Linguistics Colloquium: Navin Viswanathan (SUNY New Paltz)
SEP 12, 2013 | 4:15 PM TO 6:15 PM
The Graduate Center
365 Fifth Avenue
September 12, 2013: 4:15 PM-6:15 PM
Do listeners perceive speech gestures: A critical examination of compensation for coarticulation.
A fundamental, unsolved puzzle of speech perception is how listeners demonstrate robust speech perception despite a widely varying acoustic signal. For instance, the acoustic manifestations of a segment depends on several factors such as speech rate, speaker identity, dialect, conversational context, etc. In this talk, I will focus on the question of how listeners “cope with” variability due to the properties of neighboring segments (i.e. due to coarticulatory context). I will present a series of experiments that examine the phenomenon called compensation for coarticulation (Mann, 1980). In particular, I will examine what properties of coarticulating segments (acoustic energy relations or overlap in gestures of speech production) determine listeners’ compensation for coarticulation. Through these studies, I will evaluate whether a direct realist gestural account (Fowler, 1986) or a spectral contrast auditory account (Diehl, Lotto, & Holt, 2004) provides the best explanation of the findings. Broadly, the implications of these findings for the question of objects of speech perception (i.e. do listeners perceive vocal tract gestures or the acoustic signal itself?) will be discussed.