Motte and bailey

The cast-iron pan

Nadia: Never wash the cast-iron pan with soap. Soap ruins it.

Leo: I’ve washed mine with dish soap for years and it’s fine.

Nadia: I’m only saying cast iron needs looking after. Leaving it to soak or scouring it hard can damage the seasoning. You agree with that, right?

Leo: Sure.

A week later, Nadia tells a new roommate: “Don’t ever use soap on the cast iron. Even Leo agrees.”

The clear-cut case. “Soap ruins it” is the bailey. “Cast iron needs looking after” is the motte, and Leo agreed only to that. Nadia’s retreat was fine on its own. The error is the return: she goes back to the bold claim and counts Leo’s agreement to the modest one as support for it. Whether soap harms seasoning is a separate question that neither of them settled.

One word doing two jobs

A workshop flyer says: “Everyone is creative. Anyone can write a publishable novel once they unlock it.” A skeptical attendee asks whether that’s really true of everyone. The instructor replies: “By creative I just mean that everyone solves problems in their own way every day. Surely you don’t deny that.” At the end of the session, she reminds the room that “since everyone is creative, every one of you has a novel in you,” and hands out sign-up forms for the advanced course.

Here the retreat and the return happen through a single word. “Creative” in the flyer means something like “able to produce publishable fiction”; “creative” in the defense means “solves everyday problems”. The second is nearly impossible to deny, and the first doesn’t follow from it. Because the word stays the same, the switch is easy to miss, including for the instructor herself.

A whole method

A popular dog-training book promises on its cover that its method “will make any dog reliably obedient in a week.” When experienced trainers point out that many dogs take months, the author replies that the method is simply “reward the behavior you want, and be consistent,” and asks whether anyone disputes that. Nobody does. The next edition keeps the one-week promise on the cover.

The less obvious form, spread across a whole body of claims and two audiences instead of one conversation. The motte is a core of advice that nearly every trainer accepts; the bailey is the exciting promise that sells the book. Critics are pointed to the core, buyers are shown the promise, and the soundness of the core says nothing about the promise. No single sentence is ambiguous here. What makes it a motte and bailey is that the defense is used only when someone presses, and the bold claim is never withdrawn.

1. Assert B, the bold claim you want.
2. When challenged on B, defend M, a modest claim that is easy to defend and is worded like B (or presented as what B “really” meant).
3. Treat the successful defense of M as settling the challenge, and go back to asserting B.

The argument needs a hidden premise: M and B are the same claim, or M is enough to establish B. That premise is what the ambiguous wording hides, and it’s usually false. If it were true, there would be no need to retreat.

• One ambiguous statement. A sentence that is trivially true on one reading and bold on another. Shackel called these “Troll’s Truisms”: “a mildly ambiguous statement by which an exciting falsehood may trade on a trivial truth”. He described them as the simplest possible case of the motte and bailey.
• Retreat by redefinition. When pressed, the speaker redefines a key word so the claim becomes mundane, then drops the new definition once the pressure is off. Shackel notes that arbitrary redefinition is often called “Humpty Dumptying”, after the character in Lewis Carroll’s Through the Looking-Glass who says a word “means just what I choose it to mean”. Compare Persuasive definition and the redefinition in No true Scotsman.
• A whole body of views. A position with a core of defensible but unexciting claims, surrounded by exciting claims that are only lightly defended. Critics who attack the outer claims are pointed to the core, as in the dog-training example.
• Hedging. Bradley Dowden’s list of fallacies in the Internet Encyclopedia of Philosophy describes a close relative: refining a claim “simply to avoid counterevidence and then act as if your revised claim is the same as the original”.

Where the name comes from. The philosopher Nicholas Shackel coined “Motte and Bailey Doctrine” in “The Vacuity of Postmodernist Methodology” (Metaphilosophy, 2005), a polemical paper naming five rhetorical moves he attributed to postmodernist writers. A motte-and-bailey castle is a medieval fortification: a stone tower on a mound (the motte), surrounded by an area of land (the bailey) enclosed by a ditch or other barrier. As Shackel tells it, the motte was “dark and dank”, not a place anyone wanted to live, and existed only to protect the bailey. Under light attack the defenders held the bailey; under heavy attack they retreated to the motte, and when the attackers gave up, they reoccupied the land. Shackel applied the image to doctrines, which he said can be a single belief or a whole body of beliefs. The name has since been applied well beyond his target (a 2023 commentary in Behavioral and Brain Sciences, for instance, uses it in its title), but it has little treatment in the scholarly literature on fallacies: it isn’t among the fallacies covered in the Stanford Encyclopedia of Philosophy‘s entry on fallacies or in Dowden’s list of 231 fallacy names.

Not the same error: moving the goalposts. In Moving the goalposts, the claim stays put and the standard of evidence changes: each piece of evidence is met with a demand for more. In a motte and bailey, the claim itself changes, then changes back.

Not the same error: the straw man. The retreat often comes with an accusation: “I never said soap ruins it; you’re attacking a straw man.” If the bold claim really was made, answering it isn’t a straw man. The two can look alike from the outside, so the question is what was actually said.

Narrowing a claim under challenge is often exactly the right thing to do. It isn’t a motte and bailey when:

• You narrow and stay narrowed. Clarifying what you meant, or giving up the bold version, and then using only the narrower claim from then on.
• You say you were wrong. Dowden notes that it isn’t hedging if you “explicitly accept the counterevidence, admit that your original claim is incorrect, and then revise it”.
• You fall back openly. “I think we should do all of it, but even if you don’t accept that, you should accept this much” makes clear that the second claim is weaker and doesn’t prove the first. Arguing this way is fair because nothing is hidden.
• You hold both claims and defend each. Believing the modest and the bold claim is fine, as long as the bold one gets its own support instead of borrowing the modest one’s.

The test: after the challenge, which claim does the speaker go on asserting, and is it the one they defended?

Clarifying and sticking with it

Omar: Bread flour makes better baked goods.

Jess: Better cakes too? Mine come out tough with bread flour.

Omar: Fair point. I meant yeast breads with a chewy crumb. For cakes you want a lower-protein flour.

When Omar writes the baking guide for the school fundraiser, it says: “Use bread flour for the yeast breads and cake flour for the cakes.”

Omar retreats from a bold claim to a narrower one, which is the first half of the pattern. But he never goes back. The narrower claim is the one that ends up in the guide. That’s the narrow and stay narrowed condition above.

Proposing the smaller step openly

At a homeowners’ meeting, a board member says: “I think the whole fence along the back lot should be replaced this year. But even if we don’t agree on that, the three rotten posts by the gate have to be replaced now, because the gate won’t latch.”

There is a bold proposal and a modest one, and the modest one is much easier to defend. But the speaker presents them as two separate claims and says so. Agreeing to fix the gate posts commits no one to the whole fence, and nobody is invited to think otherwise. That’s the fall back openly condition.

The bold claim is the one worth making: it’s interesting, useful or flattering to the person making it. The modest claim is the one that survives scrutiny. Wording that can be read either way lets a speaker keep the benefits of both, and it can happen without any intent to mislead, since people often don’t distinguish the two readings of their own claims until someone presses them.

The retreat also changes how the exchange looks. Once the speaker is defending something obviously true, the critic seems pedantic or unreasonable, and the audience tends to remember that the claim was defended without noticing which claim it was. Shackel’s point about whole doctrines is that their survival can depend on this back-and-forth: retreating to the motte whenever a refutation is on the table, and reoccupying the bailey when it isn’t.

Spotting the pattern doesn’t show that the bold claim is false. It shows that the bold claim hasn’t been defended yet.

• Name both claims. “I agree that cast iron needs looking after. Are you also still saying soap ruins it?” Separating them usually does more than accusing anyone of a tactic.
• Agree to the modest claim, and say what it doesn’t cover. That keeps the agreement from being quoted later as support for the bold one.
• Ask for support for the bold claim on its own terms, if it’s the one that matters.
• When you’re the one narrowing, say that you’re narrowing, and stop using the broad version.