Engineering design may be considered to be the kernel of technology. Design is a creative process, which cannot be completely performed by deductive reasoning. At essential moments the designer has to take recourse to reductive reasoning steps. These however lead to many alternative conclusions, many of them false, several good. The remedy for this situation is presented by methodology, which greatly enhances the correctness of the reductive conclusions. This paper presents the logical structure of the various kinds of reductive reasoning, and discusses the basic design cycle as the core of design methodology. Some personal experience illustrates the argumentation.