(How) Can Infants Learn Words From Statistics?

Dr Jill Lany, University of Notre Dame, will give the next LuCiD seminar in Liverpool on Tuesday 16th January, 11am. She will discuss how children can learn words from statistics. Venue details to follow soon.

Abstract

Accumulating lexical knowledge is a fundamental achievement in early language development. Lexical development is likely to draw on multiple learning systems, including those involved in tracking sequential structure within speech, and also in detecting relations between speech and other features of the environment. I will discuss recent research highlighting how infants’ ability to track different kinds of statistical regularities may support both of these aspects of word learning. First, tracking statistics within speech, such as frequency and co-occurrence, may support learning both associative and referential mappings between words and their referents.  Second, overlap in the organization of structure within speech and in the environment may facilitate determining what words mean, and may even contribute to the unique referential character of language. I will also consider how the relative importance of these processes in word learning may change across development.

Leave a Comment

* Indicates fields are required