A Study Of Discourse Structure In The Oral Description Of Pictures By Chinese Secondary EFL Learners

This study was undertaken to explore the discourse structure generated in the oral description of the given pictures by the Chinese high school English learners. It focuses primarily on the description of the macrostructure in their oral production, and secondly on the differences and similarities in the micro-components between the high- and low-level speakers.Data employed in the present study were oral compositions made by 40 students from a high school who were required to describe a series of pictures in a given time. All of their oral work was recorded, transcribed and scored by 5 experienced middle school teachers who received a special training before doing transcribing and scoring work. Eventually, 33 valid cases were chosen for analysis, and were divided into high- and low-level groups according to the research questions. In this study, qualitative methods were used in the data analysis, in which the researcher discussed with her supervisor and some of her colleagues when uncertainties arose in the categorization of the identified discourse features. The data analysis yielded the following major findings:Firstly, three major components at the macro-level were identified in the learners'oral description of the pictures on the basis of their shared discourse features: Lead-in Description, Picture-related Description, and Coda. The major components in macrostructure indicate that the Chinese EFL learners have already developed a schema of narrative talk regardless of whether they stay at a high or low proficiency level.Secondly, Lead-in Description at the micro-level may have three types: The first type is Task-oriented Description, in which the students may start their talk by saying a bit about the task they were going to do. The second type is Theme-oriented Description, in which the students may start their talk with the generalization of a theme behind the pictures they were going to describe. The third type is the Absence of Lead-in of any kind, in which the students came to the task of the description without any preparation for the listeners. Picture-related Description at the micro-level consists of two minor categories. The first is Picture-based Description, in which the body of the speakers'talk centers around the given pictures either in a spatial sequence or in a temporal sequence. The second is Non-picture-based Description, in which the body of the speakers'talk is inspired by or related to the pictures but the speakers gave no description about the pictures.The category of Coda at the micro-level may have two types. One is Evaluation, in which the speaker may wind up his/her talk with some commentary remarks after the description of the pictures. The other is Theme Restatement, in which the speaker may finish the oral work by restating or paraphrasing the theme, if any, mentioned earlier in the beginning of the talk.The components at both macro- and micro-level provide a general picture of discourse structure in the Chinese EFL learners of the secondary level.Thirdly, the cross-level comparisons reveal: (1) More or less the same percentage of the speakers went directly to the picture describing without making any preparation for the audience. However, the high-level speakers were more flexible in starting their picture description than the low-level speakers, almost half of whom employed Theme-oriented Description to start their task. (2) In the case of Picture-based description, high-level students have a tendency to use more spatially-sequenced events in the body of the talk, while low-level speakers tend to employ temporally-sequenced description of the events. (3) In the case of Non-picture-based description, the low-level speakers tend to use more exposition-oriented description compared with the high-level ones, but such descriptions are insufficiently made, in which the low-level speakers'descriptions are either incomplete in content or poorly organized. It follows that the spatially-sequenced descriptions is more effective in the structuring of the information underlying of the described pictures.Finally, almost half of the low-level speakers did not include Coda in their description, or came to an abrupt end upon the completion of the description of the pictures. To some extent, therefore, Coda may become an indicator to the oral performance of the speakers their description of pictures.The finding s of this study may have both theoretical and pedagogical implications. Theoretically,the macrostructure identified in this study suggests that narrative units are more or less the same at a very high level regardless of different levels of proficiency, which has confirmed Rumelhart's (1975, 1977) (and many others') claim that prototypical elements are universally present in stories. The notion of sequential description in this study may also have also some theoretical implications about story grammar or story schema. Practically, this study has offered insights into the significant role of message sequencing in L2 learners'discourse competence development, and into the assessment of the learners'oral discourse ability in both testing and instructional settings.