Discourse Features In Chinese EFL Learners' Argumentative Writing: A Developmental Perspective

This study aims to investigate discourse features in Chinese EFL learners'timed argumentative essay writings. On the basis of the theories and empirical studies of cohesion and coherence in the fields of second language writing, the researcher constructs a theoretical model for the present study from the perspective of discourse analysis to seek out the developmental patterns of discourse features and how the discourse features of cohesion and coherence are related to the quality of writings by Chinese EFL learners. Review of the relevant literature shows that since its inception in late 1970s, studies of the relationship between cohesion and coherence have been the focus of researchers. However, discourse features in L2 writing have rarely received enough attention from the writing researchers. There is an urgent need to devise rating scales and other criteria for discourse elements and to assess the quality of written texts that test takers produce.To better understand the nature of textual coherence and cohesive devices used in argumentative essays produced by EFL learners across all levels,201 English majors in a university in Shandong Province participated in this longitudinal study, and their writing products, as part of the terminal exam in English Intensive Reading course in three academic stages, taken as low, intermediate and advanced levels were collected. The final 591 writing samples with a total number of 138,635 words were rated for coherence and overall writing quality respectively and coded for cohesive device use by 11 raters in all.Preliminary and exploratory analyses were conducted with Statistical Packages for Social Sciences, Version (16) to conduct this study. More specifically, One-way Analysis of Variance (ANOVA) was performed to test whether there was variation in the frequency of the use of cohesive devices and scale of coherence ratings among the three different levels; Post Hoc Multiple Comparisons LSD Tests identified how each group of students differed significantly from one another in the variables investigated. Then Pearson Product-moment Correlation was computed between cohesion and coherence to determine whether the discourse features were significantly correlated with each other. Finally, comparisons were made through Independent Samples T-test to explore the significant differences in the use of cohesive devices and ratings of coherence scales between good and poor compositions. The results indicate the following aspects. From the perspective of the use of cohesive devices, the students are able to use a variety of cohesive devices in their writings across three levels, among which lexical devices take up a larger percentage of the total number of cohesive devices than grammatical devices. The kinds of cohesive devices most commonly used across all three levels are pronouns, repetition and conjunctions. There is a significant decrease in the density of cohesive devices at the higher proficiency level. Significant differences have been found among most of the types of cohesive devices investigated across grade levels. No significant differences have been found between good and poor compositions in terms of the use of cohesive devices. From the perspective of the ratings of coherence scales, significant differences have been found across grade levels and writing abilities in their ratings of coherence scales. There is a significant increase in their coherence rating scales at the higher proficiency level. Strong endings, good organization and smooth transitions are found to be better predictors of writing quality of all the items of coherence rating scales. In terms of the relationship between cohesion, coherence and writing quality, cohesion is not significantly correlated with coherence. There is no significant correlation between cohesion and writing quality. There is a strong positive correlation between the ratings of coherence and writing quality.The study has significant pedagogical implications for both teaching and learning. L2 writing pedagogy cannot depend on the teaching of cohesive devices in order to improve students'writing ability, since the use of cohesive devices does not necessarily translate to better texts, nor does it distinguish between good and poor writings. Coherence needs to be taken as the focal point of good writing but not merely as one element of good writing. In other words, a coherent text is characterized by carefully chosen words, grammatically correct sentences, a clear thesis statement, a well-developed structure, an effective beginning and a powerful ending. It is the integration of all of these elements that will help the writer construct a coherent text. A deficiency in any one of them could possibly lead to a break in coherence. Limitations of the study and recommendations for future studies were finally provided. Owing to the complex and multi-faceted nature of cohesion and coherence, the factors explored in the present study are not enough to explore their relationships with the holistic writing quality. Therefore, more efforts are still needed to promote further research for the construction of an accurate and succinct rating scale that will make valid and efficient assessment possible.