The effect of short-term training on production and perception of s∕ʃ by Korean speakers.

Abstract

This study examines the efficacy of short‐term training on learning the English s/∫distinction by native speakers of Korean. The production and perception of Korean L2 learners was tested before and after a short computer‐controlled training session. Of interest in this study is to determine which acoustic characteristics of these fricatives (if any) improve after a focused training session, whether English listeners perceive them more “English‐like,” and whether the L2 learners can more accurately identify these fricatives after training. Korean speakers were initially recorded producing common English words containing these two fricatives in initial, medial, and final positions. Each speaker then underwent focused training on the s/∫distinction in a small set of nonsense words. Training ended when speakers met an established performance criterion. In a post‐test, speakers were again recorded producing the same words. Acoustic measurements (including duration, relative amplitude, spectral mean, variance, skewness, and kurtosis) were obtained for each fricative from both sets of recordings and from productions by native English speakers. Perception data (including both identification and discrimination) were also obtained from the English‐speaking controls and from L2 speakers both before and after the training. Results will be discussed in light of universal principles of language acquisition.