Conference: Language and thought in minds and machines

Massachusetts Institute Of Technology, Cambridge MA

Investigators

Abstract

Language is fundamental to human experience. For example, language allows humans to build deep social connections, teach each other new knowledge and skills – including accumulating knowledge across generations – and represent complex concepts in compact ways. This expands humans’ ability for complex reasoning. In recent years, advances in artificial intelligence (AI) have led to the emergence of large language models (LLMs), such as ChatGPT, which are transforming society. This project supports a workshop at a premier conference for interdisciplinary approaches to language science. The workshop brings together invited and contributed speakers who create a scientific dialogue on language and thought in minds and machines, including the relationship between language and complex cognition in both human minds and AI systems. Other benefits to society include innovative educational opportunities, including mentoring, that support workforce development for AI and other language technology sectors. It has long been proposed that language is critical for reasoning – that humans use words and linguistic structures for thinking. Others have contested this claim and instead proposed that language is primarily a communication system, not a thinking tool. Many specific questions arise from this ancient philosophical question, including (1) whether language and thought are supported by the same or distinct cognitive and brain systems, (2) if complex thought is possible without language, and (3) does knowledge of a particular language change how we think about the world? Given the rise of LLMs and other AI systems, both in cognitive science research and society at large, there is a critical need to understand whether, and if so how, these models can inform research on human language and cognition. This workshop brings together researchers across diverse fields, such as computer science, psychology, linguistics, philosophy, and anthropology to address these questions. Through interactive discussions and presentations, attendees explore the role of language in broader cognition, including how language science can inform research on AI models and how AI models can advance a scientific understanding of the language-thought relationship in humans. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.