The Relationship Between Thingking In L1 And L2 Writing Ability

The paper reports on a study of the process of English picture composition in an attempt to reveal the relationship between thinking in L1 and L2 writing ability. The subjects involved in the study were 22 college students majoring in Business English from Shandong Agricultural Administrators' College. They were asked to write an English composition in about 150 words based on the pictures given, while speaking out whatever occurred in their minds by the method of thinking-aloud. The writing task was undertaken individually and a tape recorder was used to record the whole process of thinking-aloud. The sixteen subjects were further selected out of the 22 for the final data analysis because they showed a positive attitude towards the method of thinking-aloud and silence time was less than 20 % of the total writing time. The data analysis included two tasks: (1) scoring the composition; (2) transcribing the tapes and analyzing every protocol. Two experienced teachers were invited to score English compositions independently and the internal reliability between the two raters reached 0. 80. The protocols were analyzed in three steps. First of all, the total number of Chinese words occurring in the protocols was counted against the total number of both Chinese and English words. Secondly, individual events involving the use of L1 were identified in the protocols and the function performed by L1 in each event was then defined through repeated qualitative analyses. Finally, the higher achievers were compared with the lower achievers in terms of the function categories of L1 use. The results showed that the students with higher scores on composition used L1 far less than the students with lower scores in terms of the total amount. However , considering the five functions of L1 use identified from the data ( i. e. transformation , confirmation , generating ideas , retrieving L2 forms and controlling the writing procedures) , the students with higher scores employed more L1 for transformation and confirmation , but less L1 for the other three functions. Furthermore, the mean length of each L1 utterance used by the higher achievers was longer than that used by the lower achievers in the case of transformation and confirmation, but shorter than the lower achievers in the other cases. These findings are as the same as those of Wen and Guo (1998).Based on the Hayes-Flower model of L1 writing ( see Hayes, 1996) and Bibliography L2 writing process( Wang & Wen, 2002) and incorporating the discoveries of L1 use made in this research, a model of L2 writing process is suggested. Due to the small sample size, the results yielded from this study will only serve as hypotheses to be tested in the future.