Lone Survivors: How We Came to Be the Only Humans on Earth

1

The Big Questions

It is barely 150 years since Charles Darwin and Alfred Russel Wallace presented their ideas on evolution to the world. A year later, in 1859, Darwin was to publish one of the most famous of all books, On the Origin of Species. Then, the first fossil human finds were only beginning to be recognized, and paleontology and archaeology were still in their infancy. Now, there is a rich and ever-growing record from Africa, Asia, and Europe, and I have been privileged to work in one of the most exciting eras for discoveries about our origins. There have been highly significant fossil finds, of course, but there have also been remarkable scientific breakthroughs in the amount of information we can extract from those finds. In this first chapter I will outline the evidence that has been used to reconstruct where our species originated, and the very different views that have developed, including my own. There are in fact two origins for modern human features that we need to consider. Here, I will talk about our species in terms of the physical features we humans share today, for example, a slender skeleton compared to our more robust predecessors, a higher and rounder braincase, smaller brow ridges, and a prominent chin. But there are also the characteristics that distinguish different geographic populations today—the regional or “racial” characteristics, such as the more projecting nose of many Europeans, or the flatter face of most Orientals. I will discuss their quite different origins later in the book.

In The Descent of Man (1871), Darwin suggested that Africa was the most likely evolutionary homeland for humans because it was the continent where our closest relatives, the African apes, could be found today. However, it was to be many years before the fossil evidence that was ultimately to prove him right began to be discovered. Before then, Europe with the Neanderthals, “Heidelberg Man,” and the spurious “Piltdown Man,” and Asia with “Java Man,” were the foci of scientific attention concerning human ancestry. But the 1921 discovery of the Broken Hill skull in what is now Zambia, and the 1924 discovery of the Taung skull (from South Africa), started the process that gave Africa its paramount importance in the story of human evolution, even if that process still had many years to run. By the 1970s a succession of fossils had established that Africa not only was the place of origin for the human line (that is, the continent in which the last common ancestor of humans and chimpanzees lived) but was probably also where the genus Homo (humans) had originated. But where did our own species, Homo sapiens (modern humans), originate? This was still unclear in the 1970s and remained so until quite recently.

When Charles Darwin wrote in the Origin of Species, “light would be thrown on the origin of man and his history,” he was reluctant to say any more on the subject, as he admitted twelve years later in the introduction to The Descent of Man: “During many years I collected notes on the origin or descent of man, without any intention of publishing on the subject, but rather with the determination not to publish, as I thought that I should thus only add to the prejudices against my views.” But in the intervening years he had been fortified by a growing number of influential supporters and thus felt ready—finally—to tackle the controversial topic of human origins. He then went on to say: “The sole object of this work is to consider, firstly, whether man, like every other species, is descended from some pre-existing form; secondly, the manner of his development; and thirdly, the value of the differences between the so-called races of man.” However, Darwin acknowledged that there were still many doubters, something that unfortunately remains as true today as it was then: “It has often and confidently been asserted, that man’s origin can never be known: but ignorance more frequently begets confidence than does knowledge: it is those who know little, and not those who know much, who so positively assert that this or that problem will never be solved by science.”

Darwin then proceeded to pay tribute to a number of other scientists for their work on human origins, particularly the German biologist Ernst Haeckel, and this is especially interesting as Haeckel differed from him and Thomas Huxley (“Darwin’s bulldog”) over a critical question about our origins, a question that continues to be debated even today. In The Descent of Man Darwin wrote: “We are naturally led to enquire, where was the birthplace of man at that stage of descent when our progenitors diverged from the catarrhine stock [the catarrhines group includes apes and monkeys]?… In each great region of the world the living mammals are closely related to the extinct species of the same region. It is therefore probable that Africa was formerly inhabited by extinct apes closely allied to the gorilla and chimpanzee; and as these two species are now man’s nearest allies, it is somewhat more probable that our early progenitors lived on the African continent than elsewhere.” However, he then proceeded to caution, “But it is useless to speculate on this subject … as there has been ample time for migration on the largest scale.”

Not only did Darwin have to deal with a dearth of fossil evidence in 1871, including a complete absence of any humanlike fossils from Africa, but there was also no knowledge of the concept of continental drift (the idea that landmasses migrated in the past, splitting and realigning as they moved across the Earth’s surface). This process is now known to underlie many of the present distributions of plants and animals (for example, the unique assemblages of species found in places like Australia and New Zealand). Previously, to explain puzzling links between species in different regions, now-sunken continents were often postulated. For example, lemurs are rather primitive primates that today are found only on the island of Madagascar, some three hundred miles off the coast of Africa, but ancient lemurlike fossils had been found in the Indian subcontinent, and such similarities led the British zoologist Philip Sclater to hypothesize in 1864 that there was once a large continent, which he named Lemuria, stretching across much of what is now the Indian Ocean.

Using the concept of this lost continent, Haeckel argued for a different ancestral homeland for humans: “There are a number of circumstances which suggest that the primeval home of man was a continent now sunk below the surface of the Indian Ocean, which extended along the south of Asia … towards the east; towards the west as far as Madagascar and the southeastern shores of Africa. By assuming this Lemuria to have been man’s primeval home, we greatly facilitate the explanation of the geographical distribution of the human species by migration.” Moreover, Haeckel differed from Darwin and Huxley in favoring the gibbon and orangutan of southeast Asia as better ape models for human ancestry than the gorilla and chimpanzee of Africa. And whereas Darwin followed the geologist Charles Lyell in arguing that the fossil record of human evolution was still unknown because the right regions had not yet been searched (in particular Africa), Haeckel preferred the explanation that most of the critical evidence was now sunk beneath the Indian Ocean.

Eugène Dubois and his “Pithecanthropus erectus” skull.

During Darwin’s lifetime, the Neanderthals were already known from their fossil remains as ancient inhabitants of Europe. While some scientists pushed them into the position of “missing links,” reconstructing them with bent knees and grasping big toes, others like Huxley recognized them as big-brained, upright, and unmistakably human. Darwin never lived to see the first discovery of a really primitive human fossil, announced by a Dutch doctor, Eugène Dubois, in 1891. Dubois had been inspired by Haeckel’s writings to get an army posting to what was then the Dutch East Indies (now Indonesia), to search for ancient remains. Haeckel had created the name “Pithecanthropus alalus” (“Ape Man without Speech”) for a hypothetical link between apes and humans that he believed had once lived in Lemuria. Dubois was blessed with luck in his excavations on the island of Java and soon found a fossilized and apelike skullcap and a human-looking thighbone. He named these “Pithecanthropus” (in honor of Haeckel) “erectus” (because the femur indicated this creature walked upright, as we do). We now know this species as Homo erectus, a wide-ranging and long-lived species of early human. But because this first find of the species was made on the Indonesian island of Java, it tended to reinforce Haeckel’s and Dubois’s notions of a Lemurian/southern Asian origin for humans, rather than an African one.

In naming “Pithecanthropus erectus,” Dubois was following the system laid down over a century earlier by that greatest of all classifiers, the Swedish naturalist Carl Linnaeus. The Chinese sage Confucius said that it was “a wise man” who specified the names of things, and by happy coincidence this was the name, in Latin, that Linnaeus chose for the human species: Homo sapiens. Before Linnaeus there were many different ways of naming and grouping plants and animals, often based at random on particular features that they showed—color, say, how they moved around, or what they ate. But Linnaeus believed in grouping living things by the bodily features they shared, and at the heart of his system were the two names applied to every natural kind, or species: its group or genus name capitalized, and its particular species name. Thus Homo (“Man”) and sapiens (“wise”). The system is a bit like a surname (the genus name Homo) and a first name (differentiating the different children with a particular surname, in our case sapiens). In the most-cited tenth edition of his book Systema Naturae (1758) he also named four geographic subspecies: “europaeus,” “afer,” “asiaticus,” and “americanus,” introducing some dubious anecdotal behavioral distinctions in line with then current European notions about the superiority of the European subspecies. For example, while “europaeus” was, of course, governed by laws, “americanus” was governed by customs, “asiaticus” by opinions, and the African subspecies “afer” by impulse.

Map showing early human sites.

In the early 1900s, evidence continued to accumulate in favor of a non-African origin for humans, and the focus returned to Europe. Further Neanderthal remains were found in Croatia and France, and a more ancient and primitive fossil jawbone was unearthed in the Mauer sandpit near Heidelberg in Germany in 1907. As enough material began to accumulate, scientists started to build evolutionary trees from the fossil evidence. These tended to fall into two main categories: ones where the fossils were arranged in a linear sequence leading from the most primitive form (for example, Java Man or Heidelberg Man) to modern humans, with few or no side branches (like a ladder); and others (like a bush) where there was a line leading to modern humans, and the other fossils with their primitive features were placed in an array of side branches leading only to extinction.

The combination of Darwin’s and Wallace’s publications on the transmutation of species and a proliferating Pleistocene fossil human record led to the expectation that there must have been many more ancient species of humans (the Pleistocene is a recent geologic epoch, poorly dated during Darwin’s time, but now believed to stretch from about 12,000 to 2.5 million years ago). William King had named the first fossil-based species Homo neanderthalensis in 1864, from the skeleton discovered in the Neander Valley in 1856. Within fifty years, the new European finds were being assigned to dozens of new human species in an unfortunate tumult of typology, where trivial differences were elevated to assume real biological significance. Thus, the completely modern-looking remains that had been found in the sites of Cro-Magnon, Grimaldi, Chancelade, and Oberkassel became the human species “spelaeus,” “grimaldii,” “priscus,” and “mediterraneus,” respectively, while the remains from Spy, Le Moustier, and La Chapelle-aux-Saints became “spyensis,” “transprimigenius,” and “chapellensis,” despite their resemblance to the remains already designated H. neanderthalensis from the Neander Valley. This trend for what we can call extreme “splitting” continued up to about 1950, when the pendulum swung back to the opposing tendency to “lump” fossils together in just a few species.

A replica of the jawbone unearthed in the Mauer sandpit near Heidelberg in Germany in 1907, together with one of the Boxgrove incisor teeth.

Suggestions that Europe may have hosted even more primitive human relatives started to emerge from a gravel pit at Piltdown in southern England in 1912, giving rise to yet another species called “Eoanthropus dawsoni” (“Dawn Man of Dawson”—Charles Dawson being the principal discoverer). Parts of a thick but large-brained skull, coupled with a distinctly apelike jaw, turned up there with ancient animal fossils and primitive stone tools, suggesting an age as great as that of Java Man. Africa had nothing to compare with these burgeoning finds, but that finally began to change in the 1920s. However, circumstances were such that these first finds still failed to switch the focus of human origins to Africa.

The Broken Hill (Kabwe) skull, discovered in 1921, was the first important human fossil from Africa, but it was a puzzling find. Although it was assigned to the new species “Homo rhodesiensis” by Sir Arthur Smith Woodward of the British Museum, the Czech-American anthropologist Aleš Hrdlicka dubbed it “a comet of man’s prehistory” because of the difficulty in deciphering its age and affinities. The skull was found in cave deposits that were being quarried away during metal ore mining, in what is now Zambia (then the British colony of Northern Rhodesia). It’s one of the most beautifully preserved of all human fossils, but it displays a strange mixture of primitive and advanced features, and its face is dominated by an enormous brow ridge glowering over the eye sockets. And because it was found during quarrying, which eventually destroyed the whole Broken Hill mine, its age and significance remain uncertain even today (but see the final chapter for the latest developments).

Three years later an even more primitive find was made in a limestone quarry at Taung, South Africa—a skull that looked like that of a young ape. It was studied by a newly established professor of anatomy in Johannesburg, named Raymond Dart, and in 1925 he published a paper in the scientific journal Nature, making some remarkable claims about the fossil. He argued that it showed a combination of ape and human features, but that its teeth, brain shape, and probable posture were humanlike. Dart named it Australopithecus africanus (“Southern Ape of Africa”), and he declared that it was closely related to us, and even a potential human ancestor. Dart’s claims were treated with great skepticism by the scientific establishment, particularly in England. This was partly because of judgments about Dart’s youth and relative inexperience, and partly because the fossil was that of a child (young apes may look more “human” than adult apes). Others thought that the finds from Java, Heidelberg, and Piltdown provided much more plausible ancestors than Australopithecus africanus. And finally, the location and estimated age of Taung also counted against it.

No one (not even Darwin and Huxley) had considered southern Africa as a location for early human evolution, and as the Taung skull was guessed to be only about 500,000 years old, it was thought too recent to be that of a genuine human ancestor. Instead, it was considered to represent a peculiar kind of ape, paralleling humans in some ways. We now know, of course, that the australopithecines represented a long and important phase of human evolution that lasted for over 2 million years, and which is recognized at sites stretching from Chad in the Sahara to many more in eastern and southern Africa. And we have also known since their exposure in 1953 that the misleading Piltdown remains were fraudulent and had nothing to do with our ancient ancestry.

Other finds made at this time continued to keep the focus outside of Africa, and those made in cave deposits at Zhoukoudian near Beijing from 1921 onward began to reveal a Chinese counterpart to Java Man initially dubbed “Sinanthropus pekinensis” (“Chinese Man of Peking”). Systematic excavations carried out from 1927 until the present day have yielded many skull and body parts of humans who lived there about half a million years ago, people who resembled the growing collection of fossils from Java closely enough for them to be eventually grouped in the single species Homo erectus. This species is a crucial one for studies of our origins, because it’s at the heart of radically different views of our evolution that have emerged over the last seventy years or so. Most anthropologists recognize the existence of at least two human species during the last million years—the extinct Homo erectus and our own species, Homo sapiens—but there are very different views on how these species are related.

Franz Weidenreich and some of the “Peking Man” fossils of Homo erectus that inspired him to create an early version of the Multiregional model of human origins.

What is now known as the Multiregional model of modern human origins was first proposed in the 1930s by Franz Weidenreich, a German anthropologist, who based many of his arguments on studies of the Zhoukoudian Homo erectus fossils. Weidenreich suggested that Homo erectus gave rise to Homo sapiens across its whole range, which, about 1 million years ago, included Africa, China, Indonesia, and perhaps Europe. In his view, as the species dispersed around the Old World (it’s not known from regions such as Australia and the Americas), it developed the regional variation that lies at the roots of modern “racial” differentiation. Particular features in a given region persisted in the local descendant populations of today. For example, he argued from the fossils that Chinese Homo erectus specimens had the same flat faces and prominent cheekbones as modern oriental populations, while Javanese Homo erectus had robustly built cheekbones and faces that jutted out from the braincase, characteristics argued to be especially marked in modern Australian Aborigines.

At the other extreme from Weidenreich’s Multiregional model was the view that the special features of modern humans (such as a high forehead, a chin, and a slender skeleton) would have required a long time to evolve, and hence the line leading to Homo sapiens (the “pre-sapiens” lineage) must have been very ancient and developed in parallel with large-browed and robust forms such as Homo erectus and the Neanderthals. This is an old idea, which came to prominence early in the twentieth century through influential researchers like Marcellin Boule (France) and Arthur Keith (United Kingdom), and aspects of it were taken up later by Louis Leakey, working in Kenya and Tanzania. The supporting evidence came and went through the last century, including at times specimens like Piltdown and the modern-looking Galley Hill skeleton from Kent—the former now known to be a fake and the latter wrongly dated.

Between the extremes of Multiregionalism (which potentially included every human fossil in our ancestry) and the Pre-sapiens model (which excluded most of them), there were intermediate models, ones which featured early Neanderthals in the story. The critical fossils this time were from Mount Carmel in what was then Palestine. They were discovered by an international expedition excavating a series of caves near Haifa during the late 1920s and 1930s. In two of the caves, Skhul and Tabun, they found human fossils that had apparently been intentionally buried. Moreover, they were associated with the kinds of stone tools that in Europe were associated with the Neanderthals. And yet the fossils seemed to show mixtures of Neanderthal and modern characteristics, so how should they be interpreted? In the 1930s there were no accurate methods of dating available, and so the Tabun and Skhul fossils were assumed by their describers, Theodore McCown and Arthur Keith, to be roughly contemporaneous with each other. Some suggested that the finds might represent hybrids between moderns and Neanderthals, but McCown and Keith preferred to regard them as members of a single but variable ancient population, perhaps one close to the divergence of the Neanderthal and modern lines. (In fact Keith could not quite abandon his pre-sapiens leanings and thought that they were still probably off the line leading to us, because of their Neanderthal features.)

Louis Leakey with the Olduvai Gorge “Zinjanthropus” skull, which his wife, Mary, discovered in 1959. It was the first important fossil to be dated by the potassium-argon method.

But others saw them as evidence for a pre-Neanderthal rather than pre-sapiens ancestry for modern humans, with the late or “classic” Neanderthals subsequently heading off the main line to the sidings of extinction. Following this line of argument, the American paleoanthropologist F. Clark Howell developed a neat scenario during the 1950s where “unspecialized Neanderthals” about 100,000 years ago became isolated in Europe by the last Ice Age and evolved away from Homo sapiens. At the same time, those in the Middle East (such as Tabun) evolved toward modern humans via forms like those found at Skhul. Then, to complete the story, about 35,000 years ago these Middle Eastern “proto–Cro-Magnons” migrated into Europe and replaced their European Neanderthal cousins.

In contrast to this Early Neanderthal model of modern human origins, which gave the Neanderthals at least a bit part in our evolution, there were two developments out of Weidenreich’s Multiregionalism after his death in 1948 that returned the Neanderthals to a central role in our evolution, and in one case even extended their role globally. The American anthropologist Carleton Coon used new fossil material to develop a comprehensive global scheme of the evolution of five different lineages of Homo erectus, two in Africa, and one each in Europe, China, and Australia. These five lineages evolved largely independently to become what Coon regarded as the modern races of Homo sapiens: “Capoid” (the Bushman of South Africa and related peoples), “Negroid,” “Caucasoid,” “Mongoloid,” and “Australoid.”

In this respect, Coon differed fundamentally from his mentor, since Weidenreich considered human evolution to consist of a network of lineages constantly exchanging genes and ideas, whereas Coon was quite frank about the divided lineages and the implications of their inferred different rates of evolution: “Wherever Homo arose, and Africa is at present the most likely continent, he soon dispersed, in a very primitive form, throughout the warm regions of the Old World … If Africa was the cradle of mankind, it was only an indifferent kindergarten. Europe and Asia were our principal schools.”

The American paleoanthropologist C. Loring Brace gave Weidenreich’s ideas a distinctly Neanderthal twist by arguing that Homo erectus evolved to modern humans in each part of the populated world by passing through a “neanderthaloid” phase. In essence, according to Brace, the Neanderthals and equivalent ancient people across the inhabited world used their front teeth as tools for manipulating food and materials, and this is what produced their especially prominent midfaces, large incisor teeth, and distinctive skull shape. When more advanced tools of the Upper Paleolithic (Upper Old Stone Age) were invented about 35,000 years ago, demands on the teeth and jaws were lifted, and so the face and skull were transformed into the shape we have today.

These were the main ideas about the origin of modern humans that I set out to test when I began my studies for a Ph.D. at Bristol University in 1970: the global Multiregional model and its Braceian development, the Early Neanderthal model, the Pre-sapiens model (with no place for erectus or Neanderthals), and one rather vaguer scheme developed by the anthropologists Bernard Campbell and Joseph Weiner called the Spectrum Hypothesis. This argued that ancient humans had different blends of what would become modern human characteristics, and they contributed in part, and differentially, to the evolution of Homo sapiens. So in a sense the Spectrum Hypothesis was multiregional, but some lineages had a much greater contribution to our ancestry than others. A bit of a revolution was under way around 1970, as computing power started to increase and began to influence the biological sciences. Most analyses of human evolution up to 1970 were based on direct observation, and where measurements of a fossil were taken, these were usually compared individually or through an index of just two measurements. However, multivariate programs were becoming available that could look at large numbers of measurements and specimens simultaneously, allowing more sophisticated studies of differences in size and shape. Such analyses were at the center of my studies for a doctorate, and in July 1971 I left the United Kingdom on a trip to museums and research institutes in ten European countries. The aim was to gather as many data as possible on the Neanderthals and their modern-looking successors in Europe, the Cro-Magnons, to see whether the evolutionary pattern was one of continuity or rupture. I only had a modest grant from the Medical Research Council for a four-month trip, and so I drove my old car, sleeping in it, camping, or staying in youth hostels—in Belgium I even spent one night in a shelter for the homeless. I survived many adventures, including several border confrontations and two robberies, but by the end of my 5,000-mile trip I had collected one of the largest data sets of Neanderthal and early modern skull measurements assembled by anyone up to that time.

Chris Stringer on his 1971 research trip around Europe. It’s washing day at a campsite in Yugoslavia.

Over the next two years I analyzed this information, adding comparative data on non-European fossils and modern human populations (the latter generously supplied by the American anthropologist William Howells). The measurements were transferred to data cards and fed into a computer the size of several rooms, but which had less processing power than my last mobile phone! Nevertheless, the results were instructive. Neanderthal skulls were no more similar to those of recent Europeans than they were to Africans, Eskimo, or Native Tasmanians, and Cro-Magnon skulls did not neatly slot between the Neanderthals and recent Europeans. Early modern skulls from around the world seemed to cluster with their modern counterparts rather than with any archaic skulls from the same regions. The former results provided no support for a Neanderthal ancestry for the Cro-Magnons, and the latter results contradicted Multiregional and Spectrum expectations. Studying the sequence in Europe before Neanderthal times also gave no support to the Pre-sapiens model either, because very early European fossils could not be divided into modern-like and Neanderthal-like; they seemed to show the gradual development of only Neanderthal features through time.

Things were not quite as clear in the Middle East, although there did not seem to be any “intermediate” fossils between Neanderthals and moderns there either. Skulls from Tabun and the Israeli cave of Amud seemed to be basically Neanderthal, while those from Skhul Cave seemed much more modern. But because none of these finds were well dated in the 1970s, I couldn’t exclude the possibility that, given enough time, the Israeli Neanderthals could have been transformed into early moderns, in line with the Early Neanderthal model of scientists like Clark Howell. However, a surprising alternative ancestor for the Skhul and Cro-Magnon early moderns did emerge from my results. A skull discovered in 1967 in the Omo Kibish region of Ethiopia, by a team led by Richard Leakey (the son of the famous prehistorians Louis and Mary Leakey), looked very modern in my skull shape analyses, confirming the first studies by the anatomist Michael Day; yet preliminary dating work suggested it could have been as much as 130,000 years old, more ancient than most Neanderthals. And there was an enigmatic North African skull, found in the Moroccan site of Jebel Irhoud in 1961. In skull shape it seemed Neanderthal in some ways, yet its facial shape was non-Neanderthal, partly primitive and partly modern. With an age thought to be only around 40,000 years, it was difficult to fit Jebel Irhoud into any scenario, but it and the Omo skull provided clues that Africa was going to have its own story to tell, when more data came in.

As my work developed through the 1970s and early 1980s, I gravitated increasingly toward what Bill Howells in 1976 had dubbed the Garden of Eden (or Noah’s Ark) model. This was named not because Howells was any kind of biblical creationist, but because of the implication that all modern human variation had developed from a single center of origin. A lack of fossils from many parts of the world, together with inadequate dating for many of those we did have, meant that neither Howells nor I could specify where that center of origin might have been, although we thought we could exclude the European and Middle Eastern territories of the Neanderthals. We both believed that the distinctive shared features of modern humans, such as the high rounded skull, small brows, and chin, implied a recent common origin, as otherwise there would have been much greater differentiation over time. And I started to move away from the then widespread idea that fossils as different-looking as Broken Hill, the Neanderthals, and Cro-Magnon should all be classified with us as variants of our species, Homo sapiens. Initially I agreed with some other workers in differentiating “anatomically modern sapiens” (such as Skhul and Cro-Magnon) from “archaic sapiens” forms such as Neanderthals and Broken Hill. But during the 1980s I increasingly favored limiting the sapiens term to fossils closely resembling us. Moreover, along with a few other heretics, I started to argue that the Neanderthals should be returned to the status granted them by William King in 1864 as a distinct species, Homo neanderthalensis. I also suggested that the Broken Hill skull found in 1921 could be grouped with more primitive European forms (for example, the Heidelberg jaw discovery of 1907) as Homo heidelbergensis.

As my views on our origins were developing toward a single-origin model, evidence began to accumulate that Africa was especially important in this story. The Omo Kibish find was joined by material from the sites of Border Cave and Klasies River Mouth Caves in South Africa. Moreover, new dating work hinted that Africa was not the backwater in cultural evolution that most considered it to be. Archaeologists such as Desmond Clark and Peter Beaumont argued that it might instead have been leading the way in the sophistication of its stone tools. By 1980 I was privately convinced that Africa was the main center of our evolution but, because of dating uncertainties, I could not rule out the Far East as also playing a role. It took another four years for me to take a strong “Out of Africa” stance publicly, as various lines of evidence started to fit together in my mind.

However, further confusion was sown by the strong reemergence of Weidenreich’s Multiregional views in 1984. These were given a new lease on life by Milford Wolpoff (United States), Alan Thorne (Australia), and Wu Xinzhi (China). They distanced themselves from Coon’s views by returning to Weidenreich’s emphasis on the importance of gene flow between the geographic lines, considering the continuity in time and space between the various forms of Homo erectus and their regional descendants to be so complete that all of them should be classified with modern people as representing only one species: Homo sapiens. Thus in this model there was no real “origin” for the modern form of Homo sapiens. A feature like the chin might have evolved in a region such as Africa, and spread from there by interbreeding across the human range, followed by selection for it if it was an advantageous characteristic. Another feature such as our high forehead might have developed in, say, China, and then similarly spread from there through interbreeding. Thus modern humans could have inherited their “local” features through continuity with their ancient predecessors, while global characteristics were acquired via a network of interbreeding.

But new developments in genetics research were about to have a huge impact. In 1982 I became aware of research work on a peculiar type of DNA that is found outside the nucleus of cells, in the mitochondria. These are little bodies that provide the energy for each cell, bodies that probably originated from a once-separate bacterium, which somehow survived being engulfed by a primitive cell. They then coevolved to confer mutual advantage and developed into the mitochondria that most organisms have throughout their cells. In humans, the DNA of a mother’s mitochondria is cloned in her egg when it becomes the first cell of her child, and little or no mitochondrial DNA from the father’s sperm seems to be incorporated at fertilization. This means that mitochondrial DNA (mtDNA) essentially tracks evolution through females only (mothers to daughters), since a son’s mtDNA will not be passed on to his children. This type of DNA mutates at a much faster rate than normal (nuclear) DNA, as we will discuss in chapter 7, allowing the study of short-term evolution. Early work on human mitochondria seemed promising, showing that our species apparently had low diversity and a recent origin, but the geographic patterns seemed unclear as to where that origin might be. By 1986 I had heard through the grapevine that startling new mtDNA results were on the way to publication, and a year later they appeared in the science journal Nature, shaking up arguments about recent human evolution in such a way that things would never be the same again. This seminal publication by Rebecca Cann, Mark Stoneking, and Allan Wilson put modern human origins on the front pages of newspapers, journals, and magazines for the first time.

Milford Wolpoff, an architect of Multiregionalism, with a Homo erectus skull from Java.

About 150 types of mtDNA from around the world were investigated, and their variation was determined. Then a computer program was used to connect all the present-day types in an evolutionary tree, with the most economical pattern of evolutionary change (mutations), reconstructing hypothetical ancestors for the living types. In turn, the program connected those ancestors to each other, until a single hypothetical ancestor for all the modern types was created. The distribution of the ancestors implied that the single common ancestor must have lived in Africa, and the number of mutations that had accumulated from the time of the common ancestor suggested that this evolutionary process had taken about 200,000 years. This, then, was the birth of the now-famous Mitochondrial Eve, or “lucky mother,” since the common mitochondrial ancestor must necessarily have been a female. These results seemed to provide strong evidence for a Recent African Origin view for modern humans, since the research suggested that a relatively recent expansion from Africa had occurred, replacing any ancient populations living elsewhere, along with their mtDNA lineages. However, the work was soon heavily criticized. It was shown that the kind of computer program used could actually produce many thousands of trees which were all more or less as economical as the published one, and not all of these alternative trees were rooted in Africa. Moreover, other researchers criticized the calibration of the time when Mitochondrial Eve lived, while yet others questioned the constitution of the modern samples analyzed (for example, many of the “African” samples were actually from African Americans). As a result, multiregionalists were, for a while at least, able to reject these mtDNA results as irrelevant or misleading, arguing that fossil evidence (and their interpretation of it) remained the only valid approach to reconstructing recent human evolution.

However, the results strongly supported the Recent African Origin view that people like Günter Bräuer (from Hamburg) and I had been developing from the fossils. Günter was less inclined to view Homo sapiens as a newly evolved species, and more inclined to think that hybridization had occurred with people like the Neanderthals, following the dispersal from Africa, but we both welcomed the new mtDNA data. For me, it gave greater confidence that even where the fossil evidence was patchier or more ambiguous, such as the Far East and Australasia, the story of replacement that I had read from the European record probably applied there too.

In 1987 the archaeologist Paul Mellars and I co-organized an international conference in Cambridge where recent fossil and archaeological results were compared with the new DNA data, and the discussions were electric at times, as experts got to grips with the rapidly changing landscape of recent human evolution. A year later, taking the conference discussions and DNA analyses fully on board, I wrote a review of that emerging picture for the journal Science, with my Natural History Museum colleague Peter Andrews. We laid out the two contrasting models of Multiregionalism versus Recent African Origin and what would be expected from the fossil, archaeological, and genetic data if either model was an accurate representation of recent human evolution. (I actually prefer to use the term Recent African Origin [RAO], despite the popularity of Out of Africa, because we know from more ancient fossils that there were earlier human dispersals from Africa. Hence some people distinguish them as Out of Africa 1, Out of Africa 2, et cetera, although we don’t actually know how many there were—and no doubt there were some “into Africa” events as well!)

Overall, we showed that RAO was best supported, although we recognized that the archaeological record in general and the fossil records of several regions in particular were still not adequate to test the models properly. I was shocked, though, by some of the vitriolic reactions to that paper. Both in the anonymous reviews that some other scientists sent to the journal before publication, and in letters and media comments afterward, scorn was poured on our views and interpretations, a scorn that seemed to extend to personal abuse at times. Relations became strained with a number of scientists, some of whom were people I certainly counted as my friends. Cordiality was eventually restored in most cases, but for a few people, what was seen as an extreme position, in league with the heresy of Mitochondrial Eve, was not easily forgiven or forgotten.

Two of the architects of the Recent African Origin model, Günter Bräuer (left) and Chris Stringer, pictured in the 1980s.

As more fossil and, particularly, genetic data emerged to support a recent African origin, what we can term the classic RAO model was developed by a number of researchers, including me, working separately or in collaboration. By the turn of the millennium, this had become the dominant view. Fleshing it out with the consensus view for earlier human evolution, the classic RAO model argued for an African origin of two human species—Homo erectus and Homo sapiens—and perhaps also of Homo heidelbergensis between them (in my view, though, the derivation of heidelbergensis is still unclear). Having evolved from something like the earlier species Homo habilis in Africa nearly 2 million years ago (Ma), Homo erectus dispersed from Africa about 1.7 Ma, in the event commonly known as Out of Africa 1. The species spread to the tropical and subtropical regions of eastern and southeastern Asia, where it may have lingered on, evolved into other forms, or died out. About 1.5 Ma, African erectus developed more advanced stone tools called handaxes, but these did not spread far from Africa until they turned up rather suddenly with the descendant species Homo heidelbergensis in places like southern Europe, and then in Britain, 500,000 to 600,000 years ago.

So my view was that H. heidelbergensis subsequently underwent an evolutionary split around 300,000 to 400,000 years ago: it began to develop into the Neanderthals in western Eurasia, while the line in Africa had evolved into the ancestors of modern humans by about 130,000 years ago. The origin of modern Homo sapiens must have been a relatively recent and restricted one in Africa, based on marked similarities between recent humans in both body form and DNA, and it may have been quite rapid, in one small favored area such as East Africa. Some modern humans dispersed to the Middle East (Israel) about 100,000 years ago, and they had perhaps moved on as far as Australia by about 60,000 years. However, Homo sapiens did not enter Europe until about 35,000 years ago, following the rapid development of more advanced Later Stone Age tools and complex behaviors by African moderns about 50,000 years ago. Such progress finally allowed the moderns to spread into Europe, where, as Cro-Magnons making Upper Paleolithic tools, they quickly took on and replaced the Neanderthals through their superior technology and adaptations. Bear this narrative in mind, as I will revisit it at various times later in this book.

If RAO is the most accurate model, regional (“racial”) variation only developed during and after the dispersal from Africa, so any seeming continuity of regional features between Homo erectus and present counterparts in the same regions outside of Africa must have been as a result of parallel evolution or coincidence, rather than of genes passed down from archaic predecessors, as argued in the Multiregional model. Like that model, RAO argued that Homo erectus evolved into new forms of humans in inhabited regions outside of Africa, but in RAO these non-African lineages eventually became extinct, without evolving into modern humans. Some, such as the Neanderthals, must have been replaced by the spread of modern humans into their regions, and hence the RAO model not only is popularly known as Out of Africa but is sometimes also known as the Replacement model.

As RAO gathered support and influence, it increasingly made an impact on the views of people like the American anthropologists Fred Smith and Erik Trinkaus, who believed in continuity outside of Africa but were not classic multiregionalists. Instead, they advanced what has become known as the Assimilation model, which can be seen as a moderate position between the extremes of RAO and what I have dubbed classic Multiregionalism: one where Africa dominated as the source of modern features, but where these were taken up more gradually by people outside of the continent, through a blending of populations. Modern features thus diffused out of Africa rather than being imposed through the invasion and dominance of dispersing moderns, and early moderns outside of Africa could therefore be expected to show features of the “natives” with whom they were mixing. And while the various models of human evolution were adjusting themselves to the post-mtDNA landscape, the genetic work itself was undergoing reevaluations.

I already mentioned the heavy criticisms of the 1987 “Eve” paper, from the point of view of the samples used, the methods of analysis, the rate of evolution, and the strong conclusions drawn. The team involved in the original work acknowledged that there were deficiencies, and, over the next few years, they set out to address the problems in a series of further analyses that served only to reinforce their conclusions, as we shall see in chapter 7. But as we shall also observe, most workers now agree that mtDNA, while very useful, is only one small part of the genetic evidence we need to reconstruct our evolutionary origins.

For the rest of this book, I will mainly be discussing three other human species along with our own: Homo erectus, H. heidelbergensis, and H. neanderthalensis. So how do we recognize distinct human species in the fossil record and our own ancestors? Well, that is not a straightforward question, and specialists will give differing answers. (For example, as I explained earlier, multiregionalists often regard Homo sapiens as the only human species on Earth during the last million years, so species like Homo erectus and Homo heidelbergensis have no real meaning for them.) But for me, there are features in the skeleton that, taken together, can diagnose distinct human species in the past, and that similarly characterize our species today. Because of variations in time and space, these features are rarely absolute, but in combination I think they can distinguish separate evolutionary lines that we can call species, based on their skeletal structure.

For our own species, Homo sapiens (modern humans), these features include: a large brain volume; neurocranial globularity (the curvature and doming of the bones of the braincase, and its increased height); in rear view a braincase that is wider at the top and narrower at the base; a higher and more evenly arched temporal bone at the side of the braincase; decreased height of the face and its tucking-in under the braincase; a small and divided brow ridge; a narrow area of bone between the eye sockets; increased projection of the middle of the face and nose; a bony chin on the lower jaw, present even in infants; simplification and shrinkage of tooth crowns; a lightly built tympanic bone (this contains the ear bones); a short pubic ramus that is nearly circular in cross section (this is a bone at the front of our pelvis); no iliac pillar (this is a near-vertical ridge of bone reinforcing the pelvis, above the hip socket); and femora (thighbones) that are oval in cross section and thickened most at the front and back.

In contrast, for Homo erectus, the human species that had appeared in Africa and Asia more than 1.5 million years ago, the characteristics included a small average brain volume; a relatively long and low braincase, narrow across the top but broad across the base; a lower and more triangular temporal bone; an angled occipital bone at the back of the skull, with a strong torus (ridge of bone) across it; bony ridges that reinforce the frontal and parietal bones of the braincase; a thick tympanic bone; a strong and continuous supraorbital torus (brow ridge); a strong postorbital constriction (the skull is pinched in behind the brow ridge when viewed from above); a wide area of bone between the eye sockets; a face that juts out from the braincase; a flatter and elongated superior pubic ramus; an iliac pillar; and femora that are rounded and evenly thickened in cross section.

Homo erectus seems primitive in many respects by the standards of later humans, but it represented a benchmark of change to the human condition in many aspects of its skeleton: a brain size beyond any ape or australopithecine, a human face with projecting nasal bones, small teeth, a humanlike posture for the skull, and a body frame of human rather than apelike proportions. The evolutionary biologists Dennis Bramble and Daniel Lieberman believe that erectus had made a fundamental transition to life in the open, first scavenging and then hunting over long distances. We are unique among primates in our capacity for endurance running, which may first have evolved to allow humans to get to carcasses for scavenging ahead of the competition. And people like the San today are able to gradually wear down their prey through persistent pursuit: ungulates, for example, can run much faster than humans over short distances but completely exhaust themselves over long distances, at which point they are easy to dispatch. Features of the erectus (and later human) skeleton in body shape, legs, ankles and feet, head balance and stability, and our reliance on sweating to thermoregulate could all be relics of an early adaptation to sustained running, according to Bramble and Lieberman.

H. heidelbergensis, present in Africa and Europe more than 500,000 years ago, shows combinations of features found in the more primitive erectus fossils and those found in the later Neanderthal and modern sapiens fossils, as befits a possible intermediate species: a brow ridge like erectus, but often filled with extensive sinuses (voids); an occipital bone like that of erectus; a wide interorbital breadth like erectus; a superior pubic ramus like erectus; an iliac pillar like erectus; rounder femora like erectus; brain volumes that overlap the smaller values of erectus and the larger ones of H. sapiens and H. neanderthalensis; a braincase higher than erectus, and parallel-sided in rear view; a face intermediate between erectus and later humans in its overall projection from the braincase; a temporal bone more like those of sapiens and neanderthalensis; a tympanic like Neanderthals and moderns; increased projection of the middle of the face and nose (as in Neanderthals and moderns); and, in some cases, inflated cheekbones that retreat at the sides, like those of Neanderthals.

The Neanderthals are advanced humans and thus share features with both heidelbergensis and with us. Yet there are also some retained primitive traits and those that betoken a separate evolutionary pathway. They have an elongated superior pubic ramus like erectus and heidelbergensis; rounder femora like erectus and heidelbergensis; a large brain volume like ours; a high and arched temporal bone like ours; reduced interorbital breadth; reduced total facial projection; a lightly built tympanic; in many Neanderthals, simplification and shrinkage of tooth crowns as in sapiens; weak or absent iliac pillar.

Then there are the features that seem to distinguish the Neanderthals as an evolutionary lineage. Some of these are concerned with a distinctive body shape, rib cage, and limb proportions, but the clearest ones are on the skull: a double-arched brow ridge with central sinuses; a double-arched but small occipital torus with a central pit (the suprainiac fossa); a spherical vault shape in rear view; distinctive shape of the semicircular canals of the inner ear (see chapter 3); strong midfacial projection and cheekbones that are inflated and retreat at the sides; a high, wide, and projecting nose; large and nearly circular orbits; a high but relatively narrow face; and enlarged front teeth (incisors), which are hollowed (shoveled) on the inside surfaces of the upper centrals.

(Clockwise from top left) Skulls of erectus (Sangiran, Java), heidelbergensis (Broken Hill, Zambia), sapiens (Indonesia), and neanderthalensis (La Ferrassie, France).

The feature that stands out (literally) in these comparisons of modern and archaic species is the strong brow ridge of the latter, and its absence in the former. The anatomist Hermann Schaaffhausen, one of the first describers of the original Neanderthal skull, called its strong brows “a most remarkable peculiarity,” and although there have been many scientific hypotheses to explain their presence or absence, none really convince me. The fact that many of the huge brow ridges in fossils are hollowed inside, with large sinuses (air spaces), suggests that they are not there to bear or transmit physical forces from blows to the head or heavy chewing. The eccentric anthropologist Grover Krantz even strapped on a replica brow ridge from a Homo erectus skull for six months to investigate its possible benefits, finding that it shaded his eyes from the sun, kept his long hair from his eyes when he was running, and also scared people out of their wits on dark nights. For me, that last clue might be significant, and like the paleontologist Björn Kurtén I think it may even have had a signaling effect in ancient humans, accentuating aggressive stares, especially in men. Thus its large size could have been sexually selected through the generations, a bit like antlers in deer. But if that is so, why don’t we have large brows like our predecessors? Well, I think the rest of this book will show that modern humans have developed so many other ways to impress each other, from weapons to bling, that perhaps the selective benefits of large brows wore off in the last 200,000 years.

(Clockwise from top left) Side view of skulls of erectus (Sangiran, Java), heidelbergensis (Broken Hill, Zambia), sapiens (Indonesia), and neanderthalensis (La Ferrassie, France).

If there were, in fact, different human species in the past, could they have interbred? In my view, RAO has never precluded interbreeding between modern and archaic people during the dispersal phase of modern humans from Africa. This is undoubtedly one of the main areas of confusion in studies of modern human origins: how to recognize species in the fossil record, and what this signifies. Some researchers argue that many distinct morphological groups in the fossil record warrant specific recognition, with the existence of at least ten such species of the genus Homo during the last 2 million years (that is, Homo ergaster, erectus, georgicus, antecessor, heidelbergensis, rhodesiensis, helmei, floresiensis, neanderthalensis, sapiens).

At the other extreme, some multiregionalists argue that only one species warrants recognition over that period: Homo sapiens. An additional complication is that different species concepts may become confused; for example, some multiregionalists have applied what is called the Biological Species Concept (BSC) to the fossil record to justify their belief that H. neanderthalensis and H. sapiens must have belonged to the same species and would have been fully interfertile. This concept, developed from the study of living organisms, argues that a species consists of the largest community of a group of plants or animals that breeds among itself, but not with any other community. Thus it is “reproductively isolated” with reference to other species, but its own varieties can interbreed with each other. Living Homo sapiens would be a good example of this, since people from all over the world are potentially able to mate and have fertile children, but we are apparently reproductively quite distinct from our ape cousins. I say “apparently” because there are persistent rumors that in the 1940s and 1950s scientists in the United States and/or the Soviet Union conducted unethical experiments impregnating female chimpanzees with human sperm—the results of which, so the rumors go, have been suppressed.

And what if we could still meet a Neanderthal—could modern humans interbreed with one? First this brings up the potential conflict between the BSC (which relates to living species) and the completely different concepts that I just used to recognize species in the fossil record, such as the degree of variation in the skeleton. Using the latter measure (a morphological species concept based on what is preserved in the fossil record), I and many other anthropologists recognize the Neanderthals as specifically distinct from Homo sapiens. But there is a conflict at the heart of the BSC: the fact that many closely related mammal species can hybridize and may even produce fertile offspring. Examples are wolves and coyotes, bison and cows, chimps and bonobos, and many species of monkey. So we have to recognize that species concepts are humanly produced categories which may or may not always work when compared with the reality of nature. So in my view, even if there was Neanderthal-modern hybridization (and I will discuss that thorny question in chapter 7), it would not necessarily mean that Neanderthals belonged to the same species as us—it would depend on the scale and impact of the interbreeding.

Fossils—the relics of ancient species—sparked my interest in the distant past when I began collecting them as a boy, and they still fascinate me. But on their own they are just mineralized and inert bones and teeth. In the next two chapters I will show how a range of exciting new techniques are helping us return these inanimate fossils to their ancient environments and bring them back to life.

Copyright © 2012 by Chris Stringer