NOT SO EXTINCT AFTER ALL

Paleontologists don't pretend to know everything about how Homo erectus lived, but it's a safe bet the ancient prehumans passed their days in unremarkable ways. Their language was probably little more than a system of gestures and grunts. Their diet, consisting of foraged fruits and crudely cooked animals, was not an easy one to force down--if the attachment points on their skulls for stout chewing muscles and their large front teeth suggest anything. Their skill in making tools was limited: a flaked stone or a crude ax was probably as good as it got. They arose in Africa about 1.8 million years ago and later ranged over an impressive expanse of land that included the Middle East, China, Southeast Asia and possibly Europe. But Homo erectus was no match for our species, the glamorous Homo sapiens, and not long after anatomically modern humans appeared some 200,000 years ago, the more primitive hominid died out.

Or so it seemed. According to research published last week in the journal Science, Homo erectus may not have gone so quietly into that good night. On the Indonesian island of Java, it now appears, a small group of hangers-on may have lived as recently as 27,000 years ago, thriving in a world that Homo sapiens had long before claimed as its own.

The study, based on a new analysis of fossil sites, has created a tempest in the paleontological community. Now researchers not only must explain how a single prehuman population could remain frozen in evolutionary amber for so long after its species went extinct elsewhere in the world, but also must revisit two of science's most hotly debated questions: Where on the habitable continents did modern humans first emerge, and how did they come to dominate the world? "These dates will stir up a lot of controversy," says geochronologist Carl Swisher of the Berkeley Geochronology Center in Berkeley, California, who headed the study. "Some people definitely won't believe them."

For fossil hunters studying evolution, Java has always been a good place to dig. Its equatorial climate makes it home to countless species, and periodic land bridges placed it in the middle of the migratory autobahn between Asia and Australia, making it the perfect spot to study how animals spread. Since the 1890s, numerous fossils of Homo erectus have been found on the island, but scientists were particularly intrigued by more than a dozen partial skulls found near the villages of Ngandong and Sambungmacan in the 1930s and 1970s. The skulls had unusually large braincases, and so were estimated to be anywhere from 100,000 to 400,000 years old, among the youngest Homo erectus remains ever found. For more than a generation, those estimates stood. Two years ago, however, a team headed by Swisher decided to find out just how old the fossils really were, using instruments that determine age by measuring radioactive decay in fossils that have absorbed uranium from the surrounding soil.

Unfortunately for Swisher and his group, the university on Java that is the custodian of the bones refused to allow chips to be taken from them for dating, for fear of destroying them. Instead, Swisher and his team traveled to Ngandong and Sambungmacan and unearthed animal teeth in the same stratum from which the bones were believed to have been taken, on the theory that remains found in the same spot should have been buried at the same time and should thus be the same age. When Swisher and his colleagues dated the teeth, they were stunned at the results: the animal remains were only 27,000 to 53,000 years old, suggesting that the hominid skulls were too. "We never expected to find dates this young," Swisher says.

That was an anthropological understatement. Finding Homo erectus bones that come from an era when only Homo sapiens should have been around is like finding a family of Neanderthals living in 1996. Not surprisingly, skeptics weighed in as soon as word of the discovery got out. "These dates must be wrong," says paleontologist John de Vos of Holland's National Museum of Natural History. "They just threw the fossils into a machine, and out rolled a date."

One thing that concerns scientists is that even if the hominid skulls and animal teeth wound up in the same place, there's no guarantee they started out there. Flooding, for example, could have washed the Homo erectus remains into a younger fossil bed. Swisher disagrees, arguing that the remains are too well preserved--its fragile structures are generally intact--to have been bumped around in a flood.

If the new dates are correct, it's not hard to explain how the Indonesian Homo erectus group held on for so long. Once Java's last land bridge disappeared, the island was separated from the Asian mainland by hundreds of miles; any population that took root there would thus have been well protected from interlopers. More troubling are the questions the new study raises about human evolution in general.

Paleontologists mapping the expansion of the species subscribe to one of two theories. The first, known as "out of Africa," postulates that human migration began about 1.8 million years ago, when Homo erectus left the African continent and colonized Europe, the Middle East and Asia. Roughly 1.6 million years later, anatomically modern Homo sapiens appeared in Africa, and eventually migrated out too, displacing the prehumans who preceded them. The second theory, known as multiregionalism, has no quarrel with the first wave of Homo erectus expansion, but argues that the subsequent Homo sapiens emigration never took place. Rather, it holds that Homo sapiens sprang up in many spots throughout the Old World, evolving from Homo erectus colonies already in place.

The new findings provide a boost to the out-of-Africa camp. If Homo erectus really became Homo sapiens in several places at once, they argue, why would such an evolutionary wave have simply skipped Java? What's more, multiregionalists have long contended that Javan Homo erectus gave rise to modern Australians, but if the older and newer species in fact overlapped, this becomes unlikely. "The multiregionalists will have to do some fast talking," says Homo erectus expert Philip Rightmire of the State University of New York at Binghamton. "The evidence they looked to in the Far East is falling away."

Not so, respond multiregionalists, who note that cranial capacities of Asian Homo erectus fossils sometimes vary by as much as a third. This, they say, is proof that the more primitive species was not simply supplanted by an advanced one but evolved slowly into one. Out-of-Africa proponents pass these differences off as little more than normal variations in skull size between one individual and another.

The out-of-Africa scenario was the more popular theory even before Swisher's team published its findings, and now it is only likely to find more adherents. Nonetheless, multiregionalists show no signs of folding their tents--and if anything, may have grown even more convinced of their position. "Homo erectus and Homo sapiens could never have lived together," De Vos maintains. Maybe not, counters Swisher, but for now the fossils say they did.