7th century AD

The century opens with a war that everyone expects to be decisive and no one expects to be final. Rome and Persia have fought each other for four hundred years, and the round that begins in 602 is fought with the seriousness of exhaustion: Persian armies reach Chalcedon and take Jerusalem and the True Cross, Heraclius campaigns into the Persian heartland and gets the relic back, and by 628 both empires have won and lost so completely that the frontier is roughly where it started and both treasuries are empty. To a contemporary this is a familiar story with a satisfying ending. The eastern Roman recovery is celebrated as a triumph of piety.

Within fifteen years the map is unrecognisable. Armies out of Arabia take Damascus, Jerusalem, Alexandria, Ctesiphon. Persia does not lose a province; it ceases to exist. Rome loses Syria and Egypt, which is to say most of its grain, most of its revenue and three of its five patriarchates. What makes this hard to think about, for those it happens to, is not the military fact - empires have lost battles before - but that the conquerors arrive with a book and a claim. The Qur'an is not presented as a new religion but as a correction: the same God, the same prophets, the earlier scriptures distorted in transmission and now set right. Christian writers in Syria and Palestine at first reach for the category of heresy, because it is the only one they have.

For those inside the new order the experience is different again. A single generation has gone from tribal Arabia to governing from the Atlantic to the Oxus, and the practical questions are immediate: who may rule, on what authority, and what happens to the enormous machinery of Persian and Roman administration now in Muslim hands. The answer for a generation is that it keeps running much as before, in Greek and Pahlavi, with the same clerks. The first civil wars are fought over the first question, not the second, and the divisions they open do not close.

Meanwhile the eastern edge of the world is having a wholly different century. Tang China is confident, outward-facing and cosmopolitan, its capital full of Sogdian merchants and Central Asian music, its armies in the Tarim basin. Xuanzang walks to India and back, spends sixteen years there, and returns with manuscripts and a determination to translate them properly rather than approximately - his account of the journey is also the best description of seventh-century India that survives. In India itself Brahmagupta does something quietly radical: he treats zero as a number, states what happens when you add and subtract it, and works out rules for negative quantities in terms of debts and fortunes. He is wrong about division by zero, and says so in a way that shows he has thought about it.

In the Latin west the horizon has narrowed to what one library can hold. Isidore of Seville writes the Etymologies on the assumption that if the derivation of a word is understood, the nature of the thing is understood too, and the work is a monument to a scholar trying to save everything at once with insufficient materials. It is copied everywhere for the next eight hundred years, which tells you both how useful it was and how thin the alternative had become.

What nobody can assess yet is permanence. The caliphate looks, to Byzantine observers, like the latest in a long series of eastern powers that rise and subside. The loss of Egypt looks temporary. It is not.