Water

In 1945 the Food and Nutrition Board of the National Research Council in the United States published a set of recommendations about what people should eat and drink.

It was a wartime and immediately postwar document, produced by a body set up to think about nutrition at national scale, in a period when governments had become extremely interested in whether their populations were adequately fed. The recommendations covered a lot of ground. One line in it concerned water.

The suggestion was that a person needs roughly one millilitre of water for each calorie of food they eat. For an ordinary adult diet that works out at something in the region of two to two and a half quarts a day.

Two and a half quarts is about eighty ounces. Eighty ounces is ten eight-ounce glasses. Two quarts is sixty-four ounces, which is eight.

There it is. That is where it comes from, as far as anyone has been able to establish. A ratio of water to calories, published in a technical nutrition document eighty years ago, converted into a number of drinking glasses by nobody in particular, somewhere along the way.

I find the origin story satisfying in the way these usually are, which is that the original is more sensible and more boring than the thing it became. A millilitre per calorie is a reasonable rule of thumb. It scales with the person. A large man doing physical work eats more and therefore needs more, and the ratio handles that automatically, which is exactly what a fixed number of glasses does not do.

But that is not the interesting part.

The interesting part is the sentence that came next in the same recommendation, and what happened to it.

Because the Board did not stop at the quantity. They went on to note that most of this quantity is already contained in prepared foods.

Most of it. Already there. In the food.

Fruit is largely water. Vegetables are largely water. Soup, obviously. Bread has water in it. Meat has a great deal of water in it. Milk and juice and coffee and tea are water with things dissolved in them. If you eat an ordinary diet, a substantial share of your daily fluid arrives without you drinking a glass of anything.

So the 1945 recommendation, read whole, says something like: your body needs roughly this much water, and you are already getting most of it from lunch.

That is a completely different instruction from the one that survived.

And the way it became different is not a conspiracy and not really even an error. It is just what happens to information as it travels. A recommendation has two parts. The first part is a number, which is memorable, quotable and easy to act on. The second part is a qualification, which is none of those things.

Guess which one made it to the next repetition.

Next episode: the clause falls off, and a number begins to travel on its own.