A philosophical account of causation that analyzes causal relations in terms of hypothetical or actual interventions: X is a cause of Y if manipulating X (while suitably holding background conditions fixed) systematically changes Y's distribution or values in ways that support control, explanation, or prediction.

Principle

Causal relations are those that license interventions for manipulation and control: a variable is causally upstream of another when an intervention on the first changes the second, holding background mechanisms constant and respecting relevant invariances.

Demonstration

Situation: A researcher wishes to know whether a drug reduces blood pressure. Recognition: Observational correlation exists but confounding is possible. Action: Conduct an intervention (randomized administration of the drug) that sets treatment independently of confounders and measure blood pressure outcomes. Consequence: If manipulating treatment changes the distribution of blood pressure, the interventionist claims a causal relation supporting control and policy decisions.

Misapplication

Inferring causation from any counterfactual dependence without checking whether the dependence arises from manipulable mechanisms (e.g., dependence due to a common cause or structural constraints), or treating interventions that alter background conditions as equivalent to idealized manipulations.

Consequence

Grounds experimental design and policy by linking causal claims to manipulative strategies; clarifies which variables are appropriate targets for control and connects causal language to practical interventions and predictability.

Reversal

For causal relata that cannot be meaningfully intervened upon (past singular events, abstract mathematical truths, some ecological or historical processes) or when interventions would destroy the system's relevant background conditions, the interventionist account struggles and must be supplemented by other causal concepts or counterfactual analyses.

Boundary

Applies to causal claims about systems where interventions can be coherently defined and performed and where background invariances hold. Excludes purely metaphysical accounts that separate causation from manipulability, and situations where intervention is ill‑defined or impossible.

Semantic Tension

Manipulatability/Control ↔ Explanatory breadth: privileging manipulable causes supports intervention and policy but can underrepresent non‑manipulable causal relata that remain explanatorily important.

Synthesis

The interventionist theory ties causation to the capacity to alter outcomes by interventions, making causation operational for experiment and policy; its limits highlight the need to integrate manipulability with broader explanatory and historical causal notions.