3-189 - Statistical Learning and Native-Language Learning are Linked Throughout Development

Session Type: Paper Symposium

Integrative Statement

Research suggesting that infants can learn statistical regularities in artificial languages initiated a sea-change in theories of language acquisition by incorporating a substantial (vs. incidental) role for learning. This symposium highlights three lines of research that test a key prediction of such theories: If statistical learning is a mechanism of language learning, then individual differences in statistical learning and native-language development should be related. The first presentation shows that infants’ learning of phonotactic regularities is related to native-language development, even when controlling for general cognitive development. The second presentation shows that 15-19 month-olds’ statistical learning ability is related to online language-processing. Online language-processing is related to language input and language outcomes, suggesting an intricate connection between statistical learning and language development. The third presentation reveals that behavioral and neural measures of statistical learning in children are related to native-language proficiency only in families with lower levels of caregiver education. Thus, environmental factors may buffer the effects of having weaker learning skills. The discussant has investigated how environmental and cognitive factors relate to language outcomes, and will present a critical discussion of the presentations and their implications.

While recent published studies suggest that statistical learning is related to native-language proficiency, the presentations in this symposium go beyond these preliminary findings. They refine measures of assessment of statistical learning, native-language development, and provide more stringent tests of the selectivity of the relationship. Ultimately, these presentations shed light on role of statistical learning in language acquisition.

Sub Unit

• Language, Communication

Chair

• Katharine M. Graf Estes, University of California, Davis

Discussant

• Rochelle Newman, University of Maryland, College Park

Individual Presentations

• Finding patterns and learning words: Infant phonotactic knowledge is associated with vocabulary size - Presenting Author: Katharine M. Graf Estes, University of California, Davis; Stephanie Chen-Wu Gluck, San Francisco State University; Kevin John Grimm, Arizona State University

• Infants’ Statistical Learning Ability is Related to their Real-Time Language Processing Skill - Presenting Author: Jill Lany, University of Notre Dame; Amber Shoaib, University of Notre Dame

• The Role Played by Statistical Learning in Language Development Depends upon the Quality of the Social/Linguistic Environment - Presenting Author: Christopher Conway, Georgia State University; Leyla Eghbalzad, Georgia State University; Joanne Deocampo, Georgia State Univ