The night before my drive across the Serengeti, I had an amazing dream. As it happened, it was prompted more by the side-effects of the anti-malarial mefloquine than by anything Freud would have recognised, but the form it took was to transport me back to the gates of just such an Eden. Knuckles swinging like an Australopithecus, through a dim, humid forest, I was trying to find a tree within which to take shelter from a driving, monsoon rain. A dense green canopy filtered the light, and from beyond it came the ominous dull rumble of thunder, which eventually woke me up.
If the drug was the trigger, then the mood had been set by a chilling bedtime story. On Saturday afternoon, I had descended alone into the Ngorongoro crater, climbed from the Hyundai to photograph hippos, and was confronted by an adult animal at 20 feet. Figuring - ever the liberal - that I was disturbing the wildlife, I walked slowly back to my car. But over dinner, a celebrity American photographer, Peter Beard, graphically explained that I was a lucky man indeed to be alive to pass him the yams. The kiddies favourite soft toy killed more people than crocs or cats, he said, and had this one seen a glimmer of water behind me, it would have run me down, shaken me to death and pounded my car to scrap.
So I was surging with medicine and the residue of terror when the dream formatted this story. At the heart of the new origins research has been a amazingly recent acceptance among anthropologists that humans were created by, and not merely in specific ecological settings - most notably the great primeval rain forest which once swathed the planet's tropics. To fathom us, the new thinking goes, you don't just go hunting skulls, or watching chimps, you have to start afresh by considering precisely the landscape that had surfaced in my dream.
"If you want to understand why human evolution took the course it did, then you have to understand how hominids articulated with their environment," Tom Plumber, a palaeoecologist from the University of California, Los Angeles, working in western Kenya, had explained to me the previous week. "They weren't evolving in isolation from everything else. They were part of dynamic ecosystems, which were much like ecosystems that we see today."
By this view, only once we have looked over the Garden of Eden's flora do we zero-in on the two-legged mammals. Although the fossil record is as full of gaps as 36 exposures of the D-Day landings, we can tell from bits of skull and bone that the rain forest was home to so many types of monkey (which ran horizontally through the trees), apes (which clambered upright, or swung) and Dart's australopithecines (which analyses of the famous Lucy skeleton from Ethiopia suggest slept in the branches, but waddled awkwardly on the ground), that the features of each shaded into the next: a jumble of evolutionary traits.
Even using such names for countless related hominids creates a misleading cartoon reconstruction, but accepting Australopithecus as my dream's knuckle-swinger from before The Dawn of Man, and Homo erectus as the first step after the sunrise (say, 2-2.5 million years ago), then understanding why one died off and the other took off might explain why in sleep my hands touched the ground and awake they gripped a wheel.
Perhaps another reason for my rainforest dream is that, awake, I kind of envied Australopithecus. Although he got by on a mere 450gms of brains, compared with my possible 1400gms, in the lush, steamy vegetation before today's Serengeti, he didn't have much to use it for except eating, sleeping and sex. Males were vastly bigger than females, and would have been able to live much like gorillas do now: keeping harems grouped near their favourite trees, where food was in good supply.
This was California Dreamin' without the property taxes, and you can bet he didn't bludgeon his way out. Instead, new technologies, analysing such things as cores drilled from Arctic ice-caps and ocean beds, reveal that between about 2.59 and 1.69 million years ago, long-term climate changes would have shocked Australopithecus like an eight-lane freeway down Main Street. Variations in Earth's orbit threw the planet into a cooler cycle - cutting atmospheric energy, and therefore rainfall, and causing swathes of the forest to wilt. First, patches of light woodland thinned the lush canopy. Then came grassland, the dry savannah, like parts of the Serengeti on which I drove.
This change in itself had dramatic outcomes as many plant and animal species went extinct. But the environmental engine of our species' evolution was also moving in another direction. Earth, then celebrating its 4-billionth birthday, was also shrinking, causing its crust to crunch and continents to drift, throwing landscapes this way and that. As earthquakes rumbled around the Pacific zone, a 6,000km fissure tore through east Africa (today's Rift Valley, from the Red Sea to Tanzania's Lake Manyara): creating a hotpotch of faults, buckles, domes, basins, and all kinds of contortions in the landscape.
Geological surveys of sediment beds, sometimes hundreds of metres thick, show that, over thousands of years, local climates ebbed and flowed as mountains rose and valleys sank. Lakes became rivers and rivers became lakes, or sometimes both disappeared. Changing rain shadows and drainage patterns turned forests into woodlands and woodlands into grass, and sometimes back into forests.
The result was thousands of micro-environments, each suited to different styles of life. "The eastern Rift Valley and its surroundings had a mosaic of vegetation types not too unlike those of today," wrote the late archaeologist Glynn Isaac in an essay pasted in my notebooks. "A pattern in which various types of dry thorn savannah and open woodlands predominated, was interspersed with gallery forests along water courses and perennial rivers, and with evergreen forests on the major highland areas. Substantial areas of grasslands occurred on floodplanes and probably on high plateaux. Large tracts of unbroken lowland forest probably did not occur."
Analysis in this chapter in humanity's genesis launched the first big challenge to Dart. It has been clear to specialists for at least a decade that the environmental convulsions would have been quite enough to cut the ground from Australopithecus, without any armed New Enlightenment. With its small brain (only about 50gms bigger than a chimp's) and restricted locomotion, it would, for the most part, have been too stupid to adapt to the stresses of the changes in its habitat.
"The shrinkage of forests at about 2.5 million years ago created a crisis for hominids," wrote Steven M Stanley, in the journal Paleobiology, which I found on the shelves of the Nairobi library. "Most populations should have experienced greatly increased predation pressure. Many (perhaps most) died out, while others abandoned obligate arboreal activities. One population, whether it constituted a fraction or all of the survivors of the ancestral species, evolved into Homo."
Ironically, today's investigations into this transformation are often around sites that the household-name palaeontologists have long since abandoned as barren. At Olduvai, for instance, Robert Blumenschine, professor of anthropology at Rutgers University, New Jersey, is leading a team that is unearthing material that the Leakeys could never have recognised. While the white Kenyans' primarily focused on finding and ascribing ancient dates to headline-grabbing bone fragments, these new ventures are struggling to reconstruct the landscapes and behaviours that would have been seen nearly 2m years ago.
"So much research in the past has been concerned with finding hominid bones and making a case that they are older or more important than anything else," Blumenschine told me. "But we are interested in what the hominids were doing and why. Only once we understand that can we seriously start to tackle questions about why modern humans have the peculiar traits that we do. These are the interesting questions, but they have previously been restricted to the realm of speculation and armchair philosophising because the right kind of information was not available from the fossil record."
Working with scientists from the University of Dar Es Salaam, he is investigating Homo erectus's takeoff around what was once a salty lake. Unlike traditional fossil-hunting, their studies mean digging in huge areas of land, often far from where bones were found, to cross-relate individual stone fragments, fossilised vegetation, pollens and animal remains, past geology and geography, and other clues to what life was like. These are then combined with studies of landscapes around the Rift Valley today.
With similar work at other sites, this method is firstly painting a picture of the end of the rainforest Eden. Among the mosaic of new environments, the most promising sported galleries of acacia, which lined streams and dotted the flood-plains of fluctuating lakes. Rainy seasons spawned short-lived herbaceous plants, but otherwise the ground beneath the trees was bare, and woodlands were surrounded by open grass, only a little more fertile than today.
The next task has been to decipher what was going on in such transformed settings. Enough bone fragments have been found at some east African sites to show that, far from killing each other off, at least three radically different kinds of hominid lived in close proximity for more than a million years. Although Australopithecus's rain-forest harems were wrecked by the need to forage further afield, it clung on in the woodlands until about 1.2m years ago, still sleeping in the trees. And overlapping it in time and place were the creatures evolving into Homo erectus, which spent most of their lives on the ground.
Elementary theory explains the spark for the transition. "At the genetic level, evolution occurs by accumulated substitutions of one nucleotide by another in the DNA of the organism," is how a 1995 paper in Science puts it. "Nucleotide mutations arise with constant probability, but most are lost by chance shortly after their origin. The fate of the rest depends on their effects on the organism. Many mutations are injurious and are readily eliminated by natural selection... Other mutations are favoured by natural selection because they benefit the organism."
Some such mutations would have accomplished a task that every parent today gives thought to. By comparison with our ape cousins, human babies are, in effect, born prematurely, with almost chimp-small brains which then grow at foetal rate in the first twelve months after birth. This unique feature reconciled the narrow maternal pelvis with the big head that was needed to be smart. Whose nucleotide mutations got them there first will never be known, but it was the next step on our path from the forest.