by Meave Leakey
Louise, newly pregnant with her second child, ran our fieldwork while enduring severe morning sickness. In contrast to when she ran the expedition for me in the aftermath of Richard’s plane crash when the only communication possible was by sporadic radio call, we were now equipped with e-mail and satellite phones. Samira was able to keep her updated on our medical progress, and I was able to follow the fieldwork in much more detail.
Within the space of just a few months, both of us were back on our feet again, and we returned to Kenya to resume our lives, once again thankful for the extraordinary feats that modern medicine can perform. Although Richard was officially “retired,” he still had many interests and projects. One of these was a burgeoning idea about how to help us to address the challenges thwarting our fieldwork at Ileret, which had been put on hold due to the very real question mark as to whether he would survive to execute it. When Louise and I asked if he could help us raise funds to build a new and independent research centre, his response was classic Richard: “Yes, but only if I can build something big!”
And thus the Turkana Basin Institute, an ambitious research and teaching facility, was born. The idea was to create an institute with permanent infrastructure that would enable year-round research rather than the piecemeal short seasons we had always relied on. Richard outlined his vision to Stony Brook University, and they enthusiastically endorsed the idea of TBI, committing funds and administrative support and becoming TBI’s main partner. We planned two centres on either side of the lake as well as an administrative office in Nairobi and a New York office hosted by Stony Brook.
Before long, we had constructed temporary facilities near Ileret that were conveniently close to the airstrip, and this was fully operational by the end of 2007. Construction of the first full field centre on the west side of the lake was completed in 2012, and thereafter the construction of permanent facilities began. At last, Richard and I could work together again on prehistory.
We had the capacity to conduct our own research year-round and host research projects by other scientists in fields that complemented our own interests. We also established a field school that welcomed international students who wanted to learn about human evolution, geology, and ecology in the field while earning credits towards their university degrees. At the same time, the field schools allowed us to invite Kenyan students on scholarship to foster local knowledge and talent in the earth sciences. With greater financial muscle and a permanent presence in the field, we have also built successful community projects that support health and education and employ some of the local youth whose resentment and hostility had been severely thwarting our work.
WITH OUR FIELDWORK up and running again under the auspices of TBI, we were finally able to turn our attention back to Kyalo’s skull. When Richard had first described the beautiful 992 mandible that Ngeneo had discovered, he noted its gracile slender form and its affinity “morphologically and in dental proportions” to at least some of the Australopithecus africanus specimens in South Africa. This would turn up a red herring: rather than comparing the mandible and the rest of the erectus collection to the Asian fossils, they were compared in detail to A. africanus by Colin Groves and Vratislav Mazák in 1975. Because the mandible did not resemble A. africanus that closely after all, the pair rather bizarrely and unthinkingly reclassified the entire East African collection of Homo erectus as a new species, which they named Homo ergaster (“working man”). Their argument for doing so never included a detailed study of the whole erectus sample, but because the mandible came to be viewed as a good example of East African H. erectus, the name stuck and came into common usage. I was aghast at the new name and longed to find the evidence to dispel the notion that the African erectus hominins were a different species than the Asian ones. This was my first order of business.
Fred Spoor, Louise, and I undertook a meaningful comparison of not only the African H. erectus/ergaster skulls but also the vast Asian collection of H. erectus that none of us were at all familiar with. We immediately realised that we were in over our heads, so we turned to Susan Antón, one of the very best in her field. Fred had known Susan for some time and had visited her at her farm in New Jersey that she shares with geophysicist Carl Swisher and a large family of canine friends. Not only is Susan a tremendously likeable person with a great sense of humour, she had also worked for many years in Indonesia and her forte is the Asian H. erectus. Susan readily agreed to join us in our study of Kyalo’s skull, ER 42700.
Our first task was to figure out if 42700 was an erectus/ergaster or some other early Homo. It was clear to us all that we needed to know just how old 42700 was when he or she died before any meaningful comparisons across the collections could be made. This would affect our interpretation of both the brain capacity and the morphology of the skull because both the brain and the brow ridges would be less developed in a subadult or juvenile. Without teeth, this task would have been all but impossible but for the beautifully preserved base of the skull. Unlike the bones in the skeleton, which fuse at very specific times in childhood development and give a good indicator of age, the degree of fusion of the skull bones is a more difficult indicator of age. Luckily for us, however, there is one area on the base of the skull that is a good yardstick, and this was immaculately preserved.
This small telltale area is on the part of the skull called the basioccipital, and it’s where two bones fuse together. One of these is called the sphenoid—a very important little bone right on the bottom of the skull that resembles a butterfly with outstretched wings and articulates with all the other bones in the skull to bind them together. The second is called the occipital, and it’s the bone that encompasses the bun at the back of the skull and curves round underneath it to include the foramen magnum. Spheno-occipital synchondrosis is the tongue-tangling name of the area of fusion between the two, and we can tell quite accurately how old a subadult is based on how fused they are.
With the naked eye, we could see quite clearly that the bones were not quite fully fused. The big question was: how much longer before they would have been? Our friends at Diagnostic Centre Kenya just up the road from the National Museum in Nairobi once again came to our rescue. In 2006, Fred and I worked on the skull when human patients had no need of the CT machine. Paying special attention to the spheno-occipital synchondrosis, we took a series of one-millimetre cross sections of the skull. Fred converted these data so we could view them on a computer screen. Then we stacked them together digitally, building up a 3-D picture of the whole cranial vault on the computer. When we enlarged these images and zoomed in on the spheno-occipital synchondrosis, we could see exactly at which point the fusion stopped. It was two-thirds complete when the individual died.
Homo erectus is found across most of the Old World, and yet, there are far more differences through time than physical space. Nevertheless, our study did identify key features that differentiate this species from Homo habilis.
The spheno-occipital synchondrosis is already 50 percent closed by the time humans reach puberty and the second molars are through and the wisdom teeth are not yet erupted. In chimpanzees, fusion is 50 percent complete after the wisdom teeth are already through. We don’t know whether H. erectus would have aged like a chimpanzee or a human (although we have hypothesised that it might have been halfway between), but this was enough to tell us that the individual was already a young adult or at the very least a late subadult and that its cranium would have already reached its adult size. Thus the diminutive skull held a fully grown brain that weighed in at a modest 690 cc. The spheno-occipital synchondrosis of the Dmanisi skull, which blew us away for its resemblance to 42700, is less fused than 42700’s but not by much, and its brain capacity would probably have increased slightly had it lived to be an adult. Knowing that 42700 was an adult, albeit a young one, we could be confident that our comparisons across the erectus/ergaster collection would be meaningful.
In the years since Dubois found the first H. erectus in 1891, the collect
ion had grown into one of the most comprehensive samples for any hominin species. From Africa, there were four other good skulls to compare 42700 to as well as four partial ones. In addition to the Dmanisi skulls, there were more than thirty specimens from Asia. The skulls range in age from the oldest at around 1.8 million years to the youngest, which could be a few hundred thousand years old. The whole sample spans nearly 1.5 million years—and within this span, most of the African skulls are clustered around the old end of the spectrum while the majority of the Asian ones are considerably younger. How ever were we to sort out which were valid characters that define the erectus species and which characters were inevitable differences that arose because of the enormous range—both geographically and chronologically—of this hugely successful hominin?
Looking closely at the collection of African skulls assembled under the appellation H. ergaster, we found many variations in size and features. When H. ergaster was reincarnated as a separate species based on comparison with the South African australopithecines and on a mandible alone, there was never a gold standard of defining characters put forward for the skull that could be used to determine whether a specimen belonged in this assorted group. People recognised H. ergaster skulls from informally agreed-upon characteristics. These features included thinner cranial bones and less pronounced brow ridges than those present on the Central and East Asian erectus. A third feature considered key is known as keeling: the prominence of a ridge running along the top of the skull to its base. This feature was believed to be unique to the non-African hominins and was a strong basis for separating ergaster from erectus. Common to skulls in both the African and Asian collections is the distinctive ridgelike torus at the back of the skull that differentiates erectus (or ergaster) from other early Homo. Lastly, the cranial capacity of the erectus/ergaster group is larger than that of other early Homo.
Other parallels made the Dmanisi skull very relevant to our study. When this Georgian skull was first found, some scientists latched on to its small size and gracile features to contend that it was an early intermediate hominin that linked an ancestral Homo habilis in a linear line of evolution to H. erectus/ergaster. This view was pervasive—even Mary Leakey, whom I admired as one of the most thorough and careful scientists I have ever had the privilege to work with, bought into it because H. habilis and H. erectus are in completely different beds at Olduvai. Although still cautious, Mary wrote that “it is possible that they represent two stages of human evolution.” However, this line of thinking ignores evidence that has been with us since the 1970s when we found fossils of both H. erectus and H. habilis in sediments of the same age of 1.78 million years. These fossils from Turkana suggest that H. erectus and H. habilis rubbed shoulders in the same habitat for several hundred thousand years.
There is another non-African skull that lacks the “classic” erectus characters of a pronounced occipital torus and a heavy continuous brow ridge. This skull recently made headlines after it went on a little unauthorized adventure to New York City in 1999, where it surfaced in a shop called Maxilla & Mandible, a natural-history establishment. The proprietor, Henry Galiano, was uniquely qualified to recognise the import of his acquisition from his days at the American Museum of Natural History. After cleaning the skull, he took it with him for a visit to his colleagues at the museum to see if his suspicions of its importance were well-founded. A massive hunt began for clues as to where this beautifully preserved skull could have come from. Eventually, they got to the bottom of the mystery. The skull was discovered by sand miners in 1977 on the banks of the Solo River in central Java not far from Sangiran. It was then purchased from the miner who found it, and after passing through several hands, it eventually ended up in an antiquities shop in Jakarta, where it gathered dust for some twenty years. In 1998, an Indonesian palaeoanthropologist examined the skull and published a brief description. It was this description, together with a tip-off that an Indonesian antiquities dealer had tried to sell just such a fossil, that allowed the team at the Museum of Natural History to finally put two and two together. The skull was returned to the Indonesian authorities, who gave it its rightful place in the Indonesian collection and its own accession number, Sambungmacan 3 (Sm 3).
Susan Antón works in New York, and she had followed the twists and turns of this story with great interest. When we showed her 42700, she was immediately struck by some similarities of Sm 3 to both the Dmanisi skull D2700 and ER 42700. These three skulls belonged to individuals who lived thousands of miles apart, and in the case of ER 42700 and D2700, several hundred thousand years apart. If we believe the literature, these three were most emphatically not supposed to look alike, yet they could very easily have been triplets. It was clear to us all that we would need to go over the hodgepodge collection of accepted H. erectus characters and figure out which ones held true over the huge and diverse range of specimens. This would enable us to answer two confounding questions: is H. erectus the same thing as H. ergaster, and what makes H. erectus different from H. habilis?
As he had with Flat Face, Fred set to work building a spreadsheet with all the standard measurements that defined supposed H. erectus characters. He and Susan plugged in the measurements of every skull and partial skull in the entire collection from Africa to Asia, giving us a total sample of forty-five specimens to work with. Statistically speaking, this might not amount to much, but compared to many hominin species, this is, in fact, a relatively robust sample.
Next, we divvied up the sample in different ways. To see if there was any pattern to the variation in specimens, we compared the African specimens to the non-African specimens and the younger specimens to the older ones. Then we looked at the whole sample to see if any characteristics held true for young and old and African and non-African. Fred crunched the numbers for us using bivariate analysis—comparing the relationship between two related measurements such as the height of the skull and the width between the ears—to see what simple patterns emerged. Then he conducted more complicated multivariate analyses by throwing all the measurements at the computer program at once.
Much to our delight, some very interesting patterns were evident, and they upturned conventional wisdom about H. erectus. We looked at the features that were informally considered to separate erectus from ergaster. Having a thick skull (vault thickness), prominent brow ridges, a distinctive ridge running along the top of the skull (keeling), and a rather pointed occipital torus at the back of the skull were thought to be Asian traits, but we didn’t find a single feature that was unique to African specimens, nor could we discern any that existed solely outside of Africa. Different combinations of features cropped up across the board. The conclusions were unequivocal: there didn’t seem to be any justification for maintaining H. ergaster as a separate species.
In a nutshell, there are far more differences through time than across geographical distances. In addition to changes seen over time, there are further features that have to do with the developmental age and size of the individual (which scientists call allometric differences). So a big old African male erectus who lived some 1.5 million years ago, such as OH9, will look far more like an old fellow from Indonesia or China who lived around the same time than he will a youth or maiden such as ER 42700 who lived in roughly the same neighbourhood at around the same time. Small skulls such as ER 42700 and D2700 do have features normally associated with H. erectus, but these are far less pronounced than in the bigger skulls. Without postcranial material associated with the skulls, we can’t tell precisely if there are big differences based on the sex of the individual, but sexual dimorphism likely accounts for a lot of the variation in H. erectus.
The “classic” features long held to differentiate erectus from ergaster hold true across the entire sample although they do not all appear in every individual. A thick skull and pronounced brow ridges don’t necessarily make an erectus, as these features also show up to some degree in other early Homo. There are two features on the base of the skull that do seem to
crop up every time, however. The first of these is the area on the skull that makes up the skeletal structure of the ear—the tympanic, which is basically the ear hole, and the petrous, which holds both the bony labyrinth that we use for balance and the three ear ossicles (the bones that vibrate in response to sound frequencies and allow us to hear). The tympanic and petrous are oriented in a very distinctive angle in H. erectus that is different from the angle in other early Homo. The second distinguishing feature is the surface where the lower jaw articulates with the skull. In erectus, this is rather small and narrow from side to side. ER 42700 had both of these latter unique features along with the keeling that used to be thought of as an Asian character and a small gentle occipital torus. Susan, Fred, and I unanimously agreed that there could be no doubt that this was a H. erectus. The only potentially controversial thing was that we had demolished H. ergaster in the process.
The second important fossil found that year, John Kaatho’s maxilla, also gave us a very interesting new insight. The maxilla, which was given the accession number ER 24703, had a row of rather worn teeth from the canine to the third molar on the right side, and the angle and height at which the cheekbone (zygomatic) arched gently away from the upper jaw looked unusual. Because of its very young stratigraphic age of 1.44 million years, I had simply assumed that it must be H. erectus in spite of its rather sizable wisdom tooth (third molar). But Fred took one look at that tooth and the angle of the zygomatic arch, and vehemently disagreed with me. Sure enough, when we compared the erectus sample to all the other early Homo, we found that both the third molar and the shape of the zygomatic were good markers to tell habilis apart from erectus. What was really astounding about this was that no H. habilis had been found at such a young age before. It pushed the range of time during which H. habilis and H. erectus shared the African savannah habitats to close to a half a million years and made it even more unlikely or even impossible that H. habilis was ancestral to H. erectus.