by Martin Jones
Of more than 1,000 modern Polynesian islanders surveyed, the great majority of them fitted in with the idea of a fairly tight genetic bottleneck letting through a narrow suite of mitochondrial haplotypes. However, a couple of individuals stood out. One individual from Tahiti and another from the Cook Islands had sequences that were unusual for Polynesia. The best matches were to be found in Chile and among the Mapuche Indians from Patagonia. Perhaps there was an ancient journey following the Kon-Tiki path, even if it was not the pioneer journey. Perhaps more women were braving the ocean waves than at first thought. They would need to be women to carry the mitochondrial lineages. When we move to the men, the journeys really proliferate.
the Gauguin effect
If we were really desperate to resurrect the Children of the Sun, we might muse on the similarity between the unusual mitochondrial haplotype from Polynesia and a closely related sequence from the Basque country in western Europe. Sykes makes only passing mention of this rather peripheral possibility, but when he and his colleague, Matt Hurles, shifted their attention from the maternally inherited mitochondrion to the paternally inherited Y-chromosome, Europe loomed large. What they found was that the Y-chromosomes from no less than 30 percent of modern Polynesian men descend, not from any inhabitants of the adjacent continental landmass, but from distant Europeans. The sailors, whalers, artists and missionaries, at whom Sykes and Hurles waved a stern finger, traced a varied pattern of discrete paths into the Pacific, and yet the genetic influence has been substantial. They almost certainly occupy quite a late chapter in the story, but illustrate a central element of the whole tale.
civilization comes home
Ancestral journeys and cultural identity often go hand in hand. It suits us to know where we have ‘come from’, and the stories we tell about that journey of settlement help to forge a sense of collective identity. Essential elements of those stories are the places we definitely did not come from, inhabited by people from whom we see fit to differentiate ourselves. The notion of an ancestral journey is more than a matter of dry, scientific record. It has also played a major role in forging many a cultural identity, not least in the continent that has delved deepest into its own prehistory to trace the journeys at the foundation of western civilization.
For centuries, European scholars have been intrigued by the civilizations of antiquity. They have told a number of different stories about how those ancient civilizations came about and how they led to the modern civilizations of which they themselves were part. The routes to building those stories were widespread. Researchers scrutinized ethnic traditions and notions of race, and looked at the variations between different European languages and certain Asian languages whose vocabularies seemed to display similarities. They examined archaeological remains and posited ancient territorial expansions that resonated with their own experience of European territorial expansion around the world.
In the nineteenth century, those stories had at their end point a white, Christian, civilized Europe, increasingly compartmentalized into nation states. Their starting points were the ancient archaeological sites that seemed to reveal so many of the precursors to European civilization–towns, trade, monuments and the written word–features that drew attention to the east Mediterranean and beyond, to Mesopotamia. Around this civilized core were different cultures, described by words such as ‘barbarian’, and generally associated with darker skins, simpler settlements and the absence of a written text. In telling stories of the origin of white civilized Europe, it was as important to keep these peoples and their cultural traditions outside those stories as it was to keep other peoples and traditions inside. As the prehistories of north and east Africa became better known, so would this exclusion require an increasingly vigorous editing job.
In his book Black Athena, Martin Bernal examined the editorial task facing nineteenth-century scholars in diminishing the place of Africa in the story of Europe. The greatest challenge in this was the Nile Valley, an obvious corridor of contact between the two continents. Those scholars knew of the evidence, both textual and archaeological, for influences upon Greece from Egypt and the continent to the south. This tended, however, to be put to one side in favour of evidence of influence from the east that was in line with the more palatable Aryan stories of origin. The growing study of Indo-European languages provided a new basis for charting these Aryan journeys, and marking out boundaries in tune with contemporary notions of European identity.
The various data-sets, archaeological and linguistic, that were brought together in the context of these nineteenth-century ideas remain with us, and form the starting point of our inquiries into Europe’s past. It occurred to a number of the new molecular archaeologists that the more contentious ideas about Europe and its neighbours could be tackled with modern and ancient DNA, in particular the relationship between ancient Europe and Africa. It should be said straight away that such a relation has two dimensions, only one of which is accessible through DNA. The interchange of ideas and cultural concepts must continue to be explored through text and artefact. It is only when people too are moving or intermarrying that something may show up in the molecular tachometer of DNA.
journeys from sub-saharan africa
When it comes to tracing this particular journey, molecular archaeologists have two points in their favour and one major point against. The first point in their favour is the rich diversity of African genetic sequences. Each of the human family trees assembled from molecular information has revealed a considerable breadth in the continent where humanity evolved, with only a reduced range of genetic diversity leaving Africa to populate the rest of the world. This provides a range of haplotypes to serve as specific sub-Saharan markers. The second point in favour is the long history of mummification, both natural and artificial, along the Nile Valley, the principal corridor between Africa and the Mediterranean. Indeed, it was an Egyptian mummy that provided Svante Pääbo with his first identifiable sequence of ancient human DNA. Set against these two positives, the major negative is that much of the Nile corridor has been, and remains, extremely hot -not the ideal conditions for the preservation of ancient biomolecules. None the less, it was worth a try.
Midway between Egypt and sub-Saharan Africa, the Nile passes through the ancient kingdom of Nubia. The Nubians were known to be in contact with the societies to the north and the south, and were geographically on the crossroads between distinctive human gene pools. Several population movements up and down the Nile are recorded in history and the DNA of the people living in the Sudan today will reflect that. In Barcelona, Carles Lalueza Fox wanted to reach back beyond those more recent migrations to see what movement there had been by the time of the classical civilizations of the Mediterranean. A Spanish archaeological team in Sudan had been excavating near the modern settlement of Abri, just over 100 km south-west of Lake Nuba. They had unearthed a cemetery of naturally mummified bodies, radiocarbon dated to the later first millennium BC, providing Lalueza Fox with some suitable specimens for ancient DNA research.
For a genetic marker, he chose a point mutation that appeared to be very ancient indeed. It had occurred prior to some of the earliest branching in the human mitochondrial tree, in a part of the tree that remained within Africa. The single base replacement at position 3,594 on the Anderson sequence is found throughout sub-Saharan Africa, but is now rare elsewhere in the world. It is carried by over 95 percent of Pygmies and San and !Kung bushmen, and by around 70 percent of sub-Saharan Africans as a whole. The mutation provided Lalueza Fox with an ideal tracer of movement northwards of sub-Saharan genes. His ancient Nubians lay midway between sub-Saharan populations and those in Egypt and the north. He cleaned and ground teeth and bones from twenty-nine ancient individuals, and amplified a 109 base-pair sequence spanning the mutation site. Fifteen of the individuals came up with a successful amplification, revealing that four of the fifteen carried the marker mutation that led back to sub-Saharan Africa. That 25 percent is midway between the figures for
populations to the south and to the north. Looking at these results statistically, Lalueza Fox concluded that somewhere between one-fifth and a half of the population’s genes derived from sub-Saharan Africa, and the remainder from the north.
To look at the north-east African corridor more closely, he went on to place the ancient Nubian data in the context not only of modern data from Ethiopia and Israel, but also of the continental data from Africa and Europe. The characteristic mutation was carried by almost three-quarters of those from sub-Saharan Africa, just less than half the Ethiopians examined, around a quarter of the ancient Nubians, one in eight Palestinian Arabs, and a mere trace of mainland Europeans. In antiquity, it was not just political influence and trading goods that were moving along the Nile corridor. The data fell along the corridor in a neat ‘cline’ or linear sequence of variation. The genetic cline shows that people were moving back and forth as well, and in significant numbers. Since then, Matthias Krings, who in Pääbo’s lab had led the Neanderthal analyses, has also published data on modern variation in this particular sub-Saharan marker. Mainland Europe lies at the end of this cline, and the frequencies for the characteristic mutation fall virtually to zero, but not quite.
Throughout history, the island of Sicily has been one of the major cultural crossroads of the Mediterranean. Sure enough, a survey of ninety unrelated islanders produced a frequency of the characteristic sub-Saharan mutation at just over 4 percent. This is about a third of the frequency at which the mutation is found among the only Arab data then available, from Palestinian Arabs studied in Israel. There are many episodes in the past when this gene flow could have linked the central Mediterranean ultimately with sub-Saharan Africa, and the mediaeval period of Islamic Sicily is an obvious contender. Indeed, the untested assumption when African haplotypes turn up in modern European populations is that they relate to a recent episode of immigration. However, with so little ancient human DNA study, placing the flow within a time frame remains a matter for conjecture. What the ancient Nubian mutations do indicate is that a substantial flow of genes was occurring between the south and the north of the Sahara at a time when the classical civilizations of the Mediterranean were flourishing.
ex oriente lux
For the ancestry of Europe and its civilizations, scholars have repeatedly looked to the east rather than the south. Their points of focus have been Middle East and Near East as the source of all things European, a principle summed up in a phrase that captured Childe’s imagination–Ex Oriente lux, or ‘Out of the Orient, light’. In the arguments of different scholars, significant movements of peoples from the east have clustered in one or more of three phases. A very ancient phase is associated with Palaeolithic hunter-gatherers, a subsequent phase with pioneer farmers, and a still later phase with monumental sites and the knowledge of metalworking.
Colin Renfrew had used carbon dates to query the extent of movement in the third of these phases. By the time the mitochondrial control region had been recognized as a rather interesting evolutionary tachometer, Cavalli-Sforza’s emphasis on the second phase, based on the evidence of expressed genes, was gaining adherents. Little interest had been shown during the 1970s and early 1980s in his demic diffusion argument for a burgeoning population of pioneer farmers expanding from the Fertile Crescent. It had not really struck a chord with archaeological thinking at the time. By 1986, however, his genetic database had reached impressive proportions. Furthermore, it was backed up by a comprehensive archaeological and linguistic survey. In 1987, Renfrew published a detailed argument to account for the spread of Indo-European languages. He attributed that spread, not to the later prehistoric horse-riding Kurgan cultures favoured by Maria Gimbutas, but instead to a wave of pioneer farmers, drawing directly on Ammerman and Cavalli-Sforza’s demic diffusion model. That ‘wave of advance’, spreading from the south-east across to the north-west, now seemed to enjoy the support of three quite distinct data-sets, and notable support from a range of disciplines. In 1994, Cavalli-Sforza’s global synthesis of human genetic variation was published. It was a truly exciting work, in which thousands of years of historical geography were revealed in the minute variations in the blood of the living. The great journeys of the human past could be followed over every continent. But within a couple of years another line of genetic evidence was telling a different story.
While long lists of blood group data were being brought together with other genetic information by Cavalli-Sforza and his team, a much smaller fragment of the human genome was under scrutiny by Bryan Sykes’s team in Oxford. They were engaged in the now familiar task of charting the individual variations in the control region of the human mitochondrial genome. It was an approach that had already been taken in Asia, America and Polynesia. Now it was being applied within Europe. Samples were acquired from just over 300 modern Europeans from as far afield as Portugal and Finland, together with samples from the Middle East and Turkey. One of the most variable stretches of the control region was sequenced in each case, and the sequences compared. Although the team was working with one of the fastest evolving sequences of the human genome, the 10,000 years since agriculture began remains an extremely short period of time. We would not expect a great deal of diversity to have accrued since the founder population spread across the continent–perhaps one or two base replacements in some individuals. Sykes and his colleague, Martin Richards, found that the variation was rather greater than that. Almost a third of the individuals studied deviated from the Anderson consensus sequence, some by as many as six base replacements along the short stretch of targeted DNA. This European group was looking rather more ancient than a population passing through a very recent bottleneck. Sufficient time had passed to allow a significant amount of evolutionary change. As we have seen with the colonization of the New World, however, this need not preclude a simple model of population expansion. Just as the ancient Beringians could bring a fair degree of diversity with them on their passage into America, so could the ancient farming populations from the Fertile Crescent.
This was where Sykes’s Middle Eastern data came into play. If they had brought that diversity with them, it might still be discernible in today’s populations from the Fertile Crescent. The forty-two Middle Eastern individuals in Sykes’s sample displayed something different. Rather than encompassing the European variation, they formed instead a rather distinctive group, out on a limb from the broader European pattern, and sharing their sequence pattern with only a small minority of modern Europeans.
Sykes and Richards realized that their targeted analysis of one well-studied stretch of DNA was telling a quite different story from what had emerged from Cavalli-Sforza’s masterly survey. They argued that, while a population spread of early farmers from south-west Asia had taken place, it only contributed a small fraction to the European gene pool. There was no wave of population advance sweeping across Europe, just a trickle. The sequence diversity had to be explained in terms of a more ancient population. Their hypothesis cut across the core journey within Cavalli-Sforza’s grand synthesis. The principle ancestors of modern Europeans were the hunter-gatherers who were already there, thousands of years before the spread of farming. Agriculture travelled as an idea more than as a community, and the presumed population replacement had not taken place.
A few months after the publication of the Oxford findings, Cavalli-Sforza delivered the McDonald lecture at Cambridge. He projected one synthetic map after the other, taking a captivated audience through several major episodes of the human past. We waited to hear his assessment of the Oxford paper, but when it came his characteristic charm and ease were put to one side. His feelings were unambiguously expressed first towards the Oxford group, and then towards the offending journal for letting their flawed arguments through to publication.
While a polite Cambridge audience wrestled with a certain amount of schadenfreude at so eminent a disposal of an Oxford project, we were left wondering how they had managed so to rattle the senior man. It would have been l
ess surprising to see him dispose of them with a learned smile, rather in the manner of dispatching a careless question or overly arrogant student. The challenge clearly ran deeper than that, for all the implications that the Oxford approach was carelessly flawed.
The analysis put forward by Sykes and Richards was not only different from Cavalli-Sforza’s approach in its conclusion, but also in its conception. Cavalli-Sforza had been concerned with the larger genetic picture created by extracting patterns of similarity in the expression of between 50 and 100 genes and looking for parallel patterns within linguistics and the archaeological record. The approach taken by Sykes’s team was much more narrowly targeted. The DNA sequence they examined was far shorter than even one of the many genes Cavalli-Sforza utilized. He had sought out global trends within a vast genetic data-set, while Sykes had sharpened the focus on a specific evolutionary process involving only part of a single DNA molecule. In terms of an analogy with exploring the evolution of flight, one scholar was looking across the skies at a multitude of airborne forms, while the other was dissecting the wing of an individual bird. The different styles of their conclusions were in part a consequence of that.
At first, academic opinion tended to side with Cavalli-Sforza, and various alternative reasons were put forward to explain away the Oxford group’s results. It was pointed out that the mitochondrial sequence is only a very tiny element in the human genome and should not provide a basis for argument on its own. Others pointed to the potential skewing effect of residual hunter-gatherers, reflected in the language and genetics of today’s Basque and Saami peoples. A further weak point was the mitochondrial clock, the powers of precision and accuracy of which are habitually over-used. It could always be argued that such a timepiece might vary in a way we simply do not understand, and in certain cases be several magnitudes out. Bryan Sykes, who had stuck with the work and enhanced his results, could see that the way to move forward was to look in well-dated, well-positioned contexts for ancient DNA.
