Computer-aided Error Analysis On English Word Meaning

The mastery of vocabulary meaning is difficult in foreign language learning (FLL). In Leech's terms, the types of meaning can be roughly divided into two main kinds, i.e. conceptual meaning and associative meaning. Comparatively speaking, it is more difficult to learn associative meaning than to learn conceptual meaning. Thus when confronted with culturally-loaded words, which have rich associative meanings, foreign language learners frequently make errors in comprehension and production.This thesis attempts to explore the regularities of vocabulary meaning between two languages and of semantic development in FLL. Three cases show the relationship between meanings and expressions. The first case is that two languages have the same meaning but different expressions. The second case is that the meanings overlap partly and the expressions are different. The third case is that neither meaning nor expression is the same. Some color terms and animal terms are chosen to form a questionnaire, in which the three cases are included. The statistical analyses show that the second case is the main source of systematic semantic errors at various stages of FLL. The theoretical analyses explore into the causes, and in the light of these some ways and means are proposed to solve the problem. In different languages, the conceptual meanings have semantic universals while the associative meanings have similarities and differences. Semantic comparison between the languages is therefore a necessary process.Some implications for foreign language teaching can be found from the analyses. First, the attitudes must be active in dealing with the culture factors in semantic system of the target language. Secondly, the input of cultural knowledge must followed the / +1 principle in the FLL. Thirdly, one of the better ways to master the associative errors is to learn them in contexts. Last, language teachers should hold tolerant attitudes towards the semantic errors and have different techniques for the two different kinds of meanings in error correction.