本研究藉由以韓語為母語的華語學習者發華語擦音的發音與偏誤探討學習者如何習得華語無聲擦音,華語有唇齒音[f]、齒齦音[s]、捲舌音[ʂ]、硬顎音[ɕ]、軟顎音[x]等五個無聲擦音,韓語則有送氣齒齦擦音ㅅ[sh]、送氣齒齦擦音ㅆ[s’]與聲門擦音ㅎ[h]三個。研究者邀請以韓語為母語的華語學習者(男性6位,女性18位)及華語母語者(男性3位,女性5位)共32位參與錄音實驗,錄音的內容是將華語擦音為首的詞彙以隨機的方式呈現請受試者發音,透過實驗取得的語料先由受過語音訓練的母語者進行母語者轉寫分析,再從聲學分析得到的擦音能量M1與onsetF2值,找出學習者學習擦音的難點與偏誤,並探討SLM理論、CAH提出預測是否與事實相符。研究結果顯示(1)轉寫結果顯示,發音正確率由高至低依序為[x]>[ɕ]>[f]>[ʂ]>[s]。(2)聲學分析結果,噝噝音[s]與[ʂ]的M1值重疊、[ʂ]與[ɕ]的onsetF2值重疊,非噝噝音[f]的M1值分布範圍廣,顯示學習者對[s]、[ʂ]、[ɕ]、[f]發音概念模糊。(3)學習者難點預測,實驗結果偏向SLM所提出的預測,且陌生語音[f]、[ʂ]、[ɕ]的難度比相似語音[s]、[x]低,顯示和母語相似的語音比較容易產生偏誤。希望未來能有更多相關實驗以幫助我們更了解韓籍學習者的語音習得過程。
This study investigates how Korean learners of Chinese acquire Chinese fricatives. Chinese has five voiceless fricatives: labio-dental [f], dental [s], retroflex [ʂ], palatal [ɕ] and velar [x]. Korean has three voiceless fricatives: alveolar lenis (lax) [sh], alveolar fortis (tense) [s’] and glottal [h]. The researchers invited 24 Korean learners of Chinese (M=6, F=18) and 8 native speakers of Chinese (M=3,F=5) to participate in an audio prompted repetition task. In this task, all target fricatives were placed in the word-initial position with comparable vowels. All participants were tested in two blocks with target words in different randomization orders. The data was analyzed by phonetic transcription and acoustic analyses. The phonetic transcription was done by one phonotian and the particpants’ pronounciations were first transcribed as correct and incorrect. Then, the phonotian further transcribed the incorrect productions. Another phonotian transcribed twenty percent of the total data to ensure inter-rater reliability. For the acoustic analyses, two acoustic measures, spectral centroid and onset of F2, were used to analyze the data. The results showed that based on the transcription, [x] had the highest correct rate followed by [ɕ], [f], [ʂ] and [s]. On the other hand, the acoustic analyses showed that, these Korean learners were not sure about the productions of [s] , [ʂ] , [ɕ] and [f]. It seems that SLM can better predict these learners’ Chinese fricative produtions. More research is needed to help us understand how Korean learners of Chinese acquire Chiense voiceless fricatives.