We propose a series of studies aimed at better understanding how people come to make sense of language input. We use scalp-recorded event related brain potentials or ERPs (such as the N400 component) because (1) language is a human brain function, (2) ERPs are a direct measure of electrical brain activity, offering a non-invasive look at the activity by which neurons communicate within and between the two hemispheres of the brain;(3) the size, timing, and topography of average ERPs elicited by language stimuli track factors important for comprehension, (4) and ERPs do so continuously and virtually instantaneously as neural and cognitive processes unfold moment by moment. N400 amplitude is a sensitive index of the processing that occurs in real-time as different knowledge sources converge in language comprehension. We and others have shown that language processing relies heavily on both knowledge of language and knowledge of our experiences with the world. But what world knowledge, and how and when does its organization affect online language processing? Several neuroscience models have converged on mental representations of events (including who, what, where, when, how, why) as key to organizing perception, memory, and action. We propose to investigate specific hypotheses and assumptions about events in language processing - i.e., that event knowledge plays a critical role in the incremental and dynamic construction of meaning. We propose three series of experiments using questionnaire norming, reading times (a canonical measure of language processing), and ERP and EEG frequency measures to test the role(s) of events - event-structures and sequences - in real-time language comprehension. We will compare the ERP patterns during comprehension with reading times for several critical experiments and conduct visual half field versions to assess the contributions of each cerebral hemisphere to normal language processing. The 1st set of experiments tests the hypothesis that event knowledge distinct from word-level relations contributes to incremental processing in sentence understanding. To the extent that events are an important organizing principle of human knowledge, we would expect to see incremental impacts of their sequential and temporal structures on language comprehension;we test this hypothesis in the 2nd series of studies. In the 3rd series of studies, we revisit some of these issues with frequency-based (instead of time- based) EEG measures, under the assumption that there is much processing related information in different frequencies only some of which are revealed in conventional ERP signals that are phase-locked to language input.
We all use language routinely for multiple purposes - to communicate, to learn, to persuade - some more effectively than others. Yet, we still have embarrassingly little idea of just how our brains use what we know and see to """"""""comprehend"""""""" language. Deciphering the brain's code for language processing will have untold benefits to normal and abnormal language users in all walks of life - be they taking part in conversations, delivering speeches, making demands, taking vows, offering instruction, conveying facts, reporting news, selling products, selling ideas, or just listening and reading.
Showing the most recent 10 out of 118 publications
