The plot to kill Homo habilis

A group of researchers want a new name for the 53-year-old species. Is it time to cast out habilis from our ancestry?

Naming things in human evolution is like the wildebeest who have to cross a crocodile-infested river. All of them would like to be on the other side, but no one wants to jump first.

Wildebeest crossing Mara River. Photo: Richard Toller CC-BY-ND 2.0

A handful of scientists are agitating to change the name of the venerable Homo habilis, named in 1964 by Louis Leakey, Phillip Tobias, and John Napier. Like wildebeest, they’re snorting and stamping around on the riverbank.

But they haven’t jumped.


Earth magazine published an article last year by Mary Caperton Morton about the semantics of naming genera: “Redefining Homo: Does our family tree need more branches?” The article came out in August but popped up on my Facebook feed just recently. It features several scientists who want to cast habilis out of our genus, including Bernard Wood, Ian Tattersall, Jeffrey Schwartz, and others.

If scientists were going to rebrand Homo, who would be “in” and who would be “out”? “I would allow Homo erectus in, but I would draw the line at Homo habilis,” Wood says.
The OH 7 mandible, type specimen of Homo habilis. The jaw is distorted and belongs to a juvenile. Fred Spoor and colleagues reconsturcted it by virtual methods, finding it has a surprisingly primitive shape, different from H. erectus and other members of our genus.

To make the case that the Olduvai Gorge habilis fossils belonged in Homo, Leakey, Tobias, and Napier had focused on the teeth and jaws, a bit more erectus-like than the more primitive Australopithecus africanus. They had partial skulls as well, and although the brain sizes were smaller than any erectus skull then known, they seemed a bit bigger than the small-brained africanus. With a partial hand skeleton and stone tools nearby, they had enough for the textbooks to enshrine it as ancestor: habilis begat erectus which begat sapiens.

But by the 1990s, anthropologists couldn’t ignore the small size of supposed habilis body fragments — especially the fragmented but clearly tiny OH 62 skeleton discovered by Tim White and Donald Johanson at Olduvai in 1986. Around the same time, Bernard Wood began to argue that the largest and most complete skull attributed to habilis, the KNM-ER 1470 skull from Koobi Fora, Kenya, should really belong to a different species entirely, Homo rudolfensis. In light of these changes, many scientists began to revisit the old idea that habilis might not be so different from Australopithecus after all.

More recently, fossil discoveries like Australopithecus sediba, Homo naledi, and Homo floresiensis have added whole skeletons of hominins who have skull shapes and teeth a lot like primitive Homo, but smaller brains than habilis. Together with a broader array of evidence attributed to Homo erectus, especially from Dmanisi, Georgia, such new fossils present a different story about the relationships within our genus.

Here’s one tree that shows how the picture might be changing. Mana Dembo and colleagues in a paper last year found this to be the best summary of relationships among species, based upon the anatomy of their skulls and teeth:

Everything bracketed between the blue line leading to H. floresiensis and our own species, H. sapiens, seems to belong to a single branch, or clade, with a single common ancestor. But that Homo clade includes many names intervening between habilis, erectus, and sapiens, including one — Au. sediba — that doesn’t seem to belong. Biologists try to make sure that each genus is a clade, and so if these relationships are correct, Australopithecus sediba probably ought to be recognized as Homo sediba. That was one suggestion that Dembo and her colleagues put forward in an earlier paper, in 2015.

The MH2 and MH1 skeletons of Au. sediba have long arms and primitive configurations of the femoral neck, heel, and shoulder. But they have humanlike features in their jaws, pelves, and hands.

But this tree, based only on skulls and teeth, may not be an accurate picture of relationships. Looking only at its jaw, Au. sediba may appear to belong in Homo. It even might be a sister species of habilis as the tree shows. Still, the rest of its skeleton mixes parts like short legs, long arms, and chimplike heels that don’t fit well in our genus. The mosaic anatomy of species like Au. sediba and H. naledi doesn’t fit the same picture that fragmentary jaws and teeth from other species like habilis and rudolfensis had painted.

Making Homo more inclusive by adding sediba would identify our genus with only a handful of characteristics related to jaws and teeth. That’s not satisfying to many anthropologists, who think aspects of brain shape, body shape, and locomotion were important to the way ancient humans adapted to their environments.

So scientists like Bernard Wood and Ian Tattersall favor a more radical redefinition. They want to limit Homo to the green branches on this tree. Homo erectus would be “in”, but habilis would be “out”.

Putting species into Australopithecus, here represented by the yellow branches, makes a problem worse. There’s no way to connect all those species on a single branch, without leaving out branches like Homo, Paranthropus, and Kenyanthropus.

It’s not a new idea. Eighteen years ago, working with Mark Collard, Wood suggested removing both habilis and rudolfensis from Homo and putting them into the genus Australopithecus. But that option won’t work today, at least not if we want a genus to include only species within one clade on the tree. Maybe habilis really is linked to sediba, but Australopithecus africanus doesn’t seem to be part of a clade with either of them. (Following the same logic has led some scientists to propose removing Lucy’s species, Au. afarensis, from Australopithecus.)

Writing in 2014, Wood offered the opinion that a new genus name for habilis may be the right solution:

My sense is that handy man should belong to its own genus, neither australopith nor human.

Tattersall echoed the same line as quoted in Caperton Morton’s article. Also in that article, Mark Collard suggests floresiensis is equally a problem:

“It’s really problematic, assigning those hobbits to Homo, when we’re talking about something that’s a meter tall, with a tiny brain by hominin standards,” Collard says. “It’s very difficult to see how that doesn’t stretch the boundaries of the genus Homo beyond what should be reasonable.”

For researchers like Wood and Collard, brain size and body size are fundamental to the “adaptive grade” that helps to define Homo. Fossil remains like the LB1 partial skeleton of H. floresiensis or the OH 62 partial skeleton attributed to H. habilis don’t seem to reflect the same way of life as humans. Even if you can draw a clade that includes them together with us, for some scientists they don’t seem human enough to make the grade.


Could it be that someone has already tried to rename habilis and peer review has so far kept the evil plot at bay? A taxonomic name can be published in any peer-reviewed journal — Au. afarensis famously appeared first in Kirtlandia. But you can sort of bet that the kind of scientists who want to depose habilis from Homo have Nature or Science as their aim. The shade of Phillip Tobias should be looming Raven-like over their editors.

It doesn’t take a taxonomist to see how this game ends once it starts. If we draw the boundary of Homo at erectus, then every species that branches outside that clade may soon find a scientist brandishing a new genus name. Habilis, rudolfensis, floresiensis, sediba, afarensis, garhi, anamensis, every one of them can plausibly be a genus depending on which tree of relationships you accept.

Once someone tries with habilis, all the other wildebeest will jump.


In my opinion, this is a bad moment in history to change the names of old fragmentary fossils. New, more complete, evidence is growing faster than ever before. Fieldwork is turning up new species and populations whose existence no one guessed at. The anatomy of relatively complete skeletal remains defies any simple phylogenetic scheme and suggests that parallelism or genetic exchanges were part of the evolution of most hominin species. Ancient DNA and modern genomes now tell us that Pleistocene species and populations exchanged genes in unexpected patterns.

It is not hard to imagine that habilis, rudolfensis, sediba, and erectus might have exchanged genes just as Neandertals, Denisovans, and some African archaic and modern human populations did. We may soon find ways to test such hypotheses for these ancient hominins.

Rushing to put these species into four different genera is madness.

Half these skeletons were described in the last seven years. Kadanuumuu’s tibia is not quite as long as that of the Turkana Boy skeleton, but is substantially longer than the D3901 tibia from Dmanisi, Georgia, likewise attributed to H. erectus.

Meanwhile, old ideas about “adaptive grade” can be surprisingly brittle in the face of new discoveries. The Kadanuumuu skeleton attributed to Au. afarensis, first unearthed in 2005 from Woranso-Mille, Ethiopia, has a large body and long legs that are very different from the tiny Lucy skeleton. Its describers have argued that the skeleton shares many aspects of the human body plan. If they are correct, such features are not special to “true” members of Homo like erectus.

As our team worked to describe the H. naledi material from the Rising Star cave, we thought carefully about what makes a species belong in Homo. The brain size of naledi is no bigger than any species of Australopithecus, and none of the naledi skulls is as large as estimated for OH 7, the type of H. habilis. But the naledi hand remains are more humanlike than the OH 7 hand, its foot and ankle are more humanlike than the OH 8 foot usually attributed to habilis, and the naledi legs are long. We concluded that naledi fits in the adaptive grade of Homo, defined by features related to manipulating objects, long-distance terrestrial bipedal locomotion, and chewing a generalized, high-nutrient-density diet. Brain size, a key feature of habilis, doesn’t seem to matter.

A cast of OH 24 from Olduvai Gorge, one of the most complete skulls attributed to Homo habilis. From the beginning, some scientists have maintained that it should be classified as Australopithecus.

There’s no question that such evidence is changing scientific ideas about Homo and its origin. That’s exactly why it would be a bad idea to publish a new genus name for habilis. With a better fossil record, we should be raising the standard to rely upon the whole body of evidence. It’s far too soon to be confident about the tree of relationships of hominins. We need to integrate the information from postcranial skeletons that have only recently emerged. Why in the world would we adopt a new name for old species right now, while the scientific process is still unfolding? And why would we select some of the most fragmented fossils as the basis of such a major revision of hominin taxonomy?

It’s the wrong time to rename habilis.