ABSTRACT
Contemporary research on collaborative learning (CL) lacks the deep integration between theories of learning and theories of communicative interaction that is required for understanding the contexts and processes by which knowledge is interactively elaborated. Until quite recently, CL research has been dominated by the attempt to extend cognitivist theories of human learning, centred on the individual, to the study of learning in groups (Dillenbourg, Baker, Blaye & O’Malley 1996). However, many of the learning processes that these theories propose do not readily correspond to genuine interactive phenomena. For example, results concerning the self-explanation effect (Chi, Bassok, Lewis, Reimann & Glaser 1989) were obtained with respect to individual problem solvers (with experimenter prompting). Although it has been conjectured that this phenomenon can also occur in CL situations (e.g. Ploetzner, Dillenbourg, Preier & Traum 1999), actually finding such explanations (qua individuals’ expressions of their problemsolving processes) in communicative interactions is problematic. Explanations can rarely be analysed as discrete segments of interactions: they are usually processes underlying extended sequences. Explanation is an interactive contextual reconstruction, rather than an expression of problem-solving processes that occurred in an individual’s mind (Baker 1999). A second example of lack of correspondence between postulated learning processes and interactive processes can be seen in the case of the focus of the socio-cognitive conflict paradigm (Doise & Mugny 1981) on the incidence of verbal conflicts, rather than on their interactive contexts and (possibly) associated argumentative processes (cf. Mevarech & Light 1992). Of course, not all theories of CL are cognitivist. However, theories such as Cultural-Historical Activity Theory (e.g. Leont’ev 1981; Engeström 1987) and situated learning (Lave & Wenger 1991) suffer from a different problem in this context: the link between theory, model and (interactive) data remains to be established. Although interactive processes such as ‘dialogue’, ‘appropriation in social interaction’ and ‘negotiation with the situation’ are referred to by these theories, precise models (explicitly derived from these theories) of how such phenomena can be identified and analysed in interaction corpora are in their early stages of elaboration (but cf. Wells 2002).
