By Josh Fischman
The results of a strange experiment are locked in a room on an upper floor of a nondescript building at the University of the Witwatersrand, in Johannesburg. Behind the door lie two heavy safes. Within them sit several Pelican cases, the kind used by people who don't want you to get your hands on what's inside. And nestled inside, snug in foam padding, are bones that appear to be part ape and part human.
These two-million-year-old fossils seem to be remnants of an evolutionary attempt to move from the apelike forms of our early ancestors into the first species enough like us to begin our own genus, Homo. They are not human but could mark the gateway to humanity.
On Thursday five papers published in the journal Sciencedescribe the features behind that claim: a brain barely larger than a chimpanzee's but with signs of neural reorganization; a pelvis of a bipedal animal placed over feet with narrow, apelike heels; and a hand with a toolmaker's grip. "It's got this odd combination of features that run from head to foot," says Lee R. Berger, the professor of anthropology at the university who found the bones in an eroded cave called Malapa, located in a region northwest of the city that's called, fittingly, the Cradle of Humankind.
The cave has yielded bones from at least seven individuals in the species, known asAustralopithecus sediba. And one of the things that's got anthropologists buzzing is that, in a discipline where scientists make do with isolated fragments, these include largely complete skeletons. "By having complete things, we've changed the field," says John Hawks, an assistant professor of anthropology at the University of Wisconsin at Madison who is working with Mr. Berger. "You can't use one piece anymore to say, 'We know how this organism works.' You can't play the fragment game."
Statements like that have anthropologists buzzing as well, but not with approval. Though they are unwilling to say so publicly, researchers who have worked with isolated bones say that Mr. Berger is overreaching and his science is underwhelming. One anthropologist complains that the brain reconstruction, for instance, "is little more than hand-waving." Several critics point out that sediba, dated at 1.977 million years old in one of the new papers, is 300,000 years younger than some single bones that have already been ascribed to Homo. So how could sediba give rise to a genus that was around long before sediba lived? Mr. Berger responds that, first, those older bones are too small and scrappy to be reliably called any species, and second, "this is why we published our papers. Let's let the fossils speak for themselves."

What the Bones Say

To help them testify, one of Mr. Berger's colleagues has tried to bring the long-vanished sedibabrain back to life in a computer. Though the organ is gone, its pressure against the inside of the skull of a young male found at Malapa left an impression of its contours. Kristian Carlson, a senior researcher at Wits, as the Johannesburg university is known, worked with an ultrahigh-resolution scanner at the European Synchrotron Radiation Facility, in Grenoble, France, to transform those bumps and furrows into a detailed cyber-reconstruction. It shows, like other parts of sediba, a combination of primitive features and suggestions of modern ones.
The most obvious primitive element is size. At 420 cubic centimeters, it's barely bigger than a chimp brain and comparable to other australopithecines—similar to the famous "Lucy"—that lived more than a million years earlier.
But hints of modernity appear in the brain's shape, Mr. Carlson says. "Malapa has more frontal lobe than other australopithecines do," he says. In an area just behind and above the eyes, which in modern humans is associated with the ability to make long-term plans,sediba shows a bit of expansion. Another area that looks as if it might be changing and growing is in the region now called Broca's area—a place linked to language processing in modern humans.
Dean Falk, an anthropologist and specialist in brain evolution at the School for Advanced Research, in Santa Fe, N.M., cautions that bumps and furrows are a long way from functional brain anatomy. But Mr. Carlson is sticking to his guns. "I think it's fair to talk about the protuberances as an expansion that precedes actual neural change," he says. In other words, these areas may not have had the functions they do in modern humans, but the changes were laying the groundwork.
This mix of ancient and modern features repeats in the hand of a young woman found at Malapa, the most complete fossil hand found between 4.4 million years ago—that one belonged to one of our earliest nonape ancestors, Ardipithecus—and about 200,000 years ago, among the bones of our close cousins the Neanderthals. The sediba hand is missing only three wrist bones and a few fingertips.
What it shows, says Tracy L. Kivell, who led the team that studied it, is a mix of features, "some that are remarkably derived, or humanlike, while others are more primitive than we might expect." Indeed, says Ms. Kivell, a researcher at the Max Planck Institute for Evolutionary Anthropology, in Leipzig, Germany, "if we had found these bones separately, we may not have thought to put them together into the same hand!"
She ticks off the modern-looking elements. The thumb is relatively long when held against the fingers, which themselves are rather straight. "That makes it easier to pinch the finger pads together in a grip," Ms. Kivell says. Chimps have shorter thumbs and curved fingers, good for hanging from tree branches. And the bone below the thumb also resembles a human form. On the other side of the ledger—and the other side of the hand, closer to the pinky—the bones look more apelike. And the fingers look as if they had strong muscles for curling around branches but not necessarily for pressing sideways in a pinch.

Toward Tools

It's hard to put all this together into a meaningful picture of behavior. To add to the confusion, the hand of another fossil species—called Homo habilis, or "handy man," because it was the first to be found near stone tools—is slightly younger than sediba, at 1.8 million years old, but has more curved fingers (remember the tree-swinging).
No stone tools have been linked to the sediba fossils. But "it has the precision grip needed for toolmaking," Ms. Kivell says, "because it seems able to press the thumb against the fingers to hold things." Homo habilis would have been able to do this too, just in a different manner. "So we may have two hominins that were making tools with different anatomy. We're not saying sedibawas the first toolmaker. It seems like there were lots of experiments in toolmaking around this time," she says.
Precision grips, however, can be overrated, cautions Mary W. Marzke, an emerita professor at Arizona State University who has spent her career studying hands and tool manufacture. "You need strength, too. Baboons can pinch their fingers and thumbs together to pick things up. They just don't have strong ones. And if you don't have a strong grip on a stone, it will slip out of your hands when you knock it against another stone to shape it," she says.
Sediba may not have had that forceful grip. "There's no question they could have used stone tools," Ms. Marzke says, "but they may not have been that good at it."
Of course, early experiments rarely produce results that are as good as the later ones.