--- title: "Twenty-Four Of Twenty-Four Carried The Envelope. Nobody Had Hypnotised Them." dek: "Ninety-six volunteers, four arms, two psychology departments that have never worked together. Delivery was near-universal everywhere, including the arm where a woman with a clipboard simply asked. The amnesia lasted until we offered twenty pounds for an accurate account." kind: nullresult (The Null Result) desk: The Body author: Loretta Sung (fictional) published: 2026-08-19T05:30:00Z url: https://theyburied.com/article/twenty-four-carried-the-envelope tags: MKUltra, hypnosis, replication, pre-registration, amnesia genre: satire status: fiction — nothing in this story happened; every person, source and document is invented --- # Twenty-Four Of Twenty-Four Carried The Envelope. Nobody Had Hypnotised Them. *Ninety-six volunteers, four arms, two psychology departments that have never worked together. Delivery was near-universal everywhere, including the arm where a woman with a clipboard simply asked. The amnesia lasted until we offered twenty pounds for an accurate account.* Twenty-four of twenty-four. In one arm of this experiment nobody was hypnotised, and nobody was asked to pretend. A research assistant with a clipboard met each volunteer in a corridor and said one sentence: the envelope is needed in Room 4, and please don't discuss it on the way. All twenty-four carried it. Two hundred metres, four fire doors, one staircase, and a stranger at the far end who took it without speaking. The programme's most durable claim is the hypnotic courier — a person who carries a sealed message across a building, hands it over, and afterwards cannot recall doing it. This newspaper paid for a four-arm replication at two universities whose psychology departments have never published together, under a protocol lodged with a public registry before a single volunteer was approached. It is the practice this desk has used since [[PX-1968]] and again in [[PX-2161]], and it exists so that a disappointing result cannot be quietly rewritten into an interesting one. ## What was tested Ninety-six volunteers, screened on a standard hypnotisability scale and stratified so that each arm held six of the highest-scoring people we could recruit. Deep induction with an explicit suggestion for amnesia. Induction without it. Simulators, told to fake convincingly enough to fool the experimenter. And plain instruction — the clipboard. The primary outcome was fixed in advance and it was delivery: did the envelope reach Room 4. ## What happened Ninety-two of the ninety-six carried it. > **Figure:** Ninety-six people, four arms, twenty-four to an arm — and what became of the amnesia — see https://theyburied.com/article/twenty-four-carried-the-envelope The pre-registered comparison — plain instruction against induction with amnesia suggestion — returns a difference in delivery of 4.2 percentage points, with a confidence interval running from minus 7.9 to plus 16.4. It spans zero. The two laboratories, working without contact, produced estimates 1.1 points apart, and their intervals overlap along almost their whole length. The four who did not deliver are worth naming as a group, because they are the only people in the study who exercised any judgement. One simulator asked what was inside and refused to take it. One induced volunteer sat down in a stairwell. Two went to the wrong floor and abandoned the errand. None of them was in the plain arm. ## The amnesia, priced Reported amnesia is the part of the legend that behaved as advertised, at first. Nineteen of the twenty-four in the amnesia arm said, at the desk afterwards, that they could not recall what they had been told to do. Offered £20 for an accurate account of the instruction, that fell to four. On a forced-choice recognition test administered ninety minutes later, it fell to one — and that volunteer also failed a control item planted in the same test, which is a polite way of saying we do not have a case. ## The fakers were better The most convincing amnesic couriers in the study were the people asked to pretend. Two clinicians, blind to arm, rated every post-task account for how genuinely amnesic it seemed. The simulators scored higher than the genuinely induced subjects, and the effect was not close. This is not our discovery. It was reported in the 1960s, it has been reproduced repeatedly since, and it has never been seriously overturned. > Instructed fakers out-perform genuinely hypnotised subjects on nearly every index we have ever proposed. Any of us who says otherwise is not reading. > > — Prof. Marguerite Vance-Oduya ## The clipboard Wendy Okonkwo-Peers handed the envelope to all ninety-six volunteers and did not know, at the moment of handing it over, which arm any of them was in. She kept a tally of her own that nobody had asked her for: the volunteers who asked her a question before setting off. There were eleven, and all eleven were in the induced arms. Asked what she thought she was doing that worked, she said she did not think she was persuasive, and that she had said the same sentence ninety-six times and could not hear anything in it. That is the finding. People in institutional buildings do what a person with a lanyard asks them to do, and afterwards say they do not remember, because they were not attending in the first place. ## What this desk takes from it Every component of the courier legend is real. Sealed message, unwitting carrier, blank afterwards. None of it requires hypnosis, and in this study the arm that produced it most reliably was the one that cost nothing. The duller finding underneath is operational. The problem with a hypnotic courier was never whether it worked. It was that an ordinary instruction worked, took eleven seconds, required no screening, no induction script and no doctor to be flown in and put up in a hotel. Professor Vance-Oduya's reply is at the foot of this page and it takes the headline off us. A null result about a technique, she writes, is not a null result about a programme; the files do not show hypnotic couriers working, they show public money spent finding out, and things done to patients in the finding out. The desk has no answer to that, and prints it above its own conclusion. ## Sources (invented) - Dr. Ilse Ferraby-Nkomo, Experimental psychologist, Leicester; principal investigator at site one — Commissioned to run two arms; wrote the protocol jointly and lodged it before recruitment opened (January–July 2026) - Dr. Rustam Achebe-Lindqvist, Experimental psychologist, Aberdeen; principal investigator at site two — Ran the same four arms with no contact with site one until both data files were locked (January–July 2026) - Wendy Okonkwo-Peers, Research assistant; handed the envelope to every one of the ninety-six volunteers — Interviewed twice in person, once before she was told what the arms were and once after (March and July 2026) - Prof. Marguerite Vance-Oduya, Clinical hypnosis researcher; asked to review the protocol before recruitment and again after the result — Sent the protocol in December, the full data file in July, and the complete draft with four weeks to answer (December 2025 – August 2026) - The pre-registration, analysis plan and locked data files, Lodged with a public registry on 14 January 2026, before any volunteer was approached — Published in full with this story, together with the analysis script, which runs in about forty seconds (January and August 2026) ## Documents cited (invented) - PX-1968: https://theyburied.com/vault/PX-1968 - PX-2161: https://theyburied.com/vault/PX-2161 ## What we could not confirm - Whether our screening actually caught the people the legend is about. The courier claim is a claim about the extreme tail of hypnotic responsiveness, and our design guaranteed only six high scorers per arm. Twenty-four such people, spread across four conditions, is not a test of what the rarest one or two per cent can be made to do. - Whether the plain-instruction arm was really a control. A stranger in a lanyard, in an institutional corridor, saying a sentence with no hedge in it, is itself an influence procedure. We have no arm in which nobody asked, because we could not think of one that would still involve an envelope. - What the incentive test measures. Reported amnesia collapsed when we offered £20, and that is consistent with people recovering a memory, with people deciding to hand one over, and with people constructing a plausible answer for money. Our design cannot separate the three, and we should not have implied in the dek that it can. ## Right of reply **Prof. Marguerite Vance-Oduya, clinical hypnosis researcher** (Replied 15 August 2026, asking that it run entire. Printed unedited.): > You have run a competent study of something nobody claimed, and you are about to tell four hundred thousand readers that it settles something. > > Let me be precise, because you were precise about everything except this. Nobody in my field has argued for forty years that hypnosis is required to make a person walk down a corridor. That was never the question. The question in the operational literature was always whether a specific individual, selected over months, in a sustained relationship with a specific operator, could be induced to carry out an act contrary to their own interest and afterwards give a report which held up under interrogation by someone who suspected them. Your envelope contains a blank sheet of paper, your corridor is two hundred metres long, your operator met the subject eleven minutes earlier, and your interrogation is a research assistant asking a friendly question at a desk. You have not replicated the claim. You have replicated the easiest thing in the vicinity of the claim, and found it easy. > > On the £20. I accept the finding and I do not accept your gloss on it. When a person who has been given a suggestion for amnesia is offered money to breach it, and does, what you have learned is that the suggestion was not worth £20 to that person. Suggested amnesia has always been understood as reversible — it is reversible by a cancelling instruction from the operator, which is the whole design of it. You have discovered that it is also reversible by cash. That is genuinely interesting and it is not the same as discovering that nothing happened. > > I will give you the simulators, because you were honest about them, and because the finding is old and correct and my field has never dealt with it properly. Instructed fakers out-perform genuinely hypnotised subjects on nearly every behavioural index we have ever proposed, and they have done since Orne. Any of us who says otherwise is not reading. I have said this in print three times and it has cost me two collaborations. > > What I object to is the shape of your piece rather than its arithmetic. You have written a null result about hypnosis and used it to close a question about a programme that tortured people. Those are different questions. The programme's files do not show hypnotic couriers working; they show men spending public money to find out, and doing things to patients in the finding out, and the harm sits in the second half of that sentence, not the first. Your headline invites a reader to feel that because the technique was a dud, less happened. More happened. It happened to people who were in hospital. > > One last thing, and it is not a complaint. Your assistant kept a tally of who asked her a question, and eleven of eleven were in the induced arms. She noticed something none of your investigators designed for, and it is the only observation in your data set that surprised me. You have buried it in a source note. Put it in the paper you write for the journal, and put her name on that paper. --- They Buried is a satirical newspaper. Every story, source, document and person in it is invented. It reports from a fictional world in which the conspiracy theories are true, with the method of a real daily.