UC San Diego

Young children who have difficulty learning words compared to their peers are at risk for continued language delays as they grow older. While most late talkers (LT) will move into the normal range of language abilities by school age, approximately 17-26% of LTs will be identified with SLI by 5 years of age (Paul, 1996; Rescorla, 2002, Rescorla & Dale, 2013). Accordingly, one critical question is how to best identify which LTs will have SLI at school age, in order to maximize the opportunity for early and appropriate intervention and support. Longitudinal studies of language outcomes in LTs have had only moderate success at accurately predicting which LTs will be identified as having SLI by 5-6 years of age. Research shows that the individual's lexical and cognitive processing abilities (e.g., speed of processing, memory and implicit learning) play an important role in language development (Fernald & Marchman, 2012; Graf Estes et al., 2007) and in language deficits in children with SLI (e.g., Ellis Weismer et al., 1999; Evans et al., 2009; Leonard et al., 2007; Montgomery, 2000; Montgomery & Evans, 2009; Tomblin et al., 2007). Therefore, using the lexical and cognitive processing deficits characteristic of children with SLI to inform the study of lexical and cognitive processing in infants may prove more successful than traditional static measures of vocabulary alone, in identifying LTs at the greatest risk for SLI. This dissertation utilizes fine-grained eye tracking methods to examine multiple real-time measures of lexical and cognitive processing to begin to identify characteristics of Late Talkers who may be at risk for Specific Language Impairment. Chapter 1 provides the general background and support for the questions addressed in this dissertation. In Chapters 2-5, data from four studies are presented. Specifically, speed and accuracy of processing familiar real words in a group of LT (N = 14) and TYP (N = 14) 18 month old infants is compared (Study 1). In study 2, speed and accuracy of processing of newly learned novel words in a group of LT (N =14) and TYP (N = 14) 18 month old infants is examined. Study 3 further examines infants' abilities to use transitional probabilities to discover word boundaries within a stream of speech is compared in the same group of LT (N = 14) and TYP (N = 14) 18 month old infants. Study 4 examines lexical processing, novel word learning and implicit learning in a larger group of 54, 18-months, using an alternative distributional approach. Finally, Chapter 6 informs both theoretical models and clinical implications of word learning and vocabulary development in LTs and provides insight into the mechanisms that underlie individual differences in language abilities