A rule of inference: from 'If P then Q' (P → Q) and 'not Q' (¬Q), infer 'not P' (¬P). It is the standard way to reject antecedents via failure of consequents.

Principle

The organizing idea is contraposition-based rejection: a conditional that guarantees Q when P holds licenses negation of P when Q fails, under logics that permit contrapositive reasoning.

Demonstration

Illustrative scenario: If 'If the alarm detects smoke then the alarm sounds' and 'The alarm did not sound' are accepted, modus tollens licenses inferring 'The alarm did not detect smoke'. In a scientific testing context, failing the expected consequence weakens or refutes the hypothesis.

Misapplication

Applying modus tollens where the conditional is probabilistic, defeasible, causal, or context-sensitive (e.g., 'If P then usually Q') without qualifying the inference leads to overconfident or incorrect rejections of hypotheses.

Consequence

Permits falsification-style reasoning: it is a foundational tool for refuting hypotheses, eliminating candidates, and constructing contrapositive proofs in many formal systems.

Reversal

The improper reversal is denying the antecedent (from P → Q and ¬P infer ¬Q), which is invalid. Modus ponens is its complementary affirmative counterpart.

Boundary

Relies on the conditional being sufficiently strong (material or stronger) to support contraposition; in logics that do not validate contraposition or where conditionals are nonmaterial, modus tollens may fail or require reformulation.

Semantic Tension

Tension between contrapositive validity and contexts in which consequents can fail for reasons unrelated to the antecedent (multiple causes, background conditions); the tension is practical as well as formal.

Synthesis

Modus tollens is the contraposition-based inference that when a conditional is accepted and its consequent is false, the antecedent can be rejected; it underlies falsification but must be applied with attention to the nature of the conditional.