The Accidental Species: Misunderstandings of Human Evolution

by Henry Gee

  This communiqué from beyond the realms of the known came from an international team of archaeologists working in a cave called Liang Bua, on the remote island of Flores, in Indonesia. If you want to find Flores on a map, look up the island of Java, and work your way eastward, past Bali and Lombok, and there it is. Flores is part of a long chain of islands that ends up at the island of Timor, well on the way to Australia, New Guinea, and the Pacific Ocean.

  One of the more intriguing questions in archaeology is when Australia was first settled by modern humans, the ancestors of today’s aboriginal peoples. There is much debate about this issue. Clearly, one way of illuminating the problem is to search for early modern humans living in what is now Indonesia, which can be thought of as a series of stepping-stones between mainland Asia and Australia.5 That’s where Flores comes in. Archaeologists are interested in the caves of Flores and other islands such as Timor because of their potential to yield remains of Homo sapiens, modern people caught in the act of heading toward that distant island continent later associated with cold lager, “Waltzing Matilda,” and The Adventures of Priscilla, Queen of the Desert. This is what drew an international team of archaeologists to Flores, and in particular to Liang Bua, known as an archaeological site for decades.

  Flores, though, is an island of mysteries—for it has been inhabited for at least a million years,6 and not by Homo sapiens. Stone tools have been discovered in several places on the island, and their makers are usually thought to have been Homo erectus, an earlier hominin,7 whose remains are well known from Java, China, and other parts of the world. The bones of these early inhabitants of Flores have not been found, their presence betrayed only by the distinctive stone tools they left behind.

  But whoever these early inhabitants were, their very presence is a problem. In the depths of the ice ages, when much of the earth’s water was locked up in ice caps and glaciers, the sea receded so far that many of the islands of Indonesia were connected by land bridges—they could be colonized by anything able to walk there. Not so Flores: this remained separate, cut off from mainland Asia by a deep channel. Homo erectus—if that’s who it was—must have made the crossing from the nearest island by boat or raft, or, like other animals, washed up there by accident. Once they made landfall on Flores, there they stayed—cut off from the rest of the world for a very long time.

  Isolation on islands does strange things to castaways, making them look very different from their cousins on the mainland. So it was with Flores, home to a species of elephant shrunken to the size of a pony, rats grown to the size of terriers, and gigantic monitor lizards that made modern Komodo dragons look kittenish by comparison.8

  Such peculiar faunas are typical of islands cut off from the mainland where, for reasons still unclear, small animals evolve to become larger, and large animals evolve to become smaller. Miniature elephants, in particular, were rather common in the ice ages. Practically every isolated island had its own species.9 The one on Malta lived eye-to-eye with a gigantic species of swan called Cygnus falconeri, with a wingspan of around three meters.10 Micromammoths evolved on Wrangel Island in the Russian Arctic, where they outlived their larger mainland cousins by thousands of years.11

  The fate of island faunas was an important consideration for Charles Darwin, who marveled at the creatures of the Galápagos Islands in the Pacific Ocean, when HMS Beagle visited in 1835. Darwin noted that each island had its own species of giant tortoise, as well as its own finches—different from one another yet plainly similar to finches from the mainland of South America. Had some stray finches, once marooned on the Galápagos, evolved in their own way?

  The scene is set, then, for Flores, where, at Liang Bua, archaeologists surrounded by the bizarre sought for something so seemingly prosaic as signs of modern humans.

  What they found instead was a skeleton, not of a modern human or anything like one, but a hominin shrunken to no more than a meter in height, with a tiny skull that would have contained a brain no larger than that of a chimpanzee.

  In some ways the skull looked disarmingly humanlike. It was round and smooth, just like a human skull, and with no sign of an apelike snout. In other ways it was a throwback. The jaw had no chin—the presence of a chin is a hallmark of modern humans, Homo sapiens. The arms, legs, and feet of the creature were most odd, looking less like those of modern humans than those of “Lucy” (Australopithecus afarensis), a hominin that lived in Africa more than 3 million years ago. The big surprise, though, was its geological age. Despite its very ancient-looking appearance, the skeleton was dated to around 18,000 years ago. In terms of human evolution, this is an eyeblink, hardly rating as the day before yesterday. By that time, fully modern humans, having evolved in Africa almost 200,000 years ago, had spread throughout much of the Old World. They had long been resident in Indonesia, and indeed, Australia.

  So what was this peculiar imp of a creature doing on Flores, seemingly so out of tune with its times?

  Despite the tiny brain, the creature seemed to have made tools. Pinning tools on a toolmaker is very hard (we weren’t there to see them do it), but these tools looked very like those known to have been made on Flores hundreds of thousands of years earlier, presumably by Homo erectus. The only difference was that they were smaller, as if fitted to tiny hands. Had the archaeologists discovered a hitherto unknown species of hominin, dwarfed by long isolation alongside the miniature elephants?

  Further work at Liang Bua showed that the first skull and skeleton were no flukes. The skeleton was soon joined by a collection of more fragmentary remains, though no more skulls.12 All the remains could be attributed to the same species of tiny hominin, and showed its presence at Liang Bua, off and on, from as long ago as 95,000 years ago (well before Homo sapiens arrived in the area, as far as we know) to as recently as 12,000 years ago.

  After that—catastrophe. A layer of ash found in the upper sediments at Liang Bua indicate that many of the inhabitants of Flores were destroyed in a volcanic eruption around 12,000 years ago. The calamity swept away the fairy-tale fauna of giant lizards, tiny elephants, and tiny people (though the giant rats are still there, to this day). More recent sediments, laid down after the eruption, betray the presence of modern humans, their tools, and their domestic animals.

  The account that reached my desk at Nature made it plain that the discoverers were as honestly puzzled by their discovery as anyone else would have been, in this coal-face confrontation with the absolutely unknown and unexpected—a new species of hominin that lived until almost historical times, but with a weird, antique anatomy and a very, very small brain indeed.

  To emphasize the strangeness of the creature, the discoverers gave it a scientific name that was noncommittal, yet set it apart from anything discovered hitherto. They called it Sundanthropus florianus—the Man from Flores, in the Sunda Islands. However, the panel of experts I called on to comment on the draft paper, and to make suggestions for its improvement, pointed out how relatively modern the skull looked—how much it looked like our own genus, Homo. One commentator also noted that “florianus” didn’t actually mean “from Flores” so much as “flowery anus.” Clearly, some revision was required.

  When the revised paper was published in October, the creature had become Homo floresiensis—Flores Man. The skeleton with its skull was catalogued as LB-1, but the media were quick to catch on to a suggestion of one of the discoverers that it should be known as the “Hobbit,” after the diminutive hole-dwellers of J. R. R. Tolkien’s fiction—though we in the Nature office sometimes referred to her as “Flo” (the skeleton having been described as that of a female).

  The paper—and the several commentaries that appeared in its wake—saw the Hobbit as a member of a race of humanlike creatures that had evolved in isolation, on Flores itself or nearby, perhaps descendants of the full-sized toolmakers known to have been on Flores for as long as a million years. If isolation on islands could do strange things to creatures as varied as birds and elephants, lizards and tor
toises, there seemed no reason in principle why hominins should be exempt. The Hobbit could easily be seen as a relative of Homo erectus, known from remains on mainland Asia to be almost as tall as a modern human—but dwarfed as a result of isolation, alongside the elephants whose island it shared.

  And then the fun started.

  Hardly had the ink dried on the first account of the Hobbit when the backlash began.13 Critics were exercised by two particular aspects of the discovery.

  First, that such an archaic-looking creature had existed so recently, in a region already long inhabited by modern humans.

  Second, that a creature with such an incredibly tiny brain could have made tools. The brain was so tiny, even in proportion to the tiny body, that the Hobbit must—the critics reasoned—have been suffering from a physical or genetic abnormality.

  Although criticism of the find came in various shades, critics were united, more or less, in proposing an alternative scenario for the existence of the Hobbit. Rather than it being a distinct species, a relic of an older world preserved out of time, it was a form of modern human suffering from microcephaly, a congenital disorder that produces midgets with abnormally small heads.14

  The first objection can be seen as a symptom of human exceptionalism, the erroneous yet deeply ingrained tendency that I seek to explode in this book. That is, the tendency to see ourselves as the inevitable culmination of a progressive trend of advancement in evolution. The discovery of such a primitive-looking creature living on the same planet at the same time as Homo sapiens challenges that view. It is a perhaps unfortunate fact that the only hominin that still exists on Earth is our own. This fact rather reinforces the idea that various species of hominin—the “missing links”—each more humanlike than the one before, succeeded one another with the planned inevitability of runners in a relay race, and that it is not somehow possible for several species of hominin to coexist on the same planet.

  It was not always so. As recently as 50,000 years ago, there were at least four different kinds of hominin on Earth—Homo sapiens in Africa, Neanderthals (Homo neanderthalensis) in Europe and western Asia, and Homo erectus in southeastern Asia, to which must now be added the obscure “Denisovans” from eastern Asia.15 The addition of a fifth—Homo floresiensis—would, in such circumstances, hardly be a surprise: neither should it be a surprise were yet more distinct forms of human to be discovered. Indeed, the only period in which only one species of hominin walks the earth is right now. Modern times are the exception, not the norm.

  That different hominins might live together in the same region should, likewise, not be a surprise. It is known that various kinds of early Homo coexisted with australopiths in east Africa between 2 and 3 million years ago, and that humans and Neanderthals coexisted in Europe for at least 10,000 years (between around 41,000 and 27,000 years ago). The survival of Neanderthal genes in the modern human population16 shows that the two species occasionally interbred. There can, therefore, be no objection to Homo floresiensis as a distinct species, simply on the basis that modern humans were around at the same time; nor on the basis that Homo floresiensis looks too primitive to have survived until modern times. As anachronisms go (what people like to call “living fossils”), the Hobbit is hardly a world-beater. Go tell it to the tuatara of New Zealand, the last relic of a lineage of reptiles distinct from a time before dinosaurs evolved, and hardly changed in its external appearance for 250 million years.17

  The second objection—that the very small brain of Homo floresiensis must have been pathological, a symptom of microcephaly—is likewise flawed, but much more interesting.

  Microcephalics have heads that are disproportionately small, even for very small people, such as dwarfs or pygmies. It is important to realize that microcephaly has a number of distinct causes. Microcephaly is not one single disorder. Microcephalics suffer from a variety of other disorders as well as malformations of the skull, face, and limbs, the particular suite of complaints dependent on the variety of microcephaly at issue. Some degree of mental retardation is, perhaps not surprisingly, a feature common to microcephalics in general.

  And so it was that the Hobbit was compared with various kinds of microcephalics. However, although the brain of the Hobbit is undoubtedly very small, and the skull and skeleton of LB-1 strange in many ways, its strangeness could not be mapped easily onto any variety of microcephaly recorded for modern humans. That does not mean that the microcephaly idea is ruled out. It could be that LB-1 is the only known exemplar of a hitherto unknown variety of microcephaly. After all, microcephaly of any kind is rather rare, so much so that scientists seeking to compare the Hobbit with microcephalics had to dig deep into the world’s medical museums and medical literature even to find the very few specimens of microcephalics available for examination. It is possible that LB-1 suffered from a variety of microcephaly as yet unmapped.

  Perhaps the most interesting suggestion of this sort—that LB-1 was a pathological specimen of modern human—was that it was not a microcephalic, but a cretin.18 Cretinism is not a genetic or inherited disorder, but the result of a chronic deficiency of iodine in the diet. Iodine is a vital component of a hormone, thyroxine, which the body needs for proper growth. Without thyroxine, growth is retarded, and the result is short people, with small heads and various degrees of mental impairment. Iodine is found in seafood, so cretinism is not common close to the sea. It is (or was), however, more common in isolated, inland communities. Liang Bua is in the Floresian hinterland, relatively far from the sea. It is conceivable that LB-1 could have belonged to a tribe of highlanders more prone to cretinism than fisherfolk living on the coast.

  But the more that Homo floresiensis was studied, especially once the peculiar proportions of its arms and feet became better known, the less well it fit into any known variety of pathology found in modern humans.19

  The scenarios in which Homo floresiensis was not a real species but a pathological version of a modern human were varied, but had one aspect in common: they failed to emphasize (or even mention) that LB-1 wasn’t an isolated case that could be singled out as pathological. Remains of the same kind of creature had been recovered from strata at Liang Bua representing an enormous span of time, back to a time before modern humans were known to have existed in the region. This fact alone should have been enough to question any idea that Homo floresiensis was a pathological offshoot of modern humans.

  The fundamental problem with the microcephaly idea lies less with the idea of microcephaly, or pathology, than that its proponents subscribe to an untenable view of human evolution—one that can only admit to a single pathway of evolution in which human beings stand at the head of a single line of ancestors, each one progressively improved compared with the one before. In that worldview, Flo can only be a human being—in which case one then has to explain how she came to look so odd. Proponents of this view tend to be both passionate and argumentative, and become more so as evidence mounts to discredit it. This suggests that the argument is less about one curious fossil than an attempt to shore up a view of the world that is fundamentally mistaken.

  The same problem besets the assertion that as a consequence of its small brain, Homo floresiensis would not have been able to make tools. It is now known that a wide variety of animals can make tools, many of a sophistication to rival anything made by early hominins. Some of these creatures have very small brains indeed—brain size per se need have little or no connection with technical ability. The idea that brain size matters comes from the view that human evolution is progressive, linear, and inevitably improving.

  The problem remained, however, that irrespective of its origins, Homo floresiensis really did have a disproportionately tiny head. Scaling a modern human down to Hobbit size would have created a creature with a tiny head, but only if it were pathological. The heads of Homo erectus were, in contrast, smaller than those of Homo sapiens, so perhaps Homo floresiensis would be better seen as a dwarfed (but nonpathological) Homo erectus. Homo erectus was a rema
rkably variable species with perhaps a tendency to smallness,20 something that might play in its favor as a possible ancestor of Homo floresiensis. Specimens found in the Republic of Georgia dated to around 1.7 million years ago seem to represent a sample of Homo erectus of a primitive, early kind.21 These creatures were small, some comparable in size with Homo floresiensis, but their brains were at least twice the size of LB-1. Shrinking Homo erectus down to the size of Homo floresiensis would still produce a creature with too large a brain. Flo had to have evolved from something smaller still.

  Two possible solutions presented themselves. One was a study on island dwarfism in now-extinct hippopotamuses that lived on Madagascar, showing that in some cases, the brains of animals subject to island dwarfism would be reduced more than one would expect, even when one scaled a full-sized animal down to midget size.22 This makes sense in terms of energetics. A possible cause of island dwarfism is that castaways evolve a smaller size in response to the pressure of reduced resources. The brain is, proverbially, the most expensive organ to run in terms of its mass, and so might be expected to evolve a disproportionately small size. Yet such a reduction has its limits. A brain can’t reduce to the extent that function would be impaired. However you look at it, a race of cretins or microcephalics isn’t going to survive for very long. But even when the further downsizing of brains of island species was accounted for, the brains of Homo floresiensis looked too small, even for Homo erectus.

  The second solution was that Homo floresiensis was a dwarfed version of an even earlier, more primitive hominin than Homo erectus, perhaps a creature so primitive that it would not be grouped within the genus Homo. This had indeed been an option favored by the original discovery team, but they had to some extent been dissuaded by the panel of experts I’d assembled to assess the original report, who had looked at the skull of LB-1 and said that despite its size it fit better within Homo rather than outside it.