How do children learn their first words? Psychology has adopted a myopic approach to this question suggesting either that they learn through association (Smith), or social pragmatics (Tomasello) or attention to linguistic input. This talk presents a hybrid theory of early word learning, The Emergent Coalition Model. Using a "wide-angle" lens on development, it suggests that children use perceptual, social and linguistic cues to solve the problem of mapping word to world. While all of these cues are available to learners at all times, they are differentially weighted throughout the course of development such that children first rely on associative perceptual mechanisms, then social information, and then language cues as they determine the mapping from word to world. In this process-based model, word learning principles become the products rather than the engines of development and no one theory reigns supreme. This talk not only presents the theoretical argument for a dynamic model of language development, but also provides the empirical data to support the model. Finally, we demonstrate how this broader model of language growth not only explains the changing nature of vocabulary growth in the first two years, but also insights into how atypical children (autistic children) might learn language.
Verbs are difficult to learn. Most of the studies in early language acquisition report that nouns are learned before verbs and that nouns predominate in the vocabularies of young children. Yet, verbs do appear among children's first words. How do we reconcile these facts? In this talk, we suggest a developmental story ABOUT HOW children learn verbs. Our hypothesis is based on a common hurdle that children encounter in cognitive development: mapping relations across domains. To make our case we review studies from our labs on the non-linguistic conceptual foundations of verbs and on the role that linguistic and social factors might play in the process of verb learning.