Herto fossils clarify modern human origins

Colin Groves explains the implications of the Herto fossils for human evolution and our concept of race.

The recent discovery in Ethiopia of the fossilised remains of two adults and a child by Tim White and his team at the University of California, Berkeley, have pushed back modern human origins to 160,000 years and put another nail in the coffin of the multiregional hypothesis of modern human evolution. But what does it really mean, and what is the context?

For about 30 years the origin of the modern human species has been the subject of much debate. We are divided into well-marked, if overlapping, geographic races:

* Caucasoid people in Europe, North Africa, the Middle East and the Indian subcontinent;

* Mongoloid people in eastern and South-East Asia, the Pacific and the Americas;

* Negroid people in Africa, south of the Sahara; and

* Australoid people in Australia and Melanesia.

Two competing models have attempted to explain the origin of these “major races”: Multiregionalism (or Regional Continuity) and Out-of-Africa (or Replacement).

Until somewhat after two million years before present (BP), proto- humans lived only in Africa, but thereafter they began to spread to other areas. By 1.7 million BP they were at “the gates of Europe” (Dmanisi, in Georgia), by 1.5 million BP they were in Java, and by 500,000 BP (or somewhat more) they were in China and in Europe.

But these were not anatomically modern humans, and their differences are dignified by a plethora of names. Homo ergaster is the African ancestral species, and the enigmatic Dmanisi fossils may also belong to that species. Later African and European fossils, between 600,000 and 300,000 BP, are referred to the species Homo heidelbergensis, and the Europeans who lived from then until some 30,000 BP are the famous Neandertal people (H. neanderthalensis). The Javan species (“Java Man”) is H. erectus, while the Chinese species (“Peking Man”) is either also included in H. erectus or else assigned a separate species, H. pekinensis.

For multiregionalists, each of the modern races is descended in part from one of these archaic regional forms the one living in, or close to, the region to which the modern race is indigenous. For example, they consider that Mongoloids differ from other modern people in some of the same respects that H. pekinensis differed from its contemporary archaics, while Australoids bear a special likeness to Javan H. erectus. At the same time, the model contends that there was gene flow between different regional populations, so that humanity evolved as a whole while each geographic segment retained its own racial features.

If this is how it happened it would be nonsense to divide the fossils into all those different species. All of them even the Neandertals – would rate as archaic H. sapiens.

But all those species are real for the Out-of-Africa school, and most of them are dead-ends. Java Man, Peking Man and the Neandertals arose, flourished and died out without leaving any descendants. The Out-of-Africa school contends that modern humans (H. sapiens) are descended uniquely from African H. heidelbergensis and spread around the globe, replacing the various archaic species that preceded them in different places.

The skull of Homo sapiens idaltu has placed modern human origins in Africa 160,000 years BP.

The skull of Homo sapiens idaltu has placed modern human origins in Africa 160,000 years BP.

Adherents of this school deny the existence of any features linking Peking Man and Mongoloids. For instance, multiregionalists say that shovel-shaped incisors are in common between them even though they are actually found in all the archaic species, not merely in H. pekinensis. The multiregionalists deny that large brow- ridges and a flat, receding forehead are peculiar links between Java Man and Australoids; actually, these features define almost all of the archaic species, including H. heidelbergensis.

When molecular genetics began to impinge on public consciousness in the mid-1980s, it was seen to be supporting the Out-of-Africa model to such an extent that, in the public mind, the debate was between “the fossils” and “the molecules”. This was too stark: it is true that mitochondrial DNA does point to a common human ancestor living in Africa 150-250,000 BP, but there was already a great deal of fossil evidence pointing to human origins in Africa.

Until this year, the earliest H. sapiens appeared to be a fragmentary skull from the Omo River in southern Ethiopia, which was dated (but very provisionally) at 130,000 BP. The next was the skull from Ngaloba in Tanzania, 120,000 BP. Then a set of bones, mainly mandibles, from Klasies River in South Africa dated between 120,000 and 80,000 BP.

Overlapping these at about 115,000 BP are some excellently preserved remains from Qafzeh and Skhul in Israel. While these were not in Africa, they were associated with typical African faunal remains. (It appears that from time to time in the past, as to some extent is the case today, the Levant was part of the African theatre.)

There are also several sets of remains that are anatomically intermediate between H. heidelbergensis and H. sapiens. All of them are from Africa: Guomde (272,000 or 279,000 BP), Florisbad (259,000 + or – 35,000 BP), and two skulls from Jebel Irhoud in Morocco. These are of uncertain date but are thought to be 150-180,000 BP. The more complete of the Irhoud skulls is “almost” H. sapiens, but just outside the modern range, which makes it especially unfortunate that the date is uncertain.

Now we have the new skulls from Herto, in central Ethiopia. There are three: a nearly complete adult, a less complete adult, and a child of 6-7 years of age. They date to 154-160,000 BP and, interestingly for human remains of this early date, they show evidence of sophisticated mortuary practices, including defleshing, while the heads had probably been disarticulated from the bodies, which were not present at the site.

The more complete adult skull (BOU-VP-16/1) is larger than most (and perhaps all) modern human skulls. It has a deep face, very large brow ridges, and a rugged line (the superior nuchal line) demarcating the limit of attachment of the postural muscles. In these features it is said to be outside the range of modern humans as well as of the Skhul and Qafzeh specimens. White and his colleagues describe the Herto type as a separate subspecies, H. sapiens idaltu (“idaltu” means “elder” in the Afar language). The describers did not make any special reference to the Jebel Irhoud skulls but, from the measurements and photos, BOU-VP-16/1 seems remarkably similar to Irhoud 1.

The implications of the Herto find for modern human origins are clear. Here were H. sapiens, more primitive than anyone now living but recognisably members of our own species, living in north- eastern Africa at a time when the Neandertal people were in sole occupation of Europe. Even later than Herto, the only people for whom we have evidence were still non-modern – an enigmatic Neandertal-like skull from Maba in China, and late H. erectus in Java. Just as predicted by the Out-of-Africa model, modern humans appear in Africa long before they are known from anywhere else.

There are implications for the origins of modern races, too. Herto (and Jebel Irhoud) are H. sapiens, but with primitive features. They are not, racially speaking, Africans. The later Omo and Klasies remains are more modern, but they too are archaic, and certainly show no traces of the features that characterise any modern races. Only Qafzeh and Skhul seem to lack these primitive features, and rate as “generalised modern humans”. Our species seems to have existed as an entity long, long before it began to spread outside Africa or the Middle East, let alone split into geographic races.

When, then, did H. sapiens begin to split into races? The evidence indicates that modern racial features developed only gradually in each geographic area. The earliest H. sapiens specimen outside the Africa/Levant region is from Liujiang in China, whose dating was recently confirmed at 67,000 BP by a group led by Guanjun Shen of Nanjing Normal University. Like Qafzeh and Skhul, Liujiang is a “generalised modern”; it has no Mongoloid features.

The East Asian fossil record is not good enough to show when Mongoloid features began to develop. All we can say is that they must have developed before the end of the Pleistocene (12,000 BP) because this is when people began to cross what is now the Bering Strait (which was then a land-bridge); and Native Americans are Mongoloid.

H. sapiens began to enter Europe about 40,000 BP, but it is only at 28,000 BP that we get a fossil that shows any Caucasoid features – the Old Man from Cro-Magnon, in France.

Florisbad (259,000 + or – 35,000 BP) is anatomically intermediate.

Qafzeh (115,000 BP) is a “generalised modern human”.

Within the African homeland, the appearance of Negroid features is debatable. The skull from Border Cave, on the South Africa/ Swaziland border, may be 60,000 years old and may show Negroid features, but both claims have been challenged.

And Australia? The earliest widely accepted dates for human occupation are of the order of 60,000 BP, not more, according to Bert Roberts of La Trobe University and the late Rhys Jones of the Australian National University. The claim that the Mungo Man skeleton is 62,000 BP has recently been challenged. \According to a recent study led by Jim Bowler of Melbourne University, both Mungo Man and Mungo Woman may be only 40,000 years old (AS, April 2003, pp. 18-21), but they are still the earliest skeletal remains we have from Australia. Are they Australoid?

Of all “major races”, Australoids have evidently changed least from the generalised modern human pattern, but the flat, receding forehead and angular skull vault that characterise many full- blooded Aboriginal people today are somewhat different to the Qafzeh/ Skhul pattern. A 1999 study by Susan Anton and Karen Weinstein of the University of Florida, in the process of confirming that some of the Australian fossils (including most of the famous Kow Swamp series) had undergone artificial head deformation in infancy, found unexpectedly that most of the Pleistocene fossil Australian crania are rounder-skulled than modern ones. So racial features developed late in this part of the world, too.

In summary, the new discovery at Herto does not shatter any myths, but it extends the dataset, shifts the weight of evidence yet more decisively in favour of the Out-of-Africa model of modern human evolution, and helps to place modern racial variation very firmly into context.


How many human races are there? If we look at the major geographically varying characters – hair form, skin colour, body build, facial features, and some cranial and dental features – there are four wide areas over which at least some of these characters vary more or less concordantly. These are:

* Sub-Saharan Africa, where people have “woolly” hair and tend to have elongated limbs, a wide and flat nose, and subnasal prognathism (the jaws project below the nose);

* Europe, North Africa, western Asia and the Indian subcontinent, whose people have wavy or somewhat curly hair, sharp facial features (especially a narrow, prominent nose), and abundant facial and bodily hair;

* Eastern Asia, the Pacific and the Americas, where indigenous people have straight hair, a yellow tinge to the skin, facial flatness (flat nose and forward-standing cheekbones), so-called shovel-shaped incisors, and short limbs; and

* Australia and Melanesia, whose indigenous people have very elongated limbs and prominent brow ridges.

Some anthropologists consider these to be “the major races”, and the terms Negroid (or Afrotropical), Caucasoid (Caucasian), Mongoloid and Australoid (or Austromelanesian) have been applied to them. To an extent these do represent recognisable geographic clusters whose skulls and dentitia are usually recognisable.

But each race is very heterogeneous. Skin colour plays a very minor role here: noticeably, the Bushmen of Namibia and Botswana are much lighter than most sub-Saharans, and some Indians (especially in the south) are as black as many Africans but are as Caucasian as John Howard.

The term “Caucasian” is widely misunderstood. Most westerners think it is a polite term for “white”. The term “Asian” should not be used in a racial sense – the Indian subcontinent and the Middle East are part of Asia, but Indians, Iranians and Arabs are Caucasians.

Colin Groves is professor of archaeology & anthropology at the Australian National University.

Copyright Control Publications Pty Ltd Aug 2003

Leave a Reply

Your email address will not be published. Required fields are marked *