BibSonomy ::  publication ::

We report on a study on construction of knowledge through successive argumentative activities. 120 Grade 5 students filled first a questionnaire to express individually their standpoint on a controversial issue. They then engaged in argumentative talk in triads, having at their disposal databases presenting information on the issue. Individuals filled then the questionnaire again. The triads went on in their argumentative talk. In one group G1, N1=60, triads used a computerized tool, the argumentative map, to represent viewpoints and reasons supporting them. In the second group G2, N2=60, triads used a two-column table for inserting "pro and con" reasons. The triads wrote then a collective essay. Finally, individuals filled the questionnaire for the third time. Knowledge and construction of knowledge were measured through arguments-outcomes produced in the successive activities. We observed in general that individual arguments were gradually less one-sided and more compounded. Also the reasons invoked were more relevant to the viewpoint claimed and more acceptable. More reasons supporting alternative arguments were raised. Finally the quality of the reasons invoked was higher. The analysis of collective arguments considered as produced by one cognitive entity showed that argumentative maps helped to construct significantly better collective arguments than tables. However, the measures of final individual arguments were significantly lower than collective arguments in both groups. This seems to indicate that while maps and tables helped representing arguments agreed upon members of groups, individual students only partly internalized the collectively constructed arguments via the mediation of the tools to construct own arguments.