Jewish practice holds that writing which may contain the name of God cannot be destroyed. Worn-out texts go into a geniza, a storeroom, until they can be buried.
Most communities emptied theirs periodically. The congregation at Ben Ezra in Fustat, old Cairo, did not, for something close to a thousand years.
Two things made the result extraordinary. First, the sheer duration. Second, that these Jews wrote Arabic in Hebrew characters, which meant the prohibition captured far more than liturgy. Into the chamber went business contracts, rabbinical court records, marriage settlements and divorce deeds, letters from merchants trading to India, petitions to officials, children's writing exercises, prescriptions, poems, accounts, and private correspondence that its writers would have been appalled to know survived.
The chamber had neither door nor window and was entered by ladder through a hole in the wall. It held on the order of 400,000 fragments.
The removal
In 1896 the Scottish twin scholars Agnes Lewis and Margaret Gibson returned from Egypt with fragments and showed them to Solomon Schechter at Cambridge, who identified one as Hebrew text of Ben Sira, a book known until then only in translation, and grasped what the chamber must contain.
Schechter travelled to Cairo, obtained the community's permission, and removed the greater part of the contents.
The Taylor-Schechter collection at Cambridge holds nearly 193,000 fragments. Others went to the Jewish Theological Seminary in New York, the Bodleian, the John Rylands Library in Manchester, and elsewhere, so that the two halves of a single letter can sit on separate continents, a problem it has taken a century, and latterly large-scale digitisation, to work around.
The removal is regarded less comfortably now than it was then. The documents belonged to a living community, which gave its consent, but did so to an eminent foreign scholar with institutional backing in circumstances where the terms were not really open to negotiation. None of it has gone back.
What it changed
S. D. Goitein spent decades on the documentary material and published A Mediterranean Society in six volumes between 1967 and 1993. It reconstructed, from these scraps, the ordinary life of a Jewish population inside a medieval Muslim world: how they made a living, what they owned, whom they married, how they fell out, how far they travelled, what they worried about.
That is not the usual raw material for medieval history, which mostly survives as the records of rulers and institutions. The Geniza gave a documented social history of ordinary people across the medieval Mediterranean, and it remains the single richest source for Jewish-Muslim life in that world, including how routine and unremarkable much of the interaction was.