Studies of Possible Psychoacoustic Factors Underlying Speech Perception

Abstract

As with all sounds which are presented to the ear, speech sounds are subject to processing by the auditory system. Thus, whatever their importance for the listener, whatever the message conveyed, and however the listener may use linguistic knowledge, context, or other non-acoustic cues to assist in decoding the message from a speaker, the speech sounds which are presented to the listener’s ear are subject to the same “obligatory processing” as are nonspeech sounds. In some situations, we might expect that such obligatory processing would result in the diminution of the perceptual salience of an acoustic cue—for example, as a consequence of masking between two portions of a signal. In other circumstances, we might expect that the auditory system would enhance the perceptual salience of an acoustic cue—for example, as a consequence of spectral-temporal integration. Of course the extent to which both perceptual factors and higher-order variables influence the responses which are actually observed in an experiment will reflect the demands of the particular task which is set for the listener (cf., the paper by MacMillan, Braida, Goldberg, and Khazatsky, in this volume, for further discussion of this point).

The research reported here was supported by grants from the Natural Sciences and Engineering Research Council, Health and Welfare Canada, and the Alberta Heritage Foundation for Medical Research. I am grateful to Fred Wightman and Terry Dolan for their hospitality while I was a Visiting Fellow at the Waisman Center, University of Wisconsin, Madison, and to Peter Assman, Meg Cheesman, Vince Dilollo, Bias Espinoza-Varas, Tom Johnson, Linda McEvoy, Terry Nearey, Curtis Ponton, Mike Procter, Anton Rozsypal, and Susan Rvachew for their advice and assistance at various stages of the project. Special thanks are due to Elzbieta Slawinska with whom much of this work was undertaken.