The impact of oral training on argumentative texts produced by ten-and eleven-year-old children: Exploring the relation between narration and argumentation

10 and 11-year old children were trained to improve their argumentative abilities. Tests were performed on two experimental groups to find out the effects of (1) a didactic sequence of seven lesson with oral exercises, and (2) a co-operative oral debate. To allow us to determine the improvements, the children had to produce three argumentative constrained texts (Alpha-Omega), before and after each phase of training. We compare the results with two control classes to try to determine the quality of transfer between the teaching periods. The qualitative analysis of the texts allows us to identify certain linguistic indicators of the underlying cognitive processes. The results show an effect of training only for the 11-year-old writers. In the 10-year-old children, no clear transfer from oral training was observed. With or without training, ten-year-old writers seem to have difficulties organising the text and managing multiple constraints when writing. In contrast, at 11, the subjects reorganise their activity without taking account of their narrative abilities. The vast majority produce better balanced argumentative texts. However their texts tend to be justificatory rather than negotiated. At the same time, the Alpha-Omega task carries very strong constraints which interfere with cognitive processes during on-line production.