從作文中看漢語語彙習得之研究──根據LARP at SCU之資料調查

Japanese Vocabulary Acquisition: A Study of Parts of Speech and Word Frequency in Writing

據觀察，台灣人日本語學習者從初級階段開始受母語轉移的影響，在漢語彙的習得方面以較快速度進展。筆者（2006）針對語彙的習得，曾利用LARP at SCU資料的一部，進行調查，結果發現，名詞類的使用量最大，而且其中漢語彙的正用現象十分明顯。但是，由於各學習階段有「自由變異」現象，因此有必要追蹤式調查發展情況。本稿以LARP at SCU語料持續調査。依日文漢語語彙之品詞與學習者母語的對應關係來看習得過程的特徵。品詞分名詞類、動詞類、ナ形容詞類、副詞類、接辞‧複合詞類。所謂日中語彙對應關係是採用日本文化庁（1978）的分類加以修正，分成以下五類：（1）同形同義語（簡稱S）（2）同形異義語（簡稱D）（3）同形但中日語間意義不盡相同（簡稱O）（4）唯日語有的漢語（簡稱J）（5）唯中文有的語彙（簡稱C）。以此五類之中日漢語語彙對應基準。追蹤調查日文系學生一名，自2004年3月至2007年1月共6學期，計30篇作文為探討對象。分三個學習階段探討，檢視各學習階段的作文所使用的漢語語彙，依五類漢語彙的使用狀況，於對應關係中，看出日語（L2）漢語彙的習得過程受中文母語（L1）影響的程度，以及階段性的變化。結果發現：（1）三階段皆呈S類＞J類＞O類＞D類狀況；C類極少。（2）且長期間各類別間的數量關係維持一致。可解釋為，L2學習受L1語彙之意義形態影響深厚。另外「和製漢語」類之習得尚優於同形異意語，可視為以異質性外文來接受之習得，快於部分重疊之語彙。

Vocabulary acquisition is one of the markers for judging the progress of foreign language learning. It affects the four skills of language acquisition. This paper takes interlanguage perspective and studies the vocabulary acquisition of Japanese majors from their writing assignments. The objective is to understand the characteristics of vocabulary acquisition of students from beginning to mid-level. Data collection is taken from four Japanese majors. We focus on the frequency and number of certain vocabulary and the learning pattern of parts of speech. The samples are from their monthly writing assignment, average 600 words, taken ten times in total during their first year.Our study results in the followings: 1) there is no linear progress of numbers of vocabulary acquisition but our data shows a more steady progress in the second half of the year; 2) individual variations are prominent but there is a correlation between good grades and great numbers of vocabulary; 3) the growth of MLU increases slowly but accelerates during the second half of the year.In terms of their learning pattern of parts of speech, we have found the followings: 1) the numbers of nominal and nouns account for the most; 2) the use of verb is limited to concrete description; 3) adjective usage is limited to categorical references; 4) more idiosyncrasy in adverb usage, no onomatopoeia is found; 5) serial verbs only show up for repetition.