Seminars

  • Founded
    2000
  • Seminar Number
    681

What can the study of language contribute to our understanding of human nature? This question motivates research spanning many intellectual constituencies, for its range exceeds the scope of any one of the core disciplines. The technical study of language has developed across anthropology, electrical engineering, linguistics, neurology, philosophy, psychology, and sociology, and influential research of the recent era of cognitive science have occurred when disciplinary boundaries were transcended. The seminar is a forum for convening this research community of broadly differing expertise, within and beyond the University. As a meeting ground for regular discussion of current events and fundamental questions, the University Seminar on Language and Cognition will direct its focus to the latest breakthroughs and the developing concerns of the scientific community studying language.

Seminar Website


Chair
Robert Remez
remez@columbia.edu

Rapporteur
Xinming Zhou
xz3116@tc.columbia.edu

Meeting Schedule

09/21/2023 Faculty House, Columbia University
4:00 PM
POSTPONED--How infants make sense of speech, and how we can find out
Daniel Swingley, University of Pennsylvania
Abstract

Abstract

Laboratory experiments from the early 1970s to the 1990s set much of the agenda for infant speech perception research to the present day, emphasizing infants’ cognitive skills and using theoretical tools given by the cognitive psychology of perceptual categorization. I will argue that it is time to reconsider. Infants are both more and less competent than we once thought. Even more important, parental speech does not appear to have the properties required and presupposed by the usual “statistical learning” accounts of category learning and word-form discovery. I will defend these points and propose that we should use different intuitions in contemplating the infant’s path toward language understanding. In particular, we should try to imagine the infant’s organic path, rather than setting the infant targets from the mature language.





10/26/2023 Faculty House, Columbia University
4:00 PM
Cognitive reserve in aging and dementia
Yaakov Stern, Columbia University
Abstract

Abstract

The concept of cognitive reserve posits that there are brain mechanisms that allow some people to cope better than others with age-or disease-related brain changes. There are a set of life experiences that seem to be associated with greater ability to cope with brain changes. In addition research has explored differential task related activation as a potential moderator between brain change in cognition. This talk will review studies that helped develop and that support this concept, including those using epidemiologic and functional imaging approaches. I hope that we can discuss the special role that language might play in these relationships.





11/30/2023 Faculty House, Columbia University
4:00 PM
How infants make sense of speech, and how we can find out
Daniel Swingley, University of Pennsylvania
Abstract

Abstract

Laboratory experiments from the early 1970s to the 1990s set much of the agenda for infant speech perception research to the present day, emphasizing infants’ cognitive skills and using theoretical tools given by the cognitive psychology of perceptual categorization. I will argue that it is time to reconsider. Infants are both more and less competent than we once thought. Even more important, parental speech does not appear to have the properties required and presupposed by the usual “statistical learning” accounts of category learning and word-form discovery. I will defend these points and propose that we should use different intuitions in contemplating the infant’s path toward language understanding. In particular, we should try to imagine the infant’s organic path, rather than setting the infant targets from the mature language.





02/22/2024 Faculty House, Columbia University
4:00 PM
An autoregressive conversational dynamics model for dialogue systems
Rivka Levitan, Brooklyn College; CUNY Graduate Center
Abstract

Abstract

In this talk, I describe a model trained to represent conversational dynamics by predicting prosodic and language output based on autoregressive input from the target speaker and their interlocutor. Separate attention mechanisms for each speaker allow us to examine which of the interlocutor's turns are considered influential for the prediction of a target turn. Using this novel approach, we formulate and evaluate several hypotheses about how a speaker's conversational speech influences that of their interlocutor (often called “entrainment”), and discuss how these attention scores can be used to sketch the discourse structure of a dialogue. These findings extend previous work on entrainment by expanding the hypothesis space to search for nonlocal and nonlinear dynamics.