I have been following the story of Homo floresiensis, dubbed the Hobbit by the media because of its short stature. In 2003 scientists discovered a fairly complete skeleton (skull, jaw, pelvis, arms, legs, hands and feet) on the island of Flores in Indonesia. The skeleton is 18,000 years old and is probably an adult female about 3 feet 6 inches in height and with a brain capacity of 417 cc.

There are other specimens of H. floresiensis, but none with cranial parts. The type specimen described above, designated LB1, is the only good specimen.

There has been a controversy surrounding how to interpret this fossil. In one camp are those who claim it is a modern human with dwarfism and microcephaly – a developmental disorder that results in a small head and brain. They cite as evidence morphological anomalies in the face and teeth.

The other camp maintains that LB1 is a new hominin species, separate from modern humans, deserving of its own taxonomical designation – H. floresiensis. In 2007 Dean Falk published an analysis of a virtual endocast of LB1’s skull – essentially a virtual picture of her brain. He argues that her brain does not have features in common with a microcephalic brain. It looks more like a modern human brain, but also has unique features all its own. In other words – it looks like the brain of a new and separate hominin species.

In May of this year another team published an analysis of LB1’s foot. They concluded that the foot definitely is that of a biped, but is more primitive than modern humans. It is longer in proportion to the femur, for example.

Now another team from Stony Brook University Medical Center has published a statistical analysis of all of the LB1 fossils. Rather than focusing on one feature, they looked at all the specimens and compared them to modern humans and to extant apes and humans with dwarfism. They found that LB1 does not cluster with modern humans or humans with dwarfism, but rather clusters with apes.

This finding is consistent with the interpretation that LB1 is a hominin species that branched off from the human line some time in the past. It now seems that the bulk of the evidence, and multiple independent lines of evidence, are all pointing toward this conclusion and away from the microcephaly hypothesis.

There are still two anticipated lines of evidence, however, that we do not yet have. We need to find more H. floresiensis specimens with skull parts, to confirm that LB1 is a representative specimen and not an anomaly.

Also, because LB1 is only 18,000 years old, and not fossilized (the bones were found in a wet area and are a bit mushy) this creates the potential that we may be able to extract mitochondrial DNA from the bones. If we can (perhaps not likely because of the moisture) then we can compare this to humans, Neanderthals (whose mitochondria we do have) and apes and see where LB1 falls.

Beyond confirmation that LB1 is a new species of bipedal hominins (although not ancestral to modern humans – just a side branch) there are also questions about how H. floresiensis is related to the human line.

In addition to the fossils which are 18,000 years old, there are stone tools that range from 13,000 to 90,000 years old that are associated with the fossils in such a way that they were almost definitely made by and used by H. floresiensis. These stone tools are more sophisticated than those used by H. erectus and are different than those used by modern humans. This also supports the side-branch hypothesis.

Stone tools were found elsewhere on Flores that date to about 840,000 years ago. These could be from an earlier population of H. floresiensis, or they could be from a population of H. erectus. H. erectus lived from about 1.6 million to 400,000 years ago – although specimens in Java could be as young as 50,000 years ago. In any case, this is within the 840,000 year timeline.

The reason to suspect that these tools may be from H. erectus, and that H floresiensis branched off from H. erectus, is that this was the first hominin species to spread out of Africa and across Asia. If H. floresiensis branched off earlier that H. erectus, then that would imply that some earlier hominin species spread out of Africa – something for which there is currently no evidence.

So H. floresiensis may force us to rewrite the story of hominin evolution and radiation – although not really the story of human evolution, because H. floresiensis is not on the line to humans.

The hesitation about simply concluding that H. floresiensis evolved from H. erectus is that LB1 has a really small brain, smaller than H. erectus – even when accounting for the reduction in overall stature. In other words, if you shrink down H. erectus to the height of a H. floresiensis, you don’t end up with the body proportions of H. floresiensis. H. floresiensis has a smaller brain still (about the size of a chimpanzee, but larger in proportion as chimps are bigger overall).

This suggests that H. floresiensis specifically evolved a small brain, while retaining a rather advanced tool-making ability. This has led some to speculate that H. floresiensis had a more efficient brain design – packing more cerebral power into a smaller space.

Overall, a fascinating story, that has much to teach us still. What I await most eagerly is the next high quality fossil specimens of H. floresiensis – that is almost certain to pack some surprises.