The long-term goal of this project is to improve our understanding of how children learn about the segmental (i.e., phonemic) organization of language, and to begin investigating what goes wrong in this developmental process for children with apparent delays. Studies with normally developing children are converging on the notion that these young subjects do not have as solid or as precise representations of segmental speech structure as adults do, suggesting that one aspect of learning to talk is discovering this structure in the speech signal. One model of how this developmental task is accomplished is that the weights assigned to various acoustic properties of the speech signal are returned as children gain experience with their native language. This 'developmental weighting shift,' as the model is termed, eventually leads to speech processing strategies that optimize the weightings of acoustic properties that most strongly convey information about phonetic contrasts in the child's native language. The four specific aims of this project test hypotheses derived from this theoretical account. Experiments with normal subjects will investigate the weighting schemes of listeners of different ages for stimuli with different syllable structures and phonemic compositions, and evaluate two possible causes of the developmental weighting shift: developmental changes in psychoacoustic capacities and developmental changes in the way acoustic properties are integrated. Perception experiments involving labeling and/or discrimination tasks will be used to accomplish these aims. Experiments comparing data from normal children and from children who are phonologically unaware, compared to their normal peers, will explore the possibility that their difficulty is related to a delay in the developmental weighting shift. Three kinds of experiments will be used to accomplish this aim: one acoustic analysis of speech samples, three speech perception tests, and five related psychoacoustic tests. Results should have implications for children with language impairments, including dyslexia. This work should lead to a better understanding of how normal children acquire sensitivity to the segmental structure of speech, improve our understanding of the problem faced by children having difficulty doing so, and provide a starting point for exploring the problems faced by children with several types of language impairments.
Showing the most recent 10 out of 43 publications
