Acquisition of Mandarin coronal sibilants by adult Korean speakers


Student thesis: Doctoral Thesis

View graph of relations


  • Jing SHAO


Awarding Institution
Award date15 Jul 2015


Learning to speak another language (L2) is not an easy task for adults. Their familiarity with speech sounds of the native language (L1) is more likely to make them misinterpret foreign sounds, thus causing difficulties in L2 speech perception and production. Mandarin contains a number of coronal sibilants, contrasting in three places and two manners. Korean has a set of three-way contrasted affricates and a pair of fricatives, which is unique in the world languages. The similarities and differences between Mandarin and Korean sibilants provide an interesting scenario for us to examine effects of L1 on L2 speech acquisition. This current research designed and carried out a series of experiments and also short-term training, to investigate adult Korean speakers’ performance and development in learning to perceive and produce Mandarin sibilant sounds. The first experiment analyzed acoustic characteristics of Mandarin sibilants and identified cues for optimal place and manner distinctions by native speakers of Mandarin. Then, a set of perceptual experiments were implemented to assess assimilation, discrimination, and identification of the Mandarin sounds by Korean speakers. Production experiments were carried out afterwards to examine phonetic characteristics of these Mandarin consonants produced by Korean speakers. Last, short-term perceptual training was provided to help improve Korean speakers’ perceptual ability in Mandarin. The acoustic analysis parameterized the acoustic space of Mandarin sibilants, revealing that certain spectral moments and F2 value at vowel onset provide most robust information in place distinction. Results of the perception experiments showed that Korean learners were least accurate at differentiating the contrast between alveolar and retroflex sibilants. Their accuracy increased with their Mandarin proficiency levels. However, highly experienced Korean learners of Mandarin still performed significantly worse than native speakers with the alveolar-retroflex contrast. The learners were also found to have difficulties in producing these non-native consonants. Their performance accuracy was greatly affected by vowels contexts, especially for beginning and intermediate learners and with alveolar and retroflex affricates. Acoustic data then confirmed that experience in Mandarin increased categorical saliency of acoustic parameters that resembles the native speakers’ acoustic space. Finally, results of the short-term auditory training proved that high variability in the input and customized goals can improve L2 perception significantly. The training effects can also be successfully generalized to new tokens and new tasks. In sum, this dissertation examined how adult Korean learners of Mandarin perceived and produced Mandarin coronal sibilants, and assessed how their L1 and knowledge in L2 contributed to performance accuracy and perceptual development concerning the L2 sounds. Findings from this study provide insights and important implications to in-depth understanding of L2 speech acquisition and also to Mandarin-as-a-foreign-language pedagogy.

    Research areas

  • Chinese language, Korean, Pronunciation by foreign speakers