The Sediments of Time

by Meave Leakey

  KNM-ER 1470 was discovered on July 27, 1972, along the Koobi Fora ridge by Bernard Ngeneo. It remains one of my favourite fossils ever because of the happy memories I have of reconstructing it with hippos playing in the lake and baby Louise playing at my feet in a cool basin of water. Against this backdrop and out of thousands of nondescriptive bone fragments emerged one of the most exceptional, complete, and enigmatic finds our team had ever made.

  It was evident that 1470 was not an australopithecine, and it had all the indications of being far more human than anything else we’d found at Koobi Fora because its brain was much bigger. Desperate to know just how much bigger, Richard, Bernard Wood, and Bob Campbell resorted to pouring sand into the skull cavity and transferring the sand into a rain gauge, the only implement we had on hand for measuring volume. Using this ingenious solution, we learned that the brain capacity was equivalent to eight inches of rain, and we then began the business of converting this to something meaningful. Richard was so excited that he proudly wrote the camp diary entry for September 9 himself: “Worked on the first brain size estimate for the new skull. Plasticene (play dough), masking tape and sand in conjunction with a rain gauge (inches) and a 5 cc syringe resulted in an 800 cc minimum estimate. Historic moment!” For Richard, this was incontrovertible proof that he had a very early Homo, and he rushed back to Nairobi as soon as he could to share his great excitement with his father. He was just in time as Louis would be alive for only a few days more, and 1470 brought father and son together in a way that they had not enjoyed in decades.

  One of the reasons that Richard and Louis were so excited about this large brain-capacity estimate was because the layers where 1470 was found were below the now-infamous KBS tuff, which we believed at the time to be more than 2.6 million years old. Such a date associated with such a large brain was what Louis had long wished to discover, and for that brief evening they shared together, Richard and Louis basked in the belief that this was indeed so. In fact, the KBS tuff was much younger, and 1470 would later be pegged to a date of 1.9 million years.

  One of Richard’s rare diary entries in the camp log is for June 21, 1974. I was in Nairobi, having just given birth to our second daughter, Samira, and I was desperately anxious for any news of new finds. Richard writes dismissively and notes almost with disgust the possible hominin finds that had been collected just in one week: “Very poor specimens as follows. FS 100- 2 pieces cranial vault from 6B above middle tuff. FS 101- scrap of hominin molar from 6B above middle tuff. FS 102- ? Tibia pieces close to CHARI bone. FS 103- Cranial vault fragments above middle tuff #1.” Sure enough, every one of these four was indeed hominin—but nobody saw fit to celebrate!

  After discovering something as spectacular as 1470, it was difficult to get excited about little fragments here and there—we really needed another skull. That happened at the end of July in 1973. Kamoya turned up in camp in the early afternoon to tell us he had found a piece of another skull south of the Koobi Fora peninsula in sediments close in age to those where 1470 was found. We immediately set off in Richard’s plane to have a look. All we could see was a beautiful set of teeth embedded in an upper jaw that was barely protruding from the soil next to some skull fragments lying nearby. This was the only hominin we have ever found that was so easy to reassemble—a sharp contrast to 1470. As Richard excavated each fragment over the next two days, I was able to reassemble the whole skull in the field beside him. As the skull gradually took shape, we found that we had a beautiful specimen. It was a small skull that appeared to have come from a diminutive individual. This was the skull that was given the accession number KNM-ER 1813.

  KNM-ER 1813 looks quite different from 1470. It has a smaller brain capacity (510 cc) than 1470 (752 cc). It is not only smaller, but it also lacks 1470’s flat face, and the cranium has a different profile. It is unfortunate that 1470 has no teeth preserved, but from the appearance of the tooth roots, it seems to have had larger teeth than 1813. However, we can’t actually tell where the crowns of the teeth broke off from the roots. Because upper molar roots are always splayed at the tips and taper towards the crown, this makes a big difference. If the crowns broke off partway up the roots, this would give the false appearance of much larger teeth. Until we find another specimen with teeth, we cannot be sure about how big 1470’s teeth were.

  KNM-ER 1813 closely resembles another important find from Ol­duvai, OH 13, nicknamed Cindy after Cinderella because of its delightful daintiness. Louis grouped Cindy in H. habilis along with Jonny’s Child. His decision must be seen in its historical context: at the time, only australopithecines had been recognised from Africa, so he was focusing intently on how different his new fossils were from the robustness of P. boisei rather than on differences among his gracile small-toothed fossils. Moreover, both Cindy and Jonny’s Child were quite incomplete. It probably didn’t occur to Louis that he might in fact have two rather than one additional species. Given the evidence at the time and prevailing opinions, such an assertion would never have been accepted. As more complete fossils emerged over the following decades and Africa became an acceptable cradle for mankind in public opinion, the differences among different specimens assigned to H. habilis became more apparent.

  A further problem with inferring species from these skulls is that without any reasonably complete postcranial fossils associated with the skulls, we cannot tell how large brained the individuals actually were since what we are really interested in is the relationship between brain and body size. For instance, 1813 could have been a small (probably female) individual, and if we knew what her body dimensions were, her relative brain size might not have been particularly small. Rather than a cerebral rubicon dividing true humans from other hominins, a brain-to-body-size ratio is much more informative. However, since we so rarely find postcranial and cranial fossils that clearly belong to the same individual, this is not easy to estimate.

  Most scientists prefer to err on the side of caution and therefore tend to lump everything into the minimum (preferably one) species for any age. H. habilis has therefore become a catch-all name for a collection of rather different-looking specimens. It is almost certain that they don’t all belong together, though few agree on which fossils can be grouped, which of these should belong to H. habilis, and what to do with the others that do not. One of the main problems is that the H. habilis type specimen, OH 7, is not only a juvenile but also an incomplete mandible. Most of this mandible is in beautiful condition, but the two sides have been crushed together during fossilisation, so it is impossible to see exactly what the shape of the mandible would have been. On top of these problems, mandibles can be highly variable among individuals of the same species, so they make a notoriously bad choice for a type specimen. As a result, there has been much disagreement as to whether this specimen should be grouped with the smaller hominins or with 1470. And because, like it or not, Jonny’s Child is the type specimen, this is central to the question of how researchers understand H. habilis.

  There has, therefore, been a long-standing muddle about early Homo. The controversy settled for a time when scientists largely agreed that two species are represented in the collection of fossils grouped today as H. habilis. One of these is smaller and more gracile (represented by 1813 from Koobi Fora and OH 13 from Olduvai) than the other (represented by 1470). But I and many others found it hard to place OH 7 in either of these two groupings with any conviction. At our own dinner table, this was not a subject on which agreement was ever possible, and it was considered taboo by mutual consent!

  To make things even more confusing, the Russian anthropologist Valery Alexeev muddied the waters further in 1976 by assigning 1470 to its own species,rudolfensis. When Richard and Alan Walker first described 1470, they deliberately avoided assigning it to any species because they recognized the lack of evidence to relate it properly. And for many years, the consensus among palaeoanthropologists was to live with the imperfectly understood grouping. But Alexeev changed all that, and 1470, along wit
h one or two mandibles, were assigned to Homo rudolfensis. Other than the H. erectus fossils, all the other Homo specimens from the earliest Olduvai sediments as well as all those from similar-age deposits at Turkana are lumped together as H. habilis.

  But until the placement of OH 7 could be firmly established in either of the two groupings, what should take the name habilis remained disputed. But new technology allowed us to revisit this question decades after these controversial fossils were discovered. Performed correctly by a qualified anatomist, digital reconstruction allows for far more precision in a virtual construction than our early attempts with modeling clay ever could. In 2015, Fred Spoor created a digital model using CT scanning of the OH 7 mandible in order to correct the distortions caused by the sides of the mandible being crushed together. When Fred and his colleagues compared this reconstruction to other specimens of early Homo, their results showed that the variability across the sample is inconsistent with their all belonging to a single species. The reconstruction shows that the mandible is remarkably primitive, bearing a closer resemblance to the morphology of Australopithecus afarensis than to Homo erectus. Their reconstruction of the cranial capacity of OH 7 using the parietal bones yielded a much larger brain (between 729 cc and 824 cc) compared to the 680 cc arrived at manually in the 1960s, and they concluded that 1470 could not belong to the same species as OH 7.

  But both species are very different from the lineage of robust australopithecines. The robust australopithecines very likely occupied a feeding niche of tough, fibrous vegetation (using their megadont strategy of big cheek teeth, relatively small incisors, thick enamel, and little brains with big crests for muscle attachments). In contrast, the more gracile hominins probably ate softer foods such as fruits, insects, and small mammals (using their much smaller molars, relatively big incisors, thinner enamel, small muscle-attachment areas, and small faces relative to the size of their brains). If Homo evolved from the australopithecine line, it made a radical departure from the megadont trend in an as yet undiscovered intermediate species somewhere between three and two and half million years ago. Or perhaps its origins lie in an omnivorous dietary strategy that evolved in parallel much earlier—Kenyanthropus shows that this scenario is entirely plausible.

  Since H. habilis appears rather suddenly in the East African record—fully formed with a morphology that would not evolve much over the subsequent half a million years—it is quite probable that it evolved elsewhere and then moved into the drying woodland habitats alongside H. erectus. Of course, given that there is such a huge gap in the sedimentary record in East Turkana, it is equally possible that all this evolution happened right there and we have absolutely no way of finding the evidence for it. Other sites might yield the answer to this, and perhaps one already has. There is a 2.33-million-year-old maxilla missing its third molar that was discovered by Don Johanson’s annual field expedition to Hadar in 1994. This delicate, well-preserved fossil is indisputably early Homo. It has the classic U-shaped Homo palate. Don Johanson and Bill Kimbel at the Institute of Human Origins in Tempe, Arizona, together with Yoel Rak, have extensively studied the fossil, A.L.666-1, and concluded that its closest affinities lie with H. habilis. But it is too incomplete to say any more than this. An even older mandible, at 2.8 million years from Ledi-Geraru, is currently the oldest known precursor to Homo. This specimen corresponds in age to some isolated teeth found in Koobi Fora in 1978 that share some of the same transitional traits that suggest the teeth could be from the same ancestor to early Homo. These finds are important for they represent some of the earliest evidence we have of our own genus.

  While H. habilis is mysterious because of its great antiquity and its presumed status as one of the earliest Homo, even more intriguing is the origin of H. erectus. Louis assumed that H. habilis was the direct ancestor of our own species. However, this is another example of a scientist thinking with more heart than evidence. Indeed, Louis was to miss the import of his own momentous H. erectus discovery altogether. In 1960, he came across a pile of bone fragments that looked suspiciously like a tortoise. On closer inspection, Louis elatedly realised they were the pieces of a hominin skull without the face. Glued together, the pieces of OH 9 presented a nicely preserved, thick-boned skull cap with prominent brow ridges and all the trademark features of H. erectus. But Louis had been heavily influenced by Sir Arthur Keith, who had been a victim of the Piltdown hoax. Louis was looking for a large-brained ancestor so he was completely thrown by the small size of erectus’s brain. Ironically, OH 9 is among the largest brained of all the East African erectus specimens that have subsequently been found. But other fossils that would hint at the pivotal importance of this species would be discovered only after his death. Louis had very little to compare his Olduvai specimen to, which made it all the easier to be swayed by his strong preconceptions.

  Louis and Mary never found evidence of H. habilis and H. erectus in exactly the same level. Believing that H. habilis was the maker of the numerous tools scattered about Olduvai Gorge, Louis always regarded H. erectus as a dead-end branch of the family tree and therefore not particularly interesting. We now know that Louis was wrong. H. erectus is indeed extremely interesting. As we shall see, there is currently broad consensus that H. erectus was the first species to move out of Africa—and was the ancestor to Homo sapiens.

  One thing is fairly certain about the assortment of fossils variously classified as Homo habilis and Homo rudolfensis: they can all be clearly distinguished from Homo erectus. But what distinguishes them are features they lack rather than features that are in and of themselves characteristic of this hodgepodge group. The origin of the Homo lineage remains obscure, and it continues to perplex me a great deal. We can solve this enigma only with new fossil evidence. I have long hoped to find another 1470 to try to get to the bottom of this puzzle. Indeed, the focus of the Koobi Fora Research Project today, after more than three decades of searching, is to solve this ongoing conundrum. The ultimate discovery would be a complete skeleton, which would tell us so much more than we can learn from a skull alone. Until someone can unearth such a miracle, our story continues with H. erectus.

  Part III

  13

  Becoming Grandmas

  My whole perspective on just how different we are from all other mammals stopped being mere academic curiosity when my first granddaughter arrived in August of 2004. Here before me was a totally helpless and, it has to be said, screaming and colicky infant, who turned more than just her mother Louise’s life upside down. And even though she cried almost all the time, how I loved her! In between burping and nappy changes, baby Seiyia got me thinking in a whole new light about the significance of the different stages of life history and why the differences between humans and other animals are so important in explaining how we have become what we are. I had wondered with trepidation what my new role as a grandmother would be. But Seiyia’s arrival also highlighted two unique features of the human life history. The first of these was just how dependent and undeveloped a human baby is at birth. The second was more of a surprise: that my role as a grandmother is important in the grand scheme of human evolution.

  Women are reproductively active for about as long as a chimpanzee female is—approximately thirty years. But our life histories depart in all other respects, especially in infancy and postmenopause, and these differences relate directly to our large brains.

  Strangely enough, the gestation periods for a chimpanzee and a human are quite similar—eight months for a chimp compared to nine for a human. We would thus expect the babies to have similar-sized brains. However, the average size of a newborn chimpanzee’s brain is 137 grams and a human baby’s is 364 grams. These differences increase as the infants grow, and the average volume of an adult human’s brain is 1350 cc compared to a chimpanzee’s 400 cc—more than triple. But at birth, a chimpanzee, like all other primates, is already grown enough to function independently of its mother. It can cling tightly to her as the troop moves and is quickly adept
in locomotion and basic socializing. In contrast, a human baby needs its mother for absolutely everything for many months. In general, mammals’ life histories are closely related to the size of their brains. The problem for a human baby, however, is that if it stayed in the womb for the length of time we’d expect for its brain size, it would never make it out into the world past its mother’s hips!

  Standing and walking upright imposes all sorts of constraints relating to balance and locomotion, which resulted in significant changes to the shape of the pelvis. To accommodate an efficient striding gait, the shape of the human pelvis differs significantly from that of a quadrupedal ape. Mammalian birth canals are encircled by the bones of the pelvis (the ilium, ischium, and acetabulum) and the sacrum, so changing the shape of these bones alters the shape and size of the birth canal. Our pelvis has developed big “wings” (the ilia) to attach the abductor muscles that propel our legs. The edges of the ilia also attach to muscles from our torsos that keep us upright. The angle of the ilium is much more vertical compared to the elongated, diagonally positioned ilium in a chimp. But the shape of our pelvis is constrained by the need to keep our centre of gravity as close to a vertical axis as possible. If the heads of each of our femora were farther apart, we would waddle rather than stride (which is, incidentally, why men can usually run faster than women—because women have wider hips). As a result, the acetabula, which accommodate the heads of each femur, must remain as close together as possible. This in turn confines the width of the birth canal at the outlet, and the result is that the inlet and outlet of the human birth canal are quite different in size and shape.