A CLEC-based Study On Semantic Prosody

The concept of semantic prosody was first mentioned by Sinclair in 1991. He found out that 'set in' was habitually associated with unpleasant events, and enjoyed collocating with bad words like 'rot, decay, decadences'. Some linguists took the concept and proposed that a certain word has its own particular collocating category which often conveys a writer's or speaker's evaluating meaning, and any collocate beyond it may sound unnatural. The present study focuses on Chinese English learners' performances of semantic prosody in the following five nodes, i.e. develop, serve, happen, knowledge, society.The learners to be studied are at two different proficiency levels: intermediate and advanced. A contrastive Interlanguage analysis (CIA) approach is adopted in the study. Three comparisons are made. One is made between one subcorpus of Chinese Learners' English Corpus (CLEC), ST3 representing intermediate learners' proficiency level and British National Corpus (BNC); one is between the ST3 and the other subcorpus of CLEC, ST6, representing advanced learners' proficiency level; the last one is between the ST6 and BNC. Through these comparisons, the Chinese English learners' performances of semantic prosody are revealed, in terms of quantity, the degree of nativelikeness and appropriateness. The study shows that, as far as the conception of semantic prosody is concerned, in the first comparison there are some striking differences between ST3 and BNC in the five nodes; in the second one there co-exist similarities and differences between ST3 and ST6; and in the third comparison between ST6 and BNC, 2 nodes serve, happen agree with BNC as a whole, but the rest 3 nodes develop, knowledge society do not. The further investigation indicates that although some semantic groups of a node produced by Chinese English learners in ST3 and ST6 are also shared by BNC, they are different in percentage. Some semantic groups are overused in which the collocates of a given enjoy MI values and T-scores, while some so underused that their collocates are not significant in MI values and T-score. Furthermore, some semantic groups which are absent in BNC are coined in ST3, and some semantic groups never occur in ST3 and ST6, but they do in BNC. Although 2 nodes serve, happen show great progression in terms of competence in the performance of semantic prosody by comparing ST6 to ST3 and BNC, however, the other 3 nodes develop, knowledge society do not show significant signs of development.The examination of the deviant performances of semantic prosody in the five nodes in ST3 and ST6 shows that how the learners perceive the nodes is a vital factor for misuse of semantic prosody of this kind. Some connotations of some of the nodes may be wider in BNC than in ST3 and ST6, while some are narrower. Because semantic prosody is unnoticeable, the strategies of language transfer and overgeneralization are preferred by the Chinese learners in ST3 and ST6 in the production of the nodes. It is suggested that in instructional program English learners should be made aware of the concept of semantic prosody through using English native speakers' corpora and analyzing CLEC collocational behavior.