What Makes a Good Counterexample?

Start by identifying the logical shape of the claim

Counterexamples are especially powerful against universal statements: claims that all members of a class have a property, that a rule always works, or that a certain set of conditions guarantees an outcome. If someone says every metal sinks in water, a single genuine metal that floats under the relevant conditions is enough to show that the universal statement is false.

This does not prove the opposite universal claim. Showing that “all A are B” is false does not establish that “no A are B” is true. A counterexample narrows what can reasonably be claimed, forcing the original statement to be revised rather than replaced with an equally sweeping assertion.

The example must actually meet the stated conditions

Many supposed counterexamples fail because they quietly change the subject. If a claim concerns healthy adults under ordinary conditions, a case involving children, severe disease or extraordinary conditions may be interesting but may not contradict the claim as stated. The example must belong inside the scope of the rule it is meant to test.

This is why precise definitions matter. Vague terms let people rescue a claim by moving boundaries after a challenge appears. A fair test establishes the relevant conditions before the counterexample is introduced and then asks whether the case genuinely satisfies them.

The conflict must be with the conclusion, not merely with expectations

An unusual case is not automatically a counterexample. It must produce an outcome inconsistent with what the claim entails. If a theory says an effect is common rather than inevitable, finding a case without the effect does not refute it. Probabilistic claims require probabilistic tests.

Likewise, a surprising anecdote does not overturn a statistical relationship simply because it is memorable. A person who smokes heavily and lives to one hundred does not show that smoking carries no population risk. The claim being tested must be represented accurately before an exception can count against it.

Strong counterexamples expose the hidden assumption

The most useful counterexamples do more than score a logical point. They reveal why the original reasoning failed. A case may show that a definition omitted an important category, that a causal rule ignored context, or that two properties assumed to travel together can come apart.

Once the hidden assumption is visible, the claim can often be improved. “All A are B” may become “A are usually B under conditions C and D,” which is less rhetorically dramatic but more accurate. Good reasoning treats this revision as progress rather than defeat.

Counterexamples should survive attempts at repair

After a counterexample is proposed, both sides should ask whether it is genuine or only apparent. Perhaps a measurement was wrong, the example was misclassified or the original claim included a qualification that was overlooked. A robust counterexample remains relevant after these checks.

The discipline works in both directions. Critics should avoid contrived exceptions that exploit wording without challenging the substance, while defenders should not invent ad hoc exclusions merely to protect a favourite rule. The aim is to discover the true scope of the claim.