Draft for Principia: comments welcome Downward Causation and the Autonomy of Weak Emergence

Weak emergence has been offered as an explication of the ubiquitous notion of emergence used in complexity science (Bedau 1997). After outlining the problem of emergence and comparing weak emergence with the two other main objectivist approaches to emergence, this paper explains a version of weak emergence and illustrates it with cellular automata. Then it explains the sort of downward causation and explanatory autonomy involved in weak emergence. 1. The problem of emergence Emergence is a perennial philosophical problem. Apparent emergent phenomena are quite common, especially in the subjects treated by biology and psychology, but emergent phenomena also seem metaphysically objectionable. Some of these objections can be traced to the autonomy and downward causation that are distinctive of emergent phenomena. Emergence is receiving renewed attention today, in part because the notion repeatedly arises in certain contemporary approaches to understanding complex biological and psychological systems; I have in mind such approaches as neural networks, dynamical systems theory, and agent-based models—what for simplicity I’ll call complexity science. For anyone interested in understanding emergence, two things about complexity science are striking. First, it aims to explain exactly those natural phenomena that seem to involve emergence; the range of phenomena covered by complexity science are about as broad as the examples of apparent emergence in nature. Second, the models in complexity science are typically described as emergent, so much so that one could fairly call the whole enterprise the science of emergence (e.g., Holland 1998, Kauffman 1995). A good strategy, then, for understanding emergence is to turn to complexity science for guidance. A few years ago I introduced the notion of weak emergence to capture the sort of