Evolving systems and directionality

Abstract Evolving systems in both the life and physical sciences are often thought to be directional. The processes that drive the evolution of systems are diverse, ranging from natural selection to thermodynamics. However, in many treatments of these processes, the different kinds of directionality and types of ends that evolving systems trend towards are often either poorly specified or implicitly assumed. This paper aims to provide a classification of ends and directional processes that can be used to identify and characterize directionality in evolving systems. Specifically, we propose that directional evolution can be either terminal or perpetual, with perpetual further divided into targeted or open-ended. Additionally, we caution against conflating organization, order, and complexity, as each tracks different properties of a directional system.