Speech perception in early infancy: Discrimination of speech-sound categories

Abstract

The sucking-habituation technique was used to test 4- to 16-week old infants' tendencies to discriminate changes in formant frequency (synthetic /a/ vs /i/) in the presence of irrelevant variation in the pitch contour of the stimuli, and their tendencies to discriminate changes in pitch contour (steady vs rise-fall) in the presence of irrelevant variation in the formant frequencies of the stimuli. Analysis of the sucking rates for the 2 min preceding and the 2 min immediately following the fulfillment of the habituation criteria for experimental (stimulus change) and control (no stimulus change) infants demonstrated discrimination of formant changes regardless of the presence or absence of irrelevant variation in the pitch contour of the vowel. While infants demonstrated discrimination of the pitch-contour change in the absence of an irrelevant change in the vowel, they did not discriminate the pitch change in the presence of an irrelevant change in the vowel. It took longer for infants to meet sucking-habituation criteria when an irrelevant dimension was varied. Infants ran as “silent controls” produced fewer high-amplitude sucks than infants for whom sound presentation was contingent on high-amplitude sucking [Supported by NIH Grant NS 03856 to CID.]