Lineage · in the wild
The bug outside politics
One profession teaches the move by name. One clinic measured it. One courtroom priced it at $2.5 billion. None of these speakers is a demagogue — the bug is a property of constructions, available to anyone with something to not-discuss. Politics differs in scale and stakes, not in kind.
Taught by name
Sales does not hide the move; it names it, scripts it, and mandates verbatim delivery. J. Douglas Edwards — billed within the trade as “the father of modern closing,” trainer of roughly 200,000 salespeople a year through the 1950s–70s — stated the rule outright: never ask a closing question that can be answered with yes or no.
“Would you like delivery Monday, or would Wednesday be better for you?”
Two options, both of which are ways of having already bought. The foreclosed question: Will you buy at all?
His student Tom Hopkins (How to Master the Art of Selling, 1980) taught the presupposition stack in one sentence:
“As I see it, Frank, the only decision we have to make today is how soon you’ll start enjoying the increased profits… by the way, are you going to use it in your main plant or your new warehouse?”
The buy decision is restated as settled (“the only decision… is how soon”), then attention is redirected to a trivial question — answering which building ratifies the purchase. Zig Ziglar generalized it from a script to a stance (“the assumptive attitude”: when showing a house, ask “Where would you place the sofa in this room?”); the Mike Ferry real-estate scriptbook — first instruction: “Follow scripts verbatim!” — repeats its close dozens of times: “All we need to do now is simply… sign the contract.” A 2018 car-sales manual is titled, in full, Assumptive Selling, and modern insurance training states the mechanism with no embarrassment: yes/no questions “leave the door open for an easy ‘No.’”
Sources: Edwards recordings (archive.org) · Hopkins 1980 · Ziglar, Secrets of Closing the Sale, 1984 (archive.org scan) · Mike Ferry Organization scriptbook (PDF) ↗ · Stauning, Assumptive Selling, 2018 · Ritter Insurance Marketing ↗
Measured
The bug is not folklore; it has effect sizes.
- One article, “a” → “the.” Witnesses asked “Did you see the broken headlight?” (none existed) falsely said yes at up to 20%, versus 6% for “a broken headlight” — and the definite article suppressed “I don’t know” responses from 47.5% to 15.5%. The paper says why in one line: “the” translates to “There was a broken headlight. Did you happen to see it?” — the existence question is answered for the witness. Loftus & Zanni 1975, Bulletin of the Psychonomic Society · source ↗
- A barn that never existed. One question containing “…when it passed the barn” caused 17.3% of subjects to remember a barn a week later, versus 2.7% of controls. And “How fast were the cars going when they smashed” yielded 40.5 mph estimates versus 34.0 for “hit” — plus triple the false memories of broken glass. Loftus 1975, Cognitive Psychology · Loftus & Palmer 1974, JVLVB · source ↗
- The pediatrician’s opener. In 111 videotaped visits, doctors who opened with the presumptive format — “Well, we have to do some shots” — met far less parental resistance than those who asked “What do you want to do about shots?” (adjusted odds ratio 17.5, with a wide confidence interval — flagged honestly). 74% of providers used the presumptive form. A later randomized trial training doctors in “announcement” style raised HPV vaccination about five percentage points. Opel et al. 2013, Pediatrics · Brewer et al. 2017, Pediatrics (RCT) · source ↗
- Answer a question about yourself first. “Do you consider yourself a helpful person?” before a survey request moved compliance from 29% to 77% — the request is never deliberated; the just-affirmed identity answers it. Same family: adding “even a penny will help” to a charity ask nearly doubled donation rates (28.6% → 50%) by deleting the standard refusal ground. Bolkan & Andersen 2009 (figures as reported in Cialdini, Pre-Suasion 2016) · Cialdini & Schroeder 1976, JPSP · source ↗
- The waiter’s fork. In a field experiment in two restaurants, “Would you like potato salad with that, or would you prefer fries?” significantly outsold not asking — the add-on decision never appears; only the choice between two ways of having made it. Ebster, Wagner & Valis 2006, Journal of Retailing and Consumer Services · source ↗
- A default is closure without a sentence. Organ-donor “consent”: Germany (opt-in) 12%; Austria (opt-out) 99.98%. The Austrian figure does not measure agreement — it measures the fraction of people who never encountered the question. And in a randomized trial with seriously ill patients, pre-checking the comfort-care option on an advance directive moved the choice from 43% to 77% — most patients said the default hadn’t influenced them; the data said otherwise. Johnson & Goldstein 2003, Science · Halpern et al. 2013, Health Affairs · source ↗ · source ↗
Codified
The legal system doesn’t just use the move — it regulates it, which means it understands it. Federal Rule of Evidence 611(c): “Leading questions should not be used on direct examination… Ordinarily, the court should allow leading questions: (1) on cross-examination; and (2) when a party calls a hostile witness.” A leading question lets the lawyer testify in declarative form while the witness may only ratify or contradict. The law licenses this weapon against adversaries and forbids it among friends — a rationing scheme that presupposes exactly the analysis this site makes.
FRE 611(c), Cornell LII ↗ · The impermissible extreme has been the textbook example for centuries: “Have you stopped beating your wife?” descends from a father-beating version recorded by Diogenes Laërtius; Whately’s logic textbook already called it “hackneyed” in 1844. citation trail ↗
Litigated
In 2025, the Federal Trade Commission put a price on the assumptive close. Amazon’s Prime enrollment offered one way to decline: a button reading “No, I don’t want Free Shipping” — a $139-a-year subscription decision collapsed into rejecting a gift. Its cancellation path — internally named the “Iliad Flow” — ran four pages on which Do you still want to cancel? was never the question shown; every screen presupposed staying and asked only which retention option you preferred. A button labeled “Continue to Cancel” performed no cancellation. The case settled for $2.5 billion, and the order bans that decline-button wording by name.
The retail version has a name too — confirmshaming — and receipts: Sears’ decline link read “No thanks, I hate free money”; a first-aid-kit retailer’s read “No, I prefer to bleed to death.”
Sources: FTC v. Amazon (W.D. Wash., settled Sept 2025) · FTC complaint ↗ · deceptive.design, confirmshaming ↗ · FTC Staff Report, Bringing Dark Patterns to Light, 2022
The everyday formulas
Unmeasured but recognizable — run the stages yourself:
HR
“We’ve decided to go in a different direction.”
| Presupposed frame | The decision is a fact of nature with no agent and no reasons. |
|---|---|
| Forbidden question | Who decided, on what grounds — and were they good ones? |
Institutional medicine
“A known complication occurred. The standard of care was met.”
| Presupposed frame | Membership in a named category (“known complication”) settles the question of conduct. |
|---|---|
| Forbidden question | Did this clinician, in this case, do the thing that caused this harm? |
Wellness marketing
“Toxins your doctor won’t tell you about.”
| Presupposed frame | An unfalsifiable referent plus pre-removed referees — closure sells against the medical guild as readily as for it. |
|---|---|
| Forbidden question | Which molecule, at what dose, measured how? |
Deflection to authority
“You should really ask a professional.”
| Presupposed frame | The question is beyond lay adjudication — asserted, not shown. Genuinely warranted referrals can answer the forbidden question; deflections cannot. |
|---|---|
| Forbidden question | Is this actually beyond lay adjudication, or merely inconvenient? |
Academia
“Concerns were raised.”
| Presupposed frame | Anonymous, agentless disqualification — the concern exists; no one owns it. |
|---|---|
| Forbidden question | Whose concerns, stated where, answerable how? |
Earnings calls
“We faced macro headwinds.”
| Presupposed frame | Underperformance is weather — environmental, agentless, unforecastable. |
|---|---|
| Forbidden question | Which decisions, made by whom, underperformed the sector baseline? |
These items feed the trainer’s mixed bank, so the skill generalizes beyond any one speaker. The full engine: seven stages →