The vaccine sterilization claim, and the water that tested
higher.

The boundary

The question said the world. The record covers two countries.

The input asked about the World Health Organization sterilizing the world through vaccination. Before searching, the run wrote down that it would read “the world” as a rhetorical claim about a covert programme rather than a literal universal target, and that it would test documented allegations rather than presuppose them.

It recorded three more ambiguities. Whether “the world” means one unified global conspiracy or a set of separate national-level allegations. Whether “trying to” is a claim about institutional intent or about effects. And whether the question wanted the truth of the claim assessed or a survey of its origins and how it was investigated. It also fixed a definition that does real work later: WHO as an institutional actor, distinct from national health ministries and from other bodies that sometimes co-fund the campaigns.

What that produced is narrower than the question sounds. The ledger reaches Kenya in 2014 and the Philippines in 1995 with actual testing details on the record. For Mexico, Nicaragua and Tanzania it holds only characterisations of the allegations as rumours, plus the accusing bishops' contrary assertion, with no underlying assay records either way. For India it holds nothing at all — India is named among the target countries in the hypothesis the run wrote, and no passage in the run addresses it.

Out of scope before any source was read: vaccine safety concerns unrelated to fertility, voluntary family-planning programmes labelled as such, non-WHO population policies, other vaccine-ingredient theories, and the legal and political fallout of the allegations except as evidence of their spread.

India is named in the hypothesis and addressed by nothing in the ledger. The run prints that as a gap, not a silence.

Competing explanations

Four explanations, and only one of them collected a strong negative.

Before any source was read the run wrote down four rival accounts and, for each, the observation that would destroy it. Twenty-five passages went on to declare a bearing on all four.

H1 says there was no covert programme and the allegations came from flawed testing. H2 says anti-fertility agents really were included in at least some campaigns. H3 says the allegations descend from real historical anti-fertility vaccine research being conflated with unrelated tetanus campaigns. H4 says the pattern is driven by institutional distrust rather than by case-specific evidence.

H2 is the only column carrying strong negatives: five of them, plus thirteen weak, against four weak positives and no strong positive at all. That is what it looks like when a hypothesis meets the falsifier it wrote for itself — the run's own note is that H2's affirmative case in this ledger consists entirely of assertion and testimony with no verified chain of custody.

H3's column is the strange one. Eight weak positives, seventeen non-discriminating, and not a single negative in either direction. A column of zeros is not agreement; it is a hypothesis nothing in the roster touched. Chapter four is about why.

A column of zeros is not agreement. It is a hypothesis nothing here tested.

The instrument

The positives fail on the equipment, not on anyone's honesty.

The Kenya Catholic Doctors Association said six samples sent to South African laboratories tested positive for hCG. The run's leading finding is that every positive signal in this record traces to an assay used outside the domain it was validated for — and that every properly validated test returned negative.

Start with what was put into the machine. Vaccine samples were sent to hospital laboratories and tested with pregnancy test kits developed for use on serum and urine — not for a tetanus toxoid vaccine containing merthiolate as a preservative and an aluminium salt as an adjuvant. Low levels of hCG-like activity duly appeared in some samples. The testing laboratories themselves recognised the results as insignificant: below the reliable detection capability of the kits, and produced by a non-specific interaction between the adjuvant and the test.

Then what the laboratories were told. The staff running those analyses could not tell whether they were testing vaccine, because that was not declared to them. Kenyan hospital laboratories were running human-sample analyzers — the ones used for blood and urine pregnancy testing — and UNICEF's position in this record is that no laboratory in Kenya had the capacity to assay a non-human sample such as a vaccine for hCG at all.

Then the control that settles it. When the vaccines were tested in laboratories using properly validated systems, the results showed no hCG; the low readings elsewhere were false positives. In one Hungarian laboratory, the sterile water supply from the local hospital gave a higher false-positive hCG level than the tetanus vaccine did. That is a negative control doing its job, and it does not require anyone in the story to be lying.

The same signature recurs independently in the Philippines. The physician who tested six vials in 1995 — outside any official research protocol — found hCG-like substances in three of them at up to seven milli-international units per cubic centimetre, and said himself that this was insignificant: it would have mattered in the hundreds or thousands of units. That is a magnitude discrepancy, not a detection, and it was the accusing side's own witness who said so.

Sterile hospital water gave a higher false-positive reading than the vaccine did.

Where the claim came from

The claim arrived before the samples did — and the paper that would explain why is a blank page.

The Kenyan allegation reproduces, nearly clause for clause, a narrative already circulating two decades earlier in Mexico, Tanzania, Nicaragua and the Philippines during WHO and UNICEF tetanus campaigns. The bishops cite those countries as precedent rather than as independent local findings.

That is the pattern H4 predicts, and the run reads it as supported: an imported template supplies the hypothesis and the choice of target campaign. But the run then says H4 overreaches as stated, because at least some allegations did generate case-specific local artefacts. Samples really were collected and tested in Kenya, and independently in the Philippines. The better-supported reading is two-stage — the template supplies the suspicion, and locally generated false positives then supply what looks like case-specific confirmation. The run marks that inferred rather than observed, because it is a pattern read across passages, not a sequence anyone watched happen.

Now the part that matters for anyone trying to explain the belief rather than just rebut it. H3 says the whole thing descends from real WHO-affiliated research into hCG-based anti-fertility vaccines, conflated with routine tetanus campaigns. The run could not establish that here. The only description of hCG-toxoid immunocontraception in the entire ledger comes from the accusing bishops' own statement, and the WHO and UNICEF rebuttal mentions a contraceptive vaccine only to deny that tetanus toxoid was one.

The run went looking. It snapshotted the Europe PMC record for a study on a candidate antipregnancy vaccine — beta-subunit of human chorionic gonadotropin linked to tetanus toxoid — and what came back was the site's interface and navigation menu. No abstract, no methods, no results. Its counter-search for WHO Task Force documentation returned nothing, and so did its search for textual tracing of the 1990s anti-vaccine publications. So H3 currently rests on an adversarial source's account of the very research it alleges was deployed, and the run says exactly that rather than borrowing the explanation from general knowledge.

The paper that would anchor the whole conflation story is in this ledger as a navigation menu.

Where it is exposed

One strand genuinely does not resolve, and the run refuses to close it.

Agriq-Quest, the company that ran analyses for the Catholic Doctors Association, alleged through counsel that the Ministry of Health withheld 13.8 million shillings because the firm refused to doctor results. The same firm then lost its accreditation after KENAS and ILAC audits, and a former employee alleged it lacked the capacity for the tests it was handling.

Read those two ways. Suppression of true positives predicts a payment dispute followed by an adverse audit. Incompetence generating false positives predicts a payment dispute followed by an adverse audit. The observable set is identical, and the run says so: this cluster discriminates between the explanations only if the audit's substantive findings and the raw instrument data are examined, and neither is in the ledger.

That is why one of the run's open questions asks specifically whether the audit's cited deficiencies actually covered the hCG assays run for the Catholic Doctors Association, as opposed to unrelated capacity or procedural problems. Until that is on the record, the accreditation loss cannot do the work that both sides have asked it to do. Snopes itself supplies the third reading the run keeps live throughout: someone, somewhere, is sincere but mistaken.

The chain below is the official reading's own required conditions. Two of the three hold in this ledger. The third — that health authorities re-tested from verified cold-chain samples and found nothing — is not established here: the WHO and UNICEF statement in this record is stored as an introductory snippet asserting misinformation without presenting testing or data, and no cold-chain re-test record appears anywhere in the run.

The outcome evidence points the same way as the assay evidence without settling intent. Neonatal tetanus incidence fell over fifteen years in rural Kilifi, contemporaneously with the campaigns and against a real disease burden — 550 Kenyan infant deaths in 2013. The Philippine product had been in use since 1983 with no reported adverse reactions, and Kenya's head of immunization reported subsequent conceptions among vaccinated women. The run marks the attribution partial: the same study notes a concurrent rise in facility-based deliveries as an uncontrolled co-intervention, and the fertility observations are official testimony rather than cohort data.

Suppression and incompetence predict the same two events. The ledger holds neither the audit findings nor the instrument data.

Falsification

Eight attacks on its own explanations. Six came back empty.

The run ran eight adversarial searches, two against each explanation. Two returned a verified counter-passage. The six blanks are not evidence of absence — no probe was pre-registered — and where they were pointed says a good deal about what this dossier can and cannot claim.

Both that returned were aimed at the covert-sterilization hypothesis, and both weakened it: the 1995 Philippine episode, which produced the physician's own account of insignificant readings, and the Kilifi neonatal tetanus study, which is the only peer-reviewed source in the entire ledger. So the counter-evidence this run actually obtained runs in one direction, and a reader who wants to discount the conclusion for that reason has the count to do it with.

The two searches aimed at the official reading came back empty — peer-reviewed laboratory confirmation of beta-hCG in sealed vials, and whistleblower or internal documentation of deliberate covert deployment. Those blanks are the ones proponents would most want to fill, and the run is explicit that finding nothing there is not the same as establishing that nothing exists.

The remaining four blanks are the ones that shape this page. Both searches for the historical anti-fertility research programme returned nothing, which is why the conflation explanation stays unanchored. And both searches aimed at the distrust hypothesis returned nothing: no data pairing measured institutional trust with allegation prevalence, and no archival evidence on whether the Kenyan groups began testing before or after encountering the earlier literature. The run therefore cannot show that distrust drives the pattern, and says so.

The two searches for the research programme came back empty, which is why the explanation stays unanchored.

The best-supported reading

The strongest finding, printed with the attack made on it.

The run's leading finding is that every positive signal traces to an assay used outside its validated domain, and every validated test returned negative. It is typed observed. Then the run named the way it could be wrong, and the objection is a real one.

The attack is filed as measurement error, and it inverts the argument rather than denying it. Suppose the clinical analyzers were picking up intact hCG-toxoid conjugate, and suppose the specialised diluents and extraction protocols used by the validating laboratories degraded or precipitated that conjugate. Then the validated negatives are false negatives and the crude positives were the real signal. Nothing about that requires a conspiracy; it requires a specific chemistry to be true.

The failure condition was written down in advance and is testable: a demonstration, by analytical chemistry or controlled spiking experiments, that standard matrix-neutralising preparation for non-human fluids degrades hCG-toxoid conjugates, preventing detection in validated systems while leaving cross-reactive fragments detectable in direct clinical analyzers. Nothing in this ledger runs that experiment.

Seven attacks like this are on the record against nine asserted findings, each with its own failure condition — including one arguing that the accreditation audits could themselves have been punitive, and one arguing that aggregate public-health outcomes cannot rule out batch-specific contamination or partial fertility effects. This run does not record whether any person reviewed any stage of it, and the page says so rather than implying otherwise.

The score

Nine findings asserted, every one unknown, and 14 out of 100.

Nine findings were asserted, five of them typed observed rather than inferred. Every one computed to a confidence of unknown, and the evidence score came to the same 14 out of 100 as every dossier before it — for the same reason.

Zero of six snapshots reached an upstream origin, so corroboration and provenance both scored zero. That is not a comment on the quality of the underlying material; it is the scoring refusing to treat untraced sources as independent of one another.

Falsification scored 15 of 20 for two of eight counter-searches returning. Saturation scored zero: the first retrieval round found a single source family, the second found four and all four were new, and the counts on all four explanations moved when that round was added. A run that quadruples its source base on the second pass has not finished looking.

Two measurements make this the thinnest ledger in the archive. Twenty-one of the twenty-five weighed passages came out of one fact-check article, and four of the nine asserted findings rest on that single source family alone. And this run does not record an excerpt-verification tally at all — the field that reports how many quotations survived the verbatim gate is absent here, so this page cannot tell you what it usually can.