The Incredible Human Journey (12 page)

With the return of cold, dry conditions, the Sinai and Sahara deserts would have spread out again, blocking the northern route out of Africa, and cutting off North Africa from the south. ‘The Middle East is the borderline between Africa and Europe, and it shifted back and forth,’ Yoel said. During wet periods, the border effectively moved north and African fauna (including modern humans) colonised the Levant. In dry periods, the border moved south: the African fauna shrank bank, and European fauna moved in. And that fauna included Neanderthals. ‘They felt comfortable living in the shadow of the glaciers,’ said Yoel. I asked him what he thought had happened to the modern humans in Israel – had Neanderthals pushed them out? He thought not; ‘There’s no evidence for a dramatic scenario.’

Between 90,000 and 85,000 years ago there was a vicious, cold, dry period, known as the ‘Heinrich 7 event’, or H7. Heinrich events are characterised by great icebergs breaking free from ice sheets and floating in the North Atlantic, bringing down the surface temperature of the sea. In southern Asia, these events spell a reduction of the monsoon rains and very dry conditions. Perhaps this is what drove those pioneers out of the Middle East.
¹⁰
Yoel imagined the people living around Mount Carmel migrating, following herd animals heading south as their northern pastures disappeared. ‘We could see them as a population who lived here happily, buried their dead, and then moved on.’

Neanderthal remains from Kebara and Amud in Israel date to around 60,000 years ago. But redating of the Neanderthal fossils from Tabun Cave, just around the corner from Skhul, in the same valley, has shown them to be 120,000 years old, roughly contemporaneous with the modern human burials at Skhul and Qafzeh.
¹⁴
It looks as if, for a while, between 100,000 and 130,000 years ago, the ranges of the ‘African humans’ (
Homo
sapiens
) and ‘European humans’ (Neanderthals) may have overlapped in Israel.
¹⁷
But we’re working with very broad estimates of age here, that don’t allow us to resolve whether the modern humans and Neanderthals were actually contemporaries – they could have missed each other by hundreds or thousands of years. As Yoel Rak put it, ‘Now this is not to say they were sitting in the cave and playing cards. We cannot know even if they saw each other.’ And, in fact, we don’t even know who got there first.

But what is very clear is that the modern human graves at Skhul and Qafzeh represent the earliest known symbolic burials, and that these people were therefore modern in their behavior as well as in their anatomy.
¹⁴
And it is a long time until we see evidence of modern humans in the Levant again.

Some anthropologists talk about the modern human people represented by Skhul and Qafzeh as part of a ‘failed exodus’ from Africa. But although their descendants didn’t spread into Asia and Europe, I think it’s disingenuous to call this a ‘failure’, as there could never have been an
aim
to expand into other continents. We can only see it as a failed exodus with the benefit of hindsight. However, having said that, it seems that this particular expansion did not lead on directly to the colonisation of Asia and Europe.

So that leaves us looking for a later expansion out of Africa. It is possible that the northern route, through the Levantine Corridor, remained a viable option through the cold, dry period. Indeed, it seems that some modern human populations did survive in pockets of habitable environment, or refugia, in North Africa. In 1994, a child’s skeleton was found at Taramsa Hill, on the west bank of the Nile in Egypt. OSL dating of the sands in which the skeleton was found suggested the burial was between 50,000 and 80,000 years old. The skeleton was very fragile, but enough of it – in particular the skull – was present for the anthropologists to be sure that was an anatomically modern human,
¹⁸
and some geneticists have suggested that there is Y chromosome evidence that migrations via this northern route contributed to living populations.
¹⁹

A re-emergence of modern humans from Africa, perhaps around 50,000 years ago, via the northern route, seems to fit well with archaeological and fossil evidence in Europe, but it is too late for the most recent estimates of the date of colonisation of southern Asia and Australia. Marta Lahr and Rob Foley of Cambridge University have suggested that there were at least two dispersals out of Africa: one via the southern route, from the Horn of Africa to the Arabian Peninsula, and along the Indian Ocean coastline, some 70,000 years ago, and then another exodus via the northern, Levantine Corridor up into Europe around 50,000 years ago. They suggested that each of these dispersals carried a different archaeological ‘signature’: the first migration from the Horn of Africa was associated with Middle Palaeolithic stone tools, while the later expansion was characterised by the appearance of more advanced, Upper Palaeolithic tools. They also argued that this model of multiple dispersals explained the range of anatomical variation, especially in skull shape, among both fossil and living modern humans, and that genetic studies also supported this theory.
^7
,
9

Julie Field and Marta Lahr produced an ingenious GIS-based computer model to examine potential routes of expansion out of Africa. The model was based on studies of the palaeoenvironment during the cold, dry period of OIS 4, between about 74,000 and 59,000 years ago. Growing glaciers in northern and southern latitudes would have trapped an enormous amount of water as ice: sea levels would have dropped around the globe, to around 80m below modern levels. While the Persian Gulf would have been drained dry, the Red Sea would still have existed, though its coast would have been further out than today. North Africa and Arabia would have become increasingly arid, and the deserts would have expanded.

Field and Lahr’s programme worked on the principle of finding the ‘route of least resistance’, taking into account obstacles like mountains, and wide lakes and rivers, as well as the availability of crucial water sources along the way. As colonisers would not have been heading in any particular direction or with any specific destination in mind, the programme was designed to ‘wander’, and to explore routes within a 60km radius of an origin. That starting point was set as Omo Kibish, then the computer was set free to wander. The ‘route of least cost’ across the ancient landscape, with sea levels much lower than today, took the virtual colonisers to the coast of the Red Sea, and (in a version ‘without boats’) up the west coast. Then a range of hills, level with modern-day Aswan, forced the virtual colonisers west, into the Nile Valley, and up to the Mediterranean coast. The route continued north, close to Mount Carmel, then headed east, to link up with the Euphrates, and followed that river down through the vast plain that is the Persian Gulf today. However, as Oppenheimer pointed out in a more recent paper, this northern route involved the virtual colonisers making three journeys of over 300km each, through deserts: from the Red Sea to the Nile, from the Nile to the Dead Sea, and across the Syrian desert to the Euphrates, not a mean feat for a non-desert-adapted animal that relies on plentiful supplies of water.
²
Even the modern Bushmen of the Kalahari need water.

In a second version, Field and Lahr allowed their colonisers the luxury of a boat, to cross Bab al Mandab. Then the route split, going north and coming to a stop near the Gulf of Aqaba (though presumably, if virtual boats were permitted here as well, they could have have continued around to the west coast, and there could have been a thriving virtual community all around the Red Sea coastline), or going east, along the shores of modern-day Yemen and Oman.
⁹

The computer model depends on assigning values to all sorts of variables: deciding how readily people would cross rivers, lakes and mountain ranges, for instance. It’s important to recognise that its authors didn’t suggest that it would be able to predict actual, ancient routes of migration. It was really designed as a tool that could provide a different way of looking at potential pathways through ancient landscapes. But it is slightly bizarre in that it tends to show people wandering about in the desert, rather than sticking to areas with better water supplies. However, even with its limitations, the model does suggest that, between 60,000 and 70,000 years ago, both a northern and a southern route out of Africa would have been possible.

But now, genetics comes to the table to force the palaeoanthropologists’ hand. Geneticists studying mtDNA trees have suggested that a single migration is most likely.
^{8
,
11
,
20
,
21
,
22}
All non-Africans derive from a particular lineage called L3, which originated in Africa around 84,000 years ago.
²²
The two ‘daughters’ of L3, haplogroups labelled ‘M’ and ‘N’, arose about 70,000 years ago.
²
The greatest diversity of M lineages is found in South Asia, suggesting that this is where the haplogroup originated; a branch of M, haplogroup M1, is also found in East Africa, but this may represent a late back-migration, after the Last Glacial Maximum (or ‘LGM’).
²⁰
The N lineage is almost exclusively non-African. The simplest, most parsimonious explanation for this pattern is that a branch of L3 emerged out of Africa as a single migration, some time between 85,000 and 65,000 years ago, and that M and N then sprang forth, somewhere around the Indian subcontinent. Later, the first modern human Europeans would have come, not through the Levantine Corridor from North Africa, but from populations that had established themselves in the Indian subcontinent.
²

Proponents of this model argue that the Y chromosome evidence for the northern route has been misread, and the relevant genetic markers represent a much later expansion from North Africa. It looks like there are two or three Y chromosome lineages that contribute to non-African populations (compared with the single mitochondrial L3) but this does not necessarily mean there were multiple migrations: the two or three ancestral lineages could have been carried out in a single exodus. The distribution of Y chromosome haplotypes outside Africa seems to support this proposition, and studies of genes in the other chromosomes also fit with a single exit.
¹¹

A single exit out of Africa means choosing between those northern and southern routes,
²³
but it is hard to pin down the route taken, whether via Sinai or Bab al Mandab, on the basis of the genetics alone.
²¹

For Oppenheimer, though, the choice is an easy one, if you take a look at the palaeoclimatic evidence. He argues that this suggests that, while the northern route may have been shut off during the cold of glacial stages, the southern route would have been open. Sea levels dropped significantly 65,000 years ago, at the Heinrich 6 event in the middle of OIS 4;
¹⁹
this was in fact the coldest, driest episode of the last 200,000 years. There was also a huge drop in sea levels 85,000 years ago, corresponding with the Heinrich 7 event (that Yoel thought may have forced the people of Skhul southwards). Although most of the Arabian Peninsula would have been dry, unwelcoming desert, the coast might just have still received enough monsoon rain to form a route for the beachcombers to follow. So a southern route across Bab al Mandab would have kept the colonisers close to sources of fresh water.
^2
,
8
,
11

There is evidence of people living on the East African coast by 125,000 years ago, with shell middens and MSA stone tools found at a site in Eritrea (although, in the absence of fossils, it is not definite that these beachcombers were modern humans).
²⁴
For Oppenheimer, the onset of dry conditions combined with reduced sea levels would have provided people living around the Horn of Africa (modern-day Djibouti) with both the impetus and the means to leave Africa. Increasing aridity, leading to scarcity of food and hunger, may have prompted the migration, while at its lowest the reduction in sea level meant that there would have been just 11km of sea to cross at Bab al Mandab, at the southern end of the Red Sea.
²⁵
Other climatologists suggest that the most likely time for the migration, from a climatic perspective, would have been slightly later, after the Heinrich 7 event, when the climate was warm and wet again, with the monsoon in full operation. And the last major wet phase in the Arabian Peninsula was between 78,000 and 82,000 years ago. This would still fit between the suggested dates for the emergence of L3 and her daughter haplogroups, M and N.

There is no evidence for boats from this long ago, but it seems reasonable to me to assume that modern humans living on the coast would have had the ingenuity to invent them. Small watercraft would have meant that people could have crossed the mouths of rivers and better exploited coastal resources. If they could see Arabia across the water, and if their families were struggling to survive on the African coast, as Oppenheimer suggests, that voyage seems like a very sensible option. In fact, there might even have been a maritime community all around the shores of the Red Sea, providing a base from which to spread along the southern coast of Arabia. Is it possible that we’re getting too mired in this debate about the northern versus southern migration routes out of Africa? During warmer, wetter phases modern human could have been spreading out, along coasts and rivers, and around edges of deserts. They could have made their way out of Africa via both northern and southern routes, perhaps meeting up again somewhere around the Gulf. During arid phases, populations may have become restricted to green refugia.

It’s very hard to trace movements of modern humans by archaeology alone. Before the Upper Palaeolithic or Later Stone Age, it’s difficult to tell apart tools made by modern humans and by other archaic species like Neanderthals. It would be great to have some fossils of modern humans from this important period in the Middle East, dating to between 80,000 and 50,000 years ago, but at the moment there are none. Nevertheless, some archaeologists think that they can discern the presence of
Homo
sapiens
from the tools they made, or other signals of modern behaviour, and claims have been made for modern humans being in India nearly 80,000 years ago, and in Australia by 60,000 years ago, whereas the earliest dates for the Levant (after Skhul and Qafzeh) and Europe are around 50,000 years ago. If those archaeological claims are to be trusted – and we must treat them with some caution in the absence of any accompanying modern human skeletons – then Oppenheimer’s hunch that modern humans made it out of Africa via an early, southern route of dispersal may turn out to be true.