The Effect of Talk in Argument Text Construction
Research on second language teaching and learning has to date focused primarily on the major skills of reading, writing, listening, and speaking but has treated them as relatively separate areas of investigation. By contrast this research investigates the role of one skill, speaking, in the performance of another, writing. The study investigates the effect of an instructional sequence that aimed to prepare upper high school students (Form 6/Grade 12) to write better argument essays. The sequence was experienced by the students in two ways. One way was for students to engage in talk with a peer before and during writing. The other way was for students to work in a solitary way. Qualitative data analysis compared the writing scores gained by students on two sets of variables: one to indicate the quality of text in general terms (Hamp-Lyons, 1986) and the other to indicate quality of text in terms of specific features of argument: claims, elaboration of claims, grounds and elaboration of grounds (Toulmin, 1958; Toulmin, Rieke and Janki, 1984). The results of the qualitative analysis indicate that opportunity to work with a peer before and during writing had a limited and specific effect on the texts that students wrote. Positive effects for opportunity to talk were seen in the quantity of grounds-related material, but only when students wrote texts that appeared to require more content and domain-specific knowledge (Alexander, Schallert and Hare, 1991). It appeared that talk could operate to help students access relevant prior knowledge (Alexander, Schallert and. Hare, 1991) to support the claims made in their argument texts. Working in a solitary way resulted in significantly better mean scores for linguistic accuracy and complexity. This finding is not consistent with claims made in the output hypothesis (Swain, 1985; Swain, 1995; Swain and Lapkin, 1995). One explanation is that students working together may not have been 'pushed' to consider aspects of language form in attempts to communicate meaning. Another is that this did occur but consideration of form did not transfer to subsequent writing. Another variable that appeared to influence writing quality was the level of resourcing provided for the writing task. Access to textual resources (input in the form of cohesive and linear text) appeared significantly to affect all three of the general measures of text quality, suggesting that textual input is a valuable linguistic and rhetorical resource for writers. When students' texts were analysed specifically for frequency of features of argument, different effects were found for levels of resourcing. Claims and elaboration of claims were most affected by the semi-resourced form of input represented by fact sheets (lists of propositions). Students appeared to make use of input in the form of fact sheets for meeting claim-like requirements in their texts. This might have been because the fact sheets represented information in a way that required the least amount of transformation to be accessible and useful. Analysis of transcript data was carried out on three selected pairs of students to explore the nature of the talk which produced significant and positive results. The type of talk associated with the pair that showed the greatest scores was qualitatively different in terms of the amount and topic of substantive talk and the frequency of responses to initiations. The talk also operated to push each participant, particularly the weaker of the two, to respond, explain and elaborate. The fact that the weaker student in the most productive pair made use of what he articulated suggested that, for him, the talk appeared to set the discourse parameters of the writing task. In addition, the results of the study pointed to the fact that speaking with a partner, particularly a more expert partner, before and during writing can bring positive effects particularly for drawing on relevant prior knowledge thereby enhancing content and domain-specific knowledge. A proficient and interactionally expert partner can promote discussion of relevant prior knowledge useful for supporting claims made in argument texts. The analysis of transcript data indicates that few students show interactional proficiency and that this may prove a worthwhile focus for pedagogy. The present study supports the line of research in collaborative learning (Cohen, 1994) as it has explored the conditions under which positive effects on writing are likely to occur. Research may profitably continue to explore the features of successful interaction and the conditions that successful interaction creates, particularly as it enables better writing. Not only are conditions worthy of further research, so too are effects, as they are likely to operate on different aspects of writing and in different genres. Constraints operating particularly in the area of argument need to continue to be explored empirically. The present study has concluded with the belief that there is still much to know in the relationship between speaking and writing. For this reason, teachers may do well to pay careful consideration to the way in which pair and group tasks are managed in the classroom. This entails the provision of guidance and support for the participants so that purposeful interaction occurs.