Circular Reasoning
Supporting a claim with the claim itself, dressed up in different words.
Share: fallacy.is/circular-reasoning
ยท also: fallacy.is/cr, fallacy.is/circular, fallacy.is/begging-the-question, fallacy.is/btq
In plain terms
Circular reasoning proves a claim by assuming the claim. The argument goes around in a loop and arrives back where it started, disguised by enough paraphrase that listeners don't immediately notice the path.
Also known as "begging the question" in its classical sense, though that phrase has drifted in common usage to mean "raising the question," which is the opposite thing. This entry uses the old meaning.
Why it's fallacious
A good argument moves. It takes things the listener already accepts and shows, step by step, why a new thing follows. Circular reasoning skips the move: the premise is the conclusion, only rephrased. If the listener didn't accept the conclusion to start with, they have no reason to accept the premise. The argument gives them no leverage to change their view.
It's possible to write genuinely circular arguments without realizing it, especially when a belief feels so natural that its restatements feel like independent evidence.
Canonical example
"The book is reliable because everything in it is true. And we know everything in it is true because the book says so, and the book is reliable."
The reliability claim is supported by the truth claim, which is supported by the reliability claim. The loop closes at step three. Someone who already agrees the book is reliable will nod through this. Someone who doesn't has been given no reason to change position, only a longer way to hear the original assertion.
Counter-example (not a fallacy)
"All bachelors are unmarried, because 'bachelor' means 'unmarried man.'"
This looks circular but isn't. It's a definition. The premise and conclusion are the same, yes, but the point of the sentence is to clarify the meaning of the word, not to argue that the claim is true in the world. Definitions are allowed to be tautologies. They work by convention, not by evidence.
Mathematical deductions can also feel circular when the conclusion is already latent in the axioms. That's what deduction is. The difference is whether the argument is extracting a non-obvious implication of starting points the listener accepts, or just restating the thing in dispute.
How to respond when you see it
Ask for the next step independent of the conclusion. "Without assuming [the thing we're debating], what would convince a skeptic that it's true?" If every answer routes back through the conclusion, the argument is circular, and the speaker may not have noticed. Often they're as surprised as you are.