A Tribute to Simon, and Some -Too Late- Questions, by a Cognitive Ergonomist
Résumé
Herbert Simon's proposal of a "science of design" is fundamental to the many researchers in cognitive psychology and ergonomics. This is due to Simon having applied the information processing paradigm to design problem solving and having elaborated various characteristics of design that are essential in the cognitive design problem solving processes. Empirical studies on design have also led, however, to results that lead to nuance Simon's approach. We first expose the elements proposed by Simon underlying our viewpoint on design. Then we present some reservations with respect to Simon's approach, especially concerning his tendency to under-estimate in design (i) the specificity of ill defined problem solving, (ii) the importance of problem representation building, (iii) the role of "nondetermini- stic" "leaps", and (iv) the obstacles met by the problem decomposition method. We also briefly present another paradigm, i.e. "situated cognition", adopted by design researchers in reaction to Simon's rational problem solving. Even if it makes up for some of Simon's limitations (among those cited above, especially for the role of constructing problem representations), we don't advocate this alternative approach above Simon's. We will propose an interpretation of the differences between the two, claiming that they are not contradictory, but complementary, because they focus on different aspects of design. We will conclude by some -too late- questions to Simon.