- 學位論文
中英文之聲音情感象徵探究
Affective Sound Symbolism in English and Mandarin Chinese
摘要
自索緒爾（de Saussure, 1916/2011）提出「能指」（signifier）和「所指」（signified）構面以來，語言學領域就以此語言的任意形式作為語法和語義關係的解釋（the arbitrariness of form-meaning association）。然近來的跨語言實驗和語料庫研究提出了另一解釋的觀點 － 從發音原理和目標含義兩者間所共享的物理性質進行歸因，即非任意聲音和涵義關聯之證據。本研究從跨語言的層次進一步研究語法和語義之間的關聯，以量化分析方法探究英文和中文詞彙中聲音(語法)的情感(語義)象徵。首先使用階層分群法（hierarchical clustering）對先前研究中的詞彙情感判斷之語料庫進行再分析，以研究兩語言本身與語言間的音素類別（phoneme class），於正負向性（valence）、生理激發（arousal）、自主程度（dominance）等情感向度中評比的分佈。再者進一步以線性判別分析（linear discriminant analysis）驗證本研究實驗中母語人士（中文）對於隨機假字的情感向度評比。階層分群法分析跨語言之結果表示，一些中英文舌葉阻音（coronal obstruents）在所有三個情感向度上皆有正相關之表現。而在假字實驗驗證之結果表示，聲音情感象徵之趨勢存在著性別和英語水平差異之影響。本研究之上述結論為語法和語義之間存在非任意性提供證據。
並列摘要
The field of linguistics has long appreciated the arbitrary form-meaning relationship since de Saussure’s (1916/2011) proposal of signifier and signified. Recent cross-linguistic experimental and corpus-based studies nevertheless suggest the opposite with evidence for non-arbitrary sound-meaning associations attributed to the physical quality shared by articulators and target meanings. The current study seeks to further examine the connections between forms and meanings from a cross-linguistic perspective, with a specific focus on quantitative approaches to affective sound symbolism in English and Mandarin Chinese words. Reanalysis on the affective judgment data from previous studies was conducted using hierarchical clustering to investigate cross-linguistic sound distributions over affective categories including valence, dominance, and arousal. For further verification, a linear discriminant analysis was then applied to results obtained from the experiment of affective judgments on random nonwords by native speakers of the language (Mandarin Chinese). Cross-linguistic results from hierarchical clustering analysis show that a set of coronal obstruents were present with positive correlations across all three affective dimensions. Verification results from nonword experiment show that affective sound symbolic trends presented effects from both gender an L2 English proficiency. Results above present evidence for biases between sounds and affective perception of linguistic forms.