Thomas Smith, Gregory Stevens

Paper #: 94-08-046

Over the last several decades, simulation-based theoretical research on complex systems has led to a revival of scientific and philosophical interest in emergent phenomena. In such areas as evolutionary biology and cognitive science, computational results obtained in the course of simulations have uncovered surprises about how natural systems work, typically by disclosing formerly unrecognized features of a system’s organization. Unanticipated theoretically, such patterns can be understood to have arisen “bottom up” (Seidenberg 1993; Casti 1994, 154ff). That is, they are computed on the basis of local rules presiding over the interaction of a system’s elements, as specified by investigators in a simulation’s design. Bottom up discoveries have proven to be of interest to philosophers of science because they highlight shortcomings in traditional positivist arguments about scientific theory and explanation, and because they have opened new avenues of investigation into reductionism and emergence. The same discoveries also interest working scientists because they vindicate computational strategies for theoretical research, and because they have begun to reveal how the structure of complex systems depends in fundamental ways on the interaction of a system’s parts.

PDF