No true Scotsman
The no true Scotsman fallacy is the move of rescuing a universal claim from a counterexample by redefining the category so that the counterexample no longer belongs to it. Yes, it is a logical fallacy, sometimes listed as the appeal to purity, and its signature is a word such as true, real or proper inserted into a definition after the awkward case has already been produced.
Grain of Salt is a reference on how claims are tested. The examples here are invented, and they are about clubs, bakeries and inspection rules rather than about any group of people whose membership is contested.
What the no true Scotsman fallacy is
The no true Scotsman fallacy protects a generalization by shrinking its subject. Someone asserts that every member of a group has a property, a member without that property is produced, and instead of the claim being weakened or withdrawn, the group is quietly redrawn to exclude the person who spoiled it. The claim survives untouched, and it survives by having been made unfalsifiable, which is the real damage: after the move, no observation could ever count against it.
What distinguishes it from the other logical fallacies of definition is which end of the sentence gets altered. Equivocation shifts the meaning of a word between two steps of an argument. Circular reasoning assumes what it sets out to prove. The no true Scotsman fallacy edits the subject term itself, after the evidence has arrived, and only ever in the direction that saves the speaker.
The claim in one sentence, and the two parts it always has
Every no true Scotsman has two parts: a universal claim about a group, and a redefinition introduced only after a counterexample appears. The order is everything. The same definition, offered before the counterexample, would be a perfectly ordinary piece of clarification.
The philosopher Antony Flew, who named the pattern in his writing on reasoning in the 1970s in the book Thinking About Thinking, gave the illustration the fallacy is known by. A man reads that someone in Brighton has committed an outrage and says that no Scotsman would do such a thing. The next week he reads of a man in Aberdeen doing something worse. Rather than give up the generalization, he says that no true Scotsman would do such a thing. Nothing has been learned, and the claim now covers only the people who were never going to refute it.
The three forms the move takes
The move takes 3 forms, and the third is the hardest to spot because it never uses the word true at all.
- The purity insertion. An adjective such as true, real, proper or genuine is added to the group name at the exact moment the counterexample lands.
- The retroactive criterion. A new requirement for membership is announced after the fact, phrased as though it had always been understood.
- The definition by outcome. The group is defined by the very property in dispute, so failure becomes proof of non-membership: "a properly run bakery never burns a batch, and that bakery burned a batch, so it was not properly run."
Form three is worth dwelling on because it can be stated in a tone of pure reasonableness. It converts an empirical claim about bakeries into a definition of the word properly, and definitions cannot be refuted by evidence, which is exactly why the speaker has stopped being able to be wrong.
A worked example, and the repair
Here is the raw version, from a cycling club: "Every member of this club rides with lights after dark." A committee member replies that she saw Robert riding home unlit on Tuesday. The fallacious answer arrives instantly: "No real member of this club would ride without lights." Robert has been removed from the club by a sentence, and the original claim, which was about behavior, is now a claim about vocabulary.
| Stage | The fallacious version | The repair |
|---|---|---|
| The claim | Every member rides with lights after dark | Members are required to ride with lights after dark |
| The counterexample | Robert rode unlit on Tuesday | Robert rode unlit on Tuesday |
| The response | No real member would do that | The rule was broken, so either the rule is not being followed or it is not being enforced |
| What the claim becomes | Unfalsifiable and empty | Testable, and now attached to a number that can be counted next month |
The repair has 2 legal routes and no third. Either weaken the claim to match the evidence, from every member to most members or to the rule requires, or keep the claim and state the membership criterion in advance, in writing, so that the exclusion is a rule rather than a reflex.
What it is confused with
4 labels get attached to the no true Scotsman fallacy by mistake, and one of the moves they describe is not a fallacy at all.
- Legitimate clarification narrows a term using a criterion that existed before the dispute. If the club's constitution already says that unlit riding voids membership, invoking it is administration, not a fallacy.
- Moving the goalposts changes the standard of evidence required rather than the definition of the subject. The goalposts move; here the pitch is redrawn.
- Equivocation uses one word in two senses within a single argument. The no true Scotsman fallacy uses one word in one sense and then edits the sense afterwards.
- Non sequitur and ad hominem are the two labels most often reached for when a definition tightens mid-argument, and neither fits. The conclusion here does still follow from the redefined premise, which is precisely the trouble, and nothing whatever is being said about the person who produced the counterexample.
When narrowing a definition is not a fault
Narrowing a category is legitimate whenever the criterion is constitutive rather than convenient. A vegetarian who eats a steak is not a vegetarian, and saying so is not a fallacy: not eating meat is what the word means, not an empirical generalization about people who hold the label. The same applies to any certified category, such as a registered electrician or an organic farm, where a written standard defines membership and an inspector, not the arguer, applies it.
Two conditions keep a narrowing honest. The criterion must predate the counterexample, and it must be able to exclude cases the speaker likes as well as cases the speaker dislikes. A definition that has only ever removed inconvenient members is not a definition. It is a filter with one setting. Confirmation bias is what keeps that filter invisible to the person operating it, since the exceptions that get noticed are the ones that were already unwelcome.
The test to run on the next redefinition
Ask 4 questions the moment a group's definition tightens mid-argument.
- Ask when the criterion was first stated, and whether it can be found anywhere written down before the counterexample appeared.
- Ask what evidence could now falsify the claim, and note the fallacy if the honest answer is none.
- Ask whether the new criterion would also exclude a member the speaker approves of, and watch what happens when you name one.
- Ask whether the claim is meant to be about the world or about the meaning of a word, because it cannot be doing both jobs in the same sentence.
The origin of the name, and how to avoid the move yourself
The name comes from Flew's illustration and has outlived its original context entirely, which is why the fallacy is now discussed far more often than the philosopher who described it. The Scotsman in the example is a placeholder. Any group with a name and a reputation to protect will do, which is also the reason this page keeps its examples to clubs and bakeries: the fallacy is about the shape of the sentence, and picking a contested group as the illustration teaches the shape less well, not better.
Avoiding the move in your own arguments takes one habit: quantify before you generalize. Say most, say the rule requires, say in the twelve cases we checked. A universal claim is the expensive kind, because a single counterexample destroys it, and the temptation to redefine your way out arrives precisely because you made a claim stronger than your evidence. Say less to begin with and you never need the word true.