CAREER: Mapping the neural locus of cognitive processes in word reading

点击翻译按钮获取中文摘要

The ability to read written words depends on a complex series of transformations, ranging from patterns of stimulation on the eye to neural representations of a word's spelling, pronunciation and meaning. These transformations occur quickly and automatically in literate adults, despite the fact that written language is a relatively recent cultural invention and that reading is something learned in school rather than acquired without formal teaching. Rapid advances have been made in the understanding of how people read words - both in terms of better models of the computational processes involved in these transformations and better understanding of how the brain responds to written stimuli. This project aims to bridge these computational and cognitive neuroscience approaches, using brain activity data from neuroimaging studies to answer some fundamental questions about how people read. What code does the brain use to recognize written words? How do individuals differ in how they process written words? Do people read words differently depending on the context in which they are read? In answering these questions, the ultimate goal of this research is to develop a neurocognitive theory of reading that can provide critical insights into how people read words. This in turn can benefit society by informing literacy education, the treatment of learning disabilities or the remediation of language loss after brain injury.This project's goal will be achieved by collecting functional neuroimaging data while people read, and analyzing it with an approach that maps between different cognitive reading processes and different patterns of brain activity. By using this approach, the PI can identify the pattern of brain activity elicited by individual stimuli, and calculate the similarity in the pattern between all pairs of stimuli. They can then compare these brain-based similarity measures to formal predictions of similarity derived from computational models of reading. Consider how DOUGH relates to the words TOUGH, SEW and BREAD; DOUGH is spelled similarly to TOUGH, sounds similar to SEW, and has a meaning related to BREAD. Using the logic of this analysis, brain regions in which DOUGH elicits a similar response to TOUGH, but not to BREAD or SEW can be interpreted as regions involved in the neural representation of word spellings. In this way, the current approach provides a tool for linking neural activity to cognitive operations. Using this technique, the proposed research will test competing computational models of the front end of the reading system, evaluate how differences in a task alter reading-related processes and assess individual differences in the cognitive processes used by skilled readers. This project will advance research methods in bridging computational theories of cognition and neuroscience data that could be useful for many questions in cognitive science. Along the same lines, the project includes the educational goal of developing new approaches to teaching students from high school, through college, and into graduate school how to think about the relationship between brain data and cognitive theories.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.