Introducing Composed Instruments: Technical and Musicological Implications

Source: Norbert Schnell and Marc Battier, “Introducing Composed Instruments: Technical and Musicological Implications,” Proceedings of NIME 2002, pp. 156–160. (Schnell & Battier, 2002) Local PDF

Central claim

A computer music system can function as an instrument and a score at once because its algorithms, mappings, representations and settings determine parts of a work while leaving others open to performance.

Evidence and method

Schnell and Battier develop the concept through a short historical and technical argument rather than an experiment. They move from early electronic instruments and studios to live-electronic works by Philippe Manoury, Yan Maresz and others, then describe computer music through the linked categories of instrument, machine and representation.

The sound-producing system and gestural interface are decoupled in a composed instrument, so an input can control sound through an indirect or non-obvious relation (p. 156). The performer’s freedom lies between two limits: triggering a fully determined result and controlling each detail as on a violin (p. 158).

The system may contain data structures, algorithms, procedures, mappings and configuration settings (p. 158). Streams can pass through transformation, analysis, synthesis and memorisation modules, while object-based and functional models support other forms of musical control (pp. 159–160).

Key concepts

• Composed instrument: A system in which instrument design and compositional decisions form one object.
• Instrument, machine, representation: The system supports exploration, executes computational procedures and stores musical events and rules.
• Determination continuum: The system distributes control between prior design, machine execution and live input.
• Decoupled mapping: The relation between gesture or data and sound does not have to be direct.
• Streams and modules: Inputs become audible through linked analysis, transformation, synthesis and memory processes.

Limits

The paper proposes a concept and design vocabulary without testing listener perception, creative practice or long-term use. Its examples concern specialist performance systems rather than consumer apps, subscriptions, sensor permissions or platform ownership. It cannot establish how any specific generative music app works.

Link to generative functional apps

A generative functional-music app can be read as a composed instrument whose score, sound engine and interface are partly joined. The listener may choose a function or start playback after musicians and developers have defined the assets, mappings, rules and possible behaviours.

The determination continuum does not turn each listener into a performer. Selecting “Focus” lies close to the triggering end. Body or context data can alter the result without becoming an intentional musical gesture, so the relevant question is which decisions belong to the artist, platform, runtime system and listener.

references