The Chinese language features a rich class of words called measure words that serve as units for counting objects and actions. In comparison with English and other Indo-European languages, Chinese makes much more extensive use of measure words. American students who study Chinese as a second language often find it hard to acquire the usage of Chinese measure words. To obtain a comprehensive and objective evaluation of students’ measure words acquisition, I designed an experiment where measure words as introduced in Integrated Chinese are collected. In the current study, measure words are divided into two categories by their semantic features: Concrete Measure Words and Abstract Measure Words. If a measure word directly relates to its object’s concrete exterior shape, and image thought plays an important role when people try to use this measure word, it is called a concrete measure word. Abstract measure words are those which have no obvious relation to an object’s exterior image, and whose usages mainly rely on people’s abstract thought. Students are divided into two grades based on how long they have studied Chinese: Grade 1 and Grade 2. Survey results show that students’ acquisition of concrete measure words is significantly better than their acquisition of abstract measure words. Furthermore, there is no obvious difference between measure words acquisition of the two grades; visual aids can facilitate concrete measure words acquisition to some extent. Conclusions of survey results reveal some practical principles of measure words teaching. First, concrete measure words and abstract measure words should be treated differently in classroom teaching. Second, different teaching strategies should be adopted to teach students from different grades. Third, analyzing semantic features and providing visual aids are useful methods when teaching concrete measure words. This thesis includes five chapters. Chapter One summarizes related work in previous studies and points out the importance of future research on Chinese measure words acquisition. Chapter Two focuses on the design of the survey where experimental settings, including objects, participants, survey design, and study methods, are introduced. In Chapter Three, I discuss the experimental results in more detail and summarize typically misused measure words. Chapter Four focuses on the teaching material study where I analyze the arrangements of contents related to measure words, and discuss the merits and shortcomings of the teaching materials currently used. In the last chapter, I summarize some suggestions on teaching strategies inspired by this study.

Chinese measure words are bound morphemes that can classify nouns/verbs and form a determinative-measure-word compound with numeral, demonstrative, or specifying determinatives. Dr. Yuanren Chao distributes Chinese measure words into nine categories: individual measure words (Mc), individual measure words associated with verb-object construction (Mc’), group measure words (Mg), partitive measure words (Mp), container measure words (Mo), temporary measure words (Mt), standard measure words (Mm), quasi-measure words (Mq), and measure words of verbs of actions (Mv). The complicated measure word system is a big challenge for Chinese as a second language learners. This study investigates learners’ lexical, semantic, and syntactic characteristics of acquiring Chinese measure words. Learners’ and native speakers’ spoken and written language data are extracted from three authoritative Chinese learners corpora and three authoritative native speaker corpora. Corpus linguistic methods are used in this study. To explore the three linguistic features, this study conducts quantitative and qualitative analyses. The quantitative analysis explores learners’ lexical diversity, sophistication, and accuracy. The findings reveal that learners’ lexical diversity of using measure words is lower than native speakers and English-speaking learners’ is lower than non-English-speaking learners’. Learners’ lexical sophistication of using measure words did not grow with proficiency levels and therefore cannot predict or indicate learners’ language proficiency levels. Likewise, learners’ lexical accuracy of using measure words vary from category to category. Advanced learners have lower accuracy than beginner and intermediate learners, and English-speaking learners are lower than non-English-speaking learners. The qualitative analysis examines learners’ semantic and syntactic behaviors with two case studies. The first case study analyzes nine specific measure words that belong to multiple categories that can reflect learners’ semantic expressions of word choices. The second case study dissects the syntactic structures of ge ‘an individual,’ zhong ‘kinds, species,’ and xie ‘some’ (they are the three most frequent measure words used by learners) and explores learners’ co-selection of determinatives, measure words, and nouns. The results of the first case study show that learners’ semantic expressions are not controlled by the frequency of semantic categories but are likely affected by a topic/genre. The second case study indicates that learners tend to overuse or underuse special grammar patterns without understanding correct usage. The whole study aims to bring attention to Chinese measure words in second language teaching and corpus linguistic methods as teaching and research tools. It will help second language teachers and linguists to gain a deeper understanding of how learners would grasp a classification system and their linguistic characteristics of acquisition. Keywords: corpus linguistics, Chinese measure words, second language acquisition

台灣閩南語（以下簡稱台語）的研究近年日益受到重視，但討論台語分類詞的文獻卻相當稀少。分類詞定義不清，是歷年文獻對台語分類詞看法不一或略而不談的主因。同樣地，由於台語分類詞的定位不明，也使得分類詞在台語教學中未能受到應有的重視。 　　因此，本論文將以Her & Hsieh（2010）和Her（2012）所提出的分類詞與量詞的語法測試、集合論差異、和語意數值作為基礎，重新界定分類詞，並依此定義，重新審視歷年文獻所提出的台語分類詞或量詞語料。這些文獻和語料的來源涵蓋了教育部或教育局、國科會研究計畫、語言學者、資深台灣閩南語教師、及一手田野調查的資料。本論文將文獻依其對分類詞的定位分成三大類：量詞（未提及分類詞時）、分類詞視為量詞次類（如：個體量詞等）、分類詞。 　　本論文研究目的有四：首先，藉由明確的分類詞定義，重新檢視文獻中的分類詞或量詞語料，盡可能呈現出台灣閩南語的分類詞清單。其次，以這些此分類詞清單基礎，歸納出台灣閩南語分類詞的語意歸類系統。第三點，對比台灣閩南語和台灣華語的分類詞語意歸類系統。最後，將觀察到的兩方言差異提供給台語教學者作為分類詞教學上的參考依據。