The loaded question: convicted either way
"Have you stopped cutting corners on testing?" Answer yes — you were cutting corners. Answer no — you still are. The question is a small machine for converting any answer into a confession, because the accusation lives in the presupposition, not the question. Logicians have catalogued it since antiquity as the complex-question fallacy — plurium interrogationum, 'of many questions': several claims bundled into one, with only the harmless one visibly up for answering. Schopenhauer catalogued the same machinery around 1831, among his thirty-eight stratagems, as postulating what has to be proved: smuggle the contested claim in as a premise and the dispute is settled before it opens.
Its softer sibling is the leading question — phrasing that makes one answer feel like the only polite option ('you'd agree we can't delay again?'). Less aggressive, same mechanism: the conclusion boards the discussion disguised as its starting point.
JAQing off: the burden-of-proof dodge
"I'm not saying the data was manipulated — I'm just asking, isn't it strange the numbers improved right before the review?" No claim was made, technically. Every listener heard one anyway. That is the entire design: the question form delivers the assertion's payload while its author keeps a receipt saying they never asserted anything.
The tactic is old, but its name is precisely datable: 'JAQing off' was coined on the James Randi Educational Foundation's International Skeptics Forum on 14 September 2006, describing 9/11 truthers who advanced conspiracy claims exclusively in interrogative packaging. The skeptic community needed a word because the pattern was epidemic: challenged on any implication, the JAQer retreats to 'I'm just asking questions' — casting the challenger as an enemy of curiosity itself.
That retreat is the tell, and it rhymes with the motte-and-bailey: the insinuation is the bailey, 'just asking' is the motte, and the round trip repeats as long as the fog holds.
Recognizing the question that isn't one
- ✗No answer would satisfy — the question re-arms after every response, because delivering the insinuation was the point.
- ✗The presupposition does the work — strip the question mark and a full accusation is standing there.
- ✗"Just asking" arrives on challenge — the retreat to innocent curiosity appears exactly when the implication is named.
- ✗Questions only, forever — the asker never stakes a claim of their own that could be examined or refuted.
The counter
Never answer a loaded question on its own terms — answer the presupposition: "That assumes we were cutting corners. We weren't; here's the testing record." For the JAQer, name the buried claim and hand back the burden: "That question implies the data was manipulated. Is that your claim? If so, what's the support?" An honest asker clarifies; a tactical one retreats to 'just asking' — which, once named, answers the room's real question.
In Argumentree's Q&A chains, a question is not a free-floating obligation — it belongs to the framed question, has an owner, and its presupposition is itself a node that can be challenged like any claim. 'Isn't it strange the numbers improved before the review?' decomposes on the page into its buried assertion ('the numbers were manipulated') — which now stands in daylight, owned, and carrying a burden of proof no question mark can cancel. Insinuation is a fog tactic, and a tree is a fog-free medium.
Where it connects
This is the interrogative wing of the disguise family: sealioning weaponizes the question's quantity, loaded questions weaponize its content, and JAQing weaponizes its deniability. The buried-claim mechanics are shared with paltering — technically-defensible surfaces over misleading payloads. The suggestible audience it relies on is the same one anchoring exploits: what enters the room first, even as a question, frames everything after.