Cognition and Rhetoric in English Language Learners' Writing: A Developmental Study

The present study examined the effectiveness of an instructional method in English language writing. The instruction concerned a cognitive process, i.e., Bereiter and Scardamalias (1987) knowledge-transforming, and a discourse genre, i.e., the Toulmin (1958/2003) model of argument. The instruction in the process is significant since generating discourse content was identified as a problem for novice writers. The instruction in the Toulmin model is significant since lack of attention to genre was identified as a problem in cognitive approaches to writing.

To teach and research knowledge-transforming composing and the Toulmin model, the tenets of cognitive strategy instruction in writing and sociocultural theory of mind were adopted. Instruction was adopted after Scardamalia, Bereiter, and Steinbach (1984) and had three components: explicit strategy instruction in the Toulmin model, mediation of the writing process through artefacts, and two types of verbalization: (focused) freewriting and languaging. The study had a mixed-methods design with a quasi-experimental quantitative component and a qualitative component consisting of textual analysis, dynamic assessment (DA), semi-structured interviews, and surveys.

The results indicated statistically significant gains for two of the categories of the Toulmin model, i.e., rebuttal and response to rebuttal, in the texts generated by the experimental group (EG) (n = 13) when compared with those of the comparison group (n = 13). Specifically, the gains suggested the rise above conflict criterion (Scardamalia et al., 1984) in knowledge-transforming, indicating the effectiveness of instruction. When four participants texts in EG group were analyzed developmentally, they also demonstrated knowledge-transforming and improved rhetorical structure. In particular, some discourse features which were absent in the posttest essays were indeed present in those texts. Also, the text analysis indicated the participants were able to use the mediational artefacts to generate discourse content. The DA results indicated that, with varying degrees of mediation, the participants were able to name, generate, and/or revise the discourse features, some of which were absent in the participants posttest essays. The interviews and surveys indicated the participants positive perceptions of instruction and its effect on cognitive change and rhetorical structure of argumentative texts. The study has implications for L2 academic writing instruction, assessment, and research.