A Comparative Study Of Vocabulary Knowledge Of Modern Chinese And Chinese Interlanguage Based On The Lexical Co-occurrence Within The Whole Sentence

As an important part of vocabulary knowledge, lexical co-occurrence has been attracting much attention in the previous research, which mostly focused on a span of 5 words immediately both preceding and following a target word. In this paper, we set the co-occurrence span within the whole sentence and develop an ‘Automatic Extraction Program of Chinese Lexical Co-occurrence within the Whole Sentence' based on modern Chinese corpus and Chinese interlanguage corpus, due to the actual characteristics of Chinese and research need. The Program can provide convicing material for the description of vocabulary knowledge by extracting a specific word's co-occurring words with their co-occurrence distances and entry senses in the modern Chinese corpus, or a specific word's co-occurring words with their co-occurrence distances, parts of speech, together with learners' levels and mother tongues in the Chinese interlanguage corpus. Users can get frequencies and do sorting as needed.The information from the Program can be used to record vocabulary knowledge, to serve as part of lexical representations in the computational simulation and as data in the Contrastive Interlanguage Analysis(CIA). In this paper, the co-occurrence information of ‘kan'(to see, look) is contrasted between the modern Chinese corpus and Chinese interlanguage corpus, and between the different levels in the Chinese interlanguage corpus. Then the co-occurring words of ‘kan' in both the modern Chinese corpus and the Chinese interlanguage corpus are categorized semantically according to Synonym Lexicon. Each category is calculated to what extend overused or underused in the Chinese interlanguage corpus compared with that in the modern Chinese corpus. The same work is done among the different levels in the Chinese interlanguage corpus. In this way, research on Second Language Acquisition of Chinese as an L2 can go deep into the usage differences between the Chinese interlanguage and modern Chinese in the view of lexical co-occrurrence, as opposed to confined to the Error Analysis.We can get the following conclusions: in the semantic distribution of the co-occurring words of ‘kan',(1) the most seriously overused large category is ‘Empty Words' and the most seriously underused large category is ‘Activities' in the Chinese interlanguage corpus compared with the modern Chinese corpus.(2) The top 3 overused medium categories are ‘Abstract Things/ Culture and Education', ‘Abstract Things/ Social Politics and Law' and ‘Things/ Geomorphology', while the top 3 underused medium categories are ‘People/ Proper Names', ‘Things/ Body' and ‘Activities/ Administration'.(3) Among the four learners' levels, the semantic distribution of large categories fluctuates. The differences between the overall distribution of large categories in the Chinese interlanguage corpus and that in the the modern Chinese corpus is most significant when learners have finished 1-year-and-half to 2-year's learning. After that, the overall distribution converges to the modern Chinese.