華語二語者的華語名動詞比例及其影響因素

語言習得的研究發現幼兒初期的詞彙發展存在名詞優於動詞的現象，即名詞學得更好、更快、更多。計算學習者使用的名詞和動詞的比值（名詞數量除以動詞數量，後稱「名動比」）會得到大於1的結果，表示名詞多於動詞，是謂「名詞優勢」。Gentner (1982) 認為這是一個普遍的現象，因為名詞指涉的是「東西」，有形狀、可以自背景獨立、具體、可想像的特性；動詞則指涉是東西和環境以及東西之間的「關係」，它更為複雜、隱晦，而且轉瞬即逝，因此學習動詞也比較困難。名詞和動詞在習得上的差異，被認為有跨語言的普遍性。然而，也有學者發現，在強調「關係」的語言（例如：華語和韓語），名動比與西方語言存在明顯差異。我們假設學習華語作為第二語言的學習者的詞彙發展符合第二語言習得理論的「自然順序假說」（The natural order hypothesis）和「輸入假說」（The input hypothesis），我們想知道他們的習得和輸入之間的關係以及詞彙發展的情況。我們預計探討如下因素：目標語社會自然語言環境和課堂教學環境的輸入，以及比較不同的詞彙發展階段和二語者的詞彙發展接近母語者的可能性。我們使用兩種語料來進行上述研究，第一種是既有的語料庫，包括中央研究院平衡語料庫、國教院語料庫索引典系統（含國教院華語中介語索引典系統）。第二種是自行收集、處理的語料，包括學習華語作為第二語言的主流華語教科書，以及包含二語者和華語者的碩博士學位論文摘要。我們透過詞形（type）和詞體（token）作為分析的單位以及不同名詞和動詞的界定比較語料之間的詞類分佈和名動比差異。研究發現如下：一、華語二語者的名動比低於自然語言環境及課堂教學環境，輸出並沒有反映輸入；二、從A1到C1的教科書，名動比有上升的趨勢，顯示語言程度和名動比有正相關；三、華語二語者和母語者在學位論文摘要的名動比並不存在差異，說明二語者的詞彙發展成就存在接近母語者的可能。研究顯示華語二語者的詞彙習得和使用和輸入存在差異，華語二語者的名動比遠低於華語母語者。雖然可以看到名動比隨著語言程度變高而提升以及不同名詞和動詞定義下有類似的變化跡象，但無法排除文章類型、文化以及語言教學的介入是干擾因素的可能性。以現有研究而言，比起輸入和語言特性，語言程度是較能解釋名動比變化的因素。最後，根據研究結果討論上述相關因素以及設計華語二語教學以期促進學習者的詞彙發展。

關鍵字

名動比；語言使用；語言程度；語料庫；華語二語者

並列摘要

Infant’s early language development was found that the number of learned words noun is much more than verb, called Noun Advantage, which argued consequence of a conceptual advantage in identifying objects in the word over relations among them by natural partitions/relational relativity (Gentner, 1982). A number of researchers have claimed it is not universal, and have focused on the relative frequency of nouns versus verbs, NVR (noun-verb ratio), in the speech of Children learner’s languages, like Korean and Chinese. According to Krashen’s (1982) natural order and input hypotheses, second language learners acquire language in a predictable order by receiving “comprehensible input.” Sinica Corpus and Interlanguage of beginners of CSL collected by the National Academy for Educational Research was used to sketch the vocabulary development of second language. We also collected CSL textbooks in Taiwan and Chinese thesis abstract by native speakers and advanced CSLs for comparing the vocabulary development of different stage. Type and token were both considered to count for the noun, verb percentage and NVRs under four noun and verb definitions. The NVRs comparisons among the corpus, textbook and thesis collected show that: 1. The NVR in Chinese of the CSL learners is lower than that of the input, both inside and outside the classroom. 2. There is a rising trend of NVR in textbook from language proficiency A1 to C1; the higher the proficiency of language, the higher the NVR. 3. The NVRs in Chinese in the thesis abstract of CSL learners and native speakers are no difference from each other, and CSL learners have potential to achieve native-like levels of language performance. There is no evidence shows that CSL’s Chinese vocabulary acquisition and usage are associated with input, and its NVR is far away from native speakers’. Though we can see the increasing of NVR by proficiency and almost same pattern across different noun and verb definitions, some factors, such as genre, culture and teacher interference could be confounded. Compare to language proficiency, input and linguistic requirement may contribute less to NVR. These associating factors are discussed, and a CSL teaching design based on the results are demonstrated as an alternative way to develop student’s vocabulary.