'Hobbits' turn out to be just stunted humans?

Ever since the discovery in 2004 of an 18,000-year-old fossil popularly referred to as the 'hobbit' of Indonesia, anthropologists have remained divided about whether the skeleton represents a species that was distinct from modern humans. The latest turn in the debate comes with the publication in the Proceedings of the National Academy of Sciences of a study suggesting that what was discovered was really nothing more than a deformed human.

A research team led by anthropologist Ralph Holloway at Columbia University compared brain scans and internal casts, or endocasts of skulls from children affected with dwarfism with unaffected children. What they found pointed to the presence of microcephaly, a condition resulting in abnormally small heads.

The researchers found that two specific measurements -- cerebellar protrusion (how far the base of the brain projects backwards) and relative frontal breadth -- could be used to discriminate between microcephalics and unaffected children.

The team then made similar comparisons between endocasts of skulls of 10 microcephalic humans, 79 unaffected humans, 17 individuals of the human ancestor Homo erectus, 4 individuals from the human ancestor Australopithecus, and the H. floresiensis fossil. They report that the H. floresiensis cranium overlapped most with the measurements collected from microcephalics and Australopithecus.

The fossil measurements do not fall within the range for normal modern humans or H. erectus but do fall within the range for microcephalic humans, so Holloway concludes that the fossil could have suffered from microcephaly and is not necessarily a separate species.

Needless to say, that claim has its skeptics. One anthropologist who isn't buying the hypothesis is Dean Falk of Florida State University in Tallahassee. In 2007 Falk concluded that the hobbit skull belonged to a separate species, faulted the research team's methodology. Falk said their use of handheld calipers results in less accurate measurements of rubber endocasts of skulls than she got using computed tomography scan. What's more, Falk argued, Vannucci's team wrongly included some skulls of young children that threw off his conclusions. "We had the opportunity to work with the Homo floresiensis endocast that this team used and passed it by because it was in such poor shape," Falk told Nature News. "We ran a version of this study with CT data that were far more accurate than any data that could be collected from that distorted endocast."