contrafactuals

Counterfactuals, also called contrafactuals, are conditional statements about what would be the case if circumstances different from those that actually occurred. They are typically expressed in the subjunctive mood, as in “If I had left earlier, I would have caught the train.” Counterfactual reasoning is central to philosophy, linguistics, cognitive science, and causal inference, because it licenses reasoning about alternatives, responsibility, and explanatory power.

Semantically, counterfactuals are analyzed using possible-world semantics. The canonical idea, associated with Stalnaker and Lewis, is that a counterfactual is true at a given actual world only if, in the nearest possible worlds where the antecedent holds, the consequent also holds. The precise notion of closeness and the selection of worlds differ by theorist; Stalnaker emphasizes a single most similar world, while Lewis uses a comparative similarity ordering.

Counterfactuals are contrasted with material conditionals and often require causal links: not just logical implication but dependence in a causal model. Some accounts rely on structuring assumptions or interventions to determine the truth-value of the consequent when the antecedent is imposed. Debates include whether there is a fixed strictness of similarity and how to handle impossible antecedents.

Applications include the analysis of causation and explanation, revision of beliefs based on hypothetical scenarios, and formal modeling of decision making. In linguistics, the study of counterfactuals informs mood, tense, and context. In artificial intelligence, counterfactual reasoning supports planning and causal inference under uncertainty.