A Comparative Study On Substitution Errors By Junior And Senior English Majors-A Corpus-Based Approach

This thesis mainly studies the substitution errors committed by Chinese senior and junior English majors in their compositions.The thesis, to begin with, lays emphasis on the significance and aims of the current research, together with the outline of the structure for the whole thesis. Then the thesis reviews the related literature on the current research in three subparts. The first subpart touches upon some terminologies that are involved in the thesis and the writer standardizes the conceptions of the related terms. In particular, the discrepancies between mistake and error are clarified. Likewise, some writings are also devoted to the understandings of inter- and intra-lingual factors. In addition, the writer retrospectively dwells on the definition and the researches pertaining to interlanguage, the domain of which the nature of this thesis chiefly lies in. The second subpart defines substitution error and its applicable category. In the last subpart of the literature review, the writer introduces the background of the current research by elaborating on the three hypotheses, namely, Contrastive Analysis Hypothesis, Error Analysis Hypothesis and Transfer Analysis Hypothesis, with which the methodology of corpus research pattern is to be associated.After the groundings of the theories and related researches, the writer of the thesis embarks on the research design and finding presentation, based on which some hypotheses of the current study are unveiled. In the part of research design, the writer gives a brief account of learner corpus and Chinese Learner English Corpus respectively. Besides, two different methods of classifying the substitution errors, namely the method based on part of speech and on inter- and intra-lingual division, are settled and the selected corpora are accordingly retagged. When presenting the quantitative data analysis from the retagged corpus, the writer discovers that an overwhelming majority of substitution errors made by junior English majors fall into the category of interlingual interference, and the predominance is due to such parts of speech as noun, verb, adjective and preposition. When it comes to seniors, both inter- and intra-lingual factors co-function, similar to juniors in terms of typically committed substitution errors of some parts of speech. The proportion of different substitution errors in terms of part of speech, however, varies compared with that by juniors. Therefore, the writer puts forward four hypotheses. (1) There are some marked and unmarked differences of substitution errors committed by the learners of high and low proficiencies. Some common features can be described. (2) The learners of high proficiency do not necessarily make fewer substitution errors in terms of part of speech than those of low proficiency. There are some types of substitution errors that are more likely to be committed by seniors. But the reason is not due to the lack of progress. (3) Generally, for junior English majors, substitution errors are attributable to interlingual factors, whereas, for seniors, both inter- and intra-lingual factors work. The reason is that as senior learners make constant progress and get more exposure to the target language, they will resort to the rules in target language when they are uncertain about their linguistic output. (4) For junior English majors, one substitution error of a certain part of speech can lead to another substitution error. However, seniors do not show this tendency, as their substitution errors are hardly interrelated. The reason is that juniors, one-sidedly influenced by MT, commit quite a lot of substitution errors from first language transfer. Seniors, however, tend to make separate substitution errors partly influenced by mother tongue and partly by target language; hence there is little correlation between each type.Afterwards, the research findings are discussed. Revolving around the two methods of classification of substitution errors, the writer makes an in-depth analysis and characterizes the typicality of substitution errors of various types, aided by additional illustrations from the corpora. In the final analysis, the substitution errors of commonly committed types share three features: (1) presupposed synonyms; (2) literal translation; (3) confusing sets / pairs of words / phrases. Naturally, as collocation and substitution errors overlap to some extent, another feature is due to the inappropriate collocation.After the discussion, the writer reflects upon the research findings and yields some suggestions for English language teaching. As far as pedagogical theories are concerned, the findings can generate guidance for syllabus design and material development. As for the pedagogical application, vocabulary teaching and data-driven learning are fleshed out. In teaching vocabulary, instructors should take an all-round perspective of a particular word within the learners'cognitive ability, and foster the learners'habit of associating new words with learned ones. Regarding data-driven learning, the learners are supposed to be provided with mega-corpora for instant and constant reference. And the instructors might also establish mini-corpora based on learners'compositions so that not only the acquisition process of individual learner at different stages can be tracked, but also the learners'perception or awareness of substitution errors might be promoted with the help of reinforcement exercises designed through such teaching software as Hot Potato.