Theoretical approaches to speech perception

Abstract

Theoretical approaches to speech perception have shifted dramatically during the last 50 years. Early theories, such as Alvin Liberman’s ‘‘motor Theory,’’ focused on the psychological mechanisms underlying speech perception rather than listeners’ analysis of the acoustic signal. In the 1970’s, Ken Stevens’ work on the acoustic cues for phonetic categories produced a very different view, an ‘‘invariance theory,’’ that turned theoretical attention to the acoustic signal and the detailed phonetic information that listeners gleaned from its analysis. The last decade has seen another theoretical shift. Strongly influenced by developmental studies, new theories (e.g., Kuhl’s native language magnet model, or NLM) combine the sensory and psychological approaches, arguing that listeners both perform a detailed analysis of language input and psychologically represent that information in a complex way that is unique to each language. The presentation will discuss these shifts and the kinds of data now necessary to produce the next theoretical advance. [Work supported by NICHD HD37954 and Human Frontiers RG0159 to PKK.]