Different forms of causation in dynamical systems: Determinism, pattern generation, and information

Abstract

The standpoint that three fundamentally inequivalent classes of system behavior exist, which deserve separate tools for their characterization, is developed. These classes are: determinism, pattern generation, and information. A simple explanation for the origin of these classes is given. The three classes are shown to correspond to the different possible ways in which the properties of the components of a system unfold in the course of the system's processes. Determinism will be defined as a form of causation in which the properties causing the behavior are factually the same as the ones which for the observer specify the current conditions of the system. As a consequence, deterministic systems can be characterized by state equations that describe how the observable properties change. Pattern generation is characterized here as a form of causation in which the observable properties and the behavior-generating properties correspond to each other as one to the many. Consequently, we experience in terms of the observable properties a branching behavior which is characterized by a pattern of selections at the branching points, a pattern that stands in close correspondence with the visually perceivable patterns and forms, generated by the system's dynamics. Information will be understood as a name for a process and an instance of causation where properties that cause behavior differ from those which we directly observe. The behavior of the system appears to be a consequence of the content of information hidden in the components. This information is, in turn, epistemologically equivalent to the very causal process in which it is unfolded.