Iterated singing and cultural transmission

Tracing how musical structure emerges through repeated learning and reproduction.

This project investigates how musical structure emerges through cultural transmission. Using large-scale iterated singing experiments, we track how melodies evolve as they are repeatedly learned, reproduced, and passed between individuals—revealing the inductive biases that shape musical systems.

Participants hear and reproduce unfamiliar melodies, which are then transmitted to the next generation of listeners. Over successive generations, initially random acoustic material transforms into structured, learnable musical patterns, allowing us to observe the emergence of musical regularities in real time.

Iterated singing paradigm and cultural transmission chains

What we do

  • Run large-scale iterated learning experiments using singing and vocal imitation.
  • Analyze how melodic structure, regularity, and compressibility emerge through cultural transmission.
  • Study inductive biases in pitch, rhythm, contour, and tonal organization.
  • Model musical evolution using Bayesian, information-theoretic, and computational learning frameworks.
  • Compare human results with computational simulations and animal learning models.
Emergence of musical structure across generations

Why it matters

Music is a universal human behavior, yet its structural regularities cannot be fully explained by individual cognition alone. By directly observing how musical systems self-organize through repeated cultural transmission, this project provides a powerful experimental framework for understanding the origins of musical universals, cross-cultural variation, and the evolution of auditory communication. More broadly, it sheds light on how complex symbolic systems can emerge from simple learning and imitation processes.

Cultural evolution of pitch and rhythm structure
  • placeholder: iterated singing paper
  • placeholder: cross-cultural music learning study
  • placeholder: datasets and code