recent empirical findings in psychology, anthropology and neuropsychology call into question the idea that all fields of knowledge are organised according to the same general cognitive principles. They suggest that different empirical areas correspond to different ways of cognitive structuring. On the basis of this hypothesis of structuring knowledge by field, it is possible to consider alternatives to thinking based on the dichotomy between modular peripheral cognitive systems and isotropic and quinian central systems. A rapid history of the hypothesis of structuring knowledge by field is proposed. The ongoing discussions on the interpretation to be given to experimental results are briefly presented. Cognitivist design of cognitive architecture is characterised. The contrast between central and peripheral systems is highlighted, as well as the philosophical assumptions underlying this vision of central systems. Some difficulties with this design are discussed, in particular the so-called problem of the framework. I show the type of solution that can be suggested by the assumption of the SCD and the type of change it entails of the traditional vision of cognitive architecture.