A new human species recently discovered on the island of Flores in Indonesia, dating from a mere 13 000 years ago, and remarkable for being only one metre tall. Details have just been published in Nature. It is being described as one of the most important and sensational palaeontological finds in decades.

Flores is across Wallace's Line from mainland Asia: that is, it has always been across a deep oceanic barrier regardless of fluctuating sea levels. This seems to mean that this tiny creature must have had the intelligence to make some form of raft or boat.

Homo floresiensis therefore also co-existed with our own human species, and modern inhabitants of Flores have legends of a hairy little people they call Ebu Gogo. Well, so do many peoples around the world; this does not mean much by itself, but with this discovery we now know there could be truth in these legends at least on Flores. Their legends are not just of the ancient past either: they speak of Ebu Gogo existing at the time of the Dutch colonisation, several hundred years ago, and possibly as recently as 1900. There are legends of still-extant apelike creatures on other islands, such as the orang pendek of Sumatra.

Scientists have also dubbed these creatures hobbits, a name that is sure to stick in the headlines.

Indonesian and Australian archaeologists working in limestone caves at Liang Bua found a one-metre female specimen, dubbed LB1, in September 2003. At first they assumed it was a juvenile, but wear on the teeth and growth lines on the skull confirmed it was adult. It walked upright, and had longer than expected arms. The arms haven't been fully studied yet, but may indicate it also used the trees. As Flores is home to the ravenous Komodo dragon, hiding in trees might have had survival value.

LB1 was dated to 18 000 years ago, but later discoveries of more individuals has brought the date of occupation closer. The remains are recent enough that they are good candidates for retrieving DNA to analyse their place in our family tree: it is speculated that they evolved from Homo erectus, which they resemble in skull and tooth shapes. The earliest fragments found are between 95 000 and 74 000 years old. There was a volcanic explosion on Flores 12 000 years ago: if they went extinct then, around about the time of the latest known fragment, that and competition from modern humans might have done them in.

The caves contain a few stone tools dated from about the same period, and some animal bones show signs of charring by fire. There is at present no evidence that contemporary Homo sapiens inhabitants of Flores used the caves at the same time. If not, H. floresiensis shared these two sapient abilities.

The brain size of H. floresiensis was (or is, if they're still lurking in the dense unexplored forests of Indonesian islands) about 380 cm3, like that of a chimpanzee, or a modern human baby. (An adult human brain can be four times as big.)

Islands produce strange size effects. You get both giants and pygmies. Cyprus had its pygmy hippopotamus and elephant, Wrangel Island its pygmy mammoth, but Mauritius had the dodo and Flores and Komodo have their gigantic top-predator lizard.

Later, there has been the inevitable casting of doubts on the original results. Some people have suggested the type specimen, the only reasonably full skeleton, is that of a microcephalic Homo sapiens, and the other remains are within sapiens range of sizes. There is also a bit of kerfuffle about who actually gets to keep the bits. New analysis in September 2005 of further remains of varying ages seems to confirm that the specimens do constitute a dwarfish race, not a few microcephalic individuals who happened to survive.

New Scientist, 30 October 2004
New analysis, September 2005: http://news.bbc.co.uk/2/hi/science/nature/4331252.stm

This was written for a class I took last year on prehistoric human evolution, and as such, it's hardly my opus work. In fact, I almost wanted to edit into oblivion, but I settled for scattered notes in the text to explain myself as I went along.

Those Tricksy Little Hobbitses: the Argument over H. floresiensis


  1. The Discovery

    1. Peter Brown and his team of Indonesian and Australian researchers found a fairly-complete H. floresiensis skeleton in a cave called Liang Bua on the island of Flores; they designated this specimen ‘LB1’.

    2. Soon afterwards, they found a second much less-complete skeleton, called ‘LB2’.

    3. Not long afterwards, they found the scattered remains of seven more individuals, duly noted as ‘LB3’ through ‘LB9’. Of all the remains, LB1 was the most complete, and only LB1 had an intact cranium.

  2. H. floresiensis is an australopithecine

    1. Stature (~1 m) and cranial capacity (~320 cm3) are similar to the earlier australopithecines.

    2. The anatomy of the carpals bears some resemblance to earlier australopithecines as well. (note: basically, their hands kind looked like australopithecine hands, with long, curved bones ideal for a lifestyle spent partly in trees, and partly on the ground, like many researchers believe that the earlier australopithecines lived)

    3. But — if H. floresiensis were an australopithecine, then how did it wind up all the way in Australasia? (note: as of right now, most researchers believe that the first hominid to leave Africa was Homo erectus; however, many are starting to think that there were earlier waves of migration from the continent, and my professor told me during office hours that this would become textbook knowledge in a decade or two --- stay tuned!)

  3. H. floresiensis is a Homo erectus

    1. The thickness and shape of the cranial vault suggest a relationship between H. floresiensis and H. erectus

    2. Humeral torsion is much lower than that of H. sapiens (at 110°, where the average for H. sapiens is closer to 130° - 140°); this is more similar to the Nariokotome Boy than the average for a typical modern specimen of H. sapiens. (note: this basically means that the upper arm bone, the humerus, showed a shallower angle between the shoulder joint and the elbow joint. Assuming you, Dear Reader, are an anatomically-modern Homo sapiens, your shoulder and elbow should form roughly an obtuse angle when viewed from above while you're doing your best cadaver impression. With H. floresiensis, the angle was much shallower, almost a right angle.)

    3. H. erectus was definitely able to cross bodies of water (how else could it get to Java?) (note: many think that H. erectus was capable of making rafts; they certainly used tools. Others think they drifted in a storm, hanging on for dear life onto a piece of driftwood and being washed ashore on Flores. The distinction blurs somewhat when you realize that H. erectus really wasn't likely to be the best of engineers, splendid hunters and travellers though they were.)

    4. H. erectus was the first hominid to definitively leave Africa (though there is mounting evidence to the contrary — stay tuned!)

    5. But — (note: I have no idea why I left this blank, but what I meant to say was something along the lines of 'but gosh, gee whiz, H. floresiensis sure don't look like a H. erectus!' A waffle of an answer, to be sure.)

  4. H. floresiensis is a Homo sapiens with a serious case of secondary microcephaly

    1. Microcephaly is, simply put, the cranial vault being significantly smaller than average. Primary microcephaly appears before birth and almost always results in stillbirths; the most extreme example of this is called ‘anencephaly’, and results in the complete absence of a cerebrum or cerebellum. Secondary microcephaly is the failure of an otherwise-normal brain to reach full adult size, and usually results in profound mental handicapping.

    2. The encephalization of H. floresiensis is lower than both H. erectus and H. sapiens (and lies within the lower range found for australopithecines), even accounting for insular dwarfing and the allometric effects thereof. (note: insular dwarfing is where mammals show smaller average body sizes on islands than they do on mainland areas; the opposite trend is true for reptiles, which is why you see Komodo dragons and monitor lizards on islands, but only tiny British-accented geckos on the mainland)

    3. But — Evidence of complex cultural behavior (i.e. cooperative hunting of Stegodon) (note: a species of elephant) and tool-making make H. floresiensis seem a lot more intelligent than its cranial capacity would suggest; indeed, the encephalization ratio (note: brain-to-body-size ratio) lies somewhere between the great apes and H. erectus, both of which show examples of tool-making and complex cultural behaviors.

    4. And also — analyses of a brain cast of LB1 shows an enlarged frontal lobe in proportion to the rest of the brain, which is something entirely anomalous — secondary microcephalics have more-or-less ‘normal’ cerebral proportions. This may account for how H. floresiensis could engage in such complex behaviors and have such a small encephalization quotient. Others suggest that the microstructure of the brain of H. floresiensis was different, allowing it to engage in complex mental behaviors without needing quite so much cerebral hardware.

  5. H. floresiensis is something entirely different

    1. H. floresiensis’ wrists show traits primitive to before even the split between African apes and humans. This is unlikely to have been the result of allometric changes due to their adaptations to a resource-poor insular environment; also, this is unlikely to be due to some sort of congenital deformity in the specimens analyzed.

    2. H. floresiensis exhibits a confusing mosaic of traits both primitive and derived, and there are aspects of its dentition and postcranial skeleton that are autapomorphic. While may of these traits can be explained as the allometric effects of insular dwarfism, some of them (such as the aforementioned dentition) cannot be quite so easily dismissed.

    3. The paleoanthropological community has agreed that this classification for H. floresiensis is the most likely one (though the case is still very much open).

  6. H. floresiensis and the Palau specimens

    1. Very recently, Lee Berger and a team of researchers happened upon a pair of insular-dwarfed H. sapiens skeletons on the island of Palau in Indonesia. While they show some traits reminiscent of H. floresiensis (such as small size and similar cranio-facial proportions), they also lack some traits found in the Liang Bua finds (such as iliac flaring and the dental and carpal abnormalities).

    2. The findings were published in a rush, and further analysis on the finds needs to be done to better compare the Palau finds to the Flores finds. As it stands right now, the Palau finds are fairly typical for island-dwarfed H. sapiens in a tropical environment, and Berger does not believe that they have any close link to H. floresiensis.

Discovering Homo floresiensis

Ever since Peter Brown and Indonesian-Australian research team discovered in 2003 the remains of what would become known as Homo floresiensis — or the 'Hobbits'the academic community has been embroiled in a fierce debate as to just how to place this diminutive hominid in the hominid phylogeny. Some classed it with the australopithecines, some believed it was a subspecies of Homo erectus, and some held that the sole complete cranium ('LB1') of H. floresiensis that was found was that of a modern Homo sapiens with a serious case of secondary microcephaly. But whatever H. floresiensis was, it represented a tantalizing side branch of hominid evolution, co-existing with modern H. sapiens until very recently in both time and territory — present estimates based on available fossil evidence place the extinction of H. floresiensis at a scant 18,000 years ago (Brown 2004), and maybe even as recently as 13,000 years ago, based on stone tool evidence (Morwood 2004). The H. sapiens populations on Flores arrived anywhere between 55 to 35 kya, overlapping with H. floresiensis for at least 23 kya, and most likely even longer than that (Brown 2004).

The discoveries were made in a cave called Liang Bua, located 25 kilometers from the coast of Flores. Evidence has been found on the island that indicates that the island may have been inhabited by hominins since at least 840 kya (Morwood 2004). The cave itself was the final resting place for one specimen of H. floresiensis (designated specimen ‘LB1’), dated to approximately 18 kya, as well as that of another very incomplete specimen designated ‘LB2’ (Brown 2004). Later on, remains of seven other specimens had been recovered in varying degrees of completeness, designated 'LB3' through 'LB9' (Morwood 2005). The degree of articulation still present in the skeletal remains of LB1 indicated that it was buried in fine river silt very soon after death, something consistent with evidence that the cave was flooded regularly by the nearby Wae Racang river in the past. Found along with the skeleton were the remains of several small fauna, including fish, frogs, snakes, tortoises, varanids (including Komodo dragons), birds, rodents and bats, as well as the skeletal remains of dwarf Stegodons (a species of elephant) that inhabited the island. Surprisingly, several of the bones found were charred, something very unlikely to happen spontaneously on a cave floor, indicating that H. floresiensis could very well have had the ability to cook food with fire (Morwood 2004).

Handy Little Hobbits

Additionally, a dense concentration of prepared-core stone tools (more than 5,500 per cubic meter of stratum in one area) was found surrounding the skeleton, with a date range of between 102.4 kya and 0.6 kya, but it was unable to be determined with certainty which species manufactured these artifacts. Very few tools were found in Sector VII, near LB1, but the greatest majority of the tools were found in Sector IV. Tools found in older strata were made by a much heavier and cruder hammering method than those found in newer strata, and the tools themselves were found in association with juvenile Stegodon remains, suggesting that whatever wielded these tools were engaged actively in hunting (Morwood 2004).

Was H. floresiensis an australopithecine?

The differences and similarities that H. floresiensis had between themselves and other hominid species are many-fold and confusing, and sparked a very lively debate. H. floresiensis seemed to be a mixture of features between australopithecines, H. erectus and H. sapiens. Physical stature suggested that it was an australopithecine adapted to a tropical climate and an island environment, as did its endocranial volume (which overlapped with the lower range of cranial capacities found in Australopithecus, and was basically equal to Pan troglodytes and some of the earliest hominids) (Falk 2005). But it lacked the robust masticatory adaptations seen in australopithecines and paranthropids, and had substantially reduced prognathism and facial height relative to both genera. Furthermore, it had smaller post-canine teeth and overall reduced mandibular and maxillary dimensions. Cranial capacity and stature alone are insufficient to conclusively classify these 'Hobbits' with Australopithecus.

Similarities with H. erectus

H. floresiensis had substantial similarities to H. erectus, however, given the similarities in cranial shape and thickness shared between the two species. However, the sheer thickness of the cranial vault resulted in a much-reduced brain volume relative to australopithecines and early Homo, especially when compared to similar Plio-Pleistocene hominids with similar external cranial vault dimensions, and the morphology of the mandibular symphisis was more like the LH4 A. afarensis than the Zhoukoudian and Sangiran H. erectus specimens (Brown 2004). H. erectus certainly had the capability to reach Flores, as it was well-established in Asia at the earliest horizon of the 'Hobbit' habitation of the island; furthermore, H. erectus must have had the capability to cross bodies of water, as some of the first fossils of this species were found on Java. H. floresiensis had an encephalization quotient roughly similar to H. erectus, especially when we account for its small stature.

H. floresiensis and microcephalic H. sapiens

Indonesian anthropologist Teuku Jacob and his colleagues claim that H. floresiensis was an island-dwarfed H. sapiens with secondary microcephaly, pointing to the evidence exhibited in the cranium. Indeed, H. floresiensis does show numerous cranial asymmetries that cannot be explained away as damage sustained during deposition and excavation (Jacob 2006).

Secondary microcephaly is simply a failure of the brain to grow to full adult size after birth, as opposed to primary microcephaly, which is a medical condition in which the brain is significantly below average size at birth. Secondary microcephaly would result in some distortion of craniofacial features, but would result in the brain being more-or-less of the same proportions of a very young human; this almost always results in profound mental redardation (Niven 2006).

Others who support the idea that H. floresiensis was a H. sapiens with secondary microcephaly (such as R. D. Martin) point to the proportions of the brain to the rest of the post-cranial skeleton, which falls outside of the normal range associated with H. sapiens, in such a way as to be highly unlikely to result from typical insular dwarfing (Martin 2006).

H. floresiensis was not a microcephalic H. sapiens

Opponents to the idea that H. floresiensis was a microcephalic modern human point out that there are a number of anatomical anomalies, especially in the dentition (which show the P3 with an enlarged occlusal surface as compared with H. sapiens) something that would be rather unlikely with recorded cases of secondary microcephaly, which do sometimes result in cranial deformation, but would unlikely result in significant changes in dentition (Argue 2006). There is also the consideration of the complex cultural behavior associated with H. floresiensis, which included tool-making and the controlled use of fire. While there are cases aplenty of modern microcephalic H. sapiens with intelligence quotients that fall within the normal range, certainly a microcephalic with an encephalization as tiny as that of H. floresiensis would almost certainly suffer from profound mental handicapping. Conroy and Smith suggest that perhaps the microstructure of the brain of H. floresiensis is such as to cause it to fall out of the typical range of variation for primates — a paradox, to be sure, but it is perhaps the only way to explain the low degree of encephalization found in H. floresiensis and the complex cultural behaviors associated with them. In short, Conroy and Smith state that 1 cm3 of H. floresiensis brain would necessarily be somehow 'denser' than 1 cm3 of brain matter from their nearest hominid relatives (Conroy and Smith 2006). Niven also analyzed scans of the cranial vault of H. floresiensis, and found that they exhibited an enlarged frontal lobe in proportion to the rest of the brain, something that would not be expected in a typical secondary microcephalic (Niven 2006). Paradoxical and contradictory though these findings may be, the gross features of H. floresiensis' cranium could not simply be explained away as mere secondary microcephaly; there is a host of other less-noticeable features that cannot be so dismissed.

Could H. floresiensis be Something Else Entirely?

Furthermore, when we analyze the available data on H. floresiensis on its own terms, we come to see a number of surprising facts. For one, the wrist structure is like neither Australopithecus nor Homo! Indeed, the three carpal bones analyzed in the hands of H. floresiensis appear to have traits primitive to before even the ape-human split (Tocheri 2007). Parsimony would suggest that H. floresiensis would have been descended from a species that split off and radiated before the ape-human split, making it unlikely to be descendants of populations of Australopithecus, H. erectus or H. sapiens that colonized Flores. It would be something entirely different, a sentiment that both M. J. Morwood and Peter Brown agree with. Certainly, the haphazard smattering of traits exhibited in the skeleton of H. floresiensis suggests that it is something that is neither Australopithecus nor Homo.

This notion is buttressed somewhat by some evidence in the post-cranial skeleton of H. floresiensis. The fossils' arms exhibit such a low degree of humeral torsion as to be more like that of apes and the earliest ancestors of humans, than they would of more recent hominids (Larson 2007). The pelvis shows some traits that are seemingly autapomorphic, such as an iliac flare (Brown 2004). The femur is much straighter and has less-developed attachment points for the large muscles of the leg, and both the overall femoral length and high bicondylar angle fall within the range of Australopithecus and Pan, even allowing for allometric effects related to the small size of the specimens (Brown 2004). It would seem that if we wanted to place H. floresiensis anywhere on the hominid line, it would make the most sense to place it fairly early on, certainly before Homo.

It is not beyond the realm of possibility that hominid populations left Africa before H. erectus; however, any evidence for such migrations would be nearly impossible to find, as the most likely routes for migration would have fallen victim to the slings and arrows of geology and changing sea levels. Any speculation on this topic is merely that — pure speculation — until we find more definitive evidence that H. floresiensis is a representative of a species that predates even the African ape-human split. So far, the notion that H. floresiensis predated the ape-human split is supported by an analysis of only three carpal bones, flimsy evidence at best (but conclusions of similar magnitude have been drawn on even less skeletal evidence).

However, there are other features that can be brought to bear in support of the idea that H. floresiensis forms its own clade. For one, the brain itself yields some interesting features. An analysis of the cranial vault of LB1 shows that it had a very convoluted frontal lobe, especially in the areas associated with self-awareness and other higher mental traits more often associated with late Homo (Falk 2005, Niven 2006). Despite the fact that H. floresiensis had a brain more comparable to chimpanzee than human being, the encephalization quotient places it somewhere in the mid-range between the great apes and H. erectus (Falk 2006). 'Hobbit Man' was no mere ape; it might well have been able to fashion stone tools and engage in cooperative hunting, and then bring the meat back to cave homes and cook it with fire. The general consensus in the anthropological community is that H. floresiensis is an entirely new and separate species of hominid, with a host of primitive features that place its divergence on the evolutionary timeline back to perhaps even some time before the split between African apes and hominids. This is a conclusion I am inclined to agree with, since the plesiomorphies seen in H. floresiensis seem to be numerous enough and of such a character as to be unlikely to have resulted from developmental abnormalities or the distortion in bodily proportions concomitant with island dwarfing.

Other Dwarfed Hominids

On this token, it is worth noting the very recent discovery by Lee Berger and his team of researchers of a pair of island-dwarfed H. sapiens fossils found on the Indonesian island of Palau. It is worth noting this development, because it gives us a clear picture of what a H. sapiens might typically look like under island dwarfing; while it does share some characters in common with H. floresiensis, such as similar craniofacial dimensions and especially cranium-to-body ratios, it lacks some features autapomorphic to H. floresiensis (such as a distinct flare in the ilium, or some dental traits unique to the specimens uncovered in Liang Bua) (Berger 2007).

Works Cited

D. Argue, et al.

2006 Homo floresiensis: Microcephalic, pygmoid, Australopithecus, or Homo? Journal of Human Evolution 51:360-374.

P. Brown, et al.

2004 A new small-bodied hominin from the Late Pleistocene of Flores, Indonesia. Nature 431:1055-1061. Nature Publishing Group.

Berger, Lee R., et al.

2007 Small-bodied humans from Palau, Micronesia. PloS ONE 3(3):e1780.

Conroy, G.C. and Smith, R.J.

2006 The size of scalable brain components in the human evolutionary lineage: With a comment on the paradox of Homo floresiensis. Department of Anatomy and Neurobiology, Washington University Medical School, St. Louis.

Falk, Dean, et al.

2005 The Brain of LB1, Homo floresiensis. Science 308:242-245.

Dean Falk, et al.

2006 Response to Comment on “The Brain of LB1, Homo floresiensis”. Science 312:999c

T. Jacob, et al.

2006 Pygmoid Australomelanesian Homo sapiens skeletal remains from Liang Bua, Flores: Population affinities and pathological abnormalities. PNAS 103(36): 13421-13426.

S. G. Larson., et al.

2007 Homo floresiensis and the evolution of the hominin shoulder. Journal of Human Evolution 53:718-731.

R.D. Martin, et al.

2006 Comment on “The Brain of LB1, Homo floresiensis”. Science 312:999b

M.J. Morwood, et al.

2004 Archaeology and age of a new hominin from Flores in eastern Indonesia. Nature 431:1087-1091.

M.J. Morwood, et al.

2005 Further evidence for small-bodied hominins from the Late Pleistocene of Flores, Indonesia. Nature 437:1012-1017

Niven, Jeremy E.

2006 Brains, islands and evolution: breaking all the rules. Smithsonian Tropical Research Institute, Apartado 0843-03092, Balboa, Ancón, Panamá, República de Panamá.

Tocheri, Matthew W., et al.

2007 The Primitive Wrist of H. floresiensis and its Implications for Hominin Evolution. Science 317:1743-45.

Log in or register to write something here or to contact authors.