Most of human history is prehistory. Of the 200,000 or more years that humans have spent on Earth, only a tiny fraction have been recorded in writing. Even in our own little sliver of geologic time, the 12,000 years of the Holocene, whose warm weather and relatively stable climate incubated the birth of agriculture, cities, states, and most of the other hallmarks of civilisation, writing has been more the exception than the rule.
Professional historians can’t help but pity their colleagues on the prehistoric side of the fence. Historians are accustomed to drawing on vast archives, but archaeologists must assemble and interpret stories from scant material remains. In the annals of prehistory, cultures are designated according to modes of burial such as ‘Single Grave’, or after styles of arrowhead, such as ‘Western Stemmed Point’. Whole peoples are reduced to styles of pottery, such as Pitted Ware, Corded Ware or Funnel Beaker, all of them spread across the map in confusing, amoeba-like blobs.
In recent years, archaeologists have become reluctant to infer too much from assemblages of ceramics, weapons and grave goods. For at least a generation, they have been drilled on the mantra that ‘pots are not people’. Material culture is not a proxy for identity. Artefacts recovered from a dig can provide a wealth of information about a people’s mode of subsistence, funeral rites and trade contacts, but they are not a reliable guide to their language or ethnicity – or their patterns of migration.
Before the Second World War, prehistory was seen as a series of invasions, with proto-Celts and Indo-Aryans swooping down on unsuspecting swaths of Europe and Asia like so many Vikings, while megalith builders wandered between continents in indecisive meanders. After the Second World War, this view was replaced by the processual school, which attributed cultural changes to internal adaptations. Ideas and technologies might travel, but people by and large stayed put. Today, however, migration is making a comeback.
Much of this shift has to do with the introduction of powerful new techniques for studying ancient DNA. The past five years have seen a revolution in the availability and scope of genetic testing that can be performed on prehistoric human and animal remains. Ancient DNA is tricky to work with. Usually it’s degraded, chemically altered and cut into millions of short fragments. But recent advances in sequencing technology have made it possible to sequence whole genomes from samples reaching back thousands, and tens of thousands, of years. Whole-genome sequencing yields orders of magnitude more data than organelle-based testing, and allows geneticists to make detailed comparisons between individuals and populations. Those comparisons are now illuminating new branches of the human family tree.
When the first draft of the Neanderthal genome was published in 2010, it showed that modern humans from Europe and Asia shared an average of 2 per cent of their DNA with Neanderthals, meaning that the two species must have mated at some point in the past – possibly when humans left Africa for the Middle East. Also in 2010, DNA from a finger bone led to the discovery of a whole new type of early human – the Denisovans – with whom Asians and Australians also share ancestry.
The timelapse image of prehistoric migration has only become clearer in recent years. Last October, a 45,000-year-old human leg bone from Siberia was shown to contain Neanderthal DNA in longer fragments than those found in modern humans. Longer Neanderthal DNA fragments date the DNA to a time much closer – probably within 10,000 years – to when Neanderthals and humans first swapped genes. (Imagine it like a stack of cards being cut – with each generation, the Neanderthal chromosomes were sliced into shorter and shorter pieces.) In June this year, the genome of one of the first modern-human fossil remains from Europe, a 40,000-year-old tooth from Romania, was found to contain even longer strands of Neanderthal DNA. Its owner had a Neanderthal ancestor only four to six generations in their past.
This progression gives a feel for how fast the discipline of ancient DNA studies is progressing.
In five years, we’ve gone from thinking we shared no DNA with Neanderthals, to realising that there was widespread interbreeding, to pinpointing it (for one individual) within 200 years – almost the span of a family album. But the use of ancient DNA isn’t limited to our near-human relatives. It is also telling us about the dispersal of humans out of Africa, and the origin and spread of agriculture, and the peopling of the Americas. It is also helping archaeologists crack one of the great mysteries of prehistory: the origins of the Indo-Europeans.
William Jones, a Welsh jurist and philologist, arrived in Calcutta from London in 1783, to take his seat on the Supreme Court of Bengal. As part of his duties, Jones had to familiarise himself with the precepts of Hindu law; to do that, he had to learn Sanskrit. Jones promptly hired a pandit – a Hindu religious scholar – as a tutor, and immersed himself in ancient Sanskrit texts. After three years of study, Jones arrived at a startling conclusion: Sanskrit, the classical language of India, was linked, in its grammar and vocabulary, to Greek and Latin, the classical languages of Europe. Not only that, all three were further related to Germanic, Celtic and Persian languages.
Later linguists added even more branches to this family tree, including living languages, such as Slavic, Albanian, Armenian, Baltic, and extinct ones such as Hittite and Tocharian. And if all these languages were related, logic dictated that somewhere in the distant past, they shared a common ancestor. Experts call this ancestral language Proto-Indo-European, or PIE. Its daughter languages are spoken by 3 billion people. Their natural range stretches from Sri Lanka to Portugal. English is an Indo-European language, as is nearly every other language in Europe besides Hungarian, Finnish and Basque.
In the more than 200 years since Jones announced his discovery, linguists and philologists have painstakingly reconstructed what PIE sounded like and what grammatical rules it followed. We can write in it, and some have even composed short stories in PIE, about things such as sheep, wolves and gods. Thanks to the work of students of comparative mythology and reconstructed word lists, we have some idea of Proto-Indo-Europeans’ lifestyle and what they believed. They lived among oak and beech trees, and savoured mead. They rode in wagons. They cared more for fathers and brothers than aunts and sisters-in-law. They thanked the Sky Father for sons, fat cattle and swift horses. They might or might not have been acquainted with salmon.
But in all this time, PIE has remained a learned conjecture, a hypothetical language spoken by a putative people. The Proto-Indo-Europeans are a philological phantasm and, like many ghosts, they have a habit of showing up all over the place. Would-be savants have located the PIE homeland everywhere from Scandinavia to the Tibetan Plateau to the North Pole, but in recent decades, scholarly consensus has converged around two origin stories for the PIE people.
One, most prominently argued by the British archaeologist Colin Renfrew, says that the Indo-Europeans originated somewhere on the edge of the Fertile Crescent. In this version, the key to the PIE people’s successes lay in their use of agriculture, which gave them a long-term edge over the surrounding hunter-gatherers. According to this theory, the expansion of Proto-Indo-Europeans began early, in the Neolithic, when people from Anatolia began to diffuse slowly, like a softly-heated gas.
Bronze enlarged the world, tying Eurasia in a network of trade. The horse shrunk it down, enabling a mobile lifestyle
The second leading theory of Indo-European origins, propounded most forcefully by the late Lithuanian-born archaeologist Marija Gimbutas, and more recently, by the American archaeologist David W Anthony, is known as the Steppe Hypothesis. It tells us that the Indo-Europeans originated in the steppes of southern Russia, and spread out from there thanks to a combination of new adaptations, principally the horse and the wheel. This story begins much later than the Anatolian theory, in the Bronze Age instead of the Stone Age.
The manufacture of bronze required two metals – copper and tin – which necessitated long-distance trade. Agents from the temple cities of Mesopotamia spread far and wide to acquire copper and tin, reaching all the way east to Afghanistan and north to the Russian steppes. They took with them new ideas about property, wealth and war. Meanwhile, in the steppes, someone had domesticated the horse. Wild horses were native to the plains of southern Russia. In cold weather, they knew how to kick up snow to get at grass, making them a perfect reserve (meat) animal for winter. Only later did someone realise that horses could be ridden, and yoked, to pull wagons and plows. Bronze enlarged the world, tying different parts of Eurasia together in a network of trade. The horse shrunk it down, enabling a new, mobile lifestyle. The Proto-Indo-Europeans were the chief beneficiaries of these innovations.
The Anatolian theory has the virtue of a clear demographic logic and a bit of genetic evidence behind it as well. Early studies of the genetic variation in Europe showed increasing diversity the further you moved from Turkey, which would make sense if a lot of Europe’s original population originated in the Balkans and spread out from there over time. Lately though, the preponderance of opinion has shifted towards a steppe origin for PIE. In addition to various streams of archaeological evidence, a steppe origin for the Indo-Europeans accords well with PIE’s shared vocabulary for wheels and wagons, and the seeming fascination with horses found across Indo-European mythologies. But even if we assume the Steppe Hypothesis, it’s not clear – did Indo-Europeans spread through elite capture, or mass migration?
Geneticists are beginning to chip away at this mystery. Two teams, one based at Harvard and another at the University of Copenhagen, gathered DNA from 170 Bronze Age skeletal remains found across Europe and Central Asia, and both studies detected a major population move from southern Russia into western Europe. Some time around 2500 BC, the genetic signature of the Yamnaya (Pit-Grave) people, herders from southern Russia, starts showing up all over Germany.
The Bronze Age gave us some of our earliest literature, much of which, like the Greek Iliad and Sanskrit Rig Veda, is remarkably warlike. So it’s tempting to imagine the arrival of the Yamnaya people in Europe as an armed invasion, with bands of mounted warriors looting their way across the continent, subduing everything before them with their long-shadowed spears and cutting swords. Certainly, that’s how Gimbutas saw it: to her, the arrival of the kurgan people spelled the death of ‘Old Europe’, a lost civilisation of peaceful, egalitarian, mother-goddess worshipping farmers replaced by the destructive, death-and-property obsessed patriarchy that is, in some ways, still with us today.
But the migration of the Yamnaya people could just as well have been peaceful. Anthony proposes the phenomenon of chain-migration as a rival explanation for the appearance of the PIE-speakers in Europe. In chain-migration, the movement of people is a gradual process of accretion. First, a few pioneers establish a foothold in a new territory. Imagine a Yamnaya horse-breeder finding employment with a local chief, or a herder finding good pastures, otherwise unused. Once they are settled, they send home the news of their good fortune. Before long, the region is full of their relatives and friends. In this version, the spread of the Indo-Europeans has a lot more in common with the movement of immigrants through Ellis Island than with Conan the Barbarian.
The two genetic studies didn’t detect signs of just one major migration into Europe. They also revealed an unsuspected mobility in Bronze Age populations. Central Asia appears to have been especially volatile. As the evolutionary biologist Eske Willerslev – one of the leads of the Copenhagen study – told me, it is genetically ‘the most dynamic place’ he’s ever seen, with ‘four distinct populations’ replacing one another in turn. An initial group of Eurasian hunter-gatherers were succeeded by people from the Caucasus, who were in turn replaced by Northern Europeans and then by East Asians.
Europe’s population is made up of multiple waves of migration. A similar process seems to have been at work in the Americas
Along with earlier work on the genetics of Neolithic Europeans, the two studies strongly indicate that agriculture spread into Europe by way of farmers from the Near East. The earliest farmers in Sweden seem to have been more closely related to Cypriots and Greeks than to modern-day Scandinavians. The arrival of the Near Eastern famers left a deep mark on the continent’s genetic fingerprint. In Sardinia and Sicily, the descendants of these first European farmers make up the bulk of the population. They have also been found to have a close affinity with Ötzi, the Ice Man – the natural mummy found by hikers in the Tyrolean Alps – who appears to have been a member of this first farming population.
It’s now increasingly clear that Europe’s present population is made up of the legacy of multiple waves of migration. A similar process seems to have been at work in the Americas. The first ancient body to have its entire genome sequenced belonged to a mummy from Greenland called the Saqqaq Man, who lived and died on the shores of western Greenland some 4,000 years ago. He belonged to an ancient community known as the Dorset People, who were named for the characteristic tools found at archaeological sites across the North American Arctic. When a team of Danish scientists led by Willerslev published the Saqqaq Man’s genome in 2010, it led to some startling conclusions about the history of the Arctic.
The Saqqaq Man’s DNA revealed that he wasn’t closely related either to modern-day Inuit or to Native Americans, ancient or modern. Rather, his closest living relatives seem to live in eastern Siberia. This indicates that the Bering Strait was crossed on at least three separate occasions. From the archaeological record it appears that the Dorset People colonised the Arctic sometime around 6,000 years ago. They lived there in near-total isolation (at least from their Native American neighbours to the south) for nearly 5,000 years, before vanishing. They were replaced by Thule Culture, ancestors of today’s Inuit, who expanded with lightning speed from the tip of Siberia to Greenland. As in Europe, migration seems to have happened with some frequency in the ancient world, well before the era of guns and steel.
Ancient DNA tells a similar story about the peopling of the Americas as a whole. Around 12,000 years ago, the same technology began to appear all over North America. Named the Clovis Culture – after the site in New Mexico where it was first discovered – it is characterised by especially large and beautiful spear points. The Clovis points were used to hunt mammoths and possibly other animals roaming the post-Pleistocene landscape. They were in use for a few hundred years, and then abruptly vanished, along with the mammoths. This suggested that the Clovis People were the first to arrive in the Americas. Coming down through an ice-free corridor in the Laurentide ice sheet (which covered most of what is now Canada and the northern US), they stumbled onto an empty continent, full of game. They dispersed quickly, pursuing megafauna as prey. When the megafauna became extinct, the culture vanished too, fragmenting into the hundreds of successor groups that eventually became today’s Native Americans.
But the very first Americans can’t have been the Clovis people. In the past two decades, a number of sites across North and South America have been securely dated to at least 1,000 years before the appearance of the Clovis Culture. (One of these sites – Paisley Caves in Oregon – was identified in part with the help of DNA-testing of ancient faecal samples.) The geographic distribution and material remains found in these pre-Clovis sites have raised new questions about the exact path by which people first came into the New World. Indeed, one of the oldest and best attested pre-Clovis sites – Monteverde in Southern Chile – is about as far from the Bering Land Bridge as you can get and still be in the Americas.
Some of the earliest skeletal remains found in the Americas have added to this mystery. The most notorious example is the Kennewick Man, an 9,000 year-old Paleo–Indian skeleton found by accident in the Columbia River in 1996. His skeleton has been the subject of a long-standing legal dispute between the federal government, scientists and five Columbia Plateau Indian tribes who claim him as one of their own. Scientists who examined him said that his skull morphology was clearly unlike that of modern Native Americans, and more consistent with Ainu or Polynesian descent.
Based on this evidence, two main theories now hold sway about the timing and origin of the first Americans. One theory is similar to the older Clovis-first model. The earliest Paleo-Indians arrived in America by crossing the Bering Land Bridge from Siberia. They spread across North and South America, then lived quietly (from an archaeological perspective) for 1,000 or several thousand years before abruptly giving birth to the Clovis Culture.
Another theory though suggests that the first Americans didn’t come overland, but by boat. They followed the Ring of Fire across the Northern Pacific, jumping from island to island, from the Kurils to the Aleutians, and then down the Pacific Coast all the way to Patagonia. This would account for some of the odd skull morphology seen in Kennewick and other bodies, as well as the early arrival on the west coast of South America. And seafaring on this scale so far back in the past no longer seems impossible. We now know, on the basis of archaeological and genetic studies, that Aboriginal Australians crossed the Torres Strait more than 40,000 years ago. If they could make their way across open water, why couldn’t Paleo-Indians do the same 30,000 years later?
The Anzick Boys’s genome shows that he was related to almost all living Native Americans
The Pacific Migration Hypothesis has a compelling logic behind it. However, so far it has not been borne out by the genetic evidence. The genome of the Kennewick Man, published this June, showed him to be unmistakably Native American, with no trace of Ainu, Japanese or Pacific Islander ancestry. This result helps put to rest a particularly contentious debate that’s been tinged with an unfortunate degree of racial fantasising, but for the history of the settlement of the Americas, DNA from another body might be even more important.
The Anzick Boy is our only known Clovis burial. A two-year-old child, he was buried 12,000 years ago in western Montana, surrounded by a clutch of artefacts of clear Clovis provenance. Anzick’s genome, published last year, shows that he was related to almost all living Native Americans. Oddly, it showed that he was more closely related to Native peoples in South America than to those in the North. Nearly 100 per cent of Native South Americans are his descendants, as opposed to some 80 per cent in the North. This suggests that, while Clovis people are ancestors to most Native Americans, at some point soon after their arrival on the continent, a population bottleneck cut off gene flow between North and South. Southern Americans preserved a direct line of descent to the first Paleo-Indians. In the North, the connection was slowly diluted by the arrival of later migrations, such as the one that brought Inuk’s ancestors to the Arctic.
The genes from ancient bodies make the peopling of the Americas seem like a simple story, involving a single, fairly homogenous founding population arriving all at once across the Beringian Ice, then dispersing rapidly over a virgin continent. The story told by modern genes, however, appears to be more complex. In July, researchers at Harvard and the University of Copenhagen announced that they had independently found signs of a faint, but definite kinship between Native Amazonians and Aboriginal Australians and Papua New Guineans.
Why would genes from Southeast Asia turn up in the Brazilian rainforest, but not in the rest of the Americas? The two teams who made the discovery offer different explanations. The Danish team suggests that Australasian DNA travelled to the Americas in the somewhat recent past, after the initial peopling of the Americas, but before European contact. According to them, the genes could have hopscotched across the Aleutians and then across North and Central America through chains of interbreeding individuals. By contrast, the Harvard team proposes that it was brought to the Americas by a ‘phantom’ population, nicknamed Y, who were another founding member of the group of pioneers who initially settled the Americas. If this is true, the people who first crossed the Bering Strait start to seem unexpectedly diverse and cosmopolitan.
Ancient genomics is a powerful tool for the study of prehistory, but it is still in its infancy. The first true population studies using ancient nuclear DNA – with samples numbering in the tens instead of single digits – are only a month old. For the moment, we have just two ancient genomes from the Americas. For other parts of the world, such as Africa, South and East Asia, we have one or none. With so few data points available, the world of prehistory seen through the lens of ancient DNA is like a landscape sporadically illuminated by lighting. Plenty of surprises are left in store. The situation right now is a bit like that of archaeology just after the invention of Carbon-14 dating. A revolution is on its way, but we don’t yet know what it will bring.