Researchers announced today that an excavation in Liang Bua cave on the Indonesian island of Flores has uncovered a new species of human, barely a meter tall, that lived as recently as 13,000 years ago. Christened Homo floresiensis, the hominid--known primarily from a partial skeleton known as LB1--had adult body and brain proportions comparable to those of the much older australopithecines, such as Lucy. Other features, however--including those related to chewing and walking--align it with our own genus, Homo. Describing the find in the October 28 Nature, Peter Brown of the University of New England in Armidale, Australia, and his colleagues surmise that H. floresiensis was a descendant of H. erectus. With H. sapiens arriving in eastern Asia by 35,000 years ago, and relic populations of H. erectus possibly persisting on nearby Java, three human species may have co-existed in this region not so long ago. Scientific's editorial director, Kate Wong, spoke with Brown about the discovery. An abridged, edited transcript of their conversation follows.

KATE WONG: What brought you to Flores in the first place? I assume you weren't expecting to find something like this.

PETER BROWN: The work in Flores was initiated by Mike Morwood, who's in in my department [at the University of New England in Armidale, Australia]. He has a project looking at the evolution and migration of people through the Indonesian archipelago and in Australia. He was interested in Liang Bua, which is a cave on Flores, because of previous work by R. P. Soejono [of the Indonesian Center for Archaeology in Jakarta]. It's a beautiful, large rock shelter and it clearly had a deep deposit. But previous work had only looked at the upper Mesolithic layer--no one had tried to dig to the bottom. So Mike went back with a large team, as well as with Professor Soejono, and dug down to bedrock, in the course of which they found lots of stone tools and eventually came across this hominid skeleton.

KW: There's been a fair amount of controversy about the ages of various human fossils from Southeast asia. How did you date LB1 and how confident are you in the ages that you obtained?

PB: There's a large team of experts in various fields of dating--including radiocarbon dating, optically stimulated luminescence, uranium series, electron spin resonance--and these technniques were applied to the site. Fortunately, the results are consistent. The carbon found in close association with the skeleton is about 18,000 years old. And there are optically stimulated luminescence dates above and below, and a series of other dates going down toward the bottom of the deposit as well. They all appear to be consistent and nicely stratified, so you have a range from the bottom of the site at about 94,000 years ago, give or take a little bit, to the skeleton at about 18,000, again give or take some. So we're happy with the dating results.

KW: LB1 has such a tiny body and brain--as small as the smallest australopithecine on record. Did you ever think about assigning it to a new genus rather than just a new species?

PB: In the initial letter to Nature, that's exactly what I did. But I was persuaded on further rumination that it should be placed in the genus Homo. On initial examination I was impressed by the features it appeared to share with early [hominids] like Australopithecus, but other features were more like Homo. For instance, australopithecines have large, projecting facial skeletons (large molar and premolar teeth in particular), whereas the face of LB1 is much more similar to members of the genus Homo. So you have this very humanlike looking face stuck with this very, very small braincase--a brain size which would be small for a chimpanzee. Looking at the rest of the skeleton, it had a combination of things we would consider to be humanlike features combined with things which are found in some australopithecines. I decided that some of the similarities with australopithecines were probably due to small body size and the biomechanics of locomotion, rather than just representing phylogeny. In other words, it wasn't an australopithecine located in Asia. And other things, like thickened bone in the cranial vault and the shape of the brain case, are more like those in members of the genus Homo than like Australopithecus. So I dismissed Australopithecus for a variety of reasons and then did a balancing act. And when we thought that it was most likely a dwarfed example of Homo erectus, then I leaned toward putting it in the genus Homo rather than creating a new genus.

KW: How do you know that LB1 isn't simply an aberrant individual?

PB: There are small-bodied normal modern humans just as there are small-bodied abnormal modern humans. Small-bodied normal modern humans, normally called pygmies, have small stature, but they have brain sizes and skull proportions which are similar to those of the large-bodied populations around them. So it's easy to rule out the pygmy analogy. It's more difficult to rule out, I suppose, the analogy with abnormal modern humans, like pituitary dwarfs or microcephalic dwarfs, because there you can have small-bodied people who have small brain sizes as well. Very few of these people actually reach adulthood and they have a range of distinctive features, depending upon which particular syndrome they have, throughout the cranial vault and rest of the skeleton. None of these features are found in Liang Bua. It has a suite of clearly archaic traits which are replicated in a variety of early hominids and these archaic traits are not found in any abnormal humans which have ever been recorded. We now have the remains of 5 or 6 other individuals from the site, so it's not just one. There's a population of these things now and they all share the same features.

KW: The overall trend in human evolution is one characterized by increases in body size and brain size. How do you explain this apparent reversal?

PB: There's no actual evidence for dwarfing [in Homo] on the island. We don't have any putative large-bodied ancestors on Flores for [the Liang Bua people] to have dwarfed from as yet. There's evidence of stone tools on Flores at 840,000 years ago, which was reported a couple of years ago. But from the cave itself we only have these small people. So there are a couple of possibilities. Either they dwarfed on another island and then proceeded to get to Flores one way or another; or they may have dwarfed on Flores; or maybe this lineage has greater antiquity--for which there is no evidence at present--and they represent an earlier migration, before Javan Homo erectus entered that part of Asia. In other words, an earlier migration out of Africa via Europe or wherever.

Looking at the distribution of small-bodied animals around the world today, they tend to occur in rainforests. [Editor's note: being small bodied enables animals to better regulate their body temperatures and to subsist on fewer calories, which rainforests have in limited supply.] And certainly that's where small-bodied humans tend to be found. We don't know much about the paleoenvironment on Flores yet, but everything's consistent with it being heavily rainforested back in the Pleistocene and probably heavily rainforested until agricultural humans arrived and started clearing the rainforest. The fauna is consistent with that sort of environment as well. Maybe there just wasn't a lot to eat. The island is only about 14,000 square kilometers, there's not a lot of it there. So I think the most likely scenario is that as part of their adaptation to [having fewer] calories living in a rainforest--and maybe thermoregulation as well--there was this long-term selection for smaller body size. There may be other possibilities, but this seems most likely to me at the present.

KW: What other sorts of animals lived on Flores when LB1 was alive?

PB: The faunal analysis is still ongoing, but one of the most common items in the deposit is this pygmy stegodon [an extinct elephant relative]. They tend to be baby stegodons and baby stegodons probably didn't normally live in caves, so we assume they are the result of hunting. Stegodon underwent several periods of introduction and dwarfing on Flores. They came in as the normal size stegodons (sort of an Indian elephant size), they dwarfed, became extinct for whatever reason--maybe tectonic activity, volcanoes, we don't really know--and then subsequently large stegodons came in again and dwarfed. We have these dwarfing episodes reflected not only in Liang Bua but other deposits on Flores. So there are lots of baby stegodons in the deposit and very few adult stegodons--maybe [the baby stegodons] were tastier. There's also a Komodo dragon, another large [monitor lizard], very large rodents, bats, birds and a range of other things, some of which occur naturally in caves. The main evidence of human activity, I suppose, is the Stegodon remains.

KW: Do you have any prediction about what the males of H. floresiensis will look like?

PB: I wish they'd find one, because that would make it more secure that [LB1] was a female. I'm reasonably happy with it as a female, but it'd be nice. I suspect they had fairly large canine teeth. I assume the males were larger-bodied and more heavily muscled. And you'd expect them to have a more substantial browridge and a larger, more projecting facial skeleton. In terms of stature, [the male would be] maybe 8 to 10 percent taller than the female.

KW: The LB1 find greatly extends the known range of variation in the genus Homo. Do you think this could force some researchers to rethink what's going on further back in the family tree with regard to such African genera as Kenyanthropus and Paranthropus? In light of this discovery, might those genera be better subsumed under Australopithecus?

PB: I think that's probably the case. One of the interesting things about my subject is it's only as good as the next discovery, and the outstanding feature of the broad subject during the last decade is more and more fossils and greater and greater variation. So far, the material we have from Liang Bua is fairly consistent, not showing much variation, [which is] what you would expect from a local population. But I think given the differences between Liang Bua and things like Dmanisi [H. erectus from the Republic of Georgia] and the Javan [H. erectus], people are going to have to think a lot more about what produces variation. These Asian sites--and we do expect to find more--will start to put the African things in a better context, I suspect.

KW: Paleoanthropologists tend to disagree over how bushy or neat the human family tree is. How do you see it?

PB: My view on family trees changes daily, like everyone's does. When I first was educated I went for a fairly simple family tree, but then there weren't so many fossils around at the time. But I'm more willing now to accept a more complex family tree. That was important as far as Liang Bua goes because I had no problem at all putting it in a new species and making the tree more bushy, if you like. You can take dwarf examples and stick them into new genera or just create new species, and zoologists go both ways. But it just seemed on balance more sensible to stay within the genus Homo.

KW: I want to talk a little bit now about what the Liang Bua people were doing. You have evidence that they were making points, blades, and microblades for hafting. How does this tool kit compare with that of Homo erectus?

PB: It's strikingly different from everything ever found with H. erectus. Apart from this short, small-brained thing surviving until 14,000 or 18,000 years ago, its association with these stone tools is the other most remarkable thing about the site. It's something the critics will take a very close look at because there are three possibilities. Either Homo sapiens, of which there is no [fossil] evidence, was making these stone tools; or this small [hominid] learned to make the stone tools from H. sapiens in some way; or it was actually making the tools itself. I think it was making the stone tools. We have the same tools going from 94,000 years ago until 18,000 or 14,000 years ago--no change in technology, no change in materials, consistent from the bottom to where the skeleton was found. So on account of the evidence, the association seems fairly clear. Maybe something else will turn up at Liang Bua in the end, but at present there is nothing. The only [hominid] we're finding in association with stone tools is this small thing.

Now, the stone tools are consistent with those made by Homo sapiens on various parts of the planet, but they're very different from what Homo erectus makes anywhere. Homo erectus just doesn't make tools like these. Of course in Asia, particularly East Asia and Southeast Asia, there's little association between Homo erectus and stone tools anyway. The tools are also different from those of Zhoukoudian H. erectus [in China], at least for most of the deposit. So the stone tools are surprising, they certainly aren't what you would have predicted. But the association between the tools and the skeletal material seems fairly clear.

KW: Obviously you were surprised to find a small-brained hominid making tools like those of large-brained Homo sapiens. What does this tell you about how brain size matters where intelligence is concerned? Does brain size matter relative to body size, or is it absolute brain size that is important?

PB: I think brain size in relation to body size is important because the brain has to control the body. But I think the most important thing is how the brain is wired up, how information is transferred around the brain. We can't get at this with any fossil [hominid]. We can get a gross morphology of brains from endocasts and CAT scans and things like that, but we can never determine how the brain was wired. And it seems to me that [in the case of] Liang Bua--if it was the toolmaker, and I believe it was--the important difference between it and perhaps some other [hominids] is not so much in the brain size/ body size relationship but how the brain was wired up.

KW: Do you think it's possible that H. floresiensis or other as-yet-undiscovered dwarfed human species could have contributed genetically to some of the small-bodied people who live in rainforest regions in this part of the world today?

PB: There are small-bodied people living in Melanesia and New Guinea and some of the islands in Southeast Asia, but they're just small modern humans. All of the features in the cranial vault and the skeleton reflect that they are just small modern humans. And we know from the genetic evidence and archaeological evidence when they got there and where they came from to a reasonable degree of precision. I can't see any way at all that they have anything to do with the Liang Bua people.

KW: How deep are the waters surrounding Flores?

PB: The channels are very, very deep--way more than 100 meters. Maximum sea level declination was about 80 to 100 meters in the region, but some of these channels are much, much deeper than that. One of the interesting things about these people is how the hell they got to Flores. We know Stegodon, like elephants, could swim large distances. Elephants are occasionally found five to 10 kilometers out to sea, and Flores is not so far from Komodo, and Komodo is not so far from the other islands in the chain. So you can easily think of elephant relatives swimming across here. But the currents between some of these islands are very, very strong, particularly between Flores and Komodo, and between Komodo and the other islands. The queston is whether [early humans] used watercraft, whether it was an accidental or intentional process, or whether in fact tectonic activity meant that at some time in the past there was a land bridge which isn't there today. At present there's no evidence of a land bridge, but you can't rule it out as a possibility.

KW: That's really interesting, especially if you have people on Flores by 840,000 years ago, which is what archaeological remains from other sites suggest.

PB: Yup. Java is not a problem, it was joined to the Indonesian mainland. But on current evidence these small islands never were.

KW: Certainly by the time of LB1, people had well-developed language and oral traditions. Might encounters with humans like LB1 and other, unknown, island forms be the source of mythologies that are so widespread in the world today involving very tiny people and very large people?

PB: You pick a country and there's either large Bigfoots and Yetis, or small leprechauns and Yowis, depending on which part of the world you're in. On Flores there is a mythical humanlike animal called Ebu Gogo, known for small body size, inarticulate speech and an unusual bipedal gait. Every country seems to have myths about these things. We've excavated a lot of sites around the world and we've never found them. But then last September we found LB1, so surprise, surprise. There's got to be some overlap in time between [Homo floresiensis] and Homo sapiens. But in terms of oral histories and racial memories and those sorts of things, that's still a very long time period. Whether people are thinking these things because at one time or other someone actually excavated or found the remains of something small and unexpected I don't know. But the time periods seem too long--it's not as if they're represented in rock art or anything else. Which is not to say there's not some basis. But I would be surprised. I just think humans have amazing imaginations and they see all sorts of things and create all sort of explanations for things which they half see or maybe see. And a fleeting glimpse seems to mean more than it really should.