An instrumentalist approach to causation: The case of landslide research

Landslide risk offers a particularly interesting field of study for philosophers of science working on causation. Landslides are the result of complex causal dynamics, and the object of study of a variety of disciplines which are ultimately aimed at effective risk mitigation. For these reasons, causation is central to many instruments developed to study landslides. We illustrate four such instruments, and we show how each of them assumes a different concept of causation. This observation provides the ground for the defence of a pluralist account of conceptual causation. However, in defending this position we do not aim at establishing a universally correct account of causation. Instead, we examine the adequacy of different concepts for a specific purpose and in a specific context. We therefore supplement the descriptive observation of the plurality of concepts of causation involved in landslide research with the normative claim that this plurality cannot be reduced under a monist account, because the different concepts are necessary for different epistemic purposes. While our conclusions only hold for the specific case we analyse, the instrumentalist approach we propose can be generalised to other contexts and to other concepts.