Nobody died at thirty

There is a number that does a lot of damage. You have seen it in documentaries and textbook margins: life expectancy in the Middle Ages was about thirty-five. People hear this and build a whole world out of it. A world of worn-out elders at forty, of grandmothers who were twenty-nine, of a species that simply did not last. Almost none of that world is real. The number is true and the picture is false, and the gap between them is one of the most useful things history can teach you about reading evidence.

What the average is hiding

Life expectancy at birth is an average, and an average is a single number standing in for a crowd. When that crowd contains a great many people who died in their first five years, the average collapses toward the bottom, no matter how long the survivors lived.

This is exactly what happened. In much of premodern Europe, somewhere between a quarter and a third of children died before the age of five, many in the first year. Infant mortality on that scale is a weight on the scale that nothing else can lift. Average two deaths in infancy against one death at seventy and you get an "expectancy" in the mid-twenties, a figure that describes none of the three people involved.

A low life expectancy at birth is a statement about how dangerous it was to be a baby, not about how long a grown person could expect to live.

The number that actually feels like a lifespan

The honest figure is conditional. Ask not how long a newborn could expect to live, but how long someone could expect to live given that they had already survived childhood. That number is much larger, and it is the one that matches the lives people actually led.

In medieval England, a man who reached twenty-one could often expect to see his late fifties or sixties. The demographers Edward Wrigley and Roger Schofield, reconstructing English population from parish registers, found early-modern adult lifespans that would not embarrass a mid-twentieth-century town. Survive the gauntlet of childhood and a respiratory winter or two, avoid the obvious accidents and the plague years, and old age was a normal destination, not a miracle.

The record is full of people who got there. Sophocles wrote into his ninth decade. Cato the Elder reached eighty-five. Eleanor of Aquitaine outlasted two husbands and most of her children and died around eighty. Titian was painting into his late eighties; Hokusai the same. These are not freaks of nature held up as marvels. They are ordinary old people who happened to leave a name.

The past was full of the old. It was simply also full of small graves, and the small graves got into the arithmetic.

Why the bad picture sticks

Part of the trouble is the evidence itself. Roman tombstones, a favourite source for the morbidly statistical, record age at death, but they record it unevenly. Children and the very old are under-commemorated; young adults who died with grieving spouses and money for a carved stone are over-represented. Read the stones naively and you get a population that seems to die in its twenties and thirties, which is an artefact of who could afford a memorial, not of who died when.

Part of it is that the wrong picture is more dramatic. A civilisation of doomed thirty-year-olds is a better story than a civilisation of normal lifespans bracketed by lethal infancy. The melancholy version flatters us. We get to feel that we, the long-lived moderns, are a different kind of creature.

We are not a different kind of creature. We are a creature whose children mostly survive now, which is a colossal and recent achievement of plumbing, vaccination, and germ theory, and which moved the average without changing the ceiling very much. The old upper bound of the human life was roughly where it is today. What changed is the floor.

Standing in the crowd

This matters for how you imagine a historical scene, which is the work this whole site is about. Picture a market square in 1400. If you have absorbed the bad number, you populate it with the young, a few haggard forty-year-olds passing for ancient. That square is wrong.

The real square has white-haired people in it. It has a man who remembers a king two reigns back, a widow who has buried a husband and kept a business forty years, a grandmother minding children. It also, off to the side, in a way the square does not advertise, has the absence of all the children who did not live to stand in it. Both of those truths come out of the same misread number. Once you stop reading the average as a lifespan, the past gets older and sadder and far more recognisable at once.