Verified by 
Last month (October 2009), Science magazine devoted an entire extraordinary issue to what may be our oldest ancestor: Ardipithecus ramidus. A. ramidus was discovered in the mid-1990s in Ethiopia by a research team led by paleoanthropologist Tim White. Although the team published a couple of preliminary papers in 1994 and 1995, comprehensive description of the fragmentary and fragile fossil remains had to wait until the material was first stabilized and then reconstructed. Several individuals have been identified, although many are represented only by teeth or other small remains. However, a partial skeleton preserving several long bones, hands and feet, the pelvis, and much of the skull was recovered and provides us with an extremely informative view of ramidus biology.
How does ramidus earn the title "earliest human ancestor?" First, there is its age--4.4 million years old. That puts it closer than the australopithecines (such as "Lucy) to the 5-6 million year divergence date between humans and apes suggested by molecular clocks based on genetic analysis. There are other possible earliest hominid candidates, but they are either highly fragmentary or their proposed status as a hominid has yet to be widely accepted. The hominid status of ramidus does not seem to be in doubt. According to White and his colleagues (2009, Science, 326:75-86), the upper pelvis exhibits characteristics consistent with bipedality, which are more fully developed in later hominids, including ourselves. Furthermore, the bones of the foot show an increase in robustness compared to that seen in the great apes, suggesting that it has taken on a propulsive role consistent with walking on two legs. However, in many other aspects, the foot, hand, and limb anatomy of ramidus is quite primitive. For example, although the foot shows signs of being used to push off during heel-and-toe walking, it retains an opposable big toe, thus maintaining grasping ability.
If we were to see a living A. ramidus individual, it would look very much like an ape walking on two legs. But White and his colleagues emphasize that such a resemblance could be misleading: the living African great apes and ourselves all have features that have evolved since we shared an ancestor that looked something like ramidus (although presumably that common ancestor would have been a quadruped). A. ramidus also differs from the later australopithecines in several dental features; in addition, examination of remains from several individuals suggests that ramidus was not very sexually dimorphic. Presumably, both males and females are represented in the sample, but they all seem to be about the same size. This is in stark contrast to the earliest australopithecines, who were highly sexually dimorphic, so much so that there was once vigorous debate about whether Australopithecus afarensis (Lucy's species) fossil assemblages represented one species or two.
Getting more to the point of this entry, the A. ramidus research team, led here by Gen Suwa (Suwa et al., 2009, Science, 326:68e1-7), determined that the cranial capacity of ramidus was somewhere between 280 and 350cc. Since there was no completely intact cranium that allowed the cranial capacity to be measured in a relatively straightforward way, the researchers had to make various linear measures and then use regression equations based on living apes to estimate the cranial volume. This would put ramidus in the lower range of chimpanzees, and somewhat smaller than australopithecines (or gorillas, for that matter, although their increased cranial capacity goes along with substantially increased body size). In broad terms, the cranial capacity is indicative of an "ape-sized" brain, but on the lower end of the spectrum of contemporary great apes, especially given that the estimated body size of ramidus (about 50kg for the one reasonably complete specimen, Lovejoy et al., 2009, Science 326:100-106) is also within the chimpanzee range.
The ramidus discovery fits in well with the picture of brain size evolution that has been pieced together from the human fossil record (see my book, The Lives of the Brain, Belknap Press of Harvard University Press, 2009, for an overview). Although there may have been a modest pattern of brain expansion over the first 4 million years of our evolution, it is increasingly clear that significant brain expansion, transcending the size seen in an ape, does not occur in hominids until about 2 million years ago. The quite small size of the ramidus cranium suggests that there may have been a slight increase in brain size in the transition from Ardipithecus to early Australopithecus. However, the small sample sizes, especially for Ardipithecus, means we should be wary of ascribing too much significance to relatively small (50-100cc) differences in brain size among early hominids. Among contemporary humans, variation in brain size, even within the same sex in a relatively homogeneous population, can easily be of this magnitude (on a percentage basis).
Ardipithecus ramidus, with its small cranial capacity and mix of primitive and derived traits, sits about where it should on the hominid family tree, in both time and space. The same cannot be said for the fossil known as the "Hobbit" (LB1), representing a putative new species, Homo floresiensis. Since the 2004 announcement of its discovery on the island of Flores in Indonesia (by Peter Brown and his colleagues, Nature 431:1055-1061), the Hobbit has drawn an extraordinary amount of popular and scientific attention; it deserves this attention because it is an extraordinary fossil. The Hobbit appears to represent a population of small-bodied (3-4 feet tall) and small-brained (400cc) individuals who evolved in relative island isolation and survived to as recently as 12,000 years ago. The LB1 specimen is the most completely preserved individual, and the only one retaining an intact skull with a measureable cranial capacity.
So despite being separated by 4.4 million years in time, the cranial capacity of the Hobbit is only slightly larger than that of Ardipithecus ramidus. If the Hobbit were an ape, this would not be such a big deal: chimpanzee and gorilla brain sizes have not changed all that much over 6-7 million years. But the Hobbit is clearly a bipedal hominid. The issue is not so much its separation from ramidus in time but in geographical space. Flores is a long way from east Africa. As far as we know, hominids first migrated out of Africa and into Asia starting by about 1.7 million years ago. By this time, they had brains that were already substantially larger than those seen in any ape or australopithecine. There is no evidence of an australopithecine migration out of Africa, and there is plenty of anatomical evidence to establish that the Hobbit is not just a misplaced australopithecine.
This suggests then that Homo floresiensis evolved in its island isolation as an offshoot of Homo erectus (or forms similar to it) or modern humans. Evolving from modern humans seems quite unlikely--there are plenty of relatively small-bodied populations in the world, but cranial capacity is more or less maintained in these groups. The socio-cultural dynamics within human groups probably mitigate against selection for reduced brain size (i.e., too much cognitive competition), even if there were substantial energetic gains to be had from not growing a large brain. But that may not have been the case for Homo erectus. Perhaps a small group of erectus individuals reached Flores, and in the absence of competition from other hominid groups, evolved both reduced body and brain size. Island dwarfism has occurred many times around the world in many mammal species, even on Flores, so it would not be a zoologically unique event.
Another alternative hypothesis has been enthusiastically offered to explain the Hobbit: that it represents a pathological individual, a microcephalic dwarf from an already small-bodied and probably somewhat inbred modern human population (see, for example, Jacob et al. 2006, PNAS 103:13421-13426). Dean Falk led the team that published the description of the Hobbit's skull and brain endocast (Falk et al., 2005, Science 308:242-245), and because many of the claims for pathology focus on the skull, she has also been one of the primary defenders of the Homo floresiensis species hypothesis. In a recent issue of the American Journal of Physical Anthropology (2009, 140:52-63), Falk and her colleagues took on the hypothesis that the Hobbit suffered from Laron Syndrome, a condition resulting from primary growth hormone insensitivity that leads to dwarfism and other distinctive skeletal features (Hershkovitz et al., 2007, American Journal of Physical Anthropology 134:198-208). In the same issue of the journal, Yousuke Kaifu and colleagues (American Journal of Physical Anthropology, 2009, 140:177-185) countered the hypothesis that the Hobbit skull was so asymmetric that it must be pathological (see Jacob et al. 2006). Without going into the anatomical details, suffice it to say that the supporters of the Hobbit-as-new-species concept have reasonable counter arguments to the Hobbit-as-pathological-individual advocates. Falk and her colleagues make the important general point that it is not really helpful to just say that the Hobbit is pathological, but that such claims must be tied to a known and recognizable condition in order to test competing hypotheses.
Cranial sample size is an issue for both A. ramidus and the Hobbit. But since ramidus fits in better with what we know about human evolution, it is a much less critical issue for that species. It would be different if it were claimed that ramidus had a 1300cc brain rather than 350cc--that would be truly unbelievable. The Hobbit represents just the opposite situation, and some researchers have essentially argued that it is unbelievable that it represents a species rather than a pathological individual. I think that no specific pathology hypothesis has been convincing enough so far to dismiss the Hobbit species hypothesis; conversely, a larger sample size of Hobbit crania would go a long way towards convincing me, and a lot of other anthropologists, that there really was a species of small-brained and -bodied cousins of ours living in isolation on an island in southeast Asia. I want to be convinced, since if Hobbits really did live on Flores, the broader story of human evolution would be enriched by their presence, even in a small, supporting role.
