| Over the past decade, the possible effect of a variety of task variables, ranging from task type, task structure, and task complexity to task processing conditions and the like, on L2learners’oral performance has been investigated by L2researchers both abroad and in China. Most of these studies have been made in light of cognitive models represented by Skehan, P.’s (1998) and Robinson, P.’s (2001). Some of the task variables studied have been found relatively consistent predictors of L2performance, while others not, among which is the variable of task structure. Furthermore, the effect of interactive mode of task as implementing condition on L2oral performance has been relatively less examined and got no consistent findings.The present study, which adopted the cognitive models as a main framework but was also inspired by some sociocultural perspectives, aimed to explore the impact of the inherent structure of narrative tasks and interactive mode as task implementing condition on English L2learners’oral performance. Thereinto, the tight/loose structure and the dialogic/monologic mode were designed as two independent variables respectively and the complexity and accuracy of English L2learners’oral performance as the dependent variables. The participants were37freshmen in Nanjing Normal University Taizhou College who were not English majors and whose average English level was intermediate or lower intermediate. The participants were randomly arranged into monologic group (15participants) and dialogic group (22participants). The two groups performed two narrative tasks of different inherent structures. The whole process of task performance was recorded and transcribed by the researcher and her assistants. The data, which had been inspected for its validity, was analyzed for accuracy measure (EFC ratio) and complexity measures (D-index and MLU and SC/AS-units ratio) by using CLAN (V27-Sep-2012) and Excel2007; and the differences within and between groups were tested by SPSS17.0. The major results are as follows:a) The inherent narrative structure was not found to have any significant impact on the accuracy and complexity of participants’English L2oral performance.b) The interactive mode (dialogue and monologue) as task implementing condition was found to have no significant influence on accuracy of the participants’ English oral performance. However, dialogic condition was found to cause significant lower syntactic complexity with both structure types. c) A combined effect of the Inherent narrative structure and interactive mode as task condition was found:Tightly structured task in a dialogic mode led to significantly higher lexical diversity, but it did not affect syntactic complexity.The research results were explained and discussed from cognitive views of processing time, cognitive demand, and online-planning on the one hand; and were analyzed supplementarily with a sociocultrual concept of "collaborative dialogue" on the other hand. The study has partially confirmed the cognitive models and related previous studies, but also reveals that cognitive models alone and quantitative approach only cannot adequately reflect the complex nature of task variables and their influences on learner performance, particularly true with the accuracy and complexity change under the dialogic task condition. Other implications of the study include more appropriate choice of task material difficulty according to learners’level and the need of task planning when using task structure as a predictor of task performance. |
