Prof. Rachel Theodore gives Colloquy

ABSTRACT

If we can learn anything from binge-watching Law and Order, it is that eye-witness testimony can be very unreliable. This is not surprising given the literature on memory for spoken language, which shows that although memory for the meaning of our interlocutors' messages is quite good, memory for the specific form of the message is relatively poor. Poor memory for the form of language is striking when considering that for early stages of mapping from speech to meaning, listeners show exquisite sensitivity to the acoustic-phonetic structure of speech. In this talk, I will review findings from on-going work in our laboratory that examines listeners' ability to dynamically adapt to structured phonetic variation, focusing on variation associated with individual talkers' idiolects. These studies examine the mechanisms that allow listeners to exploit structured variation for speech perception and voice recognition. In doing so, these findings will help explicate a theoretical framework that can account for the tension in a linguistic architecture that uses both abstract and instance-specific representational knowledge to guide language comprehension.