---
slug: the-glasses-that-were-never-prescribed
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: 84
confirmed: 52
wrong: 31
unverifiable: 0
stale: 1
minor: 11
---

This record follows the 31 triage items and 52 source claims from the N2 checker, plus repair finding R1. The fixer independently re-read the triage evidence; carried-forward source claims identify their N2 provenance. The repair worker re-read Valtin, the 1945 Water paragraph, EFSA and Spence, inspected the two share images and authored copies named by the fix-checker, and executed the page script. C49 is now WRONG for its unsupported hydration-origin comparison; R1 records the introduced display-rounding slip as MINOR. Counts are 52 CONFIRMED, 20 WRONG, 11 MINOR counted under wrong, and one STALE, totaling 84. Existing local fixes remain recorded, while unresolved media and other-layer copies are held under open board notes until the board issues note ids. X25 retains the triage MINOR level and its supplied board note d24c18.

## Claims

- MINOR | X1 [title-never-prescribed]: The Eight Glasses That Were Never Prescribed | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R993-R995 distinguishes a review finding no supporting scientific studies from documented advice by physicians and writers. The text now names the 2002 review and its scope.
- MINOR | X2 [no-source]: It has no source. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R993-R995 distinguishes a review finding no supporting scientific studies from documented advice by physicians and writers. The text now names the 2002 review and its scope.
- MINOR | X3 [rule-with-no-source]: a rule with no source | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R993-R995 distinguishes a review finding no supporting scientific studies from documented advice by physicians and writers. The text now names the 2002 review and its scope.
- WRONG | X4 [deleted-dek]: That sentence got deleted, and a total-turnover figure became a demand for 1.89 litres of plain water on top of food. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X5 [dropped-body]: That sentence got dropped, and a total-turnover number quietly became a demand for 1.89 litres of plain water on top of food. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X6 [deleted-visible]: and the very next sentence got deleted. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X7 [deleted-second]: The second sentence is the one that vanished from the retelling. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X8 [whole-myth]: That deletion is the whole myth. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X9 [cut-twitter]: and the sentence got cut. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X10 [closest-source]: The closest thing to a source is a single line from the 1945 US Food and Nutrition Board | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994 identifies Stare's obituary as his closest reference; the Board appears as a different proposed route on R995.
- WRONG | X11 [food-quote-main]: A suitable allowance of water for adults is 2.5 litres daily in most instances. Most of this quantity is contained in prepared foods. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The original 1945 scan, printed p. 11, PDF page 12, was visually read. It uses liters; the per-calorie sentence intervenes; the prepared-food sentence contains this quantity is. The display and metadata now preserve the source words.
- WRONG | X12 [food-quote-description]: the 1945 figure was 2.5 litres of TOTAL water, 'most of this quantity contained in prepared foods.' | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The original 1945 scan, printed p. 11, PDF page 12, was visually read. It uses liters; the per-calorie sentence intervenes; the prepared-food sentence contains this quantity is. The display and metadata now preserve the source words.
- WRONG | X13 [food-quote-og]: the 1945 figure was TOTAL water, 'most of it contained in prepared foods,' | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The original 1945 scan, printed p. 11, PDF page 12, was visually read. It uses liters; the per-calorie sentence intervenes; the prepared-food sentence contains this quantity is. The display and metadata now preserve the source words.
- WRONG | X14 [food-quote-twitter]: The 1945 figure was TOTAL water ('most of it in prepared foods') | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The original 1945 scan, printed p. 11, PDF page 12, was visually read. It uses liters; the per-calorie sentence intervenes; the prepared-food sentence contains this quantity is. The display and metadata now preserve the source words.
- WRONG | X15 [deleted-sentence-accounting]: then does what the deleted sentence asked: it subtracts the water already in your food, and the water your own body makes by burning that food. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The entire 1945 Water paragraph includes food but gives no metabolic subtraction, 20% food share, or 300 mL metabolic amount. EFSA sections 2 and 6.1 distinguish later dietary and available-water quantities.
- WRONG | X16 [remainder-as-need]: What's left is what you'd have to drink. | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | EFSA sections 2, 4 and 6.1 distinguish dietary from available water and place the food fraction on dietary water. The fixed illustrative model uses D = T - M, F = fD, B = D - F and labels B as water from all beverages. Default B = 1360 mL, not the old 1300 mL or a personal plain-water need.
- WRONG | X17 [food-and-metabolic-already-in-total]: the ~20% food-water and ~300 mL metabolic water that the 2.5 L already contained | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf ; https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | The entire 1945 Water paragraph includes food but gives no metabolic subtraction, 20% food share, or 300 mL metabolic amount. EFSA sections 2 and 6.1 distinguish later dietary and available-water quantities.
- CONFIRMED | X18 [turnover-rate]: Total water turnover ≈ 1 mL per kcal | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf ; https://www.nationalacademies.org/read/10925/chapter/6 | REJECTED with evidence: EFSA section 4, printed p. 27, explicitly reports a 1 mL/kcal total-available-water convention for adults; section 2 includes metabolic production in available water. IOM p. 88 uses an intake convention, which does not invalidate this alternative. The approximate turnover proxy remains in a balanced illustrative model. This confirms only X18, not the old food denominator or plain-water interpretation.
- MINOR | X19 [population-averages]: The three inputs are population averages with real spread: | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The Board gives a standard/allowance. EFSA supplies separate conventions and ranges, not a measured population mean for all three defaults. They are now illustrative inputs with their scopes named.
- MINOR | X20 [no-documentation]: the specific instruction to drink eight glasses of plain water on top of food has no documented source | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R993-R995 distinguishes a review finding no supporting scientific studies from documented advice by physicians and writers. The text now names the 2002 review and its scope.
- WRONG | X21 [footer-valtin-attribution]: The "no evidence for 8×8" finding, the 1-mL-per-calorie origin and the deleted "most of this quantity is contained in prepared foods" sentence: | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- MINOR | X22 [footer-metabolic-attribution]: Water balance and metabolic-water production (~250–350 mL/day): | https://pmc.ncbi.nlm.nih.gov/articles/PMC6315424/ | Armstrong/Johnson Table 3 contains sex-specific examples spanning 0.2-0.4 L/day. EFSA section 6.1 directly gives the sedentary 250-350 mL/day estimate, and is now the cited source for that range.
- WRONG | X23 [readout-meaning]: Plain water you'd actually drink | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | EFSA sections 2, 4 and 6.1 distinguish dietary from available water and place the food fraction on dietary water. The fixed illustrative model uses D = T - M, F = fD, B = D - F and labels B as water from all beverages. Default B = 1360 mL, not the old 1300 mL or a personal plain-water need.
- MINOR | X24 [verifier-recomputes]: The verifier recomputes all of these \u2014 the 8×8 = 1.89 L conversion, the 1945 2.5 L total at 2500 kcal, the full ledger (total − food − metabolic), that it sums back to the total, and Valtin's zero-studies finding. | git show f5e5f84f34:research/the-glasses-that-were-never-prescribed/verify-the-glasses-that-were-never-prescribed.mjs | The pinned verifier assigns a study count of zero and tests it against zero. That does not repeat the review. The edited explanatory prose and verifier identify this as a source note, while arithmetic is recomputed.
- MINOR | X25 [check-capacity]: It can settle the finding; it cannot notice a figure that drifted on this page. | git show f5e5f84f34:research/the-glasses-that-were-never-prescribed/verify-the-glasses-that-were-never-prescribed.mjs | NOT FIXED. The triage explicitly retains MINOR for this checking-scope overclaim and assigns [open note:d24c18]. The generated placard and its shared generator are left for landing; the record no longer labels an accepted, deferred finding as declined.
- WRONG | X26 [dropped-relationship]: '8 × 8' came from a dropped second sentence. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R994-R995 calls these possible origins, reports Thomas's suggestion that the food sentence was overlooked, and also discusses Stare/McWilliams. It does not establish a deletion or the claimed causal history.
- WRONG | X27 [single-mistranslation]: Two health-and-body misconceptions traced to a single mistranslation of a source. | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8956797/fullTextXML ; https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Spence PMC8956797, Introduction and Figure 1, describes replotting and normalization of tongue-map data. Valtin R994-R995 gives an uncertain omission hypothesis. Neither supports the asserted shared single-mistranslation history.
- MINOR | X28 [eight-no-basis]: eight glasses a day is shown to have no basis at all. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Valtin R993-R995 distinguishes a review finding no supporting scientific studies from documented advice by physicians and writers. The text now names the 2002 review and its scope.
- STALE | X29 [modification-metadata]: "dateModified":"2026-06-27" | git show a80269e78367cf0f36f259b5ece60f62934fa7dc:public/strata/the-glasses-that-were-never-prescribed/index.html ; git show 2e60bcb5dd1374a2e5ef778ddb265bdd9ab3fee9:public/strata/the-glasses-that-were-never-prescribed/index.html | The initial a80269e78367cf0f36f259b5ece60f62934fa7dc HTML and the pinned page have dateModified 2026-06-27. Read-only git log and 2e60bcb5dd1374a2e5ef778ddb265bdd9ab3fee9 show a later HTML edit on 2026-10-01. Set to this revision's date, 2026-10-14, under the metadata-only exception.
- WRONG | X30 [added-body-next-sentence]: (added) recommended about 2.5 litres of *total* water a day and then, in the very next sentence, said "most of this quantity is contained in prepared foods." | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | The original 1945 scan visibly places the per-calorie sentence between the 2.5-liter allowance and the food sentence. The markdown body now includes that intervening standard and calls the food sentence third.
- MINOR | X31 [added-coffee-direction]: (added) The direct hydration counterpart: that page dismantles the eight-glasses rule, this one adds that coffee also counts toward the fluids you do need. | git show f5e5f84f34:src/content/strata/does-coffee-dehydrate-you.md | Read the pinned coffee frontmatter and glasses HTML. This/that came from the coffee page's viewpoint. A new local, direction-neutral relationship supplies the generated glasses teaser without editing the coffee layer.
- MINOR | R1 [rounded-display-arithmetic]: At 1450 kcal, 11% food and 200 mL metabolic water, food = 11% × 1250 = 138 mL; beverages = 1250 − 138 = 1113 mL. | public/strata/the-glasses-that-were-never-prescribed/index.html ; research/the-glasses-that-were-never-prescribed/verify-page.mjs | Reproduced from the pre-repair inline script. The exact amounts are 137.5 mL and 1112.5 mL; independent integer rounding made the displayed subtraction false. The repair keeps fractional millilitres in the equation and beverage sub-readout. The underlying model is unchanged.
- CONFIRMED | C01: The rule is everywhere: eight glasses of water a day, on top of whatever else you drink. | https://geiselmed.dartmouth.edu/news/2002_h2/08aug2002_water.shtml | N2 source check carried forward: [rule-everywhere] Dartmouth explicitly describes the exhortation as widespread. Valtin R993-R994 describes plain-water advice excluding caffeinated and alcoholic drinks. This checks a reported rule, not its validity.
- CONFIRMED | C02: You were told to drink eight glasses of water a day, on top of everything else you drink. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [plain-on-top] Valtin R993-R994 reports the recommendation and its exclusion of caffeinated and alcoholic drinks. The page is describing that advice, not endorsing it.
- CONFIRMED | C03: The 1945 US Food and Nutrition Board recommended about 2.5 litres of total water a day on a one-millilitre-per-calorie basis | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [1945-recommendation-dek] Read the original Water paragraph on printed p. 11, PDF page 12, including the rendered scan. The 2.5-liter allowance, one-milliliter standard, and inclusion of prepared foods are all present.
- CONFIRMED | C04: But the number everyone traces it to said the opposite. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [source-opposite] The substantive comparison holds: the 1945 allowance includes prepared foods and is not an instruction to drink that volume as plain water. The certainty of the origin story is assessed separately in the deletion entries.
- CONFIRMED | C05: the 1945 source said the opposite | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [source-opposite-meta] The 1945 paragraph expressly includes prepared foods. This verifies the contrast between total dietary water and plain drinking water, not a demonstrated historical chain of transmission.
- CONFIRMED | C06: Heinz Valtin's 2002 review found zero studies behind '8 × 8.' | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [valtin-zero-dek] The abstract states "No scientific studies were found in support of 8 ⫻ 8." This is a claim about the 2002 review, not a new claim that no later research exists.
- CONFIRMED | C07: Valtin (2002) found zero studies behind '8×8.' | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [valtin-zero-meta] Same 2002 abstract finding. The author and publication year agree with the paper.
- CONFIRMED | C08: In 2002, the physiologist Heinz Valtin went looking for the science behind | https://geiselmed.dartmouth.edu/news/2002_h2/08aug2002_water.shtml | N2 source check carried forward: [valtin-scientist] The dated Dartmouth announcement identifies the physiology professor, the invited review, and its August 8, 2002 online publication.
- CONFIRMED | C09: and found none. No study. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [valtin-no-study] Confirmed as the result of Valtin's search for supporting studies, within the sedentary, healthy-adult scope stated in the abstract.
- CONFIRMED | C10: 1 mL of water per calorie of food | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [board-rate] The scan reads "An ordinary standard for diverse persons is one milliliter for each calorie of food."
- CONFIRMED | C11: roughly 2 to 2.5 litres a day of total water. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [board-total-range] The original gives 2.5 liters and one milliliter per calorie. The 2-to-2.5-liter range follows for 2,000-to-2,500-kcal examples; Dartmouth also describes roughly two to two-and-a-half quarts.
- CONFIRMED | C12: Food & Nutrition Board, National Research Council, 1945 | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [fnb-label] The document title and authorship read "RECOMMENDED DIETARY ALLOWANCES REVISED 1945" and "Food and Nutrition Board, National Research Council".
- CONFIRMED | C13: its very next sentence read 'most of this quantity is contained in prepared foods.' | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [food-quote-dek] The dek explicitly includes the one-milliliter-per-calorie basis before saying "very next sentence". In that context the food sentence does follow it. The body version omits that intervening sentence and has a separate added finding.
- CONFIRMED | C14: The 1945 Food and Nutrition Board really did recommend ~2.5 L of total water on a 1-mL-per-calorie basis and really did add "most of this quantity is contained in prepared foods" as the next sentence. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [food-next-sentence] This sentence includes the 1-mL-per-calorie basis before its next-sentence assertion. That sequence agrees with the three-sentence original; it is different from the visibly abbreviated quotation.
- CONFIRMED | C15: Heinz Valtin's 2002 invited review in the American Journal of Physiology really did find no scientific studies supporting "8 × 8." | https://geiselmed.dartmouth.edu/news/2002_h2/08aug2002_water.shtml | N2 source check carried forward: [valtin-invited-review] Dartmouth and the paper identify an invited American Journal of Physiology review in 2002 and its no-supporting-studies result.
- CONFIRMED | C16: More food, more water moving through you. | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [more-food-more-water] The cited IOM excerpt establishes metabolic production proportional to energy expenditure. The same chapter, p. 134, also states that daily fluid requirements increase with metabolic rate and heat stress. This is a broad physiological relationship, not a fixed personal requirement.
- CONFIRMED | C17: Fruit, soup, yoghurt, even bread carry water. | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [foods-have-water] EFSA section 3.1 describes water in bakery products, hot meals, fruit, vegetables, and milk. These food examples are consistent with that account.
- CONFIRMED | C18: EFSA/IOM put food at ≈20% of total intake; | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [food-share-efsa] EFSA section 3.1 explicitly gives beverages 80% and food 20%. Section 6.1 gives the wider 20-to-30% food range. The page also acknowledges dietary variation.
- CONFIRMED | C19: EFSA/IOM put food at ≈20% of total intake; | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [food-share-iom] The IOM consumption section explicitly gives approximately 20 percent from food and 80 percent from fluids.
- CONFIRMED | C20: a fresh-food diet sits higher. | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [fresh-food-share] EFSA contrasts fruit and vegetables containing more than 80% water with relatively low-water fast foods and states that diet changes the food contribution.
- CONFIRMED | C21: the water your own body makes by burning it | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [body-produces-water] The IOM section on metabolic water states that oxidation of carbohydrate, protein, and fat produces water.
- CONFIRMED | C22: Burning carbohydrate, fat and protein releases water | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [oxidation-water] The cited IOM passage directly names all three substrates and their water yields.
- CONFIRMED | C23: 250–350 mL a day. | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [metabolic-range] EFSA section 6.1 gives 250-to-350 mL/day for sedentary people and up to 600 mL/day for very active people. The page describes the input as an approximate model parameter and separately warns about activity.
- CONFIRMED | C24: Free, and never counted by the myth. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [never-counted] Read as the distinction between water generated internally and the rule's externally drunk volume. The rule demands a fixed number of glasses; metabolic production is not a glass drunk. The excerpt alone is weak, but the rule definition and physiology together support that distinction.
- CONFIRMED | C25: 1 mL/kcal is the 1945 board's own basis (it gives 2.5 L at 2500 kcal, matching their figure); | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [model-board-basis] The Board's stated standard is one milliliter per calorie; 2,500 times one gives 2,500 mL. This checks the stated standard and arithmetic, not the later turnover interpretation.
- CONFIRMED | C26: food contributes ≈20% of total intake per EFSA/IOM but ranges widely with diet; | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [food-share-varies] EFSA section 6.1 expressly says the food-to-beverage relationship is not fixed and depends on food and beverage choices.
- CONFIRMED | C27: metabolic water is ≈250–350 mL/day. | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [model-metabolic-range] EFSA gives this range for sedentary people. Its larger active-person figure does not contradict the page's explicitly idealized, variable model parameter.
- CONFIRMED | C28: Real hydration needs vary enormously with heat, exercise, body size, pregnancy, and illness, and can be far above any of these numbers. | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [needs-vary] Read the primary IOM passage and both additional sources. IOM names exercise, heat, fever and diarrhea; EFSA proposes 300 mL/day extra in pregnancy; Armstrong Table 3 names body size. The claim and qualifications are supported.
- CONFIRMED | C29: the figure it's traced to counted food-water in, not out. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [food-counted] The 1945 paragraph includes water in prepared foods. The proposed historical origin remains a hypothesis, addressed by the deletion findings.
- CONFIRMED | C30: For healthy adults thirst is a reliable regulator | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [healthy-thirst] The IOM chapter summary and thirst discussion describe daily hydration maintained by thirst and drinks at meals. The page also names limits for older people and exercise in heat.
- CONFIRMED | C31: you don't need to hit a number. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [no-target] Supported in the surrounding healthy-adult, ordinary-condition context by Valtin's scope limitation and IOM's thirst discussion. This is not treated as advice for illness or extreme heat.
- CONFIRMED | C32: thirst is a weaker guide for the elderly and for athletes sweating in extreme heat, where deliberate intake genuinely matters. | https://www.nationalacademies.org/read/10925/chapter/6 | N2 source check carried forward: [thirst-limits] IOM discusses reduced thirst responses in older people and incomplete short-term replacement during exercise in heat; EFSA describes the need to replace extreme losses. The scope is contextual, not a universal thirst-only prescription.
- CONFIRMED | C33: The "no fixed number" finding is about the general healthy adult, not every body in every condition. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [scope-healthy] The cited abstract limits its conclusion to healthy adults in a temperate climate leading a largely sedentary existence. The page's neighboring sentences explicitly exclude extreme conditions.
- CONFIRMED | C34: The 2.5 L figure is total, not plain. | https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/36074/GIPE-031998.pdf | N2 source check carried forward: [total-not-plain] The original prepared-food sentence directly supports this distinction.
- CONFIRMED | C35: The US fluid ounce (29.5735 mL) is the standard unit conversion. | https://www.nist.gov/pml/special-publication-811/nist-guide-si-appendix-b-conversion-factors/nist-guide-si-appendix-b8 | N2 source check carried forward: [ounce-unit] NIST Appendix B.8 lists the US fluid ounce as 2.957 353 E+01 mL. The page rounds this to 29.5735 mL.
- CONFIRMED | C36: Eight 8-oz glasses is 64 fl oz = 1.89 L of plain water | https://www.nist.gov/pml/special-publication-811/nist-guide-si-appendix-b-conversion-factors/nist-guide-si-appendix-b8 | N2 source check carried forward: [eight-glasses-volume] 64 times 29.57353 mL equals 1,892.70592 mL, which rounds to 1.89 L. The page's rounded conversion factor gives 1,892.704 mL, the same displayed result.
- CONFIRMED | C37: Eight 8-oz glasses is 1.89 litres; | https://www.nist.gov/pml/special-publication-811/nist-guide-si-appendix-b-conversion-factors/nist-guide-si-appendix-b8 | N2 source check carried forward: [eight-eight-meta] Same independently recomputed US-volume conversion; no disagreement at the quoted precision.
- CONFIRMED | C38: Valtin, H. (2002), "Drink at least eight glasses of water a day." Really? Is there scientific evidence for "8 × 8"? Am. J. Physiol. Regul. Integr. Comp. Physiol. 283(5): R993–R1004 | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [valtin-citation] Read the title, author, journal, year, and R993-R1004 pagination in the PDF.
- CONFIRMED | C39: Dartmouth Geisel summary | https://geiselmed.dartmouth.edu/news/2002_h2/08aug2002_water.shtml | N2 source check carried forward: [dartmouth-citation] The Dartmouth page carries the matching announcement, dated August 8, 2002.
- CONFIRMED | C40: EFSA, Scientific Opinion on Dietary Reference Values for Water (2010) | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [efsa-citation] The PDF title and first page identify the 2010 scientific opinion, EFSA Journal 8(3):1459.
- CONFIRMED | C41: Armstrong & Johnson, "Water Intake, Water Balance, and the Elusive Daily Water Requirement," Nutrients (2018) | https://pmc.ncbi.nlm.nih.gov/articles/PMC6315424/ | N2 source check carried forward: [armstrong-citation] The tool encountered a challenge, but curl fetched the article itself. Its header names Lawrence E Armstrong and Evan C Johnson, the exact title, Nutrients 2018, 10(12):1928, and the matching DOI. Only HTML spacing prevents a literal normalized match to the combined excerpt.
- CONFIRMED | C42: Metabolic water (overview) | https://en.wikipedia.org/wiki/Metabolic_water | N2 source check carried forward: [metabolic-citation] The named overview exists and its lead contains the quoted definition. The physiological definition is independently supported by IOM.
- CONFIRMED | C43: Total-water reference values and the ≈20% food contribution: | https://backend.orbit.dtu.dk/ws/files/6350877/5.Scientific.pdf | N2 source check carried forward: [footer-food-attribution] EFSA gives total dietary water reference values and a 20% example, with a 20-to-30% range elsewhere. The attribution is sound.
- CONFIRMED | C44: The ledger is a clean accounting identity, not a clinical prescription. | git show f5e5f84f34:public/strata/the-glasses-that-were-never-prescribed/index.html | N2 source check carried forward: [accounting-identity] The formula is algebraically an identity under its definitions, and the nearby text explicitly says it is not a clinical prescription. That qualification does not validate the separate claims about what the source allowance included.
- CONFIRMED | C45: Both are record-corrections of a 'fact' that survived by repetition, not evidence. | https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | N2 source check carried forward: [repetition-relationship] Valtin documents widespread repetition of the recommendation and lack of supporting studies; the linked portal explicitly discusses repetition. This is the page's comparative interpretation, not evidence of the asserted deletion history.
- CONFIRMED | C46: The tongue map came from a mis-rendered German paper; | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC8956797/fullTextXML | N2 source check carried forward: [tongue-relationship] The scientific review describes Boring's replotting of Hanig's German thesis and the misleading interpretation of regional sensitivity. It supports a mis-rendered graph account; it does not establish a literal linguistic mistranslation.
- CONFIRMED | C47: six breaths a minute is shown to be a real resonance with a derivable basis; | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC7578229/fullTextXML | N2 source check carried forward: [breaths-relationship] The review places human resonance frequencies at 4.5-to-6.5 breaths/minute. The linked layer explicitly models a peak near 5.4 and treats six as the rounded explanatory value. It does not claim every person resonates at exactly six.
- CONFIRMED | C48: coffee also counts toward the fluids you do need. | https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0084154 | N2 source check carried forward: [coffee-relationship] The cited randomized crossover study reports similar hydration with moderate coffee and water in habituated men. IOM and Valtin also count caffeinated beverages toward fluids. This checks the factual coffee clause; the reversed page references are an added finding.
- WRONG | C49: The same shape of scare: a health rule the culture treats as settled that traces back to one thin, over-read source, corrected by actually reading the evidence. | https://www.ebi.ac.uk/europepmc/webservices/rest/PMC4870486/fullTextXML ; https://perruchenautomne.eu/wordpress/wp-content/uploads/2014/12/Valtin-AJPR-review-2002.pdf | Repair review overturns the carried-forward CONFIRMED verdict for the hydration comparison. Valtin R994-R995 presents competing possible origins; the MSG review cannot establish a single-source history for the eight-glasses rule. The authored source is src/content/strata/does-msg-cause-headaches.md:12, outside this repair scope; the local Nearby copy is rebuilt at landing after its source is corrected.
- CONFIRMED | C50: Nothing here may lie about anything real; that rule never bends. | git show f5e5f84f34:WAKE.md | N2 source check carried forward: [site-honesty-rule] The source expressly states "It must never lie about anything real." This is a documented rule, not a claim that every page is already error free.
- CONFIRMED | C51: When a check finds something false, the page says what changed and when. | git show f5e5f84f34:assay/README.md | N2 source check carried forward: [site-corrections] The source policy assigns dated Corrected lines to WRONG claims and Updated lines to STALE claims. This checks the stated correction policy, not a census of compliance.
- CONFIRMED | C52: These are facts of the record, not model outputs. | git show f5e5f84f34:research/the-glasses-that-were-never-prescribed/verify-the-glasses-that-were-never-prescribed.mjs | N2 source check carried forward: [facts-not-model] The described unit conversion, original Board wording with the per-calorie clause, and Valtin's review result are independently supported by the external documents. The verifier alone would not suffice.

## Findings

- [fixed] X1 [no-source-overstatement]: Editable title fixed locally. On 2026-10-18 the authored gallery-card title in public/strata/repeated-until-true/index.html:360 was changed to "The Eight Glasses Valtin Found No Evidence For" (MINOR, no dated line), and public/og/the-glasses-that-were-never-prescribed.png and public/og-thumbs/the-glasses-that-were-never-prescribed.webp were inspected the same day and already show the corrected title. Still left: this layer's card in The Anatomy of Error (public/strata/the-anatomy-of-error/data.json:2587) still carries the old title "The Eight Glasses That Were Never Prescribed". Round 3 on 2026-10-18 gave the glasses card in the-anatomy-of-error its current title; board note 2f38e0 is closed.
- [open note:db84a2] X2 [no-source-overstatement]: The local markdown body is fixed. On 2026-10-18 the copy in src/content/strata/does-swallowed-gum-stay-in-your-stomach.md:14 and its Nearby teaser (public/strata/does-swallowed-gum-stay-in-your-stomach/index.html:470) were narrowed to Valtin's 2002 finding of no scientific studies supporting the rule for healthy, largely sedentary adults in a temperate climate (MINOR, no dated line). Still left: the authored relationship copies in src/content/strata/is-the-5-second-rule-real.md:18, src/content/strata/is-your-attention-span-shorter-than-a-goldfish.md:11 and src/content/strata/does-spicy-food-cause-ulcers.md:14, and the old notes of all four layers in the preserved quotations of There, Here (public/strata/there-here/index.html:250 and research/lineage-edge-grammar/page.json:1). The family shared with X1/X3/X20/X28 remains open for those copies; Valtin reports absent scientific support within his review scope, not an absence of recommendations or documentary sources. Round 3 on 2026-10-18 corrected the relation notes in is-the-5-second-rule-real, is-your-attention-span-shorter-than-a-goldfish and does-spicy-food-cause-ulcers; what remains is the preserved quotations in There, Here.
- [fixed] X3 [no-source-overstatement]: The kicker now introduces the evidence behind eight glasses; metadata and research copies no longer claim no source.
- [open note:b76f58] X4 [deletion-origin-as-fact]: Local dek and historical copies are fixed, covered by Corrected line 1. On 2026-10-18 the authored copies in src/content/strata/repeated-until-true.md:53 and public/strata/repeated-until-true/index.html:359 were corrected: the overlooked prepared-food sentence is one possible origin reported by Valtin, not a documented deletion, and it is the third sentence, after the per-calorie standard (X5-X9/X21/X26, X30), with two dated Corrected 2026-10-18 lines there. Still left: this layer's TV narration at public/tv/lib/narration/batch-15.js:374 ("Then one sentence got dropped"), this layer's card in The Anatomy of Error (public/strata/the-anatomy-of-error/data.json:2592-2593: "a sentence later dropped", "the deleted sentence is the whole error"), and the preserved quotation in There, Here (public/strata/there-here/index.html:250 and research/lineage-edge-grammar/page.json:1, notes[1041]). Round 3 on 2026-10-18 corrected the eight-glasses card in the-anatomy-of-error (dated lines there), and the directing instance corrected the TV narration line; what remains is the preserved quotation in There, Here.
- [fixed] X5 [deletion-origin-as-fact]: Markdown body now attributes the suggestion and preserves uncertainty. Corrected line 1.
- [fixed] X6 [deletion-origin-as-fact]: Visible dek, descriptions, JSON-LD and authored comments now qualify the origin. Corrected line 1.
- [fixed] X7 [deletion-origin-as-fact]: Quote caption identifies the third sentence and an attributed suggestion. The demonstrative deletion animation was removed. Corrected line 1.
- [fixed] X8 [deletion-origin-as-fact]: The whole-myth assertion is replaced with an explicit statement that the proposed origin is not a complete history. Corrected line 1.
- [fixed] X9 [deletion-origin-as-fact]: Twitter description now gives a possible overlooked-sentence origin. Corrected line 1.
- [fixed] X10 [board-as-closest-source]: Opening paragraph names Stare's obituary and the competing 1974 book route. Corrected line 2.
- [fixed] X11 [altered-board-quotation]: Display restores liters and the intervening per-calorie sentence in the original order. Corrected line 3.
- [fixed] X12 [altered-board-quotation]: Description and JSON-LD restore the exact prepared-food sentence, including is. Corrected line 3.
- [fixed] X13 [altered-board-quotation]: OG description restores this quantity is and the rest of the sentence. Corrected line 3.
- [fixed] X14 [altered-board-quotation]: Twitter description restores the original prepared-food sentence. Corrected line 3.
- [fixed] X15 [board-metabolic-water-attribution]: Calculator introduction distinguishes later assumptions from the Board paragraph; README and verifier do too. Corrected line 4.
- [fixed] X16 [model-as-plain-water-need]: Model uses the dietary denominator and calls the remainder illustrative beverage water. Default, examples, live text, canvas labels, README and verifier assertions corrected. Corrected line 5.
- [fixed] X17 [board-metabolic-water-attribution]: Details remove the claimed historical 20%/300 mL decomposition and identify those as later model inputs. Corrected line 4.
- [fixed] X19 [defaults-as-population-averages]: Details describe illustrative inputs, a convention, ranges and chosen defaults, not three measured averages.
- [fixed] X20 [no-source-overstatement]: Details, metadata and research copies use the dated, scoped review finding instead of no documented source.
- [fixed] X21 [deletion-origin-as-fact]: Footer now attributes possible origins to Valtin and links the original Board scan. Corrected line 1.
- [fixed] X22 [metabolic-range-attribution]: Footer cites EFSA directly for the sedentary metabolic range; Armstrong/Johnson is retained for balance and variation.
- [fixed] X23 [model-as-plain-water-need]: Readout and dependent text identify water from all beverages; corrected denominator gives 1.36 L at default. Corrected line 5.
- [fixed] X24 [verifier-settles-source-claims]: Check panel says arithmetic is recomputed and the review result is restated. The verifier removes the assigned-zero assertion and prints an explicitly untested source note.
- [open note:d24c18] X25 [verifier-settles-source-claims]: The triage retains MINOR and assigns this existing census note to the generated checking-scope overclaim. The verifier checks conditional arithmetic and restates the review; it does not establish the history. No local placard or shared generator edit; landing handles the generated copy.
- [fixed] X26 [deletion-origin-as-fact]: Frontmatter relationship now says the third sentence may have been overlooked, with attribution to Valtin. Corrected line 1.
- [fixed] X27 [single-mistranslation-origin]: Relationship source distinguishes replotting from a possible omission; its Nearby teaser was regenerated. Corrected line 6.
- [fixed] X28 [no-source-overstatement]: Frontmatter relationship now states the 2002 review result instead of no basis at all.
- [dated] X29 [stale-date-modified]: JSON-LD dateModified set to 2026-10-14. No Updated or Corrected line for this metadata change; the seven substantive correction lines already date the revision.
- [fixed] X30 [board-sentence-order]: Markdown body restores the intervening per-calorie standard and names the prepared-food sentence as third. Corrected line 7.
- [fixed] X31 [reversed-coffee-relationship]: This layer's own coffee relationship now supplies a direction-neutral note; the existing Nearby generator regenerated only this page's block.

- [open note:08dd3f] C49 [deletion-origin-as-fact]: Corrected the record verdict to WRONG. On 2026-10-18 the authored relation note at src/content/strata/does-msg-cause-headaches.md:12 was corrected to keep the single 1968 letter for MSG only and to present the overlooked 1945 sentence as one possible origin of the eight-glasses rule; that page's Nearby text matches and it carries a dated Corrected 2026-10-18 line. Apart from this page's generated Nearby copy (public/strata/the-glasses-that-were-never-prescribed/index.html:338), which the landing rebuild regenerates from the corrected source, the only thing left is the preserved quotation in There, Here (public/strata/there-here/index.html:250 and research/lineage-edge-grammar/page.json:1, notes[306]), which still says the rule traces back to one thin, over-read source and needs an annotation.
- [fixed] R1 [rounded-display-arithmetic]: Food and beverage amounts retain one decimal place where needed. The reported 1450/11/200 setting now displays food = 11% × 1250 = 137.5 mL; beverages = 1250 − 137.5 = 1112.5 mL. The beverage sub-readout preserves 137.5 mL too. Added a regression scenario executing the actual page script. MINOR precision repair, with no new dated line.

## Not done

Four findings use open board notes: X1 for title pixels and the repeated-until-true title, X2 for four other-layer relationship notes, X4 for the repeated-until-true history, and C49 for the MSG comparison. The claims gate requires six hexadecimal characters, so the requested placeholders intentionally leave this record awaiting board integration.

X25 remains a listed MINOR checking-scope overclaim under the existing board note d24c18, as triage directed. Shared generated surfaces are rebuilt at landing; no shared generator or placard was edited by the repair worker.

The old title is still baked into both share images. It is a title claim, not a dek-derived claim; regeneration is requested in the repair report. No film, audio or image was rendered or re-encoded.

The original fixer report is research/the-glasses-that-were-never-prescribed/fix-report-2026-10-14.md. The subsequent evidence audit, every fix-checker outcome, all seven dated lines, commands and directing-instance handoffs are in research/assay-line/cited-1014b/the-glasses-that-were-never-prescribed/repair-report.md. The repair worker changed only this record, the layer HTML, its page regression check and the repair report. Other layers and public/checks were not changed by the repair worker.