華語量詞與名詞語意配對之多向度分析

本研究探討華語母語者的量詞與名詞語意配對，以及成人和兒童的語言能力差異。利用統計分析方法和MDS視覺化模型，深入分析華語母語者在量詞與名詞語意配對上的特徵和差異，並選擇25至45歲的成人組和10至12歲的兒童組作為受試者。研究者根據《國語日報量詞辭典》的分類，設計了包含15個量詞和225個名詞的問卷，共有3375個量詞-名詞配對，涵蓋華語的各種量詞使用情境。結果分析上，使用SPSS 26進行資料處理與分析，並用描述統計法分析兒童和成人在量詞使用上的多樣性，再用歐幾里德距離計算MDS（多向度量尺分析法）的相似性度量，呈現語意結構的視覺化模型。研究發現，高專一性名詞和少量詞搭配的現象較明顯，且成人和兒童在語意結構發展上有多方面的差異。特別是在形狀量詞搭配上，發現了清楚的群集結構，但抽象名詞的配對可接受性較複雜。MDS的視覺化模型也印證了語意結構的立體關係。

This study investigates the semantic pairing of classifiers and nouns in Mandarin Chinese native speakers, along with language proficiency differences between adults and children. Employing statistical analysis methods and the MDS visualization model, it delves into the characteristics and distinctions in the semantic pairing of classifiers and nouns among Mandarin native speakers. The study focuses on participants aged 25 to 45 in the adult group and 10 to 12 in the children's group.The researcher, following the classification in "國語日報量詞辭典" designed a questionnaire consisting of 15 classifiers and 225 nouns, resulting in 3375 classifier-noun pairs that encompass various contexts of Mandarin classifier usage. Data processing and analysis were conducted using SPSS 26. Descriptive statistics were applied to analyze the diversity in classifier usage between children and adults. Subsequently, Euclidean distance was calculated for Multidimensional Scaling (MDS) similarity measurements, presenting a visual model of semantic structure.The study revealed a distinct occurrence of high-specificity nouns paired with fewer classifiers, emphasizing multifaceted differences in semantic structure development between adults and children. Particularly in the pairing of shape classifiers, a clear clustering structure was identified, while the acceptability of pairings with abstract nouns appeared more intricate. The MDS visual model also confirmed the three-dimensional relationships within the semantic structure.