Abstract EN:

This work is located at the crosspoint of two problematics: rule inducing and problem solving. Implicit learning has been widely studied in rule inducing, but never really in problem solving. Solving a problem, is to start from an initial state for reaching a final state through a space search. In many problems, this process involves rule inducing. We assume that reaching the goal requires a planification of moves that obscures the rule inducing mecanism. For studying this two process, we adopted the chinese ring puzzle (pac) which contains an initial and a final states, a space search and rules governing the moves. In the two first experiments, we have insured on the possibility of rule inducing and their verbalization (experiment 1), on understanding and use of that rule (experiment 2) for both adults and children. The third experiment demonstrates that adults, when planification is not required, performed better the task when they see how to reach the goal, than when they don't perceive it. On the other hand, whenever the planification of moves is required, the knowledge of the goal obscures rule inducing (experiment 4) because it is necessary to define other subgoals for reaching the main one. We demonstrate that learning and verbalization are dissociated; indeed we discovered a great difference between performance and related verbalizations (experiment 5). Finally, last experiment demonstrates that when the task is semantically perceived through the links that bind objets; the more the problem involves semantic links, the easyer are problem solving and related verbalizations.