詹曉蕙Chan, Shiao-Hui沈昀Shen, Yun2024-12-172024-07-232024https://etds.lib.ntnu.edu.tw/thesis/detail/150f72c70a20238f101538d891ff8c0f/http://rportal.lib.ntnu.edu.tw/handle/20.500.12235/123100本研究從：語境為基礎之意面(context-based meaning facet) (Ahrens, Chang, Chen,& Huang, 1998)、屬性結構(qualia structure)及詞意誘迫(meaning coercion) (Huang & Ahrens, 2003; Pustejovsky, 1998)、及概念的變動性(conceptual flexibility) (Yee & Thompson-Schill, 2016)之觀點，以事件相關電位方法(event-related potential, ERP)探討中文個別(如：支、顆）及事件(如：場、班）分類詞片語之語意分別。從意面的角度解釋，個別及事件分類詞為名詞概念提供了不同的語境訊息，用以強調名詞概念中某一特定意面；而從屬性結構及詞意誘迫的角度來說，個別及事件分類詞所分別強調之屬性角色(qualia role)：語意類別角色(FORMAL role)及功能角色(TELIC role)，會誘迫(coerce)後方的名詞意義，並同時激發與其意義相對應的語意特徵：感官特徵(perceptual feature)及功能特徵(functional feature)。若從神經認知歷程的角度理解，概念的形成及處理具變動性，而此變動性有賴於概念在何種語境中呈現。不同名詞類別、語意特徵、及語意關聯性之操弄會誘發不同之事件相關電位成分(ERP components)，代表其背後不同的認知處理機制。根據以上理解，本研究探討個別及事件分類詞是否會誘迫/促發後方的名詞概念表現出相應的個別及功能意義解讀。實驗材料由觸發及目標(prime-target)詞彙組成。觸發詞彙為數字－分類詞及名詞(以兩個連續畫面呈現），而目標詞彙為與觸發詞彙相關或不相關之語意特徵。受試者需判斷目標詞彙是否合適於描述前面的觸發詞彙。目標詞彙之ERP結果呈現過去文獻中時常出現之N400語意關聯性效應 (relatedness effect)；而後，在前腦區出現一致性效應(congruity effect，指分類詞及語意特徵的屬性角色一致性)所誘發之晚期正向波 (frontal post-N400 positivity, f-PNP)。此電生理證據說明分類詞確實會誘迫後方的名詞意義，亦支持了以語境為基礎之意面，以及屬性結構及詞意誘迫的理論。更具體來說，當一個與分類詞屬性角色一致之語意特徵接連在名詞後出現時，受試者會對此進行更深層之後續分析。此結果同時也支持了神經認知領域中，對於概念之變動性的推論。名詞概念之處理可以分階段進行：在350-500毫秒時，名詞概念及語意特徵之間的關聯性會先被處理，而在500毫秒後，分類詞及語意特徵之間的屬性角色一致性才會被進一步分析。The current study explored the semantic distinction between Mandarin individual (e.g., 支zhi, 顆ke ) and event (e.g., 場chang, 班ban) classifier phrases from the perspectives of context-based meaning facets (Ahrens, Chang, Chen,& Huang, 1998), qualia structure and coercion (Huang & Ahrens, 2003; Pustejovsky, 1998), and conceptual flexibility (Yee & Thompson-Schill, 2016) with the event-related potential (ERP) technique. From the semantic point of view, individual vs. event classifiers provide distinctive contextual information for choosing a particular meaning facet of the nominal concept, or alternatively, the different qualia roles (FORMAL vs. TELIC) expressed by two types of classifiers coerce the meaning of the noun and also highlight the corresponding features (perceptual vs. functional) of the nominal concepts. From the perspective of neurocognition, concepts are processed flexibly in terms of the context they are presented in, and the manipulations of nominal categories, semantic features, and semantic relatedness would induce different ERP components, suggesting the processing differences behind them. Based on the previous understandings, this study aimed to explore the differences between individual vs. event classifiers by investigating whether these two types of classifiers would coerce/activate the corresponding individual vs. event reading of the following nominal concept. The experimental materials consisted of prime-target pairs, with the prime being the numeral-classifier and the noun (appearing in two consecutive frames) and the target being a semantic feature related or unrelated to the prime. Participants had to judge whether the target appropriately described the prime. The ERP response to the target showed the commonly reported N400 relatedness effect followed by a frontal post-N400 positivity (f-PNP) congruity effect (i.e., the congruity of the qualia roles between classifiers and features). The electrophysiological evidence demonstrated that classifiers did coerce the following nouns, strengthening the theories of context-based meaning facets and Mandarin classifier coercion. Furthermore, when the upcoming feature matched the highlighted meaning facet/qualia role, participants would perform a deeper, post-lexical analysis. The findings also supported that nominal concepts were processed flexibly in different stages: the semantic relatedness between classifier phrases and features during 350-500ms and the congruity of qualia roles between classifiers and features after 500ms.中文分類詞意面屬性結構詞意誘迫概念的變動性事件相關電位神經語言學Mandarin classifiermeaning facetqualia structuremeaning coercionconceptual flexibilityERPneurolinguistics學術論文