Grammar learning across languages

Children learn how subjects, objects, and verbs combine into sentences irrespective of the language. The cognitive mechanisms underlying this incredible flexibility, however, are still poorly understood.

Current theories suggest that pattern tracking and cognitive biases underlie grammar learning, but fail to account for the full breadth of grammatical diversity.

In this project, I will set aside common theories and assumptions, and use a data-driven approach to investigate the mechanisms for grammar learning by sampling a diverse set of twenty languages.

From this larger, representative perspective, I will re-characterize the roles of (1) pattern tracking and (2) cognitive biases to advance theory, and synthesize my findings into (3) a database to facilitate cross-linguistic research.

I leverage cutting-edge technological advances in natural language processing and online experimentation to provide a window into the early flexibility of human cognition.