The Earth of 8000 BCE was warm and growing warmer, but the great ice sheets had not finished retreating. A vast ice sheet still covered much of modern Canada and would not surrender its last remnants for another three thousand years; the ice over modern Scandinavia was likewise withdrawing northwards. All this meltwater was raising the sea. The most detailed reconstructions, such as those of Kurt Lambeck and colleagues, place global sea level around this date some thirty-five to forty metres below its present mark, and rising rapidly — at times by more than a metre a century. Coastlines that people had known for generations were being drowned.

The consequence for these islands was that Britain was not yet an island. It remained joined to the European mainland across the bed of what is now the southern North Sea, by a low, well-watered country that archaeologists have named Doggerland. Britain would not finally be severed by the rising sea for another two thousand years, around 6200 to 6000 BCE. Elsewhere the same process was at work: the land bridge between modern Russia and Alaska had only recently been submerged, and the broad continental shelf off Southeast Asia was steadily drowning into the islands we know today.

Most striking of all, the largest hot desert on Earth did not, in any meaningful sense, exist. During the African Humid Period — driven by changes in the Earth’s orbit that strengthened the monsoon — the Sahara was green, a landscape of grassland, seasonal rivers, and great freshwater lakes that endured roughly from 10,000 to 5,000 years ago. Elephants, giraffes, hippopotami, and crocodiles lived where there is now only sand, and human communities thrived along the water. The desert we associate with North Africa was a thing of the future.

Any figure for the human population of 8000 BCE is an educated estimate built on archaeological site densities and modelling, and the honest range is wide. The most commonly cited central figure is around five million, with serious estimates spanning from perhaps one to ten million. The whole of humanity, in other words, was fewer in number than the inhabitants of a single large modern city, scattered across six continents.

The geographical distribution of those millions is far shakier ground than the global total, and it is worth being blunt that any regional breakdown is largely guesswork; the figures that follow are illustrative orders of magnitude, not measurements. The most-cited modern reconstruction, the History Database of the Global Environment (HYDE), gives the world of 10,000 BCE a total of roughly 4.5 million and distributes it in a way that startles most readers: around half in the Americas (about a quarter in North America and a quarter in South America), only some 480,000 in the whole of Europe, and — implausibly to many specialists — as few as 230,000 in Africa, with the remainder, on the order of a million and a half, in Asia. These particular figures should be treated with real scepticism: Africa was the species’ ancient homeland and the green Sahara was well watered and peopled, so the suggestion that it held the fewest inhabitants of any continent strains belief, and the American totals strike many as far too high for a hemisphere settled comparatively recently from small founding groups.

Older and more conventional estimates, such as those compiled by Colin McEvedy and Richard Jones or by the demographer Massimo Livi-Bacci, instead place the great bulk of humanity in Africa and Asia — the long-settled heartlands of the species — and likewise put Europe at only a couple of hundred thousand. That two serious reconstructions can disagree by an order of magnitude over which hemisphere held most of humanity is the clearest illustration of how little is really known. What can be said with confidence is only this: the global total was a few million; people lived in small, mobile bands on every habitable continent; numbers were densest where food was richest — coasts, lakes, great rivers, the well-watered uplands of the Near East, the monsoon lands of southern and eastern Asia, and the green Sahara; the Americas and Australia were sparsely settled; and no region on Earth held more than the most trivial fraction of the numbers it holds today. Beyond that, the regional shares are genuinely uncertain, and the published figures deserve a corresponding humility. What is certain is that the planet was, by any modern standard, almost empty: even the largest settlements of the age held only a few hundred souls, and the overwhelming majority of people lived in groups of a few families.

Before turning to the peoples themselves, it is worth fixing the regions through which we shall view them, for these are the seedbeds from which the later, divergent course of human history — including the first true civilisations, some five thousand years further on — would grow. At their crudest the divisions are simply the inhabited continents; but a more useful map distinguishes a handful of culture-areas, each with its own environment, its own emerging way of life, and its own destiny. Chief among them is the Near East — the lands of south-west Asia about the eastern Mediterranean — and, at its heart, the Fertile Crescent, the well-watered arc that was then the cradle of farming. Beyond it lie East Asia, the basins of the great Chinese rivers and the islands of Japan; South Asia, the Indian subcontinent; North Africa and the Nile, then fused with a green and populous Sahara; sub-Saharan Africa, the species’ ancient homeland; Europe, a forested peninsula of hunter-gatherers; the Americas, peopled comparatively recently and developing in isolation; and Australia with New Guinea, the surviving fragments of the drowning continent of Sahul. These are the roots of the regional stories that follow — the starting points against which the later rise of distinct civilisations can be measured. The survey below treats each in turn.

The Near East and the Fertile Crescent: modern Turkey, Syria, Lebanon, Israel, the West Bank, Jordan, and northern Iraq

The most innovative societies of the age lived in what scholars call the Near East — broadly, the lands of south-west Asia at the eastern end of the Mediterranean, running from modern Turkey and the eastern Mediterranean coast across to Iran; in everyday speech much of this is now called the Middle East. More precisely, they lived in its watered land that curves from the eastern Mediterranean shore up through modern south-eastern Turkey (a region the ancients called Anatolia) and down towards the Persian Gulf, so named for its crescent shape and its fertility. The era in question is what archaeologists call the Pre-Pottery Neolithic — the first farming phase of the region, before pottery came into local use. (The Neolithic, or “New Stone Age”, is the archaeologists’ name for the era of early farming and polished stone tools; the “Levant”, referred to throughout, is the eastern Mediterranean seaboard, roughly modern Israel, the Palestinian territories, Lebanon, western Syria, and Jordan.) This world had deep local roots: for two thousand years and more before any farming, the sedentary and semi-sedentary Natufian foragers of the modern Levant had already built stone-founded dwellings, harvested wild cereals with flint sickles, and buried their dead in fixed places.

Its most arresting monument is Göbekli Tepe, on a limestone ridge in modern south-eastern Turkey, near the city of Şanlıurfa. There, hunter-gatherers — not farmers — raised great enclosures of carved, T-shaped limestone pillars, some weighing many tonnes and decorated with reliefs of foxes, boars, snakes, scorpions, and cranes. Radiocarbon dating places its construction and use between about 9600 and 8200 BCE, with the earliest measured dates falling near 9530 BCE, just as the Younger Dryas gave way to the warm Holocene; one of its painted boar sculptures is dated to roughly 8700–8200 BCE. What we can say with confidence is that the structures are monumental, deliberately built, and richly symbolic. What we cannot say is what they were for: the popular label “temple” is an interpretation, not an established fact, and recent attempts to read calendars or astronomy into the carvings are speculative and not accepted by the wider field. It is now clear, however, that Göbekli Tepe was not unique. It belongs to a cluster of contemporaneous sites across the region known collectively as the Taş Tepeler, the “Stone Hills” — among them Karahan Tepe, Nevalı Çori, and Sayburç — sharing the same distinctive pillars and pointing to a connected tradition across many communities. This recognition, the work of the past two decades, overturned the old assumption that monumental architecture had to await agriculture.

Some 600 kilometres to the south, at Jericho — the site of Tell es-Sultan, in the modern West Bank, in the Jordan Valley, and one of the oldest continuously occupied places on Earth — the inhabitants raised a massive stone tower and an encircling wall around 8000 BCE, structures whose purpose, defensive or communal, is still genuinely argued over. Across this whole region, people were taking the decisive early steps of farming. The “founder crops” of the Old World — emmer and einkorn wheat, barley, lentil, pea, chickpea, bitter vetch, and flax — were under cultivation, though full domestication was a slow business worked out over many generations. Sheep and goats were being domesticated in the highlands in roughly this window, with cattle and pigs to follow; the dog, by contrast, had been a human companion for many thousands of years already.

These were villages of at most a few hundred people, their round houses giving way over the following centuries to rectangular ones.

The picture is fuller than these two famous sites alone. The same centuries saw a constellation of early villages and ritual centres strung across the northern Fertile Crescent — at Çayönü and Aşıklı Höyük in modern Turkey, alongside the Taş Tepeler sites, and at Jerf el-Ahmar, Mureybet, Tell Qaramel, and Abu Hureyra along the rivers of modern Syria. Archaeologists divide the era into an earlier Pre-Pottery Neolithic A (roughly 9700–8700 BCE), when communities were still largely foraging while experimenting with cultivation, and a later Pre-Pottery Neolithic B (roughly 8700–6500 BCE), when farming villages, rectangular houses, and a shared repertoire of symbols and goods — including obsidian carried hundreds of kilometres — knit the region into what is sometimes called the “PPNB interaction sphere”. The art is extraordinary for its antiquity: the life-sized human statue known as Urfa Man, from modern Şanlıurfa and dated to around 9000 BCE, is the oldest known statue of a person, and within the following two thousand years the inhabitants of ’Ain Ghazal, near modern Amman in Jordan, were modelling large, wide-eyed human figures in lime plaster. As elsewhere, we can describe what these people made and how they lived with growing precision, while remaining honest that the meaning of their imagery is lost to us.

The later significance of this single arc of land is difficult to overstate, and it bears directly on why it commands our attention. The same Fertile Crescent that, ten thousand years ago, gave humanity its first farms and its first monumental architecture would, in the millennia that followed, give it the first cities and the first writing — in the Sumer of southern Iraq — and later still the three Abrahamic faiths, Judaism, Christianity, and Islam, whose adherents today number more than half the living world. It is essential, however, to keep the chronology straight, for the distances in time are immense. The figures and narratives of the Abrahamic scriptures are set in these same lands — Ur and Harran, Canaan, Egypt, and Arabia — but in ages utterly removed from the world of this essay: where historians attempt to place the age of Ibrāhīm (Abraham) in time at all, they look no earlier than the early second millennium BCE, more than six thousand years after the world described here, while the scriptures themselves belong to later ages again — the Hebrew Bible taking shape across the first millennium BCE, the Christian Gospels written in the first century of the Common Era, and the Qur’an revealed in the seventh. Popular writing that ties Göbekli Tepe to the Garden of Eden or to the Flood should be set firmly aside: such claims rest on neither archaeological evidence nor scriptural warrant, and belong to speculation rather than to history or to orthodox tradition. What is true, and remarkable enough without embellishment, is that no region on Earth has been so continuously consequential to the human story — yet every one of those later glories lay in the unimaginably distant future of the foragers and first farmers who lived here ten thousand years ago.

East Asia: modern China and Japan

In the lower reaches of the Yangtze River, in Zhejiang province in eastern China, the people of the Shangshan culture (dated to roughly 10,000 to 8,200 calendar years ago) were among the first societies anywhere to manage and cultivate rice. The evidence, much of it published in the Proceedings of the National Academy of Sciences, includes rice phytoliths (microscopic particles of silica that form inside living plant tissue and survive in the soil long after the plant itself has rotted away), pottery deliberately tempered with rice husks and chaff, settled occupation, and even residues indicating the fermentation of rice into an early alcoholic drink. As in the Near East, the domestication of the staple was a protracted process; fully domesticated rice emerged only some millennia later. Notably, these early rice cultivators already made pottery — a reminder that the elements we lump together as “the Neolithic” did not arrive everywhere as a single bundle. In the drier north of modern China, other communities were beginning to exploit and manage millet, the second great cradle of East Asian agriculture. In the Japanese archipelago, meanwhile, the Jōmon — settled, prosperous hunter-fisher-gatherers — were producing some of the oldest pottery in the world, made long before farming reached the islands.

South Asia: modern Pakistan and India

The subcontinent at 8000 BCE was a land of hunter-gatherers, and it is important to be clear about the chronology, because this is where the Indus civilisation would one day arise. The earliest known farming settlement in the region, Mehrgarh, on the Kacchi plain of Balochistan in modern Pakistan near the Bolan Pass, dates from around 7000 BCE — roughly a thousand years after our window. At 8000 BCE, in other words, even the deepest agricultural roots of South Asia had not yet been put down. The great cities of the Indus — Harappa and Mohenjo-daro, in modern Pakistan — lay some five and a half thousand years in the future, in their mature form between about 2600 and 1900 BCE. The people living in the region ten thousand years ago foraged, fished, and hunted, and left a faint archaeological trace by comparison with the spectacular record of their distant successors.

The Nile and the Green Sahara: modern Egypt, Sudan, and Niger

Here too the chronology corrects a common assumption. At 8000 BCE there were no pharaohs, no pyramids, and no farming along the Nile; agriculture would reach the Egyptian Nile valley only around 6000 BCE, and the unified Egyptian state not until about 3100 BCE — some five thousand years later. The people of north-east Africa at this date were hunter-fisher-gatherers, and many of them did not live along the Nile at all but across the then-green Sahara to its west. The richest record of them comes from Gobero, in the Ténéré desert of modern Niger, the oldest and largest known cemetery in the Sahara, founded around 8000 BCE. Its earliest occupants, the Kiffian, were tall hunter-fisher-gatherers — some men and women reaching close to six feet — who colonised the greened desert during its wettest phase, between roughly 10,000 and 8,000 years ago. They speared metre-long Nile perch with barbed bone harpoons in deep freshwater lakes, made decorated ceramics, and shared their world with elephants, giraffes, hippopotami, crocodiles, and pythons. The discovery, dated by nearly eighty radiocarbon measurements, is among the most vivid windows we have onto African life during the African Humid Period.

The Americas: modern United States, Mexico, and Chile

By 8000 BCE the peopling of the New World was already complete, from the Arctic to the southern tip of South America; the oldest securely dated site, Monte Verde in southern Chile, is older still, at about 10,500 BCE. The great Ice Age mammals of the Americas — mammoth, mastodon, giant ground sloth, and the rest — had by now largely vanished. The famous Clovis and Folsom traditions of fluted-point hunters had passed, succeeded across the Great Plains of modern North America by the Late Palaeo-Indian, or Plano, complexes (“Palaeo-Indian” being the standard term for the earliest inhabitants of the Americas; the named complexes — Agate Basin, Hell Gap, Scottsbluff, Eden — are simply distinctive styles of stone spear-point). These were highly skilled hunters who pursued a now-extinct giant bison, Bison antiquus, using the spear-thrower, or atlatl (a hooked throwing-stick that extends the arm and hurls a dart with far greater force), and communal techniques to drive and trap herds. Elsewhere people were turning increasingly to smaller game, fish, and plant foods, the shift that marks the opening of the Archaic period (the long era of diversified, broad-based foraging that followed the age of the big-game hunters). Crucially — and this bears on the question of contact addressed below — the Americas after the drowning of the land bridge developed in near-total isolation from the rest of humanity.

Europe and the British Isles: modern United Kingdom, Denmark, and beyond

Across a continent steadily reforesting as the climate warmed — birch and pine giving way to hazel, oak, and elm — lived the foragers of the Mesolithic (the “Middle Stone Age”, the term for the post-glacial hunter-gatherer cultures falling between the Ice Age Palaeolithic and the farming Neolithic). Across the North European Plain and the drowning lowlands of Doggerland these were largely the people of the Maglemosian tradition (the dominant hunter-gatherer culture of the region at this time, named after a bog site in Denmark), who fished, fowled, and hunted with microlith-tipped weapons and barbed antler points. In Britain, still tethered to Europe by Doggerland, the outstanding site is Star Carr, a lakeside settlement in North Yorkshire occupied between about 9300 and 8480 BCE, in the centuries immediately before our window. Its waterlogged deposits have preserved an extraordinary record: a substantial timber structure among the oldest known in Britain, red-deer antler “frontlets” that may have been worn on the head, an engraved shale pendant regarded as the oldest known art in Britain, worked amber, haematite, and iron pyrites, and the carcass of a domestic dog.

The most celebrated individual of the British Mesolithic is Cheddar Man, recovered from Gough’s Cave in Cheddar Gorge, Somerset, and directly radiocarbon-dated, on two occasions, to around 8300 BCE — squarely within our window. He is the oldest near-complete human skeleton found in Britain: a hunter-gatherer about 1.66 metres tall who died in his twenties. Nearby Aveline’s Hole, also in Somerset, is Britain’s oldest dated cemetery, holding the remains of at least twenty-one men, women, and children placed there around 8400 to 8200 BCE.

Australia and New Guinea

In the far south, Aboriginal Australians — present for fifty thousand years or more, among the longest continuous human cultures on Earth — were living through the breakup of their world by water, as rising seas divided the single great continent geologists call Sahul into modern Australia, New Guinea, and Tasmania. And in the highlands of modern Papua New Guinea, at Kuk Swamp, we find one of the genuine surprises of world prehistory: the earliest evidence for the deliberate working of the land — a buried channel and old land surface associated with plant management — dates to about ten thousand years ago, with unambiguous drained-field cultivation of taro and banana following by some seven thousand years before the present. New Guinea was, independently, one of humanity’s cradles of agriculture.

Because the question is so natural, it deserves a direct and consolidated answer: at 8000 BCE, the great river-valley civilisations of antiquity did not exist, and would not for thousands of years. Placing each in time is the single most effective way to grasp the antiquity of the period we are describing.

In modern Iraq, the land that would become Sumer — the world’s first urban civilisation — was, at 8000 BCE, not yet even settled by farmers. The southern Mesopotamian plain (the flat lowland of southern Iraq watered by the Tigris and Euphrates rivers, “Mesopotamia” being Greek for “between the rivers”) was first colonised by irrigation farmers only in the Ubaid period, from around 6500 BCE; the earliest city, Eridu, was founded around 5400 BCE; true urban life flowered in the Uruk period, around 4000 to 3100 BCE; and writing — the proto-cuneiform of Uruk, the earliest, picture-based form of cuneiform, the wedge-shaped script later pressed into clay tablets — appeared only around 3400 to 3200 BCE. The famous Sumerians, in short, postdate our window by some four and a half to five thousand years. (It is also worth noting that the Pre-Pottery Neolithic sites described above lay in the northern and western reaches of the Fertile Crescent, in modern Turkey, Syria, and the Levant — not in the southern Iraqi plain that would later cradle Sumer.)

In modern Egypt, the picture is much the same. Farming reached the Nile valley only around 6000 BCE, brought in part by people from the Near East; the predynastic Badarian and Naqada cultures flourished in the fifth and fourth millennia BCE; hieroglyphic writing emerged around 3400 to 3200 BCE; and the unified Egyptian kingdom of the pharaohs began around 3100 BCE.

Pharaonic Egypt postdates 8000 BCE by roughly five thousand years.

In modern Pakistan, as already noted, the earliest farming village, Mehrgarh, dates from around 7000 BCE, and the mature cities of the Indus civilisation from around 2600 to 1900 BCE — between one and five and a half thousand years after our window.

The lesson is plain. The civilisations that dominate the popular imagination of “the ancient world” are, from the vantage of 8000 BCE, events of the distant future. What we are looking at ten thousand years ago is the world before all of that: a planet of foragers and the very first farmers, in which the most elaborate human achievements were a stone enclosure in modern Turkey and a tower in the Jordan Valley.

The question is reasonable but it conceals a trap, and the honest answer must acknowledge both. “Advanced” embeds a hierarchy of values — it tends to reward sedentism, monuments, food production, and dense settlement, and to overlook the deep knowledge, mobility, and sophistication of foragers. With that caution stated plainly, the answer by the conventional measures of social complexity is reasonably clear. The communities of the Fertile Crescent — the builders of Göbekli Tepe and its sister sites, and the early farming villages of the Levant and modern Turkey and Syria — possessed the fullest combination of innovations found anywhere on Earth at the time: permanent or near-permanent settlement, the cultivation of crops and herding of animals, monumental architecture, sculpture and complex symbolism, and participation in long-distance exchange networks. By that yardstick they were the most elaborate societies of the age, and they sat at the leading edge of the transformation that would, far in the future, lead to cities and states.

But the yardstick is ours, not theirs. The Jōmon of modern Japan built settled, populous communities and made some of the world’s first pottery while remaining hunter-fisher-gatherers; the Kiffian of the green Sahara mastered a rich aquatic and savannah environment; the Aboriginal Australians sustained one of the longest continuous cultural traditions known. These peoples were not “behind” the farmers of the Fertile Crescent on a single ladder of progress; they were following different paths suited to different worlds. The framing of a race towards civilisation, with a clear leader, is a modern imposition the evidence does not require, and it is more accurate to speak of several sophisticated societies, of which the Fertile Crescent’s was the most socially complex by the particular measures we happen to prize.

Here the evidence is unusually good, and it points to a clear answer at two different scales. Within regions, contact could be surprisingly extensive. The best evidence comes from obsidian — the black volcanic glass prized for tools before the use of metals — whose chemistry betrays its precise volcanic source. In pioneering work in the 1960s, Colin Renfrew and his colleagues J. E. Dixon and J. R.

Cann showed that obsidian was moving across the Near East along established routes as early as 8000 BCE. The hard particulars are striking: at the Pre-Pottery Neolithic site of el-Hemmeh in modern Jordan, dated to the mid-to-late ninth millennium BCE (around 8500–8000 BCE), the obsidian came from sources in Cappadocia, in central Turkey, over 900 kilometres to the north. Cappadocian obsidian travelled the length of the Levant and even reached the island of Cyprus across open water. These are not chance encounters but sustained networks of exchange, carrying not only goods but, almost certainly, ideas and techniques — the means by which the elements of the Neolithic spread from community to community.

Between the great land masses, however, there was effectively no contact at all. After the drowning of the land bridge from Asia, the peoples of the Americas developed in isolation from the rest of humanity, as did, to a large degree, the inhabitants of Australia and New Guinea. There was no connection between the rice cultivators of China and the farmers of the Near East, still less between the Old World and the New. The “known world” of any person alive in 8000 BCE was a regional one — wide enough, in the Near East, to span hundreds of kilometres of exchange, but nowhere approaching the global. The independent invention of agriculture in at least four separate centres — the Fertile Crescent, China, Mexico, and New Guinea — is itself the clearest proof of this isolation: the same idea arose more than once because no one carried it between them.

The short and well-founded answer is no: the racial categories of modern usage are a recent social construction, and applying them to the people of 8000 BCE is both anachronistic and misleading. Several things can be said with confidence, and the boundary between them and what cannot be known is worth drawing carefully.

What is established is that the people of ten thousand years ago were a single biological species, Homo sapiens, exactly as we are, and that the human variation of the period was clinal — that is, it graded continuously across geography rather than falling into discrete, sharply bounded types. The neat boxes of folk racial taxonomy do not correspond to any natural division of humanity, then or now. It is also established, and important, that the visible traits we casually attach to those boxes were distributed quite differently in the past. Ancient DNA shows that the genes responsible for pale skin in Europeans had not yet become widespread; the western European hunter-gatherers of this period, Cheddar Man among them, appear from their genomes most probably to have had dark skin together with pale or blue eyes. Pale European skin spread later, carried by the Near Eastern farmers and steppe pastoralists who arrived in the following millennia. The modern association of particular skin colours with particular continents is, in large part, a development of the last several thousand years.

Furthermore, the genetic work of recent decades — much of it synthesised by David Reich and others — has shown that the broad population groupings we recognise today were largely formed after 8000 BCE, by the mixing of more ancient and more sharply distinct ancestral populations. The Europeans of historical times, for example, are a blend of Mesolithic hunter-gatherers, incoming Neolithic farmers, and later steppe peoples — a fusion that had not yet happened in 8000 BCE. To project a modern label such as “white” or “Asian” back onto the inhabitants of the early Holocene is therefore to read a recent and contingent arrangement of humanity into a past that did not contain it.

What we cannot say is what most of these people looked like in any detail, because securely analysed ancient genomes from 8000 BCE remain few and unevenly distributed across the globe. We know something of European and Near Eastern hunter-gatherers; we know far less, genetically, of the peoples of Africa, southern and eastern Asia, the Americas, and Australia at this date. Confident statements about the appearance of most of the world’s population ten thousand years ago go beyond the evidence. The safe conclusion is the negative one: our racial vocabulary does not fit them, and was not built for them.

This is the question on which we must be most candid: we do not know, and in large part we cannot know. The reason is straightforward. Writing — the only direct record of a language — did not exist anywhere in 8000 BCE and would not appear for another four and a half thousand years. Everything we know about unwritten ancestral languages we recover by the comparative method — the technique of reconstructing a vanished parent language by working backwards from the systematic resemblances among its surviving descendants (the way the shared features of French, Spanish, and Italian point back to Latin). But that method has a horizon. The deepest language families that can be securely identified are on the order of six to eight thousand years old; Proto-Indo-European and Proto-Uralic are usually dated to around 4000 BCE, and Proto-Afroasiatic is somewhat older.

Beyond roughly ten thousand years, the signal of common descent is swamped by chance resemblance and the accumulated noise of change, and most historical linguists regard the attempt to reach further — the “long-range” macro-families such as “Nostratic” or “Amerind” — as methodologically unsound.

The upshot is that the languages actually spoken in 8000 BCE lie at or beyond the very edge of what reconstruction can reach, and in nearly every case beyond it. The tongues of the builders of Göbekli Tepe, of the Shangshan rice cultivators, of the Kiffian of the green Sahara, of Cheddar Man — all are irretrievably lost, unrecorded and unreconstructable. We cannot name them, group them, or describe their grammar.

What we can say rests on firmer ground. The people of this age were anatomically and cognitively modern, and possessed full human language exactly as we do; there is no question of grunts or proto-speech. We can also reasonably infer that linguistic diversity was very great — quite possibly greater than today’s several thousand languages — because the world was populated by many small, scattered, and relatively isolated communities, precisely the conditions under which languages multiply and diverge. But the particular words and structures of that vanished diversity are gone. On the languages of ten thousand years ago, intellectual honesty requires the admission of near-total ignorance.

It follows from the above that the communication of 8000 BCE was overwhelmingly oral and gestural, conducted in fully developed spoken languages, and that nothing was written down. This matters enormously for the period’s character: it was a world without records, in which all knowledge — of the land, the seasons, kinship, craft, story, and the proper order of things — was held in living memory and passed from person to person by speech, demonstration, and ritual. A society without writing is not a society without intellect or law or history; it is one whose intellect, law, and history leave no direct textual trace.

There are, however, two important qualifications, and both must be stated with care about what is fact and what is interpretation. The first is that symbolic and artistic communication certainly existed and is directly attested. The carved animals of Göbekli Tepe, the engraved shale pendant of Star Carr, and the personal ornaments of beads and pierced shells found across the world all transmitted meaning of some kind — of identity, status, group membership, or belief. That such things carried meaning is hard fact; what specific meaning they carried is, in almost every case, beyond us. The most remarkable single object is the so-called Urfa Man, found at Şanlıurfa in modern Turkey and dated to around 9000 BCE — the oldest known life-sized sculpture of a human being. That it represents a person, or a being, is plain; whom or what it was meant to be, we do not know.

The second qualification concerns the deep prehistory of writing itself. From around 8000 BCE, small clay tokens of simple geometric shapes begin to appear across the Near East. On the influential interpretation associated with Denise Schmandt-Besserat, these served as counters for commodities — so many tokens for so many measures of grain or head of livestock — and the system slowly elaborated over the following millennia until, around 3400 to 3200 BCE, it gave rise to the written numerals and proto-cuneiform of Mesopotamia. If that reading is correct, then the humble accounting tokens of 8000 BCE are the most distant ancestor of all writing. It must be stressed, however, that this is an interpretation of the tokens’ function, not a certainty, and that the tokens of our period are emphatically not writing: they are, at most, the first faint seed of a technology that would not flower for four and a half thousand years.

This is a domain where the distinction between observation and inference is sharpest, and where it is easiest to say too much. What we observe — the hard evidence — is a set of behaviours, and a striking set it is. The careful, deliberate treatment of the dead is the clearest. Across the Near East, the green Sahara, and Europe, people of this age buried their dead with evident intention: in fixed cemeteries such as Gobero in modern Niger and Aveline’s Hole in Somerset; with grave goods of ornaments, tools, and sometimes food; sometimes sprinkled with red ochre (a natural reddish-brown earth pigment used by humans for tens of thousands of years); and sometimes with the body returned to and rearranged after the flesh had gone. In the centuries just after our window, the communities of the Levant took to detaching the skulls of the dead and remodelling their faces in plaster — the famous “plastered skulls” of Jericho and ’Ain Ghazal in modern Jordan — a practice whose roots in careful mortuary treatment are already present around 8000 BCE. To these we may add the monumental enclosures of Göbekli Tepe and the Taş Tepeler, with their charged animal imagery and their T-shaped pillars carved, in some cases, with human arms and hands, as though the pillars themselves represented beings; and the figurines and statues, the Urfa Man foremost among them.

What all this demonstrates, beyond reasonable doubt, is that the people of 8000 BCE engaged in ritual — in patterned, symbolic, non-utilitarian behaviour surrounding death, the body, animals, and certain special places — and that they invested great labour in it. This is the firm conclusion, and it is a profound one: the religious impulse, broadly understood as ritual and symbolism, is plainly older than farming and older than cities.

What we cannot do is read their beliefs. We do not know whether they conceived of gods, or spirits, or an afterlife; what the animals on the pillars signified; whether the dead were venerated, feared, or simply mourned; or whether anything resembling organised religion, with doctrine and priesthood, existed. The temptation to fill the silence by reading later, historically attested religions back into the period is strong and must be resisted, for there is no evidence to support it. We observe the acts of what may have been a religious life; the content of that life — its theology, if it had one — is lost with the languages that would have expressed it. Here, as with so much of this remote age, the most important intellectual discipline is to hold firmly to what the ground reveals and to refuse to invent the rest.

The world of 8000 BCE was, in its fauna, markedly poorer than the one of a few millennia earlier. Across northern Eurasia and the Americas, the spectacular megafauna (literally the “big animals” — the giant land mammals) of the late Pleistocene (the long Ice Age epoch that immediately preceded our own Holocene) — woolly mammoth and rhinoceros, the giant sloths and the elephantine creatures of the New World — had largely disappeared. Whether the decisive cause was the rapid climatic and vegetational upheaval of the deglaciation, the pressure of skilled human hunters (the “overkill” hypothesis), or some compounding of the two, remains one of the genuinely open questions of prehistory, and it would be dishonest to present any single answer as settled. A few isolated populations lingered in island refuges for thousands of years more, but on the continents the great beasts were, by this date, essentially gone.

Materially this was a world of stone, bone, antler, wood, fibre, and hide. Its characteristic toolkit relied on microliths — small, finely worked flint blades set in series to make composite weapons and implements — together with the bow and the spear-thrower. In the farming regions, the ground or polished stone axe was coming into its own as the essential instrument for clearing forest. Pottery, often taken as the hallmark of the era, was in fact unevenly distributed: already made by the rice cultivators of China, the Jōmon of Japan, and the hunter-fishers of the green Sahara, yet absent among the emphatically “Pre-Pottery” early farmers of the Near East, who stored and cooked without fired clay for centuries. Metals were effectively unused; the smelting of ores lay several thousand years in the future. The dog was the one near-universal domestic animal.

The people of 8000 BCE were fully modern in body.

Where their health can be read from bones and teeth, foragers were generally robust and varied in diet, while the earliest cultivators were beginning to show the first signs of the dental and skeletal stresses that a narrower, cereal-heavy, more crowded and sedentary life would bring. One small universal is worth recording: like every human being then alive, they could not digest milk into adulthood. The genetic adaptation of lactase persistence, which allows many adults today to drink milk, had not yet arisen, and would spread only later, alongside the keeping of dairy herds.

How can anyone speak with confidence about a world ten thousand years gone? The dates that anchor this essay rest chiefly on radiocarbon dating, the single most important tool in the archaeologist’s kit. Its principle is elegant. Cosmic rays striking the upper atmosphere continually create a radioactive form of carbon, carbon-14, which living plants and animals absorb along with ordinary carbon throughout their lives. At death the intake stops, and the carbon-14 begins to decay back into nitrogen at a fixed, known rate — half of it disappearing roughly every 5,730 years. Measure the proportion of carbon-14 left in a scrap of charcoal, bone, or grain, and the time elapsed since the organism died can be calculated; modern accelerator mass spectrometry performs this on samples of only a few milligrams, and the method reaches back about fifty thousand years. Because the atmosphere’s carbon-14 has not been perfectly constant, raw “radiocarbon years” are converted into true calendar years by calibration against independently dated records — above all the growth rings of ancient trees — which is why such dates are quoted as “calibrated years before present” (cal BP), each carrying a stated margin of error rather than a single, falsely precise figure.

Radiocarbon does not stand alone, and it is the agreement of several independent methods that turns plausible estimates into firm knowledge. Dendrochronology — the counting and cross-matching of tree rings, each ring a single year — yields exact calendar dates and underpins the radiocarbon calibration curve, with oak and pine chronologies now reaching back more than ten thousand years. The great ice sheets of Greenland and Antarctica preserve their own annual layers, countable like tree rings, while the ratio of oxygen isotopes locked within the ice serves as a thermometer of the ancient climate; it is in the Greenland ice that the beginning of the Holocene is formally fixed. Annual sediment layers in lakes, called varves, and the growth bands of corals tell the same story by other means; luminescence and uranium-series techniques date materials and spans of time beyond radiocarbon’s reach; and the recovery and sequencing of ancient DNA from old bone now traces kinship and ancestry directly. None of these methods delivers perfect, unconditional certainty — every date carries a quantified uncertainty — but where charcoal from a hearth, the rings of a buried timber, and a band of polar ice independently converge on the same age, the result is about as secure as knowledge of the deep past can be.

A closing word on the foundations of all this, for the subject demands humility. The firm ground consists of the dated and the physical: the radiocarbon-dated settlements and skeletons named throughout, the annual climate record of the Greenland and Antarctic ice, the sea-level reconstructions drawn from dated shorelines, the genomes recovered from ancient bone, the plant and animal remains recovered by excavation, and the traded materials such as obsidian whose sources can be chemically fingerprinted. From these we know, with real confidence, where people lived, what they ate and built, how the climate and coastlines stood, that they exchanged goods across hundreds of kilometres, that they buried and memorialised their dead, and — in broad terms — how some populations were related.

What remains inferred, and properly uncertain, is the harder and more human matter: the function of the monuments, the meaning of the symbols, the content of belief, the structure of society, and the precise size and distribution of the population. And what is, on present evidence, simply unknowable is the inner life of the age — above all the languages spoken and the beliefs held, which vanished with the breath of the people who spoke and held them, leaving no direct trace.

It is fitting to end on that frontier between knowledge and ignorance, because it is the most honest place to stand. New discoveries continue to push it outward, and occasionally to overturn an assumption, as the recognition of the Taş Tepeler sites did within a single generation. But the broad and evidenced shape of the world ten thousand years ago is not in serious doubt: a planet freshly released from the ice, its seas rising and its coasts dissolving, its deserts in bloom and its great beasts mostly gone; peopled, sparsely, by perhaps five million fully modern human beings in small communities on every habitable continent; a few of them — in the uplands of modern Turkey, Syria, and the Levant, and along the rivers and wetlands of China, Mexico, and New Guinea — beginning, independently and without knowing where it would lead, the slow work of farming that would, thousands of years later, give rise to everything we call the ancient world.

The Earth of 3000 BCE was, in its broad physical geography, very nearly the world we know. The great ice sheets that still lay over modern Canada and Scandinavia ten thousand years ago had at last melted away; the post-glacial rise of the sea, which a few millennia earlier had been drowning coastlines within a single lifetime, had slowed almost to a halt. The most detailed reconstructions, such as those of Kurt Lambeck and colleagues, place sea level by this date within a few metres of its present mark, with some regions standing fractionally higher than today — a mid-Holocene high-stand whose traces survive in raised beaches around the world. The coastlines a sailor of 3000 BCE would have known are, in the main, recognisably ours.

The consequence for these islands is that Britain was now, unambiguously, an island. The low country of Doggerland, across which one could once have walked dry-shod from Yorkshire to the Netherlands, had been finally submerged around 6200 to 6000 BCE, some three thousand years before our window; the North Sea had closed over it, and Britain has been severed from the Continent ever since. The map of the seas had, in most places, settled into its modern form.

Most striking of all is a reversal of the picture we drew for ten thousand years ago. Then, the Sahara was green — a landscape of grassland, seasonal rivers, and great freshwater lakes, grazed by elephants and hippopotami and peopled along its waters. By 3000 BCE that African Humid Period was ending. As the Earth’s orbital geometry shifted and the monsoon withdrew southward, the lakes shrank, the grasslands browned, and the largest hot desert on Earth came into being more or less as we know it. The desiccation was, in places, abrupt; and its human consequence was momentous, for as the Sahara dried it drove its herders and fishers eastward and southward, many of them towards the one reliable water left to them — the Nile. The concentration of population along that river, as the archaeologists Rudolph Kuper and Stefan Kröpelin have argued, is bound up with the very rise of Egyptian civilisation. The desert that had been a garden in our last survey was, by this one, helping to make a kingdom.

Any figure for the human population of 3000 BCE remains an estimate, but it is a firmer one than for ten thousand years ago, because the centres of dense settlement can now be identified and, in a few cases, counted. The commonly reproduced reconstructions — the History Database of the Global Environment (HYDE) and the older syntheses of Jean-Noël Biraben and of Colin McEvedy & Richard Jones — place world population around this date on the order of fourteen million, within a plausible range of perhaps seven to twenty million. That is roughly a trebling since the world of ten thousand years ago, and the cause is no mystery: farming feeds more mouths from a given acre than foraging can, and where farming took hold, people multiplied.

As before, the distribution of those millions is far less certain than the total, and the published regional figures deserve humility. But the broad shape is now clearer than it was, because agriculture had created, for the first time, regions of genuine density. The thickest concentrations of humanity lay where irrigation or rich rains supported intensive farming: along the Tigris and Euphrates in southern Mesopotamia; up the Nile; across the plains of the Indus and its tributaries; and in the millet lands of the Yellow River and the rice lands of the Yangtze in modern China. Around these cores the older world persisted — herders on the steppe and the drying Sahara, farmers thinly spread across Europe, and hunter-gatherers still holding the greater part of the planet’s surface, from the forests of Siberia and the Americas to the whole of Australia. What had changed since our last survey was not that the world had filled — by any modern measure it remained almost empty — but that it had begun, in a few river valleys, to crowd.

Before turning to the peoples themselves, it is worth recalling the regions through which we viewed them ten thousand years ago, for they are the same culture-areas, and the point of returning to them is to watch them diverge. Then they were seedbeds, scarcely distinguishable in their level of complexity. Now their paths had begun to separate sharply. Chief among them is still the Near East, and at its heart Mesopotamia, where the seedbed had become a field of cities. Beyond it lie Egypt and the Nile, by now a unified kingdom; South Asia, where the towns of the future Indus civilisation were forming; East Asia, the basins of the great Chinese rivers; the drying Sahara and sub-Saharan Africa; Europe, a continent of farmers, megaliths, and the first metal; the Americas, where in one place on the Peruvian coast monuments were rising in complete isolation; and Australia with New Guinea. These are the same roots we traced before — but five thousand years on, and grown into strikingly different things. The survey below treats each in turn.

Mesopotamia and the First Cities: modern southern and central Iraq

The most consequential development anywhere on Earth at this date was unfolding on the flat alluvial plain of southern Iraq, between the Tigris and the Euphrates — the land the Greeks would later call Mesopotamia, “between the rivers,” and that history knows as Sumer. In our last survey this plain was not even settled by farmers; it was colonised only from around 6500 BCE, in the Ubaid period, by people who learned to lift the rivers’ water onto the fields. By 3000 BCE that long preparation had culminated in something without precedent: the city. Uruk, the greatest of them, sprawled across some two and a half square kilometres and may, at its height in these centuries, have held a population in the tens of thousands — in all likelihood the largest single settlement then existing on the planet. It was a true urban society, with monumental temple precincts raised on great platforms (the ancestors of the later stepped towers, or ziggurats), with specialised crafts, with social classes, and with the beginnings of kingship.

Above all, it had writing. From the temple administrations of Uruk, around 3300 to 3100 BCE, come the oldest written documents in the world: the proto-cuneiform tablets, in which signs pressed into clay recorded quantities of grain, livestock, and labour. They are, almost without exception, accounts — receipts and ration lists, not chronicles or poems — and that is itself a profound fact about the dawn of literacy, to which we shall return. Over the following centuries the script was elaborated, by about 2900 to 2700 BCE, into the wedge-shaped cuneiform that would serve the region for three thousand years. The language it first recorded was Sumerian, a tongue with no known relatives anywhere on Earth — the first human language ever committed to writing, and for that reason the earliest whose words we can actually read. To this world we owe, on the evidence presently available, not only the first city and the first script but the first wheel set to vehicles, the ox-drawn plough, and the earliest large-scale bureaucratic state. The famous Sumerian kings of legend — Gilgamesh of Uruk among them — belong to the centuries just after our window, around 2700 BCE; the world of 3000 BCE is the threshold across which they would shortly step.

Egypt and the Nile: modern Egypt and Sudan

Some twelve hundred kilometres to the west, along the Nile, the second of the world’s pristine civilisations was taking shape, and on a strikingly compressed timetable. Farming had reached the Egyptian Nile valley only around 6000 BCE; yet by about 3100 BCE the many communities of the river had been welded — whether by conquest or by slower coalescence is still debated — into a single kingdom stretching from the Delta to the first cataract, traditionally under a ruler the later Egyptians remembered as Narmer or Menes. By 3000 BCE, therefore, Egypt was a unified state in its First Dynasty, governed from the new capital at Memphis, its kings buried in great mud-brick tombs at Abydos in the south and at Saqqara. Hieroglyphic writing appears in Egypt at almost the same moment as cuneiform in Sumer — the earliest securely dated signs, on small ivory labels from a royal tomb at Abydos, fall around 3300 to 3100 BCE — and it served from the first the cult of a divine king, whose person was identified with the falcon-god Horus. It is essential to keep the monuments in their right order, for popular imagination runs ahead of the evidence. At 3000 BCE there were no pyramids. The first of them, the Step Pyramid raised for King Djoser by his architect Imhotep at Saqqara, lay a little over three centuries in the future, around 2670 BCE; the great pyramids of Giza, later still, around 2560 BCE. The Egypt of our window is the kingdom before the pyramids — already literate, already ruled by a god-king, already monumental in its tombs, but with its most celebrated works yet to come. To the south, in Lower Nubia, the contemporary communities archaeologists call the A-Group traded gold, ivory, and incense northward and stood in a relationship with early Egypt that was by turns commercial and hostile.

South Asia: modern Pakistan and India

In the valley of the Indus and along its tributaries, the chronology once again corrects a natural assumption, for this is where the third of the Old World’s great river civilisations would arise. At 3000 BCE it had not yet arisen. The period is what archaeologists call the Early Harappan — a landscape of substantial farming villages and small towns, such as those of the Kot Diji and Ravi phases, with craft production, long-distance exchange, and the first hints of the planning and standardisation that would later become the hallmark of the region. But the mature cities of the Indus — Harappa and Mohenjo-daro, with their grid-planned streets, their fired-brick drains, and their enigmatic seal-script — lay some four centuries ahead, flourishing between about 2600 and 1900 BCE. The earliest farming settlement of the region, Mehrgarh on the Kacchi plain of Balochistan, was still occupied at this date. South Asia in 3000 BCE was a civilisation in gestation: the towns were there, the foundations were being laid, but the cities had not yet been built.

East Asia: modern China and Japan

In China the picture at 3000 BCE is of complex Neolithic societies that, in one case, stood on the very edge of statehood. In the lower Yangtze, around the marshes of modern Zhejiang, the Liangzhu culture (roughly 3300 to 2300 BCE) had built something extraordinary: a walled city of some three hundred hectares, surrounded by the earliest large-scale water-management system known anywhere in the world — a network of dams and reservoirs whose construction has been dated to around 3100 BCE. Liangzhu was a stratified society of rice farmers ruled by an elite that controlled the working of jade, and whose finest tombs were furnished with exquisitely carved jade discs and tubes (the bi and the cong) bearing a recurring deity-and-beast emblem. Many archaeologists now regard it as an early state in all but name, and it is on the strength of such evidence that the claim of “five thousand years of Chinese civilisation” is most seriously advanced. In the drier north, the millet-farming cultures of the Yellow River — the late Yangshao giving way to the Longshan — were raising rammed-earth walled towns and showing their own marks of growing hierarchy.

Two cautions are in order. First, China at 3000 BCE had no writing: the oracle-bone script of the Shang dynasty lay more than a millennium and a half in the future, around 1250 BCE, and the scattered marks on Neolithic pottery, though sometimes called proto-writing, do not constitute a system that can be read. Second, the historic dynasties did not yet exist: even the semi-legendary Xia is placed by tradition no earlier than about 2070 BCE.

Meanwhile, in the Japanese archipelago, the Jōmon continued the way of life we met ten thousand years ago — settled, prosperous hunter-fisher-gatherers, makers of some of the world’s oldest pottery, who would not take up farming in earnest for many centuries yet.

North Africa and Sub-Saharan Africa The drying of the Sahara is the great African fact of this age. The lakeside peoples we met at Gobero in the green desert ten thousand years ago belonged to a world that was, by 3000 BCE, closing; the cemeteries fell silent as the waters retreated, and human life withdrew towards the Nile, the southern Sahel, and the highlands. Along the Nile and in Nubia, as we have seen, this concentration helped feed the rise of states. Across the rest of the continent the evidence is thinner, and honesty requires us to admit how much less we know of sub-Saharan Africa at this date than of the literate river valleys. What can be said is that it remained a world of hunter-gatherers, fishers, and — increasingly — herders of cattle, sheep, and goats across the savannahs. The domestication of West Africa’s own crops, above all pearl millet in the Sahel, lay some centuries ahead, around 2500 to 2000 BCE; the great Bantu expansion that would carry farming and, later, iron across the continent’s south was further off still, beginning around the end of the second millennium BCE. Africa in 3000 BCE was, outside the Nile, a continent whose deep transformations were yet to come.

Europe and the British Isles: modern United Kingdom, Ireland, Malta, and the Alps

Europe at 3000 BCE was a continent of farmers without cities or writing, but it was neither static nor poor, and three things mark it out. The first is monumental stone. This is the great age of European megaliths: in Britain, the first phase of Stonehenge — a circular bank, ditch, and ring of pits — was laid out around 3000 BCE, the towering stone settings following some five centuries later; in Ireland, the vast passage tomb of Newgrange, with its midwinter-sunrise alignment, had been raised around 3200 BCE; in Orkney, the stone village of Skara Brae was inhabited; and on the islands of Malta the remarkable temple culture was building the great sanctuaries of Ġgantija and its fellows, among the oldest free-standing stone monuments on Earth. The second mark is metal: this is the Copper Age, and copper tools and ornaments were now in use across much of the continent. The third, and least visible to the eye but most consequential in the long run, was a great movement of people — the expansion, from the Pontic-Caspian steppe, of the pastoralist populations associated with the Yamnaya culture, who from around 3000 BCE began to move west and to reshape the genetic and probably the linguistic map of Europe, as we shall see.

The age has its celebrated individual, as the British Mesolithic had its Cheddar Man. He is the so-called Iceman, Ötzi, whose freeze-dried body emerged from a melting glacier in the Ötztal Alps and who died around 3230 BCE — squarely within our window. No human being of comparable antiquity is so intimately known. He was a man of about forty-five, some 1.6 metres tall; he carried a bow, a quiver of arrows, and an axe with a blade of nearly pure copper; his body bore more than sixty tattoos; his last meals included ibex and red deer meat with einkorn wheat; his arteries were already hardening, and his gut carried the ulcer bacterium and intestinal parasites. He did not die peacefully: an arrowhead lodged in his shoulder, and a recent restudy of his genome suggests he was dark-skinned, probably balding, and descended overwhelmingly from the early Anatolian farmers who had brought agriculture into Europe. He is the period made flesh.

The Americas: modern Peru, the United States, and Mexico

The Americas at 3000 BCE were still developing, as they had since the drowning of the land bridge, in complete isolation from the rest of humanity — which makes the events on one stretch of the Peruvian coast all the more remarkable. There, in the river valleys of the Norte Chico, or Caral-Supe, region north of modern Lima, the oldest civilisation of the Americas was taking form. Its people raised monumental platform mounds and great sunken circular plazas; the largest centre, Caral, with its complex of pyramidal mounds, reached its monumental peak around 2600 BCE, while the broader tradition reaches back towards 3000 BCE and earlier. What astonishes is the manner of it. This was a society that built monuments on an enormous scale yet made no pottery and, so far as we can tell, possessed no writing — though it may have kept records on knotted cords, an early ancestor of the later Andean quipu. That a civilisation should arise here, in the same centuries as Sumer and Egypt and in total ignorance of them, is among the most eloquent proofs we have of the independence of the human story across the hemispheres.

Elsewhere in the Americas the older patterns held, with their own surprises. In the lower Mississippi valley, at Watson Brake in modern Louisiana, hunter-gatherers had raised a complex of earthen mounds around 3500 BCE — monumental architecture built without farming, and far older than the better-known Poverty Point. In Mesoamerica, the long, slow domestication of maize was advancing, though the civilisations it would one day feed — the Olmec earliest among them — lay more than a thousand years ahead. The hemisphere was, as ever, on its own timetable.

Australia and New Guinea

In the far south, the Aboriginal peoples of Australia continued one of the longest unbroken cultural traditions on Earth, sustaining across an entire continent a way of life founded on hunting, gathering, and the deep management of land by fire — and, notably, never taking up farming, by what was in all likelihood a considered adaptation to their environment rather than any want of capacity. The dingo, the one notable newcomer to the continent, would arrive only a little after our window, around 1500 BCE. In the highlands of New Guinea, by contrast, the independent agriculture we glimpsed in its infancy ten thousand years ago at Kuk Swamp was now mature: drained-field cultivation of taro and banana was long established, making New Guinea one of the world’s genuine and wholly separate cradles of farming.

Because the question is so natural, it bears a direct and consolidated answer — and here the contrast with our last survey is the whole point. Ten thousand years ago, not one of the great river-valley civilisations existed, and all of them lay thousands of years in the future. Five thousand years ago, the picture is transformed: two of them have arrived, a third is forming, and a fourth is still distant. In southern Iraq, Sumer is a living civilisation: cities, temples, kingship, and the world’s first writing are all present at 3000 BCE, with the legendary kings only a century or two ahead. In Egypt, the unified kingdom of the pharaohs has just been founded, around 3100 BCE, literate and god-ruled, though the pyramids are still some three centuries away. In the Indus, the great cities have not yet risen — the Early Harappan towns of 3000 BCE would mature into Harappa and Mohenjo-daro only around 2600 BCE. And in China, complex society is far advanced — Liangzhu in the south stands at the threshold of statehood — but writing and the historic dynasties remain well over a millennium in the future. The lesson is the mirror image of the one we drew before. Where the world of ten thousand years ago was the deep prehistory that preceded the river valleys, the world of five thousand years ago is the very moment of their birth — the hinge on which prehistory begins, unevenly and in a few places only, to turn into history.

The question carries the same trap it did before — “advanced” smuggles in a hierarchy of values that rewards cities, writing, and states, and slights the deep sophistication of those who chose other paths — and the same caution must be entered. But with that said, the answer at 3000 BCE is less ambiguous than it was at 8000 BCE, because by now a genuinely new kind of society had appeared. By any conventional measure of social complexity, the city-states of Sumer and the unified kingdom of Egypt stood apart: they alone combined dense urban or centralised settlement, writing, organised states with kings and administrations, monumental architecture, craft specialisation, and metallurgy. They were, in that strict sense, the most complex societies the world had yet produced.

Yet the older humility still applies, and more pointedly than the textbook story allows. Liangzhu in the Yangtze delta built a great city and a vast hydraulic system without writing; the Norte Chico of Peru raised monuments to rival any in the Old World, in total isolation and without pottery or script; the megalith-builders of Atlantic Europe moved stones of staggering size; the herders of the steppe were mastering the horse. These were not laggards on a single road with Sumer in the lead. They were distinct experiments in how human beings might live at scale, and several of them arrived at monumentality and hierarchy by routes that bypassed the very things — the city, the written word — by which we are tempted to measure “advancement.” It remains more accurate to speak of several remarkable societies, of which the river-valley states were the most complex by the particular yardstick we happen to prize.

The same idea arose more than once because, between the hemispheres, no one carried it.

At the regional scale, contact in 3000 BCE was more extensive than ever before, and the evidence for it is unusually hard. The clearest witness is the movement of precious materials over great distances. Lapis lazuli, the deep-blue stone whose only significant ancient source was the mountains of Badakhshan in modern Afghanistan, was already reaching both Mesopotamia and Egypt — a journey of thousands of kilometres, passing from hand to hand along established routes. So too were copper, tin, silver, timber, and turquoise: Egypt drew cedar from Byblos on the Lebanese coast, and the bronze of the age depended on tin whose sources lay far from where it was worked. The networks carried more than goods. In late predynastic Egypt, around 3300 to 3100 BCE, scholars have long recognised Mesopotamian influence — in cylinder seals, in a distinctive style of niched mud-brick architecture, and in motifs such as the intertwined long-necked beasts and the “master of animals” that appear on carved objects like the Narmer Palette and the Gebel el-Arak knife. Whether by trade, travelling craftsmen, or the borrowing of ideas, the two earliest civilisations were not wholly unaware of one another.

Between the great land masses, however, the older isolation held absolutely. There was no contact whatever between the Old World and the New, nor any between the civilisations of western Asia and the rice-farmers of China, still less with the peoples of Australia. The single most powerful proof is the one we noted before, now even stronger: the independent invention of civilisation itself. Cities, monuments, and social hierarchy arose separately in Mesopotamia, in Egypt, on the Peruvian coast, and — in their own forms — in China, in the same broad span of centuries and in mutual ignorance. The same idea arose more than once for the same reason it did ten thousand years ago: because no one carried it between the continents. The “known world” of a Sumerian or an Egyptian had widened to embrace much of western Asia and the eastern Mediterranean; but it was a regional world still, and the globe remained, to every soul alive, unknown.

The answer is the same as it was for ten thousand years ago, and rests on the same well-founded grounds: the racial categories of modern usage are a recent social construction, and to apply them to the people of 3000 BCE is anachronistic and misleading. Humanity was then, as now and as ten thousand years before, a single biological species whose variation graded continuously across geography rather than falling into discrete, bounded types. The neat boxes of folk racial taxonomy answer to no natural division of mankind.

What the intervening five thousand years had begun to change is the very thing that makes the modern projection so tempting — the distribution of visible traits and of the deep ancestries beneath them. The genetic work of recent decades, much of it synthesised by David Reich and his colleagues, has shown that the broad population groupings we recognise today were still in the act of formation around 3000 BCE, and that this was one of the great ages of their mixing. The expansion of Yamnaya steppe pastoralists into Europe from about 3000 BCE is the clearest case: it poured a new ancestry into a continent that had been a blend of older hunter-gatherers and incoming farmers, and the modern European populations are the result of that very fusion — a fusion still under way in our window, not completed before it. Even the spread of pale skin in Europe, as the case of Ötzi reminds us, was incomplete at this date. The peoples of 3000 BCE were not yet sorted into the continental “types” of later folk imagination; they were being remade, by migration and admixture, into the populations from which those later appearances would emerge. To read modern racial labels back onto them is to mistake a process for a finished thing.

Here, for the first time in this series, the honest answer is not one of near-total ignorance — and the reason is writing. Because Sumerian was being committed to clay in the temples of Uruk, and Egyptian to ivory and stone along the Nile, we can for the first time name, and in part read, specific human languages of the period. Sumerian, as noted, is a language isolate, related to no known tongue living or dead; Egyptian belongs to the great Afroasiatic family, whose other branches include the Semitic, Berber, and Cushitic languages. A third early script, the still-undeciphered Proto-Elamite of south-western Iran, appears around 3100 BCE and records a language — Elamite, or a relative of it — that we can locate but not yet understand. For the first time, in other words, the linguistic map of part of the world is not a blank.

There is more. The middle Holocene is also, by the reckoning of most historical linguists, roughly the era of the languages that the comparative method can actually reach. Proto-Indo-European — the reconstructed ancestor of tongues from English and Latin to Persian and Sanskrit — is usually dated to around the fourth and third millennia BCE and associated, on the now-dominant steppe hypothesis, precisely with those Yamnaya pastoralists who were beginning to move outward in our window. Proto-Indo-European was therefore not some impossibly remote ancestor but, in a real sense, a language of roughly 3000 BCE, spoken on the steppe and on the verge of its great dispersal. The deepest securely reconstructed families lie at just this depth.

And yet the candour of the earlier essay must be preserved for almost everything else. Writing illuminated two river valleys; it left the rest of the planet as dark as before. The languages of Liangzhu and the Longshan towns, of the Norte Chico mound-builders, of the megalith-raisers of Atlantic Europe, of the herders of the green-turning-brown Sahara, of every Aboriginal Australian nation — all are unrecorded and, in the great majority of cases, unreconstructable. Linguistic diversity in 3000 BCE was surely still immense, very probably greater than today’s. But of that vast unwritten ocean of speech, only two or three coastlines had, by this date, been charted.

It follows that 3000 BCE was the moment writing began, and it is worth being precise about how meagre that beginning was, for the popular image of ancient “scribes” setting down laws and epics is, at this date, simply wrong. The earliest writing, both the proto-cuneiform of Uruk and the first hieroglyphs of Egypt, was overwhelmingly administrative. Its subjects were quantities of barley and beer, heads of livestock, allotments of land and labour, the names of officials and the years of kings. It was, in its origins, a technology of accounting and administration — the book-keeping of temple and palace — and not, for some centuries yet, a vehicle for literature, history, or law. The great Sumerian hymns and the epic of Gilgamesh, the Egyptian funerary and wisdom texts, all lay in the future. What 3000 BCE possessed was the seed of the written word, and the seed was a tally.

For the overwhelming majority of humanity, even this did not exist, and communication remained what it had always been: oral, gestural, and embodied, conducted in fully developed spoken languages, with knowledge held in living memory and transmitted by speech, demonstration, and ritual. But the symbolic record was now richer and more various than ever. Beyond the two scripts there were the cylinder seals of Mesopotamia, rolled to leave a signature in clay; the standardised marks and tokens of administration; the recurring deity emblem carved on Liangzhu jades; the knotted cords that may have counted the harvests of Caral; and the great alignments of stone — Stonehenge, Newgrange — that recorded, in architecture, the turning of the solar year. The deep human urge to fix meaning outside the fragile vessel of memory was finding many forms at once; in two places, one of those forms had become writing.

Of all the domains transformed by the arrival of writing, religion is among the most striking, for here the silence of the earlier age begins, partially, to speak. Ten thousand years ago we could observe ritual — careful burial, grave goods, red ochre, the carved enclosures of Göbekli Tepe — but we could not read a single belief behind it. Five thousand years ago, in the literate river valleys, we can for the first time put names to gods. The temple-cities of Sumer were dedicated to deities we can name: Inanna at Uruk, Enlil at Nippur, Enki the god of the sweet waters, An the sky. Their cult was the organising principle of the city, whose great household was the god’s, and whose priests and priest-rulers (the en and the later ensi) administered in the deity’s name. In Egypt, the king himself was divine, the living Horus, and the elaborate care of the royal dead at Abydos points already towards the mortuary religion that would later raise the pyramids. For the first time, we glimpse not merely the acts of worship but something of its content, recorded by the worshippers themselves.

Two qualifications keep this in proportion. First, even in Sumer and Egypt the earliest texts are administrative, and a full theology must still be reconstructed from later writings and from inference; what we have in 3000 BCE are names, temples, and offices, not scriptures. Second, and more important, writing illuminated only those two corners of the world. Everywhere else — on Malta, where the great temples imply a cult we cannot name; in Peru, where the mounds of Caral surely served some sacred purpose now lost; at Stonehenge and Newgrange, whose builders plainly cared about the movements of the sun — religion is still attested only by its acts, and its beliefs remain as unrecoverable as they were in our last survey. What is certain, here as before, is that the religious impulse was old, deep, and universal; what is new is only that, in two river valleys, we can at last hear a few of its words.

The great Ice Age megafauna, already much diminished ten thousand years ago, were by 3000 BCE long gone from the continents — though one haunting exception deserves record. On Wrangel Island, in the Arctic Ocean north of Siberia, a relict population of woolly mammoths still survived at this date, and would endure until around 2000 BCE: mammoths, that is, still walked the Earth while the first Egyptian kings ruled and the city of Uruk wrote its tablets. On the mainland, however, the age of the great wild beasts had given way to the age of the domestic ones. The defining animal fact of 3000 BCE is not the wild fauna but the herds: cattle, sheep, goats, and pigs were now the mainstay of life across the Old World, and to them had been added new servants — the donkey, beast of burden along the Nile and across the Near East, and, most momentously, the horse, which was being brought under domestication on the Eurasian steppe in roughly this window, as recent genomic work on the western steppe has shown. The horse would, in time, remake war, trade, and migration; in 3000 BCE its career had only just begun.

Wild creatures of great size still ranged the settled lands — lions in Greece and the Near East, elephants in Syria, the now-vanished wild aurochs across Eurasia — but the human relationship with the animal world had tilted, decisively, from the hunt to the herd.

The material world of 3000 BCE was no longer the pure Stone Age of our last survey, though stone tools remained in everyday use everywhere. Its great innovation was metal. Copper was worked across much of the Old World, and in the Near East the first true alloys had appeared — arsenical copper, and then the harder, more useful tin-bronze that would, in the following centuries, give its name to an entire age. With metal came a suite of technologies that together reshaped daily life: the wheel, both the potter’s wheel and the wheeled vehicle, the latter attested across Eurasia by around 3500 to 3000 BCE; the ox-drawn ard, or scratch-plough, which let a farmer work far more land than the hoe; the sail, which carried Egyptian boats up and down the Nile; and large-scale irrigation, on which the wealth of Mesopotamia and Egypt alike depended. Textiles of wool and linen were woven on the loom; the first cities were built of sun-dried mud-brick in the river valleys and of dressed stone in Egypt and on Malta. And in two places, the most consequential material of all was now in use — not a metal or a fibre but a medium: clay tablets in Mesopotamia, and, in Egypt, the reed-paper we call papyrus. The toolkit of ten thousand years ago had been one of stone, bone, antler, and hide; that of five thousand years ago added metal, the wheel, the plough, the sail, the loom, and the written page.

The people of 3000 BCE were, like those of every age in this series, fully modern in body and mind. But their bones now record, more sharply than before, the mixed bargain of the farming and urban life that most of them led. Skeletons from dense agricultural and city populations show the costs of a cereal-heavy diet and crowded living: more dental decay, more evidence of episodic malnutrition and of the infectious diseases that thrive where people and their animals press together, and, in some populations, a decline in average stature from the robust foragers of earlier ages. Social inequality, the new companion of the state, is legible in the bones as well, in the differing health of rich and poor. One small and telling adaptation was still incomplete: the genetic change called lactase persistence, which allows many adults to digest fresh milk, remained rare even at this date, although the keeping of dairy herds was by now widespread — recent analyses of fats preserved in ancient pottery show that people drank milk for millennia before they evolved the ability to do so comfortably. The match between the dairy and the digestion would be made only later. And in Ötzi the Iceman, frozen in the Alps around 3230 BCE, we possess the single most detailed portrait of an individual human body from the whole of prehistory — his ailments, his diet, his tattoos, and the manner of his death all preserved by the ice.

How can we speak with any confidence of a world five thousand years gone? The methods are those set out in the earlier essay, now joined by one entirely new kind of witness. The dated framework still rests, above all, on radiocarbon dating — the measurement of the slowly decaying carbon-14 absorbed by every living thing, calibrated against the growth-rings of ancient trees to yield true calendar dates, each with its stated margin of error.

Radiocarbon is buttressed, as before, by the cross-matched tree-ring chronologies of dendrochronology, by the annual layers and oxygen isotopes of the Greenland and Antarctic ice, by lake varves and coral bands, by luminescence and uranium-series methods for spans beyond radiocarbon’s reach, and — increasingly the most powerful of all for questions of ancestry and migration — by the recovery and sequencing of ancient DNA from old bone, which has rewritten our understanding of the Yamnaya expansion and much else.

What is new at 3000 BCE is the written record itself.

For the first time, the spade is joined by the text: the administrative tablets of Uruk and the labels and inscriptions of early Egypt give us names, quantities, and — soon — the reigns of kings, the raw material from which chronology in the fuller, historical sense is built. But the new witness must be handled with as much care as the old. The earliest texts are partial, formulaic, and skewed towards the concerns of temple and palace; some early scripts, such as Proto-Elamite and the later Indus seals, cannot yet be read at all; and writing speaks for only the two or three societies that possessed it, leaving the rest of the world as dependent on archaeology as ever. It is the convergence of these independent lines — a radiocarbon date, a tree-ring sequence, a band of polar ice, a strand of ancient DNA, and now a dated tablet — that turns plausible reconstruction into secure knowledge.

A closing word on the foundations, for the subject still demands humility, even as the ground beneath it firms.

What is securely known of 3000 BCE is more than for any earlier age in this series: the dated and physical record of settlements, skeletons, climate, sea level, and ancient genomes, and now the first written documents, together tell us, with real confidence, where people lived and how they fed themselves, that two river valleys had built cities and states and begun to write, that long-distance exchange spanned much of the Old World, that the hemispheres remained wholly apart, how the great population movements of the age reshaped human ancestry, and even the names of a handful of gods and the words of two or three languages.

What remains inferred, and properly uncertain, is the harder human matter: the inner workings of the new states and cities, the lives of the unrecorded majority within them, the meaning of the monuments at Caral and on Malta, the content of belief beyond the two literate valleys, and the precise size and distribution of the world’s people. And what is, on present evidence, simply unknowable is what it always was — the inner life of the unlettered world: the languages of all but a corner of humanity, and the beliefs and thoughts of the overwhelming majority who left no written trace, and whose words vanished with their breath.

It is fitting to end, as before, on that frontier between knowledge and ignorance — but to note how far it had moved. Ten thousand years ago the whole of human experience lay on the unknowable side of the line, recoverable only through the mute testimony of stone and bone. Five thousand years ago, in two river valleys and in a few words pressed into clay, the line had begun, for the very first time, to give way — and history, in the strict sense of a past that speaks to us in its own words, had been born. The world of 3000 BCE is the threshold of that long conversation, of which this essay, written five thousand years further on, is a distant continuation.

The Earth of two thousand years ago was, in its physical geography, essentially the world we inhabit. The great ice sheets had long since melted from Canada and Scandinavia; the postglacial rise of the sea, which ten thousand years ago had been drowning whole landscapes within a lifetime, had effectively ceased, and the coastlines a sailor of the year 1 would have recognised are, in nearly every particular, our own. Britain had been an island for some six thousand years, severed from the Continent when the North Sea closed over the lost land of Doggerland around 6200 BCE. The Sahara was a desert, as it is today: the green and well-watered Sahara of our first survey, with its lakes and its hippopotami, had dried into sand some two thousand years before this date, and its peoples had long withdrawn to the Nile and the southern grasslands.

One physical fact, however, bears directly on the human story, and it is climatic. The centuries on either side of the turn of the era fell within an interval of relatively warm, stable, and benign climate across the Mediterranean and much of the northern hemisphere — a phase that historians of Rome, following the work of scholars such as Kyle Harper, have come to call the Roman Climatic Optimum, or Roman Warm Period, lasting roughly from 200 BCE to 150 CE. Reliable rains, mild winters, and good harvests are no guarantee of empire, but they are a powerful help, and it is no accident that the rise of Rome and the flourishing of Han China alike coincided with this favourable spell. The climate that had made farming possible ten thousand years before now helped, in its quieter way, to make empires possible; and its later deterioration would, in turn, form part of the long story of their decline.

For the first time in this series, the question of how many human beings were alive can be answered with something better than a guess — because, for the first time, somebody counted. World population at the turn of the era is generally reckoned to have stood on the order of two hundred to three hundred million, with figures around 200 to 250 million the most commonly cited; the standard older synthesis of Colin McEvedy and Richard Jones gives about 170 million, while the demographer Jean-Noël Biraben and the HYDE database place it higher. Whichever figure one prefers, the order of magnitude is secure: roughly a fifteen- to twenty-fold increase over the fourteen million we estimated for five thousand years ago, and the engine of that growth was, as ever, farming — now practised across vast territories and feeding the densely settled empires.

What sets this date apart is the Han census of the year 2 CE. The administration of the Western Han, needing to reckon its taxes and its soldiers, counted its people commandery by commandery, and the result was preserved in the dynastic history, the Book of Han: 57,671,400 individuals in 12,366,470 households. It is the oldest surviving census in the world, and modern scholars hold it to be substantially accurate. At a stroke it lifts the population of one society, on one date, out of the realm of estimate and into the realm of record — and it reveals an East Asian population of nearly sixty million, very probably the largest under any single government on Earth. The Roman Empire, the other great agglomeration of mankind, is not so exactly known, but is generally put at somewhere between forty-five and sixty million in the reign of Augustus. Between them, then, the two great empires of the Old World — wholly ignorant of each other’s true nature — held something approaching half of all the human beings then alive.

The distribution of the rest is clearer than for any earlier age, because the dense populations now lay in identifiable places: along the rivers and across the irrigated plains where intensive agriculture flourished. The thickest concentrations were in Han China, in the Roman Mediterranean, and on the plains of northern India, with further substantial populations in Parthian Mesopotamia and Persia. Around and beyond these cores the older world persisted — herders on the Eurasian steppe and the drying margins of the Sahara, farmers thinly spread across temperate Europe and much of Africa and Southeast Asia, and hunter-gatherers, fishers, and small-scale cultivators holding the whole of Australia, the greater part of the Americas, and the northern forests of Eurasia. The decisive demographic fact of the age is concentration: for the first time a large fraction of humanity lived not in small mobile bands or modest villages but as the taxpaying subjects of enormous states, in a world that still, by any modern measure, lay all but empty between its few crowded centres.

As in the earlier surveys, it is worth recalling the regions through which we shall view the peoples of this age, for they are broadly the same culture-areas we have followed throughout — and the point of returning to them is, once more, to watch how far they had diverged. Ten thousand years ago they were scarcely distinguishable seedbeds; five thousand years ago two of them had pulled sharply ahead. Now, at the turn of the era, several had been welded into empires and kingdoms of a scale never before seen, while others followed paths that bypassed the state altogether. Chief among them are the Roman Mediterranean; Han China; Parthian Persia and the Near East; the Indian subcontinent; Egypt and the rest of Africa; temperate Europe and the still-independent British Isles; the Americas, developing as ever in isolation; and Australia with the islands of the Pacific. The survey below treats each in turn.

Rome and the Mediterranean: modern Italy and the lands about the Mediterranean Sea

The most powerful state in the western half of the world was the Roman Empire, which at the turn of the era had just emerged, after a century of civil war, into the long peace that historians call the Pax Romana. Its architect was the first emperor, Augustus — the former Octavian, great-nephew and heir of the murdered Julius Caesar — who from 27 BCE ruled as ‘first citizen’ over a domain that ran from Spain and Gaul to Egypt and Syria, embracing the whole Mediterranean, which the Romans could call mare nostrum, ‘our sea’. He boasted, with some justice, that he had found Rome a city of brick and left it one of marble, and the great works of the Augustan age — the forums, the temples, the Altar of Peace — set the style of imperial display. The capital itself was approaching a million inhabitants, almost certainly the largest city in the world, fed by grain fleets from Egypt and watered by aqueducts. Latin and Greek were the languages of a literate governing class, and the age produced one of the supreme flowerings of Western literature, in the epic of Virgil, the verse of Horace and Ovid, and the vast history of Livy.

Two cautions are needed against the popular image.

First, the Rome of the year 1 had not yet reached its greatest extent: that came under Trajan more than a century later, around 117 CE, and the famous monuments by which we picture the city — the Colosseum, the Pantheon in its present form — were not yet built. Second, the relentless expansion was already meeting its limits. In 9 CE, within our window, three entire legions under Publius Quinctilius Varus were annihilated by the Germanic chieftain Arminius in the forests east of the Rhine, a catastrophe that fixed the empire’s northern frontier on the Rhine and the Danube for centuries and from which, it was said, the ageing Augustus never fully recovered. Rome was mighty, but it was not yet the Rome of the textbooks, and it had learned that even its power had an edge.

Han China: modern China

At the other end of Eurasia lay an empire of comparable scale and greater population: Han China, founded in 202 BCE and destined to give the majority people of China their enduring name. Its sophistication was extraordinary. The Han state was administered by a salaried, partly examined bureaucracy schooled in the doctrines of Confucius, which the dynasty had adopted as its official creed; its historians had created, in the Records of the Grand Historian of Sima Qian, the foundation of two thousand years of Chinese historiography; and its craftsmen worked cast iron in blast furnaces on a scale the West would not match for well over a millennium, drew the silk that the whole Old World craved, and were on the verge of inventing cheap paper. Yet at the precise moment of the year 1 the dynasty was in crisis. A succession of infant emperors had left power in the hands of the regent Wang Mang, who in 9 CE would set the boy-emperor aside and seize the throne for himself as the founder of a short-lived ‘Xin’ dynasty, before the Han were restored in 25 CE. The empire that the census of 2 CE found teeming with sixty million souls was, politically, on the brink of upheaval.

Parthian Persia and the Near East: modern Iran, Iraq, and Syria

Between Rome and India stretched the third of the Old World’s great powers, the Parthian Empire of the Arsacid dynasty, which ruled Persia and Mesopotamia and was Rome’s perennial rival in the East. The Parthians had inflicted on Rome one of its most humiliating defeats, the destruction of Crassus and his legions at Carrhae in 53 BCE, where the heavy armoured lancers and mounted archers of the steppe tradition shattered the Roman infantry; thereafter the two empires contended, on and off, over the buffer kingdom of Armenia and the trade of the East. The Parthian world was heir to the Hellenistic kingdoms that Alexander’s conquests had left behind, and to the far older civilisation of Persia; its religion was the ancient faith of Zoroaster, and it sat astride the overland trade routes, growing rich as the indispensable middleman between the silk of China and the markets of the Mediterranean — a position it guarded jealously.

South Asia: modern India, Pakistan, and Sri Lanka

The Indian subcontinent at the turn of the era was politically fragmented but culturally and intellectually among the most vital regions on Earth. The great Mauryan Empire, which under Ashoka in the third century BCE had united most of the subcontinent and spread Buddhism, had collapsed in 185 BCE, and no single power had replaced it. The north-west was ruled by a succession of foreign dynasties — Indo-Greeks, then Indo-Scythians (the Sakas) and Indo-Parthians — and was about to be united by the Kushans, a people of Central Asian origin whose great empire, linking India to the Silk Road, was beginning to form. The Deccan plateau was held by the Satavahanas, and the far south, beyond the reach of northern empires, by the Tamil kingdoms of the Cheras, Cholas, and Pandyas, whose ‘Sangam’ poetry and brisk maritime trade with Rome are well attested. Across this patchwork flourished a civilisation of the first rank: the great epics, the Mahabharata and the Ramayana, were reaching their developed form; the foundations of classical Hinduism were being laid; Buddhism and Jainism were widely practised and patronised; and Indian mathematics, soon to give the world the decimal place-value system, was in vigorous development. India lacked a single empire, but it lacked nothing in vitality.

Egypt and the African Kingdoms: modern Egypt, Sudan, Ethiopia, and Nigeria

Egypt — the most ancient of the literate civilisations described in our last survey — was now a province of Rome, annexed by Augustus in 30 BCE after the death of Cleopatra, the last of the Greek Ptolemies; it served the empire as its indispensable granary, even as its people went on writing in Demotic and worshipping their immeasurably old gods under Roman rule. Up the Nile, beyond the first cataract, lay the independent kingdom of Kush, ruled from the city of Meroë, with pyramids of its own and a script, Meroitic, that we still cannot read; its armies clashed with Rome on the southern frontier. Further south-east, in the highlands of modern Ethiopia and Eritrea, the kingdom of Aksum was rising into the trading power it would later become. In West Africa, the Nok culture of modern Nigeria was working iron and modelling the remarkable terracotta heads that are among the oldest sculpture of sub-Saharan Africa; and across the centre and south of the continent the long, slow Bantu expansion was carrying farming, and in time iron, into new lands. It must be said plainly that, lacking written records of its own across most of this vast region, sub-Saharan Africa at the turn of the era is far less well known to us than the literate empires, and much about it remains to be recovered.

Temperate Europe and the British Isles: modern Germany, France, and the United Kingdom Beyond the Roman frontier, across temperate Europe, lived the Iron Age peoples whom the Romans lumped together as Celts and Germani: farmers and warriors organised in tribes and chiefdoms, some of them dwelling in substantial fortified towns, or oppida, minting their own coinage and trading briskly with the empire to the south. Their priests, the druids, and their largely oral culture have left us tantalisingly little in their own words. Britain, in particular, was at the turn of the era still wholly independent: Julius Caesar had raided the island in 55 and 54 BCE but had not held it, and the Roman conquest under Claudius lay more than four decades in the future, in 43 CE. Powerful native kingdoms were taking shape, notably that of the Catuvellauni under Cunobelinus — the ‘Cymbeline’ of Shakespeare — who struck Roman-style coins and dominated the south-east. It is a measure of how far the record has advanced that, where our earlier surveys could name a British individual only through his skeleton, as with Cheddar Man, this age yields British kings known by name, from coins and from the histories of their Roman neighbours.

The Americas: modern Mexico, Guatemala, Peru, and the United States

Across the oceans, in complete isolation from the Old World, the peoples of the Americas were building civilisations of their own. In the lowland forests of modern Guatemala, the Maya of the Late Preclassic period had raised the vast city of El Mirador, whose great pyramid, La Danta, is among the largest structures by volume ever built by human hands — and all of this in the centuries around our window, long before the famous ‘Classic’ Maya cities of Tikal and Palenque, which belong to a later period from about 250 to 900 CE. In the Valley of Mexico the immense planned city of Teotihuacan was beginning the rise that would make it one of the largest cities on Earth, its Pyramid of the Sun raised in the first centuries of the Common Era, while the older centre of Cuicuilco was soon to be overwhelmed by a volcanic eruption. On the Peruvian coast the Nazca were creating the colossal ground-drawings that still mark their desert, and the Moche were emerging; in eastern North America the Hopewell built great earthen mounds and ran exchange networks that spanned half a continent. And here recurs the standing marvel of this series: the Americas achieved monumental cities and complex societies with neither iron, nor the wheel for transport, nor any large draft animal — a wholly independent route to grandeur that the Old World’s measuring-sticks were never designed to take.

Australia and the Pacific: modern Australia and the islands of the western Pacific

In the far south, the Aboriginal peoples of Australia continued one of the longest unbroken cultural traditions on Earth, sustaining across an entire continent a way of life founded on hunting, gathering, and the deep management of land by fire, and — as in every earlier survey — never taking up farming, in what was almost certainly a considered adaptation to their environment rather than any want of capacity. The dingo, the one notable newcomer, had arrived perhaps a thousand years or so before this date. Out in the Pacific, one of the boldest movements in all of human history was under way: the Austronesian expansion, which by way of the seafaring Lapita peoples had carried farming communities out across thousands of kilometres of open ocean to Fiji, Tonga, and Samoa by around 1000 to 500 BCE. At the turn of the era these islands of western Polynesia were settled, but the final, astonishing leaps eastward — to Tahiti, Hawaii, Easter Island, and New Zealand — still lay many centuries in the future. The most thinly peopled quarter of the globe was still, slowly, being discovered by its first human beings.

Because the question is so natural, it bears a direct answer — and here the contrast with the earlier surveys completes a pattern. Ten thousand years ago, not one of the famous ancient civilisations existed; all lay in the future. Five thousand years ago, two had arrived and a third was forming. Two thousand years ago, the great empires are not in the future at all: they are the present. Rome is a living empire under its first emperor; Han China is a living empire of sixty million; Parthia holds Persia; India teems with kingdoms. The reframing this time is subtler, and it concerns not whether the famous things existed but exactly when, for popular imagination still runs ahead of the evidence in revealing ways. Rome had not yet reached its greatest extent — that came under Trajan around 117 CE — and the long decline of the Western empire to its fall in 476 CE lay nearly five centuries ahead; the Colosseum (opened 80 CE), the Pantheon in its present form (rebuilt under Hadrian around 126 CE), and Hadrian’s Wall (begun around 122 CE) were none of them yet built. Britain was not Roman and would not be for over forty years. The classic temple-pyramids of the Maya belong to a Classic period from about 250 to 900 CE, still centuries away; the Maya of our window were the monument-builders of an earlier, Preclassic golden age. The lesson is the gentle inverse of the earlier ones: where before the famous things lay unimaginably far in the future, now they cluster close on either side of our date, and the work is to set them in their right order rather than to defer them wholesale.

The question carries the same trap it has carried throughout this series — ‘advanced’ quietly rewards cities, writing, states, and monuments, and slights the deep sophistication of those who chose other ways of living — and the same caution must be entered. But at the turn of the era the answer is, by the conventional measures, less a matter of a single leader than of a near tie between two giants. The Roman Empire and Han China stood, on any reckoning of scale and complexity, far ahead of the rest of the world and remarkably level with each other. Each governed some fifty to sixty million people; each ran a literate bureaucracy, a professional army, a coined money economy, a network of roads, and great cities; each commanded engineering of the first rank — Rome its concrete, aqueducts, and arches, Han China its cast iron, its crossbows, and its waterworks; each produced a sophisticated literature and a tradition of history-writing, in Livy on the one hand and Sima Qian on the other. That two such comparable civilisations should have arisen at opposite ends of Eurasia, in almost complete ignorance of one another, is among the most instructive facts of world history.

That two such comparable civilisations should have arisen at opposite ends of Eurasia, in almost complete ignorance of one another, is among the most instructive facts of world history.

Yet the older humility applies here too, and pointedly. Parthia and India were no barbarian hinterlands but heirs to ancient and brilliant traditions of their own — India in particular a powerhouse of mathematics, grammar, philosophy, and religion whose influence would in time reach across Asia. And beyond the literate Old World entirely, the builders of El Mirador and the rising lords of Teotihuacan raised monuments to rival any in Rome, in total isolation and without iron, the wheel, or a single draft animal; the Hopewell moved earth on a continental scale; the Aboriginal Australians sustained across a continent a way of life of profound ecological depth. These were not laggards on a single road with Rome and China in the lead. They were distinct answers to the question of how human beings might live at scale. It remains more accurate to speak of several remarkable civilisations, of which the two great empires were the most complex by the particular measures we happen to prize.

Here the change since our last survey is profound, for the turn of the era is the age in which the Old World became, for the first time, a single connected system — loosely, indirectly, and at second or third hand, but really. Two great arteries of exchange did the work. Overland, the network of routes we call the Silk Road ran from the western frontier of Han China, through the oasis cities of Central Asia and across Parthian territory, to the Mediterranean; the Han had pushed into Central Asia in the previous century, after the embassies of the explorer Zhang Qian, and the road carried Chinese silk westward in such quantity that Roman moralists complained both of the gold drained away to buy it and of the immodesty of the transparent fabric. By sea, an even greater volume of trade ran across the Indian Ocean, where Greek and Roman merchants had mastered the rhythm of the monsoon winds and sailed directly from Egypt to the ports of southern India and beyond — a trade described in remarkable detail in a surviving Greek merchant’s handbook, the Periplus of the Erythraean Sea, written about the middle of the first century CE, and confirmed by the hoards of Roman gold coins still turned up in the Indian south. Goods, and with them ideas, religions, and diseases, now moved across the whole breadth of the Old World: Chinese silk reached Rome, Roman glass reached China, Indian pepper seasoned Roman kitchens, and Afghan lapis and Arabian incense travelled as they had for millennia.

Rome and Han China between them held perhaps half the living world, yet knew of each other only as a rumour at the edge of the earth. And yet the two ends of this system knew of each other only barely, through a fog of rumour and through intermediaries with every interest in keeping them apart. The Romans called the distant source of silk Seres, the ‘silk people’, and knew almost nothing true about them; the Chinese called the Roman world Daqin, a kind of counterpart-empire imagined at the western edge of the earth, and a Han embassy under Gan Ying, sent towards Rome around 97 CE, was turned back at the Persian Gulf by Parthians unwilling to lose their place as middlemen. No Roman of the year 1 had seen China, and no Chinese had seen Rome. The two empires that between them ruled half of humanity faced each other across a continent neither could cross, linked only by a chain of traders no single one of whom travelled its whole length.

Between the hemispheres, meanwhile, the ancient isolation held as absolutely as ever. There was no contact whatever between the Old World and the New, nor between either and Australia, nor with the Pacific islands then being settled from the west. The proof is the same as it has been throughout this series, and by now overwhelming: the civilisations of Mesoamerica and the Andes arose, flourished, and built their monuments in complete ignorance of Rome, China, and India, and did so along wholly independent lines — without the wheel, without iron, without Old World crops or beasts. The ‘known world’ of even the most travelled person alive in the year 1 had widened enormously, to embrace much of a single hemisphere; but the globe entire remained, to every soul on Earth, unknown.

The answer is the same as it was for ten and for five thousand years ago, and it rests on the same well-founded grounds: the racial categories of modern usage are a recent social construction, and to apply them to the people of two thousand years ago is anachronistic and misleading.

Humanity was then, as throughout this series, a single biological species whose variation graded continuously across geography rather than falling into the discrete, bounded types of folk taxonomy. What had changed since five thousand years ago is that the great population mixtures of the middle Holocene — the fusions of hunter-gatherer, farmer, and steppe ancestries that the genetic work of David Reich and others has traced — were now largely complete, so that the broad ancestral make-up of the world’s regional populations had come to resemble, in its outlines, the distribution we know today. In that limited sense the peoples of the year 1 were closer to us than those of the earlier surveys. But the modern racial vocabulary still does not fit them, and for a reason that goes beyond biology: the people of antiquity did not, on the whole, think in its terms. The Romans, who ruled an empire of staggering diversity, drew their fundamental distinctions not by skin colour or ‘race’ in the modern sense but by citizenship, language, and culture — between Roman and non-Roman, civilised and ‘barbarian’, free and enslaved. A man might be dark-skinned and a Roman citizen, or pale-skinned and a barbarian slave; what mattered was his standing, not his complexion. Such colour-prejudice as existed was real but unsystematic, and mapped onto nothing resembling the rigid racial hierarchies constructed in much later centuries. To sort the inhabitants of the Roman, Parthian, Indian, and Han worlds into ‘white’, ‘black’, ‘Asian’, and ‘Oriental’ is therefore to impose on them both a biology that does not correspond to nature and a way of dividing humanity that they would not have recognised. The categories are ours, and they were not built for them.

Here, at last, this series can speak with confidence rather than confess ignorance — for two thousand years ago much of the Old World was not only speaking but writing, and we can name, read, and in many cases still study the languages of the age. The Roman Mediterranean was effectively bilingual at the top: Latin was the language of law, army, and administration in the west, while Greek remained the language of learning, commerce, and culture across the east, the common tongue of the educated from Sicily to Syria. The everyday speech of the Near East, heard in the streets of Judaea and across Parthian Mesopotamia, was Aramaic — the mother tongue, among countless others, of the Galilean teacher whose life is discussed below. In Persia they spoke Parthian and the other Iranian tongues; in Egypt, the ancient Egyptian language in its late form, soon to be written in the Greek-derived alphabet we call Coptic; in India, Sanskrit as the language of learning and liturgy, alongside the many Prakrits of daily speech and, in the south, the wholly unrelated Tamil of the Sangam poets; and in China, the Old Chinese of the Han, already recorded in a script whose direct descendants are still written today.

Two deeper points follow. The first is that the great language families reconstructed by the comparative method — the technique, described in our earlier surveys, of working backwards from related living tongues to their vanished parent — were by now no longer buried in unreachable prehistory but splayed across the map in their attested daughters: Latin, Greek, Sanskrit, Persian, and the Celtic and Germanic tongues are all branches of the single Indo-European family whose reconstructed parent we placed, in the last survey, on the steppe around 3000 BCE. Two thousand years ago we can watch that family in its maturity, its members already as different from one another as Latin is from Sanskrit. The second point is the old humility, much reduced but not gone: writing illuminated the literate civilisations and left the rest of the world as dark as before. The languages of the Hopewell and the Nazca, of the builders of El Mirador, of every Aboriginal Australian nation, of most of sub-Saharan Africa and the European barbarian north, were unwritten, and the great majority of them are lost beyond recovery. Linguistic diversity remained immense; but where five thousand years ago two or three coastlines of that ocean of speech had been charted, two thousand years ago a whole hemisphere’s worth had been mapped, while the other hemisphere remained a blank.

The contrast with our last survey is, in this domain, the sharpest of all. Five thousand years ago writing had just been invented, and it recorded almost nothing but tallies of grain and livestock; the idea of writing down a poem, a law, or a history lay centuries in the future. Two thousand years ago, in the literate civilisations, writing did all of these things and more. Rome had a vast literature — the epic of Virgil, the verse of Horace and Ovid, the history of Livy, the speeches and letters of Cicero a generation before — and it had archives, libraries, published books copied and sold, and a public life conducted in part on paper. Ordinary people wrote too: the painted graffiti of Pompeii and, a little after our window, the wooden letters of the soldiers and their families at Vindolanda on the British frontier preserve the unguarded voice of the non-elite as no earlier source can. China had its own magnificent tradition, crowned by the Records of the Grand Historian of Sima Qian, completed around 94 BCE, and stood on the verge of a revolution in the very medium of writing, for although paper existed in a coarse early form, its refinement into a cheap writing surface — traditionally credited to the official Cai Lun in 105 CE — lay just beyond our date. India, for its part, transmitted a vast sacred and learned literature with a fidelity that owed as much to prodigies of memorised recitation as to writing. Two qualifications keep this in proportion. The first is that, for the overwhelming majority of humanity — the unlettered subjects of the empires, and the entire populations of the Americas, Australia, and most of Africa and northern Europe — communication remained what it had always been: oral, gestural, and embodied, with all knowledge held in living memory and passed on by speech, demonstration, and ritual. Literacy, even within the empires, was the accomplishment of a minority. The second is that the symbolic record beyond writing was richer than ever: the coinage now circulating across the whole Old World carried images and messages to people who could not read; the Andean peoples kept records on knotted cords, the ancestors of the later Inca quipu; the Maya were developing a true writing system of their own, which would flower in the centuries after our window; and great monuments, from Roman triumphal arches to the desert drawings of Nazca, recorded meaning in forms that needed no script. The deep human urge to fix meaning outside the fragile vessel of memory had, in the literate world, achieved its full and familiar form; elsewhere it continued to find the many other forms it always had.

Of all the domains transformed by the spread of writing, religion is the most consequential, for the turn of the era is one of the great hinges in the religious history of mankind. Ten thousand years ago we could observe ritual but read not a single belief; five thousand years ago, in two river valleys, we could at last put names to a few gods. Two thousand years ago we stand amid the fully documented theologies of the literate civilisations — and, more momentously, at the very birth of religious traditions that would shape the next two millennia and that command the allegiance of much of the living world today.

The Roman world was home to a tolerant, capacious polytheism, its old gods increasingly identified with their Greek counterparts and overlaid now by a politically potent novelty: the imperial cult, in which the deified Julius Caesar was worshipped and his heir Augustus honoured as ‘son of a god’, divi filius, with full divine honours to follow after death. Educated Romans, meanwhile, increasingly sought meaning in the great philosophical schools of the Greeks — Stoicism above all, with its ethic of duty and self-mastery, and Epicureanism — which functioned for many as something close to a religion of conduct. From the eastern provinces, mystery cults promising personal salvation and rebirth — those of Egyptian Isis, of Persian-derived Mithras, of Anatolian Cybele — were spreading through the empire and would soon contend for the soul of the Roman world.

It was in one small and turbulent corner of that world that the most consequential religious development of the age took shape. In Roman Judaea the Jewish people maintained, around the great rebuilt Temple in Jerusalem, the most fully developed monotheism of antiquity — the worship of a single God under a revealed Law — divided into rival parties, among them the Pharisees, the Sadducees, and the Essenes of the Dead Sea Scrolls, and charged with a fervent expectation of a deliverer, a Messiah. Into that expectant world, in the last years of the reign of Herod the Great and so around 6 to 4 BCE, was born the figure whose life would prove, by any measure of historical consequence, among the most influential ever lived: Jesus of Nazareth, known in the Islamic tradition as the prophet ʿĪsā, son of Maryam (Mary) (peace be upon him). Of the bare historical outline most scholars are confident: a Jewish teacher and healer from Galilee, who preached in Aramaic, gathered disciples, and was put to death by crucifixion under the Roman prefect Pontius Pilate around 30 to 33 CE. From his followers’ proclamation grew Christianity, which over the following three centuries spread across the Roman Empire and in time became the largest religion on Earth.

A point of method and of respect must be made plainly here, for the figure stands at the meeting-place of faiths, and on his meaning the great traditions differ. The bare facts of a Galilean teacher’s life belong to history; the significance placed upon them belongs to faith. In Christian belief Jesus is the incarnate Son of God, crucified and risen. In Islam — whose orthodox Sunni understanding is this series’ frame on such matters — ʿĪsā is one of the greatest of the prophets and the Messiah, al-Masīḥ, born by the command of God of the Virgin Maryam without a father, as set out in the Qurʾan in Sūrat Āl ʿImrān and Sūrat Maryam; yet he is emphatically not divine and not the son of God, and the Qurʾan teaches that those who sought his death neither killed him nor crucified him, but that it was made to appear so to them, and that God raised him to Himself (Sūrat al-Nisāʾ, 4:157–158). The historian’s task is not to adjudicate between these convictions but to record a remarkable fact of perspective: from the vantage of the year 1, this most consequential of lives was, to the literate empires of its day, almost wholly invisible. No contemporary Greek or Roman writer records it; the vast apparatus of imperial administration took no notice; and the event passed, at the time, entirely beneath the gaze of history. Of all the lives lived around the year 1, the one that would shape the future most was, to the literate empires of its day, almost wholly invisible. The same age was momentous for the religions of the East. Buddhism, already some five centuries old and carried across India by the patronage of the emperor Ashoka in the third century BCE, was now developing the strand known as the Mahayana and beginning the long missionary journey, along the Silk Road, that would bring it into China within a century or so of our date and eventually across the whole of East Asia. In India the foundations of classical Hinduism were being laid, the great epics reaching their developed form and with them devotional currents that endure to this day. In China the state had embraced Confucianism — less a religion than an ethical and ritual order, centred on the family, the ancestors, and the right ordering of society — as the official doctrine of the Han, while the contemplative tradition of Daoism flourished alongside it and the veneration of ancestors underlay them both. In Persia the ancient faith of Zoroaster, with its vision of a cosmic struggle between truth and falsehood, light and darkness, was the religion of the Parthian world. Within a few centuries on either side of the year 1, in short, the religious map of the Old World assumed an outline we can still recognise — Christianity born, Buddhism spreading, Hinduism and Confucianism and Zoroastrianism in their classic forms — even as, beyond the reach of writing, in Mesoamerica and the Andes, at the weathered circle of Stonehenge and along the Australian songlines, religion continued, as in our earlier surveys, to be attested only by its acts.

The great Ice Age megafauna, already long gone from the continents in our last survey, were a distant memory by the turn of the era — though here too there is a haunting coda, for the last mammoths on Earth, the dwarfed population marooned on Wrangel Island in the Arctic, had themselves died out only around 2000 BCE, so that they had outlived the building of the Egyptian pyramids but were gone more than a thousand years before Rome was an empire. The animal story of this age, however, is not the wild fauna but the tamed. The domestic herds — cattle, sheep, goats, and pigs — were the mainstay of life across the Old World, and to them the previous millennia had added the beasts that now defined the age: the horse, which under the Parthians and on the steppe had remade war into a contest of mounted archers and armoured cavalry; the camel, the Bactrian in Central Asia and the dromedary in Arabia and North Africa, whose endurance made possible the desert stretches of the silk and incense roads; and the patient donkey and ox of farm and caravan. Large wild animals still ranged the settled lands — lions survived in the Near East and North Africa, the wild aurochs still roamed the forests of Europe, and the elephant was hunted in Syria and put to war from India to the Mediterranean — but the decisive relationship of human beings to the animal world had long since tilted, from the hunt to the herd, and the beasts that mattered most were now those that pulled, carried, and were eaten.

The material world of two thousand years ago was an Iron Age world fully matured — iron tools and weapons were in everyday use across the whole of the Old World, from the legions of Rome to the smiths of the Nok culture in West Africa — and it had added to the long inheritance of stone, bronze, and iron a suite of refinements that gave the age its particular character. Rome’s signature was in building: opus caementicium, the concrete that let her raise domes, vaults, aqueducts, harbours, and the vast public works that still stand, and the network of paved roads that bound the empire. A recent and brilliant innovation was glassblowing, invented in the Syro-Palestinian region around the middle of the first century BCE, which transformed glass from a luxury into a commonplace. China’s genius ran along different lines: the Han worked cast iron in blast furnaces on a scale the West would not match for well over a millennium, drew fine silk, lacquered wood, and stood on the threshold of cheap paper. India produced the finest steel and the most prized cotton textiles of the ancient world and was developing the decimal numeral system whose later perfection would change mathematics everywhere. And the most universal token of the age was money: coined currency, unknown five thousand years ago, now circulated across the entire Old World, so that a Roman denarius, a Parthian drachm, an Indian punch-marked coin, and a Han cash all bought and sold in their own spheres, and sometimes far beyond them. Only across the oceans did this world end: the Americas, raising their monuments in stone, knew neither iron, nor the wheel, nor coined money, nor the written page, and yet achieved a complexity that puts the very category of ‘material progress’ under honest strain.

The people of two thousand years ago were, like those of every age in this series, fully modern in body and mind. But their bones and their texts together now record, more sharply than ever, the mixed bargain of the urban and agrarian life that so many of them led. The great cities were reservoirs of disease: crowded, often filthy, and fed by long-distance trade that carried infections as readily as goods, they suffered epidemics whose toll the Roman sources begin, for the first time, to describe — though the most devastating, the Antonine and later plagues, fell after our window. Diets were heavily cereal, and the skeletons of the poor show the familiar marks of episodic hunger, dental decay, and overwork, while the inequality that was the companion of the state is legible in the differing stature and health of rich and poor. Roman medicine, soon to be codified by the encyclopaedist Celsus around 25 CE and later by Galen, was a curious mixture of shrewd observation and humoral theory. Life expectancy at birth was low — perhaps in the mid to high twenties, dragged down by a fearful toll of infant and child mortality — though those who survived childhood could reasonably hope for a vigorous middle age and sometimes a long one. One small adaptation, traced through this series, was still spreading rather than universal: the genetic change called lactase persistence, which allows adults to digest fresh milk, remained patchy even now, although dairying was ancient — for, as recent analysis of milk-fats in old pottery has shown, human beings drank milk for thousands of years before they evolved the comfortable ability to do so.

How can we speak with such confidence of a world two thousand years gone? The methods of the earlier surveys all still serve — radiocarbon dating, calibrated against tree rings, to fix the age of organic remains; dendrochronology itself, often now precise to the single year; the annual layers and isotopes of the polar ice, recording the ancient climate and even, in their lead and other pollutants, the rhythm of Roman mining and metalworking; and the recovery of ancient DNA, which has illuminated the migrations and the kinship of the age. But at the turn of the era these are joined, and in many places overtaken, by the written record itself — by histories, inscriptions, coins, papyri, and the dated reigns of kings and consuls — which allows us, for the first time in this series, to speak not merely of centuries but often of single years, and occasionally of single days. We know the year Augustus took his title and the year the legions were lost in Germany; we can date an emperor’s reign, a battle, or an eclipse. The new witness, however, demands the same wary handling as the old, and more, for the written record of antiquity is partial, partisan, and the work of a narrow elite: it speaks for the victors, the rulers, the cities, and the men, and is silent or hostile on the conquered, the enslaved, the rural, and the female, to say nothing of the wholly unrecorded worlds beyond writing’s reach. It is, as ever, the convergence of independent lines — a dated inscription, a radiocarbon measurement, a tree-ring sequence, a coin hoard, a strand of ancient DNA, a layer of pollution in the Greenland ice — that turns testimony into secure knowledge.

A closing word on the foundations, for the subject still demands humility even as the ground beneath it has firmed beyond anything available to the earlier surveys. What is securely known of the turn of the era is vast: the dated and physical record of settlements, skeletons, climate, and ancient genomes, now joined by a deep and various written record, tells us — with real confidence — where the great mass of humanity lived and how it was governed, that two empires between them held perhaps half the world’s people, that literate civilisations flourished from the Atlantic to the Pacific, that goods and ideas and faiths moved across the whole breadth of the Old World while the hemispheres remained wholly apart, what languages much of mankind spoke, what gods it worshipped, and even the names, the words, and the deeds of many thousands of individual men and women.

What remains inferred, and properly uncertain, is the inner life of the silent majority: the daily experience of the peasants and the enslaved who made up the bulk of every empire and left almost no record of their own; the true workings of societies we know mainly through the self-flattering testimony of their elites; the meaning of the monuments and the content of the beliefs of the unlettered peoples — the builders of El Mirador, the makers of the Nazca lines, the nations of Aboriginal Australia. And what is, on present evidence, simply unknowable is what it has always been in this series — the interior worlds of those who left no written trace, the languages of half the planet, and the thoughts of the overwhelming majority of human beings then alive, which vanished with their breath. It is fitting to end, as in the earlier surveys, on that frontier between knowledge and ignorance — but to mark how far, in five thousand years, it had moved. Ten thousand years ago the whole of human experience lay on the unknowable side of the line. Five thousand years ago, in two river valleys and a handful of clay tablets, the line had just begun to give way. Two thousand years ago, across the entire breadth of the literate Old World, it had given way for good: history, in the strict sense of a past that addresses us in its own words, was now not a novelty confined to two cities but the normal condition of civilised mankind. The world of the year 1 is the one in which that long human conversation — of which this essay, written two thousand years further on, is a distant continuation — had become, at last, a roar rather than a whisper. And yet, as every page of it reminds us, the roar was the voice of a privileged few; behind it, then as now, stretched the vast and patient silence of the unrecorded many.

As in the earlier essays, three things must be kept apart: what is securely known, what may reasonably be inferred, and what cannot be known at all. The advantage of the sixth century over the deep past is enormous. This is a fully historical world across most of the literate Old World — a world that names itself, dates its own events, writes its own laws and chronicles and poems, and argues, often at ferocious length, about its own beliefs. We can follow the reign of an emperor year by year, read the dispatches of his generals, and inspect the coins struck in his name. But the gift is uneven, and in one celebrated case — the Britain of this very period — the written record very nearly goes dark, so that we are thrown back upon archaeology and legend much as we were ten thousand years before. The frontier between knowledge and ignorance has not vanished; it has merely moved, and in places it doubles back upon itself.

The physical world of 570 CE was, in its broad outline, the world we inhabit: the great ice sheets long melted, the sea at essentially its present level, Britain an island these six thousand years, and the Sahara the desert it remains. In our previous survey we dwelt upon the climate, because the centuries around the year 1 had enjoyed a warm, stable, and benign spell — the Roman Climatic Optimum — that had helped, though it did not cause, the flourishing of Rome and Han China alike. The sixth century offers the mirror image, and the contrast is one of the most striking facts in this whole survey.

In the year 536 CE something extraordinary happened to the sky. A veil of dust spread across the Northern Hemisphere and dimmed the sun for some eighteen months. The Byzantine historian Procopius, an eyewitness, recorded that ‘the sun gave forth its light without brightness, like the moon, during this whole year’, and that it seemed as though the sun were in perpetual eclipse. We now know the cause. Ice cores and the rings of ancient trees record a cluster of colossal volcanic eruptions — in 536, again in 540, and a third around 547 — whose ash and sulphur dimmed the sun and chilled the earth. The work of the dendrochronologist Ulf Büntgen and his colleagues, who in 2016 christened the episode the Late Antique Little Ice Age, identifies the resulting cold spell as among the most severe of the past two thousand years, lasting from 536 to roughly 660 CE. One historian has called 536 the beginning of the worst period to be alive in recorded history.

The consequences were not abstract. Across Europe, the Mediterranean, the Near East, and China, the chronicles speak of failed harvests, summer frosts, snow in high summer, and famine. A cold and hungry world is a weakened world; and into that weakness, almost at once, came the pestilence.

In 541 CE plague arrived in the Egyptian port of Pelusium, and the following year it reached Constantinople. It was, so far as the record allows us to say, the first pandemic of bubonic plague in human history — caused by the bacterium Yersinia pestis, the same organism that would return eight centuries later as the Black Death, and confirmed as such in our own day by the recovery of its DNA from sixth-century graves. Contemporaries described the swellings, the fever, and the heaped and hasty burials with a horror that leaps off the page. The outbreak is named, after the reigning emperor, the Plague of Justinian, and it recurred in waves for two hundred years.

How many it killed is one of the liveliest controversies in the field, and it is worth pausing upon, because it bears directly on the character of the age. The older, ‘maximalist’ view — associated with the historian Kyle Harper — holds that the plague carried off between a quarter and a half of the population of the Eastern Roman Empire, perhaps tens of millions across Eurasia, and that it helped to end antiquity itself. A ‘minimalist’ school, led by Lee Mordechai and Merle Eisenberg, has argued in a series of studies since 2019 that the evidence for catastrophe is thin: the papyri, coins, and inscriptions show striking continuity, the plague pits are few, and at two great cemeteries in Bavaria, of nearly two thousand graves, only eleven have yielded the plague bacterium. The honest position, and the one this series adopts, is that the plague was real, recorded, and genetically confirmed, but that its scale is genuinely uncertain — a salutary reminder that even in a literate, well-documented age, the most basic human facts can elude us.

World population around the year 600 CE is usually reckoned at something close to two hundred million — much the same figure we gave for the turn of the era, six centuries before. This near-stagnation is itself the headline. In every previous survey the population had risen many-fold between one instalment and the next; here, for the first time, the line is flat, and may for a generation or two have dipped. The standard synthesis of Colin McEvedy and Richard Jones places the world of 600 CE at around two hundred million; other estimates run somewhat higher or lower. The combined drag of the cold decades and the recurring plague is the obvious suspect for the stall, though, as we have seen, the magnitude of the plague is disputed, and the figures are in any case coarse.

The distribution, as always, can be sketched only in the roughest terms. The two demographic heavyweights of the Old World were, as at the turn of the era, China and the Indian subcontinent, each home to some tens of millions; the lands of the two great western empires — the Eastern Roman and the Sasanian Persian — held perhaps thirty to fifty million between them; Europe beyond the Mediterranean, sub-Saharan Africa, and the Americas accounted for the rest, with the New World’s population, concentrated in Mesoamerica and the Andes, a particular unknown. These are educated guesses dressed in numerals, and should be read as such. What can be said with confidence is that no single state now governed anything like the share of humanity that Han China had governed six centuries earlier; the age of the truly universal empire was, for the moment, in abeyance.

For two and a half centuries the politics of the Near East had been dominated by a single, grinding rivalry between two empires of comparable might, and the sixth century found the contest at its height. To the west lay the Eastern Roman Empire, which historians by convention call Byzantine, after the old Greek name of its capital, Byzantium — refounded by the emperor Constantine in 330 CE as Constantinople, ‘the city of Constantine’, and known today as Istanbul. This was the Roman Empire still: it called itself Roman, its law was Roman, its emperor sat in unbroken succession from Augustus. But its language was now increasingly Greek rather than Latin, its faith was Christianity, and its centre of gravity had shifted decisively to the eastern Mediterranean.

The towering figure of the age was the emperor Justinian I, who reigned from 527 to 565 CE and had died just before our datum. His ambition was nothing less than the restoration of the Roman Empire in the West, and for a time he seemed to achieve it: his general Belisarius destroyed the Vandal kingdom of North Africa, recovered Italy from the Ostrogoths after a ruinous twenty-year war, and seized a foothold in southern Spain. Justinian left two monuments that have outlasted every conquest. The first is the Corpus Juris Civilis, the great codification of Roman law completed in the 530s, which became the foundation of the legal traditions of much of the modern world. The second is the church of Hagia Sophia in Constantinople, raised in a mere five years and dedicated in 537, whose vast floating dome remained the largest enclosed space in Christendom for the better part of a thousand years.

Yet the restoration was a mirage, and it was already dissolving as our window opens. The wars had bankrupted the treasury and the plague had thinned the armies. In 568 CE, only three years after Justinian’s death, a Germanic people called the Lombards poured into Italy and tore away most of what Belisarius had so painfully won, leaving Byzantium clinging to Ravenna, Rome, and the south. The empire that bestrode the eastern Mediterranean in 570 was magnificent, ancient, and dangerously overstretched.

Its great antagonist lay across the Euphrates: the Sasanian Empire of Persia, which ruled modern Iran and Iraq and reached deep into Central Asia, and which was heir to a Persian imperial tradition a thousand years old. Its capital was Ctesiphon on the river Tigris, near modern Baghdad, where the colossal brick vault of the royal audience hall — the Ṭāq Kasrā, still standing today — proclaimed the majesty of the King of Kings. Its faith was Zoroastrianism, the ancient dualist religion of the prophet Zarathustra, with its sacred fires and its cosmic struggle of light against darkness. And in 570 it too had a great king: Khosrow I, called Anūshirwān, ‘of the immortal soul’, who reigned from 531 to 579 and whom later Persian and Islamic tradition would remember as the very pattern of the just ruler. He reformed the taxes and the army and made his court a haven of learning: when Justinian closed the thousand-year-old philosophical Academy of Athens in 529 as a pagan relic, several of its scholars found refuge in Persia, and the academy of Gondēshāpūr became a meeting-place of Greek, Syriac, Persian, and Indian knowledge.

These two empires had fought each other to exhaustion over Mesopotamia, Armenia, and the Caucasus, and would soon fight the most terrible of all their wars, from 602 to 628, which left both so spent that neither could withstand what came out of the desert between them. That, however, is to run ahead. In 570 they were the two poles of the civilised world; and between them, unregarded, lay Arabia.

The Arabian peninsula in the sixth century was, to the great empires, a margin: a vast expanse of desert and oasis, mostly beyond the reach of any state, peopled by tribes who lived by herding, raiding, trade, and an intricate code of honour, and who spoke a family of closely related dialects we call Arabic. Islamic tradition would later call this pre-Islamic age the Jāhiliyya, the ‘age of ignorance’ — a religious judgement upon its idolatry rather than a description of its culture, for the Arabs of the Jāhiliyya possessed one of the supreme oral literatures of the ancient world. Their poetry, composed and recited in a highly refined common tongue and preserved in odes known as the muʿallaqāt, was the pride of the tribes and the arbiter of reputation; the Arabic language was a magnificent instrument long before it became the vehicle of scripture.

The margins of Arabia were, however, drawn firmly into the orbit of the great powers. Along the desert frontier of Byzantine Syria, the Christian Arab dynasty of the Ghassanids served as Rome’s clients and shield against Persia; along the Persian frontier of Iraq, the rival Arab kingdom of the Lakhmids, based at al-Ḥīra, did the same for Ctesiphon. And in the fertile south, in modern Yemen, the ancient and wealthy kingdom of Ḥimyar passed through a religious upheaval that previews, in miniature, the larger drama to come. In the early sixth century its rulers had adopted Judaism, and a Jewish king, Yūsuf — remembered as Dhū Nuwās — massacred the Christians of the city of Najrān around 523 CE. In reprisal, the Christian kingdom of Aksum, in modern Ethiopia, invaded across the Red Sea and made Yemen a Christian, Aksumite domain. Within fifty years, around 570, the Sasanians in turn drove out the Aksumites and made Yemen a Persian province. Jewish, then Christian, then Zoroastrian: the south of Arabia changed its official faith three times in two generations. The religious ferment that would soon produce Islam was not a bolt from a clear sky.

It is the Aksumite occupation of Yemen that brings us to the threshold of our subject. The Aksumite viceroy, a Christian general named Abraha, governed Yemen and built at Ṣanʿāʾ a great cathedral. According to the Arab and Islamic tradition, Abraha resolved to extend his power northward and to destroy the shrine that was the religious and commercial heart of the Hijaz — the Kaʿba at Mecca — and he marched against it with an army that included at least one war-elephant, a beast never before seen in those parts. The expedition ended in disaster before Mecca, and the year of its failure became known to the Arabs as the Year of the Elephant — ʿĀm al-Fīl. It is the year in which the most widely held tradition places the birth of the Prophet Muḥammad ﷺ, and the destruction of the elephant army is the subject of one of the shortest and most vivid chapters of the Qurʾān, the hundred-and-fifth, Sūrat al-Fīl, ‘The Elephant’.

In the understanding of the tradition, the proud army that came to destroy the House of God was annihilated by a power against which elephants and armies are as nothing — and the sanctuary was preserved for the Prophet who was to be born that very year.

A note of scholarly caution is in order, of the kind this series tries always to supply. The traditional date of 570 CE for both the Year of the Elephant and the Prophet’s birth is the one most widely transmitted in the Islamic sources, but some modern historians, weighing the South Arabian inscriptions, place Abraha’s campaigns somewhat earlier in the century. The precise year is uncertain; the significance is not.

Mecca itself was a sanctuary town in the arid valley of the Hijaz, owing its prosperity to two things: trade and the sacred. It lay upon the caravan route that carried the goods of the Indian Ocean and Yemen northward to Syria and the Mediterranean, and it was governed by the tribe of Quraysh, who were both its merchants and the custodians of its shrine. That shrine was the Kaʿba, a cube of stone which Islamic belief holds to have been raised by the patriarch Ibrāhīm (Abraham) and his son Ismāʿīl (Ishmael) as the first house of the worship of the one God. By the sixth century, however, it had become the centre of a thriving polytheism: the tradition records that it housed some three hundred and sixty idols, presided over by a chief deity called Hubal. The Arabs of the Hijaz already acknowledged a supreme creator-god whom they called Allāh — the word means simply ‘the God’ — but they associated with Him a host of lesser deities and intercessors, and venerated in particular three goddesses regarded as His daughters: al-Lāt, al-ʿUzzā, and Manāt, whom the Qurʾān would name and reject (Sūrat al-Najm 53:19–20).

To this shrine the tribes came on pilgrimage in the sacred months, when raiding and bloodshed were forbidden — a rite, the ḥajj, far older than Islam, which Islam would purify and reclaim. And amid the prevailing idolatry there were, the sources tell us, a few seekers known as the ḥunafāʾ — men such as Zayd ibn ʿAmr and Waraqa ibn Nawfal — who had rejected the idols and groped after the pure monotheism of Ibrāhīm. Jewish and Christian communities, too, were scattered through the peninsula: Jews at the northern oasis of Yathrib, the town later called Medina, and Christians at Najrān and among certain tribes. The religious soil of Arabia, in short, was neither barren nor uniform; it was a ferment of paganism, half-remembered monotheism, Judaism, and Christianity, waiting upon a catalyst.

Into this world, in the Year of the Elephant, the Prophet Muḥammad ﷺ was born, in Mecca, to the clan of Banū Hāshim of the tribe of Quraysh. His father, ʿAbd Allāh, had died before his birth; his mother, Āmina bint Wahb, would die when he was a small child, leaving him in the care first of his grandfather ʿAbd al-Muṭṭalib and then of his uncle Abū Ṭālib. He would grow to manhood as a respected and trusted member of his community — known, the tradition records, as al-Amīn, ‘the trustworthy’ — and it would not be until around the year 610, in his fortieth year, that he received in a cave above Mecca the first of the revelations that became the Qurʾān. All of that lay in the future at the moment of his birth. The point for this survey is the sheer magnitude of what was beginning, unseen, at the margin of the great powers’ world. The wealth and the armies of the age were concentrated at Constantinople and Ctesiphon; yet within a single century of the birth we have just described, the followers of the man born that year would have overthrown the Sasanian Empire entirely, stripped Byzantium of Syria, Egypt, and North Africa, and carried a new faith and a new language from the Atlantic to the frontiers of India and China.

East of Persia, the Indian subcontinent in 570 CE had recently lost its unifying power. The Gupta Empire, whose three centuries of rule are remembered as the classical golden age of Hindu civilisation — the age of the great Sanskrit poet Kālidāsa and of extraordinary advances in mathematics, astronomy, and the arts — had fragmented by the middle of the sixth century, broken in part by the invasions of a people the Indians called the Hūṇas, a branch of the Hunnic peoples who also troubled Persia and Rome. In the north, the land had dissolved into a patchwork of regional kingdoms; the conqueror Harsha, who would briefly reunite much of the north, was not yet born. In the south, the great dynasties of the Deccan and the Tamil country — the Chālukyas, the Pallavas, the Pāṇḍyas — pursued their own vigorous careers.

Politically divided, India remained an intellectual powerhouse without rival. It was here, around the turn of the sixth century, that the mathematician Āryabhaṭa composed his Āryabhaṭīya (499 CE), a work that gave a strikingly accurate value for π, set out a table of sines, and proposed — a thousand years before Copernicus was heard in Europe — that the apparent daily rotation of the heavens is caused by the rotation of the earth upon its axis. It was in India, too, that the decimal place-value system of numerals and the use of a symbol for zero were being perfected — the single most consequential invention in the history of arithmetic, which would pass by way of the Arabs to the whole world, and which we still, with unconscious accuracy, call the Arabic numerals. The cave-temples of Ajanta and the dawning rock-cut shrines of Ellora belong to this brilliant, fractured age, in which Hinduism in its classic form and a still-flourishing Indian Buddhism shared the religious landscape.

At the far end of Eurasia, the China of 570 CE presents a mirror to the fallen West. In our previous survey China had been a single Han empire of sixty million, the equal of Rome. But the Han had collapsed in 220 CE, and China had then endured nearly four centuries of division — the long era known as the Northern and Southern Dynasties. In 570 the country was split: in the north, two rival states, the Northern Zhou and the Northern Qi, ruled over a population much intermarried with the steppe peoples who had conquered it; in the south, the Chen dynasty held the Yangtze lands from its capital at modern Nanjing. As Rome had been one and was now many, so China had been one and was now many — but, crucially, the Chinese division was about to end. In 581 CE, only a decade beyond our window, the Sui dynasty would seize the north and, in 589, conquer the south, reuniting China after three and a half centuries and inaugurating the great medieval flowering of the Tang. In 570 that reunification was imminent but unaccomplished.

Buddhism, which had entered China along the Silk Road in the Han twilight, had by now sunk deep roots, and the period of division was in many ways its golden age in China — an age of vast monasteries, colossal cave-sculptures, and pilgrim-scholars. It was not without its convulsions: the emperor of the Northern Zhou launched a sweeping persecution of Buddhism in the 570s, the better to harness the realm’s wealth and manpower for the coming wars of reunification. And from China the faith was even then reaching the islands of Japan, to whose court Buddhism is traditionally said to have been introduced in 552 CE — a date that falls squarely within our window, and that marks the beginning of one of the world’s great religious and cultural transformations.

Between the settled empires, across the immense grasslands from Mongolia to the Black Sea, a new power had just arisen that would shape the next thousand years of Eurasian history. In 552 CE a chieftain named Bumin overthrew his overlords and founded the First Turkic Khaganate — the earliest state in history to bear the name of the Turks. Within a decade the Turks, in alliance with Sasanian Persia, had destroyed the Hephthalites, the so-called White Huns who had dominated Central Asia and harried both Persia and India; and the Turkic realm, briefly the largest land empire the world had yet seen, came to straddle the Silk Road and to share a frontier with Persia, with China, and — by way of the steppe corridor north of the Black Sea — with Byzantium itself.

The Turks understood their commercial advantage, and in 568 CE a Turkic embassy, led by a Sogdian merchant named Maniakh, reached Constantinople to propose a joint commerce in silk and a joint war against Persia — one of the more remarkable diplomatic episodes of the age, linking the courts of the eastern steppe and the eastern Mediterranean directly for the first time. Meanwhile another steppe people displaced by the Turkic expansion, the Avars, had ridden westward into the heart of Europe and established in the Hungarian plain a khaganate that would terrorise the Balkans and the Franks for two centuries. The appearance of the Turks as a named power upon the world stage is one of the quietly momentous facts of the sixth century.

In the lands where the Western Roman Empire had stood until its formal extinction in 476 CE, a new political order had taken shape: a mosaic of kingdoms ruled by the Germanic peoples who had settled within the old imperial frontiers. The most durable of these was the kingdom of the Franks in Gaul — modern France and the Rhineland — under the Merovingian dynasty founded by Clovis, who around the year 500 had united the Franks and, decisively, converted to Catholic rather than heretical Christianity, binding the new Frankish monarchy to the Roman Church. By 570 the Frankish realm was divided among Clovis’s quarrelsome descendants, whose savage and colourful dynastic feuds are chronicled, with appalled relish, by Bishop Gregory of Tours, our indispensable witness to the sixth-century West.

In Spain ruled the Visigoths, still adherents of the Arian form of Christianity that the wider Church deemed heretical, but upon the verge — in 589, just beyond our window — of converting to Catholicism and founding a durable Hispano-Gothic kingdom. In Italy, as we have seen, the Lombards had overrun the north and centre in 568, leaving a patchwork of Lombard duchies and shrinking Byzantine enclaves. And in the city of Rome itself, shrunken and impoverished, shorn of the empire it had once ruled, a new kind of authority was rising in the person of the bishop — the Pope: within twenty years the formidable Gregory the Great would assume the office and, in 597, dispatch the mission that began the conversion of the English. The West of 570 was poorer, more rural, and more fragmented than the West of the year 1; but it was not the lightless ‘Dark Age’ of legend, and the institutions — the Church above all — that would build medieval Europe were already in place.

Nowhere does the unevenness of the historical record show more starkly than in Britain, which in this period nearly vanishes from written history altogether. Roman rule had ended around 410 CE, and the fifth and sixth centuries saw the gradual settlement of eastern and southern Britain by Germanic peoples from across the North Sea — the Angles, Saxons, and Jutes whose descendants would become the English — and the slow retreat of the native, Christian, Brittonic-speaking Britons into Wales, Cornwall, and the north. The sixth century is the most poorly documented stretch of British history since the Iron Age. Our chief contemporary source is a furious sermon, On the Ruin of Britain, written around the 540s by a British monk named Gildas, who lamented the calamity that had befallen his people but recorded almost no names and no dates.

This is the shadowed century to which legend assigns King Arthur — the British war-leader who, in the much later stories, held back the Saxon tide. The honest verdict is that there is no contemporary evidence for Arthur as a historical individual; he may rest upon some real fifth- or sixth-century commander, but he belongs, as we can recover him, to legend rather than to history — a striking thing to have to say of a period so close to the well-lit empires of the Mediterranean. By 570 the patchwork of small Anglo-Saxon kingdoms that would coalesce into England was forming, and it was pagan; Christianity would not return to the English until the close of the century. Ireland, by contrast, never Roman and already Christian, had become a beacon of monastic learning, and in 563 CE the Irish monk Columba founded the monastery of Iona off the Scottish coast, whence the conversion of much of northern Britain would proceed.

South of the Sahara, the dominant power of the age was the kingdom of Aksum, in the highlands of modern Ethiopia and Eritrea, whose intervention in Yemen we have already met. Aksum was a Christian state — it had adopted the faith in the fourth century — that struck its own gold coinage, controlled the Red Sea trade, and was reckoned by a contemporary writer among the four great kingdoms of the world, alongside Rome, Persia, and China. Its conquest of Yemen around 525 marked the height of its power; by the century’s end, with the Persians installed across the strait, that power was beginning its long contraction, but Aksum remained one of the most sophisticated societies of the age and the seedbed of the enduring Christian civilisation of Ethiopia.

Elsewhere in Africa the picture is drawn more from archaeology than from texts. Across the Sahara the camel had by now made regular trans-desert trade possible, and the West African societies that would later crystallise into the great empire of Ghana were taking shape along its southern edge. Throughout central, eastern, and southern Africa the long migration of Bantu-speaking farmers and ironworkers — one of the largest movements of people in human history — continued to carry agriculture, cattle, and iron metallurgy into new lands, a vast and consequential process conducted entirely beyond the horizon of the literate world.

On the far side of an ocean that no one would cross for another nine hundred years, the peoples of the Americas pursued their histories in complete isolation, ignorant of the Old World as it was ignorant of them. And here the sixth century offers its sharpest inversion of the Old World’s fortunes, for while the Mediterranean reeled under cold and plague, the great civilisations of the Americas stood at or near their height. In the forests of modern Guatemala and southern Mexico, the Classic Maya were in their full flower — a constellation of literate city-states, dominated by the long rivalry of the two superpowers Tikal and Calakmul, whose kings raised towering temple-pyramids and recorded their dynasties and wars in the only complete writing system ever developed in the pre-Columbian Americas. The Maya tracked the heavens with great precision, sustained a sophisticated mathematics, and, like the Indians on the other side of the planet and quite independently, employed a symbol for zero.

In the highlands of central Mexico the vast city of Teotihuacan — laid out upon a grid around its colossal Pyramids of the Sun and Moon, and home, at its peak, to well over a hundred thousand people, which made it one of the largest cities on earth — was in this very century passing the summit of its power and approaching the violent collapse that would overtake it around 550 to 650 CE. In the Andes of South America the Moche of the northern Peruvian coast were raising their great adobe platforms and producing metalwork and ceramics of astonishing artistry, while the highland states of Tiwanaku and Wari were expanding toward the first true empires of the Andean world. None of these peoples possessed iron, the wheel for transport, or — beyond the Andean llama — large domestic animals; and yet they built cities, mapped the stars, and kept records to rival much of the Old World. They were not, as an older condescension imagined, primitive laggards on a single ladder of progress. They were a wholly independent answer to the question of how human beings might live at scale.

The question recurs in every instalment of this series, and it carries, as always, a concealed bias: the word ‘advanced’ quietly rewards cities, writing, states, and monuments, and slights the profound sophistication of those who chose to live otherwise. With that caution entered, the conventional answer for 570 CE points to no single leader but to a handful of peers. The two great western empires — Byzantium, with its Roman law, its domed architecture, and its Greek learning, and Sasanian Persia, with its administration, its scholarship, and its imperial grandeur — stood at the summit of the western world. China, though divided, was a literate, bureaucratic, iron-working civilisation about to reunite into the most powerful state on earth. India, though equally divided, led the world in mathematics and astronomy. And in total isolation the Maya and the builders of Teotihuacan raised monuments and kept records that no contemporary in the Old World would have dismissed.

The honest conclusion is the one this series has reached at every date: it is more accurate to speak of several remarkable civilisations, each brilliant by its own lights, than of a single vanguard of mankind. What distinguishes 570, perhaps, is that the centre of gravity of the most powerful states was beginning one of its great historical shifts — away from the exhausted Mediterranean and toward the about-to-be-reunified East, and toward a desert margin that no one yet thought to watch.

By the sixth century the Old World was, as it had become around the turn of the era, a single connected system — loosely, indirectly, and at second or third hand, but really. The overland Silk Road still carried goods and ideas across Eurasia, now under the watchful control of the new Turkic khaganate; the monsoon-driven trade of the Indian Ocean still linked the Mediterranean, by way of the Red Sea and the Persian Gulf, to India, South-East Asia, and ultimately China. A vivid emblem of this connection is the story, recorded by the Byzantines, of how around 552 CE two monks smuggled the eggs of the silkworm out of China — concealed, it was said, in a hollow staff — and so founded the Byzantine silk industry and broke the long Persian monopoly upon the trade. Religious ideas travelled the same roads: the Church of the East, the so-called Nestorian Christians whose home was in Sasanian Persia, sent missionaries far along the Silk Road into Central Asia and India, and would reach the Chinese capital within two generations.

And yet the limits of this connectedness must be kept in view. The contact was that of trade goods and travellers’ tales, not of governments or peoples who understood one another; Byzantium and China remained, to each other, lands of rumour at the edge of the map. And between the hemispheres there was still no contact whatever. The Americas and the Old World knew nothing of each other’s existence, and would continue in mutual ignorance for nine centuries more. The connected world of 570 was a single Old World system; the planet as a whole was not yet one.

As in every survey in this series, it must be said clearly that the racial categories of the modern world do not apply to the sixth century and were not built for it. Humanity then, as now and always, was a single species varying continuously across the globe, and the peoples of 570 sorted one another not by the crude colour-coded races of later centuries but by the things that mattered to them: religion above all, then language, citizenship, and tribe. A Greek-speaking Christian of Constantinople and a Latin-speaking Christian of Carthage felt themselves kin across a difference of complexion; a Persian Zoroastrian and an Arab idolater were divided by faith, not by phenotype. To project the racial map of the nineteenth or twentieth century back upon this world is to misunderstand it at the most basic level.

For most of the literate Old World we can now name, and read, the languages of the age. Greek was the tongue of the Byzantine state and Church and of a great Christian literature; Latin, receding in the East, remained the language of the Western Church and of learning in the post-Roman kingdoms. Syriac, a form of Aramaic, was the great lingua franca of Near Eastern Christianity and scholarship, the vehicle in which much Greek philosophy and science would pass to the Arabs. Middle Persian, or Pahlavi, was the language of Sasanian Persia and of the Zoroastrian scriptures. Sanskrit and its derived Prakrits carried the learning and literature of India; a middle form of Chinese served the world’s oldest continuous literary tradition; Geʿez served Christian Aksum, Coptic the Egyptian Church, Armenian its own ancient nation. And across the steppe the Turkic language was now spoken from Mongolia to the west, though its earliest inscriptions lay a century and more in the future.

Two tongues deserve special mention on the eve of their transformation. The first is Arabic, already a magnificent vehicle of oral poetry, which was about to become — through the Qurʾān — one of the great scriptural, literary, and administrative languages of the world, and to spread across three continents. The second is the silence that still surrounds most of humanity: the unwritten languages of sub-Saharan Africa, of the Americas, of the Pacific, and of the European and Asian countryside, spoken by the great majority of the living and recorded by almost none of them, and so largely lost to us. Even in this most articulate of ages, history speaks chiefly for the literate few.

It is in the history of religion that the sixth century reveals itself as one of the true hinges of the human story, and the survey is best ended there. Christianity was now the dominant faith of the Mediterranean world, but it was a Christianity bitterly divided — not yet between Catholic and Orthodox, a rupture centuries away, but over the nature of Christ. The Council of Chalcedon of 451 CE had defined Christ as one person in two natures, divine and human; great Churches rejected the formula, and by 570 the Christian East was split between the Chalcedonian establishment of Constantinople, the Miaphysite Churches of Egypt, Ethiopia, Armenia, and much of Syria, and the Church of the East in Persia. These were not academic quarrels; they shaped the loyalties of whole provinces, and the disaffection of Miaphysite Egypt and Syria would help to explain how readily those lands fell to the Arabs in the following century.

Beyond the Christian world, Zoroastrianism remained the state religion of Persia, and Judaism endured, scattered and resilient, across the empires, its rabbinic learning crowned in this very period by the completion of the Babylonian Talmud in the academies of Mesopotamia. Buddhism flourished from India through Central Asia to China and, newly, to Japan; Hinduism stood in its classic form across the subcontinent. The great religious map of the Old World was, in most respects, the map we should recognise today — with one immense exception. The youngest of the world faiths did not yet exist. Islam was not a religion in 570 CE; it was an event waiting in the womb of time, bound up with the life of the child whose birth that year the tradition records. Within sixty years of that birth the Arabian peninsula would be united under the worship of the one God; within a century, as we have noted, the new faith would have toppled one of the two great empires of this survey and torn the richest provinces from the other. No survey written from the vantage of 570 could have predicted it. That is the abiding lesson of the date.

It is fitting to close, as the earlier essays did, upon the moving frontier between what is known and what is not — and to mark how strangely that frontier runs in the sixth century. Ten thousand years ago the whole of human experience lay on the unknowable side of the line, recoverable only through stone and bone. Five thousand years ago, in two river valleys and a handful of clay tablets, the line had just begun to give way. Two thousand years ago, across the breadth of the literate Old World, it had given way for good, and history had become a roar rather than a whisper. Fifteen hundred years ago that historical world endured, fuller and better documented than ever across most of Eurasia — and yet, in the fogs of post-Roman Britain, the light of the written record guttered almost to nothing, so that within sight of the best-recorded empires in human history a whole society slipped back, for a century, into the half-darkness of legend. The frontier of knowledge, this date reminds us, is not a single advancing line but a ragged and uneven thing, bright here and dim there even within one age.

And the deeper lesson of the sixth century is one of humility before history itself. The men who held power in 570 — the heirs of Justinian in their domed and golden city, the King of Kings in his vaulted hall at Ctesiphon — could see, with the clarity of the well-informed, the shape of the world and the location of its wealth and its danger. Not one of them was looking at Mecca. The hinge upon which their world was about to turn lay not at the centre, where the armies and the treasuries were, but at the unregarded margin, in the life of a child born that year in a trading town in the desert. It is the kind of thing history does, and rarely announces. This essay, written fifteen hundred years further on, is a distant link in the long human conversation that the sixth century carried forward through cold and plague and war; and if it has a moral, it is that the conversation’s most decisive turns are seldom heard, by those alive at the time, above the noise of the powerful.

The question of how human diversity arose was, for the classical scholars, a less troublesome matter than one might suppose. The Companions and their successors moved amongst Abyssinians, Persians, Byzantines, and Turks; they were aware of more distant peoples, including the Chinese (al-Ṣīn); and they observed in their own communities the considerable variation in colour, build, and temperament that the human form admits. Diversity was not a phenomenon awaiting explanation. It was a settled fact about the world, to be accommodated rather than puzzled over. The principal textual basis for that accommodation is the ḥadīth narrated by Abū Mūsā al-Ashʿarī (raḍiya Allāhu ʿanhu): that Allah created Ādam (Adam) from a handful He took from the whole of the earth, so that the children of Ādam came forth in accordance with it — "some red, some white, some black, and some between that; some gentle, some harsh, some wicked, and some good." The narration is recorded in Sunan Abī Dāwūd (no. 4693) and Jāmiʿ al-Tirmidhī (no. 2955), the latter grading it ḥasan ṣaḥīḥ; it attributes physical and temperamental diversity directly to the substance of Ādam's creation, framing it as divinely intended rather than anomalous. The reader of the earlier surveys will recognise the picture at once. The ḥadīth's gradation — red, white, black, "and some between that" — is, in the idiom of the seventh century, a description of precisely the clinal variation that ancient DNA has since confirmed: a single humanity shading continuously across the earth, with no sharp lines and no privileged centre. What modern genetics states as the falsity of discrete racial types, the tradition had stated as the unity of a family taken from the whole earth.

Qurʾānic reinforcement comes in Sūrat al-Rūm (30:22), where the diversity of tongues and colours is numbered amongst the signs of Allah, and in al-Ḥujurāt (49:13), where the existence of peoples and tribes is presented as a deliberate ordering for mutual recognition — litaʿārafū — rather than for ranking. For the post-Deluge dispersal, the classical historians constructed a genealogy in which Sām (Shem), Ḥām (Ham), and Yāfith (Japheth), the three sons of Nūḥ (Noah), peace be upon him, were the progenitors respectively of the Semitic, African, and northern and eastern peoples — a framework preserved in al-Ṭabarī's Tārīkh, al-Masʿūdī's Murūj al-Dhahab, and Ibn Kathīr's al-Bidāya wa'l-Nihāya, though it draws on Isrāʾīliyyāt material that Ibn Kathīr himself cautioned should not be relied upon uncritically.

The more analytically interesting contribution came from Ibn Khaldūn in the Muqaddima. In his discussion of the seven climes he attributed physical differences — most prominently skin colour — to climate and environment rather than to ancestral curse, directly refuting the "curse of Ḥām" narrative that had entered some Muslim writings through Jewish and Christian intermediaries, and arguing that the darkness of sub-Saharan Africans was a function of intense solar exposure over generations. This is, in modern terms, a recognisably proto-environmental account of phenotypic variation, and it sat entirely comfortably with a monogenic origin from Ādam and Nūḥ. An earlier and more striking work still is al-Jāḥiẓ's Fakhr al-Sūdān ʿalā al-Bīḍān — "The Boast of the Blacks over the Whites" — composed in the third Hijrī century, which not only assumes the common Adamic origin but mounts an explicit defence of the virtues of black peoples. The classical synthesis, in short, was perfectly capable of handling diversity. Unity of origin and diversity of appearance posed no tension whatever — the very conclusion, reached from the opposite direction, that closed the racial section of every survey before this one.

The tension the educated Sunni encounters today is therefore not over diversity at all. It is over a single population-genetic constraint: the finding, emerging from the past quarter-century of genomic research and sitting at the five-sigma end of scientific reliability, that the present human gene pool cannot derive from a single breeding pair within any plausible recent timeframe. The argument rests on no one piece of evidence; it is the convergence of independent lines that closes the question. And it is, pointedly, the same body of work the earlier surveys leaned upon — the ancient-DNA revolution associated with David Reich and others, which showed the modern populations forming by admixture — now turned to a sharper end.

The simplest version turns on combinatorial constraint. A diploid individual carries at most two alleles at any locus, so a founding pair carries at most four; yet at many loci humans exhibit far more than four distinct alleles. The point becomes decisive at the Human Leukocyte Antigen genes, where thousands of alleles persist in the modern population, and where many specific variants are demonstrably older than the human–chimpanzee split, sharing closer kinship with particular chimpanzee or gorilla alleles than with other human ones. These "trans-species polymorphisms" can be maintained only if the human lineage preserved them continuously from before that split, some six to seven million years ago. A bottleneck of two individuals at any point in that interval would have erased them; their presence is a direct molecular record that no such bottleneck occurred.

The picture is corroborated by whole-genome coalescent reconstruction. Methods such as PSMC and its descendants recover ancestral effective population size as a function of time from a single genome; the output is a curve, not a number, and across the entire reconstructable range — hundreds of thousands of years — it oscillates between roughly ten thousand and thirty thousand, never dropping near unity. Founder bottlenecks have occurred — the Out-of-Africa migration around sixty thousand years ago, the very event the earlier surveys invoked to explain why the Americas and Australia were peopled from small founding groups — but their floor lies in the low thousands, never the units.

The popular invocation of "Mitochondrial Eve" and "Y-chromosomal Adam" deserves explicit correction, for the names mislead. These are statistical constructs of coalescent theory: the woman to whom all living humans' mitochondrial DNA traces, and the man to whom all Y chromosomes trace. They lived at different times, were not partners, and — most importantly — were not the only humans alive in their respective eras, but members of populations numbering in the thousands, whose other uniparental lineages happened, by stochastic extinction, to die out. The autosomal genome tells an entirely different story, in which each individual's ancestry fans out to thousands of distinct ancestors at any given depth.

How is the believer to hold such a finding? The classical framework for apparent conflicts between revealed text and demonstrated reason was articulated most lucidly by Imām al-Ghazālī and later by Ibn Rushd. Al-Ghazālī — for all his celebrated refutation of the philosophers — was clear that genuinely demonstrated rational truths cannot contradict revelation, and that where an apparent contradiction arises one must find either that the supposed demonstration is flawed or that the apparent meaning (ẓāhir) of the text admits of interpretation (taʾwīl). Ibn Rushd, in Faṣl al-Maqāl, made the principle explicit: truth cannot contradict truth, and where a demonstrative proof contradicts the literal sense of scripture, the text must be read figuratively.

This is the discipline the series has, in its own way, observed throughout. When the first survey set aside the facile identification of Göbekli Tepe with Eden, it was refusing to read revelation into the ground where revelation had said nothing; the present essay refuses the opposite error, of reading the ground out of revelation where reason has demonstrated something. The two refusals are one principle. The working epistemology is not credulous capitulation to whatever science currently asserts, but disciplined engagement with what is genuinely demonstrated. The five-sigma findings constrain the field of viable readings; they do not, on the classical methodology, dictate any particular one.

The five-sigma findings constrain the field of viable readings; they do not, on the classical methodology, dictate any particular reading.

Contemporary Sunni scholarship has produced a spectrum of positions, all within orthodoxy. At one pole, Shaykh Nūḥ Ha Mim Keller and many traditional ʿulamāʾ reject macroevolution entirely, treating Ādam's creation from clay (Qurʾān 3:59, 15:26, 38:71) as a non-negotiable literal datum. A middle position accepts biological evolution in general but exempts Ādam (peace be upon him) from it, treating him as a special, de novo creation; David Solomon Jalajel's Islam and Biological Evolution (UWC, 2009) is a sustained treatment in this register, and Shoaib Ahmed Malik's Islam and Evolution: Al-Ghazali and the Modern Evolutionary Paradigm (Routledge, 2021) is the most rigorous English-language treatment to date.

At the other pole, Nidhal Guessoum (Islam's Quantum Question, I.B. Tauris, 2010) and a smaller group argue for full theistic evolution, with the Ādam narrative read theologically rather than biologically. Between these, and most useful for the present purpose, is the position Malik categorises as "Ādam exception with polygenesis" — or, in its Christian-tradition analogue, the Genealogical Adam framework articulated by S. Joshua Swamidass (The Genealogical Adam and Eve, IVP Academic, 2019). On this view Ādam and Ḥawwāʾ (Eve) are real historical individuals, specially created and divinely ensouled, who entered amongst an existing population of anatomically near-modern hominins; their lineage spread through that population over subsequent millennia, theological humanity propagating along genealogical lines whilst the biological gene pool drew on a much wider ancestral base. It is no accident that this is the reading the genetics leaves standing: it is the one that distinguishes genealogical descent from genetic descent, and so absorbs the coalescent findings without strain.

The polygenesis model accommodates the genetic data straightforwardly. It also requires two precisifications that often go unmade in popular discussion. The first concerns the indigenous population. By the time of any plausible Adamic moment — somewhere in the last one to three hundred thousand years — the hominins amongst whom Ādam and Ḥawwāʾ entered were nothing like chimpanzees.

They were anatomically modern or near-modern Homo sapiens, Neanderthals, and Denisovans, with brain volumes comparable to our own, complex tool use, art, burial, plausibly language, and elaborate social organisation — the very peoples, indeed, whose deep ancestry the earlier surveys traced. The human–chimp common ancestor lies six to seven million years deep; the populations relevant to the model lie within the last few hundred thousand years and are, biologically, our cousins rather than our remote relatives. The model is best stated thus: Ādam and Ḥawwāʾ as ensouled individuals introduced into a population of biologically near-identical but theologically non-human hominins.

The second precisification concerns the nature of the distinction. The popular shorthand — that Ādam was the first conscious human — is not quite right in the philosophical register. Phenomenal consciousness, the bare capacity to experience, is present in many animals and was almost certainly present in the indigenous hominin populations. What Islamic theology marks out as constitutive of theological humanity is more specific: the rūḥ whose inbreathing is referred to in Qurʾān 15:29 and 38:72, which brings with it taklīf (moral accountability), the capacity for maʿrifa (knowledge of Allah), the reception of divine address as in the primordial covenant of Qurʾān 7:172, and the moral and spiritual faculties that follow. Ensoulment, not consciousness, is the right register: it allows the pre-Adamic hominins to have been intelligent, social, and conscious in the phenomenal sense without being insān in the theological sense.

Ensoulment, not phenomenal consciousness, is the right register: it allows pre-Adamic hominins to have been intelligent and social without being insān.

The model needs a position on the origin of Ḥawwāʾ.

The Qurʾān (4:1, 7:189) refers to her creation minhā — from him, or from it. Three readings circulate. The first is that Ḥawwāʾ was created directly from Ādam, the classical reading supported by the ḥadīth on the creation of woman from the rib (Ṣaḥīḥ al-Bukhārī no. 3331; Ṣaḥīḥ Muslim no. 1468). The second is that she was selected from the existing population and ensouled alongside Ādam, with minhā read as her being of the same kind. The third is that she was a de novo creation alongside him. The model is not internally specified until this choice is made.

The doctrinal commitments that remain qaṭʿī — definite, and binding — are the existence of Ādam (peace be upon him) as a real historical individual, the first prophet, the father of humanity in some meaningful sense, and the recipient of divine address; and the historicity of Nūḥ (peace be upon him) and the Deluge in some form. What is ẓannī — open to considered interpretation — is the precise biological mechanism of Ādam's coming into being, and whether "father of humanity" must mean the sole biological progenitor of the present gene pool, or whether it may mean the first ensouled and accountable human and the federal head of a covenant. Each reading has serious scholarly proponents within orthodoxy; the choice amongst them is a question of ijtihād in hermeneutics, not a binary of faith against apostasy.

It is worth being exact about what the science does and does not do here, for the temptation to ask it for more than it can give is strong. The genomic findings corroborate the viability of the genealogical reading and, in the same motion, foreclose the strict claim that Ādam was the sole biological ancestor of all living people within a recent span. They support, that is, the position the educated Sunni is most likely to hold, against its rivals within the tradition. What they cannot do is establish the theological content itself — the ensoulment, the taklīf, the covenant, the prophethood — for these lie beyond the reach of any genome, on the far side of the line the surveys called unknowable. The demonstrated reason of Ghazālī's equation is supplied by the science; the revealed text supplies the rest; and the believer who holds both, in the manner of Ghazālī and Ibn Rushd, has not strained his faith but exercised it.

The series began with a planet of mute foragers and ended, four surveys later, on the birth of the Prophet ﷺ and the coming of the final revelation. It is fitting that it should close on the oldest question of all, and fitting too that the answer should be the one its first pages already implied. The deepest science and the plainest sense of the tradition converge on a single proposition: that humanity is one — one species, drawn from the whole earth, descended into a single family, and addressed by a single Creator. Ghazālī and Ibn Rushd would have recognised the present question as a perfectly ordinary instance of the accommodation they were trained to perform; and the foragers of the first survey, could they have been asked, would have understood the conclusion better than they would have understood the taxonomy. The frontier between what we know and what we do not has moved, in this essay, almost to the very beginning of the human story — and there, at its edge, reason and revelation are found not at war but in concord.