What happens when a voice and its machine-made copy keep changing each other?
What happens when a voice and its machine-made copy keep changing each other?
The performer repeatedly says "spoken words." A machine listens to their voice and returns its pitch and movement as synthesized sound. Each response becomes denser, until imitation develops into an exchange that neither side fully controls.
Contemporary algorithms often act like mirrors, learning our expressions and returning amplified versions of them. Spoken Words stays with that mirror. The performer keeps answering their imitation until the exchange becomes an argument, a collision, and eventually something close to a dance.
Spoken Words began as word play: repeating a phrase until it loses its meaning, a game familiar to most people. Returning to it alone, the voice starts to sound like something that asks for an answer.
That reversal pointed to imitation as a structure running through many media, where a machine learns what we say and returns it slightly changed. Reading around the idea suggested it could carry a performance, and the work took shape with a machine as the other side of the exchange.
Created by Yohta Kitagawa and Nathan Petree. Performed by Yohta Kitagawa.