One of the most contentious questions in psycholinguistics is whether language can be divided into separate rule-based and memory-based systems-the so-called dual-system model-or whether, alternatively, the entire system can be subsumed under one kind of computation. The answer has implications for the way language is acquired by children, used by adults, built into machines, and remediated after brain injury.
The aim of this project is to test the dual-system model and its alternatives in a novel domain: genetics. Based on previous work using aphasics, Parkinson's patients, and neuroimaging with healthy subjects, it is known that the left frontal lobe, and in particular striato-cortical dopaminergic circuits in the left frontal lobe, play a role in rule-based but not memory-based processes within language. We propose to test whether normal variation in genes that encode the brain's response to dopamine (including DRD2, DRD4, DAT, and COMT) is correlated with variation in performance on rule-based but not memory-based language processing in a large, population-based sample. A positive finding would provide new evidence for the dual-system model as well as suggesting new avenues for thinking about the structure and genetics of language.
