Using written language to probe speech recognition models. Speech recognition models fall into two principal classes, with fundamentally different processing architectures. Feedback models (e.g. TRACE, McClelland & Elman, 1986) allow lexical knowledge to exert top-down control over phonemic analysis. Feedforward models (e.g. Merge, Norris, McQueen & Cutler, 2000) assume that information flow is entirely bottom-up. Our project adopts an innovative approach to testing between these model classe ....Using written language to probe speech recognition models. Speech recognition models fall into two principal classes, with fundamentally different processing architectures. Feedback models (e.g. TRACE, McClelland & Elman, 1986) allow lexical knowledge to exert top-down control over phonemic analysis. Feedforward models (e.g. Merge, Norris, McQueen & Cutler, 2000) assume that information flow is entirely bottom-up. Our project adopts an innovative approach to testing between these model classes, by examining the influence of written-word knowledge on speech perception. To distinguish the models, contrasts must test different processing levels and examine strategy effects. TRACE favors broad effects with limited strategic influence; Merge favors lexical effects that are necessarily sensitive to strategic factorsRead moreRead less
The seeds of literacy in infancy: empirical specification of the acoustic determinants of language acquisition. Reading is one of the most difficult skills we learn, and while the process is largely forgotten by adults, any minor difficulty can have lasting effects. This project will follow speech, vocabulary and reading in infants at or not at risk for dyslexia from six months to five years with implications for parent-child interaction and language delay intervention.