Some Thoughts on "Normalization" in Speech Perception

Author: David B. Pisoni

Abstract:
One of the central problems in speech perception concerns stimulus variability, specifically, the mapping relations between acoustic attributes of the signal and linguistic categories resulting from perceptual analysis. Previous accounts of speech perception have treated variability as an undesirable source of noise in the signal that must be reduced or eliminated through a process of perceptual normalization. In this paper, I address what people mean by the term normalization in speech perception and examine the definitions and some of the implicit assumptions that have been made about it in the past. Then I summarize the results of several recent experiments of talker variability and perceptual learning which show that listeners encode fine stimulus details and use indexical attributes of speech in word recognition and sentence perception. Finally, I describe an alternative view of speech perception that is based on ideas from nonanalytic cognition. According to this approach, stimulus variability is "lawful" and "informative" for perceptual analysis. Perceptual normalization, as discussed in past theoretical accounts, may not involve a true "loss" of information but rather may entail the encoding of specific instances and the details of perceptual analysis. The generation of "equivalent forms" may occur at the time of retrieval from memory as a result of computational processes rather than early in perceptual analysis and encoding as previously assumed. This approach to speech perception and word recognition provides a new way of dealing with a number of long-standing problems in the field.