A PHILOSOPHICAL FOUNDATION FOR UNIFICATION OF DYNAMIC MODELING METHODS BASED ON HIGHER-ORDER POTENTIALITIES AND THEIR REDUCERS

In the development of disciplines addressing dynamics, a major role was played by the assumption that processes can be modeled by introducing state properties, called potentialities, anticipating in which respect a next state will be different. A second assumption often made is that these state properties can be related to other state properties, called reducers. This paper proposes a philosophical framework in terms of potentialities and their reducers, which can be used to obtain a common philosophical foundation for methods in AI, cognitive science and beyond to model dynamics. Based on this framework a metamodel for dynamic modeling approaches is described. The philosophical framework and the metamodel together provide a unified foundation for numerical, symbolic, and hybrid dynamic modeling approaches used in a large variety of disciplines.