Coarticulatory cues in the recognition of spoken words

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Speech comprehension involves discrete categorization at multiple levels of analysis based upon ambiguous information. Explaining the extraordinary speed and accuracy with which these choices are implicitly made requires close examination of the temporal dynamics of the process; the prior information listeners bring to bear, and the analytical mechanisms that guide listeners to the speaker's intended meaning. In the proposed research, the nature of these processes is investigated empirically using anticipatory nasalization in vowels as a test case, both in children and adults. This case is ideal for these purposes because variation in vowel nasalization, which has clearly demonstrable effects on interpretation in adults, can be experimentally manipulated. The contrast between nasal and oral vowels offers a window into how listeners' defaults and biases are exercised, as speech is comprehended. The project focuses on the time course with which evidence is evaluated with respect to prior expectations. To this end, the proposed research will make use of a recently developed paradigm, the "visual world" eye-tracking paradigm. In this paradigm, participants' eye movements are monitored as they follow spoken instructions to click on one of four pictured objects displayed on a computer screen (e.g., "Click on the can"). The pattern of eye movements generated as the sound pattern of the referent object is heard is taken to reflect the ongoing lexical interpretation of the acoustic signal. Furthermore, eye gaze to visual referents is a measure that can be used with young children, permitting to compare the processing of coarticulatory cues in adults and young children.