Interventionist Account of Causation
The interventionist account of causation is a philosophical framework that defines causation in terms of interventions and counterfactuals. Associated most closely with James Woodward's "Making Things Happen" (2003), the account holds that \(X\) is a cause of \(Y\) if and only if there is a possible intervention on \(X\) that would change the probability or value of \(Y\), where an intervention is an exogenous change that sets \(X\) to a specific value without directly affecting other variables except through \(X\)'s causal consequences.
The interventionist account emerged from dissatisfaction with regularity theories (Hume) and probabilistic theories (Suppes), which struggled to distinguish genuine causation from spurious correlation. By grounding causation in manipulability, the framework connects philosophical analysis to experimental practice: a scientist's claim that \(X\) causes \(Y\) is vindicated by the ability to produce changes in \(Y\) by intervening on \(X\).
The framework has been influential in philosophy of science, causal inference, and the foundations of causal emergence. In causal emergence, the interventionist criterion provides the test for whether a macro-level variable is genuinely causal: can we intervene on the macro-state and produce reliable changes in the system's future? If so, the macro-state is a cause, regardless of its micro-implementation.
Critics argue that the interventionist account is circular — it defines causation in terms of interventions, but interventions themselves presuppose causal knowledge about how to manipulate the system. Defenders respond that the circularity is not vicious: the account provides a systematic framework for testing causal claims, not a reductive definition of causation in non-causal terms.
The interventionist account is not a theory of what causation is in the mind of God. It is a theory of what causation is in the mind of a scientist with a pipette, a lever, or a budget. It replaces ontology with methodology — and in doing so, it may capture everything about causation that actually matters.