James Crutchfield
Paper #: 94-03-011
This brief essay reviews an approach to defining and then detecting the emergence of complexity in nonlinear processes. It is, in fact, a synopsis of Reference [1] that leaves out the technical details in an attempt to clarify the motivations behind the approach. The central puzzle addressed is how we as scientists--or, for that matter, how adaptive agents evolving in populations--ever “discover” anything new in our worlds, when it appears that all we can describe is expressed in the language of our current understanding. One resolution--hierarchical machine reconstruction--is proposed. Along the way, complexity metrics for detecting structure and quantifying emergence, along with an analysis of the constraints on the dynamics of innovation, are outlined. The approach turns on a synthesis of tools from dynamical systems, computation, and inductive inference.