Last week's number cited a study that never existed. This week's is stranger: it cites a real document — a genuine government recommendation, printed and archived. The forgery isn't the source. It's the edit. Someone took a true sentence, cut it in half, and sold you the first part as a rule.

Read this autopsy on the web →

THE BODY "Drink eight glasses of water a day."

Eight glasses, eight ounces each — sixty-four ounces, call it two liters, of water you must consciously pour and drink, on top of everything else you consume, every single day, or fall short. It rides in your water bottle's measurement marks, in the app that buzzes to remind you, in the wellness column and the doctor's offhand advice and the number "8×8" that sounds like a formula because it is shaped like one. It is the rare health rule with a friendly face: no restraint, no discipline, nothing to give up — just more. Which is exactly why nobody thought to check it.

BROUGHT IN BY Everyone, and no one in particular. Unlike most bodies that come through here, this one has no single author to name — no book, no guru, no viral post. It arrived by consensus, the way "everybody knows" things do. But there is a paper trail, and it does lead somewhere: not to a laboratory, but to a single sentence in a wartime nutrition pamphlet, with its second half torn off.

TIME OF DEATH Call it 2002 — the year a kidney physiologist, a man who had spent his career studying exactly how the body handles water, went looking for the evidence behind the rule and came back empty-handed. But the belief had been dead on the page since 1945, the moment its source was misread. It just took fifty-seven years for anyone qualified to perform the reading. First, though, its strongest case — because the honest ones deserve a full hearing, and this one is more honest than most.

THE STEELMAN (at full strength — this is the honest case FOR)

Do not mistake this for a stupid belief. It sits on a foundation of things that are simply true. Water is not optional — you are mostly made of it, every cell runs on it, and real dehydration is genuinely dangerous: it degrades attention, triggers headaches, strains the kidneys, and at the extreme, kills. That is not in dispute anywhere in what follows.

And the rule does real good. Most people are careless drinkers, not careful ones; a simple, memorable target gets a sedentary office worker who would otherwise run on coffee and forget to drink until 4 p.m. to actually pick up a glass. For some people it is medicine: anyone prone to kidney stones is told, correctly, to drink more, and a concrete number makes that prescription stick. A visible quota builds a real habit, and a habit of drinking water is a good one to have. Against "drink whatever, whenever," "eight glasses" is a clear improvement — it errs toward a substance that is free, harmless in ordinary amounts, and easy to get.

That is the body at full strength: a true premise, a useful nudge, a harmless-seeming target that helps the people most likely to under-drink. Now watch what happens when you ask where the number eight came from.

THE AUTOPSY

Exhibit A — the sentence that was cut in half. The trail ends at a 1945 bulletin from the Food and Nutrition Board of the U.S. National Research Council. Here is what it actually said: "A suitable allowance of water for adults is 2.5 liters daily in most instances. An ordinary standard for diverse persons is 1 milliliter for each calorie of food." So far, this looks like the ancestor of the rule — 2.5 liters, roughly eight-and-a-half glasses. But read the very next sentence, the one that got amputated in transit: "Most of this quantity is contained in prepared foods." That is the whole autopsy in a single clause. The 2.5 liters was never a drinking target — it was total water, most of which the Board said you already get from the food on your plate and the coffee in your cup. The recommendation came with its own correction attached, and someone tore the correction off and reissued the first half as an order to drink. The number you obey is a real sentence with its ending deleted. [1]

Exhibit B — the specialist who went looking. In 2002, Heinz Valtin — a physiologist who wrote the standard textbooks on the kidney and water balance, a man with every reason to want the rule to be true — published a review in the American Journal of Physiology with a blunt title: "Drink at least eight glasses of water a day." Really? Is there scientific evidence for "8 × 8"? He searched the literature for the trial that justified the number. There was none. His finding, in his own careful phrasing: no scientific proof that every healthy person must drink eight glasses a day, and no evidence that mild, thirst-corrected fluctuations in hydration do any harm. He also demolished the rule's most-repeated corollary — that coffee and tea "don't count" or actively dehydrate you. They count. In the vast majority of people, caffeinated drinks contribute to daily fluid intake like any other. The rule was not only unsourced; the folklore built on top of it was wrong too. [2]

Exhibit C — what the measurement actually shows. When the counting was finally done properly, it produced not a number but a range. The U.S. Institute of Medicine's 2004 reference intakes set adequate total water — from all food and drink combined — at about 3.7 liters a day for men and 2.7 for women, and stated plainly that roughly a fifth of that arrives in food. Announcing the report, the Academies put it more bluntly still: "the vast majority of healthy people adequately meet their daily hydration needs by letting thirst be their guide." [3] And in 2022, a study in Science put deuterium-traced water through the bodies of 5,604 people across 26 countries, ages eight days to ninety-six years — the largest measurement of human water turnover ever made. Daily turnover ran from about 1 to 6 liters, with outliers near 10, driven by body size, climate, altitude, and activity. There is no universal number because the body does not have one; a small woman in a cool office and a large man roofing in Phoenix are not running the same plumbing. And crucially: turnover is not a drinking prescription — it counts every drop, including the water in food and the water your own metabolism makes. The honest answer to "how much water should I drink?" is enough that you're rarely thirsty and your urine is pale — a figure that moves with your size, your heat, and your day. True, useful, and printable on no water bottle. [4]

So why did this number win, and keep winning? Here the autopsy stops being about water and becomes about the sale. Most myths ask you to give something up; this one asks you to consume — and consumption can be sold. A quota needs a vessel. It needs the branded bottle with the hourly markings, the case of water that costs hundreds of times the tap it was drawn from, the electrolyte powder to make the water you didn't need "work better." The rule is the rare falsehood that feels like virtue: carrying the bottle makes you the disciplined one, so no part of you wants to check it. And it is nearly impossible to falsify from the inside — you cannot feel yourself being pointlessly over-hydrated, so nothing ever corrects the number downward. A belief you can't disprove by living, that flatters you for holding it, and that moves product — that is not a health rule that survived on evidence. It is one that survived because everyone in the chain, including you, had a reason not to look. And a closing note the friendly face hides: the rule is not even perfectly safe. People who force water past thirst to hit a target — most visibly endurance athletes — can dilute their blood sodium into hyponatremia, which at the extreme is fatal. The dose that helps the careless drinker can harm the obedient one.

CAUSE OF DEATH Amputation. A true sentence — "a suitable allowance of water is 2.5 liters, most of it contained in prepared foods" — had its second clause cut off, and the surviving half was hardened into a drinking quota it explicitly never was. The murder weapon was an edit: not a lie added, but a truth removed. Sixty-four ounces, poured daily in tribute to a footnote whose own author told you that you'd mostly already eaten it.

VERDICT FALSIFIED.

What survives, honestly stated: water is essential, real dehydration is genuinely dangerous, and if a fixed target is what gets you to drink, it has done you a service — keep it. But the specific claim — that every person must consciously drink eight glasses, sixty-four ounces, on top of food and other drinks, each day — has no study behind it. It is one sentence of a 1945 recommendation with the other sentence removed, unsupported when a kidney specialist finally checked, and contradicted by the only large measurement we have, which found not a number but a range from one liter to six. Drink when you're thirsty. Drink more when it's hot, when you're working, when your urine runs dark. Count the coffee; it counts. And retire the number — not because water is bad for you, but because the eight was never prescribed by anyone. It was left over from a sentence nobody finished reading.

THE LEDGER Entry 005 — "Drink eight glasses of water a day." Status: FALSIFIED. Cause of death: amputation — a recommendation cut in half and the wrong half kept. Beliefs built on a real source read only halfway: 1. Running total opened: 5.

This week's job: find one rule you follow that came from an authority — a guideline, a label, an official number. Read past the headline to the sentence after it. The correction you need is often sitting right there, in the half nobody quotes.

—M.

Next week: you use only ten percent of your brain. The other ninety is not idle — and we have the scans to prove where it went.

The weekly autopsy is free. The Ledger — every belief I've opened, searchable, plus the methods — is for paid subscribers.

The body of evidence (I show my work)

[1] Food and Nutrition Board, National Research Council. Recommended Dietary Allowances. National Academy of Sciences, 1945. (Origin of the figure: "A suitable allowance of water for adults is 2.5 liters daily in most instances… 1 milliliter for each calorie of food. Most of this quantity is contained in prepared foods." The final sentence — that most of the 2.5 L is already supplied by food — is the clause dropped in transmission. Provenance traced by Valtin 2002.) [2] Valtin H. "'Drink at least eight glasses of water a day.' Really? Is there scientific evidence for '8 × 8'?" American Journal of Physiology — Regulatory, Integrative and Comparative Physiology. 2002;283(5):R993–R1004. (No scientific support for 8×8 in healthy people; no evidence thirst-guided hydration harms; caffeinated beverages do count toward fluid intake.) [3] Institute of Medicine. Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate. National Academies Press, 2004. (Adequate Intake total water ≈3.7 L/day men, ≈2.7 L/day women, from all food and drink; ~20% from food; "let thirst be your guide" quote from the National Academies news release announcing the report, Feb 2004.) [4] Yamada Y, Zhang X, Henderson MET, et al. "Variation in human water turnover associated with environmental and lifestyle factors." Science. 2022;378(6622):909–915. (5,604 participants, 26 countries, ages 8 days–96 years; daily water turnover ~1–6 L, outliers to ~10 L; turnover includes food and metabolic water and is not a drinking target.)