Complexity : a review of the classics

This text was written as part of the project Modelling of Complex Systems for Public Policy. It reviews the classical authors who jointly contributed to establish the elements of what could constitute a "science of complexity". Based on the original writings of these authors, the text discusses the central concepts of complex systems: i) the interaction between (homogeneous or heterogeneous) agents and the environment; ii) emergence and self-organization; iii) the importance of nonlinearity and scales; iv) the determinism of rules; v) the emphasis on dynamics and feedback; and vi) the notions of adaptation, learning and evolution. Finally, contemporary criticisms are presented. They suggest that the arguments of complex systems do not support the epistemological establishment of a supposedly new science, but they do not reject the advances proposed by complexity studies.