2011-08-012024-05-18https://scholars.lib.ntu.edu.tw/handle/123456789/710239摘要：嬰兒語音知覺能力在六到十二個月大之間快速成長，並呈現母語經驗的影響：對母語語音的敏感度日漸增長，而對於外語語音敏感度日漸下降。有哪些學習因素會影響語音知覺發展？這還是一個具有爭議的議題。特定而言，是否「領域一般」機制，如「統計式學習」及「分散式學習」，以及「語言領域限定」的因素，如「語音聲學特徵顯著性」，都會影響語音知覺發展？本研究預計以五個實驗，檢視這些因素如何單獨或與其他因素合作，影響嬰兒形成精細的語音知覺表徵。實驗一檢驗「條件式統計學習」（語音組合規律限制），對七到十一個月大嬰兒輔音知覺區辨表現的影響。實驗二再次檢驗實驗一的目的，並利用語音聲學特徵顯著性較低的語音對比，檢視「統計式學習」及「聲學顯著性」對輔音知覺發展的影響。實驗三檢視語音組合規律的條件式統計學習對聲調知覺發展的影響。實驗四檢視統計分配式學習及分散學習，對於嬰兒學習外語語音類別及語音組合規律的影響。九個月大嬰兒將在實驗室中和英語說話者互動，而按外語互動頻率分成「分散式學習」和「集中式學習」兩類，以探索「分散式學習」效果。實驗四也操弄嬰兒聆聽特定語音組合規律的出現頻率，以檢視「統計分配式學習」對語音知覺發展的影響。實驗五檢視是否簡短的語言經驗，就足以引發嬰兒學習從未聽過的語音組合規律，並檢視這樣的學習對輔音區辨的影響。綜合五個實驗的結果，將全面檢視影響嬰兒建構精細語音表徵的「領域特定」及「領域限定」學習因素，並進而推論影響嬰兒掌握母語特性的「嬰兒學習特性」及「語言環境」因素。<br> Abstract: Infant speech perception abilities show a divergent trend between 6 and 12 months ofage: perceptual improvement for native speech sounds and a perceptual decline for non-nativespeech sounds. What learning mechanisms would account for such rapid development ofspeech perception? Listening experiences with the native language shape infants’ perceptualworld for speech sounds. Are there other factors facilitating infants to construct phoneticrepresentations from the speech inputs? Specifically, could both domain-learning mechanisms,e.g., statistical learning and distributed learning, and domain-specific mechanism, e.g.,acoustic salience, affect speech perception development in infancy? Five experiments will bedesigned to assess effects of these learning mechanisms on infant speech perceptiondevelopment.Experiment 1 will assess the effect of conditional statistical learning of speechsound sequences (i.e., phonotactics) on consonant perception development between 7 and 11months of age. Experiment 2 will extend the findings of Experiment 1 by testing infants usingan acoustically less salient phonetic contrast, and also examine the effect of acoustic salienceon consonant perception development. Experiment 3 will examine if the conditional statisticallearning of speech sounds affects lexical-tone perception development between 7 and 11months of age. Experiment 4 will expose 9-month-old Mandarin-learning infants to English ina short-term foreign language exposure condition. Infants will be exposed to English in twolearning styles, distributed-learning and massed-learning, and the distributional probabilitiesof speech sound sequences will be manipulated. Therefore, Experiment 4 will examine theeffects of both distributional statistical learning and distributed learning on infants developingnew phonetic categories and phonotactic patterns. Experiment 5 will assess if infants learnnew speech sound sequences of the native language from a brief language exposure. TheExperiment 5 also examines the role of learning new phonotactics on consonant perception.Results of this proposed study would provide critical evidences to asses if bothdomain-general and domain-specific factors affect infants to form detailed phoneticrepresentations from speech inputs during infancy.嬰兒研究語音知覺發展塞音區辨統計式學習Infant studyspeech perception developmentstatistical learningphonetic discriminationstop consonant perception.