the Syntagm Engine (v1.91)
The Syntagm Engine is an attempt to solve the following problem: How does one combine a performative practice of non-idiomatic free improvisation with an aesthetic of critical sound collage (critical in the sense of detournement, the removal of a piece of culture from its original context and its placement in a new context with the purpose of commenting on or critiquing that material or its origins)?
The answer so far (for me): a relational database of sounds categorized by several variables (like intensity, "rhythmicness", "semiotic density", etc.) which can be easily queried to present source material that is connected. Hence a wide variety of material is available to be juxtaposed and organized via many possible relationships, though the relationships are chosen dynamically by the player.
The name of this new instrument derives from the field of semiotics, in which a syntagm is defined to be a sequence of signs. A traditional example of a syntagm is a sentence, such as "man bites dog." Of course the words, or signs, in this sentence can be rearranged to form a new syntagm with a different meaning, like "dog bites man." Similarly, the Syntagm Engine is a tool for quickly forming sequences of sonic signs.
The Syntagm Engine consists of about 7000 lines of Perl, SQL, and Csound code which run on a Linux laptop. I first performed with it in May 2000 at the San Francisco Electronic Music Festival, and though it has been used in about a dozen shows since then, it is still considered to be software in development.