive at won't be what you wanted exactly.

What I ended up doing was creating a large Kontakt library containing hundreds of samples, with each key able to trigger several different sounds depending on velocity. This allowed me to take all the time I wanted doing sound design of the individual samples, and then to have the ability to improvise freely with the material in performance. I enjoyed the act of playing a keyboard where each key made several different sounds, and the imprecision of velocity values kept an aspect of randomness involved that made it surprising to play. I also found the on-stage setup of a guy jamming a shitty midi keyboard plugged into a computer to be pretty funny.

The first few performances used just electronic sounds, but after a couple sets I decided to try incorporating recorded words into the sample libraries, to see if they could function as units within the musical phrases. I incorporated some ideas I like from computer science like comparison operators, which allowed me to create phrases like "{sound} is equal to {other sound}", along with absurdist value-laden ones like "is worse than", and the copious use of the philosophically vague term "this". I put the words on the highest velocity layer of the keyboard so that I could at least decide when to trigger them, but I rarely knew which key would trigger which word or phrase so the grammatical constructs that arose were pretty nonsensical as language.

After a few shows with this setup people kept asking me to explain what the idea was exactly. They'd get it once I articulated what I was trying to do, but it was clear that nobody else was hearing the music the way I was aiming for. I made a couple tracks attempting to force the issue, the clearest of which is included here as "Operator Grammar". Eventually I decided to just embrace how the brain seems to want to connect words together at the exclusion of other types of sounds, and began recording what I thought of as "poems": brief pieces that used repeated, grammatically coherent phrases rather than my imaginary, solipsistic unified syntax.

The release is a collection of my favorite tracks I made using this setup. The "poems" all have titles reflecting their motifs, whereas the improvisations have names gleaned from the academic study of how linguistic expressions are determined to be syntactically coherent. I hope it brings you some form of pleasure.