Appeal to Purity




Also Known As: No True Scotsman Fallacy

Description: This fallacy is an attempt to protect a generalization about a group from a counterexample by an unprincipled change to the definition of the group to exclude the counterexample. This is a fallacy because the tactic does not refute the counterexample, but only asserts without support that it does not apply.  The fallacy is also known as the No True Scotsman fallacy thanks to the philosopher Anthony Flew. The fallacy has the following form:


Premise 1: Counterexample E is made against Claim C about group G.

Premise 2: Counterexample E does not apply to any true member of G.

Conclusion: C is true (and E is false).


Like many fallacies, it draws its persuasive power from psychological factors. A someone with a favorable view of the group has a psychological, but not logical, reason to reject the counterexample. Few are willing to believe negative things about groups they like or identify with. In Amthony Flew’s example, a Scotsman refuses to believe a story about the bad behavior of other Scotsmen on the grounds that no true Scotsman would do such things. People can also reject counterexamples on pragmatic grounds, such as when this would provide a political advantage.

The fallacy can also be used in the opposite way to reject positive counterexamples about negative claims. For example, if someone claims that all video games are senselessly violent and rejects counterexamples of non-violent video games, then they are committing this fallacy.

This variation is also fueled by psychological factors, in this case negative ones: a person dislikes the group in question and hence is motivated to reject positive counterexamples against negative claims. This can also be done for pragmatic reasons; for example, a politician might refuse counterexamples that go against their negative rhetoric about a group they are trying to demonize.

Sorting out who or what belongs in a group can be a matter of reasonable debate. For example, when members of religious groups do awful things, the question arises as to whether these people are true members of these groups. For example, the Westboro Baptist Church is infamous for its slogan “God Hates Fags” and its hate speech. Some might contend that they are not true Christians because their beliefs seem counter to mainstream Christianity. Others assert that they are Christians because they claim to be and back up their views with scripture.

Debates over group membership need not be fallacious, so it should not be assumed that every argument rejecting a counterexample must be a fallacy. For example, if someone contends that true Christians do not hate LGBT people and rejects the counterexample of the Westboro Baptist Church by providing reasons why they do not meet a good definition of “Christian”, then this fallacy has not been committed. This is because they have provided reasons to support their claim rather than simply rejecting the counterexample out of hand. Their argument could still fail, but not because it is this fallacy.

Providing a guide to settling such disputes goes far beyond the scope of this work, but this fallacy is not a tool that should be used in rational efforts to address such matters.

While it is an error to dismiss counterexamples out of hand, it is also an error to simply accept that what is claimed about some members of a group applies to all or most members of a group. For example, someone might note that a migrant committed a crime and then assert that most migrants are criminals. As another example, one might assert that most police officers are prone to excessive violence because some have been involved in high profile cases of police violence.  These would be example of the Hasty Generalization fallacy. This is leaping to a conclusion too quickly from a sample that is too small to support it properly.


Defense: The main defense against this fallacy is to consider whether a counterexample is rejected on principled grounds or is rejected without evidence, such as on psychological or pragmatic grounds. One way to try to overcome a psychological bias is to ask what evidence exists to reject the counterexample. If there is no such evidence, then all that would be left are psychological or pragmatic reasons. These have no logical weight.



Example #1

Bill: “Islam is a religion of peace. No Muslim would harm another person.”

Sally: “What about the Muslims who are fighting in Syria and Yemen right now?”

Bill: “They are not true Muslims.”


Example #2

Bill: “Christianity is a religion of peace. No Christian would harm another person.”

Sally: “What about all the Christians that killed each other in the world wars and other conflicts?”

Bill: “They were not true Christians.”


Example #3

Mark: “Republicans are not racists and certainly not white supremacists.”

Hector: “What about those racists and white supremacists who support Republican politicians? What about the Republican politicians who are racist and sexist?”

Mark: “We don’t accept them in our party; we are not racists.”


Example #4

Mark: “Democrats are not sexists; we are all for equal rights and respect women!”

Hector: “So, what about those Democrats who got outed by #MeToo for assaulting women?”

Mark: “They are obviously not real Democrats; no real liberal would do such things!”

Originally appeared on A Philosopher’s Blog Read More