Critical Thinking

The Motte and Bailey

The term “Motte and Bailey” has become rather popular in recent years to describe a certain style of argument wherein an arguer strategically equivocates between a boring but easily defensible claim and a more radical but indefensible claim. “Motte and Bailey” is a reference to a type of medieval defensive warfare, where an impregnable fortress (the motte) overlooks some desirable but lightly defensible territory (the bailey). The bailey can be defended against light skirmishes. But against a sustained attack, defenders would fall back to the motte and hold out there, raining arrows on the attackers until the attackers give up and retreat, at which point the bailey could be occupied again. Similarly, in argument, one might retreat to a narrow set of defensible commitments if challenged, but then resume saying indefensible things once the skeptic is out of earshot.

This concept was popularized by the SlateStarCodex blog. If you’ve heard the term, it’s probably because of the work done on that blog. But many don’t realize that the term was coined by a philosopher, Nicholas Shackel of Cardiff University, in a paper published in 2005, “The Vacuity of Postmodernist Methodology.” Shackel’s paper has been cited only 41 times since then – not bad, to be sure, but one suspects that the number would be higher if more philosophers were aware of it. It’s great.

Shackel’s goal in the paper is to expose the fallacies on which the arguments for postmodernist positions rely. In the process, he gives name to a number of different fallacious rhetorical strategies. That makes Shackel’s paper an invaluable contribution to our understanding of informal fallacies, and not just a delightful skewering of Foucault, Lyotard, and their fellow travelers. And that’s why I want to talk a bit about the paper.

The first concept Shackel introduces us to is Humpty Dumptying. Humpty Dumptying is when someone redefines a word that is in common usage to mean somthing new. It gets its name from the following passage from Alice Through the Looking Glass:

“I don’t know what you mean by “glory,”” Alice said.

Humpty Dumpty smiled contemptuously. “Of course you don’t—till I tell you. I meant “there’s a nice knock-down argument for you!’”

“But “glory” doesn’t mean “a nice knockdown argument,’” Alice objected.

“When I use a word,” Humpty Dumpty said in rather a scornful tone, “it means just what I choose it to mean—neither more nor less.”

“The question is,” said Alice, “whether you can make words mean different things.”

“The question is,” said Humpty Dumpty, “which is to be master—that’s all.”

Given the conventional nature of language, we can use our words to mean whatever we want. Humpty Dumpty is right about that. But as Alice notes, Humpty Dumptying is a source of confusion. And it enables further mischief.

The mischief that it enables is the construction of Troll’s Truisms. A Troll’s Truism is a claim that uses a Humpty Dumptied term, and which is a truism if we are bearing in mind the new definition of the term in question. But it is an absurdity if the term is taken with its traditional meaning.

For instance, suppose someone wants to arbitrarily redefine the term ‘knowledge’ to mean ‘power.’ That would be an instance of Humpty Dumptying. Having thus redefined the word ‘knowledge,’ one could then go on to assert claims like “Knowledge is structured by social relations of dominance” or, simply, “Knowledge is power.” These are Troll’s Truisms. ‘Knowledge’ in this context just means ‘power,’ so these two claims simply state that power is structured by social relations of dominance and that power is power. These are truisms. But they look like radical critiques of objective epistemology. In this way, Troll’s Truisms can be used to mislead.

By riffing on Troll’s Truisms, we can establish the infamous Motte and Bailey doctrine. String together enough of these Troll’s Truisms, and you’ll have what looks like a critique of objective epistemology. That’s the Bailey: the radical conclusion that a postmodernist might want his audience to arrive at. But, if challenged, the speaker can claim that all of his statements are mere truisms about ‘knowledge,’ i.e. power. Surely you don’t dispute these trivialities? That’s the Motte.

Humpty Dumptying is a stark example of what Shackel calls an Equivocating Fulcrum. Equivocating fulcra are systematic patterns of equivocation on a particular word or phrase. These equivocating fulcra can be used to state a variety of Troll’s Truisms and thereby construct a Motte and Bailey defense of indefensible views.

Humpty Dumptying establishes the equivocating fulcrum by arbitrary redefinition, but subtler means of establishing equivocating fulcra are available. For instance, we could use a word that already has multiple recognized meanings without specifying which of the two meanings that we intend, and allow our talk to shift back and forth between one definition and another depending on our rhetorical needs.

This can be most effective if we’re using a word that has both a descriptive meaning and a normative meaning. Consider JS Mill’s famously bad argument at the beginning of Chapter 4 of Utilitarianism: (1) If we are capable of desiring something, that thing is desirable. (2) Everything that is desirable is good. Therefore, (3) Everything that we are capable of desiring is good. His proof is obviously fallacious, since he explicilty moves from a descriptive premise to a normative conclusion in one step by equivocating on the word ‘desirable.’ But imagine if Mill had never mentioned either goodness or our ability to desire, and instead carried out a whole discourse regarding ‘desirability’ which sometimes suggested the descriptive reading and sometimes suggested the normative reading. That would make ‘desirable’ an equivocating fulcrum. Shackel accuses Lyotard of doing something very much like this using the term ‘legitimation’ in The Postmodern Condition.

Given the amount of mischief that’s advanced in popular discourse that relies on the use and abuse of language, the concepts of Humpty Dumptying, Troll’s Truisms, Equivocating Fulcra, and the Motte and Bailey Doctrine are ones that every philosopher should have in their toolkit. We should be teaching them in undergraduate critical thinking classes. And we should be alert to these tricks when we see them used in contemporary philosophical discourse.

Shackel argues that these tricks are rampant in postmodern discourse. I fear they’ve begun to creep into analytic philosophy as well. I’ll write more on this soon.

+19

4 Comments

  1. JP Thomas

    Great piece, thanks for sharing it with us all. A tiny typo that has some consequence is here:
    For instance, suppose someone wants to arbitrarily redefine the term ‘knowlede’ to mean ‘power.’

    Putting ‘knowlede’ in quotes makes the reader think the misspelling is intentional.

    +10
  2. Troy

    I’m sympathetic to the worry that a lot of postmodern discourse, and a lot of contemporary political rhetoric, employs the Motte and Bailey tactic. But one response I sometimes see to Motte and Bailey accusations (especially in political contexts), that I think is worth considering, is that they wrongly treat one “side” as an undifferentiated mass rather than a group of people with diverse though related views. For example, suppose person 1 believes that all racial disparities are due to structural racism, and person 2 believes that structural racism causes many racial disparities but that there also other causes of racial disparities. Both claim to endorse “critical race theory.” When the latter acknowledges that not all racial disparities are due to racism someone might respond with a Motte and Bailey accusation, when what’s really going on is just that there’s a somewhat vague theoretical framework that person 1 and person 2 are interpreting differently.

    +3
    • Matt Lutz

      This is a good point. The Motte and Bailey accusation only makes sense when leveled against the statements of a certain person. It can’t be used against a group, because different people in the group might just have different beliefs!

      The interesting thing here, I think, is the way that the Motte and Bailey is enabled by redefinition of common terms. Beware the Humpty Dumpty!

      0

Leave a Reply