ON THE ANCESTOR TRAIL

The wonderful thing about human origins is that they change so fast. Although the scientists' models are increasingly sophisticated, they still resemble Zeppelins: vast quantities of theoretical gas in a thin envelope of data, floating above a tiny capsule of physical evidence. They may be punctured by a few chipped rocks, a single fossil tooth or a new laboratory measurement. Just when scientists think they have put together a theoretical structure that squares with the evidence, they are presented with a new piece that doesn't fit. It happens every few weeks. In pop music, it's the same old tunes; in the arts, it's irony as usual. But the action in human origins is like a new punk rock and a new Armoury Show every month.

Over the past year, there have been half a dozen dramatic announcements that promised to make humankind look different at a stroke. It was claimed that a pierced bone from Slovenia was a Neanderthal flute. Three spears from a mine in Germany, shaped and balanced like javelins, provided unprecedented evidence of technological sophistication in hominids 400,000 years ago. Spanish scientists christened a new hominid species, Homo antecessor. Meanwhile, other workers announced that they had successfully extracted DNA from the skeleton that gave the Neanderthals their name, proving that they were not the direct ancestors of modern Europeans.

Claims like these tend to get a mixed reception: wild enthusiasm from those whose ideas they support, scepticism from those whose perspectives they threaten. Neanderthals have a devoted band of supporters, particularly in the US, who are distinctly cool about Europeans who deny their Neanderthal heritage. Some of them have formed a Neandertal [sic] Anti-Defamation League (they decided that dropping the 'h' was palaeoanthropologically correct). While the spears command general respect, the jury is still out on the flute. Only a bold scholar, or a gambling one, would predict how any of these findings will look in a year's time.

The past year's two most extraordinary claims about early humanity came from Australia. Their fate, told in a Horizon film entitled "Out of Asia", is a cautionary tale. Richard Fullagar, a rising star in Australian archaeology, had measured the age of samples at Jinmium, a remote site in west Australia. In Aboriginal mythology, Jinmium is a female ancestor-being, and the spot is where she turned to stone after being caught by Djibigun, her male pursuer. Paintings and cup-like indentations on rocks affirm that the place has been important to the First Australians for a very long time. But scientists and the public alike were unprepared for the dates Fullagar and his colleagues calculated. The cup-marks, they reported, could be nearly 60,000 years old. That would make them twice as old as the most ancient European cave paintings.

Whether they should be regarded as art could be debated, but at least they could just about be squeezed into current models of Australian settlement. It had been accepted for some time that people had reached the continent by around 40,000 years ago, and the use of a dating technique known as thermoluminescence was pushing the estimated time of arrival back past 50,000 towards 60,000 years. But the other half of the Jinmium double whammy was the claim that humans were there over 116,000 years ago. The occupation span was doubled at a stroke.

In their formal publication, Fullagar and his colleagues emphasised that their thermoluminescence tests had not established the dates conclusively, and must be confirmed. But the paper appeared several months after the press had got wind of the findings, which were splashed in the Sydney Morning Herald. The archaeologists were forced, as Rhys Jones of the Australian National University puts it in the Horizon film, to ride the bronco. Other scientists were sceptical, and their doubts were confirmed when subsequent tests indicated that the dates were wrong.

One of the reasons why the claims attracted so much attention is that the question of original settlement is highly politicised in Australia. When recent Australians celebrated 200 years of occupation, Aborigines could retort that they had 40,000 years or more to commemorate. These millennia are of rhetorical rather than legal significance. It's not a question of precedence - the English would hardly be impressed if Italians claimed land rights on the grounds that the Romans were here before the Anglo- Saxons. But tens of thousands of years do imply a moral right to tenure.

This appropriation of scientific data represents a marked contrast with earlier times, when scientific claims did Aborigines no favours. Scientists noted similarities between the shapes of Aboriginal skulls and those of fossils from Java, assigned to the Homo erectus species. It was suggested that the long, low, thick crania, with their heavy brow-ridges, were the marks of descent from Homo erectus. Although nowadays anthropologists prefer to explain such features as environmental adaptations, like the stocky build of polar peoples, their predecessors regarded them as visible signs of internal primitiveness. In 1962, the American anthropologist Carleton Coon achieved notoriety for expressing such views just as the political climate was beginning to turn against them. His book The Origin of Races featured juxtaposed photos of an Aboriginal woman and a Chinese "sage". The former had a brain less than a litre in size, Coon claimed, while the latter was twice as large; they thus represented "the Alpha and Omega of Homo sapiens".

The after-effects of Coon's frank scientific racism continue to reverberate today. At opposite ends of the world, the Neanderthal question and the ancient Australian question form the two poles of the same controversy, that of modern human origins. One camp, whose most prominent member is Chris Stringer of the Natural History Museum in London, maintains that all modern humans are descended from a single population which emerged from Africa around 100,000 years ago. This population, says the "out-of- Africa" school, replaced all the other hominid populations that by then occupied the Old World. The modern populations did not interbreed with the older populations to any significant extent, so we do not carry any Neanderthal or Homo erectus genes. The Neanderthal DNA results delighted the Africa hands, since they supported the idea that, in Western Europe at least, Neanderthals and modern humans did not interbreed.

The other school of thought is known as multiregionalism, and its leading advocate is Milford Wolpoff of the University of Michigan. As Wolpoff and his wife, Rachel Caspari, recounted in a recent book, Race and Human Evolution, multi-regionalists have had to struggle to distance themselves from Coon's ideas. On top of that, the out-of-Africa model has been promoted to the public as the anti-racist one. "Human unity is no idle political slogan," wrote Stephen Jay Gould. "All modern humans form an entity united by physical bonds of descent from a recent African root."

Wolpoff and Caspari's book is an attempt to dislodge their opponents from the moral high ground. They place themselves in the tradition of Franz Weidenreich, who developed the multi- regional model in the 1940s. Weidenreich emphasised interbreeding and the flow of genes between populations. Coon abandoned this emphasis in favour of the implausible idea that different populations reached the modern human condition independently. Wolpoff and Caspari argue that in treating humanity as one great Internet of gene flow, multi-regionalism relegates features like brow ridges to the equivalent of regional accents.

The contest goes back and forth. New dates for Javan fossils, considered to be Homo erectus, suggest they may be younger than 30,000 years, which would not leave enough time for them to evolve into moderns. On the other hand, genetic studies by Rosalind Harding, based in Oxford, suggest that modern humans - including ones in the Oxford area - carry DNA sequences which evolved in Asia 200,000 years ago. That was long before modern humans emerged from Africa, according to the replacement scenario.

Despite the false dawn of Jinmium, the archaeology won't lie down either. The vegetation of northern Australia changed around 40,000 years ago, although the climate was stable then. The presence of charcoal suggests that the cause was fire. Perhaps this, rather than rock carvings, was the first mark humans made on Australia. And they may have done so much earlier. Similar evidence appears in a sample up to 150,000 years old, from the Great Barrier Reef. That implicates Homo erectus.

Although the Reef and large areas of the Australian continental shelf were dry land in the past, Australia was always at least 90km from the nearest island. According to multi-regionalist Alan Thorne, it would have taken eight or nine days to make the crossing by raft; and it would also have required language to organise the voyage. Multiregionalists tend to be generous in their estimates of ancient hominid capacities, but this is expansive even by their standards. It implies that erectus was capable not just of island-hopping, but of conceiving an invisible goal, beyond the horizon.

'Out of Asia' is on BBC2 this Thursday at 9.25pm