Soft control: Furthering the case for Modified Interventionist Theory

Not all testing interventions that we might want to perform, or need to be performable in principle, fail to cause off-path variables. This is a problem for Woodward's Orthodox Interventionist Theory of causation, but not the ‘Modified Interventionist Theory’, which I proposed in a previous issue of this journal (Friend, 2021). As I explain here, this is because only the modified theory permits ‘soft control’. I will survey three different kinds of case (beyond the case considered previously) in which soft control is necessary for a reasonable application of interventionism. These include cases where soft control makes intervention more practical, physically possible, and causally probative in the context of mechanisms. I'll also take the opportunity to remove some of the confusing aspects of my original formulation of the modified theory. The result, I believe, constitutes a strong case for it.