Variational causal claims in epidemiology.

This article examines definitions of cause in the epidemiological literature. Those definitions describe causes as factors that make a difference to the distribution of disease or to individual health status. In philosophical terms, they are "difference-makers." I argue that those definitions are underpinned by an epistemology and a methodology that hinge upon the notion of variation, contra the dominant Humean paradigm according to which we infer causality from regularity. Furthermore, despite the fact that causes are defined in terms of difference-making, this doesn't fix the causal metaphysics but rather reflects the "variational" epistemology and methodology of epidemiology. I suggest that causality in epidemiology ought to be interpreted according to Williamson's epistemic theory. In this approach, causal attribution depends on the available evidence and on the methods used. In turn, evidence to establish causal claims requires both difference-making and mechanistic considerations.