Matching Evidence to the Claim

A reason has to be about the thing you claimed. Say the soup is too salty. The bowl is blue is not proof — the color of the bowl has nothing to do with salt. Good proof: you taste it, or three people all reach for water. Same subject is not the same as real proof. Ask: does this actually show what I said?

Evidence is relevant when, if it were true, it would make your specific claim more likely. That last word matters: specific. The claim fixes what counts as proof. If you argue that a later school start raised test scores, then evidence has to be about scores — not about attendance, not about how much students enjoyed sleeping in, however true those things are.

The common mistake is to match the topic instead of the claim. A paragraph can be full of true sentences, all about your subject, none of which raises the odds of the one thing you asserted. Before reaching for a fact, say the claim's key word to yourself and ask whether the fact touches it.

The claim decides in advance what would count as proof, so name it precisely first: what exactly does your sentence assert? Relevant evidence bears on that assertion, not merely on its subject. Screen time harms teenagers' sleep is not supported by data showing teens own more phones than they used to — ownership is about how common phones are, not about harm to sleep.

Test each fact with one question: if this were true, would my exact claim be more likely? If yes, it is relevant and you keep it. If the claim would be just as likely either way, the fact is off-target, however interesting — and interesting off-target facts are the ones that quietly pad weak paragraphs.

You say your dog is fast. To show it, you race him and he wins. Saying he is soft and fluffy does not show he is fast. Pick the proof that fits what you said.

Relevance is a relation between a piece of evidence and a claim, not a property of the evidence alone. Evidence E is relevant to claim C when knowing E changes how likely C is — when the probability of C given E differs from the probability of C by itself. Writers routinely match on topic instead: they assemble true, on-subject sentences, none of which shifts the probability of the proposition the paragraph is meant to defend.

The corrective is unglamorous. Fix the claim's predicate in view — the exact thing being asserted — and audit each sentence against it. A dense, well-sourced paragraph can still be irrelevant end to end if every fact is about the subject and none is about the assertion.

In informal logic, relevance is one of three criteria for premise adequacy, alongside acceptability and sufficiency — Johnson and Blair's RSA triad. A premise can be true, and so acceptable, and still fail relevance if it does no work toward the specific conclusion at issue. Relevance is dialectical rather than merely topical: it is a premise's function in establishing or undermining the conclusion actually in dispute.

The classical fallacies of relevance name the failure precisely. Ignoratio elenchi — missing the point — proves a conclusion adjacent to the one required; the red herring substitutes an easier claim. Each can be built entirely from truths. This is why relevance cannot be read off a fact in isolation: it exists only in the relation between what you offered and what you undertook to show.