成杰　李佳

【Abstract】This study tries to reveal the semantic prosody of the word totally in Chinese learners' written English in comparison with native English speakers by examining the semantic prosodic features of totally in Spoken and Written English Corpus of Chinese Learners and Corpus of Contemporary American English. The study shows that Chinese Learners show significant differences with native English speakers for the word totally. The possible reasons for the differences may lie in negative language transfer.

【Keywords】semantic prosody；corpus；totally

1. Introduction

Semantic prosody focuses on the mutual influence of words in one sentence through the study of language collocation. Native English speakers， through long time using of mother language， can hardly use words with wrong semantic prosody. However， this may not be true when it comes to Chinese EFL learners who are often not aware of the collocation of English words. This paper tries to reveal the semantic prosody of the word “totally” in Chinese learners English through the comparison of native English speakers. The Spoken and Written English Corpus of Chinese learners and Corpus of Contemporary American English are adopted in this paper.

2. Semantic Prosody

Semantic prosody refers to the phenomenon that collocates before or after a word tend to form a specific semantic category with the same emotional significance. The term “semantic prosody” is coined and defined by Bill Louw. According to him， semantic prosody is “a consistent aura of meaning with which a form is imbued by its collocates.” （Louw， 1993） Sinclair reveals that the word happen habitually collocates with unpleasant events， such as accidents happen， something terrible has happened， etc. So he describes the phenomenon that “many use of words and phrases show a tendency to occur in a certain semantic environment.” （Sinclair， 2003）

Stubbs （1995） classifies semantic prosody into three categories in his works， namely negative prosody， positive prosody and neutral prosody. This way of classification is widely accepted in the field of linguistics. According to Stubbs， within a given context， a favorable or pleasant meaning is considered positive， while an unfavorable or unpleasant meaning is considered negative. If collocates convey neutral or mixed information， or the context does not offer enough proves， it is considered neutral.

Semantic prosody is closely linked with corpus data. It is contingent upon concordance and lexical profiles， apparently depending upon them for its recognition. Data-based approach is widely applied in the studies of semantic prosody. The researcher observes， summarizes and describes the semantic prosody according to the collocates in the corpora.

3. The semantic prosody of “totally” in Chinese learners English

3.1 Data Processing Procedures

In this paper， the Spoken and Written English Corpus of Chinese Learners （SWECCL） and Corpus of Contemporary American English （COCA） are adopted.

Firstly， the author queries the word “totally” in two corpora and discovers significant collocates. The span is set 10 words （-5/+5） in order not to missing any possible collocates. Then the author examines the collocates respectively and judges the semantic prosodic feather of the word in the two corpora. Collocates with favorable or pleasant meaning are judged as positive while collocates with unfavorable and unpleasant meaning are judged as negative. When collocates convey neutral or mixed information， or the context does not provide enough evidence， they are regarded as neutral. Finally， the author runs the Chi-Square test with SPSS and compares the proportions of the same kind of collocates between the two corpora. The author judges whether the differences in semantic prosodies between the two groups of subjects are significant or not according to the results.

3.2 Results and Discussion

The significant collocates of “totally” in Spoken and Written English Corpus of Chinese learners are presented below：

The significant collocates of “totally” in Corpus of Contemporary American English are presented below：

After judging the semantic prosody of these collocates， the numbers and proportions of each category of semantic prosody for collocates of totally in the two corpora are shown in the following tables：

collocates Positive collocates Negative collocates Neutral collocates

The two tables above show that 61.54% of the collocates in SWEECL are positive， and 23.08% are negative. As for COCA， there is 26.32% positive and 52.63% negative.

Since the detailed statistics are available， Chi-square test on the proportions of the same kind of collocates between the two corpora is completed and the results are listed below：

From the tables above， it can be seen that more than 60% of the Chinese English learners use totally in a positive way， as the following examples show：

She totally earned my trust.

He regarded my work as totally good.

I believe this statement is totally right.

However， it is worth noting that there is 50% of the native speakers use totally in a negative way， such as

… that is totally unacceptable and intolerable， and we must stop it.

Gaza's position as a “hostile entity” is a highly politicized and totally false statement.

Both cars were totally destroyed.

But I will find it very hard if I am totally blind....

To conclude， the semantic prosody of totally in SWEECL and COCA is different. The Chi-square test proves this conclusion. The P-value of Chi-square test 0.00 is lower than 0.05， so it is safe to say the Chinese English learners do not use the word totally with appropriate semantic prosody.

4. Conclusion

From the above analysis， it can be conclude for the word totally， Chinese English learners show significant differences with native English speakers by using them with positive semantic prosody， while native speakers tend to use this word with negative semantic prosody. The statistics show that Chinese English learners are not aware of the knowledge of semantic prosody.

The possible reasons for this difference may lie in negative language transfer. According to Biskup （1992）， the semantic field of a word in L1 will interfere with language users L2， causing their misunderstanding of the semantic filed in L2. The Chinese translation equivalents for totally include 整個(gè)地，完全地，which bear positive connotations. Therefore， Chinese learners tend to collocate it with positive collocates. Negative transfer of their L1 interferes with Chinese English learners recognition of the wordsmeanings.

Semantic prosody is often hidden from human intuition and can only be revealed with the help of computer and corpora. Therefore， special attention to semantic prosodic features should be paid in teaching practice， so that students could raise the consciousness of semantic prosody and thus avoid the misuse of words.

References：

[1]Louw B. Irony in the Text or Insincerity in the Writer？ --The Diagnostic Potential of Semantic Prosodies （p. 157， pp. 159-167） [J].Text and Technology， 1993.

[2]Sinclair J. Corpus， Concordance， Collocation[M].Oxford： Oxford University Press， 1991.

[3]Stubbs M. Text and Corpus Analysis[M].Oxford： Blackwell Publishers Ltd， 1996.

[4]Corpus of Contemporary American English. url： http：//www.americancorpus.org.

[5]文秋芳，梁茂成，晏小琴.中國(guó)學(xué)生英語(yǔ)口語(yǔ)筆語(yǔ)語(yǔ)料庫(kù)2.0版[M].北京：外語(yǔ)教學(xué)與研究出版社，2008.

[6]衛(wèi)乃興.語(yǔ)義韻研究的一般方法[J].外語(yǔ)教學(xué)與研究，2002.