What this models
This activity treats each participant as an adaptive agent in a shared acoustic environment. There is no conductor and no central plan: each person responds locally to what they hear. The group-level pattern emerges through feedback, adjustment, perturbation, and mutual listening.
v.1 is designed for a co-located group, or for a remote group using a shared audio call with echo cancellation carefully managed. The web page itself does not synchronise participants through a server.