The Third Man

For over a century the low-browed Homo erectus has sparked scientific fascination about our origins--and not-so-scientific ramblings about the meaning of race.

By Matt Cartmill
Sep 1, 1997 5:00 AMNov 12, 2019 5:11 AM


Sign up for our email newsletter for the latest science news

Java is a big, densely populated tropical island northwest of Australia. To most of us, it’s just another word for coffee, but to scientists who study human origins, Java is more richly evocative as the first place ancient human fossils were discovered in the 1890s. Recently some of those dusty fossils have come off the museum shelves and, after more than a century, started to make headlines again. New techniques for dating fossils indicate that some of those Javanese remains are a lot older, and others are a lot younger, than we had thought. These new dates pose problems for all our big-picture theories about human evolution-- though the news is worse for some theories than for others.

The story of Java in paleoanthropology goes back to 1891, when a young Dutch army doctor named Eugène Dubois found the top of a simian- looking skull in a Javanese riverbank near the town of Trinil. At first he thought he had found a chimpanzee fossil. Like a chimpanzee, the Trinil creature had a low braincase with a steeply sloping forehead and big bony ridges over the eye sockets. But the braincase was a bit too large, and the forehead a bit too bulging, to belong to an ape.

The next year Dubois dug up a human-looking thighbone a few yards away from where he had found the skullcap. He put the two specimens together and announced that he had found the missing link between apes and humans: a primate that had long legs and an upright gait like ours, but whose brain was only about 50 percent larger than a big gorilla’s. Dubois named his creature Pithecanthropus erectus, meaning upright ape-man.

Other Dutch scientists went back to the area in 1931 and uncovered 11 more skulls from a site near the town of Ngandong, downriver from Trinil. The Ngandong skulls were younger than the Trinil skullcap, and somewhat less primitive-looking. Although they too had thick bones, big brows, and receding foreheads, their braincases were 21 percent bigger than Trinil’s, averaging around 1,100 cubic centimeters--which is at the low end of today’s normal human range.

Over the years, more Pithecanthropus fossils have emerged. We now have a dozen more-or-less fragmentary skulls, several lower jaws, and a number of smaller bits and loose teeth. The deposits they come from had been dated from more than a million years ago to around 700,000 years ago-- about midway through the Pleistocene Epoch (1.6 million to 10,000 years ago). Pithecanthropus skulls are found throughout the sequence. They didn’t change much through the years, although on average, the more ancient skulls have slightly smaller braincases.

Nowadays anthropologists consider Pithecanthropus a species in our own genus, Homo. We are Homo sapiens, and Pithecanthropus erectus is now Homo erectus. Experts argue over which species the Ngandong skulls belong to; and they argue even more vigorously about the precise relationship between the two species. Many scientists, from Dubois’s time to the present, have embraced Homo erectus as our direct ancestor. Others have dismissed it as merely a retarded cousin and have searched for more ancient people with bigger brains to put at the base of the human family tree.

The basic facts about human origins are pretty much agreed upon these days. Around 2.5 million years ago in Africa, Homo evolved from one of the smaller-brained, bipedal man-apes called Australopithecus. The first stone tools also show up in Africa around this time, and some researchers think the two events are connected. There were probably two species of early Homo--H. rudolfensis and H. habilis. The former had bigger brains than Australopithecus; the latter had smaller molars. These advanced traits suggest that one or both early Homo species were making those tools, since toolmaking takes brains, and using them takes some of the load off your teeth.

By 1.9 million years ago, these two species had been joined in Africa by a third: our old Javanese acquaintance, Homo erectus. The newcomer had a thoroughly human body build, with relatively long legs and short arms that made it look less apelike than previous hominids. H. erectus also had a bigger brain--around 900 cubic centimeters, versus 600 to 700 for the earlier Homo. But its skull is curiously brutish, with thickened braincase walls, massive browridges, and evidence of powerful neck muscles. All in all, H. erectus fits the familiar stereotype of the chinless, thickheaded, beetle-browed, bullnecked caveman. It must have been a fearsome competitor for the earlier Homo types, which disappeared some 300,000 years after erectus arrived on the scene.

After its competitors died out, African Homo erectus made two important breakthroughs. One was a new kind of stone technology. Earlier toolmakers had just banged a couple of pebbles together to get a short broken edge. The new tools were made more economically and artfully, by flaking big flat chips off boulders and then reworking the detached flakes, yielding a sharp edge all around.

The other breakthrough was moving out of Africa. By one million years ago, Homo erectus was in China and Asian Georgia. By 500,000 years ago, there were erectus-like populations throughout the Old World, from Germany to the Far East and down into Africa. These creatures were improvements on the original H. erectus, sporting enlarged, 1,200 cubic centimeter brains under their persistently low foreheads and thick skull bones. Some people call this new, improved model archaic Homo sapiens. Others call it advanced Homo erectus, and still others put it into a species of its own, Homo heidelbergensis. In Europe this intermediate type seems to have evolved into the distinctive-looking, bigger-brained Neanderthal. The first fossils recognized as modern Homo sapiens of our own sort, with proper foreheads and protruding chins, showed up in the Middle East around 90,000 years ago.

All the experts agree on the fundamentals of this story. But they disagree on what it all means. The simplest interpretation of the fossils is that all the Homo erectus populations, from Africa to Java, evolved together as a single entity into modern Homo sapiens. By this so-called regional-continuity interpretation, there’s no real distinction between sapiens and erectus, and Homo heidelbergensis is just a vague label for the populations in the middle of this process.

The other leading interpretation is the out-of-Africa theory, which sees human evolution as a series of two or three waves of advancement emanating from Africa. In this view, erectus populations were replaced by a wave of heidelbergensis populations, including their Neanderthal offshoot in Europe. All these were replaced in turn by a wave of fully modern Homo sapiens--with no interbreeding between the old natives and the new immigrants.

The out-of-Africa theory has some shortcomings. The biggest problem with it is that nobody can reliably distinguish all these supposed species--Homo erectus, heidelbergensis, neanderthalensis, and sapiens--from one another. There are a lot of fossils that straddle the lines between them, and no two experts agree on just where one species ends and another starts. But the theory also has some facts on its side. Several lines of evidence suggest that the genetic differences between human populations today date back no further than some 200,000 years. As many geneticists see it, their data just don’t fit the picture of a gradual, million-year-long evolution of erectus into sapiens throughout the whole Old World. To these researchers, the genetic facts suggest that modern populations (or at least modern genes) spread more recently from a single center, just as the out- of-Africa model would have it.

One piece of paleontological evidence for the out-of-Africa theory is that some archaic Homo populations seem to have lingered beyond their time, alongside more modern-looking people--implying that the two types weren’t interbreeding and therefore must have belonged to different species. And this brings us back to the new dates from Java.

Determining just how old the Javanese fossils are has always been a problem. Most of the early finds were unearthed by local laborers who were paid for each fossil brought in--and therefore had financial incentive to conceal the exact spot where they had struck pay dirt. As a result, none of the early finds can be placed exactly in the local stratigraphy. And even knowing exactly where a Javanese fossil came from is no guarantee of its age. When a river cuts through ancient deposits, the fossils that weather out of its banks can tumble downslope, fall into the river, and get reburied in new sediments. If you date them from those fresh sediments, you underestimate their age. All the Javanese fossils come from riverbank deposits, and a lot of them look as if they were knocked around for some time in a flowing river. If that’s indeed the case, they may originally have been buried and fossilized in sediments older than the ones they were found in.

In 1971, Garniss Curtis from the University of California at Berkeley put a date on some volcanic minerals found around one of the oldest Pithecanthropus fossils from Java, the skull of a child from a site called Mojokerto. He analyzed how the potassium in the rock had changed over time into argon and came up with a surprisingly ancient date--around 1.9 million years ago. In 1992, Curtis and his colleague Carl Swisher used a more sophisticated technique to date the minerals taken from inside the skull itself and got an age of 1.8 million years.

These dates are almost a million years older than most experts had expected. If the new dates are correct, Homo erectus appears in the fossil record of Java around the same time it first shows up in Africa. This is hard to square with any of our theories about early human evolution. If erectus evolved in Africa, why doesn’t it show up there before we encounter it in Java? And if erectus evolved somewhere in between Africa and Java and spread from there to reach both Africa and Java 1.9 million years ago, then why haven’t any earlier fossil hominids been found outside Africa?

Having dropped their Mojokerto bombshell, Curtis and Swisher went back to take a look at the other, upper end of the Javanese fossil record-- those later but still primitive-looking braincases from Ngandong. (The Ngandong deposits don’t contain the volcanic minerals needed for argon dating, so the researchers used other techniques. Buried teeth pick up uranium salts from groundwater, and by measuring the ratio of uranium to the products of its fission in tooth enamel, you can estimate how long the teeth have been buried.) For these fossils they obtained age estimates ranging from 56,000 to only 21,000 years ago.

These dates are stunningly late. By 20,000 years ago, Neanderthals were long gone and humans around the world were essentially indistinguishable from people living today. By 10,000 years ago, people in the Middle East and Southeast Asia were starting to experiment with agriculture. But if the new dates for Ngandong are correct, a lonely surviving population of backward, slope-headed creatures that most experts would call Homo erectus was still clinging to a Lower Pleistocene way of life in Java long after the last continental glaciers had melted and everybody else had straightened up their foreheads and started painting pictures on cave walls. Like any other date from Java, these new dates are subject to various doubts; but if Curtis and Swisher are right, it’s a blow to the regional-continuity theory.

The blow isn’t necessarily a crushing one. Pleistocene Java might be the Land That Time Forgot: a primitive pocket, colonized early on and then cut off from all the currents of gene flow that were steadily transforming the rest of erectus into modern sapiens. One reason for thinking that Java might have been a peculiar backwater is that there are almost no stone tools in the erectus deposits. Everywhere else in the world where we find Homo fossils, they’re far outnumbered by the stone tools found alongside them. But the only tools ever found with Javanese erectus are a few dubiously associated flakes and hammerstones. Some scientists argue that erectus didn’t need stone tools in Java because they could make all the knives, spears, and scrapers they needed out of the flinty stems of bamboo. But then what were they using to cut the bamboo?

Why do anthropologists get so excited about these questions? Why should any of us care whether we’re descended from late-surviving archaic humans, or from equally primitive types who lived somewhat earlier and somewhere else? The electricity surrounding these issues flows partly from the clash of scientific egos and partly from the sheer fascination of stories about things long ago and far away. But it also flows out of the long, sordid history of scientific racism.

The study of human evolution has been contaminated with the politics of race from its very beginnings. In the late 1800s it seemed pretty clear to most educated Europeans that white folks were taking over the world because they were better than everybody else. Naturally, Darwinians saw this as the survival of the fittest. Many scientists regarded the native peoples of colonial Africa, Australia, and the New World as living fossils: leftovers from earlier stages in human evolution, doomed by Nature’s iron laws of competition to go down to extinction with the Tasmanian wolf and the Tasmanian aborigines.

It was easy to turn this racist thought the other way around and apply it to the fossils. If the Tasmanian natives had been exterminated because they were too low and primitive to compete, then still lower and more primitive extinct forms like Pithecanthropus must have been wiped out because they couldn’t compete, either. And whoever wiped them out must have been the ancestors of modern humans. Therefore we can’t be descended from Pithecanthropus. Nobody ever laid this argument out quite so nakedly as that, but a lot of experts were thinking along similar lines. Down through the late 1940s, leading textbooks of human evolution portrayed almost all the Pleistocene human fossils as doomed, blind-alley offshoots of the mainstream (white European) human lineage. Some depicted the tree of human evolution as fir-shaped, with a thick central trunk leading to the Europeans, and a lot of mostly extinct side branches coming off: first Pithecanthropus (which dies out), then the Neanderthals (which die out), and then the native Australians and Africans, who haven’t died out--yet.

After World War II, in the aftermath of Nazi horror, anthropologists hastened to throw out all these ideas about racial hierarchy. The collapse of the European empires in Africa and Asia also helped cure white scientists of the old habit of thinking of themselves as members of the master race. By 1960, no reputable anthropologists were still talking about racial differences in evolutionary status.

The old hierarchical interpretation of the fossil record--and the prehistoric genocides implicit in it--got thrown out as well. During the 1950s and 1960s it came to be regarded as vaguely racist to exclude any fossil humans from our own ancestry. All the taxonomic doors flew open. Australopithecus joined the human family, Pithecanthropus was welcomed into the genus Homo, and the Neanderthals were recognized as just an extreme racial variant of modern Homo sapiens. And if the low-browed Neanderthals were the ancestors and equals of today’s Europeans, it was ludicrous to think of assigning a lower status to the high-browed modern natives of Africa, Australia, or America. Scientific racism looked like a thing of the past.

It wasn’t. In his 1962 book The Origin of Races, the American anthropologist Carleton Coon came along and turned all this blossoming egalitarianism upside down. Coon accepted regional continuity and agreed that all the erectus fossils were our ancestors. But by juggling his dates and his taxonomy, he made out that the transition from erectus to sapiens had happened at different times for different races. Not surprisingly, he thought that whites had gotten sapienized first, then the Oriental peoples (probably by gene flow from Europe), and finally the Africans and Australians out on the periphery of Eurasia, who were the last people to become fully human. Coon hinted that these dark-skinned latecomers are still a bit retarded when compared with whites and Asians. His book features side-by-side photos of a small-headed Native Australian woman and a big-domed Chinese sage, captioned The Alpha and Omega of Homo sapiens.

Coon’s book aroused furious controversy. Dismayed by the racial hierarchy implicit in his model, many anthropologists attacked it as a theoretical impossibility. They insisted that a single interbreeding population couldn’t belong to one species at one end and a different species at the other. (In fact, several examples of this sort of arrangement have been described in some species of birds, salamanders, and other animals, although scientists argue about the truth of these descriptions.) The upshot of the controversy over Coon’s book was that some anthropologists began to view the whole regional-continuity model of human origins as a theory tainted by racism.

For this reason, some fossil experts hail every piece of evidence for the out-of-Africa theory as further proof of modern human equality. (If we’re all descended from an African Eve who lived just 100,000 years ago, we can’t be very different from one another, can we?) Others see the regional-continuity model, with its ages-long pattern of gene flow back and forth among all human populations, as a bulwark against racial typology. And some of us, who were trained back in the sixties and got all those old tree diagrams of racial divergence stuck in our heads, still have a vague feeling that real egalitarians shouldn’t discriminate against Neanderthals. (If Neanderthals are fully human, the differences between modern races are too trivial to bother about, aren’t they?) This debate may be more vehement than it needs to be because partisans on each side see themselves as defending the unity and equality of the Family of Man against attacks by the opposing camp.

The scientific jury is still out on those dates from Java. But while we wait for fresh evidence, we would do well to remember that the equality of today’s human beings is not really on the line. We are what we are, not what our ancestors were. If my grandmother had type A blood, that doesn’t mean I have it or am even carrying a gene for the A blood group. And even if my great-great-to-the-nth-power grandfather 30,000 years ago was a Neanderthal, that doesn’t change the slope of my own forehead by a single degree or imply that I’m carrying some taint of Neanderthal blood. The truth of racial egalitarianism hinges on the facts about living people. Their genealogies are irrelevant.

Understanding these simple truths might help alleviate some of the uneasiness many people feel about human evolution. A lot of that uneasiness springs from a mistaken notion that deep down, underneath all the cultural varnish, we are still what our ancestors were--that if we’re descended from apes, we must somehow be apes, and have license to behave like apes whenever we feel like it. We aren’t, and we don’t. Knowing where we come from doesn’t tell us where we are right now. If we can all accept that principle, it may help make future debates about human origins less heated and more illuminating.

1 free article left
Want More? Get unlimited access for as low as $1.99/month

Already a subscriber?

Register or Log In

1 free articleSubscribe
Discover Magazine Logo
Want more?

Keep reading for as low as $1.99!


Already a subscriber?

Register or Log In

More From Discover
Recommendations From Our Store
Shop Now
Stay Curious
Our List

Sign up for our weekly science updates.

To The Magazine

Save up to 40% off the cover price when you subscribe to Discover magazine.

Copyright © 2023 Kalmbach Media Co.