An Empirical Study On The Application Of Cohesion Theory In Improving Writing Teaching In Advanced English Majors' Context

Cohesion, a crucial notion in text linguistics, is an important construct in English texts. However, cohesion has long been neglected in English writing teaching in China. A questionnaire investigation revealed that lack of cohesion is the feature appearing to cover a large number of perceived weaknesses in advanced English majors' compositions.Cohesion has drawn much of linguists' attention. Linguists and researchers agree that cohesion is an indispensable condition for the creation of a coherent English text. There are some characteristic features about the relevant studies: (a) most of the researches on cohesion have been directed on its effects on the reader instead of on the writer; (b) the linguists are mainly concerned with establishing theoretical models for cohesion, and few tried to teach cohesion in classroom for its practical usage; (c) experiment was not conducted in most of the studies; (d) very few studies were aimed at Chinese English learners. Taking these features into account, the author decided to engage an empirical study to try to improve writing teaching in Chinese advanced English majors' context, and it is expected that this study will provide reference data for the pedagogical reform of English writing course. Halliday and Hasan's cohesion theory, which has had a major impact on the understanding and the teaching of cohesion, was employed in the present study. According to them, cohesion refers to the semantic relations between sentences by which a text is bound together as a unified whole; besides, they explicate that cohesion is created by five important devices: reference, substitution, ellipsis, conjunction and lexical organization.This study was designed to explore: (a) whether cohesion is teachable in AEMs1 composition class; (b) what procedures are effective if cohesion is teachable; (c) whether the teaching of cohesion contributes to the improved writing performance. In this study, junior English majors in the experimental class (EC) were taught cohesion in addition to a process method, while the students in the control class (CC) were taught through only aprocess method. Both classes did almost the same amount of exercises, but with different focus: the CC focused on writing skills, structures, etc.. while the EC on textual cohesion. Each class wrote six papers, and the first (pretest), the third (Draft 3 and Draft 4) and the last papers (posttest) were rated by four experienced teachers independently with the rating scale for the writing item in TEM4. The SPSS procedure was used to do the calculations. First, t-test was employed to make a general examination on the overall quality of the pretest, posttest, Draft 3 and Draft 4 in both classes. The results of the t-tests demonstrated the EC and the CC did not differ from each other significantly at the beginning of the experiment (p>.005), but they did differ statistically at the end of the experiment (p< .005, and the observed t < the given t). This result indicated, with all the other variables controlled, the instruction on cohesion facilitated the progress by the students in the EC; to make the result more reliable, a case study was undertaken to make a specific examination on the use of cohesion. The analysis on the randomly selected papers in the two classes further showed that the EC's improvement could be attributed to the use of cohesive features which reinforced the textual meaning, made the papers coherent, thus, enabled the readers to interpret the papers easily and correctly.The results of the t-tests and the case study reinforced each other, suggesting that the EC students make improved performance in their papers after the instruction on cohesion. The students' success can be attributed to a number of factors: (a) teaching the students cohesion made them more sensitive to the reader's needs; (b) the knowledge of cohesion helped the students in their revising process; (c) teaching the students cohesion facilitated their peer editing activity. The findings demonstrate that teaching cohesion to advance...