Training new fricative contrasts

Abstract

A perceptual fading technique [cf. Jamieson and Morosan, Percept. Psychophys. 40, 205–215 (1986)] was used to train two types of listeners to perceive new fricative contrasts. One group consisted of young adult francophones who had difficulty producing and perceiving the voiced/ voiceless “th” distinction in English. With this group, training using the fading technique improved identification accuracy both for the specific, synthetic targets used in training, and for multiple, natural tokens of these sounds, spoken by different individuals, both men and women. However, this learning was quite specific, as performance was poor when listeners were tested with the target sounds in new word positions, or when /d/ tokens were used in the distractor set. The second group of listeners were young children who consistently mispronounced at least one English fricative sound, and had been diagnosed as having a functional articulation disorder. A sizeable proportion of these children, all of whom had been selected on the basis of production difficulties, displayed atypical perceptual skills; for these children, training produced a measureable improvement in identification accuracy, and in some instances, perceptual training (without explicit production training) also improved the productions of the target sounds. [Work supported by NSERC AHFMR and HWC:NHRDP.]