An experimental study on the production of Korean affricates by Chinese learners of Korean

Abstract

The purpose of this study is to analyze pronunciation errors and acoustic characteristics on Korean affricates produced by Chinese learners of Korean. To do so, listening judgement tests by Korean natives and acoustic analysis were conducted against produced tokens by Chinese learners. Listening judgement results showed that Chinese learners did not articulate lenis accurately in the ‘zha’ and ‘zhi’ groups and had difficulties in pronouncing lenis with being confused with fortis sounds in word-initial. They did show some confusions with lenis and fortis sounds reciprocally in the ‘azha’ and ‘azhi’ groups in word-medial position. On the aspect of pronunciation improvement, produced tokens of the ‘zha’ group showed an improvement to some degree with the accumulation of learning experience but those of the ‘zhi’ group rarely showed the effect of learning experience. In word-medial, produced tokens of the ‘azha’ group showed some improvement with the accumulation of learning experience. The ‘azhi’ group, however, was not affected by learning experience at all. According to the results of acoustic analysis Chinese learners of Korean distinguished types of affricates in the ‘zha’ group by CF-VOT. For the ‘zhi’ group, CF-VOT was the main factor to distinguish aspirated from lenis and fortis sounds while distinguishing lenis from aspirated using f0 (Chinese females). In word-medial, The Chinese learners’ production cues for the affricates in ‘azha’ and ‘azhi’ groups were almost the same as those of the Korean group. One more thing to mention is that Chinese males distinguished lenis from fortis and aspirated sounds by means of f0.