Through a presentation of today's cognitive design research, the text introduces our view of design as a construction of representations and extends Goel and Pirolli (1989)'s generic-design hypothesis, based on Simon (1969)'s approach to design. Simon's view of design as ill-structured problems solving has been amended in many cognitive design studies. Design involves complex problems, rarely decomposable into independent sub-problems. Designers tend to start a project by generating a few simple objectives, creating a solution kernel to which they are going to stick. Rather than one, "correct" solution, problems have several more or less satisfying solutions. Problems and solutions lack pre-existing, objective evaluation criteria. Reuse is a fundamental strategy. The opportunistic organisation of design remains under discussion. According to our view, designing consists in specifying an artefact (the artefact product), given requirements that indicate functions to be fulfilled, and needs and goals to be satisfied by the artefact. At a cognitive level, this specification activity consists of constructing representations of the artefact until they are so precise, concrete, and detailed that the resulting representations specify explicitly and completely the artefact product's implementation. This construction proceeds through generating, transforming, and evaluating activities.