Java Man by Garniss Curtis, Carl Swisher and Roger Lewin, Little, Brown, 拢18.99, ISBN 0316648604
WE ARE certain as anyone can ever be in biology that modern humans are more closely related to chimpanzees than any other living animal. We are also reasonably sure that it was in Africa that the earliest exclusively human ancestor split off from our common precursor between 8 and 5 million years ago. What is much less certain is when-and in what form-our ancestors first ventured out of Africa. Java Man aims to solve that mystery.
Discoveries of important fossil evidence for human origins quite rightly hit the headlines. For all but the past 100,000-at most 200,000-years, we are utterly dependent on the fossil record for reconstructing the history of human evolution. Without fossil teeth, skulls and limbs, and the prodigious efforts to find and prepare them, the cupboard of palaeoanthropology would be bare. From their size and shape we can extract clues about how many species are represented in the human fossil record. Then we can frame hypotheses about how they are related.
Advertisement
Of course, if all the fossil evidence had come from undisturbed rock layers at the same place, we would have little difficulty working out the temporal order of the species. You just start at the bottom and work upwards. But although some of the famous early fossil sites in Africa, such as Koobi Fora, Sterkfontein and Hadar, have each contributed hundreds of specimens to the human fossil record, their strata are rarely neatly layered. How do you translate those sequences of fossils into dates?
A date for the rock can be a reliable proxy for the geological age of the fossil. As potassium undergoes radioactive decay, much of it is converted into argon. The rate of this reaction gives us the potassium/argon (K/Ar) method of dating rocks.
Forty years ago, Louis Leakey invited Garniss Curtis-an unsung hero of scientific palaeoanthropology-and Jack Evernden to test out the K/Ar method on the potassium-rich layers of volcanic ash in Olduvai Gorge, at the edge of the Great Rift Valley. It provided the first absolute age, a then scarcely credible 1.8 million years, for an early human fossil from Africa. Now a single rock crystal can yield reliable dates.
Curtis abetted by his prot茅g茅 Carl Swisher fulfilled a long-standing ambition to provide an absolute date for a child鈥檚 cranium from a site in Java called Mojokerto. They also teamed up with colleagues to date an enigmatic collection of broken skulls from Ngandong. If the Ngandong fossils prove to belong to Homo erectus, and to date from between 25,000 and 50,000 years ago, then H. erectus survived long enough to overlap in time with modern humans.
Java Man resembles a Russian doll: story fits inside story inside yet another story. Roger Lewin is the author, but writes in the 鈥渧oice鈥� of Swisher. Garniss鈥檚 long and distinguished association with K/Ar dating provides the scientific story. He begins with the search for early remains in Java and how these fit into the human fossil record. Then he adds one side of the scuttlebutt about the schism between Don Johanson and Gordon Getty. Their difference of opinion led to the 鈥渄ating鈥� scientists splitting off from those devoted to the search for new fossil evidence. (I have good friends on both sides of the rift, so 鈥渘o comment鈥� from me.)
This enjoyable book skilfully reviews a lot of hard science about dating and palaeoanthropology. It emphasises that we need to pay as much attention, time, energy and intellectual rigour to the context of the human fossil record as we do to the collection and analysis of the fossils themselves.