ABSTRACT
Science is practised by locating invariant patterns in the world and using these as the basis for attributing complexes of causal capacities to particular cases. Talk of capacities, powers and the like is endemic to common-sense reasoning as well as to scientific reasoning. It is commonly believed, and scientifically established, that smoking can cause heart disease and that aspirins can cure headaches. Non-idealized capacity claims can be tested against counterfactual intuitions, which can be used to investigate the robustness of causal attributions. Causal relations are typically attributed to the world as a result of the identification of invariant relations, which are then idealized to yield causal laws. However, common-sense causal reasoning cannot be more than merely approximate. In order for a discipline to progress to a stage where it can be used to put forward powerful explanatory and predictive theories, it needs to be transformed into a discipline of precise measurement.
