Small skull, huge controversy: Saga of the Flores 'hobbit' continues

In October 2004, while working in his lab, Bob Eckhardtheard a report on National Public Radio: A team of archaeologists had unearthed bones of a three-foot-tall humanlike creature on the island of Flores in Indonesia. Based on the shape and size of the skull and other skeletal remains, the archaeologists, led by Michael J. Morwood of the University of New England in Armidale, Australia, claimed they had discovered a new species of human.

The diminutive biped had a cranium no larger than a chimpanzee's, yet its bones had been found along with abundant stone tools. Radiocarbon dating of charcoal in the same stratum, along with luminescence dating of surrounding sediments, implied that the skeleton was only 18,000 years old. Considering other earlier archaeological finds on Flores, Morwood and his colleagues concluded that a new human species had evolved from a preceding population of Homo erectus that had been isolated for over 840,000 years on Flores, in the archipelago between Asia and Australia.

Eckhardt, a professor of developmental genetics and evolutionary morphology in Penn State's department of kinesiology, added it up. Three feet tall. A tiny brain. Complex stone tools. Evolved in complete isolation in 40,000 generations. He says: "It just didn't ring true."

Eckhardt read the scientific papers, published in the British journal Nature, setting forth the findings and conclusions of Morwood's group. "A lot of things didn't make sense," he says. "For instance, the overall height seemed to be off. I took the long-bone measurements from the paper and plugged them into standard regression formulas." Where Morwood and colleagues estimated an overall height of 1.06 meters for their specimen, Eckhardt came up with figures ranging from 1.15 to 1.33 meters, with an average of 1.25 meters—more than seven inches taller than Morwood's estimate. Eckhardt also wondered about the proximity of the small cranium to sophisticated stone tools, including points, perforators, blades, and microblades. Over a century of research by anthropologists has established a rough correlation between an increasing brain size and advances in stone-tool technology. The kinds of tools described in the Nature article matched those made elsewhere by Homo sapiens. Says Eckhardt, "It seemed very unlikely that a human with a chimp-sized brain would have invented such tools independently and in total isolation."

A "hobbit" is born

Courtesy R.B. Eckhardt

Front view of LB1 skull.

That the Morwood find represented a new species also seemed doubtful to Maciej Henneberg. Henneberg works at the University of Adelaide in Australia, where he is the Wood Jones Chair of Anthropological and Comparative Anatomy and heads the division of Anatomical Sciences. The day the Morwood papers appeared in Nature, Henneberg announced during a radio interview that the most complete skeleton recovered by the Morwood group likely came from a developmentally abnormal individual, a member of Homo sapiens whose tiny head exhibited microcephaly, a condition in which a person's braincase remains very small because the brain fails to attain a normal adult size.

A flurry of e-mails passed between Eckhardt and Henneberg. (The two have known each other for years and currently are co-investigators on a project funded by the Australian Research Council.) Says Eckhardt, "Maciej's hunch complemented my own conviction that the 'new species' scenario didn't make sense. And it dovetailed with my belief that the Morwood group had exaggerated the size of their specimen downward." Eckhardt notes that the apparent novelty of the Flores skeleton was enhanced by comparisons with populations from Europe and other major continents where the "normal" stature approaches six feet.

Peter Brown, also of the University of New England, had worked with Morwood in analyzing the Flores remains. They named the purported new species Homo floresiensis, since it had been found on Flores.
The nearly complete skeleton (the arms were missing, but they turned up in a later dig) was categorized as LB1, in reference to the expansive limestone cave, Liang Bua, where the bones had been unearthed about six meters below the cave floor. (Liang Bua means "cool cave" in the local language.) Less formally, members of Morwood's team dubbed the creature a "hobbit"—capitalizing, Eckhardt believes, on the popularity of the film adaptation of J. R. R. Tolkien's fictional trilogy The Lord of the
, in which small humans known as hobbits do heroic things.

The press, both popular and science-oriented, latched onto the name. And they embraced the new-species notion enthusiastically.

In February 2005, Scientific American ran an article accompanied by a color illustration of a band of pint-sized, spear-toting hunters overwhelming a Stegodon, an extinct dwarf elephant. (Stegodon bones had also been found in Liang Bua, bearing marks made by bladed tools.) The article, by Kate Wong, was entitled "The Littlest Human" and included as a subhead: "A spectacular find in Indonesia reveals that a strikingly different hominid shared the earth with our kind in the not so distant past." It continued: "Conventional wisdom holds that Homo sapiens has been the sole human species on the earth for the past 25,000 years," but the remains found on Flores "have upended that view."

The cover of the May 2005 National Geographic presented a mockup portrait of the hobbit—dark-skinned, big-eyed, startled-looking. Morwood, in a feature article inside, wrote: "We had discovered a new kind of human ... We had stumbled on a lost world: pygmy survivors from an earlier era, hanging on far from themain currents of human prehistory." Jared Diamond, a UCLA evolutionary biologist, stated in a Public Broadcasting System interview: "This is the most amazing discovery in any field of science in the last ten years." Others touted the find as the most important discovery in human evolution and paleoanthropology in half a century.

Alone on an island?

Those characterizations of the importance of the Flores skeletons only intensified Eckhardt's interest. In the kinesiology department he teaches a graduate course in Experimental Design and Methodology. "The course stresses a key principle articulated by Sir Peter Medawar, who shared a Nobel Prize for pioneering work in immunology," Eckhardt says. "Scientists, particularly those of us with decades of experience, are supposed to work on the most important problems that we have a reasonable chance of solving." To Eckhardt and his colleagues, the Flores find represented precisely such a problem. They would attack it, Medawar-style, not through armchair theorizing but by testing hypotheses.

Courtesy CIA World Factbook

Flores is one of the Lesser Sunda Islands, in the Malay archipelago.

Morwood and his colleagues speculated that a founding cohort of Homo erectus individuals had reached Flores from a nearby island, probably during a period of intense global glaciation, when huge volumes of
water would have been tied up in the polar icecaps, lowering sea level and exposing a greater amount of land. It was unclear how the hominids had gotten to Flores, whether by using primitive rafts or clinging to flotsam. Stegodons had also colonized Flores at about the same time. (Elephants are known to be
strong swimmers.) Once isolated on the island, both hominids and elephants shrank. The stegodons went from being slightly larger than modern African elephants to about the size of a water buffalo. The hominids supposedly dwindled as well from their more robust Homo erectus ancestors.

The so-called island rule is a widely accepted biological precept holding that mammals larger than approximately rabbit size tend to become smaller over millennia in an adaptive response to an island's limited food resources. Most paleoanthropologists, however, believe that our culture and behaviors buffer humans against some of the factors that cause other mammals to evolve rapidly; where another species might develop a thick pelt to ward off the cold, we make clothes and harness fire. In his National Geographic article, Morwood said that the small human skeletons provided "powerful evidence" for hominid evolution in isolation on Flores.

But had the island really been isolated? In the 1950s and 1960s, evidence of an early human presence had been found on Flores. Theodor Verhoeven, a Dutch priest and amateur archaeologist, had excavated crude stone artifacts near the fossilized bones of stegodons thought to be around 750,000 years old.
On nearby Java, others had found 1.5-million-year old Homo erectus remains, which led Verhoeven to conclude that erectus had somehow made the crossing to Flores.

Morwood and his colleagues had unearthed a number of hominid bones in Liang Bua, although only the one complete cranium. They noted the sloping forehead, arched brow ridges, large jawbones, and receding chin on LB1, which, they said, mirrored Homo erectus traits. However, as Morwood wrote in National Geographic, "The tiny skull is most reminiscent not of the hefty Homo erectus from elsewhere in East
Asia but of older, smaller erectus fossils." The Morwood team stated in their Nature article that a CT scan demonstrated a congenital absence of a third molar, and they noted a unique positioning of other teeth. They also pointed to an unusual robustness of the leg bones and a low degree of humeral torsion, the twisting of the upper arm bone between the shoulder and the elbow. All of these characteristics were advanced as proof of a new species.

"Hobbit wars" heat up

Eckhardt knew that populations still living in parts of the world near Flores—on the Malay Peninsula, in the Philippines—were short-statured. He checked his impressions against a book he had read decades
before, The Origin of Races, by the anthropologist Carleton Coon, published in 1962. There Eckhardt found a footnote describing two small skeletons excavated in separate caves on Flores in the 1950s by
the amateur archaeologist Verhoeven. Deciding that he needed to see those previous finds, Eckhardt tracked down the skeletons at Naturalis, the Dutch national museum of natural history in Leiden. In January 2005, Eckhardt flew to the Netherlands to examine the skeletons. "The two measured 1.5 and
1.6 meters in length—quite small but somewhat larger than the height Morwood's group was proposing for LB1," Eckhardt says. He realized something else: The Verhoeven skeletons differed not only from
Morwood's Liang Bua specimen but also from each other. He says, "To me, those differences clearly suggested that Flores, far from being isolated, had been reached repeatedly by people from other regional populations."

Courtesy R.B. Eckhardt

At left, side view of Liang Momer E skull in Naturalis (Netherlands National Natural History Museum, Leiden). Right, side view of Liang Togé skull in Naturalis.

By that time, Radien Soejono of the National Archaeological Research Center in Jakarta, listed as one of the coauthors of the Morwood Nature paper, had asked the Indonesian paleoanthropologist Teuku Jacob to restudy LB1. Jacob is with Gadjah Mada University in Yogyakarta; his entire career has centered on the analysis of ancient human remains. Says Eckhardt, "Radien and Teuku are considered to be the two grand
old men of Indonesian archaeology. Radien works mainly with stones, Teuku with bones."

A number of scientists had begun questioning the new-species designation through letters and comments in a range of scientific journals. The group that included Eckhardt and Henneberg was at the forefront of the critics, while other specialists had lined up behind Morwood and his team. What the press started
calling "the hobbit wars" had begun to heat up.

Following preliminary analysis of LB1, Jacob also concluded that the skeleton was not normal and did not represent a new species. Says Eckhardt, "Morwood's team reacted in an odd manner for scientists, who
are supposed to believe in the value of independent study of evidence and replication of results." Instead, through the popular scientific press, "They made numerous charges, including that Jacob was holding on to LB1 and would restrict access to the bones in the future."

Continues Eckhardt: "Just the opposite was the case. Teuku had repeatedly invited me to examine the bones myself. Then, early in February 2005, I got an e-mail from Teuku saying he was under intense pressure to return the remains. If I wanted to see them firsthand, it had better be now." Eckhardt rearranged his Penn State classes and flew to Yogjakarta in mid-February, where he joined a group that included Jacob; Henneberg; Etty Indriati, a University of Chicago-educated anthropologist specializing in dentition, and Jacob's colleague at Gadjah Mada University; and Alan Thorne, a paleontologist with the Research School of Pacific and Asian Studies at the
Australian National University in Canberra.

Look at the bones

"There we were," recalls Eckhardt, "sitting around a four-by-four table covered with a batch of plastic trays holding the remains of several small, long-dead Indonesians. We were picking up the bones, examining them, putting them back down. Every once in a while, looks would be exchanged across the table, and then one of us would articulate something we had all probably noticed. For instance, Maciej held up one of the femurs and said, 'The Nature paper says this is a right femur. But it is a left femur.'"

Indriati handed the LB1 skull to Eckhardt. "She said, 'Look at the back of the maxilla.' She whisked off some bits of dirt. Where the third molar was supposed to be congenitally absent, instead we have a socket with a piece of tooth in it." Discussions were intense and wide-ranging as the scientists drew upon their collective knowledge of mammalian evolution, human variation, and regional conditions in Indonesia and Southeast Asia.

The international team came up with four key areas of evidence disproving the assertion that LB1 represented a new species: geographical factors; a pronounced asymmetry of the skull and face of LB1; dental traits; and abnormalities in bones other than the cranium. The Proceedings of the National Academy of Sciences (PNAS) published their findings on September 5, 2006.

Team of researchers
Courtesy C.D. Eckhardt

Team including Eckhardt in Yogjakarta, discussing LB1 fossils before Indonesian media. Left to right: Eckhardt, Indriati, Henneberg, Thorne, Soejono.

Morwood and his colleagues had theorized that Homo erectus individuals traveled to Flores around 840,000 years ago and subsequently evolved in isolation to become Homo floresiensis. That claim assumed no additional influx of humans to the island until just before or just after the "hobbits" had died out around 15,000 years ago, perhaps following a volcanic eruption that also led to the extinction of the stegodons. Jacob's team pointed out that other studies showed that the dwarf elephants had been able to reach the island on at least two separate occasions. Fluctuating cycles of glaciation at the earth's poles would have repeatedly enlarged the land mass of Flores and adjacent islands, leaving water gaps
of only a few kilometers. (That conclusion was based on research by K. Hsu of the National Institute of Earth Sciences in Beijing, a specialist in Pleistocene geology and a coauthor of the PNAS paper.)

Says Eckhardt, "There could have been numerous arrivals of humans during glacial stages with low sea levels, before final higher sea levels around ten thousand years ago widened the water gap separating Flores from neighboring islands. But by then, watercraft made crossings easy."

According to Eckhardt and his colleagues, the 14,200-square-kilometer island would not have offered food resources sufficient "for sustaining in isolation an adequate effective population" of hominids that would have provided enough genetic diversity to allow for survival and adaptation over hundreds of thousands of years. Rather, sporadic immigration from other Homo sapiens groups was far more likely.

After the Jacobs team had noticed that the LB1 skull was highly asymmetrical, they brought in David Frayer, an anthropologist at the University of Kansas. Using a set of photographs of the skull taken by a professional photographer, Frayer worked up computerized composite images of the hobbit's face. The combining of two left and two right side images of the face allowed for a comparison that made the asymmetry in the actual specimen strikingly obvious. The researchers also compared seven data points of left and right side measurements on the skull to quantify the asymmetry.

Evidence for abnormality

Jaw Bone
Courtesy R.B. Eckhardt

Base of LB1 skull showing socket of alleged "congenitally missing" upper right third molar.

"It turns out there's a huge forgotten body of literature on facial asymmetry, including many papers published nearly a century ago, based on studies at England's prestigious Galton Laboratory," Eckhardt says. "Everybody's face is asymmetrical to some extent. But when asymmetry exceeds about 1 percent, you're over the line into abnormality." In studying LB1, Eckhardt and Adam Kuperavage, a graduate student in kinesiology at Penn State, found that six of seven measurements taken on the skull's right side were larger than corresponding measurements on the left side by as much as 40 percent, while the seventh was 6 percent larger on the left side.

"Craniofacial asymmetry that extreme demonstrates that LB1 did not develop normally," Eckhardt says. "When we pointed out the asymmetry—which the Morwood group said in their original paper wasn't present—they backpedaled and said, sure, there's a small amount of asymmetry, but it was probably caused by pressure from sediments." Eckhardt cites a rebuttal of this explanation by the University of Wisconsin paleoanthropologist John Hawks, who writes in his weblog: "Yes, it is true that any archaeological specimen is likely to be distorted to some extent by reconstruction or postdepositional deformation. That might be true of this skull also. But in this case, the asymmetry clearly extends to morphological characters that should be relatively unaffected by such distortion."

Anthropologists frequently cite a unique shape or placement of teeth when describing a new species. According to Morwood's team, a CT scan had demonstrated the absence of a third molar for LB1. Etty Indriati had found the existing socket and a tooth fragment where the "missing" molar should have been. But LB1's teeth displayed other peculiarities, including enlarged wear surfaces, long roots, and an unusual rotated position of premolars in the upper jaw. "Those traits were characterized as unique," says Eckhardt. "But it turns out that the rotated premolars are shared by about 20 percent of the people still living in Rampasasa, a village near Liang Bua." This particular Australomelanesian population is
short-statured enough to be known as the Rampasasa pygmies. Many individuals in the population show receding chins (another supposed species-distinguishing characteristic), leading Eckhardt and his colleagues to state in their PNAS paper: "Absence of a chin cannot be a valid taxonomic character for the Liang Bua mandibles." The Jacob team contends that Morwood and his research group should have compared LB1's teeth with those of other populations in the same region, such as the Rampasasa cohort, rather than with Homo sapiens from other geographic areas of the world, principally Europe and Africa.

Femur Bone
Courtesy Etti Indriati

LB1 humerus, showing torsion between shoulder and elbow that is low but within the range of living humans.

Morwood's group had cited an unusual robustness of the leg bones of LB1. Eckhardt's team had CT scans done on the bones. "We paid for the scans with a few thousand dollars in traveler's checks that I was
carrying," Eckhardt says. "Others in the group combined funds and paid the professional photographer. Our involvement in this project moved so rapidly that there wasn't time to apply for conventional grants."

The CT scans showed that the cortex, or outer solid bone, was actually quite thin: "Those femurs are not robust at all," says Eckhardt. On the bones, the location of attachment points for the muscles suggest at least some paralysis. The LB1 skeleton also showed a low degree of humeral torsion, the twisting of the upper arm bone between shoulder and elbow. Normal humeral torsion in Homo sapienscommonly is about 140 degrees; LB1's arms show 110 degrees of torsion. "When a limb develops with serious muscle weakness, torsion is usually only about 110 degrees," Eckhardt says. "Many points of evidence combine to suggest that this individual probably had severe movement disabilities."

How did humans evolve?

Was LB1 microcephalic? According to Eckhardt, around two hundred medically distinctive disease conditions can produce microcephaly. The malady can be genetic in origin, and it can be caused by various diseases and by infection. Asymmetries in the face and other bones often accompany microcephaly. Microcephaly exists in skeletons from the Upper Pleistocene and the Holocene periods. The ratio of LB1's small cranial capacity and short stature are similar to ratios found over several generations of microcephalics studied by physicians in the twentieth century. Scientists have also traced the
condition through succeeding generations of humans.

Says Eckhardt, "The archaeologists who dug up LB1 made serious mistakes in characterizing what they found, and they drew conclusions that were not supported by the balance of evidence. Altogether, they have one complete skull, plus a second mandible, which is similarly small, and assorted other bones from perhaps eight individuals. You cannot designate a new species based largely on an abnormal individual.

Team of researchers
Courtesy C.D. Eckhardt

Group photo in Yogjakarta. Left to right: Thorne, Indriati, Henneberg, Jacob, Soejono, Eckhardt.

"In summary, the normal traits of LB1 were not unique but rather are characteristic of human populations in the region. The degree of humeral torsion, the structure of the long bones, the facial asymmetry, and the unusually small braincase all point to developmental abnormalities of the sort that often accompany microcephaly."

Continues Eckhardt: "We may be dealing with a population of individuals who went through a period of food shortages that made them smaller than they might otherwise have been. LB1 was about 1.25 meters
tall, and abnormalities of the sort from which that individual suffered commonly reduce stature markedly. The Rampasasa pygmies living near Liang Bua average just under 1.5 meters. That's not a huge stature difference."

In an apparent reaction to the Jacob group's PNAS paper, "the proponents of Homo floresiensis have now switched to arguing that the small humans must have originated elsewhere," Eckhardt says. "It seems that whenever we test one hypothesis and disprove it, they reinterpret the hypothesis into a less
readily testable form."

Since 1971, when Eckhardt earned his Ph.D. in anthropology and human genetics from the University of Michigan, many bones have passed through his hands. He has studied skeletal material in many of the major museums in the world, and spent five summers working with samples at the Institute of Anthropology and Human Genetics at the University of Frankfurt in Germany. In 1992 he published a comprehensive study of
skeletal changes in native Peruvians, based on samples ranging in age from 10,000 years before present to living populations. He has studied variation in skeletons as it is affected by age at death; sex; pathologies and developmental abnormalities; and evolutionary changes over time. He has also worked with the bones of other mammals, including chimpanzees, gorillas, orangutans, macaques, and baboons.

"Most people in the fields of anthropology and archaeology believe the process of human evolution has been one of intense splitting over time," he concedes. "My belief, based on studying thousands of specimens during my career, is the opposite. Variation within any given species seems to be consistently underestimated."

The study of human evolution has always been a notably contentious field, and the Flores skeletons remain the focus of an intense and not always collegial debate. Currently scientists from different disciplines are studying the volume and shape of LB1's cranium, and trying—unsuccessfully, so far—to recover mitochondrial DNA from the skeletal material. (If found, this DNA might be compared to samples taken from both Neanderthal fossils and modern humans.) Morwood's team continues digging on Flores. The
scientific press and mainstream media seem happy to keep the controversy alive.

Says Eckhardt, "My estimate is that 80 percent or so of paleoanthropologists want this new species to be real—so much so that they are willing to overlook glaring errors and inconsistencies in the studies
and conclusions of the archaeologists who found the bones. This is not surprising, since LB1 was proclaimed to be our generation's critical test case of paleoanthropological theory. By disproving the validity of this new species, we have called a central dogma into question.

"Core beliefs are incredibly resistant to change. But testing and rejecting cherished hypotheses is how science moves forward."

Robert B. Eckhardt, Ph.D., is professor of developmental genetics and evolutionary morphology in the Department of Kinesiology and the College of Health and Human Development. His email is His research interests include the interaction
of genetic and environmental influences on growth and development in human populations, particularly in musculoskeletal structures.

Last Updated April 23, 2007