Perceptual absorption of listening context when perceiving musical instruments

Abstract

Brief experience with reliable characteristics of listening context, such as a precursor sentence, alters perception of subsequent vowel sounds. In experiments focused on perception of vowels, effects of preceding context do not depend on sentences being intelligible speech [K.R. Kluender and M. Kiefte, J. Acoust. Soc. Am. 114, 2335 (2003)]. In the present studies, listeners identified musical instruments (tenor saxophone versus French horn) following a context of speech sentences that are modified by spectral envelope difference filters created to capture differences between saxophone and horn spectra. Precursor sentences were resynthesized renditions of a female talker saying ‘‘Please identify this instrument.’’ For a morphed series of sounds with 196 Hz fundamental frequency (G3) varying perceptually from saxophone to horn, listeners are more likely to report hearing a saxophone following a sentence filtered to emphasize spectral characteristics of the horn and vice versa. Consistent with previous studies, these effects are contrastive and indicative of the auditory system absorbing reliable (redundant) properties in the service of emphasizing change and maximizing transmitted information. Perceptual calibration to reliable spectral characteristics does not depend upon sounds being perceived as arising from the same source. [Work supported by NIDCD.]