---
slug: the-date-that-can-only-say-no
passed: 2026-10-18
by: the assay line (research/assay-line/claude-1018b/), Claude fixer (claude-opus-5-5) directed by claude-assaying-boyle-1014; source record by a Claude agent (claude-sonnet-5-5), checked by a Claude agent (claude-opus-5-5)
checked: 108
confirmed: 90
wrong: 9
unverifiable: 2
observed: 6
stale: 1
---
# Claims pass: The Date That Can Only Say No

This pass re-read the portal against its sources, source first, in the assay line's batch claude-1018b (tier B): a Claude agent wrote the source record (N1-sources.json), a second Claude agent checked 108 claims against it and the members (N2-claude.md), and the fixer re-read every WRONG item at its source before deciding and fixing it (triage.json). Five WRONG items were fixed: the bench called 1852 the OED's first citation (the OED now gives 1829, in the Court Journal; it gave the Dickens 1852 in 2024); the limit case quoted a sentence its member withdrew on 2026-10-14; the member-overlap note misdated what The Anatomy of Error added on 2026-08-02, undercounted the overlap (ten pairings, not six, recomputed over git at c3110c79443 and at the working tree), and said its verifier could not let that count go quietly out of date when its patterns could not read the JSON cards. The verifier (research/the-date-that-can-only-say-no/verify.mjs) now matches "slug": "x", expects the ten pairings, and checks the new quotation in the member; it prints 207 passed, 0 failed, and goes red when the page's tally word or the new quotation is changed. One STALE item (19 + 12 assertions, now 19 + 13) is dated with an Updated line. Four MINOR items are listed, not edited (tier B). No finding was rejected.

## Claims
- CONFIRMED | A lexicographer finds the word boredom in a newspaper from 1829, and Dickens is no longer its inventor. | https://www.dictionary.com/articles/charles-dickens-boredom | [lex-albion] "Boredom can be found as far back as 1829, when it appears in the August 8th issue of The Albion". Context: per Wiktionary (citing OED Online) the Albion text is a reprint of "Autobiography of an Opera Ticket", Court Journal (London), 6 June 1829, and the OED files it under the obsolete sense "The state of being a bore".
- CONFIRMED | The Latin name for it is terminus post quem, the limit after which. | https://en.wikipedia.org/wiki/Terminus_post_quem | [tpq-term] "A terminus post quem ('limit after which', sometimes abbreviated TPQ)".
- CONFIRMED | Cobalt aluminate is documented in 1777; comparing that milestone with an illustrative 1665 setting flags a question, not a proved forgery. | https://www.webexhibits.org/pigments/indiv/history/coblue.html | [cobalt-1777] "In 1777, Gahn and Wenzel found cobalt aluminate during research on cobalt compounds." The flag-not-proof half is the page's own reasoning.
- CONFIRMED | The word boredom sits in a New York weekly in 1829, so Dickens did not coin it in 1852. | https://en.wiktionary.org/wiki/boredom | [albion-newyork] Wiktionary: "republished in The Albion ... or British, Colonial, and Foreign Weekly Gazette ... New York, N.Y. ... 8 August 1829". A reprint of the London Court Journal of 6 June 1829.
- CONFIRMED | Dickens coined the noun boredom in Bleak House (the popular claim) | https://en.wikipedia.org/wiki/Boredom | [dickens-bleak-house-myth] "A popular misconception is that Charles Dickens coined the term "boredom" in his work Bleak House".
- WRONG | the coinage would have to be 1852, the OED's first citation | https://www.oed.com/dictionary/boredom_n?tab=factsheet | [boredom-oed-1852] The OED today: "OED's earliest evidence for boredom is from 1829, in Court Journal"; entry history: "The following sections of this entry have been updated: Sense 1 (2025) Sense 2 (2025)", "last modified in June 2026". It said "from 1852, in the writing of Charles Dickens" in July 2024 (web.archive.org/web/20240709153059/https://www.oed.com/dictionary/boredom_n). Wiktionary recorded the OED's 1829 and 1831 dates on 2025-11-20, so the 1852 label was already out of date when this page shipped (2026-07-28). The record's excerpt ("1852") does not mention the OED at all. Fixed 2026-10-18.
- CONFIRMED | Rousseau's manuscript held the brioche sentence three years before Marie Antoinette reached France, so the line did not begin with her there. | https://en.wikipedia.org/wiki/Confessions_(Rousseau) | [rousseau-1769] The record's excerpt ("finished ... in 1769") supports only one year; the three years rests on "Books I to VI were written between 1765 and 1767" (Wikipedia, Confessions).
- CONFIRMED | the sentence is finished in Rousseau's Confessions manuscript by 1767 at the latest. | https://en.wikipedia.org/wiki/Confessions_(Rousseau) | [rousseau-1767-latest] "Books I to VI were written between 1765 and 1767"; Let them eat cake: "book six ... whose first six books were written in 1765".
- CONFIRMED | she reaches France in 1770, the earliest year it could have begun with her there | https://en.wikipedia.org/wiki/Marie_Antoinette | [ma-1770] "On 14 May 1770 she met her husband at the edge of the forest of Compiègne."
- CONFIRMED | Lemaitre derived the law two years before Hubble. | https://en.wikipedia.org/wiki/Hubble%27s_law | [lemaitre-1927] "In 1927, two years before Hubble published his own article, ... Lemaître was the first to publish research deriving what is now known as Hubble's law."
- CONFIRMED | the discovery would have to begin in 1929 | https://en.wikipedia.org/wiki/Hubble%27s_law | [hubble-1929] Hubble, PNAS 1929, in the article's references.
- CONFIRMED | Slipher published the first galaxy radial velocity in 1913, sixteen years before Hubble | https://en.wikipedia.org/wiki/Hubble%27s_law | [slipher-1913] Slipher (1913) "The Radial Velocity of the Andromeda Nebula", Lowell Observatory Bulletin; "In 1912, Vesto M. Slipher measured the first Doppler shift of a spiral nebula". 1929 minus 1913 is 16.
- CONFIRMED | Vasari praises it in 1550, the earliest year the story could start | https://en.wikipedia.org/wiki/Mona_Lisa | [vasari-1550] "his biography of Leonardo published in 1550".
- CONFIRMED | supremacy over its Louvre rivals is attested only from 1911, the year of the theft | research/a-minor-leonardo/coverage.json | [theft-1911] The record's Wikipedia excerpt supports only the 1911 theft and its role in fame; the member's saved counts carry the claim: Mona Lisa first tops both rivals in 1911 (1183 pages against Venus de Milo 154 and Winged Victory 56; 1910 was 143 against 169 and 65).
- CONFIRMED | the Louvre guidebooks in which two rivals outranked it | https://web.archive.org/web/2024/https://www.smithsonianmag.com/arts-culture/stolen-how-the-mona-lisa-became-the-worlds-most-famous-painting-16406234/ | [baedeker-1907] "in 1907 he had a mere two sentences, much less than the other gems in the museum, such as Nike of Samothrace and Venus de Milo".
- CONFIRMED | Ned Pennant-Rea publishes the conditional extrapolation into hundreds in 2018 | https://publicdomainreview.org/essay/the-dancing-plague-of-1518 | [pennant-2018] "By Ned Pennant-Rea ... Published July 10, 2018 ... The final toll is unknown but, if such a daily death rate was true, could have been into the hundreds."
- CONFIRMED | Waller's 2009 article separately reports a chronicle's claim of about fifteen deaths a day. | https://www.medievalists.net/2014/05/forgotten-plague-making-sense-dancing-mania/ | [waller-2009-fifteen] "One chronicle states that it claimed, for a brief period at least, about 15 lives a day"; Crossref 10.1016/s0140-6736(09)60386-x, Waller, The Lancet 373, 2009.
- CONFIRMED | The contemporary Strasbourg record documents the dancing and mentions no deaths. | https://en.wikipedia.org/wiki/Dancing_plague_of_1518 | [strasbourg-no-deaths] "the sources of the city of Strasbourg at the time of the events did not mention the number of deaths, or even if there were fatalities."
- CONFIRMED | of the beam pulled from a wall in Show Low, Arizona | https://en.wikipedia.org/wiki/Andrew_Ellicott_Douglass | [showlow-hh39] "on June 22, 1929, a beam labeled HH-39 was extracted at the Show Low site in Arizona." "From a wall" is the member's wording ("a charred roof beam pulled out of a wall"), not in this source; not independently checked.
- CONFIRMED | The work is dated 1942; the sale to Göring's representative was in 1943. | https://www.cultureelerfgoed.nl/actueel/weblogs/kunstwerk-van-de-maand/2014/h.a.-van-meegeren-christus-en-de-overspelige-vrouw-1942 | [goering-1943] "Van Meegeren verkocht het werk in 1943 aan een vertegenwoordiger van ... Hermann Göring". Sources differ: English Wikipedia (Han van Meegeren) says "In 1942 ... one of Van Meegeren's agents sold ... Christ with the Adulteress to ... Alois Miedl".
- CONFIRMED | the paint medium tests as phenol-formaldehyde resin, 1907 | https://en.wikipedia.org/wiki/Bakelite | [bakelite-1907] "developed by chemist Leo Baekeland ... in 1907"; Han van Meegeren article: "He came up with a way to use phenol formaldehyde (Bakelite) to make paints harden".
- CONFIRMED | the 1514 date of Dürer's Melencolia I engraving | https://en.wikipedia.org/wiki/Melencolia_I | [melencolia-1514] "Melencolia I is a large 1514 engraving by ... Albrecht Dürer".
- CONFIRMED | Janine Rivière ... writes that segmented sleep "is not discussed in manuals of health or discussions of the proper regimens of sleep." | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC10404514/fullTextXML | [riviere-quote] Boyce 2023 (doi 10.1017/mdh.2023.14): "Janine Rivière notes that 'although A. Roger Ekirch suggests ... segmented sleep, this is not discussed in manuals of health or discussions of the proper regimens of sleep.'"
- CONFIRMED | slepe fyrst on thy syght side ... turne ayen on thi ryght side and ther slepe all nyght forth | https://raw.githubusercontent.com/textcreationpartnership/A01993/master/A01993.xml | [gouernayle-quote] Verbatim with the page's ellipses; the checker's PARTIAL comes from the page break (pb facs tcp:10418:15) between "ryght" and "side".
- MINOR | Gouernayle of Helthe, London, ca. 1490 | https://raw.githubusercontent.com/textcreationpartnership/A01993/master/A01993.xml | [gouernayle-london] TCP imprint: "[Westminster :" "William Caxton," "1490?]". The member corrected its own caption from "London" to "Gouernayle of Helthe, Westminster, ca. 1490" on 2026-10-14 (commit 30dc50b1fa6); the portal still says London. Listed (tier B MINOR), not edited.
- CONFIRMED | That is a real counterexample to the sentence quoted above ... says nothing about an hour of wakefulness. | public/strata/what-the-first-sleep-was-first-of/index.html | [portal-quote-counterexample] Verbatim at 4fcb19ad5e.
- CONFIRMED | That layer swept the Text Creation Partnership corpus, 61,287 hand-keyed books and 1,547,130,448 words printed between 1473 and 1810 | public/strata/what-the-first-sleep-was-first-of/numbers.json | [corpus-size-books] "texts": 61287, "words": 1547130448, "yearMin": 1473, "yearMax": 1810.
- CONFIRMED | the corpus is "a transcribed subset of what survives" | public/strata/what-the-first-sleep-was-first-of/index.html | [limit-subset-quote] Verbatim.
- WRONG | no list of spellings catches a paraphrase, so "every second-sleep count on this page is a floor" | public/strata/what-the-first-sleep-was-first-of/index.html | [limit-floor-quote] The quoted words are not on the member page. They were (329feda618a) until commit 30dc50b1fa6 (2026-10-14) replaced them with: "\"Lye to sleep again\" is not searched. The fixed search can miss other descriptions of returning to sleep, and phrase counts do not count the practice." The portal quotes, as the member's own words, a sentence the member withdrew. Fixed 2026-10-18.
- CONFIRMED | an argument from silence "cannot distinguish a thing nobody did from a thing nobody bothered to mention" | public/strata/what-the-first-sleep-was-first-of/index.html | [limit-silence-quote] Verbatim.
- CONFIRMED | the transcribed books effectively stop in 1800 | public/strata/what-the-first-sleep-was-first-of/numbers.json | [limit-1810] Member: "The harvested transcribed books extend through 1810"; numbers.json byBand: 3,099 texts in the 1775 band, 344 in the 1800 band, so "effectively" is a fair gloss.
- CONFIRMED | no list of spellings catches a paraphrase | public/strata/what-the-first-sleep-was-first-of/index.html | [limit-paraphrase] "they cannot see paraphrase ... no list of spellings will catch that."
- CONFIRMED | "a stretch of prehistory acquired exact years. Not a range. Years." | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-years-quote] Verbatim.
- CONFIRMED | all 48 Arizona cores land on their published year against a Colorado chronology 187 km away that shares not one tree | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-48-48] "Every one of the 48 Arizona cores lands on its published year against the Colorado chronology"; "shares not one tree".
- CONFIRMED | the weakest genuine match still reaching t = 7.21; against an Alaskan control 3,718 km away, zero of 48 land correctly | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-721] Member text matches; the portal verifier checks the figures (mutating 7.21 in the prose turns it red).
- CONFIRMED | 23 of those 48 certainly-wrong Alaskan matches clear the discipline's conventional threshold of t ≥ 3.5 | src/content/strata/every-tree-wrote-the-same-bad-year.md | [tree-23-of-48] Dek: "the discipline's own significance threshold cleared by 23 of those 48 impossible matches".
- CONFIRMED | "what happens when you take the best of several hundred tries and then apply a cutoff designed for one try." | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-best-tries-quote] Verbatim.
- CONFIRMED | best false 5.17 against weakest genuine 7.21 | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-5-17] "The best false match ... is 5.17. The weakest genuine match ... is 7.21."
- CONFIRMED | Bark still on the sample gives the felling year. Sapwood without bark gives a felling range about a decade wide for oak. | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-ladder] Member's ladder table.
- CONFIRMED | "A floor and nothing else: the tree cannot have been felled before this. There is no upper bound at all." | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-floor-quote] Verbatim.
- CONFIRMED | "The bottom row is the one that gets flattened in retelling, because 'dated to 1486' reads the same on the page whether it means a felled year or a floor." | public/strata/every-tree-wrote-the-same-bad-year/index.html | [tree-1486-quote] Verbatim apart from nested quotation marks.
- CONFIRMED | That layer's own heading is "One measurement is not the discovery" | public/strata/who-discovered-the-expanding-universe/index.html | [redshift-heading-quote] Heading present.
- CONFIRMED | which that layer prints as "two years before Hubble" | public/strata/who-discovered-the-expanding-universe/index.html | [redshift-two-years-quote] "in 1927 ... two years before Hubble".
- CONFIRMED | its headline object is blueshifted | public/strata/who-discovered-the-expanding-universe/index.html | [redshift-blueshift] "it's a blueshift, coming toward us at −300 km/s."
- CONFIRMED | it says Slipher read his shifts as ordinary Doppler motion rather than as space stretching | public/strata/who-discovered-the-expanding-universe/index.html | [redshift-doppler] Verbatim sense.
- CONFIRMED | The Redshift Before the Law lays out the chronology and splits the credit three ways without ever stating a rule about what dates can and cannot do. | public/strata/who-discovered-the-expanding-universe/index.html | [redshift-no-rule] "So the credit splits three ways"; a search of the member text for refute, can never, cannot confirm, one-sided, a date can, terminus returns 0 (the same search finds "three ways", the positive control).
- CONFIRMED | is the second of that layer's four honest limits, a caution about newspaper OCR | public/strata/a-minor-leonardo/index.html | [leonardo-ocr-limit] "Four honest limits ... (ii) It is OCR of century-old newsprint".
- CONFIRMED | the coverage instrument its layer built runs from 1905 to 1920 | research/a-minor-leonardo/coverage.json | [leonardo-1905-1920] "window": from 1905 to 1920.
- CONFIRMED | Supremacy over the Louvre's rivals is attested from 1911, the year of the theft, 361 years after Vasari first praised the painting | public/strata/a-minor-leonardo/index.html | [leonardo-361] "in 1550, 361 years before the theft"; 1911 minus 1550 is 361.
- CONFIRMED | that "not famous" would be a lie | public/strata/a-minor-leonardo/index.html | [leonardo-not-famous-lie] "\"Not famous\" would be a lie".
- CONFIRMED | "The dates mix attestations, approximate eras, and the date of an engraving; none establishes a word's birthday." | public/strata/boredom-word/index.html | [boredom-quote] Verbatim.
- CONFIRMED | The boredom layer flags its own 1829 antedating as soft, reported by a dictionary publisher and not inspected at the original | public/strata/boredom-word/index.html | [boredom-soft-1829] "the 1829 Albion antedating is reported by Dictionary.com and not inspected at the original".
- CONFIRMED | its solid floor is Poe's 1844 ... which still refutes the Dickens story by eight years | https://www.eapoe.org/works/tales/thingma.htm | [boredom-poe-1844] "Southern Literary Messenger ... December 1844 ... the boredom which we should cause our readers in so doing"; 1852 minus 1844 is 8.
- CONFIRMED | the boredom floor has already been lowered twice on the record: 1852 to 1844 to 1829 | research/boredom-word/build.mjs | [boredom-lowered-twice] Timeline entries 1852, 1844, 1829.
- CONFIRMED | its build.mjs asserts that span, 4 of 4 | research/boredom-word/build.mjs | [boredom-build-4of4] recreated: node research/boredom-word/build.mjs printed four OK lines, the last "OK  OED first citation (1852) above its lowest antedating (1829) = 23".
- CONFIRMED | "The chronology puts Rousseau's text before her arrival in France; it cannot prove that she could never have repeated an older sentence." | public/strata/let-them-eat-cake/index.html | [cake-quote] Verbatim.
- CONFIRMED | "This comparison does not establish public availability or authenticate the painter." | public/strata/pigment-that-wasnt-born/index.html | [pigment-current-quote] Verbatim.
- CONFIRMED | "No contemporary toll is documented; a later report and a conditional total are not a contemporary count." | public/strata/the-dancing-plague/index.html | [plague-quote] Verbatim at 4fcb19ad5e.
- MINOR | A dictionary's first citation is the earliest quotation its editors happened to be sent, with no denominator at all | https://en.wikipedia.org/wiki/Oxford_English_Dictionary | [oed-readers] The source has readers sending quotations ("receives about 200,000 a year") but also editors searching: "New text search databases offered vastly more material for the editors of the dictionary to work with". The boredom entry itself moved to 1829 in an editorial revision (2025). "Happened to be sent" is narrower than the source; the member says "happened to find". Listed (tier B MINOR), not edited.
- CONFIRMED | Sapwood without bark gives a felling range about a decade wide for oak. | https://www.frontiersin.org/journals/ecology-and-evolution/articles/10.3389/fevo.2022.823622/full | [sapwood-decade] "the felling date of the tree can be estimated with a precision of approximately 10 years for oaks".
- UNVERIFIABLE | It is one of the oldest tools in scholarship and one of the least examined | tried Wikipedia Terminus post quem; no source measures how old or how seldom examined the method is | [tpq-oldest-tools] An unsourced judgement.
- UNVERIFIABLE | There is a move that historians, chemists, lexicographers and astronomers all make, and each of them thinks of it as belonging to their own field. | tried Wikipedia Terminus post quem and Dendrochronology | [everyone-thinks-theirs] An unsourced generalisation (the dek lists conservators instead of chemists).
- CONFIRMED | A proved anachronism in original paint rules out a date only after later restoration and contamination are excluded. | public/strata/pigment-that-wasnt-born/index.html | [pigment-original-paint] "A proved \"no\" is fatal to the claimed date, once later additions and contamination are excluded."
- CONFIRMED | The portal verifier runs the identical sweep offline and asserts the confirmed count is zero. | research/the-date-that-can-only-say-no/verify.mjs | [verifier-sweep-offline] recreated: lines "ok  B. boredom: 197 refuted, 0 flagged, 430 silent, 0 confirmed over 627 years" and five more.
- CONFIRMED | sweep every year from 1400 to 2026 across all six cases, and the third lamp never lights. | research/the-date-that-can-only-say-no/verify.mjs | [verifier-3762] recreated: "ok  B. 3,762 settings swept   6 cases x 627 years = 3762".
- CONFIRMED | The bench's sweep runs every integer year from 1400 to 2026 inclusive against all six collision cases | public/strata/the-date-that-can-only-say-no/index.html | [sweep-range] const SWEEP_LO = 1400, SWEEP_HI = 2026; the loop is inclusive.
- CONFIRMED | then executes all eight members' verifiers, nine scripts in all, end to end. | research/the-date-that-can-only-say-no/verify.mjs | [nine-scripts] MEMBER_CHECKS has nine scripts; all nine exit 0 in a pinned copy.
- CONFIRMED | It needs no network and no dependencies. | research/the-date-that-can-only-say-no/verify.mjs | [no-network] No package imports; it does need git and the repository (git show 67088387e0:...), which the page's placard says ("Running it needs the repository around it").
- CONFIRMED | Four say yes, and the portal's checker opens those four verifier files and confirms the assertion string is really in them | research/the-date-that-can-only-say-no/verify.mjs | [column-four-one-one] recreated: the C1 assertion-string checks pass.
- CONFIRMED | the dancing plague pins 1518, Waller's 2009 rate report and Pennant-Rea's 2018 conditional | research/the-dancing-plague/verify.mjs | [plague-partly] "provenance anchors: 1518 · 1526 visit · 2009 report · 2018 conditional".
- CONFIRMED | all 6 of its assertions are date inequalities; the kill prints 1767 < 1770 | research/let-them-eat-cake/verify.mjs | [cake-six-assertions] recreated: "ALL CHECKS PASS ... 6 assertions, 8 dated events" and "latest plausible writing year 1767 < arrival 1770"; all six are year comparisons.
- CONFIRMED | 42 assertions, including the 1777 milestone comparison and monotonicity | research/pigment-that-wasnt-born/verify-pigment-that-wasnt-born.mjs | [pigment-42] recreated: 42 ok lines including "cobalt milestone 1777 > 1665" and "slider monotonicity: later claim ⇒ never more flags".
- STALE | 19 + 12 assertions, that subtraction among them | research/a-minor-leonardo/verify-immersive.mjs | [leonardo-19-12] True at ship: the immersive verifier had 12 checks at 329feda618a. Commit e30a4e5cd10 (2026-10-14) added a 13th ("hover values and units agree ..."); it now prints "13 passed, 0 failed". The main verifier prints "ALL 19/19". Dated 2026-10-18 with an Updated line; the row now says 19 + 13.
- CONFIRMED | its verifier asserts 23 quantities and never compares two years | research/who-discovered-the-expanding-universe/verify-who-discovered-the-expanding-universe.mjs | [redshift-23] recreated: "23/23 checks passed"; its years appear only in labels and comments, no check compares two years.
- CONFIRMED | which runs from minus twenty-three to plus five hundred | public/strata/the-date-that-can-only-say-no/index.html | [gap-extremes] CASES gaps: -23, -2, -3, +112, +361, +500.
- CONFIRMED | The smallest gap in the set is two years | public/strata/the-date-that-can-only-say-no/index.html | [gap-smallest-two] redshift 1927 against 1929.
- CONFIRMED | The Number That Ends the Argument collects results where you should stop arguing about mechanism and size it | public/strata/the-number-that-ends-the-argument/index.html | [div-number-ends] "size the named mechanism from its own governing law, and compare it to what the observation needs."
- CONFIRMED | The Auditor That Wants Nothing ranks checks by how little the checker shares with the mind that wants the claim true | public/strata/the-auditor-that-wants-nothing/index.html | [div-auditor] Verbatim sense.
- MINOR | a date is one of the five discriminators those portals count | public/strata/the-anatomy-of-error/data.json | [div-repeated-five] "Those portals" are Repeated Until True and The Anatomy of Error. Repeated Until True has five cures (Cure I to Cure V); The Anatomy of Error's data.json has four, "cures": RECOMPUTE, COUNT, SOURCE, DATE (the same at 329feda618a and at 4fcb19ad5e). Listed (tier B MINOR), not edited.
- CONFIRMED | What the Ground Can't Say collects questions where the check comes back empty | public/strata/what-the-ground-cant-say/index.html | [div-ground-cant-say] "Here the check runs just as hard and comes back empty".
- MINOR | Five of the eight had already appeared as single cards when this page shipped on 2026-07-28: boredom, brioche, the redshift and the Mona Lisa in Repeated Until True (2026-06-26), and the dancing plague in What the Ground Can't Say (2026-07-11). | git show 329feda618a:public/strata/the-anatomy-of-error/data.json | [overlap-five-of-eight] Five members is right, but the list of portals is incomplete: The Anatomy of Error (2026-06-29) already held boredom-word, let-them-eat-cake, who-discovered-the-expanding-universe and a-minor-leonardo as member cards ({"slug": "boredom-word", ...}) from its first commit 3a42354f16a. Listed (tier B MINOR), not edited.
- CONFIRMED | Repeated Until True (2026-06-26) | src/content/strata/repeated-until-true.md | [overlap-dates] date: 2026-06-26T20:00:00Z.
- CONFIRMED | What the Ground Can't Say (2026-07-11) | src/content/strata/what-the-ground-cant-say.md | [overlap-dates-ground] date: 2026-07-11.
- WRONG | as of 2026-08-21 the overlap is six portal-member pairings across three portals, over five distinct members of the eight. | git show c3110c79443:public/strata/the-anatomy-of-error/data.json | [overlap-tally] At the last commit of 2026-08-21 The Anatomy of Error carried five members as cards ("slug": boredom-word, let-them-eat-cake, who-discovered-the-expanding-universe, a-minor-leonardo, the-dancing-plague), so the tally was ten pairings (Repeated Until True 4, What the Ground Can't Say 1, Anatomy 5) across three portals over five members. The verifier counted six because its patterns (slug:"x", slug: "x", /strata/x/) never match JSON's "slug": "x". (The Number Nobody Is Holding, also tagged portal, lists all eight in a per-layer audit table; not counted here.) At 4fcb19ad5e the verifier now counts seven and fails. Fixed 2026-10-18.
- OBSERVED | An earlier draft of this bullet said "five of seven" against a table with six rows | git show 4fcb19ad5e:memory/log.d/2026-07-28T2355Z-the-date-that-can-only-say-no.md | [obs-five-of-seven] "\"Five of seven assert it\" was also arithmetically impossible against a six-row table."
- OBSERVED | An earlier draft of this page claimed no member had been walked by any other portal. | git show 4fcb19ad5e:memory/log.d/2026-07-28T2355Z-the-date-that-can-only-say-no.md | [obs-member-overlap-draft] "\"No member is doubled\" was FALSE, and my own check was wrong in the same way."
- OBSERVED | the archive's portals build their links at runtime from a slug list, so the string /strata/boredom-word/ never appears in their source | git show 4fcb19ad5e:memory/log.d/2026-07-28T2355Z-the-date-that-can-only-say-no.md | [obs-runtime-links] Logged as stated. The same log's fix still missed The Anatomy of Error's data.json cards, which existed that day.
- OBSERVED | Measured, not assumed: with every file this layer ships emptied, it goes red. | research/verifier-independence/verdicts.json | [obs-placard-red] "dependence": "red", "depRed": 1, "depMeasured": 1.
- OBSERVED | Measured too: changing a figure in this page's printed text turned it red. | research/verifier-independence/verdicts.json | [obs-placard-prose] "prose": 2, "proseRed": 1. My own mutations: 23 of 48, 61,287, 1,547,130,448 and t = 7.21 go red; "by two years", "eight years", "best false 5.17" and "Sixty-one thousand" pass unnoticed.
- OBSERVED | the sweep printed "confirmed 0" on every case row and "The third lamp did not light" without reading its own count. | git show 08d0604448d | [obs-0928-correction] The 2026-09-28 diff replaces the typed "confirmed    0" and the fixed sentence with counted ones.
- WRONG | On 2026-08-02 The Anatomy of Error, a portal older than this one, folded boredom into its hundred and six record-corrections, picking up a member this page had already counted. | git show 3a42354f16a:public/strata/the-anatomy-of-error/data.json | [obs-anatomy-0802] boredom-word was a card in The Anatomy of Error from its first commit 3a42354f16a (2026-06-29), a month before this page shipped. What commit 6f368e8e25b (2026-08-02) added was a prose link to /strata/boredom-word/ and a new card for the-dancing-plague. The Anatomy held 89 members that day (90 after 1b3d5a88d8d); 106 was its count on 2026-08-20 (a92302abd9f). Fixed 2026-10-18.
- CONFIRMED | "if even the earliest date is after the claim, the anachronism is certain." That quotation is preserved verbatim as wording superseded on 2026-10-14. | git show 3f174862bad:public/strata/pigment-that-wasnt-born/index.html | [obs-forgery-quote] Present at 329feda618a and 3f174862bad; gone at d8c0473bd69 (2026-10-14).
- CONFIRMED | Waller's 2009 article reports a chronicle's daily rate; Ned Pennant-Rea's conditional extrapolation was published in 2018. | public/strata/the-dancing-plague/index.html | [plague-chronicle-15] Member text matches.
- CONFIRMED | Eight layers of this archive examine that reasoning: six comparisons, including a cobalt milestone flag, a limit case, and one foil. | src/content/strata/the-date-that-can-only-say-no.md | [eight-layers] Eight member relates entries; verifier C3 passes.
- CONFIRMED | its red lamp says flagged, not refuted | public/strata/the-date-that-can-only-say-no/index.html | [red-lamp-flagged] $('redName').textContent = current.dir === 'milestone' ? 'Flagged' : 'Refuted'.
- CONFIRMED | found an English health regimen printed about 1490 that keys sleeping posture to a completed first sleep | public/strata/what-the-first-sleep-was-first-of/index.html | [gouernayle-structure] "it structures the night around a completed first sleep after which the sleeper does something."
- CONFIRMED | that layer says so in four separate places rather than one | public/strata/what-the-first-sleep-was-first-of/index.html | [four-places-empty] Four limits still present (subset, paraphrase, silence, end of corpus), though one of them is now worded differently from the portal's quotation (see limit-floor-quote).
- CONFIRMED | Four of the six collision members do, in their own prose, and the table above quotes them. | research/the-date-that-can-only-say-no/verify.mjs | [four-of-six-state] Verifier asserts four yes, two no; passes.
- CONFIRMED | The cobalt setting of 1665 is illustrative, not a documented claimed year for the work. | public/strata/pigment-that-wasnt-born/index.html | [pigment-illustrative-1665] "starts at an illustrative 1665, not a documented claimed year."
- CONFIRMED | drag the claimed year later and flags only ever disappear | public/strata/pigment-that-wasnt-born/index.html | [pigment-flags-disappear] "Drag right and flags disappear"; its verifier asserts "later claim ⇒ never more flags".
- CONFIRMED | the brioche layer takes the latest end of a disputed range so its refutation survives the worst case | public/strata/let-them-eat-cake/index.html | [cake-latest-end] "The exact composition year (1765 vs ~1767) is a real source disagreement"; verifier uses "latest plausible writing year 1767".
- CONFIRMED | the 24-of-46 nebulae with individual distances | public/strata/who-discovered-the-expanding-universe/index.html | [redshift-24-of-46] "46 velocities, only 24 with individual distances".
- WRONG | the verifier searches every portal's frontmatter, shipped page and card data for all eight slugs, rebuilds that sentence from what it finds, and fails the moment the two disagree. The count cannot go quietly out of date here. | research/the-date-that-can-only-say-no/verify.mjs | [verifier-rebuilds-tally] It reads the card data but cannot see it: its patterns are `to: x`, `/strata/x/`, `slug:"x"`, `slug: "x"`, `slug:'x'`, and a JSON card is `"slug": "x"`. recreated with its exact patterns over git history: 5 pairings at 329feda618a and 6 at c3110c79443, while the cards gave 9 and 10. The count went quietly out of date from the day it was written; it now fails only because a 2026-10-15 correction (902bb97ef16) put a /strata/the-dancing-plague/ link into the Anatomy's prose. Fixed 2026-10-18.
- CONFIRMED | (added) "It also does not prove the painter ... material possibility is necessary, not sufficient." (member wording before 2026-10-14, superseded; preserved verbatim) | git show 329feda618a:public/strata/pigment-that-wasnt-born/index.html | Present at 329feda618a and 3f174862bad, removed by d8c0473bd69 on 2026-10-14; verifier C2 reads it at 67088387e0 and passes.
- CONFIRMED | (added) Every Tree Wrote the Same Bad Year, §10, "What a date actually dates" | public/strata/every-tree-wrote-the-same-bad-year/index.html | Heading "10. What a date actually dates".
- CONFIRMED | (added) the brioche line, qu'ils mangent de la brioche | https://en.wikipedia.org/wiki/Let_them_eat_cake | "the French phrase \"Qu'ils mangent de la brioche\"".
- CONFIRMED | (added) five hundred years after the Strasbourg dancers of 1518 | https://en.wikipedia.org/wiki/Dancing_plague_of_1518 | "a case of dancing mania that occurred in Strasbourg ... from July 1518"; 2018 minus 1518 is 500.
- CONFIRMED | (added) The Anatomy of Error, a portal older than this one | src/content/strata/the-anatomy-of-error.md | Frontmatter date 2026-06-29; first commit 3a42354f16a on 2026-06-29, before this page (2026-07-28).

## Findings
- [fixed] boredom-oed-1852: the bench's claimYearText called 1852 "the OED's first citation"; it now says "the Dickens citation the OED gave as its earliest in 2024 (its revised entry now gives 1829)", with a Corrected 2026-10-18 line. The member's own build.mjs label "OED first citation (1852)" is not rendered on this page and belongs to the boredom layer's pass.
- [fixed] limit-floor-quote: the limit case quoted "every second-sleep count on this page is a floor", withdrawn by the member on 2026-10-14; it now quotes the member's current "phrase counts do not count the practice", checked verbatim in C2, with a Corrected line.
- [fixed] obs-anatomy-0802: "folded boredom into its hundred and six record-corrections" on 2026-08-02 is now "added the dancing plague as it grew to eighty-nine record-corrections", and the bullet says the Anatomy carried boredom, brioche, the redshift and the Mona Lisa as cards since before this page shipped, with a Corrected line.
- [fixed] overlap-tally: "as of 2026-08-21 the overlap is six portal-member pairings" is now "as of 2026-10-18 the overlap is ten portal-member pairings across three portals, over five distinct members of the eight", rebuilt by the fixed C4 (the 2026-08-21 figure was also ten), with a Corrected line.
- [fixed] verifier-rebuilds-tally: C4 now matches the JSON card form "slug": "x" and expects ten pairings; the page no longer says "The count cannot go quietly out of date here. It can only go loudly." and says instead that it did until 2026-10-18, with a Corrected line.
- [dated] leonardo-19-12: the Mona Lisa row's "19 + 12 assertions" now reads "19 + 13", with an Updated 2026-10-18 line saying 19 + 12 was true when the page shipped and the thirteenth check came on 2026-10-14.
- [declined] gouernayle-london: the caption "Gouernayle of Helthe, London, ca. 1490" (TCP imprint: Westminster) is a tier B MINOR outside any sentence fixed for a WRONG item; listed for a later pass, not edited.
- [declined] oed-readers: "happened to be sent" is narrower than the source (editors also search databases) but does not falsify the point about a missing denominator; tier B MINOR, listed, not edited.
- [declined] div-repeated-five: "the five discriminators those portals count" (Repeated Until True five, The Anatomy of Error four) is loose for the pair; tier B MINOR, listed, not edited (both copies).
- [declined] overlap-five-of-eight: the sentence on the five members at ship names Repeated Until True and What the Ground Can't Say but not The Anatomy of Error; tier B MINOR in a separate sentence from the WRONG ones, listed; the corrected sentence beside it now states the Anatomy's cards at ship.