False but open
“Candidate X received 51%.”
The count can settle it. A correction has a recognized path into the argument.
The mechanism
An ordinary argument can be wrong. A closed argument changes the rules so being shown wrong counts as being attacked. The small version is a premise slipped in as background — the cart before the horse, the assumptive close. The big version begins when the check comes back adverse and the checker, not the claim, is what changes.
Treat a disputed premise as common ground.
Make contrary evidence inadmissible or its source corrupt.
Make reopening the premise feel like betrayal, stupidity, or complicity.
A precision that matters: people do not literally have to accept a presupposition in order to understand a sentence. They often accommodate it—let it pass so the conversation can continue—and they can explicitly reject it. Closure exploits the ordinary cost of doing so. Nor is every closure device a presupposition: some work by disqualifying referees or by using a different standard for wins and losses.
Feel the small version
Have you stopped exaggerating your résumé?
Both offered answers accommodate a proposition you were never asked to evaluate: that you have exaggerated your résumé. “Yes” places it in the past; “no” places it in the present. The accurate move is to reopen the upstream question: what evidence suggests I exaggerated it?
Exactly. Notice that rejecting the premise takes more words and can sound evasive. That small asymmetry is what larger closure systems exploit.
“Cash or card?” before you have said you are buying is a fallacy Aristotle catalogued: the conclusion — you are buying — used as a premise. The schools called it petitio principii; English has called it begging the question since 1581; the plain name is the cart before the horse. The conclusion is placed where the premise should be, and the question that would have tested it is never asked. Old names follows what became of the English phrase.
The essential distinction
“Candidate X received 51%.”
The count can settle it. A correction has a recognized path into the argument.
“Election errors happen.”
True. It becomes closure only if every adverse result is then classified as an error without a satisfiable test.
“I won; every institution saying otherwise is corrupt.”
The factual claim is false, while the referee rule protects it from correction.
The name this site uses is for the third column, and for the second once the classifying starts: discursive closure — a claim is closed when an adverse check produces not revision but the reclassification of the checker. A false claim held open is only wrong.
The Overton window is moved by years of campaigning, and what was moved can be moved back. Closure does not move the window; it removes the question, and a removed question leaves no vacancy to campaign for. Nobody agitates to restore what nobody remembers asking.
The diagnostic question is not “Do I dislike this claim?” It is: what evidence is allowed to count against it, and who is allowed to present that evidence?
The corollary is the site’s most important limit. A true claim can be used in a closed way, and a false claim can remain open to correction. Rhetorical form therefore cannot decide whether an underlying allegation is true; that has to be established separately, on the record. The 2020 case is built in two layers for exactly this reason, and Objection 8 holds the site to it.
Single devices rarely do this alone. How devices reinforce each other shows the loop in which a fraud claim, a corrupt-referee claim, and a persecution claim each protect the other two.
Worked example
“The only way we’re going to lose this election is if this election is rigged.”
The utterance · The words
“The only way we’re going to lose this election is if this election is rigged.”
The timing matters: this was said on August 17, 2020, before the election could supply evidence either way.
Surface claim · What is asserted
Fraud is the only possible cause of a Trump defeat.
This is not merely a warning that fraud can happen. “Only way” rules out ordinary defeat in advance.
Presupposed frame · What is treated as settled
In a fair election, Trump necessarily wins.
That premise is not established in the sentence. It is the background needed to make the conclusion work.
Hidden warrant · The unstated rule
An outcome that conflicts with the expected victory is evidence that the process was corrupt.
The result is judged against confidence in the candidate, not against evidence about how votes were cast and counted.
Forbidden question · The question made costly
On its own, the sentence is a pre-commitment: it announces the rule but has not yet applied it. It becomes closure when the rule is kept after recounts, audits, and rulings go the other way and the people who ran them are reclassified as part of the fraud. That later record is the subject of the case page; here, the sentence only shows the rule being announced in advance.
Emotional cue · The emotional pressure
Vigilance against dispossession: your victory is about to be stolen.
The emotion may be sincere. Its rhetorical function is to make suspension of judgment feel passive or disloyal.
Permitted / excluded conclusions · The output
A win validates the candidate; a loss validates the fraud story.
Both branches preserve the same conclusion. The election supplies drama, but no longer supplies a test — provided the rule survives contact with the count.
The quick test
The same four questions, answered one at a time with a result you can send: the sixty-second check →
The site’s full catalog uses seven fields so examples can be compared consistently. It is a coding aid. The candidate psychology — why accommodation is cheap and challenge is expensive for a hearer — is separated out on Why it works on people.
| Stage | Name | What to look for |
|---|---|---|
| 1 | The utterance | What is literally said |
| 2 | Surface claim | The reading the speaker can defend |
| 3 | Presupposed frame | What the wording invites you to treat as already settled |
| 4 | Hidden warrant | The unstated rule that makes the conclusion follow |
| 5 | Forbidden question | The question made socially or evidentially costly |
| 6 | Emotional cue | The affect that makes asking feel like betrayal |
| 7 | Permitted / excluded conclusions | The lower-cost conclusion and the alternative made costly |