The Frame Problem in Speech Communication: Defining the Dimensional Space for Phonetic Categorization
Date of Original Version
Abstract or Description
Much theoretical and empirical work has been focused on how language learners/users parse multi-dimensional auditory spaces into the phonetic categories of a native or second language. A more fundamental question is how the listener determines the relevant dimensions for the perceptual space in the first place. Harvey Sussman has offered one of the only principled theoretical accounts for the existence of particular perceptual dimensions – neural columnar encoding that leads to relative dimensions. The framework of Sussman’s theory – interactions of neural processing constraints with statistics of the input – provides insights into potential answers to the more general questions about how listeners can solve the “frame problem” of which dimensions are most relevant to an auditory categorization task.
This document is currently not available here.
Topics in Speech Processes: Language Production and Perception, Festschrift in Honor of Harvey Sussman, forthcoming.