Motte and Bailey
Defending a bold claim by retreating to a weaker, related claim when challenged, then returning to the bold claim later.
Share: fallacy.is/motte-and-bailey
ยท also: fallacy.is/mb, fallacy.is/mab, fallacy.is/motte
In plain terms
The name comes from medieval fortress design. A motte was a fortified tower on high ground, small and defensible but not very useful. A bailey was the surrounding enclosed area where people actually lived and farmed, large and valuable but hard to defend. In peacetime you lived in the bailey. When enemies came, you retreated into the motte, waited them out, and returned to the bailey when they left.
The rhetorical move works the same way. A speaker maintains two positions. The bailey is the bold, appealing, controversial claim they actually want to advance. The motte is a much weaker, almost-unobjectionable claim. When pushed, they defend the motte. When unchallenged, they advocate for the bailey and act as though defending the motte amounted to defending the bailey.
Why it's fallacious
The two claims are not the same claim. Proving the motte doesn't prove the bailey, and pretending it does relies on the conversation flowing fast enough that nobody notices the swap. The speaker gets credit for the strength of the easy claim and the appeal of the hard one, without ever having to defend the hard one directly.
This is especially common in ideological arguments where a weak, almost tautological version of a position is held alongside a strong, more ambitious version. Critics end up in a frustrating loop: attacking the strong version feels urgent, but the defense keeps retreating to the weak one.
Canonical example
Bailey (when not challenged): "All traditional institutions are fundamentally illegitimate and should be dismantled."
A: "That seems extreme. Dismantling all traditional institutions would include schools, hospitals, courts, families โ are you really advocating for that?"
Motte (when challenged): "I'm just saying traditional institutions sometimes embed biases we should be willing to examine."
A: "Oh, sure. That's uncontroversial."
Bailey (three paragraphs later): "And that's why the entire system needs to be torn down."
The motte ("we should examine biases") is not the same as the bailey ("tear down the system"). The speaker wins a small, easy victory on the motte and then uses the feeling of victory to keep riding the bailey.
Counter-example (not a motte and bailey)
"My broad claim is that exercise improves health. When pushed for specifics, I clarify that the evidence is strongest for cardiovascular benefits, with smaller effects elsewhere. My broader framing and my narrower defense are the same claim at different levels of detail, not two different claims."
This isn't motte-and-bailey. Generalizing and specifying are normal operations in argument. The fallacy requires the two claims to be substantively different โ different scope, different strength, different implications โ such that proving one doesn't prove the other. Scaling a single claim up and down in detail is fine.
How to respond when you see it
Keep both claims on the table at the same time. "When you said [bold version], were you advancing that claim or the weaker [narrow version]? Because they're not the same thing." Force the speaker to pick. If the speaker keeps toggling between the two, say that out loud. Once the pattern is named, the move stops working. The motte was only a motte as long as nobody mentioned the bailey.