台大心理系

The explanatory power of dynamical models in systems science (e.g., systems biology and systems neuroscience) has stirred a heated debate recently. Proponents of mechanistic explanation argue that dynamical models are explanatory only if they are coordinated with mechanistic models. By contrast, others argue that dynamical models can be explanatory without this coordination, and that they can provide explanations that mechanistic models cannot provide. In this paper, I argue that there is some common ground that both kinds of models share, and that their explanatory powers can be subsumed under this common ground in a way that maintains the distinctness of their explanatory powers. The common ground lies in the fact that both kinds of models satisfy what I will call the manipulability criterion, and the distinctness lies in the fact that they satisfy it in different ways.