Out of Eden: The Peopling of the World
Page 16
Asian men in Europe
The Y chromosome, being so much larger, should hold far more secrets of our past than does mitochondrial DNA, that tiny trace of our ancient microbial invader. But geneticists are only beginning to unravel it, and have yet to fully master the trick of Adam’s genetic clock. Dates of branches vary enormously from one research group to another, and are generally anything from twice to five times younger than the equivalent dates on the mtDNA tree. In other words, there is a good Y trail but researchers often underestimate how old it is. Luckily, the careful analysis of founder lines and mtDNA dating carried out by Martin Richards and his colleagues, and others such as Toomas Kivisild, provide a framework of genetically dated routes on which to fit the Y trail. Here for the moment we can use the Y tree to look much more closely at the geographical routes of spread.29
Stanford geneticist Peter Underhill’s extraordinary achievement at the end of 2000 of building a world tree with 150 Y markers was quickly followed by a more detailed look at Europe by Ornella Semino and colleagues in Underhill’s team.30 This showed even more clearly than the mtDNA story that Europe must have had several routes of colonization.
We have seen how, in parallel with the daughters of Europa, Jahangir, a grandson of the out-of-Asian founder line through Seth, may have spread up the Fertile Crescent from the Gulf, thus making the first southern entry through Turkey and into Europe. Can we see any other male lines – apart from Inos – to parallel HV, who came out of the Caucasus, or even further east in Russia, and travelled through the Ukraine to western Europe? The answer is yes, and with much clearer trails. I have already mentioned Inos as type who could mirror HV in the north Caucasus. There are three more candidate genetic lines for an eastern entry to Europe that tease geneticists with their distribution, age, and origin. Like Inos (and Jahangir), they are all ultimately descended from Seth, the third son of Out-of-Africa Adam. Two of these lines very clearly come from the East, probably Central Asia, but how recently did they arrive, and where did they ultimately come from?’ They are both descendants of the most prolific of Seth’s five genetic sons, ‘K, whom I shall call Krishna after the Indian deity of the same name.31 (See Figures 3.4 and 3.5)
One of these two eastern genetic lines is called TAT, after its signature mutation, and by remarkable coincidence echoes the name of the Oxford-based geneticist Tatiana Zergal, whose name is most associated with him. In Europe he is confined to the eastern fringe, where he is most evident among eastern Baltic Finns and Russians. His origin is most likely to be Central Asian, although his genetic father and grandfather derive ultimately from Kashmir and South Asia. He is found in low frequency in the Far East as well. Again the dates are vague because we are looking at Y chromosomes, but the very local distribution (mainly among Uralic speakers) and low diversity in Europe suggest a migration from Central Asia.32
The other European intruder from Krishna’s line is defined by the M17 mutation. The Hungarians achieve the highest frequency of this M17 line, at 60 per cent. The M17 line dominates Eastern Europe, and although he spread to all parts of western and southern Europe and the Levant, he is always found at low frequency in those places. His low frequency in the Levant tends to undermine that region as a route into Europe. Lluís Quintana-Murci, Peter Underhill and their colleagues see Central Asia, especially the Altai, as the most likely source of the European M17 line. This is a necessary argument for Underhill and co-workers since they see the whole Seth clan as coming separately out of North Africa into the Levant and proceeding directly to Central Asia.33 (See Figure 1.3.)
For me and for Toomas Kivisild, South Asia is logically the ultimate origin of M17 and his ancestors; and sure enough we find highest rates and greatest diversity of the M17 line in Pakistan, India, and eastern Iran, and low rates in the Caucasus. M17 is not only more diverse in South Asia than in Central Asia, but diversity characterizes its presence in isolated tribal groups in the south, thus undermining any theory of M17 as a marker of a ‘male Aryan invasion’ of India. One estimate for the age of this line in India is as much as 36,000 years while the European age is only 23,000. All this suggests that M17 could have found his way initially from India or Pakistan, through Kashmir, then via Central Asia and Russia, before finally coming to Europe. As an aside, these and similar observations for other Seth lines in South Asia form an important part of my counter-argument to the view of Seth entering Central Asia from North Africa and the Levant (rather than by my preferred single, southern out-of-Africa route).34
To return to Europe, while the origins of the European M17 line to the east of Europe in the Altai or beyond seem clear, the dates of that east–west migration are anything but. Semino and colleagues dispense with genetic dating and instead have both M17 and his father, M173, as part of an archaeologically dated Palaeolithic movement from the east to the west 30,000 years ago. Which brings us to M173, the other strong candidate male Asian line in Europe. He belongs to a very large clan that I shall call Ruslan, after a Russian folk hero.35
Ruslan: Asian progenitor of half of Europe’s men
Spread throughout the Old World from England to South and Central Asia, Siberia, and beyond – even to Australia and America, the Ruslan clan is well travelled, and his root type is the commonest single Y type in the world according to one study. Yet Ruslan’s genetic father, P, whom I shall call Polo after the family of Silk Road explorers, is confined to India, Pakistan, Central Asia, and America. Study of the geographical distribution and the diversity of genetic branches and stems again suggests that Ruslan, along with his son M17, arose early in South Asia, somewhere near India, and subsequently spread not only south-east to Australia but also north, directly to Central Asia, before splitting east and west into Europe and East Asia (see Chapters 5 and 6).36
Unlike his son M17, Ruslan does not lurk at the threshold in Eastern Europe. In fact he is very much commoner in the west, reaching 86 per cent in Basques and similar high rates in the British Isles. This back-to-front distribution of father and son suggests that the former may have arrived in Europe earlier. As we shall see in Chapter 6, this can be explained by the ice age. Suffice it to say here that although M17 may be a relative newcomer to Europe, his father Ruslan is possibly the strongest male marker line for the original Early Upper Palaeolithic invasion of western Europe from the east around 33,000 years ago. Between them, this father-and-son team account for 50 per cent of extant European male lines today.37
Europe’s Asian roots
This trip through genetic and human time has suggested two extraordinary conclusions: first, that the Europeans’ genetic homeland was originally in South Asia in the Pakistan/Gulf region over 50,000 years ago; and second, that the Europeans’ ancestors followed at least two widely separated routes to arrive, ultimately, in the same cold but rich garden. The earliest of these routes was the Fertile Crescent, which opened 51,000 years ago as a corridor from the Gulf, allowing movement up through Turkey and eventually to Bulgaria and Southern Europe. This seems to coincide with the Aurignacian cultural movement into Europe. The second early route from South Asia to Europe may have been up the Indus into Kashmir and on to Central Asia, where perhaps more than 40,000 years ago hunters first started bringing down game as large as mammoths. Some of these hunters with their elaborate technical skills may then have moved westward across the Urals to European Russia and on to the Czech Republic and Germany. A more conservative view of this eastern invasion might be that the Trans-Caucasus, rather than Central Asia, was the earliest route of modern human entry into Russia.
4
FIRST STEPS INTO ASIA, FIRST LEAP TO AUSTRALIA
IN CHAPTER 1 WE SAW EVIDENCE for a single out-of-Africa genetic line, L3, whose two daughter lines, Manju and Nasreen, jointly peopled the rest of the world. This single genetic line is central to the logic of a single southern exodus. As we saw in Chapter 3, the fact that, uniquely, Europeans are descended solely from the Nasreen clan fixes the origin of their branch to a colony early on the
route out of Africa, probably near the Gulf. The absence of any of the root daughter genetic branches of either Manju or Nasreen in North Africa or the Levant, and their abundance in India, excludes the northern route into Europe and confirms the southern route across the mouth of the Red Sea. We shall now look in more detail at the genetic and other evidence for the earliest primary colonization of South Asia and the coastline of the Indian Ocean beyond. The Y chromosome now enters the story more fully, along with some genetic markers other than the Adam and Eve genes. South Asia includes all the countries initially found along that pioneering beachcombing trail round the northern shores of the Indian Ocean.
Survivors of the great trek: place and time
If all non-Africans share one ancestral origin, the date of exit for the ancestors of aboriginal Australians is the same as for Europeans, Indians, and Chinese. All their trails should lead back to one point in space and time; and all the colonies, left behind en route, should hold genetic and even physical keys to who went that way. This is the case. We can look at the mtDNA or Y chromosomes from anyone outside Africa, and find them a place on their respective single branch of the out-of-Africa genetic tree. Exactly where they are on their out-of-Africa branch very often tells us how they got there and even, sometimes, when.
We should not be tempted into trying to draw firm conclusions from genetic dates in isolation. The exact branching structure of the tree and the geographical distribution of the branches are often much more revealing than are dates for them derived from the molecular clock.1 Those dates are rather approximate and should always be weighed against of the rest of the evidence, namely the climatic record, which tells us when there were windows of opportunity for or constraints against migration, and of course the archaeological record, where present.
The archaeological beach trail is inevitably very faint. Between 60,000 and 85,000 years ago the sea level fluctuated between 40 and 100 metres (130 and 330 feet) below its present level, so the beaches and even the hinterlands our ancestors wandered along and frequented are now mostly deeply submerged. But there are exceptions to the effects of this watery blanket, as we shall see.
Along the coastline of the Indian Ocean we still find small colonies of so-called aboriginal peoples who may be descended locally from those first beachcombers. Long before the study of mtDNA, the first reasons for calling these groups aboriginals were that their cultures and appearance marked them out from the people surrounding them: some had features in common with Africans, such as frizzy hair and very dark skin. I shall discuss the more objective genetic and physical evidence to support these impressions here and in the next chapter, but it is worth first naming some of the groups in question.
Such peoples are often called by controversial and presumptive terms such as ‘Australoid’, ‘Negroid’, and ‘Negrito’ to indicate how they differ from surrounding peoples. Starting from the South Arabian coast in the west there are the Hadramaut, who have been described as Australoid but almost certainly contain an element of much more recent African admixture. On the coast of Pakistan around the mouth of the Indus are found the Makrani Negroid ethnic groups. Again, there is genetic and historical evidence to support a major recent African admixture as a result of the slave trade. This may not be the case for other so-called Negroid types, such as are found in India, including the Kadar and Paniyan. There are also a number of other South Asian aboriginal groups including the so-called Proto-Australoid ethnic groups, such as the Korava, Yanadi, Irula, Gadaba, and Chenchu of India, and the Veddas of Sri Lanka. Recent work on two of these groups, the Chenchus and Koyas, has strongly suggested not only that their ancestral mtDNA and Y chromosomes were uniquely shared with other South and West Asians, but that they are characteristic of the earliest genetic heritage of the region. Their beach-settling ancestors from Africa would have provided the genetic seeds for the subsequent differentiation of the distinctive East and West Eurasian gene pools, and they would have received only limited gene flow from other regions since then.2
The Andaman Islands, situated between India and Southeast Asia in the Bay of Bengal, include several groups with preliterate traditional cultures and a very distinctive, so-called African appearance. The groups least affected by outside intrusions are the Jarawa and the Onge, who live in the most southerly Andamans. Recent genetic studies still being analysed may help to elucidate their maternal and paternal origins. On the maternal side, the Onge and Jarawa feature two distinct mitochondrial groups which nevertheless both place them in the Manju super-clan, thus confirming fellowship of the single out-of-Africa migration. These genetic lines show connections to the base of two ancient and unique Indian Manju clans, M2 and M4. M2 is the oldest and most diverse Indian Manju group and is also the commonest mtDNA component among the Indian aboriginal groups mentioned above. On the paternal side, the Onge and Jarawa possess only the Abel clan (the rarest of the three out-of-Africa Y lines, and known as Group D or Asian YAP). This is all consistent with the view that the Onge and Jarawa have remained as isolated groups since the early beachcombing sweep round the Indian Ocean. The other aboriginal groups in these islands, known as the Greater Andamanese, are slightly different culturally and physically. Genetically, while they share branches of the same two unique Manju genetic lines, significantly their Y chromosomes all come from another of the three sons of Out-of-Africa Adam, Seth (see Figures 4.2–4.4).3
Farther along the trail are the so-called Negrito peoples of the Malay Peninsula, known generally as the Semang (see Plate 16), and perhaps the best known of the candidate remnants of the old beachcombers. Another relict group possibly left over from the beachcombers in Indo-China and the Malay Peninsula are the so-called Aboriginal Malays, who are physically intermediate between the Semang and Mongoloid populations (see Chapter 5). Crucially, they hold clues to the earliest branches of an mtDNA line now characteristic of Mongoloid populations, known as the F clan. This founder East Asian maternal pre-F line in Malay aboriginals is also shared with groups across the water in the Greater Andamans, discussed above, as well as with people in the Nicobar Islands a little farther south.4
I shall have more to say, in this and the next chapter, about the various peoples and colonies thought to be left over from that first great trek along the coastline of the Indian Ocean. As always, what is most important to bear in mind is that the genetic tree tells us this was not an earlier out-of-Africa venture, as has been thought by some archaeologists and palaeontologists. It was the vanguard of colonization of the entire Old World.
Curiously, some of the best, if not the only archaeological evidence for dating the beachcombers’ trek along the coast of the Indian Ocean comes not from India, South Arabia, or Africa, but from the later parts of the trail – the Malay Peninsula, New Guinea, and Australia. So we shall take a trip there first to look at those who travelled farthest round the Indian Ocean, the New Guineans and the Australians.
The first Australians
Perhaps more academic heat has been generated over which humans first got to Australia, and when, than over any other question on the archaeology of the region. The ‘which?’ question, a rallying call for multiregionalists with their repetitive arguments about different regional skull shapes derived from earlier humans (see Chapter 1), is looking less and less relevant in the face of increasing genetic evidence that all modern Australians and New Guineans belong to either the Nasreen or the Manju clan.5 In other words, they are not a special case, nor are they an earlier type of human or modern humans mixed with local Homo erectus. The mounting evidence has not stopped rearguard, Spielberg-like attempts by multiregionalists to use extremely ancient DNA to resuscitate their case. However, the evidence now available comes down on the side of antipodeans belonging to the same single out-of-Africa L3 migration as everyone else (for a discussion of recent developments on this thorny issue, see note 6). The ‘when?’ question is now coming much more to the fore. Historians like to peg events on dates rather than be satisfied with a simple chronol
ogical order, so I shall deal with dates first, although they will probably be the last issue to be finally resolved.
The earliest generally accepted archaeological evidence of modern human colonization outside Africa has, until recently, been Australian – but this is a rapidly changing field. Until the 1990s, there was no clear evidence for humans in Australia or New Guinea before 40,000 years ago. It now seems that, as with datings for Europe, this was largely due to the limitations of the radiocarbon method of dating. But then new methods of dating began to be applied. One approach, the so-called luminescence dating of silica, enabled researchers to probe beyond the radiocarbon limit of 40,000 years. In 1990, the Australian geologist-cum-archaeologist team of Richard Roberts and Rhys Jones reported dates between 50,000 and 60,000 years ago for the first occupation of a rock shelter (see Figure 4.1) on the coast of Arnhem Land, in northern Australia. Arnhem Land is directly opposite Timor, the nearest island of the Indonesian archipelago and therefore the most likely casting-off point for the first Australians.7
Then Australian ‘dates’ took a rather more dramatic turn and, as it were, came off the wall. In 1996, archaeologist Robert Fullagar examined the rock art site of Jinmium. The wall of this shelter is covered in artefactual dents known as pecked cupules, and Fullagar reported that a fallen fragment of engraved sandstone, buried in sediments, had been dated in two independent estimates to 50,000 and 75,300 years ago. He even recovered stone artefacts from levels dated by thermoluminescence to between 116,000 and 176,000 years ago, which would have made it the oldest known human occupation site on the continent, and two to three times as old as the Arnhem Land shelters mentioned above. These new dates caused a furore. But the problem appeared to be solved when it was shown that contamination by sand grains from fallen rubble had produced the wildly old dates for the earth surrounding the artefacts. In other words, the dates were wrong, and analysis of individual grains of sand helped to detect and overcome inaccuracies caused by contamination. The oldest dates of human occupation in Australia have therefore remained around 60,000 years.8