2-154 - From input to uptake in early language development: Insights from bigger data

Session Type: Paper Symposium

Integrative Statement

A major challenge in the study of early language is the astonishing variability of children’s developmental trajectories. This variability can make it difficult both to build quantitative theories and to identify atypical outcomes. In recent years, one major approach to overcoming this variability has been to leverage larger datasets. This symposium brings together researchers with diverse questions and perspectives who take this approach. The papers center around two core questions: How is lexical development shaped by visual and parental input (Papers 1 and 2), and which factors influence the trajectory of both typical and atypical language development (Papers 3 and 4)?

The first paper examines how the natural statistics in infants’ learning environments provide the foundation for learning object words, synthesizing data from a large corpus of visual scene data. The second reports data from an intensive longitudinal study, examining the intersection between parental input and infant word comprehension. The third draws upon a large-scale data aggregation project, highlighting the influence of environmental factors and word-level factors. The fourth uses a large longitudinal population cohort study to examine factors influencing the language development of late talkers, with specific focus on the trajectory of late talking and the nature of late-talkers’ semantic categories.

Together, these papers showcase the breadth of research in this area in terms of questions and methodology (i.e., behavioural, eye movements, report measures) while all highlighting how large datasets can yield critical empirical and theoretical insights into key questions about early language.

Sub Unit

• Language, Communication

Chairs

• Susan Graham, University of Calgary

• Michael Frank, Stanford University

Individual Presentations

• Visual object training datasets for babies - Presenting Author: Caitlin Fausey, University of Oregon; Elizabeth Murphy Clerkin, Indiana University; Linda Smith, Indiana University

• Semantic Relatedness Effects in Early Word Comprehension - Presenting Author: Shannon Dailey, Duke University; Elika Bergelson, Duke University

• Building quantitative theories of early language using large-scale parent report data - Presenting Author: Mika Braginsky, MIT; Dan Yurovsky, University of Chicago; Virginia Marchman, Stanford University; Michael Frank, Stanford University

• Late talkers at 36 months: Findings from a longitudinal, population-based cohort - Presenting Author: Beverly Collisson, University of Calgary; Susan Graham, University of Calgary; Cassia Tremblay, St. Francis Xavier University; Sarah Rose, Alberta Health Services; Sheila McDonald, Alberta Health Services; Suzanne Tough, University of Calgary