A Phonetic And Phonological Analysis Of The Production And Perception Of English Post-alveolar Fricatives By Mandarin And Mandarin/Shaoxing Wu Speakers

This dissertation presents an experimental investigation of how Mandarin and Mandarin-Shaoxing Wu speakers produce and perceive the English post-alveolar fricatives,to establish the relation between production and perception in L2 English.The Shaoxing Wu speakers in the current research are Mandarin-Wu bilinguals who are proficient both at Mandarin and Shaoxing Wu,so a comparison was made between the monolinguals and bilinguals,to check differences between L2 perception and production.In the production experiment,the researcher examined the spectral moments of the English post-alveolar fricatives produced by Mandarin and Mandarin-Wu speakers,and compared them with that produced by native English speakers and also the post-alveolar fricatives in the learners’ L1 s.ANOVA and LDA analyses both demonstrated that Mandarin speakers replaced the English voiceless post-alveolar fricative /(?)/ with Mandarin voiceless retroflex /(?)/,and the voiced post-alveolar /(?)/ with the voiced one /(?)/(?)/.On the other hand,Mandarin-Wu speakers used the voiceless alveolo-palatal /(?)/ to substitute both the English voiceless /(?)/ and voiced /(?)/.The production patterns do not completely conform to the sound learning model(SLM)proposed by previous researchers.Next,native English speakers were invited to rate the accentedness of the L2 production by Mandarin and Mandarin-Wu speakers,and they rated Mandarin-accented /(?)/ as most accented while the Mandarin/Wu-accented/(?)/ was rated least accented.The pattern of accentedness can be linked to the degree of violation on the phonological features in L2.Therefore,I infer that the production of the L2 sound is a complex process,which relates not only to L1-L2 acoustic distance but also to the L1-L2 phonological distance,and the perception of the accentedness should be rated both from an L2 learner and a native speaker point of view.Since the voiced alveolo-palatal fricative /(?)/ is very unstable in its acoustic space and thus cannot function as an ideal substitution in the production experiment,in the perception task,only the voiceless English post-alveolar fricative /(?)/ produced by Mandarin and Mandarin/Wu speakers were examined.In the discrimination task,Mandarin and Mandarin-Wu speakers were required to discriminate between English/(?)/ and Mandarin /(?)/ or Mandarin/Wu /(?)/ in a CV structure.The results showed that both groups of participants perceived /(?)/ to be closer to /(?)/ when followed by the vowel of /a/ but closer to /(?)/ when followed by /i/ and /u/.This perception pattern,similar between Mandarin and Mandarin-Wu speakers,does not align with their respective L2 production patterns.However,in the identification task Mandarin speakers perceived the target English segment to be more like Mandarin retroflex /(?)/across all vowel contexts(/a/,/i/,/u/),while the Mandarin-Wu speakers perceived it to be more like Mandarin/Wu /(?)/ in most vowel contexts(with the exception of the /a/context,which is thought to be influenced by educational factors).Therefore,the perception in this task is correlated with the results of production when the target segment is embedded in the word form.Finally,the researcher performed a phonological analysis on Mandarin and Mandarin/Wu speakers’ production patterns of the English post-alveolar fricatives,and proposed relevant constraints on L2 sound production in Optimal Theory.The current study also sheds some light,in a pedagogical sense,on how to improve the learning of the L2 segments.