As a 2025 Schmidt Science Fellow, Luke will explore the ecological and behavioral origins of hierarchical syntax, a defining feature in the evolution of human language.

Luke will use “DeepWild”, a deep learning algorithm developed for wildlife identification and analysis, to create novel training data for tracking the gestures and movements of wild chimpanzees.

Luke aims to determine whether the arm movements of gestural communication are similar to those used when building nests in trees, a deliberative behavior that seems to emulate the hierarchical reasoning and structure of syntax.

Luke will pivot from Evolutionary Biology to Machine Learning