The library of Alexandria did not burn

The story is familiar. A great library, the collected knowledge of the ancient world, destroyed in a single fire, setting human progress back by centuries. It is a powerful image and it is mostly wrong.

There was a fire in 48 BCE, during Caesar's campaign, and it damaged part of the collection. There were other losses later. But the library did not end in an afternoon. It declined over roughly three centuries, and the causes were mundane: funding withdrawn, scholars no longer invited, the loss of a political arrangement that had paid for copying. Papyrus does not last. A scroll needs recopying every century or two, and copying requires paid copyists. When the money stopped, the collection began to disappear at the speed of decay rather than the speed of fire.

Historians are careful to say this is not a minor correction. The two stories suggest completely different lessons. The fire version says knowledge is fragile in the face of violence, and the answer is to protect it from enemies. The slow version says knowledge is fragile in the face of indifference, and the answer is to keep paying for the boring work of maintenance.

The second lesson is less dramatic and considerably more useful, because indifference is a far more common threat than arson. Most archives that have been lost in the last century were not attacked. They were simply not funded, and then the building leaked, and then nobody checked.