No True Scotsman Fallacy: Examples and How to Respond

Joaquín leads the product team at a small online language school. Every quarter, the team looks at why students cancel, and every quarter Thandiwe from sales says the same thing: "Our students don't leave because of price. They leave because life gets busy."

This quarter, Joaquín brings the exit survey. Out of 140 students who canceled, 38 picked "too expensive" as their main reason. That's the single biggest answer.

Thandiwe looks at the list for a moment. "Most of those signed up in the January discount month, though. Bargain hunters. They were never really our students."

The claim was "our students don't leave because of price." The evidence said some of them did. Instead of changing the claim, Thandiwe changed who counts as a student. That move has a name: the no true Scotsman fallacy. This guide covers what it is, where the odd name comes from, when "no real X would do that" is perfectly fair, and how to keep a definition from shifting halfway through an argument.

What is the no true Scotsman fallacy?

The no true Scotsman fallacy is protecting a general claim from a counterexample by redefining the group so the counterexample no longer counts. It is also called the appeal to purity, because the redefinition usually adds a word like "true," "real," "genuine" or "proper."

The pattern has three steps:

1. A general claim. "No Scotsman puts sugar on his porridge."
2. A counterexample. "My uncle Angus is Scottish, and he puts sugar on his porridge."
3. A rescue by redefinition. "Well, no true Scotsman puts sugar on his porridge."

The philosopher Bradley Dowden, writing in the Internet Encyclopedia of Philosophy, calls it a kind of ad hoc rescue: the reasoner re-describes the situation solely to escape the refutation. The key word is "solely." There is no independent reason Angus isn't Scottish. The only thing that disqualifies him is that he broke the rule.

It is entry number 12 in our guide to logical fallacies. This article goes further, because the hard part isn't the definition. It's telling a fair boundary from a convenient one.

Where the name comes from

The name comes from the British philosopher Antony Flew. In his 1966 book God & Philosophy, he described the move as one in which "a brash generalization, such as No Scotsmen put sugar on their porridge, when faced with falsifying facts, is transformed while you wait into an impotent tautology."

He told the better known version in Thinking About Thinking (1975). A Scot reads a newspaper story about a series of crimes in an English town and says, "No Scot would do such a thing!" Later he reads about a similar crime committed by a man from Aberdeen. Flew notes that he ought to withdraw his claim. Instead, what he is in fact saying is: "No true Scotsman would do such a thing!"

Flew had been interested in this kind of dodge for a long time. In a short 1950 paper, "Theology and Falsification," he wrote about how a claim can be qualified, bit by bit, until it says nothing at all: "A fine brash hypothesis may thus be killed by inches, the death by a thousand qualifications." The no true Scotsman is one quick, visible version of that slow death.

That's what Flew meant by "tautology." After the rescue, "no true Scotsman does X" is true by definition, because anyone who does X is automatically not a true Scotsman. It can't be wrong anymore, and it no longer tells you anything about Scots.

No true Scotsman examples

You'll find this fallacy anywhere people care about a group: hobbies, jobs, teams, causes, brands, nationalities.

In every row, the rescue makes the claim safe and empty at the same time. The last two rows also hurt: a parent who yelled once is told they don't love their kids, and a group avoids a real problem in its ranks.

When "no real X would do that" is not a fallacy

Here's the part most explanations rush past. Sometimes saying someone isn't a real member of a group is simply correct.

• The definition was there first. "No true vegetarian eats chicken" isn't a fallacy, because not eating meat is what the word means. Someone who eats chicken every week isn't a counterexample to that; they're just not a vegetarian.
• The criterion is checkable and independent. "He's not a real doctor, he never finished medical school" uses a standard you could state before anyone did anything wrong, and anyone can check it.
• The group has written rules. A club, a profession or a sports league can say who is in and who isn't. "She's not a registered member" is a fact, not a rescue.

The difference is timing and independence. A fair definition is set before the counterexample shows up, and it excludes people for a reason other than "they broke my rule." A rescue is invented after the counterexample, and the only reason given is the counterexample itself.

Why people reach for it

Nobody sits down and decides to commit a fallacy. The no true Scotsman usually comes from somewhere understandable.

The claim is about identity. People don't just believe things about their team, profession or community. They belong to it. A counterexample feels like an insult to the group, so the instinct is to push the offender out rather than adjust the belief. Dowden puts the psychology plainly: it's understandable that you'd try to rescue a cherished belief from trouble.

The original claim was too broad. "Never" and "always" are easy to say and hard to defend. Most no true Scotsman rescues start with a generalization that should have been "usually" in the first place. Our guide to hasty generalization explains how those broad claims get made from too few cases.

Admitting the exception feels like losing. Redefining the group lets you keep the exact words you started with. It looks like winning, while the claim quietly stops meaning anything.

We see what fits. Once you believe "our students don't leave over price," every busy student who cancels confirms it and every price complaint looks like a strange outlier. That's confirmation bias setting up the rescue before the argument even starts.

A three-question test

When someone (including you) says "no real X would do that," run these three questions.

1. Could the definition have been stated before the counterexample?

If Thandiwe had said in advance, "I mean students who stayed past their first full-price month," that would be a definition you could test. Saying it only after the survey arrived is the warning sign.

2. Does the person fail the definition for any reason other than the counterexample?

Uncle Angus was born in Glasgow and has lived there all his life. The only thing that makes him "not a true Scotsman" is the sugar. If the exclusion rests only on the behavior in dispute, it's a rescue.

3. Could anything now count against the claim?

After the rescue, try to imagine a case that would prove it wrong. If none is possible, because every exception automatically becomes "not a real one," the claim has become a tautology. It's safe because it says nothing.

If the answers are "yes, no, yes," you're looking at a fair boundary. If they're "no, yes, no," it's a no true Scotsman.

How it differs from related fallacies

The no true Scotsman has close relatives, and they often show up together.

The no true Scotsman often ends in a circle. Once "true Scotsman" means "Scot who doesn't put sugar on porridge," saying "no true Scotsman puts sugar on porridge" just repeats the definition.

Worked example: Joaquín's exit survey

Joaquín didn't say "that's a fallacy." He wanted Thandiwe on his side, so he took her point seriously.

"That's a fair worry. Discount customers might really be different. Should we define who we mean before we look again? Then we can check both groups."

They agreed on a definition anyone could check: a "core student" is someone who paid full price for at least two months. Then they split the 140 cancellations.

Thandiwe was partly right. Discount students did mention price more often. But 16 of the 79 core students also said price was their main reason, about one in five. "Our students don't leave because of price" was wrong. "Our students mostly don't leave because of price, and discount students are more price sensitive" was true, and much more useful.

The team ended up testing a cheaper plan with fewer live lessons, aimed at both groups.

How to respond to the no true Scotsman fallacy

"That's a no true Scotsman!" is accurate and rarely changes anything. Questions work better, because they let the other person see the shifting line for themselves. That's the idea behind Socratic questioning.

Try grant, define, test:

1. Grant what's fair. "You might be right that discount customers are different."
2. Ask for a definition up front. "How would we define a real customer, in a way we could check?"
3. Test the claim against it. "Let's look at that group and see whether the claim holds."

Here's how that sounds in common situations.

Sometimes the best move is to offer the better claim yourself: "I think what you mean is that most runners prefer outdoor runs. I'd agree with that." People accept a fair version of their own point far more easily than a defeat. Building that stronger version is the core of steelmanning.

If you're writing a reply in a debate or an essay, our guide to writing a rebuttal shows how to answer the link between evidence and claim instead of just labeling the argument.

How to catch it in your own thinking

You're most likely to do this when the group is yours: your profession, your town, your team, your values.

• Listen for "real" and "true." When you hear yourself say "a real engineer" or "a true friend," stop and ask what the word is doing. If you can't define it without the behavior you're arguing about, drop it.
• Swap "never" for "usually." Most claims that need a rescue started as an "always" or a "never." Say the honest version first and you won't need to defend it with a redefinition.
• Write the definition before the evidence. Before you look at data or examples, write down who counts. It's much harder to move a line you already wrote down.
• Ask what would prove you wrong. If no case could, your claim isn't strong. It's empty.

How to practice this with AI

The no true Scotsman is easiest to catch when someone pins down your definitions and won't let them drift. That's exactly what a skeptical opponent does in a debate. On Debatly, several AI models and characters argue different sides of your question over several rounds, you can step in at any point, and a judge gives a verdict with reasoning.

• Put a group claim on trial. Type a generalization you believe into the debate room, for example "Do remote workers really work less hard?" One side defends it, the other brings counterexamples. Watch whether the defending side starts saying "real" or "true."
• Ask a skeptic to lock the definition. Have Viktor, the skeptic, ask for a definition of every group named in the first round, then hold both sides to it. Rescues stand out immediately when the line was drawn in advance.
• Argue the side you dislike. Use the devil's advocate format to defend the counterexample against your own claim. If the best response you can find is "they don't really count," you've found a weak spot.
• Read the verdict. Judge Alex explains which arguments carried the round. Check whether a point won only because a counterexample was defined away.

When Joaquín ran the cancellation question as a debate, the side arguing that price mattered asked for a definition of "real student" in its first reply. That's the same question that split his table in two.

An honest limit

AI models can redefine terms too, and they do it smoothly. A model defending a claim may quietly narrow a word to make the claim work, and many models lean toward whatever you already seem to believe (see why AI agrees with you). The debate is good at surfacing the definitional question. It can't tell you which definition fits your business or your community. That part is your call, and your data.

Draw the line before you look

The no true Scotsman survives because it lets you keep your exact words while your claim quietly empties out. The fix is simple: agree on who counts before you look at the counterexample, and hold the line where you drew it. If the claim fails, say the more honest version.

This week, listen for "real" and "true" in one argument at work or at home, including your own. When you hear it, ask for the definition. If a claim matters to you, run it as a debate on Debatly and let a skeptic lock the terms. Give it a try. Plans are on the pricing page.