UC San Diego

A growing number of functional MRI studies have examined age-related changes in language organization. However, existing studies have predominantly examined differences between children and adults using cross-sectional designs and have been limited to a single language component studied at a single point in time. Thus the mechanisms by which cognitive changes occur over time are still uncertain. A better understanding of developmental changes in the brain organization for language might broaden our understanding of cognitive development, elucidate the causes of atypical development, and could potentially inform our diagnosis, treatment, and prevention of developmental disorders. The current project is founded on a comprehensive theory of language development, which predicts organization from predominantly sensorimotor- based (bottom-up) processing in early stages to more strongly top-down controlled processing in later childhood and adulthood. Applied to two language tasks that tap into the two primary components of language development (lexicosemantic and morphosyntactic), this theoretical perspective generates specific hypotheses about developmental change, which were tested both in cross- sectional and longitudinal analyses. Subjects completed two language tasks: lexical semantic decision (assessing the semantic congruency of sentences describing objects) and morphosyntactic judgment (assessing the grammaticality of sentences) during functional MRI. Forty-one children in two age groups (7 and 9 years) were scanned and brought back for a second scan (n=30) after twelve months, providing a longitudinal component to the study. An adult group (n=15) was scanned at one time point as a comparison. Longitudinal and cross-sectional analyses of the data were performed to examine the interactions and main effects of group (7-year olds, 9-year olds, adults), task (lexicosemantic or morphosyntactic) and time. We found that both the lexical-semantic and morphosyntactic judgment tasks were associated with left lateralized fronto-temporal networks that were overall similar to those seen in adults. Our findings of age- and time- dependent activation increases in left frontal and parietal networks are consistent with a model of age- dependent strengthening of 'top-down' control mechanisms during language processing. However, we did not find evidence of language development emerging from sensorimotor abilities. In fact, we found inverse results with age- and time-dependent increases in bilateral middle and superior occipital gyri for both tasks. This dissertation is only the second longitudinal study of neurolinguistic development in children and the first to utilize sentence-embedded tasks for two language domains. While our results provide additional evidence for increased top-down control in the process of language learning, more time points are needed in future longitudinal studies to examine possible non-linear development in sensorimotor regions. The present study, which was largely exploratory due to its novel methodology, provides regions of interest for correlation analysis with behavioral developmental markers for future longitudinal studies