The Genomic Ancient DNA Revolution

The Genomic Ancient DNA Revolution

A New Way to Investigate the Past
David Reich [2.1.16]


My experience collaborating with Svante since 2007, has been that the data we’ve looked at from the incredible samples they have has yielded surprise after surprise. Nobody had ever gotten to look at data like this before. First, there were the Neanderthals, and then there was this pinky bone from Southern Siberia. At the end of the Neanderthal project, Svante told me we have this amazing genome-wide data from another archaic human, from a little pinky bone of a little girl from a Southern Siberian cave, and asked if I'd like to get involved in analyzing it.

When we analyzed it, it was an incredible surprise: This individual was not a Neanderthal. They were in fact much more distantly related to a Neanderthal than any two humans are today from each other, and it was not a modern human. It was some very distant cousin of a Neanderthal that was living in Siberia in Central Asia at the time that this girl lived.

When we analyzed the genome of this little girl, we saw that she was related to people in New Guinea and Australia. A person related to her had contributed about 5 percent of the genomes to people in New Guinea and Australia and related people—an interbreeding event nobody had known about before. It was completely unexpected. It wasn’t in anybody’s philosophy or anybody’s prediction. It was a new event that was driven by the data and not by people’s presuppositions or previous ideas.

This is what ancient DNA does for us. When you look at the data, it doesn’t always just play into one person’s theory or the other; it doesn’t just play into the Indo-European steppe hypothesis or the Anatolian hypothesis. Sometimes it raises something completely new, like the Denisovan finger bone and the interbreeding of a gene flow from Denisovans into Australians and New Guineans. 

DAVID REICH is a geneticist and professor in the Department of Genetics at the Harvard Medical School. David Reich's Edge Bio Page

THE REALITY CLUB


ROBERT TRIVERS
​Evolutionary Biologist; Professor of Anthropology and Biological Sciences, Rutgers University; Author, Wild Life: Adventures of an Evolutionary Biologist

Absolutely superb ancient genomics Edge piece: Reich is riveting. I know most of the work—I teach a course on Human Evolutionary Genetics—but I knew nothing about the degree of Indian genetic admixture, for example, most fascinating as it interacts with wide-scale cousin marriages, themselves associated with such horrors as “honor killings.” As such, the very large sample sizes that genetics easily generates can then be analyzed for ancient components in ways that are often astonishing.

How else would we know that Neanderthal gene admixture of at least 40,000 years ago (~3% of all genes in modern Euro-Asian populations) is associated with 1) light skin tone, 2) blue eyes, 3) straight hair, and 4) novel parasite-resistant genes to the level possibly of allergies?! God bless archaeology and paleo-anthropology, but genomics can now tell us more about what was going on in Yorobu-land 8500 years ago and Europe 10000 years ago than can these fields. Indeed, genomics can give us a far broader picture as well, showing recently that there was modest human genetic admixture into Neanderthals some 100,000 years and some slight movement of modern human genes back into northern Africa some 4000 years ago. Of course as Reich makes clear, integration of genetics with linguistics and other fields is most helpful but genomics will continue to get much more powerful quickly, dominating other forms of evidence, whenever there is conflict.



THE GENOMIC AND ANCIENT DNA REVOLUTION

I’m a population geneticist. I study patterns of genetic variation and differences amongst human populations around the world. My big interest is in how people got to be the way they are today and how the differences amongst human populations came to be the way they are today. I was a student of David Goldstein who was, in turn, a student of Luca Cavalli-Sforza. 

Luca Cavalli-Sforza is the grandfather of the field of studying human evolution and human population history. He realized, as did Allan Wilson, that you can take differences amongst the genomes of different people in the world and use those differences to learn which people are most closely related to which other people, which populations are most closely related to which other populations.

At the time I did my PhD, which was in the late '90s, the amount of data that was available was about 100,000 times less than is the case today, and that was the type of data that Luca Cavalli-Sforza had available. The ability to learn about history was very thin at the time, but already, he and others could see that people had come out of Africa from a single founder population around 100,000 or 50,000 years ago, and that people moving out of Africa dispersed around the world. But the details were impossible to discern from the data that was available at the time.

Luca is someone who is deeply immersed in archeology, history, and linguistics and was reading onto those rich fields the little additional information that genetics could provide. What’s happened since that time—the last fifteen years—is genetics has become an extremely powerful type of information. It’s as important in terms of the information it can convey about prehistory as archeology and linguistics, and it’s emerging as a third way to make inquiries about the past before writing. That’s what I am interested in learning about—trying to take advantage of that massive new source of information to understand how people are related to each other.

My career began as a Luca Cavalli-Sforza student, who was my academic grandparent, working initially on some data that was available and generated by him and his colleagues, and then taking advantage of the genomic revolution and using that information to learn about history and, in particular, with ancient DNA in the last five to seven years.

There is an interesting phenomenon to do with this new type of information, this new type of data suddenly coming into the room where it wasn’t available before. Previously, the understanding of what population transformations were like, what the peopling of certain parts of the world was like, what lifestyles were like, was the province solely of archeologists and cultural anthropologists, or biological anthropologists and paleontologists. Now, this new information is coming into the room and is speaking to many of the same questions. In part, there is an interesting friction related to the new type of information that is coming into a field which previously hadn’t had access to that information. It’s a little bit threatening to people who are already in those fields—that there is a new type of information.

Another thing that’s going on is that population geneticists, such as myself, are a bit unschooled. We haven’t gone through graduate school in anthropology, or linguistics, or history, and yet, we’re making very strong statements about these people’s fields. It's a little bit like barbarians are walking into your room, and you can’t ignore barbarians because they have information, weapons, and technology that you didn’t have access to before.

A concrete example of this is work on population history in India that I have been involved in and that I continue to be very intensively involved in. The history of Indian populations is very rich; it’s one of the most incredibly diverse places in the world, with all these ethnic groups.

There are more than 4000 well-defined ethnic groups that practice mutual endogamy and don’t mix with each other in practice. It’s a very complex place. People look very different, they have extremely different cultures and histories and traditions. There’s a great deal of anthropology and anthropological study that has gone on in India to try to understand the population history and the context of the relationships to places outside of India.

In the beginning of 2007, we started studying at the whole genome level, the whole organism level, the DNA from initially twenty-five diverse Indian populations. It’s now more than 200 that we’ve studied. We picked these populations to be as diverse as possible, capturing the linguistic diversity of India. In the south of India, people speak languages called Dravidian, which are not related to languages outside the Indian subcontinent. In the north, people speak Indo-European languages, which are related to the languages of Europe and Armenia and Iran. There are some other language groups, but we picked people to represent the diversity of these language groups, diversity of social status as encoded in the caste system, and we studied the genetic variation.

What we saw was an amazingly simple pattern: The simple pattern was that the great majority of Indian groups today are descended from a mixture of basically just two ancestral populations, one which we call the ancient ancestral North Indian and one which we call the ancestral South Indian. Everybody is mixed in India without exception. Even the most isolated groups, which are hunter-gatherers living in the forest or isolated places, everybody is mixed with at least 20 percent of each of these ancestries.

This is a surprise that comes from the genetics. There is no pure unmixed ancestral population of Indians. People who are Dravidian, who come from the south of India, tend to have more of the ancient South Indian ancestry. The people from the north, the people who speak Indo-European, tend to have more of a North Indian ancestry. But there is variability in proportion: people who have traditionally higher caste status both within Southern and within Northern India tend to have more of the ancient Northern Indian ancestry. So what was this reflecting?

The other thing you can see in the data is that there is an intense sex bias to the event. If you look at people’s proportions of ancient North Indian ancestry, which ranges from about 20 percent to 80 percent depending on which group you’re in, the ancient North Indian ancestry is coming primarily from your paternal side. Most of the ancestry that you get that is North Indian, which is related to West Eurasians, related to Middle Easterners and Central Asians and Europeans even, most of that ancestry is coming from your male side. Most of your ancient South Indian ancestry is coming from your female side. That type of ancestry is not related to anything outside the Indian subcontinent; it is completely local.

What this reflects is an amazingly profound and convulsive historical mixture event. This was very interesting because what it implied was some historical phenomenon that was not clearly documented by archeology or anthropology, about which there were debates, but there was no question that this had occurred.

A lot of our research on Neanderthal population history, which I have worked on very closely with Svante Pääbo, was driven in terms of the methods we developed by this India question and trying to understand what happened in population history in India. When we followed this up, our big question was, when had this mixture occurred between people of this ancient North Indian and ancient South Indian ancestry? because it was a profound and convulsive mixture that affected every population in India without exception.

What we developed is a set of statistical techniques that used a very simple principle to try to estimate the age when the mixture occurred and the ancestry of the present day person. The way this idea works is very simple: When you are a mixture of two people of different ethnicities, for example, you have two copies of each chromosome, you have 23 pairs of chromosomes, and a first generation offspring of a person, say, of ancient South Indian and ancient North Indian ancestry will have one chromosome of three entirely ANI ancestry (ancient North Indian), and the other chromosome three of entirely ASI (ancient South Indian ancestry).

When they perform a mixed offspring, a kid, they’ll break those chromosomes once or two times per chromosome per generation and send a hybrid, a mosaic chromosome, down to their kid, which will have perhaps the first third ANI, next third ASI, next third ANI. So you have a regular breaking one or two times per generation until today, you have a broken-up chromosome where the number of breaks reflects how many generations it’s been. You just count the breaks to see how long ago the mixture happened.      

When we did that we found that the mixture in India happened between 2000 to 4000 years ago. Prior to that, there were unmixed populations in India. Amazingly, what you could see from the genetic data was an event in which you could see cultural change. You could see that in India, prior to 4000 years ago, the populations there looked nothing like they do today. There were unmixed populations related to West Eurasians, and they were related to ancient South Indians.

After that, there was a profound and convulsive population mixture event that affected every group without exception, even the ones that are isolated and outcasts outside the tribal system. Then, beginning about 2000 years ago, the whole system locked in and the mixing stopped, and you could see that because there are these founder events—a relatively small number of people giving rise to the large number of people that exist in any one group today. There is very strong endogamy in India, so you can see that a lot of people in a group today will descend from the same founders. And if there was even a little bit of genetic input into those groups from outside over that 100 generations, it would be disrupted, that signal.

What we can see is a cultural change where there is initially massive acceptance of mixture or massive occurrence of mixture—it’s a sex-biased event, which probably means it has to do with power—and then it locks in and the caste system sets in. This is very interesting in the context of the anthropology because one of the leading views in biological anthropology was a kind of revisionist view about the caste system in India.

The caste system was an invention of colonialism. Nicholas Dirks, for example, was arguing that the caste system wasn’t very important prior to the British. The British found this system in India, they strengthened it as a way of ruling India, they put themselves on top, they used it as a way of organizing society, and they systematized it; they imposed it across different parts of India.

To many extents, that is true, they did use the caste system, they did systematize it. But it’s wrong to say that it was not strong, and you can see in the genetic data that it’s been around for thousands of years. You could see this in the genetic data.      

There is an incredibly old longstanding debate in India about the events that led to the amazing mixes that have occurred in India. India is a mixture of Indo-European languages related to European languages, and also of Dravidian languages not related to languages outside of Europe, as well as other language groups. It’s a mix of agricultural systems. It’s at the collision of the Chinese and Mesopotamian agricultural systems with rice and certain other domesticates coming from China, and wheat and barley and goats and sheep coming from the West. It’s a mix in all sorts of ways, and it’s a mix genetically.

The question is how this came about, so the linguistics is an important line of evidence here, and it’s a very contentious and interesting line of evidence. The question is still not resolved and that’s what genetics and what I am most engaged in trying to understand right now.

There is this phenomenon of the Indo-European languages. These are these languages like English, like almost all the languages of Europe except for Basque, Hungarian, Estonian, Finnish, and a couple of others, which are spoken across this extremely broad region of the world even before the movement of these languages to the Americas and elsewhere through colonialism. These languages have a funny distribution: They're spoken in Europe and they’re spoken in Iran and India and Armenia, but they haven’t been spoken in the Middle East and the Near East for 5000 years, mostly. The reason we know that is Near Eastern languages—the Near East is the place where writing was invented and that we know these languages weren’t Indo-European. You have this strange bi-lobed distribution with a lobe in Iran and India, and a lobe in Europe, and a gap in-between. How did this come about? This is one of the great mysteries of the West.

It was discovered in the late 18th century by a British magistrate working in India who was schooled in Greek and Latin from his public school education in Britain and realized that the Sanskrit he was studying was just like Greek and Latin in its grammar and its structure—these were closely related languages. This mystery of how Indo-Europeans spread over such a vast region and what the historical underpinnings of it would have been is ongoing and remains a mystery. The fact that these languages are in India has led to the hypothesis that they came in from somewhere else, from the north, from the west, and that perhaps maybe this would be a vector for the movement of these people.

Another reason that people think that is that when you have languages coming in, not always but usually, they're brought by large movements of people. Hungarian is an exception. The Hungarians are mostly not descended from the people who brought Hungarian to Hungary. In general, languages typically tend to follow large movements of people.

On the other hand, once agriculture is established, as it has been for 5000 to 8000 years in India, it’s very hard for a group to make a dent on it. The British didn’t make any demographic dent on India even though they politically ruled it for a couple of hundred years.

It’s a mystery how this occurred, and it remains a mystery. What we know is that the likely timing of this event is probably around 3000 to 4000 years ago. The timing of the arrival of Indo-European language corresponds to the timing of the mixture event.

This debate is happening now in a substantial way, and the reason is DNA. Linguistics is making amazing and interesting continued progress, but the DNA evidence is very radical and important. Colin Renfrew had a very important argument, and until this year, it was the primary working hypothesis for most people of the origin of Indo-European languages. He called it the Anatolian hypothesis.

The idea of the Anatolian hypothesis is that the Indo-European language is the language spoken by the people who invented farming in Anatolia.

We know from the archeology that farming expands radically and rapidly into Europe beginning about 8500 years ago, first in Greece and then radiating out from Greece until it reaches Scandinavia and Britain about 6000 years ago. Between 8500 years ago and 6000 years ago, it spreads right across Europe in a wave of events.

What he thought was a plausible explanation is that the movement of Indo-European languages into Europe was due to the expansion of farming; it came hand-in-hand with farming. Similarly, the movement of Indo-European languages into Iran and India was due to the similar expansion of farming from Anatolia and the Near East to the East. That was the hypothesis.

The linguistics tended to lean the other way—to support that there was more of a connection of these languages to the northern steppe. But the evidence based on archeology was winning the day.

The best evidence Colin Renfrew and colleagues had was that people typically change languages due to force of numbers, due to large numbers of people moving. What he argued was that we know there is a mass movement of people into Europe beginning 8500 years ago; that was confirmed by genetics beginning in 2009. But we don’t know of any major movement of people into Europe since the arrival of agriculture. And not only that, it would have been very difficult for there to be a major demographic impact on Europe after that point.     

~ ~ ~

Coming into this field for me has been made possible by my collaboration with Svante Pääbo, who brought me on as one of the primary analysts looking at population history, interpreting the Neanderthal data that they produced beginning in 2007 to try to learn about history and the relationships of Neanderthals to modern humans.

What Svante invented was a technology for looking at the whole genetic code, basically, of ancient humans. Once we had that data, we could compare it to present-day humans and other ancient samples to learn about history.

This technology is like the invention of new scientific instruments. It’s a momentous new scientific instrument, like the invention of a telescope or a microscope. When you have a new instrument, anything you look at is new. For example, when you look with a microscope into a bit of pond water, you see cells for the first time; that’s what people saw. Leeuwenhoek, when he first used microscopes, observed cells, observed microbes, observed cell walls, all these things one couldn’t see before were great surprises. When you have a new instrument, you can see new things.

When I began working with Svante my experience, when we had whole genome data from a Neanderthal which is an archaic human—the ones we were studying initially lived 40,000 years ago in Croatia and Europe—was that those samples were more closely related to non-Africans than to Africans. There were no ifs, ands, or buts about it; it was very clear genetically and there were multiple ways that we could see that this had occurred.

Since that time, line of evidence after line of evidence, way of looking after the data, way of looking after the data has shown that non-Africans today are descended from a mixture of Neanderthals like the ones that we have data on from Croatia, and modern humans. Somewhere between 1 to 3 percent of the ancestry of non-Africans today is from Neanderthals. That was a big surprise in that it was against the orthodoxy, it surprised us because Svante and I were both from a world in genetics which was very much the "out of Africa" world.

We were part of the triumphant march of people saying that it wasn’t lots of different independent origins of modern humans around the world, which is what some people thought before. Allan Wilson, in Svante’s lab, and Luca Cavalli-Sforza were part of this triumphant march of geneticists saying we all have a common origin of about 50,000 years ago from an exit from Africa, and there are deeper roots of human variation in Africa; that’s what’s going on.      

Multiregionalism, the idea of multiple independent origins and parallel evolution with gene exchange in different places, basically, that view is wrong in the sense that the basic story is that most of our lineage comes from Africa—98 percent of it in most cases for non-Africans—around 50,000 or 100,000 years ago, somewhere in that range. That’s basically the story. This is the case where DNA just cuts through, and in an extremely powerful way, provides new information.     

It’s a radically new type of information about the past; it’s a great gift to be able to have access to it. It’s a great surprise. Who would have thought DNA survived that long? It provides direct information about population relationships, and it allows you to connect the skeletons you get data from to archeological cultures and to different other ancient and modern people. It’s an amazing technology.      

The access we have to prehistory currently is through linguistics and through archeology, both important, exciting fields. Linguistics is the study of languages and their relationships and the reconstruction of ancestral languages, and it provides information about what languages were like prior to the invention of writing because bits of language are inherited through the word systems.

Similarly, archeology provides records of the material culture—the stone tools, and other artifacts people left behind, as well as some skeletons of people.

But understanding how people are related to each other has been nearly impossible based on archeological and linguistic evidence.

My experience collaborating with Svante since 2007, has been that the data we’ve looked at from the incredible samples they have has yielded surprise after surprise. Nobody had ever gotten to look at data like this before. First, there were the Neanderthals, and then there was this pinky bone from Southern Siberia. At the end of the Neanderthal project, Svante told me we have this amazing genome-wide data from another archaic human, from a little pinky bone of a little girl from a Southern Siberian cave, and asked if I'd like to get involved in analyzing it.

When we analyzed it, it was an incredible surprise: This individual was not a Neanderthal. They were in fact much more distantly related to a Neanderthal than any two humans are today from each other, and it was not a modern human. It was some very distant cousin of a Neanderthal that was living in Siberia in Central Asia at the time that this girl lived.

When we analyzed the genome of this little girl, we saw that she was related to people in New Guinea and Australia. A person related to her had contributed about 5 percent of the genomes to people in New Guinea and Australia and related people—an interbreeding event nobody had known about before. It was completely unexpected. It wasn’t in anybody’s philosophy or anybody’s prediction. It was a new event that was driven by the data and not by people’s presuppositions or previous ideas.

This is what ancient DNA does for us. When you look at the data, it doesn’t always just play into one person’s theory or the other; it doesn’t just play into the Indo-European steppe hypothesis or the Anatolian hypothesis. Sometimes it raises something completely new, like the Denisovan finger bone and the interbreeding of a gene flow from Denisovans into Australians and New Guineans.

Another example that we encountered along these lines was part of our work on India and part of our work on Neanderthals. We were developing tests of population mixture, tests of whether present-day groups today are the result of a mixture of other ancient groups. This is something that Luca Cavalli-Sforza wasn’t able to do because his data and the data that people like him were looking at was too thin at the time, but now we can do it.

When we developed these tests of population mixture we applied them to different people around the world, and in 2012 we applied them to Northern Europeans, for example, French, but it would also apply to English or Germans or Scandinavians or to many other populations in Northern and even Central and Southern Europe. We saw that everybody there is mixed. One of the mixing populations looks today most like Southern Europeans, like isolated people from Southern Europe like Sardinians, and the other population of all people is Native Americans. This was a huge surprise, completely unexpected—why Native Americans and Southern Europeans give rise to Northern Europeans. It was definitely Native Americans, not East Asians, not present-day Siberians; it was definitely Native Americans.

What we proposed in 2012 was that what we were seeing evidence of was an ancient mixture event that had affected Northern Europeans, and between early European farmers related to ice- related Southern European populations, and the group which we call the ancient North Eurasians.

The ancient North Eurasians were a population once distributed across parts of North Eurasia. Before 15,000 years ago they went into the Americas, in mixed form probably, but went into the Americas and became Native Americans, and also, at some point, went into Europe. You have this ancient North Eurasian population that once existed there but doesn’t exist today because it was replaced after the Ice Age.

What we had proposed was a ghost population, a population that we were predicting statistically based on the patterns that’s left on present-day people but that doesn’t exist anymore in the place where it once was.      

The Ice Ages are very profound events. There has been an ice age in Eurasia in the last 45,000 years, so there is a very severe cold period with glaciers covering much of Europe and much of North America with the maximum being between about 26,000 and 19,000 years ago. This was a profound event that made it impossible to live in many places in the world and radically changed the climate everywhere else, so people’s habitats were disrupted and people presumably moved around.

We can see that very clearly in the archeology; you don’t need the genetics for that. You see that there’s big transformations, whole periods when there are people missing from Northern Europe during this time. Then, there are re-peopling events following the retreat of the glaciers into North America, there is a re-peopling from the north after the retreats of the glaciers. And in Europe, there is a re-peopling from the south.

These are profound events and they are naturally expected to be accompanied by population changes and transformations in Central Asia. There are also dramatic population changes. The present-day people of Siberia are almost certainly post-Ice Age re-peopling from the south, replacing these ancient North Eurasians who migrated to the Americas and also to Europe. There are very dramatic changes. We are currently working on Ice Age Europe and genetics of Ice Age Europe, and we see extremely dramatic changes associated with the Ice Age and population replacements and changes and turnovers.      

It’s a new type of information. I’ve been at a series of interdisciplinary meetings between linguists and archeologists and geneticists recently, and the genetics give succor to some people but it doesn’t make everybody happy; nobody is completely happy with the genetic data. It doesn’t perfectly play into anyone’s theory. In general, there is this battle line that’s been drawn between people who support what’s called the Anatolian hypothesis, the arrival of Indo-European languages from the Near East with farming, and people who think it arose later through a homeland in the steppe.

The genetics is tending to support the steppe hypothesis because in the last year, we have identified a very strong pattern that this ancient North Eurasian ancestry that you see in Europe today, we now know when it arrived in Europe. It arrived 4500 years ago from the East from the steppe, and it now constitutes about half the ancestry of Northern Europeans.

What we showed in the genetic data is that around 4500 years ago, at least in Central Europe, there is a massive population replacement with new people coming in. The old people disappear or get marginalized, and the populations ever afterward have very large percentages of ancestry from the East and, in particular, they have ancestry that is closely related to a group called the Yamnaya, which is the first mobile population of the steppe.

They had to use the recent domestication of the horse and the recent invention of the wheel to be able to herd their cattle and horses in the steppe lands which were previously inaccessible to humans. These people spread all over the steppe, they spread into parts of Siberia, and they also spread west into Europe. What we’ve shown in the genetics data is they actually replaced, or their descendents replaced, much of the population.

We now have a massive population replacement in Central Europe beginning at least 4500 years ago that is so late that it must have brought new languages to the people, and therefore, it’s highly likely that at least some of the Indo-European languages in Europe owe their origin to migrations from the steppe. This is a very exciting development and it means that the Anatolian hypothesis cannot be an explanation for all the languages of Europe because some of them have to come through the steppe, and it also means that possibly there are ways that one needs to reconsider the possible spread processes in light of the genetic data.

~ ~ ~

I opened up my own ancient DNA laboratory in Boston in 2013 with the help of Svante Pääbo, to study some of the topics that he wasn’t interested in studying himself, mostly population transformations after the Ice Age, and that’s what the work of the last three years has been. It’s been turning ancient DNA into an industrial process, studying very large numbers of samples, moving away from the model of studying just one or two or three amazing interesting samples and studying dozens or hundreds of people and understanding how carefully constructed time transects through different places in the world, and how populations have changed over time.        

It’s important to be extremely careful and to say true things that are extremely well supported by genetic data. That cuts out 95 percent of what one might say. What we are trying to do is to be very cautious and careful about the things we say and that when we publish something, to say things that are very clear, strong, and correct, as much as we can.

Surely, we’ll make mistakes sometimes, but we try to not be speculative and to say very confident things. Once we do that, we are in a place where we can defend what we say, and we have an obligation to report what we find regardless of how it plays into a discussion. In general, the effect of the genetics and the genetic discoveries of the last few years has been to make it harder to make an argument about the superiority of one group or another.

One of the things that’s emerged from the genomic ancient DNA revolution from this new science of the human past is the realization that human populations today are mixed. This is a very unexpected thing and it was not in people’s way of thinking about the world before, but when people think intuitively about human population differences, they think, "Oh, there are these differences that I recognize intuitively amongst groups that I see in the world, and they must reflect group differences that go back deep in time to the time that we all share a common ancestral population." But in fact, that’s not true. If you look, for example, 10,000 years ago at the population structure of Eurasia, and you compare it to the population structure of Eurasia today, at that time, the populations were just as differentiated from each other as they are today, but the structure was nothing like what it is today.

For example, today, West Eurasia, Europe, the Near East, Central Asia, Iran is a region of very low differentiation, the populations genetically are quite similar to each other, but in fact, that reflects over the last 10,000 years a dramatic collapse of four very different populations into each other: hunter-gatherers of Europe, ancient Near Easterners both in the West of the Near East and the East of the Near East, and it’s people from the steppe.

None of these populations disappear, but they all mixed in with each other such that it’s a large region of low differentiation. There are these people who are ancient North Eurasians who used to be spread over Siberia; they don’t exist anymore. They don’t exist in unmixed form, but they left huge numbers of descendants in India and Europe and the Americas.

There were groups that were highly differentiated at that time, but they’re very different from the groups today. And what you see instead is a lattice of groups in the past which are quite differentiated from each other but form mixtures of other quite differentiated groups, going back and back and back and back in time. It’s mixture all the way down; it’s a very different picture. That’s difficult to reconcile with people’s intuitive sense of the differences amongst groups which are more of these pictures of static difference going back a long way.     

What’s happened very rapidly, dramatically, and powerfully, in the last few years has been the explosion of genome-wide studies of human history based on modern and ancient DNA, and that’s been enabled by the technology of genomics and the technology of ancient DNA. Basically, it’s a gold rush right now; it’s a new technology and that technology is being applied to everything we can apply it to, and there are many low-hanging fruits, many gold nuggets strewn on the ground that are being picked up very rapidly. That’s what’s happening now.      

What’s very clear and what the archeologists who are really scientists, embracing scientific technology, are realizing is that this is a new way to investigate the past. This is going to be embraced by archeology, by the humanities, by linguistics, as a new way of inquiring into the past. In five or ten years, this will be integrated properly into departments of linguistics, especially into departments of archeology, as a central way of inquiring into the past.

If you are excavating a site and you have skeletal remains of humans or of animals, the DNA will tell you how the people who inhabited the sites, or the animals who inhabited the sites, or the plants that were raised at the sites, relate to those at other sites that have been excavated. It will allow you to tell the sexes of the individuals, it will allow you to tell the family relationships of the individuals, it will allow you to understand the details of the population relationships of other groups. It’s a little bit like the radiocarbon revolution that happened beginning in 1949.

Walter Libby invented this technology for estimating directly the dates of samples by the ratio of carbon-14 to carbon-12. This was quickly recognized by archeologists to be a transformative discovery because it made it possible to obtain an absolute direct date on samples where previously it was only possible to obtain relative dates.

For example, Colin Renfrew’s career is built on the radiocarbon revolution; it’s built on the idea that you can get absolute dates on things. Colin Renfrew was very impressed and involved in work that demonstrated that the megalithic structures of Europe—these big stone structures of Europe—preceded the pyramids and the ancient large stone structures of the Near East. The previous archeologists had argued all ex-Orient looks, the best big ideas are always coming from the East.

The big structures of Europe must be derived from the East; they must have come afterward. But it’s not true. It was radiocarbon dated before, so it gave an absolute timescale for history. Archeologists have fully integrated dating into their work, and archeologists will also fully integrate ancient DNA into their work.

I have a lot of hope here because archeologists are scientists. Even though archeology is often embedded in the humanities, archeologists as part of their training, learn how to interpret radiocarbon dates, they learn how to interpret isotopic information, and they have embraced science.

They are desperate to learn about the past with scientific methods and other methods. It's not physics, so they don’t have equations very much, but they are desperate to learn about the past. It's very clear from talking about them that they are embracing this technology.

Now, how this relates to anthropology is a more complicated question because anthropology has a grounding in the humanities, and what we’re beginning to learn about, as in the case of our work on India is about past interactions of people. We can understand, for example, from the genetics, we can see that mixtures of events were mediated by sex bias.

I mentioned the one in India where most of the West Eurasian related ancestry in India is coming through male ancestors, but more recently, we see evidence of this in African-Americans. It’s 20 percent European ancestry, but it’s coming three-to-one from the male side, the European ancestry. If you look at Colombia in Mestizos, in people of mixed native and European ancestry and a bit of African ancestry, the European ancestry is coming twenty-to-one from the male side.

What you’re seeing in the imprint of these populations genetically is the history of power inequality, which is usually males of power from one group having preferential access to local females, and that’s what you see in these groups. You see this again in Iceland. Iceland today is a mixture of Scandinavian men and British women who were stolen from Britain by, as is documented in the sagas, by these tax evaders from Norway who were male, and you see this imprinted in the DNA. You can see sex bias in the genetic data.

Now, that’s very interesting because it tells you about something about the cultural nature of these interactions. In India, you can see, for example, that there is this profound population mixture event that happens between 2000 to 4000 years ago. It corresponds to the time of the composition of the Rigveda, the oldest Hindu religious text, one of the oldest pieces of literature in the world, which describes a mixed society where there is people with not entirely Indo-European names who are being incorporated as poets and kings.

You have that happening, and then you have the ossification of that system into a kind of caste system, which you can see both in the genetics also reflected in the text. But you can establish this with the genetics, especially with ancient DNA, but even in the case of India with present DNA because we don’t yet have ancient DNA from India. Anthropologists will eventually need to use this information, and they will because they’ve embraced also radiocarbon data.