Kafkatrap

In plain terms

A kafkatrap is an accusation packaged with a rule that makes denial self-incriminating. Agree, and you've confessed. Deny, and your denial is the evidence. Both doors open into the same room.

The name comes from Franz Kafka's The Trial, where Josef K. is prosecuted for a crime that is never specified, and his protests of innocence only deepen the court's certainty. The term was popularized by Eric S. Raymond in 2010.

The pattern is not always deployed cynically. Some accusations really are about things people cannot easily see in themselves, and someone who believes that sincerely will find the denial genuinely unpersuasive. That's what makes the structure tempting rather than obviously absurd.

Why it's fallacious

A claim that treats every possible response as confirmation is not a claim about the world. It's a closed loop. If no observation, no statement, and no piece of evidence could count against it, then nothing counts for it either, because there was never a test to pass.

It also quietly reverses the burden of proof. The accused is handed a task with no completion condition and blamed for failing to finish it.

Canonical example

A: "You're only defending that policy because you're biased against people like me."

B: "I don't think I am. I've thought about it, and I'd hold the same view if the roles were reversed."

A: "That's exactly what someone with that bias would say."

B is now stuck. Denial confirms it. Silence confirms it. Agreement would obviously confirm it. Getting visibly frustrated confirms it twice.

The underlying concern may be real. People are poor witnesses to their own motives, and that's a genuine problem in reasoning. But as A has stated it, there is no version of B that could be innocent, and a charge no one can answer isn't an argument, it's a verdict with a debate attached.

Counter-example (not a fallacy)

A: "You interrupted her three times in that meeting and nobody else once."

B: "I really didn't notice doing that."

A: "I believe you. That's what makes it a habit rather than a decision. The call was recorded, watch the last ten minutes."

Superficially similar: an accusation about behavior the accused didn't perceive, and a denial that doesn't end the matter. It isn't a kafkatrap.

The claim points at something observable, the denial isn't treated as evidence of guilt, and there is a clear way B could turn out to be right. Not-noticing is offered as an explanation of the mechanism, not as proof of the charge.

The line: is there any response, or any piece of evidence, that would count as disconfirming? If nothing could, it's a trap rather than a claim.

How to fix it

If you've been linked here, the thing you're pointing at may be entirely real, and you may be right about it. The problem is the shape of the argument, not the subject. Make the claim testable: name the behavior instead of the interior state, cite the specific instance instead of the disposition, and say in advance what would count as evidence you're mistaken. If your honest answer is that nothing would, then you're not making a case, and the other person's inability to satisfy you isn't evidence of anything. Dropping the "your denial proves it" clause costs you nothing if the underlying observation is sound.

If you're on the receiving end, name the structure calmly rather than protesting harder, since protesting harder is what the argument is designed to feed on. "As you've set it up, there's no answer I could give that you'd accept as a no. Can we talk about something specific I did instead?" Then consider the substance separately. An unfair framing doesn't automatically make the underlying observation wrong.