---
slug: how-many-ever-lived
passed: 2026-10-14
by: the assay line (research/assay-line/cited-1014b/), Codex fixer and repair worker directed by claude-assaying-boyle-1014; source record by Codex, checked by Codex
checked: 103
confirmed: 81
wrong: 21
unverifiable: 0
observed: 0
stale: 1
---

This was the directed correction pass for how-many-ever-lived against the cited-1014b triage at f5e5f84f34. The fixer re-read each triage item's evidence, checked PRB's published table and the relevant primary texts, recalculated the disputed quantities, and inspected the actual interface handlers before editing. The record contains all 21 triage items, all 81 CONFIRMED claims re-read by checker N2 with N2's methods and qualifications attributed below, and one additional MINOR research-facts correction checked by the repair worker. The nine WRONG items cover eight distinct errors because X3 and X16 share one claim; eleven MINOR items remain listed under the directing decisions. The wrong count includes those eleven MINOR lines and the additional repair finding. This pass does not promote N2's separate OBSERVED or UNVERIFIABLE findings into confirmed facts.

## Claims

- MINOR | X1 [never-majority] "The living have never been a majority of humankind." | git show f5e5f84f34:public/strata/how-many-ever-lived/index.html; https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Read PRB Table 1 and the pinned HTML at f5e5f84f34, lines 220-221 and 342-343. The initial two-person row is explicitly a modelling simplification, not evidence of a historical living majority. The populated historical benchmarks support a much narrower claim than never. Retain MINOR as an unsupported universal extension, rather than assert a historical counterexample from the seed. Tier C: list. This is a separate sentence from the WRONG superlative in X2, so the same-sentence exception does not apply.
- WRONG | X2 [fastest-thinning] "But the ghosts are thinning faster than at any time in history." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | PRB Table 1 and the page formula (everBorn - pop) / pop give annual average declines of 0.660624948 in 1650–1750, 0.301970483 for the page's 1968 estimate through 2022, and 0.151179699 in 2010–2022. These interval averages contradict the displayed absolute-ratio superlative and establish no alternative all-history maximum.
- WRONG | X3 [no-simple-growth-law] "the published endpoints can’t yield those births by any simple growth law" | git show f5e5f84f34:research/how-many-ever-lived/build.mjs | The verifier tests exponential population growth with one birth rate per interval. A fitted counterexample is P(t)=2+1999998*(t/140000)^q, q=1999998/(7856100000/(0.08*140000)-2)-1=1.851292813095431. Its endpoints are 2 and 2000000, and 0.08*140000*(2+1999998/(q+1))=7856100000 births. This disproves the universal impossibility without recovering PRB's method.
- MINOR | X4 [errors-shown] "We show by how much, and don’t pretend otherwise." | git show f5e5f84f34:public/strata/how-many-ever-lived/index.html; git show f5e5f84f34:research/how-many-ever-lived/build.mjs | Read the pinned HTML, lines 250-268 and 473-479, and build.mjs, lines 101-110. The page supplies reconstruction bands and the success threshold; the numerical failing discrepancies are printed in the named verifier, rather than displayed alongside the page's claim. Retain MINOR as imprecise wording about where the checking is shown. Tier C: list. This sentence follows, but is not part of, the sentence containing the WRONG growth-law claim in X3.
- WRONG | X5 [oldest-fifth] "the deepest, oldest fifth is an admitted guess, and every cell is named for what it is." | git show f5e5f84f34:research/how-many-ever-lived/build.mjs; https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | PRB's first two intervals sum to 8993889769 births, or 7.685742% of the 2022 total. The six agrarian intervals contain 95517087185 births, 81.624270%; the four intervals after 1900 contain 12509471619 births, 10.689988%. The approximately 18.4% unreproduced remainder includes early and modern births, not an oldest fifth.
- WRONG | X6 [assumptions-operable] "operate the assumptions and the answer moves." | git show f5e5f84f34:public/strata/how-many-ever-lived/index.html; git show f5e5f84f34:memory/log.d/2026-06-29T0600Z-how-many-ever-lived.md | The HTML handlers select fourteen fixed TABLE rows or reveal twelve fixed birth intervals. Neither start date nor birth rate is adjustable. The author's 2026-06-29 log, final bullet, explicitly describes the sensitivity dial as a future addition. Changing assumptions would change an estimate, but the page offers no such control.
- MINOR | X7 [any-moment-picker] "pick any moment from 190,000 B.C.E. to today and read how many humans had been born and how many were then alive" | git show f5e5f84f34:public/strata/how-many-ever-lived/index.html | Read the unrendered Markdown body at f5e5f84f34:src/content/strata/how-many-ever-lived.md:26-27 and the HTML table and button handlers at lines 341-370 and 420-438. There are fourteen selectable benchmarks, ending in 2022; any moment is loose wording for those discrete choices. Retain MINOR and list under tier C. This sentence is separate from X6's relation note and from every sentence being fixed for a WRONG item.
- MINOR | X8 [all-numbers-transcribed] "Every number is transcribed from PRB Table 1 and re-derived in `research/how-many-ever-lived/build.mjs`" | git show f5e5f84f34:research/how-many-ever-lived/build.mjs; git show f5e5f84f34:public/strata/how-many-ever-lived/index.html | Read the Markdown at lines 33-38, build.mjs at lines 30-31, 45 and 137-150, and the HTML at lines 362-364 and 411-412, all at f5e5f84f34. The 1968 cumulative figure is interpolated, its population is attributed separately to UN WPP, and the reconstruction substitutes 35 for PRB's published 31-38 range. Retain MINOR for the blanket provenance and reconstruction wording under the rule for overclaims about the page's own checking. Tier C: list; no WRONG item requires fixing this sentence.
- STALE | X9 [modification-date] ""dateModified":"2026-06-29"" | git log f5e5f84f34 -- public/strata/how-many-ever-lived/index.html; git show 2e60bcb5dd1374a2e5ef778ddb265bdd9ab3fee9 | The June 29 publication date was initially accurate. Commit 2e60bcb5dd1374a2e5ef778ddb265bdd9ab3fee9 changed this page's og:image:alt on October 1 without moving dateModified. This revision sets it to 2026-10-14 under the explicit JSON-LD exception; the other corrections date the revision.
- WRONG | X10 [benchmark-note-bce50000] "180,000 years gone, ~2 million alive"; "yet already ~8 billion births. Deep prehistory is the table’s softest guess." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | PRB Table 1 starts at 190000 BCE and next lists 50000 BCE: 190000-50000=140000 years. Its population of 2000000 and cumulative births of 7856100002 agree. The distinct body paragraph rounding 182000 years through 8000 BCE to 180000 years is not this error.
- MINOR | X11 [benchmark-note-bce8000] "Farming begins. From here on, the model and PRB agree to a fraction of a percent." | git show f5e5f84f34:research/how-many-ever-lived/build.mjs | Read the pinned HTML benchmark note at line 347 and build.mjs at lines 97-129. The successful reconstruction band is 8000 BCE to 1900; four later intervals fail. Retain MINOR as an overstatement of the coverage of the page's own checking. Listing the missing ending date is the tier C disposition. This is a separate benchmark note from the WRONG demographic-transition date in X14.
- WRONG | X12 [benchmark-note-1750] "The eve of industry. Three-quarters of all humans ever born are already in the past." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | PRB Table 1 gives 97564499091 births by 1750 and 117020448575 through 2022: 83.373889% of the latter total. Even subtracting the 795000000 alive in 1750 gives 82.694521%, not 75%. The replacement explicitly states births through 2022 rather than the ambiguous already-in-the-past wording.
- MINOR | X13 [benchmark-note-1850] "The first billion alive at once"; "but still barely 1% of everyone who ever lived." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Read PRB's paragraph immediately before Assumptions and Table 1, and the pinned HTML at lines 356-357. PRB places the billion milestone around 1800; 1850 is the first listed benchmark above it, at 1.265 billion. The label is imprecise about first reaching versus first displaying a billion. Retain MINOR and list under tier C; the accompanying roughly one-percent share is not the error. This sentence is not otherwise being fixed for a WRONG item.
- WRONG | X14 [benchmark-note-1950] "The demographic transition begins"; "birth rates start to crash, so one rate per step gets coarse." | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC2781829/fullTextXML | Bongaarts (2009), section 2, second paragraph, places the start in the nineteenth century in what are now economically developed regions, with major birth-rate reductions following early in the twentieth century. The transition was therefore already under way at the selected 1950 benchmark.
- MINOR | X15 [benchmark-note-2000] "Six billion alive"; "the largest living share of humankind there had ever been." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Read PRB Table 1 and the pinned HTML at lines 342-370. The 2000 share is larger than the earlier populated historical percentage cells, but those sparse benchmarks cannot establish an all-history maximum. The initial two-person model seed is an artificial starting condition, not a literal historical refutation. Retain MINOR for the unbounded superlative and list under tier C; this is a separate sentence from X2's rate-of-decline claim.
- WRONG | X16 [bar-bce50000] "7,856,100,000 births (6.7% of all 117 billion). Model-soft: the published endpoints can’t reproduce this by any simple growth law." | git show f5e5f84f34:research/how-many-ever-lived/build.mjs | Same impossibility as X3 in the amber-bar handler. The fitted power curve described in X3 reproduces 7856100000 births at the first published endpoints. The correct count and 6.7% share remain; both amber bars now describe only failure of the tested exponential reconstruction.
- WRONG | X17 [bar-label-2000] "1968–2000: 6,064,994,884 births, 5.2 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | PRB assigns 6064994884 births to 1950–2000. The page's extra interpolated 1968 row has births:null, but the old TABLE predecessor lookup used it as the interval start. Choosing the preceding birth-bearing row, with the seed for the first interval, restores 1950–2000 in the visible and accessible labels.
- MINOR | X18 [nearby-ruler] "A number with no single true value, only a ruler." | git show f5e5f84f34:src/content/strata/incommensurable.md | Read the pinned population-page relation note at src/content/strata/how-many-ever-lived.md:13, its rendered first-sentence caption at index.html:305, and src/content/strata/incommensurable.md:3-5 and 43-54. The full relation note discusses the uncertain population estimate; the truncated caption sits under a page about the fixed value and irrationality of sqrt(2), making its referent misleading. Retain MINOR as ambiguous relationship metadata, rather than attribute a new mathematical assertion to the linked proof. Tier C: list; the sentence is not already being fixed for a WRONG item.
- MINOR | X19 [nearby-estimate] "That page builds a population estimate and shows what it rests on." | git show f5e5f84f34:src/content/strata/cut-only-what-they-could-not-guess.md; git show f5e5f84f34:public/strata/how-many-ever-lived/index.html | Read the pinned HTML at line 306 and src/content/strata/cut-only-what-they-could-not-guess.md:19-20. In the original relation, That page means How Many People Have Ever Lived; reusing the same words on the reverse link makes the pronoun point toward the epitaph page. The source and destination metadata confirm the note's direction was lost. Retain MINOR as relationship metadata/pronoun imprecision and list under tier C. It is not in a sentence already being fixed for a WRONG item.
- MINOR | X20 [nearby-denominator] "This page needs a denominator and that page built it." | git show f5e5f84f34:src/content/strata/everyone-or-no-one.md | Read the pinned HTML at line 307 and src/content/strata/everyone-or-no-one.md:13-14. The original relation says the genealogy page needs the population page's denominator; the reverse-link caption retains this page and that page, swapping their referents. Retain MINOR as relationship metadata/pronoun imprecision and list under tier C. The existing content supports the original direction, and no WRONG item requires editing this sentence.
- MINOR | X21 [verifier-summary] (added) "growth model reproduces 99%+ of births; one deep-prehistory cell flagged soft." | git show f5e5f84f34:research/how-many-ever-lived/build.mjs | Read git show f5e5f84f34:research/how-many-ever-lived/build.mjs:94-129 and 176-177, and independently summed the successful intervals as 95,517,087,185 births, or 81.624270% of the total. The final 99%+ summary and one-soft-cell count overstate what the same code checks: two prehistoric and four modern intervals fail. Retain MINOR under the explicit rule for overclaims about the page's own checking. This contradiction existed within the original file, so it is not STALE. Tier C: list; its final console-log sentence is separate from the WRONG growth-law assertion in X3.
- CONFIRMED | "But nobody counted them." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [not-census] PRB describes an estimate made with assumptions, rather than an enumeration. Its introduction says "part science and part art".
- CONFIRMED | "built from guessed prehistoric birth rates and a chosen day to start calling us human." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [model-assumptions] The assumptions section specifies initial people, population benchmarks and assumed birth rates; Table 1 supplies the chosen starting year "190,000 B.C.E".
- CONFIRMED | "117,020,448,575" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [headline-total] Table 1, 2022: "117,020,448,575". Independently summing births through 2022 and adding the seed 2 gives exactly 117020448575.
- CONFIRMED | "people ever born, by 2022" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [headline-year] The source labels its snapshot "World population in mid-2022" and gives the corresponding cumulative births.
- CONFIRMED | "Population Reference Bureau, 2022." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-short-citation] The fetched article gives the stated title, authors Toshiko Kaneda and Carl Haub, and publication date November 15, 2022.
- CONFIRMED | "part science and part art." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-art-quote] The quoted words occur in the first sentence, in the same description of estimating total births.
- CONFIRMED | "It is an estimate, not a census" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [estimate-not-census] PRB describes an estimate made with assumptions, rather than an enumeration. Its introduction says "part science and part art".
- CONFIRMED | "the crowd of the dead standing behind each of us is thinning." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [ghosts-thinning] Derived from the modern benchmarks: (ever born - alive)/alive is 42.18 in 1950, 17.53 in 2000 and 13.69 in 2022. This is the modern trend, not monotonic over all prehistory.
- CONFIRMED | "In 1968 Arthur C. Clarke opened 2001: A Space Odyssey" | https://www.sovereignflowinc.com/uploads/1/2/5/9/125930484/2001.pdf | Checker N2 re-read: [clarke-book-date] Opened the PDF manually with pdftotext, pages 1-3. Page 2 names Arthur C. Clarke and "Original copyright year: 1968"; the cited sentence opens the Foreword. The PDF is a transcription of the novel, not a publisher facsimile.
- CONFIRMED | "Behind every man now alive stand thirty ghosts, for that is the ratio by which the dead outnumber the living." | https://www.sovereignflowinc.com/uploads/1/2/5/9/125930484/2001.pdf | Checker N2 re-read: [clarke-foreword-quote] Opened the PDF manually. Page 2, Foreword, contains the full quoted sentence, including "thirty ghosts". Whitespace-normalized excerpt comparison succeeds.
- CONFIRMED | "Clarke counted thirty." | https://www.sovereignflowinc.com/uploads/1/2/5/9/125930484/2001.pdf | Checker N2 re-read: [clarke-thirty] Opened the PDF manually. Page 2, Foreword, contains the full quoted sentence, including "thirty ghosts". Whitespace-normalized excerpt comparison succeeds.
- CONFIRMED | "Today there are about fourteen." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [today-fourteen] The surrounding selector explicitly identifies the 2022 endpoint. Table 1 gives births 117020448575 and population 7963500000, yielding 13.6946 dead per living person, approximately fourteen. Read "Today" as that dated endpoint, not a new 2026 census.
- CONFIRMED | "by less every year, because the living population more than doubled in one lifetime while the all-time total grew only a few percent." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [less-every-year] For the stated lifetime comparison, 1968 to 2022: living population rises from the page's 3.551 billion to 7.9635 billion, a factor 2.243; cumulative births rise from 110.0846 to 117.0204 billion, 6.30%. Modern benchmark ratios decline; the table does not measure each individual year.
- CONFIRMED | "Despite 180,000 years of prehistory" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prehistory-duration] The model's pre-agricultural span is approximately 182000 years, 190000 BCE to 8000 BCE. Its cumulative births by 8000 BCE are 8993889771, 7.6857% of the 2022 total; "180,000" is reasonable rounding in this paragraph.
- CONFIRMED | "the deep past is a thin sliver: tiny populations, however long they lasted, make few people." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [small-populations] The model's pre-agricultural span is approximately 182000 years, 190000 BCE to 8000 BCE. Its cumulative births by 8000 BCE are 8993889771, 7.6857% of the 2022 total; "180,000" is reasonable rounding in this paragraph.
- CONFIRMED | "the ten thousand years of farming, high birth rates, and short lives between the first villages and 1900." | https://web.archive.org/web/20180317034411id_/http://www.prb.org/Publications/Articles/2002/HowManyPeopleHaveEverLivedonEarth.aspx | Checker N2 re-read: [agrarian-ten-millennia] The archived PRB article places the dawn of agriculture at about 8000 BC. From that benchmark to 1900 is approximately 9900 years, consistent with ten thousand.
- CONFIRMED | "high birth rates, and short lives" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [agrarian-short-lives] The source describes low life expectancy at birth and a required birth rate of 80 per 1000 in early periods. This supports the broad contrast; it should not be read as the typical adult lifespan.
- CONFIRMED | "by the year 1, almost half of everyone who would ever be born already had been" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [half-before-one] Table 1, 1 CE: "55,019,222,125" ever born. Dividing by the 2022 total gives 47.0168%, consistent with almost half and 55 of 117 billion. The comparison ends in 2022, not the future extinction of humanity.
- CONFIRMED | "55 of the eventual 117 billion." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [fifty-five-billion] Table 1, 1 CE: "55,019,222,125" ever born. Dividing by the 2022 total gives 47.0168%, consistent with almost half and 55 of 117 billion. The comparison ends in 2022, not the future extinction of humanity.
- CONFIRMED | "The 8,000 years between farming and Rome alone hold 46 billion births." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [farming-rome-births] Table 1 assigns "46,025,332,354" births to the interval ending 1 CE, beginning 8000 BCE. That is 39.3310% of the 2022 total. Agriculture is the approximate PRB benchmark, not a claim that Rome was founded in 1 CE.
- CONFIRMED | "The modern billions are real, but they are recent, and they are still a minority of the whole." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [modern-minority] Births after 1900 total 12509471619, 10.6900% of the 2022 cumulative total, a minority.
- CONFIRMED | "The whole table and both tests live in research/how-many-ever-lived/build.mjs" | git show f5e5f84f34:research/how-many-ever-lived/build.mjs | Checker N2 re-read: [both-tests-location] The pinned file contains the column-sum and reconstruction tests, plus a third ghost-ratio calculation. Reader access is a separate finding: the printed research path returns 404, while the unadvertised /checks/ mirror runs.
- CONFIRMED | "every source and quote is in facts.md." | git show f5e5f84f34:research/how-many-ever-lived/facts.md | Checker N2 re-read: [all-sources-location] Read the pinned facts.md: it has named sections for PRB, its 2011 edition, the model assumptions, Clarke, UN WPP and quotation cautions. This confirms the stated location of apparatus, not the correctness of every assertion inside it. Missing public access is recorded separately.
- CONFIRMED | "it is the sum of PRB’s own published births-per-interval, plus the two humans they seed at 190,000 B.C.E." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [seed-plus-intervals] Independent addition of all published births through 2022 plus "2" gives "117,020,448,575" exactly.
- CONFIRMED | "Applying PRB’s recipe (births = birth-rate × population, integrated over each interval)" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-recipe-attribution] The source assumes population sizes, constant growth rates within premodern intervals, and birth rates. Integrating birth rate times population is the ordinary mathematical implementation of that stated recipe.
- CONFIRMED | "no demographic data exist for more than 99% of the span of human existence," | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-no-data-quote] The words and their prehistoric scope agree with PRB. The tool's literal quotation failures arise from capitalization/terminal punctuation and ending the second quotation before "at the time"; neither changes the proposition in its page context. No substantive misquotation finding.
- CONFIRMED | "may underestimate the average population size." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-underestimate-quote] The words and their prehistoric scope agree with PRB. The tool's literal quotation failures arise from capitalization/terminal punctuation and ending the second quotation before "at the time"; neither changes the proposition in its page context. No substantive misquotation finding.
- CONFIRMED | "The myth: “most people who ever lived are alive today”" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [majority-myth-false] PRB's 2022 living share is "6.8" percent. Using Table 1 gives 6.8052% and a dead/living ratio 13.6946, confirming the myth is false. Table 2 differs from Table 1 by 500000 people, immaterial at this rounding.
- CONFIRMED | "Those alive in 2022 are 6.8% of everyone ever born" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [living-share] PRB's 2022 living share is "6.8" percent. Using Table 1 gives 6.8052% and a dead/living ratio 13.6946, confirming the myth is false. Table 2 differs from Table 1 by 500000 people, immaterial at this rounding.
- CONFIRMED | "the dead outnumber the living roughly 14 to 1." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [dead-living-fourteen] PRB's 2022 living share is "6.8" percent. Using Table 1 gives 6.8052% and a dead/living ratio 13.6946, confirming the myth is false. Table 2 differs from Table 1 by 500000 people, immaterial at this rounding.
- CONFIRMED | "somewhere, at some time back in the 1970s, a writer made the statement that 75 percent of the people who had ever been born were alive at that moment. This factoid has had a long shelf life, even though a bit of reflection would show how unlikely it is." | https://www.prb.org/resource/distilled-demographics-how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-rumour-quote] The fetched 2011 PRB resource contains the full quotation and attributes it only to an unnamed writer in the 1970s, exactly the limited attribution used on the page.
- CONFIRMED | "Across the entire table, the share of all-time humans alive at once never exceeds 8.1%" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [maximum-share-projection] Among PRB's printed percentages, "8.1" is the maximum and belongs to "2050", a future row in the 2022 publication. The initial seed row has a blank percentage; do not infer a universal historical upper bound from that blank.
- CONFIRMED | "and that is a projection to 2050, not a record." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [projection-not-record] Among PRB's printed percentages, "8.1" is the maximum and belongs to "2050", a future row in the 2022 publication. The initial seed row has a blank percentage; do not infer a universal historical upper bound from that blank.
- CONFIRMED | "PRB starts the human story at 190,000 B.C.E." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-start-year] PRB distinguishes the Homo genus from modern Homo sapiens, and starts its estimate at "190,000 B.C.E". Adding earlier Homo individuals would add births; this is a scope statement about the model, not verification of the current fossil chronology.
- CONFIRMED | "just two people (a minimalist approach!)" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-two-people-quote] The source explicitly starts with two people and calls the approach a simplification. The page preserves that qualification.
- CONFIRMED | "an admitted simplification, not a claim about Adam and Eve" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [two-people-simplification] The source explicitly starts with two people and calls the approach a simplification. The page preserves that qualification.
- CONFIRMED | "their 2011 edition started at 50,000 B.C. and got ~108 billion instead" | https://web.archive.org/web/20180317034411id_/http://www.prb.org/Publications/Articles/2002/HowManyPeopleHaveEverLivedonEarth.aspx | Checker N2 re-read: [earlier-start-total] The archived 2011 edition starts "50,000 B.C." at population "2" and gives total "107,602,707,791", approximately 108 billion.
- CONFIRMED | "so the start date is a choice worth a few percent." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/; https://web.archive.org/web/20180317034411id_/http://www.prb.org/Publications/Articles/2002/HowManyPeopleHaveEverLivedonEarth.aspx | Checker N2 re-read: [start-date-choice] The 2011 source says an earlier starting date would raise the total. The new 190000-50000 BCE interval contributes 7856100000 births, about 6.7% of the 2022 total. Not all of the 2011-to-2022 difference comes from the start date; the endpoint and modern estimates also change.
- CONFIRMED | "And these are modern Homo sapiens only; count the whole genus and the total climbs." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [modern-sapiens-only] PRB distinguishes the Homo genus from modern Homo sapiens, and starts its estimate at "190,000 B.C.E". Adding earlier Homo individuals would add births; this is a scope statement about the model, not verification of the current fossil chronology.
- CONFIRMED | "“how many continents” also has no single true answer, only a convention you can operate" | https://education.nationalgeographic.org/resource/Continent/ | Checker N2 re-read: [continents-convention] National Geographic explicitly explains that separating Europe and Asia depends on cultural perceptions despite their single landmass. The pinned companion page implements several counting conventions.
- CONFIRMED | "change where you start counting humans (190,000 vs 50,000 B.C.E.) and the total shifts by billions" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/; https://web.archive.org/web/20180317034411id_/http://www.prb.org/Publications/Articles/2002/HowManyPeopleHaveEverLivedonEarth.aspx | Checker N2 re-read: [different-horizons-billions] The 2011 source says an earlier starting date would raise the total. The new 190000-50000 BCE interval contributes 7856100000 births, about 6.7% of the 2022 total. Not all of the 2011-to-2022 difference comes from the start date; the endpoint and modern estimates also change.
- CONFIRMED | "An immersive layer; it renders as its own standalone page at the `href` above" | git show f5e5f84f34:public/strata/how-many-ever-lived/index.html | Checker N2 re-read: [standalone-page] curl returned HTTP 200 for the href. The served page matches the pinned page except for injected analytics scripts.
- CONFIRMED | "its body is not rendered." | git show f5e5f84f34:src/pages/strata/[...slug].astro | Checker N2 re-read: [body-not-rendered] The pinned route filters strata with ".filter((s) => !s.data.draft && !s.data.href)". This entry has an href and serves its separate HTML.
- CONFIRMED | "Kaneda & Haub, *How Many People Have Ever Lived on Earth?*, 2022" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-manuscript-citation] The fetched article gives the stated title, authors Toshiko Kaneda and Carl Haub, and publication date November 15, 2022.
- CONFIRMED | "Full sources and the careful do-not-misquote flags are in `research/how-many-ever-lived/facts.md`." | git show f5e5f84f34:research/how-many-ever-lived/facts.md | Checker N2 re-read: [sources-and-flags] Read the pinned facts.md: it has named sections for PRB, its 2011 edition, the model assumptions, Clarke, UN WPP and quotation cautions. This confirms the stated location of apparatus, not the correctness of every assertion inside it. Missing public access is recorded separately.
- CONFIRMED | "Every figure sourced; the soft cells named." | git show f5e5f84f34:research/how-many-ever-lived/facts.md | Checker N2 re-read: [every-figure-sourced] Read the pinned facts.md: it has named sections for PRB, its 2011 edition, the model assumptions, Clarke, UN WPP and quotation cautions. This confirms the stated location of apparatus, not the correctness of every assertion inside it. Missing public access is recorded separately.
- CONFIRMED | "Toshiko Kaneda and Carl Haub, How Many People Have Ever Lived on Earth?, Population Reference Bureau, 2022-11-15" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [prb-metadata-citation] The fetched article gives the stated title, authors Toshiko Kaneda and Carl Haub, and publication date November 15, 2022.
- CONFIRMED | "PRB seeds the model with just two people"; "an admitted simplification, not a literal claim." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [benchmark-note-bce190000] The source explicitly starts with two people and calls the approach a simplification. The page preserves that qualification.
- CONFIRMED | "Nearly half of everyone who would ever be born already had been"; "55 of the eventual 117 billion." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [benchmark-note-1] Table 1, 1 CE: "55,019,222,125" ever born. Dividing by the 2022 total gives 47.0168%, consistent with almost half and 55 of 117 billion. The comparison ends in 2022, not the future extinction of humanity.
- CONFIRMED | "High birth rates, short lives: the medieval world adds 27 billion in a single step." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [benchmark-note-1200] Table 1 assigns "26,591,343,000" births to the interval ending 1200, approximately 27 billion. Its assumed rate is 60 per 1000, with short life expectancy discussed in the text.
- CONFIRMED | "Plague and famine: population barely grows for 450 years, yet 13 billion are born and die." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [benchmark-note-1650] Population rises from "450,000,000" in 1200 to "500,000,000" in 1650; interval births are "12,782,002,453". Births minus population increase gives 12.732 billion deaths; both round to 13 billion. PRB discusses plague and, elsewhere, famine-driven fluctuations.
- CONFIRMED | "The ghosts keep thinning as the living crowd swells." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [benchmark-note-2010] Table 1 shows living population rising from 6149000000 in 2000 to 6986000000 in 2010, and the calculated dead/living ratio falling from 17.5341 to 15.5085.
- CONFIRMED | "Today: ~14 ghosts each"; "less than half of Clarke’s thirty, in one human lifetime." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [benchmark-note-2022] The surrounding selector explicitly identifies the 2022 endpoint. Table 1 gives births 117020448575 and population 7963500000, yielding 13.6946 dead per living person, approximately fourteen. Read "Today" as that dated endpoint, not a new 2026 census.
- CONFIRMED | "190,000 BCE: 2 people had ever been born, 2 were alive"; "about 0 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-bce190000] Table 1 values and independent arithmetic agree: births 2, living 2, ratio display "0×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "50,000 BCE: 7,856,100,002 people had ever been born, 2,000,000 were alive"; "about 3927 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-bce50000] Table 1 values and independent arithmetic agree: births 7,856,100,002, living 2,000,000, ratio display "3927.1×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "8,000 BCE: 8,993,889,771 people had ever been born, 5,000,000 were alive"; "about 1798 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-bce8000] Table 1 values and independent arithmetic agree: births 8,993,889,771, living 5,000,000, ratio display "1797.8×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1 CE: 55,019,222,125 people had ever been born, 300,000,000 were alive"; "about 182 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1] Table 1 values and independent arithmetic agree: births 55,019,222,125, living 300,000,000, ratio display "182.4×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1200: 81,610,565,125 people had ever been born, 450,000,000 were alive"; "about 180 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1200] Table 1 values and independent arithmetic agree: births 81,610,565,125, living 450,000,000, ratio display "180.4×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1650: 94,392,567,578 people had ever been born, 500,000,000 were alive"; "about 188 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1650] Table 1 values and independent arithmetic agree: births 94,392,567,578, living 500,000,000, ratio display "187.8×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1750: 97,564,499,091 people had ever been born, 795,000,000 were alive"; "about 122 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1750] Table 1 values and independent arithmetic agree: births 97,564,499,091, living 795,000,000, ratio display "121.7×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1850: 101,610,739,100 people had ever been born, 1,265,000,000 were alive"; "about 79 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1850] Table 1 values and independent arithmetic agree: births 101,610,739,100, living 1,265,000,000, ratio display "79.3×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1900: 104,510,976,956 people had ever been born, 1,656,000,000 were alive"; "about 62 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1900] Table 1 values and independent arithmetic agree: births 104,510,976,956, living 1,656,000,000, ratio display "62.1×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "1950: 107,901,175,171 people had ever been born, 2,499,000,000 were alive"; "about 42 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-1950] Table 1 values and independent arithmetic agree: births 107,901,175,171, living 2,499,000,000, ratio display "42.2×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "2000: 113,966,170,055 people had ever been born, 6,149,000,000 were alive"; "about 18 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-2000] Table 1 values and independent arithmetic agree: births 113,966,170,055, living 6,149,000,000, ratio display "17.5×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "2010: 115,330,173,460 people had ever been born, 6,986,000,000 were alive"; "about 16 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-2010] Table 1 values and independent arithmetic agree: births 115,330,173,460, living 6,986,000,000, ratio display "15.5×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "2022: 117,020,448,575 people had ever been born, 7,963,500,000 were alive"; "about 14 dead for every living person." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [readout-2022] Table 1 values and independent arithmetic agree: births 117,020,448,575, living 7,963,500,000, ratio display "13.7×". The initial seed is explicitly an assumption. The rounded announcement also matches actual pinned JavaScript output.
- CONFIRMED | "190k–50k: 7,856,100,000 births, 6.7 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-bce50000] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "50k–8k: 1,137,789,769 births, 1.0 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-bce8000] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "8k–1: 46,025,332,354 births, 39.3 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "1–1200: 26,591,343,000 births, 22.7 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1200] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "1200–1650: 12,782,002,453 births, 10.9 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1650] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "1650–1750: 3,171,931,513 births, 2.7 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1750] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "1750–1850: 4,046,240,009 births, 3.5 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1850] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "1850–1900: 2,900,237,856 births, 2.5 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1900] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "1900–1950: 3,390,198,215 births, 2.9 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-1950] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "2000–2010: 1,364,003,405 births, 1.2 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-2010] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "2010–2022: 1,690,275,115 births, 1.4 percent of all" | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [bar-label-2022] The published interval birth count and its share of 117020448575 agree with Table 1, and the label matches the pinned JavaScript output.
- CONFIRMED | "holds 46 billion people, 39% of everyone who has ever lived." | https://www.prb.org/articles/how-many-people-have-ever-lived-on-earth/ | Checker N2 re-read: [widest-bar] Table 1 assigns "46,025,332,354" births to the interval ending 1 CE, beginning 8000 BCE. That is 39.3310% of the 2022 total. Agriculture is the approximate PRB benchmark, not a claim that Rome was founded in 1 CE.
- CONFIRMED | "Twins in the verification venue: a question the internet answers with one flat number that turns out to be a convention or a model, not a fact." | git show f5e5f84f34:src/content/strata/how-many-continents.md | Checker N2 re-read: [nearby-conventions] The companion metadata describes multiple continent-count conventions; the present page describes a population model. This confirms the subject connection, not an independent census of search-engine wording.
- CONFIRMED | "Both take a flat confident demographic number that the internet repeats without its apparatus, and open the model underneath it." | git show f5e5f84f34:src/content/strata/nobody-died-at-forty.md | Checker N2 re-read: [nearby-demography] The target metadata explicitly contrasts a widely repeated life-expectancy figure with the underlying ONS life table; the present page opens PRB's total-births model. The subject connection is accurate.
- CONFIRMED | "Nothing here may lie about anything real; that rule never bends. Claims are being checked against their sources, layer by layer, and the working is shown on the page. When a check finds something false, the page says what changed and when." | git show f5e5f84f34:assay/README.md | Checker N2 re-read: [shared-help-2] Read the Assay Office policy and its procedures for source checks and dated corrections. The sentence states the site's rule and ongoing checking practice, not that this unaudited page is already free of errors.

- MINOR | R1 [facts-2022-ratio] The research facts describe the 2022 dead-to-living ratio as "~14.7". | research/how-many-ever-lived/facts.md; https://www.prb.org/news/how-many-people-have-ever-lived-on-earth/ | PRB Table 1 gives 117020448575 ever born and 7963500000 alive in 2022. The dead-to-living ratio is (117020448575 - 7963500000) / 7963500000, approximately 13.694600185, rounding to 13.7. The repair corrects this separate research-facts typo reported by the fixer; the page already uses 13.7.

## Findings

- [declined] X1 [MINOR] Tier C: the universal never goes beyond the finite historical benchmarks; the two-person seed is a modelling simplification, not a historical counterexample. Its sentence is separate from X2.
- [fixed] X2 [WRONG] Replaced the historical fastest-thinning superlative in the HTML with the supported decline across PRB's modern benchmarks; added the dated X2 correction in HTML and markdown.
- [fixed] X3 [WRONG] Narrowed the growth-law claim in HTML prose, verifier source, and facts.md to the tested exponential reconstruction; removed the inferred unpublished prehistoric schedule from that explanation. Refreshed this layer's published verifier as a byte-for-byte source copy. Added one dated correction shared with X16.
- [declined] X4 [MINOR] Tier C: the failing percentage discrepancies are printed by the named verifier rather than shown beside the page claim. This imprecise checking-location sentence is separate from X3 and was retained.
- [open note:f3a037] X5 [WRONG] The markdown dek, page and facts.md conclusion are corrected, and the dated correction gives the independently calculated 7.7% and 10.7% shares. Nothing was corrected on 2026-10-18: the copies pass found no remaining copy in how-many-ever-lived itself. The only copy left is in the frozen measured corpus research/every-cell-costs-the-same/corpora/wasteland-own.txt:3961 ("the deepest, oldest fifth is an admitted guess"), which must keep its bytes; it needs an adjacent provenance/correction note in every-cell-costs-the-same linking /strata/how-many-ever-lived/#corrections.
- [open note:f09b32] X6 [WRONG] The frontmatter relation and two matching promises in facts.md are corrected; HTML and markdown carry the dated correction. The gallery copy at public/strata/there-here/index.html:250 and research/lineage-edge-grammar/page.json:1 still promises assumption controls. This other layer needs an adjacent correction through research/lineage-edge-grammar/copy-corrections.mjs, preserving the quoted note. The directing instance must obtain the board note ID.
- [declined] X7 [MINOR] Tier C: the unrendered markdown loosely promises any moment to today, but the interface selects fourteen benchmarks ending in 2022. It is separate from the relation note fixed for X6.
- [declined] X8 [MINOR] Tier C: the blanket transcription/re-derivation wording overlooks the interpolated 1968 births, separately attributed population, and 35 used for the published 31–38 rate range. It is a separate sentence about the page's checking.
- [dated] X9 [STALE] Set JSON-LD dateModified to 2026-10-14. No separate Updated line was added: the user's explicit exception permits the dated Corrected lines for this revision to carry the date.
- [fixed] X10 [WRONG] Changed 180000 to 140000 years in the 50000 BCE benchmark note. Added the dated correction without altering the separate approximately 180000-year prehistory paragraph.
- [declined] X11 [MINOR] Tier C: the successful reconstruction ends at 1900; from here on omits that endpoint. This checking-coverage note is separate from X14.
- [fixed] X12 [WRONG] Changed the 1750 benchmark note from three-quarters already in the past to about 83% of estimated births through 2022 already occurring by 1750. Added the dated correction.
- [declined] X13 [MINOR] Tier C: PRB places the billion milestone by 1800; 1850 is the first displayed benchmark over a billion, not its historical onset. The accompanying approximately one-percent share is not the error.
- [fixed] X14 [WRONG] Changed the 1950 benchmark note to say the transition was already under way and birth rates had begun falling earlier in what are now developed regions. Added the dated correction, with Bongaarts section 2 linked in the correction sources.
- [declined] X15 [MINOR] Tier C: the 2000 percentage exceeds earlier populated cells, but sparse benchmarks establish no all-history maximum. The artificial seed is not a literal historical refutation, and this sentence is separate from X2.
- [fixed] X16 [WRONG] Changed the shared amber-bar text to failure of the tested exponential growth model with one birth rate per interval, preserving its count and share; the same dated line as X3 covers this copy.
- [fixed] X17 [WRONG] Changed the predecessor from TABLE[TABLE.indexOf(r) - 1] to k === 0 ? TABLE[0] : intervals[k - 1]. Both labels now describe 1950–2000 for 6064994884 births. Added the dated correction.
- [declined] X18 [MINOR] Tier C: the full relation discusses uncertainty in the population estimate, while its truncated caption sits under a proof of sqrt(2)'s irrationality. Retained as ambiguous relationship metadata, not a new assertion that sqrt(2) has no fixed value.
- [declined] X19 [MINOR] Tier C: the reverse-link caption retains a pronoun originally referring to the population page, so its direction is misleading. Relationship metadata is separate from every WRONG sentence.
- [declined] X20 [MINOR] Tier C: the original genealogy relation needs the population page's denominator; reusing this page and that page in reverse swaps the referents. Relationship metadata is separate from every WRONG sentence.
- [declined] X21 [MINOR] Tier C: the verifier's separate closing summary says 99%+ and one soft cell, although the same run shows 81.624270% reproduced, two failed prehistoric intervals and four failed modern intervals. This original checking overclaim is MINOR, not STALE, and the directing triage explicitly says to list it.
- [fixed] R1 [MINOR] Changed the 2022 dead-to-living ratio in research/how-many-ever-lived/facts.md from about 14.7 to about 13.7, matching PRB Table 1 and the page formula. No quotation was altered.

## Validation

- `node research/how-many-ever-lived/build.mjs`: exit 0 before and after the changes. The column sum is exact and the six agrarian intervals reconstruct within 0.3%; the separate incorrect closing 99%+ summary remains listed under X21 by direction.
- Fixer validation before repair: `node scripts/check-claims-pass.mjs` exited 1 on 18 other layers' records; the exported `checkRecord` reported no errors for the then-current 102-claim record. Repair has since marked X5 and X6 open and added R1. The current counts are 103 claims: 81 CONFIRMED, 9 WRONG, 12 MINOR, 1 STALE.
- `node /tmp/how-many-ever-lived-assay-1014/ui-check.mjs`: exit 0. Runs the existing nine-check DOM smoke with its unavailable absolute Playwright import redirected in memory to the installed package, then checks all 14 benchmark readouts, all 12 visible and accessible bar labels, the three corrected benchmark notes, both amber messages, all eight correction mirrors, dateModified, published-source equality, the table's cumulative sums, the six successful reconstruction intervals, and the fitted power-curve counterexample. The first temporary harness run failed because JavaScript string replacement collapsed its double-dollar selector; the harness was corrected and the rerun passed. No repository smoke-test file was changed.
- Scoped `git diff --check`: exit 0. A snapshot comparison found no added U+2014 characters in changed lines or the new record.
- Repair validation, 2026-10-14: the layer verifier and the fixer's existing DOM harness both exit 0. All eight dated correction pairs match and retain the finder; six HTML quotation blocks and all existing anchors are unchanged. The claims gate exits 1 for the two required PENDING dispositions here and problems in 20 other layers; the scoped check reports only those two disposition errors, with counts consistent. The em-dash gate, run with explicit base 9306f663bb5bb7f447014082bd7f342ea22e982d to avoid its default Git fetch, exits 1 for four lines in two other layers and none in this layer. Full commands, evidence, and board-note drafts are in research/assay-line/cited-1014b/how-many-ever-lived/repair-report.md.

## Not done

- All eleven listed MINOR findings remain unchanged, including X21's separate verifier summary, as directed. No MINOR correction lines were added. X9 has no separate Updated line under the explicit metadata exception.
- The other layer's relation-data copies at `public/strata/there-here/index.html:250` and `research/lineage-edge-grammar/page.json:1`, and the research-corpus copy at `research/every-cell-costs-the-same/corpora/wasteland-own.txt:3961`, remain open under X6 and X5 respectively. The original author log remains historical evidence. The live source relation and dek were corrected here.
- No shared publisher, placard writer, full-site build, or media render was run. Only this layer's already-published verifier was refreshed from its source. Shared manifests and other layers' generated copies remain for their owners.
- XML and archived HTML fetched by the fixer were read in memory; no bulk downloads were saved.