
Abstract
Theories of language acquisition generally fall into one of two camps. On one side, there are domain-general constructivist theories which argue that children break into language with very little beyond a capacity to learn, an interest in others, and some knowledge of the world. On the other side, there are domain-specific nativist theories which propose that children come equipped with rich innate knowledge of syntax, semantics, and the principles that allow them to map between them. Jesse Snedeker and I have recently proposed an alternative: a theory that roots language acquisition in rich, abstract conceptual structures (akin to the semantic structures of linguistic theories), but posits that syntax arises from domain-general learning abilities and the belief that the communicative signals of others convey thoughts which are structured like our own. On this hypothesis, both the learning abilities that drive syntax acquisition and the conceptual structures themselves are available to prelinguistic infants and evolved prior to the emergence of external language. The critical evolutionary step was the insight that motivates children to link conceptual structures to patterned external signals. I describe how this theory builds on empirical insights from language acquisition, developmental psychology, linguistic typology, computational linguistics, and learning theories. Finally, I attempt to translate between Conceptual Nativism and other theoretical perspectives, delineating where the theories differ in their core commitments from where they differ solely in their framing, terminology and emphasis.
Given by:
Dr. Joshua Hartshorne
Research Associate Professor, Communication Sciences & Disorders, MGH Institute of Health Professions