received from chris brown on aug 16 1995 copyright 330w
by Chris Brown
written for the Abel-Steinberg-Winant Trio
"Often an animal's ... behavior is not immediately traceable
to any particular input event. The number of triggers
needed to fully control every sonic movement of an
'interactive' composition of the complexity of a housefly
would already be quite high" George Lewis
I decided to respond to this remark by developing an
interactive computer music program to emulate the behavior
of a fly. My composer-fly has extremely limited sensory
inputs; it hears only loudness and tone color, but it hears
them directly (through the numbers of a DSP chip). When it
invents a response to what it hears, it holds it, circling
around its host, searching for the right moment to alight.
It is stimulated by motion, heat, taste, and touch. It is
tenacious, with a strong will to live, to feed, to fly, and
to reproduce. When irritated, it could bite. Mostly, its
hearing, like its vision, is local.
As for its musical education, my fly knows only one scale
with two generating intervals (a major second followed by a
minor third), and it doesn't know about octaves. It
transposes simple patterns very well though, and can
retrograde effortlessly. In short, it has only a few
stimulus-response behaviors; but it is sensitive to the
proportions of sound and silence, and uses this primitive
perception of phrasing to make its decisions about when and
how to respond.
I wrote a violin piece full of dynamics and timbre changes
to feed my flies. My flies program made the percussion and
piano parts by listening to a recording of the violin piece.
In performance, the same violin part drives the motion of
the electronic flies that circle around the stage from
speaker to speaker.
[FLIES aesth. purpose]
READ THE SURFACES OF MUSICAL EVENTS
TRANSLATE THEM INTO OTHER LANGUAGES
TRACING THE CONTOUR OF ONE MEDIUM INTO ANOTHER,
THE TIME DIMENSION MAY BE FREELY RESCALED
TOP OF DOCUMENT