4. The Wild West:
The Colonisation of Europe

On the Way to Europe: Modern Humans in the Levant and Turkey

Considering that Europe is geographically so close to Africa, it seems remarkable that modern humans made it all the way to Australia some 20,000 years before we find any evidence of their presence in Europe. Why did it take so long? The answer is likely to be complex, involving geographical and environmental barriers, and, perhaps, the presence of other humans already occupying Europe. Because, whereas in most of Asia (with the notable exception of Flores) earlier humans had vanished long before moderns arrived on the scene, Europe was the domain of the Neanderthals.

There is a huge gap between the appearance of the first anatomically modern humans in the Near East – in the Skhul and Qafzeh caves in Israel, some 90,000–120,000 years ago – and the first evidence of modern humans in Europe – at around 45,000 years ago. After Skhul and Qafzeh, modern humans disappear from the Levant for around 50,000 years, although it seems that, during this time, modern humans were making their way eastwards along the coast of the Indian Ocean.

From Arabia and the Indian subcontinent, it may seem that the colonisers should have been able to spread north into Europe with ease. Stephen Oppenheimer¹ suggests that, just as deserts may have blocked the northern route out of Africa for much of the last 100,000 years, the way from the Indian subcontinent and Arabian Peninsula to the Levant was also sealed off by geographical barriers: by the Zagros Mountains, and the Syrian and Arabian deserts. While the beachcombers surged eastwards, their northwards expansion into Europe was blocked. But around 50,000 years ago, the climate warmed up briefly, for a few thousand years. Oppenheimer argues that this warming opened up a green corridor from the Arabian Gulf to Syria, a gate into Europe.

The colonisers could then have spread north-east, skirting the Zagros Mountains, up the coast of what is now Pakistan and Iran, and up the River Euphrates into modern-day Iraq and Syria, making their way from the Persian Gulf to the Mediterranean coast. Some archaeologists argue that Upper Palaeolithic sites along the Zagros Mountains support this route, and indeed suggest that Upper Palaeolithic technology may have originated around the Zagros Mountains, with some dates in excess of 40,000 years.² However, these dates need to be treated with caution: firstly, these are dates at the very extreme of what is considered reliable for radiocarbon dating, and, secondly, the dates were reported in the 1960s, long before the new sampling and calibration techniques were applied. However, the types of tools found at the Zagros sites are similar to the earliest Upper Palaeolithic tools found around the eastern Mediterranean, known as the ‘Levantine Aurignacian’.

Routes into Europe. The black footprints represent the first modern humans to reach Europe, bringing with them the Aurignacian culture, starting around 45,000 years ago; the grey footprints represent the later incursion of the Gravettian people, beginning around 30,000 years ago.

The route from the Persian Gulf to the Mediterranean, following an earlier southern dispersal from Africa, seems a reasonable suggestion, but other researchers still favour a simple northern route out of Africa, from Egypt. But whichever of these routes was taken, we would expect to find part of the archaeological trail in the Levant and in Turkey; in other words, in those countries that border the eastern Mediterranean.

There is a growing body of archaeological evidence for the earliest modern humans in the Levant and Turkey, in the form of Upper Palaeolithic stone tools, and ornaments – and some bones.

In the 1940s, archaeologists began to excavate down through 19m of deposits at the site of Ksar ‘Akil, near Beirut in Lebanon. They found twenty-five layers containing Upper Palaeolithic archaeology. In the deepest layers, they found Levallois-type technology (stone tools made from prepared cores), typical of the Middle Palaeolithic, alongside classic, reshaped Upper Palaeolithic tools, like end-scrapers and burins. In later layers, the Levallois cores are replaced by cone-shaped prismatic cores, from blade manufacture – a hallmark of the Upper Palaeolithic. Dating of layers above and below the earliest Upper Palaeolithic stratum at Ksar ‘Akil suggests that those tools were being made there somewhere between 43,000 and 50,000 years ago.³ And the discovery of a skeleton at Ksar ‘Akil confirmed that it was modern humans who were making those tools.⁴ Dating of the Upper Palaeolithic at Kebara also suggests the presence of modern humans in the area by 43,000 years ago.⁵

Tracing the northwards expansion of people into Turkey has been problematic. The Palaeolithic of Turkey has long been in the shadows, mostly because comparatively little archaeological research has been carried out there.⁶ Many Palaeolithic sites in Turkey are just ‘findspots’, where stone tools have been spotted on the surface. Relatively few have been excavated, but in the last twenty years archaeologist have been striving to plug this gap, and some interesting finds have emerged, which help us to trace the journey of those early European colonisers.

The archaeological site of Üçagizli lies on the rocky south-west coast of Turkey, about 150km north of Ksar ‘Akil, and near the city of Antakh (ancient Antioch). The partially collapsed cave was first discovered in the 1980s, and excavations began there in earnest in the 1990s, leading to the discovery of Upper Palaeolithic artefacts in red clay sediments. The tool manufacture at Üçagizli seems to follow a very similar pattern to Ksar ‘Akil: the oldest Upper Palaeolithic layers in fact contain a mixture of ‘Middle Palaeolithic’ technology alongside more classic, retouched tools. Just as at Ksar ‘Akil, the Levallois cores disappear in the higher layers, to be replaced by prismatic cores, where blades have been knocked off by soft-hammer or indirect percussion. But at Üçagizli there are also tools made of bone and antler. The earliest Upper Palaeolithic layers date to between 41,000 and 44,000 years.³

Throughout the Upper Palaeolithic layers at both Ksar ‘Akil and Üçagizli, the archaeologists discovered classic signs of the Upper Palaeolithic – and of modern humans: ornaments. The vast majority are small seashells, pierced through to be used as beads or pendants. More than five hundred shell beads have been found at Üçagizli alone. It’s possible to see how the holes have been created in the shells – some by scratching away, while others seem to have been punched through with a pointed tool – and it’s quite clear that these holes were made by humans and not by other, natural processes. There were shells from marine snails Nassarius gibbosula, Columbella rustica and Theodoxus jordani, as well as the pretty, ridged bivalve Glycymeris. From the large collection of shells at Üçagizli, it looks as if those hunter-gatherers also enjoyed the taste of seafood: bigger shells, from limpets (Patella) and the edible snail Monodonta also appear, unpierced, in the archaeological strata. The archaeologists were sure that these shells had been collected for food as they were not wave-worn like the empty seashells that wash up on beaches, and, in addition, many of them were burnt.

The shell beads from Üçagizli are not – by a long stretch – the earliest ornaments: the pierced shells from Skhul in Israel date to between 100,000 and 135,000 years ago.⁷ Shell beads were also found in Blombos Cave, dating to around 75,000 years ago, and the evidence for ochre use goes back to beyond 160,000 years ago at Pinnacle Point. These finds suggest that art and ornamentation is probably almost as old as the human species itself. But the Üçagizli beads are useful as a marker for modern human presence. They show that modern humans were bringing with them a shared culture, and perhaps an awareness of identity and a system of communication that did not seem to have been there among archaic human populations.³

There is a scarcity of Upper Palaeolithic sites in Turkey. Kuhn⁶ argues that the Anatolian Plateau, most of which lies more than a kilometre above sea level, would have been a cold, unwelcoming place during the late Pleistocene, and that modern humans (as well as other animals) would have gravitated towards the warmer coast. With a higher sea level today, any Pleistocene coastal sites would now be submerged. Nevertheless, Üçagizli is an extremely important site, a stepping stone into Europe, with Upper Palaeolithic artefacts going back more than 40,000 years, anticipating the spread of the classic Upper Palaeolithic culture, the ‘Aurignacian’, from east to west across Europe, between about 40,000 and 35,000 years ago.⁸

This all seems to fit together very neatly, but we need to be aware that these initial movements into Europe are occurring at an awkward time for radiocarbon dating, and some dates obtained during the twentieth century might need reassessment. And not only do archaeologists and anthropologists argue about the exits from Africa, there is also debate about the route into Europe, and where Upper Palaeolithic culture began. Kuhn⁶ seems convinced by the dating of Üçagizli, but also suggests that this culture may represent a spread southwards from Europe, rather than northwards from the Zagros Mountains and the Levant. Other researchers have argued that Upper Palaeolithic culture may have arisen in the Russian Altai, north of the Zagros Mountains, with a dispersal of modern humans, carrying this technology, coming into Europe from around the Caucasus Mountains and the northern coast of the Black Sea.⁹

However, most researchers seem to agree that Üçagizli and Ksar ‘Akil fit very well with a model where modern humans, bearing an Upper Palaeolithic, pre-Aurignacian toolkit, arrive in the Levant, spread north into Turkey, and then westwards through Europe.⁸

Crossing the Water into Europe: the Bosphorus, Turkey

Making my own way up through the Asian part of Turkey, I reached a watery barrier: the Bosphorus. This narrow strait connects the Black Sea in the north to the Sea of Marmara, and at its southern end the Sea of Marmara narrows down again to form the Dardanelles, connecting through to the Aegean Sea.

In Istanbul, I took the ferry to cross the glittering Bosphorus, from the Asian to the European side. I thought about the early colonisers reaching this waterway, and Bulbeck’s ideas about coastal and estuarine adaptations, and the possible use of watercraft in the Palaeolithic. It didn’t seem to me that the modern Bosphorus and the Dardanelles would have constituted much of a barrier to the early pioneers.

In fact, when I looked into it, it turned out that the Bosphorus was dry during the Pleistocene. It wasn’t until after the Ice Age that the sea level rose sufficiently to flood the Bosphorus and connect them. Boreholes drilled down through the sediments at the bottom of the strait show how and when the connection between the two bodies of water was established: the Sea of Marmara spread northwards and eventually opened into an estuary at the southern end of the Black Sea, around 5300 years ago, and the Bosphorus was formed. Interestingly, it looks like there was another, intermittently open, connection between the Sea of Marmara and the Black Sea during the Pleistocene, to the east, from what is now the Gulf of Izmit.¹ So even if the Bosphorus wasn’t there to cross, there may have been times when the colonisers would have got their feet wet crossing from Asia to Europe.

From the Bosphorus region colonisers could have headed northwards to the coast of the Black Sea, or continue beachcombing westwards along the Mediterranean coast, and it appears that they did both. There are a scattering of sites with Upper Palaeolithic stone tools, on or close to the Mediterranean coast, in Italy, France and northern Spain, and there are also sites dotted along both the European and Asian shorelines of the Black Sea. An important early Upper Palaeolithic site is Bacho Kiro, in Bulgaria, where a ‘pre-Aurignacian’ toolkit – with lots of blades – has been found, dating to 43,000 years ago. Moving northwards up the west coast of the Black Sea, the colonisers would have reached the great delta of the Danube, in modern-day Romania. They could have then used the waterway as a superhighway into the heart of Europe: there are many Upper Palaeolithic sites along the Danube and its tributaries. The conventional radiocarbon dates for this westwards spread across Europe suggested it took place between 45,000 and 35,000 years ago. New, calibrated radiocarbon dates for these sites suggest that this dispersal into central and western Europe happened fairly rapidly, between 46,000 and 41,000 years ago. This speedy spread across Europe may have been helped by another episode of global warming, the Hengelo interstadial.^{2, 3, 4, 5}

Face to Face with the First Modern European:
Oase Cave, Romania

My next destination was a site along that Danube corridor. I travelled to Romania, to meet geologist and speleologist (cave expert) Silviu Constantin, who was going to introduce me to the cave where the earliest known modern human fossil in Europe had been discovered.

From Bucharest, we drove east, following the Danube fairly closely, like our predecessors 40,000 years before. We drove and drove, and eventually reached a small village in the south-western Carpathians, the name of which I cannot disclose because the exact locality of the cave was secret. A couple of stray dogs chased our car up the hill, excitedly barking, but when we stopped and got out they ran away. We were staying in the 7 Brazi (fir trees) Pensiune, high up on the hill overlooking the village.

The following morning, we met up with the team of cavers who were accompanying us – Mihai Bacin (leading the team), Virgil Dragusin and Alexandra Hillebrand – and sorted out caving gear for the trip before heading off to the cave itself. We drove through wooded valleys and past abandoned factories, and eventually turned off on to a dirt track. About a quarter of a mile later, we reached a point where the road had collapsed. We all got out to take a look, and, after hauling some large stones into the hole, decided that it was passable, and cautiously drove the cars over the roughly stopped-up breach. It held. We pulled up just around the corner from the collapsed track. The cave itself was below us in the steep-sided valley, and we scrambled down the wooded slope, lugging our equipment down to the stream bank at the bottom. Once we were down, I could see the tall, slit-like entrance of the cave, with the stream emerging from it.

This was Peştera cu Oase, the Cave of Bones. It was incredibly exciting to be standing there, in front of a place that I had read so much about. A colleague from Bristol University, João Zilhao, had been part of the team that excavated there in 2003–5, so I had heard a lot about the cave from him. Unfortunately, he was in Portugal investigating another cave, but that left me in Silviu’s capable hands – he had also been one of the excavating team, and he had dated the finds from the cave.

On 16 February 2002, a group of intrepid cave divers were exploring the cave. The cavers had made their way into the depths, past a duck-under and through a longer, underwater section, and up a steep ramp to an area littered with animal bones. ‘They found a human mandible, right on top of the flowstone. It was probably dug out recently by an animal. It was just sitting there, waiting for someone to discover it,’ said Silviu.

When the mandible was radiocarbon dated, it was found to be 35,000 years old, the oldest known remains of a modern human in Europe.

‘You must have been pretty excited to get that date back,’ I said.

‘I remember everyone was excited,’ said Silviu. ‘The oldest human was here, in Romania. We were proud that it was in one of our caves.’

The cave was an exciting proposition for scientists studying this period of prehistory. In 2003, an international team of archaeologists visited the cave, and discovered masses of bones – mostly cave bear, but also some other human material: fragments of a skull. These pieces were found further down the slope than the mandible, in part of the cave that was subsequently, and evocatively, named Panta Strãmoşilor (the Ramp of the Ancestors). The skull and mandible were from two different individuals.

The place where the bones were found is hard to get to now – and involves that dive.

‘The archaeologists were trying to figure out how such a massive bone deposit could come into the cave,’ said Silviu. As a geologist and a caver, he had joined the team to answer this question. He could also help with dating – using uranium series dating on stalagmites. From his investigation, it seems that Peştera cu Oase once had another entrance, allowing the cave bears access. In fact, the two galleries that lead off from the Ramp of the Ancestors both seem to have had openings in the past. Today, these entrances have collapsed, and are practically blocked off, although some small animals – such as rodents – still fall down into the cave through the sinkholes.¹

The majority of the bones in Oase belonged to cave bears. There were also bones of other cave-dwellers such as wolf and cave lion, which had presumably made their dens in there at various times. However, also found in the cave were skeletal remains of distinctly non-cave-dwellers such as ibex and red deer. These could have been brought into the cave by people, but there were no signs of such habitation in the cave, nor any cut marks on the animal bones to suggest that they had been eaten by humans. That left geological processes and carnivores to explain the accumulation of bones in the cave. In 2005, the archaeologists returned to excavate the cave; they found many more bones, but also important clues as to how the skeletal remains had ended up there. Underneath a covering of stalagmite, they found a 30cm-thick layer of cave bear and other animal bones, many bearing the tooth marks of bears and wolves. This looked like a collection of bones of animals eaten by carnivores denning in the cave, as well as the cave-dwellers themselves. Under that was a layer of bones mixed with sand, gravel and cobbles, and these were sorted by size – larger bones at the top of the slope, smaller ones at the bottom, and many of the bones in this layer had rounded-off edges. These bones had been swept into this part of the cave by flooding. So it seemed that the animal bones in Oase had ended up there through a combination of cave-dwelling carnivores bringing their dinner home, and flooding.

‘So how do you think the human bones ended up in the cave?’ I asked Silviu.

‘Most probably they had been washed in.’

There were no gnaw marks on these bones, so perhaps a person, or even a buried skeleton, had fallen in through a sinkhole and then the bones had been washed further down the cave by floods.¹ Nevertheless, it seems rather strange that there are only cranial remains – a skull and a mandible, and no other parts of the skeleton – but there are still a lot of bones in Oase. Perhaps the rest of the skeleton will yet be found.

I followed Silviu into the cave – wading along the stream and into the first great gallery, with Mihai, Virgil and Alexandra following behind. The roof was very high, and hung with enormous stalactites. Towards the back of this gallery were a series of pools with stalagmite edges. We followed the stream around to the left, where the roof plunged down. We were lucky that day: there hadn’t been much rain the previous weeks and days, so the water level was low enough to enable us just to keep our heads above water. In a wetter period this would have been a duck-under. Still, it was an awkward manoeuvre: under the water, the cave floor sloped away from the gap I was aiming for, and I couldn’t get a foothold. As my feet slipped, I gave up trying to walk through the gap, and swam through it instead. On the other side, I was wading shoulder-deep, hanging on to a rope to keep me close to the shallower right-hand wall. The floor of the tunnel gradually rose up until the water was just knee-high.

We all made our way along this narrow, tunnel-like part of the cave. Sometimes the floor dipped down and we would plunge in up to our chests again. Eventually, we reached a wider part of the tunnel, but the roof sloped down and down until it touched the water: this was the siphon, and the end of the line for me. I would have had to dive to get to the Ramp of the Ancestors. It was frustrating in a way, but I had known that I would only be able to get this far. I still felt privileged to have come so close, and to have explored the cave where Europe’s earliest modern human had been found. Silviu and I sat down on a convenient stalagmite and I asked him about excavating in the cave. It sounded arduous and difficult – getting to the site was one thing, taking in tools and bringing back bags of sediment and bones, through the siphon and through the duck-under another altogether – but the results had been worth it. When the skull had been dated, it was even older than the mandible: around 40,000 years old. And there were some things about the skull and mandible that seemed a little odd.

Leaving Oase, Silviu and I drove back to Bucharest, where I would be able to see the skull and mandible. We had travelled close to the Danube on the way out to Oase, but on the way back we took a route through the beautiful wooded gorges of the Carpathians. Where the gorges opened out into valleys, fields were visible in which people were cutting hay and heaping it on to three-legged wooden frames to make stacks. The haystacks varied from field to field and village to village. Some were tall and thin, others squat and conical. We slowed down to pass a heavily laden hay cart pulled by two horses.

The centres of the villages were often lined with low, terraced cottages, but on the outskirts there were more often than not massive, ugly tower blocks. These buildings seemed so incongruous among the fields. ‘Why build blocks of flats in villages?’ I asked Silviu. ‘They were built by Ceauşescu,’ he said. ‘He wanted to create more land for farming.’ ‘But it seems like there’s plenty of land,’ I suggested. ‘Yes. There is. Ceauşescu wanted to destroy the villages.’ It was strange to think that it really wasn’t that long ago – 1989 – that Ceauşescu had been removed from power. Romania was a country in recovery.

Silviu told me about how he had travelled around the countryside as a young man. He recounted how, if he’d got stuck for somewhere to stay overnight, he would find a barn and sleep in the hay. He usually looked for someone to ask beforehand, and had often been invited in to dinner. ‘People treated the occasional backpacker as a traveller: to be given hospitality,’ he said. ‘Now they want money. They think tourists have money and they want some of it.’ Silviu was worried about the people and the countryside being spoilt by tourism. He liked the wilderness.

‘What would this landscape have looked like 40,000 years ago?’ I asked Silviu.

‘I don’t know,’ he said, and paused for thought. ‘It would have been [Oxygen Isotope] Stage 3. So – colder than now. And wetter summers. Like – perhaps Norway today, at the coast; maybe like Bergen.’

I tried to imagine the hunter-gatherers in the foothills of the Carpathians, dressed warmly against the cold, in a country with red deer, ibex, wolves and cave bears. Although cave bears are long gone, there are plenty of brown bears roaming Romania today: nearly half of all European bears are in the Carpathians. Some hunting of bears goes on today; indeed, one restaurant in Bucharest even had ‘bear paw’ on the menu. There were times when I was truly glad to be vegetarian.

The following day, I visited Silviu in his natural habitat at the Emil Racovita Institut de Speologie in Bucharest. He brought a series of cardboard boxes into a lab next to his office and we carefully unpacked some of the Oase bones. There were pieces of animal bone embedded in speleothem – a real gift for a geologist like Silviu who could date the stalagmite, and therefore the bone. He also brought out a huge and formidable looking cave bear skull. And then there were the human remains: the mandible and the skull.

Again, these were quite clearly modern human. The skull was globular in shape, without any of the more obvious features of an archaic skull, like big browridges, or a protruding occiput at the back of the head. The jaw was quite gracile and modern-looking too, with a definite chin. But there were some oddities – particularly in the mandible or jawbone. The chin was very straight, the ramus (the part of the mandible ascending up to the jaw joint) was very wide, and the mandibular foramen (the hole on the inside of the ramus where the nerve supplying the lower teeth enters) was a little strange. Then there were the teeth, set in a very wide arc in the mandible, and with absolutely enormous wisdom teeth. These molars are usually smaller than the ones in front, but in the Oase jaw they were huge.^1,2 I had read about these odd features in the scientific articles announcing the discovery and dating of the Oase bones, but it was something entirely different to hold the actual bones in my hands and look at them myself.

Now, everyone is unique and we all have ‘anatomical traits’ – little variations on a theme – in our bodies and our bones. In our skulls, some of us may have one hole for a particular nerve, where others have two or three. We might have little islands of bone, or ‘ossicles’, embedded in the zigzag joins or sutures where the plates of the skull come together. Some of these traits are genetic, whereas others appear during our lifetimes, and might be related to diet – or other things we do to our bodies. For instance, lumpy bits of bone inside the ‘external auditory meatus’ – the bony part of the ear canal – can be caused by swimming in very cold water. But, like skull shape and size generally, we’re not sure about how or why some of these traits occur, and the various influences of genes and environment. It’s all part of that great question in developmental biology: how are our bodies shaped by our genes and our environment?

Having said all that, however, there are traits that do at least seem to hark back to an ‘earlier’ form. If you were being unkind, you might call them ‘throwbacks’. If you were a bit nicer about it, you might think of these characteristics as echoes of evolutionary history, glimpses of where we’ve come from. And those odd traits in the Oase jaw seem to fall into this category: archaic features. But could this be more than just an echo of a more distant evolutionary past in these 40,000-year-old bones? Is it possible that, instead, the Oase bones show a mixture of modern and archaic traits because that person was a mixture of a modern and an archaic human – some sort of hybrid? This isn’t such a preposterous idea, because, when modern humans started making their way into Europe, someone else was already there: the Neanderthals.

As we have seen from the archaeology and fossil record of East Asia, ours was not the first human foray out of Africa. A series of archaic human species made the leap before us.

In a review article that appeared in Science in 2003, Ann Gibbons wrote: ‘The long-legged, relatively big-brained hominin called Homo erectus has long been considered the Moses of the human family – the species that led the first exodus out of Africa more than 1.5 million years ago.’³

The biblical analogy is great. I can imagine that striding, big-browed man leading his people out of Africa, across the Red Sea, but it’s a sleight of pen in two ways. Firstly, there’s a wry and unwritten ‘but of course it wasn’t like that’, as we all have this tendency to promote our ancestors to heroes and imagine their lives as epic struggles against adversity, winning through so that we could be alive today. And secondly, Ann Gibbons goes on to write about Homo georgicus, one of the recent and somewhat cheeky surprises in European palaeoanthropology. The three paradigm-nudging and diminutive fossil skulls were recently discovered in Dmanisi in Georgia, and dated to 1.75 million years ago. Then another small skull was found in Kenya, dating to about 1.5 million years ago – perhaps belonging to a particularly small population of Homo erectus, perhaps linked with the Georgian hominins. The Kenyan skulls were the same age as another famous fossil: Turkana Boy. This young man, with a largish brain, is sometimes classified as Homo erectus, sometimes as Homo ergaster. (Remember that the world of palaeoanthropology is interpreted differently by ‘lumpers’ and ‘splitters’ – see page 3.)

Fossils can be extraordinarily slippery when you’re trying to pin a species name on them. The Kenyan and Dmanisi skulls are no exception. They all look a bit like a small Homo erectus without browridges, but also bear similarities to another, earlier hominin species, Homo habilis. Although the discoverers of the Dmanisi skulls claim that they warrant a new species name, Homo georgicus, most researchers place the skulls in Homo erectus. So maybe erectus was the first hominin to get out of Africa after all.³

Georgia lies west of the Caspian Sea, but it is hard to know whether it really counts as part of Europe or Asia. Certainly, Dmanisi was sidelined when another, intriguingly ancient fossil, this time from the Sierra de Atapuerca, in northern Spain, was reported as ‘The first hominin of Europe’. It dated to 1.2 million years ago, and its discoverers suggested that it was Homo antecessor (a category that some lump into Homo heidelbergensis).⁴

By about 300,000 years ago, Homo heidelbergensis in Europe had morphed into Neanderthals. And when modern humans got to Europe, these other hominins, their distant cousins, were still around in the landscape.^5,6

The first fossil of these ancient Europeans was found in 1848, in Gibraltar, but nobody paid it much attention. The bones that gave the species its name were found in 1856 in Germany, near Düsseldorf – in the Neander Valley, or Neanderthal. The valley was being quarried for limestone, and workmen clearing out mud from caves in the cliffs prior to quarrying found what they thought were cave bear bones. But a local teacher recognised that they were human and collected them up.⁷

The following year, rather bravely for the time, Professor Schaafhausen of Bonn University published a report on the skull and bones from the Neander Valley, saying that they were normal – non-pathological – but seemed to be from an ancient inhabitant of Europe as the remains were found alongside bones of extinct animals. This interpretation was challenged by Professor Mayer, also at Bonn, who said the bones were probably much more recent, probably those of a Russian Cossack dying from rickets who had crawled into the cave, with great browridges from frowning in agony. But a few years later the find had been widely published, and there was a growing consensus that the bones were very ancient. The Irish anatomist William King proposed that the skeleton should be given a new species name: Homo neanderthalensis. It was the first known species of a fossil human.^7,8,9

Since the discovery of the first Neanderthal fossil more than 150 years ago, several thousand bones have been found from more than seventy different sites. And there are over three hundred sites where Neanderthal stone tools have been found.

Neanderthal characteristics start to appear in European Homo heidelbergensis. For instance, the Sima de los Huesos specimens from Atapuerca, which are over 350,000 years old, already have some ‘Neanderthal’ features such as protruding faces and gaps behind their wisdom teeth, as well as a characteristic shape of the browridge, and a ridge across the back of the head: the ‘occipital torus’.¹⁰ By about 130,000 years ago, ‘classic’ Neanderthals, with full-blown features, lived right across Europe – and beyond.¹¹ Their territory extended from Portugal in the west to Siberia¹² in the east, from Wales in the north to Israel in the south. And they persisted in some parts of Europe and western Asia until less than 30,000 years ago – after modern humans had arrived in Europe.

Given that we now know so much about Neanderthals, there are still many questions about their disappearance. Although there is no evidence of Neanderthals and modern humans actually living in precisely the same places at the same time, there was certainly a period when both species were present in Europe. It used to be thought that this period of overlap lasted around 10,000 years, but new calibrated radiocarbon dates suggest that the overlap was shorter: about 6000 years in north and central Europe, and perhaps only one or two thousand years in western France.¹³ But why did the Neanderthals disappear? Did we kill them off or out-compete them? Or perhaps they are actually still around – could Neanderthals have been assimilated into the expanding modern human population as it flowed westwards across Europe?

There are certainly some researchers who think so. They put forward specimens – mostly skulls – like Oase and Cioclovina from Romania, the Mladeč fossils from the Czech Republic and the Lagar Velho skeleton from Portugal1 – as physical evidence for interbreeding between Neanderthals and modern humans. Palaeanthropologists João Zilhao and Erik Trinkaus suggest that archaic traits in these fossils are not just ‘throwbacks’: they may be evidence of Neanderthal genes in the early modern human populations of Europe.

Neanderthal Skulls and Genes: Leipzig, Germany

So I made my way to Germany, not to the Neander Valley, but to the Max Planck Institute for Evolutionary Anthropology in Leipzig, where I had arranged to meet Dr Katerina Harvati, who had recently analysed the Cioclovina skull. Katerina met me on the other side of the revolving door at Max Planck, and we walked into an enormous space, some three floors of atrium with light streaming in through glass walls on two sides. Stairs and ramps to the upper floors seemed to float in the air. Katerina led me up to the labs on the second floor, where she was going to show me CT scans of the Cioclovina skull.

But my attention was first drawn to a composite skeleton, put together from casts of fossil bones from different sites, standing in the corner of the lab. It was the first time I had laid eyes on a complete, assembled Neanderthal skeleton, and it was interesting to see just how stocky he looked. The ribcage flared out at the bottom, quite different from the modern human chest shape. Individual bones were generally quite similar to modern human bones, but nonetheless very rugged.

‘We can tell from their body form and proportions that Neanderthals were showing some level of cold adaptation: they were stocky, with short limbs,’ said Katerina.

But how much of an advantage would this have given them, compared with modern humans?

‘It has been calculated that the advantage would be – perhaps not as great as we originally thought – maybe the equivalent of one business suit.’

It didn’t sound that impressive. Cultural adaptations, like clothing and use of fire, must have been more important to the Neanderthals’ survival in Ice Age Europe.

But the most ‘different’ part of all the Neanderthal skeleton was the skull. Neanderthals have very long, low skulls, whereas modern human human crania are much rounder. Neanderthal faces are big: they have massive browridges, large, goggly orbits (eye sockets), large nasal openings and projecting, prognathic jaws.

So what about this Cioclovina skull that had been suggested to be a Neanderthal/modern human hybrid? The skull itself had been discovered in the cave that gives it its name in southern Romania, in 1941 – during phosphate mining – and had recently been radiocarbon dated to about 29,000 years old. The skull was really just a braincase: most of the face was missing. Although its general shape was definitely modern, some researchers had suggested that the shape of the browridge and the back of the skull were Neanderthal-like.¹

Obviously, before you can confirm or reject a claim that a skull represents a hybrid, you have to have an idea of what a hybrid might look like. Is it likely to have an even mix of features from each parent? Or might it be mostly like one parent with just a few features from the other? Katerina had looked into the features of hybrids in other primate groups and she found that a common feature of hybrids seemed to be a size change – either bigger or smaller – than would be expected from the parent populations. Some hybrids – like a gibbon–siamang cross in Atlanta Zoo, and hybrids from different macaque and baboon species – looked anatomically like a mixture of the two species they came from. It also seemed that hybrid populations tended to be more variable than the parent species, and also had rare anomalies popping up more often than usual.¹

So Katerina had analysed the Cioclovina skull to see if it showed any of these signs of being a hybrid: an appreciable size difference, a mixture of features, a high level of variability, or any strange anomalies. But she had also measured the skull so that she could compare its size and shape with those of other modern human and Neanderthal skulls. Describing features in skulls, even measuring them, is fraught with problems, as I had seen so vividly in China, but Katerina had also used a technically sophisticated and perhaps more objective approach to the problem of comparing skull shape and size.

The first step was to convert a real skull into a mathematical model, a cloud of points in 3D space that described the shape and size of the skull, using features or ‘landmarks’ that could be recognised on any skull. Rather than measuring a skull with calipers to get distances and angles, Katerina showed me how she had captured the 3D shape and size of skulls in two ways: using an electronic digitiser and CT scans. The digitiser was an elegant piece of equipment – an articulated arm ending in a stylus that could be placed on the surface of a skull – and points could be captured in 3D space, with x, y, z coordinates. It was a piece of apparatus that was widely used in design and engineering – and was now beginning to be applied to the study of old bones. 3D coordinates could also be taken from detailed CT scans of skulls, which would allow points on the inside as well as the outside of the skull to be recorded.

Having captured and quantified all that information, Katerina could then compare different skulls, and she did this in the context of variation among different primates.

‘The difference between Neanderthals and modern humans is not similar at all to the differences that you would find between subspecies of primates living today,’ she said.

‘It is much more similar to the distances you’d find between closely related species.’

‘So you can be absolutely sure that Neanderthals are a separate species?’ I asked.

‘Yes, that is what I’d say. They are too different to be another population or even a subspecies of modern human. They were our sister species. Closely related – but a different species.’

This seemed to refute the multiregionalist idea that all species since Homo erectus have essentially been one.

‘So what about Cioclovina?’ I asked. Katerina showed me a 3D computer model of the Cioclovina skull, based on CT scans that had been done at a local hospital. She spun the model round on the screen, and pointed out the relevant features. The browridge was big, but it was broken in the middle, unlike the uninterrupted ‘monobrow’ of Neanderthals. The occipital bone at the back of the skull did bulge out, and the nuchal line where neck muscles attached was well marked, but not really Neanderthal-looking. There didn’t seem to be anything unusual about its size, nor were there any odd anomalies in the skull.

So what about the results of the shape analysis? Katerina had compared the 3D ‘landmark configuration’ of the Cioclovina skull to Neanderthal and modern human (including Upper Palaeolithic) skulls. Using different sets of statistical analyses to make the comparisons, Cioclovina always came out closer to modern humans.¹

‘From my analysis, I wasn’t able to see any resemblance to Neanderthals,’ Katerina told me. ‘There is no evidence to support the claim that this is hybrid. It actually turns out to be very typically modern human in its anatomy.’

It was clear that Katerina couldn’t wait to look at the other proposed hybrid specimens, like Oase. She was open-minded about what her results meant, and what she still might find.

‘Of course this doesn’t mean that hybridisation didn’t happen. It could have happened and we just haven’t found the hybrids yet. Or, some of the other proposed hybrids that I haven’t examined yet might fit the criteria. Or it could be that it was so rare that it hasn’t left a trace in the fossil record. And the genetic evidence to date suggests that if admixture happened, it was so low that it was really not significant in an evolutionary sense.’

Indeed, it wasn’t just the shape and size of Neanderthal bones that was being studied in Leipzig, it was genes as well. In 1997, a team of scientists led by Svante Pääbo of the Max Planck Institute published the first analysis of DNA from an extinct human. They had managed to extract mtDNA from one of the original fossils from the Neander Valley. Pääbo chose to look for a non-coding, fast-mutating section of mitochondrial DNA that had already proved useful in studies into evolutionary relationships between living species.

Getting DNA out of an ancient bone was always going to be a huge challenge – DNA starts to fall apart after death – but Pääbo and his team had hoped that some tiny fragments might still be there. The extraction was done in a sterile room, to try to reduce the possibility of contamination with modern DNA. The bone sample was ground into powder and then the sample was treated to amplify up any DNA – by getting any fragments to make copies of themselves. Then the sequencing could start, and the results were quite stunning: when they compared the Neanderthal sequence with the equivalent mitochondrial DNA sequence from nearly 1000 modern humans, they found that it was distinctly different. The modern human mtDNA sequences differed from each other by an average of eight different base pairs out of almost four hundred. But the Neanderthal sequence had an average of twenty-six base pair differences compared with the modern human samples. This difference suggested that Neanderthal and modern human mtDNA had been evolving along separate pathways for about 600,000 years. Although this seems a very long time ago, compared with the dates of the earliest known Neanderthal (about 300,000 years ago) and the earliest known modern human (about 200,000 years ago) it still makes sense, as the lineages would have started to diverge within an ancestral population of Homo heidelbergensis.²

This result seems to support the theory of a recent African origin of modern humans, and a replacement of any earlier human populations. In contrast, the multiregional hypothesis suggests that archaic populations in Africa, Europe and Asia developed into modern human populations. A halfway house theory has modern humans originating in Africa, then spreading into Europe and Asia and interbreeding with existing archaic humans.

Pääbo’s findings suggested that the mitochondrial DNA lineages, at least, had separated (and stayed separate) hundreds of thousands of years before modern humans appeared in Europe. Even if you ignore the timings, then the multiregional model with hybridisation suggests that Neanderthals should be genetically closest to modern Europeans, but there was no evidence of this in the mitochondrial DNA: the Neanderthal sequence was equally different from all modern humans across the globe. Another study compared ancient DNA extracted from two 25,000-year-old European modern human fossils, and found that the Cro-Magnon mtDNA fell in the modern human range of variation, and was very different from the Neanderthal sequences.³

Looking at mtDNA variation as well as modelling the population expansion of modern humans in Europe, researchers in Switzerland came up with a maximum interbreeding rate between the two populations of less than 0.1 per cent. Statistically, this is so low as to be practically non-existent, and the Swiss scientists go as far as to say that this suggests the two species were biologically separate – and could not produce fertile offspring even if they had seized upon the chance to have sex with each other.⁴

So do these mtDNA results represent definitive evidence that the Neanderthals could not be counted as among the ancestors of modern Europeans? Well, they certainly seem to point in that direction, but, actually, it’s impossible to completely rule out any hybridisation between modern and archaic populations. Neanderthal genes could have entered the human gene pool, but those lineages might have died out, leaving no trace of them today. And what if only Neanderthal men, not women, had interbred with the incoming modern humans? That wouldn’t show up in the mitochondrial DNA – which is inherited only from the mother. So although these Neanderthal mtDNA studies are amazing, and suggest that hybridisation didn’t happen, they can’t rule it out. So would it be possible to probe further, to go after more Neanderthal DNA – perhaps nuclear DNA?

When Svante Pääbo was interviewed for Science magazine after the publication of the Neanderthal mtDNA paper in 1997, he was very pessimistic about the chances of anyone ever managing to recover and sequence nuclear DNA from Neanderthal bones.⁵ But just over a decade later, I was visiting his lab at the Max Planck Institute – and they were doing just that.

The genetics labs were just along the (very beautiful, sky-lit, gently curving) corridor from the bone lab. The Institute felt like a modern monastery, with an all-pervading calm and scholarly atmosphere. But instead of monks painstakingly copying out biblical passages, scientists were locked away in their high-tech scriptoria, sequencing the Neanderthal genome.

I met up with Ed Green, one of the geneticists hard at work on the Neanderthal Genome Project. Ed had brought along some casts of the original fossils from which DNA had been extracted.

‘How do you go about trying to extract DNA from these fossils?’ I asked Ed.

‘Well, the first thing is to find the fossil that has ancient DNA that can be extracted. Then the way it’s done is to simply to take a dentist’s drill, drill a bit, get some bone powder, and then use a standard extraction method where you bind DNA to silica beads.

‘Then the really fun part begins – trying to sequence this DNA, and see what is there. Is this DNA from the individual that owned this bone originally? Or DNA from bugs that have crawled into the bone since then?’

‘And presumably there’s quite a lot of modern human DNA knocking around as well – from the archaeologists who excavated them,’ I suggested. Ed agreed. He was very keen to encourage archaeologists to excavate fossils in a ‘sterile’ way today, but there were many bones that had been discovered decades ago, and handled by scores of archaeologists and curators.

The team had looked at more than seventy Neanderthal fossil bones, and tested them first to see if they were likely to contain any usable DNA by checking the condition of other organic molecules: amino acids. Six of the specimens had good levels of these protein building blocks, so there was a good(ish) chance that some DNA might be in them as well. They went on to extract DNA, but, always aware that this genetic material could come from modern people, they checked for contamination before going any further.

A sample from a fragment of Neanderthal bone from Vindija Cave in Croatia looked particularly promising. ‘Luckily for us, this shard of bone was not interesting enough morphologically to have been handled and looked at a lot – so this guy is nearly free from contamination by modern humans,’ said Ed.

So the geneticists chose to try out DNA sequencing procedure on the extract from the Vindija fossil. This technology is advancing at an astonishing rate. Inside insignificant looking white boxes in genetics labs there are small trays holding hundreds of wells of DNA fragments. And the genetic material they are dealing with is very fragmentary: over time, long stretches of DNA that start off with millions of base pairs become broken and broken again into short sections of just a few hundred or tens of base pairs each. So the process involved sequencing those fragments and then virtually sticking them back together. New technology meant that many different fragments could be sequenced at the same time. ‘The throughput for DNA sequencing is hundreds of times more than it was just three or four years ago,’ Ed told me.

He explained the sequencing method in a very visual way (considering you can’t actually open up the box and watch it in action). In each well, there were many copies of one strand of DNA, and the machine worked out the sequence by ‘asking’ each strand what nucleotide base (A, C, T or G) was next. It did this by flowing a solution over the wells containing each base in turn. If the ‘next’ base was T, the solutions of A, C and G would flow over uneventfully. When the solution containing T was introduced, enzymes would grab the base and at the same time emit a flash of light. This is called ‘pyrosequencing’. ‘Every flow, you’ve got different wells lighting up, like a firework display,’ said Ed. Every time a nucleotide solution passed through, some of the wells would answer ‘yes’ by emitting a flash. The machine cycled on and on, until all the strands in all the wells had been sequenced. This technique can read segments of 100–200 nucleotides in length: perfect when you’re looking at tiny fragments of an ancient genome.

Many of the sequences had turned out to be bacterial, but that’s exactly what the geneticists expected. But comparing the sequenced fragments with human, chimpanzee and mouse genomes, a good percentage of them looked primate. Then came the work of assembling those sequenced fragments into longer pieces. Eventually, if they managed to extract enough fragments, the geneticists would be able to sequence the entire Neanderthal genome.⁶

Analysis of Neanderthal DNA should be able to cast light on many areas of enquiry, not only the question of hybridisation. By comparing the differences between Neanderthal and modern human DNA, the geneticists can estimate the time of the ‘split’ between the lineages. At the moment, in Leipzig, that’s looking as though it happened some time around 516,000 years ago. This is older than the split suggested by fossils, at about 400,000 years ago – but that’s unsurprising. The genetic split would have happened in a population that was still ‘together’.

This is ground-breaking science, so it’s not surprising that there are still problems that need to be ironed out. And probably the most tricky one is that problem of contamination with modern DNA, which could skew results. Pääbo’s Leipzig lab isn’t the only place where Neanderthal genome sequencing is going on. A team led by Edward Rubin, in California, are also at it – and they published their first chunk of Neanderthal sequence in the same week as Pääbo’s team. But they came up with different results and a different – even earlier – prediction for the divergence of Neanderthals and modern humans, of around 706,000 years ago.⁷ So it seems that, even with all that careful screening, some contamination may have crept in, explaining the earlier dates coming out of the Leipzig lab.⁸ With each lab acting as a check on the other, though, the scientists hope that they will be able to overcome these teething problems.⁹ The Californian dates may seem very early indeed, but it’s important to remember that this is the predicted date of divergence of the mtDNA lineages, not of the actual populations. Based on this genetic data, Rubin’s team estimated that the population split happened about 370,000 years ago, which is quite a good match with the fossil data.

Another potential application for ancient DNA is in identifying bone fragments that are too small to characterise on the basis of size and shape. In fact, this has already been applied to fossils from at least two sites. A child skeleton from Teshik Tash in Uzbekistan has often been held up as the most easterly example of a Neanderthal, but some have disputed its credentials. Even further east, bones and teeth from Okladnikov Cave in Siberia, found alongside Mousterian tools, were too broken up for it to be decided if they were modern human or Neanderthal. Genetics to the rescue, then. Scientists working in labs in Leipzig and in Lyons independently extracted and analysed the mtDNA from the bones from both sites. The results showed that the Teshik Tash child had Neanderthal mtDNA, and so did two of the bone fragments from Okladnikov.¹⁰ This study was very significant: it hugely extended the known range of Neanderthals to the east, right into Central Asia. Maybe they even got to Mongolia and China. Genetic analysis is clearly an exciting addition to the toolkit of the Palaeolithic archaeologist.

There is also exciting potential for finding out – at some point in the distant future, when we know a lot more about the functions of genes in us and other animals – more about Neanderthal biology.⁶ But even now we know that at least some Neanderthals possessed a version of a gene that probably gave them red hair. The gene in question is melanocortin 1 receptor (or ‘mc1r’). In modern humans today, mutations that impair the function of this receptor gene produce red hair and pale skin. A team of geneticists managed to extract DNA – including part of the mc1r gene, from two Neanderthal fossils, one from Spain and another from Italy. Both fossils contained a mutated version of the mc1r gene, different from any of the variants seen in modern humans. To see what effect this gene would have, scientists inserted it into cells in the lab and found that it had a partial loss of function – like the other variations in the mc1r gene that produce red hair in humans today.¹¹ It is important to note that this is a different mutation from that in modern human redheads. It doesn’t imply any genetic mixing between Neanderthals and modern humans, and it certainly doesn’t suggest that the redheads among us are Neanderthals!

Another particular gene that has been identified in Neanderthals is FOXP2. This is a gene that has two specific differences in humans compared with other living primates. People missing out on those human-specific changes to FOXP2 have problems in both producing and understanding speech. Analysis of FOXP2 in living people suggested that it appeared and swept through the human population about 200,000 years ago, which seemed to fit quite well with the appearance of modern humans in Africa. It suggests that ‘modern’ language and symbolic behaviour are uniquely human attributes, with a biological basis. Eric Trinkaus took issue with this interpretation. He argued that there was evidence for symbolic behaviour in the Neanderthal archaeological record, with intentional burial, for instance. And he found it hard to imagine how complex subsistence strategies would have appeared – from around 800,000 years ago – without complex social communication. And yet the ‘human’ version of FOXP2 was initially estimated to have arisen well after the split between modern human and Neanderthal lineages.¹² But a recent DNA study of two Spanish Neanderthal fossils showed that they both carried the ‘human’ form of FOXP2.13 For Trinkaus, this showed that the ‘much maligned Neanderthals’ had a degree of human behaviour that was reflected in the archaeological record but that he felt had often been played down. But how can we explain the same version of FOXP2 existing in both modern humans and Neanderthals? Either it is much older than the earlier studies suggested, and was present in the ancestors of modern humans and Neanderthals, or it has passed from one population to the other by gene flow. The latter seems very unlikely as no other genetic studies to date had produced any evidence of gene flow.¹³

But what about the ambitious Neanderthal Genome Project? Was there any evidence for hybridisation emerging from the nuclear DNA? The key to looking for evidence of hybridisation was to concentrate on genes or other bits of chromosomes that are specific to modern Europeans (and this is a tall order as most genetic differences are shared between populations across the globe rather than being specific to one area), keeping an eye out for these sequences in the Neanderthal genome. If any European-specific DNA sequences were found in Neanderthals, this would strongly imply that there had been some sharing of genes between Neanderthals and modern humans in Europe.

When I visited the Max Planck Institute in the early summer of 2008 Ed told me that they had managed to sequence about 5 per cent of the Neanderthal genome. I asked him a difficult question, considering that the Neanderthal Genome Project was still such a long way from completion: ‘If chimpanzees are about 1.3 per cent different from us, in terms of the sequence of DNA, do you have a feeling for how different the Neanderthal genome is going to be from ours?’

‘Yes, we do,’ he replied. ‘It’s looking about ten times closer than the chimpanzee. But Neanderthals are so closely related to us, it’s hard to speak in terms of percent differences. It really depends on which Neanderthal and which human you’re talking about.’

‘And have you seen any suggestion at all of hybridisation with modern humans?’ I asked.

‘No. There’s no evidence to date of any hybridisation between modern humans and Neanderthals,’ he replied. ‘But by the end of the summer we should have 65 per cent of the Neanderthal genome, so we’ll be able to give a much more definitive answer then.’

This question of what happened when modern humans walked into Neanderthal territory was fascinating. I asked Ed what he would have done if he’d met one of our cousins.

‘If I came face to face with a Neanderthal, the first thing I would do is ask for a DNA sample,’ said Ed, ever the scientist.

So far, then, Neanderthal genetics has shed light on how far this ancient species ranged across Europe and Asia, has shown that they possessed the same ‘language gene’ as modern humans (although it must be stressed that the development of language cannot be linked to just one gene), and that some of them had red hair. And, bearing in mind that there was still a lot of genome left to sequence, there was no evidence – yet – for any mixing between Neanderthals and modern humans in Europe. (Nearly a year after I visited Leipzig, Svante Pääbo announced the completion of the first draft of the Neanderthal genome – 63 per cent of it, over three billion bases – at the annual meeting of the American Association for the Advancement of Science in Chicago. There was still no sign of interbreeding with modern humans.)¹⁴

But it’s also important to remember that the conclusions from genetic studies like this can never rule out any hybridisation. Perhaps it’s just that Neanderthal lineages have not survived to the present day, and maybe some Neanderthals had modern human genes – just not the ones whose genomes were being sequenced.

Does this make the whole endeavour futile? Far from it. If there is no evidence of mixing, then we can at least say that hybridisation didn’t happen at a level that we could consider to be significant, and so it cannot explain the apparent disappearance of Neanderthals from the fossil and archaeological record: they cannot have been absorbed and assimilated into ‘modern’ populations.

Thus far, all the genetic studies suggest that any hybridisation was, at the most, insignificant. And, actually, when you take a closer look at when and where Neanderthals and modern humans were living in Ice Age Europe, this makes some sense. There are only two areas where the dates for modern humans and Neanderthals actually coincide: in southern France and in south-west Iberia, in the period between 25,000 and 35,000 years ago.¹⁵ Even then, they could have missed each other by hundreds or thousands of years, so the opportunities for inter-species sex would have been extremely few and far between anyway. So it’s not really surprising that no ‘Neanderthal’ genes have been found in the modern gene pool – or vice versa.

So that means that the Neanderthals – whether or not rare liaisons led to hybrids whose existence has now been expunged from the modern gene pool – really did disappear. But why did the Neanderthals, who had been living in Europe for hundreds of thousands of years, fade away when modern humans arrived on the scene?

I needed to look more closely at the archaeological evidence: was there any difference in the way modern humans and Neanderthals were subsisting in their environment? Was there anything that could have given modern humans ‘the edge’ in Europe?

Treasures of the Swabian Aurignacian: Vogelherd, Germany

In a complete contrast to the ultra-modern Institute in Leipzig, I next visited the medieval university town of Tübingen. I walked up cobbled roads to a castle where I passed through a great arch into a courtyard, then on past a fountain and up stone steps, then turned a corner to enter the Department of Early Prehistory and Quaternary Ecology. At the end of a corridor plastered with posters of wonderful carved animals and birds, I found Professor Nick Conard in his office.

Nick’s office was lined with red cupboards on one side, dark wooden bookshelves on another, and wooden filing cabinets. There were two desks, each piled high with papers and books, and in one corner was a large grey safe with a map of the Swabian Jura hanging on it. Nick had spent years excavating sites around Tübingen, where he had discovered evidence of the earliest modern humans in Europe. But it wasn’t just stone tools that he’d found: there had been some rather wonderful pieces of art and musical instruments. And he had some of them in the safe. I had to look away while he found the key and then started bringing small cardboard boxes over to a low table, where we sat down to open the boxes of treasures.

The first object Nick took out, dating to around 35,000 years ago, was an ivory flute. It was discovered in 2004, at a cave site called Vogelherd, lying beneath two other flutes that had been made from hollow swan bones. The ivory flute had taken much more craftsmanship, though: it had been carved out of a mammoth tusk, then split to hollow out the inside, and joined back together with something like birch pitch. There was a row of incised notches down each side, crossing the join, perhaps made to help when putting the two halves back together.

The ivory flute had been smashed up into fragments, which archaeologists had found and carefully pieced back together; the notches had also helped the archaeologists when it came to reconstructing the flute. Nick explained that, using mammoth ivory, the instrument-maker wouldn’t have been constrained by the dimensions of a hollow bird bone and so could make a much larger, longer instrument. But it also seemed to be an exhibition instrument – designed to show off the technical skill of the instrument-maker. Nick had been completely taken by surprise by this discovery. They had found mammoth ivory carvings in Vogelherd before, but this was the first indication of music that had emerged from the site. The three small flutes represented the first real evidence of music – anywhere in the world. Nick had a replica of one of the swan-bone flutes, which I tried to play with less than impressive results, not being any sort of musician. But I could at least get a series of notes out of it. More accomplished musicians have tried and produced music that sounds quite harmonious to the modern ear, with tones comparable to modern flutes or whistles.

Opening the other boxes, Nick brought out some finds from the 2006 digging season at Vogelherd, and from the nearby cave of Hohle Fels – beautiful things nestled into cut-to-fit shapes in foam inside each box. Nick lifted out a tiny ivory mammoth, just 3cm long. It was carved in the round, with naturalistic detail, its trunk hanging down and curving over to the right, and there was a tiny spike of a tail. The hind legs were shorter than the front. It seemed perfectly proportioned. The bottom surfaces of the feet were scratched in a crisscross pattern.

Then there was a lion carved in relief, again in ivory, with hatching along its back. It had a long body, and its hackles were raised. And a tiny, beautiful bird. The body of the bird had been discovered in earlier digs, and there had been much speculation about it. Was it a human torso? But then the archaeologists had discovered the head and neck – a minute fragment that could so easily have been passed over. But it fitted the body, and, suddenly, there was a bird, perhaps a duck or a cormorant, with its neck outstretched. Finally, from another small box, Nick carefully lifted out a minute lion-man. Standing just over 2cm tall, he looked like a miniature version of the famous lion-man from Hohlenstein-Stadel, near Ulm – around the corner from Hohle Fels. All of these objects dated to more than 30,000 years ago.¹

But there was one more surprise. Nick opened a long box, and inside it was a long, smooth piece of stone, unmistakably carved into the shape of a penis, with the foreskin and glans carved into it at one end. We contemplated this bizarre object. Was it a hammer stone, carved in a phallic shape as a joke? Or could it be that this stone had a functional use more related to its shape? Nick was quietly amused by the find. It suggested that the people of the Swabian Aurignacian had, at the very least, a healthy sense of humour, and perhaps an even healthier sexual appetite.

The art of the Swabian Jura was fascinating, and this really is the earliest evidence of something that we can properly appreciate as art. I had seen pierced shells and ochre ‘crayons’, leaving us guessing what was drawn with them, but here were carefully executed carvings of animals, and strange therianthropic beasts – men with heads of lions. Nick said that the styles of these Aurignacian carvings were similar across different sites in the Swabian Jura, although there were many different themes. It seemed to be a time of some artistic experimentation. But recurring imagery like the lion-men from Hohlenstein-Stadel and Hohle Fels also suggested very strongly that they were made by people from the same cultural group in the Lone Valley. Many different ideas have been put forward about the meaning and function of these artefacts: some have suggested that they indicate hunting magic, and the therianthropic figures in particular have been linked to shamanism. For Nick, the discovery of the tiny waterfowl carving challenged previous interpretations of Aurignacian carvings from the Swabian Jura as representing fast and dangerous animals, with whom Palaeolithic hunters may have identified.¹

‘I think the combination of these symbolic artefacts, ornaments, figurative representations and musical instruments, shows us these people have the mental sophistication of ourselves, the same creativity that we have,’ he said. ‘And we can even get insights into the system of beliefs. For instance, the examples of human depictions combined with lion features show that, at least in their iconography, they were engaging in transformation: people having a connection with the animal world, being depicted as mixed animal/human figures.’

But how could these small ivory carvings hold any clue to the survival of modern humans – and the demise of the Neanderthals? Well, certainly the Neanderthals, however intelligent and whether or not they had language like us, never produced anything like the objects found at Vogelherd. I asked Nick about the differences between modern humans and Neanderthals, and it became clear that he thought culture had played a key role in the expansion of modern human, and contraction of Neanderthal, populations, during the late Pleistocene.

On their own in Europe, the Neanderthals seemed to have been getting along just fine.

‘The Neanderthals were the indigenous people of the area. They had very sophisticated technology, certainly command of fire, and knew how to get along in their environment. They had everything 100 per cent under control, and they were doing very well,’ said Nick.

‘So if they were so good at surviving in Ice Age Europe, why did they disappear?’ I countered.

‘Well, I would approach that question from an ecological point of view. If you have one organism occupying a niche, it’s going to stay there until something drives it out of its niche: either environmental change that makes it impossible to occupy the area, or another organism coming in and competing for resources.’

‘So you’re saying that modern humans were that competing organism?’

‘Well, yes. It’s very clear that Neanderthals and modern humans were really occupying the same niche. We see that unambiguously in the archaeological sites: the diet consists of the same foods – especially reindeer, horse, rhinos and mammoth.’

‘But why did we modern humans survive and not Neanderthals?’

‘Well, there’s no question the Neanderthals were very effective hunters, and really were at the top of the food chain. But we do see some differences in technology. I think that the innovations that modern humans developed in Europe, the Upper Palaeolithic toolkit, organic artefacts, but also figurative art, ornaments and musical instruments – these are all things that seemed to help give them an edge against the Neanderthals.’

I found it hard to imagine why art and music might have given modern humans an advantage.

‘Well, think of the lion-man,’ said Nick. ‘There’s a lion-man from this valley, and a lion-man from the Auch Valley. It’s the same iconography, the same system of beliefs, the same mythical structure, and they’re the same people. And we don’t see those kinds of symbolic artefacts with Neanderthals, so it seems that their social networks were much smaller than those of modern humans.

‘And from my point of view,’ he continued, ‘the evidence even at this time, 35,000 years ago, is completely unambiguous: music was a really key part of human life. It’s not entirely clear how that would give you a major biological advantage over the Neanderthals, but it seems to fit into this complex of symbolic representation, larger social networks. Perhaps music helped to form the glue that held these people together.

‘When the competitor arrived, the Neanderthal way of doing things wasn’t as effective in the face of people who had new ways of doing things, new technology, new culture and social networks,’ explained Nick.

Whereas competition for an ecological niche seemed to have spurred modern humans on to develop wider social networks, the Neanderthals appeared to be ‘culturally locked in’. It was a competition that modern humans would eventually win. Nick explained that, while their respective territories probably shifted back and forth over the centuries and millennia, Neanderthals were, on average, retreating while modern humans expanded.

‘In regions like the Levant, we have good evidence for movement back and forth of the two populations. It’s certainly not the case that modern humans always immediately expanded at the cost of Neanderthals; there are some good examples of Neanderthals displacing early modern humans, too.

‘When the new people came in, resources got tight, and modern humans were able to develop new technologies and new solutions quicker than the Neanderthals. In a sense there was a continual cultural arms race going on. And here, in this setting, it seems like a lot of innovations took place that gave the modern humans a bit of an edge. But it wasn’t a sudden, blanket devastation of the Neanderthals: there was a lot of give and take, but, ultimately, they were pinched out demographically.’

‘So do you think modern humans and Neanderthals were actually in contact with each other?’ I asked.

‘Well, in some areas, there were fairly dense populations of Neanderthals. And I think they did meet. And I think they would have been checking each other out from a distance, often avoiding each other. That was probably the most common scenario, but there may have been times when they came together, in peaceful co-existence, and times when there was quite a bit of conflict.’

‘What do you think about the question of interbreeding?’

‘Any place where people come together, interbreeding is the most normal thing in the world. So I think there were occasionally encounters where interbreeding took place, but not very often, and so it didn’t contribute very much to our genetic make-up or our anatomy.’

While it’s difficult – even impossible – to summarise the interactions that may have occurred between the two populations over so many thousands of years, the question ‘Why did we survive to the present while Neanderthals disappeared?’ is still relevant. Even if members of the two species never came face to face, they were in competition with each other in the landscape. And there was archaeological evidence for different subsistence strategies, which, for modern humans, included different and possibly more flexible technology, as well as culture and complex social networks, which may ultimately explain why we are here today and the Neanderthals aren’t.

Later that day, Nick took me to Vogelherd itself, in the lush Lone Valley, where excavations were ongoing. A team of archaeological technicians and students were busy digging down through the spoilheap (or ‘backdirt’) of the original excavation, finding plenty of evidence that had been discarded by the first archaeologists who dug there.

‘The site was first dug in 1931, and all of the material was dumped outside the cave,’ explained Nick as we walked past the cave entrance. ‘We’re systematically digging through it all to find out what they missed.’

It looked like a very pleasant place to be digging. The cave was set on a hill above an idyllic, lush, green valley. I asked Nick what it would have been like 35,000 years ago.

‘If you’re talking about the Ice Age, you think of ice: white, stark, and inhospitable. That’s wrong. I mean, it was cold in the winter, but in the spring and summer it would be more like it is today: lots of grass, greenery, really abundant fodder for the animals. Just think about a mammoth: the archetypal animal of the Ice Age. A mammoth eats about one 150 kilos of grass every day to stay alive. The mammoth steppe was a very rich environment, and at these sites in the Lone Valley we see abundant remains of woolly rhino, mammoth, reindeer, horses, all kinds of animals.’

‘But it must have been very cold during the winter here?’ I suggested.

‘Well, yes. But humans – and Neanderthals – can live almost anywhere as long as there’s something to eat, and materials, particularly hides, to make clothing out of, and controlled use of fire.’

Nick wandered around the site, visiting trenches to see what finds had emerged that day. These included fragments of flint blades, quite typical of Aurignacian toolkits. The sediment was being bagged up as it was removed, and would be sieved. It was only through this careful sifting of the soil that Nick’s team had found the fragments of ivory that made up the ivory flute, and the head of the bird carving.

But Vogelherd had contained disappointments as well as revelations. On first excavating it in 1931, archaeologists dug out some 300m³ of sediment from inside the cave, finding Middle and Upper Palaeolithic artefacts in distinct layers. The latter included a rich collection of Aurignacian tools and artefacts, which have since been radiocarbon dated to 30,000 to 36,000 years ago. The archaeologists also found modern human remains including two crania and a mandible, embedded in the Aurignacian layer. These bones, in association with the Aurignacian tools, seemed to provide conclusive evidence that modern humans were the makers of this technology. The findings from Vogelherd tallied with the discoveries at the Cro-Magnon rockshelter in France, where modern human skeletal remains – ‘Cro-Magnon Man’ – had been found associated with Aurignacian tools.² This link between modern humans and a particular technology meant that archaeologists could assume the presence of modern humans, in the absence of skeletal remains, when they found Aurignacian tools and artefacts. At other sites, Neanderthal remains had been found associated with Mousterian (Middle Palaeolithic) tools. So it seemed that each of these populations had a clear ‘signature’ that archaeologists could use to map out their sites and territories in Europe.

In 2004, Nick Conard and his colleagues published radiocarbon dates for the skeletal remains from Vogelherd: they dated to a mere 4000 to 5000 years ago. It looked as if they were intrusions from late Neolithic burials near the cave entrance. The ‘association’ with the Aurignacian layers in the cave was incidental. This disappointing result had wide implications. Vogelherd had been a key site for demonstrating that modern humans made the Aurignacian. And in 2002, radiocarbon dates had been published for the Cro-Magnon skeletal remains as well, showing them to be about 28,000 years old: too young for the Aurignacian tools in the rockshelter, although nowhere near as young as the Vogelherd bones had turned out to be.² The identification of modern human sites through Aurignacian tools alone was starting to look decidedly shaky.

Vogelherd, along with a generous scattering of other Aurignacian sites, including Hohle Fels and Geissenklösterle in Germany, and Willendorf in Austria, had also been used to support the theory that the ‘Danube Corridor’ provided a route for the early modern human colonisation of central Europe.³ But the radiocarbon dates of the Vogelherd and Cro-Magnon bones meant that archaeologists could no longer assume that the spread of Aurignacian technology represented the ingress of modern humans in Europe. As shocking a suggestion as it may seem to Palaeolithic archaeologists who have relied on Aurignacian tools as signs of modern humans, there was now nothing to suggest that this was a reasonable assumption to make. Indeed, there was nothing now to refute the hypothesis that Neanderthals might have made those tools and even those beautiful ivory carvings and flute from Vogelherd.²

When I interviewed Nick, however, he did not seem like a man about to consign a lifetime’s work to the dustbin of prehistory. There was still reason to think that it was modern humans who made the Aurignacian. The Oase skull and mandible showed that modern humans were in Europe, close to the Danube, by 40,000 years ago. The Aurignacian appears suddenly in Swabia, and is always ‘on top of ’, i.e. later than, Mousterian (Middle Palaeolithic) archaeology, and in some places where both Middle and Upper Palaeolithic artefacts have been found, there has been a distinct gap, an ‘occupation hiatus’ between them.¹ It seems too much of a coincidence to think that the existing, Neanderthal population of Europe would have started manufacturing a completely new-looking toolkit just as modern humans arrived on the scene: it seems more likely that it was the moderns who brought that technology in.

And there were still some sites that seemed to hold up that association between modern humans and the Aurignacian. The site of Ksar ‘Akil in Lebanon is important in this respect, as it produced a modern human skeleton alongside that ‘transitional’ half-Middle, half-Upper Palaeolithic ‘pre-Aurignacian’ industry. The burial dated to 40,000–45,000 years ago, and the layers overlying the skeleton were full of classic Levantine Aurignacian tools.^4,5

The site of Mladeè in the Czech Republic, which was first excavated in the nineteenth century, has produced over one hundred modern human fossils, in association with classic Aurignacian tools, including bone points. In 2002, a date was published for the calcite overlying the skeletal remains: 34,000–35,000 years ago.⁶ This looked like a good candidate for reaffirming the link between moderns and the Aurignacian, but a direct date on the bones themselves was needed to clinch it. A few years later and a radiocarbon date for the fossils themselves was published, placing them at around 31,000 (uncalibrated radiocarbon) years old.⁷ Although less well dated than Mladeè, there are also several French sites where modern human remains and Aurignacian tools have been found together: Les Rois and La Quina in the Charente, and Brassempouy in the Pyrenees.⁵

Palaeolithic archaeologists everywhere must have breathed a huge sigh of relief. Once again – and actually even more surely than before the Cro-Magnon and Vogelherd redating upset, now that it was based on direct dating of human remains – the assumption that the Aurignacian – everywhere – was made by modern humans seemed well founded.

‘We don’t have Neanderthal bones, or modern human bones, here,’ said Nick. ‘We have dates that correspond to the period of the last Neanderthals and the first modern humans – in theory, it could be either. But the most plausible explanation is that it’s modern humans.’

Having seen the beautiful artefacts from Vogelherd, I left the dig and Nick Conard, and drove a mile or so down the valley to meet experimental archaeologist Wulf Hein for a practical lesson in the differences between Middle and Upper Palaeolithic technology. I wanted to get to grips with the Mousterian and the Aurignacian – literally.

Wulf Hein was an expert flint knapper and aficionado of all things Palaeolithic. His car was full of crates of flint, various spears, spear-throwers, bows and arrows. The finished objects were beautiful, but I really wanted to know how the flint tools were made, and to see the Levallois and prismatic core techniques in action.

We sat in a field in the idyllic Lone Valley, having spread out a blanket to catch any stray bits of flint. In this archaeologically rich area, Wulf was very concerned that he didn’t confuse any archaeologists in the future by adding new, twenty-first century flint tools to the archaeological record. He took a Levallois core out of a crate; it was flattish and he had made it into the shape of a tortoise shell by knocking flakes off the periphery – making it into a ‘prepared core’. He struck it with a pebble – expertly – and a large flake detached itself from the middle of the disc-like core. Middle Palaeolithic technology starts off with simple flakes struck from cobbles. The Mousterian industry of the Middle Palaeolithic takes this a stage further, with flakes produced from a prepared (Levallois) core, just as Wulf Hein had demonstrated. It is named after the site of Le Moustier in the Dordogne, where Neanderthal fossils were found alongside their characteristic tools.

Wulf then took a stone that had been carefully prepared into a cone shape out of the crate, and handed it to me: I was to make a blade from this prismatic core. Under his expert guidance, I gripped the core between my knees, held the tip of a piece of antler close to the edge of the stone, and struck the antler with a pebble. A long, thin and extremely sharp blade detached itself from the side of the core and fell to the ground.

‘Oh, nice one!’ exclaimed Wulf.

‘Are you proud of my work?’

‘Yes, I am. I’m astonished. It took me forty years.’

I think my success had more to do with beginner’s luck – and having a good teacher.

There it lay, a blade, the hallmark of the Upper Palaeolithic, the foundation of the classic Aurignacian industry. This technology was named after the site of Aurignac in the lower Pyrenees, excavated in 1860. The sort of tools that characterise the Aurignacian include long, thin slivers of flint called lames Aurignaciennes (Aurignacian blades) – like my blade but retouched all around the edges, as well as end-scrapers and burins, carinate (keeled) scrapers and tiny bladelets. The functions of these tools are being debated – especially the carinate ‘scrapers’ and the bladelets that are made from them. Are the unretouched bladelets waste products from making such a scraper? Or are the bladelets really what the maker was after, and the ‘scraper’ is actually just a small core – not a tool in itself?⁵ Even though the precise functions of all these pieces of flint aren’t yet understood, the shapes are very characteristic.

Archaeologists have long believed that knapping blades from a prismatic core was much more efficient than the old-fashioned Levallois technique: the manufacturer can produce many blades from one prismatic core. In comparison, the Middle Palaeolithic Levallois technique just gives you a few flakes from one tortoise-shaped core. However, I later met up with experimental archaeologist Metin Eren, in Exeter, who had spent years making and comparing Middle and Upper Palaeolithic flakes and blades. His results had been surprising: the initial preparation of the prismatic cores actually produced more waste than discoidal cores, and thin blades didn’t last as long as flakes. So, in terms of producing usable cutting surfaces, it seems that there actually wasn’t much to choose between discoidal and prismatic cores, flakes and blades, in terms of efficiency.⁸

Now I had seen very early evidence of (small) blade production, in Africa, Europe and Arabia, going right back into the Middle Palaeolithic. But there’s more to the Upper Palaeolithic than prismatic cores and long blades. Bone and antler tools are also seen to be characteristic of this culture. These materials do (rarely) pop up in other Middle Palaeolithic contexts – like Blombos Cave and Howiesons Poort in South Africa. Perhaps this isn’t surprising for those are modern human sites.⁹ (Just to throw a spanner in the works, there’s some evidence of Neanderthals making bone points as well.) Other elements that are considered particularly characteristic of the Upper Palaeolithic are grinding and pounding stones, suggesting more processing of vegetables was going on, and widespread use of body decorations like shell, tooth and ivory beads. (Now, I had seen evidence of much earlier use of ochre and ornamentation, but before the Upper Palaeolithic it is very patchy.) There appears to have been much more long-distance transport of raw materials, sometimes over hundreds of kilometres, compared with generally shorter distances in the Middle Palaeolithic. Carved figurines – as I had seen from Vogelherd and Hohle Fels – appear in the Upper Palaeolithic, along with cave painting (more of which later). It’s not straightforward, but, as a package, the Upper Palaeolithic does seem to be something special: it is still a useful category.

Moving beyond the Aurignacian, improved hunting tools appear, like spear-throwers in the Gravettian, and, eventually, bows and arrows and boomerangs.⁹ Wulf had brought some of these along with him. The atlatl – or spear-thrower – was a very simple tool, essentially a stick, about half a metre in length, with a hooked end that fitted into a recess in the end of the spear.

‘This spear-thrower has a beautiful carved antler end. We don’t know what the original shaft looked like, but this is a reconstruction. It’s interpretation, but it works.’

‘I’d like to see how far you can throw these slender spears, with and without a spear-thrower,’ I challenged him.

‘OK, and you will be astonished.’ Wulf took up the gauntlet.

He threw a spear without the spear-thrower first. ‘That was 220 grams thrown by hand,’ he said. Then he got ready to throw another spear, fitting it into the hook and laying it above the spear-thrower, gripping them both with one hand. ‘And this is 220 grams thrown with a spear-thrower – with the same power.’

He threw it – rotating the spear-thrower so that it became like an extension of his arm – and the spear flew off …

It was extraordinary: it had double the range. I had a go as well and amazed myself by how much further I could throw the spear with the atlatl. It was such a simple but impressive piece of kit.

‘If you’re trained you can get it even further. The record is 180 metres,’ said Wulf.

Wulf had more examples of Stone Age projectile weaponry with him.

‘At the end of the line stands the bow and arrow,’ said Wulf. He had a beautiful replica of a Mesolithic bow with him. It was found in Denmark, preserved in a bog called Holmegaard, near Copenhagen. ‘The original is about 8,600 years old. It’s the oldest bow we’ve ever found.’

Even earlier arrows have been found, dating to more than 11,000 years ago, and some archaeologists argue that there is evidence, albeit fragmentary, going even further back, into the Upper Palaeolithic. It seems that the development of bows and arrows was related to ecological changes: they appear as the world was warming up after the Ice Age, and Europe was becoming wooded. ‘If you’re hunting in the woods, a bow and arrow is much more effective than a spear-thrower. And it’s easier to aim,’ said Wulf.

I enjoyed playing with Wulf’s ‘Mesolithic’ bow and flint-tipped arrows, and I am very sorry to say that two were left behind – somewhere in the long grass by the stream in that field; after so much time had been spent looking for them, we had to accept that the arrows were gone. Perhaps they will be found by archaeologists one day.

Neanderthals, as far as we know, never made spear-throwers or bows and arrows, although they did progress from using thrusting spears to throwing spears. But does that really indicate an innate technological superiority of modern humans? Stone Age weapons expert John Shea has argued that the development of true projectile technology was key to our species’ ecological success, giving our ancestors an advantage in hunting and even providing them with long-range weapons that could be used to eliminate rivals, of our own species and perhaps others as well.¹⁰ But there’s no direct evidence to suggest that spear-throwers were ever used against Neanderthals. And the first sign of the bow and arrow is well after the LGM, when Europe was warming up and woodland was returning – long after Neanderthals had disappeared from the landscape.

But I’m getting diverted here by later developments of the Upper Palaeolithic. Sticking with the Aurignacian, which was contemporary with the Mousterian, there is still a distinct difference between Middle and Upper Palaeolithic. Bar-Yosef ⁹ argued that the characteristic elements of the Upper Palaeolithic are ‘evidence for rapid technological changes, emergence of self-awareness and group identity, increased social diversification, formation of long distance alliances, [and] the ability to symbolically record information’.

Across Europe, then, the change from Middle to Upper Palaeolithic cultures, between 40,000 and 30,000 years ago, is taken to represent the replacement of Neanderthals – bearing Middle Palaeolithic technology – with modern humans, carrying with them Upper Palaeolithic tools and artefacts. It has been called the ‘Upper Palaeolithic Revolution’, but this is a problematic label as it suggests that the late Pleistocene, in Europe, is the time and place of some kind of emergence of ‘fully modern’ behaviour. This is difficult to argue, as this pits modern humans in Europe not only against the Neanderthals but also against modern humans elsewhere. Were the people in Africa and Asia, still making Middle Palaeolithic stone tools, cognitively inferior to the Europeans? This is obviously a very divisive and Eurocentric viewpoint, and recalls that comment by Movius about Eastern Asia being an area of ‘cultural retardation’. It seems much more likely that modern behaviour was essentially born at the same time as our species, and that modern humans came out of Africa, as Oppenheimer puts it ‘painting, talking, singing, and dancing’.¹¹ And, as we have seen, the African evidence supports this idea.

But there is no doubt that the Aurignacian represents a new sort of culture. So how can we explain it? Well, culture is something that represents human interaction with the environment, and with other humans, so changes in culture can be seen as being driven by changes in climate and environment, as well as changes in society – and rather than a biological change. The Aurignacian didn’t require a new brain, or a few new genes. It was a new product of a brain that was already well equipped to develop behavioural solutions to environmental challenges. We have seen that the hallmark of modern humans elsewhere was ingenuity, adaptability and inventiveness. So it seems utterly reasonable to suggest that, in Europe, the new, modern human culture of the Aurignacian represented adaptations to a new environment – perhaps including the existence of a competitor in the landscape.

The Aurignacian objects that have been passed down to us from our earliest ancestors in Europe show us that much more was going on on top of the change to a new method of tool manufacture. They show us that those people were flexible and adaptable, and forming complex social networks. They seem to indicate that these people felt part of something which extended far beyond their immediate families and familiar landscapes. Archaeologists argue endlessly about classification and naming of toolkits. The arguments can seem – especially to someone more schooled in bones than stones – incredibly complicated and esoteric. But the arguing itself means that this was an interesting time: several things are happening all at once: new stone tools are appearing in Europe, and at the same time there are new social structures emerging.⁹ Maybe the new styles of stone tools are part of a badge of identity, a fashion, for those early modern humans in Europe.⁸

The Aurignacians spread inexorably across Europe. Perhaps at times the Neanderthals regained territory, but, ultimately, they were to disappear from the landscape. Their traces gradually contract and disappear, and the last Neanderthals seem to have occupied a lonely outpost in the very south-west corner of Europe: Gibraltar.

Tracking Down the Last Neanderthals: Gibraltar

The first Neanderthal fossil to be found was in Gibraltar – a skull blasted out of Forbes Quarry, in 1848¹ – but no one recognised it at the time, so the German find, six years later, had the honour of giving the species its name.

The Gibraltar skull, probably that of a female, remains one of the best-preserved Neanderthal fossils. In 1926, another Neanderthal fossil – parts of the skull of a four-year-old child¹ – was excavated from the Devil’s Tower site on Gibraltar. So Neanderthals certainly lived on the Rock – and more recent excavations have revealed quite challenging details about their way of life, and even why they might have died out.

As I flew into Gibraltar, it was much smaller than I expected it to be: a 6km-long rocky headland sticking off the south-west corner of Spain into the Mediterranean. The runway of the airport was crossed by the main road into Gibraltar. From the plane window I could see the white cliffs of the limestone outcrop, dotted with square caves. The Rock is riddled with caves and tunnels, some natural, but many man-made. A great deal of them date from the Second World War, when civilians were evacuated and Gibraltar became a military fortress. The military presence on Gibraltar is still very evident today, but the Rock is also home to nearly 30,000 people. Sheer cliffs, more than 400m tall, rise out of the sea on its east side, but on the west the houses and hotels of the town cluster at the foot of a more gentle slope.

I had travelled to Gibraltar to see Clive and Gerry Finlayson, and we met up on the terrace of the Rock Hotel, looking out over the harbour. They had studied the Gibraltar Neanderthals for years, and were keen to set the record straight when it came to perceptions of these ancient people as brutish, cold-adapted savages. In fact, the evidence emerging from Gibraltar seemed to indicate people much more like ourselves.

Clive’s background as a zoologist and ecologist informed his approach to Palaeolithic archaeology. He wasn’t into constructing theories based on stone tool typologies, but favoured a landscape-first view where humans (modern and Neanderthals) were firmly placed in their environmental context. He saw differences in technology as very much driven by ecological and social changes – not the other way around. From this perspective, technological and cultural differences represented quantitative rather than qualitative differences between the people making them. He was clearly not a supporter of the Upper Palaeolithic revolution, and argued that there was no sudden appearance of ‘modern behaviour’. He wasn’t suggesting that there were no differences between modern humans and Neanderthals, but rather that those differences were shaped by the environments in which cultures emerged, and by the social structures adopted by each population.² The Finlaysons’ research in Gibraltar had provided evidence not only for late survival of the ancients in this corner of Europe, but for something which I had, perhaps naively, previously thought of as being exclusively ‘modern’ behaviour: the Neanderthals had beachcombed.

Early the following morning, I met up with Clive down at the marina, and we took a boat out, heading around the point to the cliffs on the east. As we set out, the sea was as smooth as glass, and we seemed to glide along, the Rock bathed in golden sunlight and looking magnificent. There were natural caves, half in, half out of the water, all along the east side of the Rock.

‘There are over a hundred and forty caves in the whole of the Rock, and there’s probably twenty or thirty just along this cliff,’ Clive told me. ‘We get the feeling this was almost like a “Neanderthal city”. On Gibraltar, we’ve got ten sites, two with fossils and the other eight with tools: occupation sites. That’s probably the highest density that’s been recorded anywhere. And bear in mind, that’s just the sites where the archaeology has been preserved to today. It’s difficult to say how many people were living here, but I imagine the Rock would have had maybe a hundred Neanderthals living on it.’

Clive wanted to show me one site in particular where he had been excavating in search of Neanderthals for several years – with very successful results: Gorham’s Cave.

‘Most of the evidence is submerged or has been lost. So we’re lucky to have Gorham’s Cave: the sea hasn’t washed the sediments away there. And we’ve found a lot of evidence of Neanderthal occupation: stone tools, animals that they’ve butchered, hearths – Neanderthal barbecues, if you like.’

The site had first been excavated in the 1950s, then had lain forgotten for decades. Clive had started excavations again in 1991.

‘We come every year and excavate the site. It’s huge. There’s eighteen metres’ depth of archaeology in there, and there are new results coming out every year.’

The results of radiocarbon dating on charcoal from the cave showed that it was occupied until 28,000, and perhaps even as late as 24,000 years ago. It made this cave the last known Neanderthal outpost.³

Some archaeologists have argued that these late dates for the uppermost Mousterian level in Gorham’s Cave are due to contamination with charcoal from even higher, Upper Palaeolithic layers. But Clive has argued that the stratigraphy inside the cave is reliable, and that indeed, after the last Middle Palaeolithic occupation, the cave seemed to have lain abandoned for 5000 years. After 19,000 years ago, the cave was reoccupied, by modern humans bearing a later Upper Palaeolithic toolkit called the Solutrean.³

Clive didn’t like the idea of Gibraltar as a ‘refugium’ for Neanderthals. ‘When we talk about a “refuge”, the impression is almost that this is a place that Neanderthals came because there was nowhere else to go,’ he said. ‘The reality is that this was a good place to be. And for a period of 100,000 years, there were Neanderthals living here.’

Even when Europe was beginning to feel the full chill of the LGM, between 28,000 and 24,000 years ago, south-west Iberia was experiencing a mild but still balmy Mediterranean climate, with mean annual temperatures of around 13–17 degrees C – in fact, very similar to today.^3,4

The large caves that now lie right on the rocky coast would have been set back from the seashore 25,000 years ago. Offering protection from the elements and from predators, I imagined the caves would have made quite sought-after Palaeolithic homes. Hearths had been lit deep inside Gorham’s Cave – but it was tall enough for smoke to rise to the ceiling and find its way out without getting into the eyes and throats of the cave-dwellers. And there were plenty of resources just on the doorstep. The sea would have been much lower than it is today, with a mosaic of habitats including woodlands and wetlands between the foot of the cliffs and the coast. The sandy coastal plain would have been dotted with sand pines and junipers, and liberally watered with streams. ‘The vegetation also indicates seasonal pools,’ explained Clive, ‘and that’s also borne out by animal remains that we’ve found, including newts and frogs, and waterfowl like ducks and coots.’

We had hoped to access the cave from the sea, but just an hour or so after we had set out on a millpond, the winds and currents had conspired to create a sizeable swell that would make any landing attempt dangerous. We continued northwards in the boat, Clive pointing out other caves that had been home to Neanderthals. But, of course, with the change in sea level, it was likely that a lot of evidence was under the waves. And Clive and Gerry were looking there, too.

Turning around, we headed back around Europa Point and into the Bay of Gibraltar, to rendezvous with the dive boat. Already on board were Gerry and marine ecologist Darren Fa. As we transferred on to the boat, I learnt that work had already started: somewhere beneath us, two underwater archaeologists were busy trying to move a large rock, using air-filled lift bags. Gerry was hoping that removing this rock would reveal an undisturbed layer of sediment. Surveys of the seabed had revealed a large reef, some 20 to 40m beneath the waves. Caves in this reef would have been on the coast – rather than underwater – during the time of the Neanderthals. The archaeologists were interested in collecting samples which would provide even more information about the palaeoenvironment, but they were also hoping that they might find signs of ancient occupation hidden beneath the seabed. Mounting an archaeological investigation of this type requires expertise in diving, marine ecology, underwater archaeology – and patience.

‘The logistics involved are really difficult,’ said Gerry. ‘It’s expensive and very time-consuming. If you’re excavating in a cave and your pencil breaks, for example, you can just go to your box of tools and get another one. If you’re on the bottom of the sea and your pencil goes, and you don’t have another, that’s the end of your dive. We take two of as many things as possible. We plan very carefully. But sometimes things go wrong, and you have to come back to the surface.’

Each diver could stay down for only an hour at most – so each day’s digging was about meticulous planning, and small but effective steps towards collecting data. They worked in shifts; Gerry and Darren got into their wetsuits and dive gear ready to step in as the first team resurfaced.

Out on the boat, I could clearly see the mountains of Morocco in the distance: the Rock of Gibraltar lies just 21km across that narrow strait from Morocco. Once again, I found myself surprised that there was no evidence that this had ever been a crossing point in Palaeolithic times. Modern humans appeared to have stayed firmly put on the African side, and there’s nothing to suggest that Neanderthals ever made the journey over to Morocco either.

When Gerry and Darren came back up, I asked them what they’d found.

‘We actually managed to move the rock using the lift bags,’ said Darren. ‘And we got some sediment,’ said Gerry. She was very pleased that they had made steady progress and managed to safely shift the rock, opening up the previously covered sediments for excavation. I had sort of hoped that one of the divers might bring up a nice Mousterian tool from the seabed on the day I was there, but that was a bit too much to hope for. It was slow and painstaking work – but essential if we are to learn more about the ancient environments that our ancestors and Neanderthals inhabited.

Back on dry land, Clive was keen to show me some of the animal bones that had been found in Vanguard Cave, the neighbouring cave to Gorham’s, occupied by Neanderthals over 50,000 years ago.

I picked up a metapodial, a foot bone. It looked a bit like a sheep bone to me. ‘That one,’ said Clive, ‘well, either it’s their favourite, or it’s easier to catch, or there’s a lot more of them than other herbivores. It’s a wild goat: Iberian ibex. Eighty per cent of the large mammal bones belong to this species.’

Red deer bones were also fairly common. ‘They were quite partial to those.’ There was also a bone from an aurochs: a massive, ancestral cow. Clive thought that bringing down an animal of this size, with its huge horns, would have required a certain degree of courage, planning and cooperative hunting. ‘It’s probably not remarkable that we don’t find that many of them,’ he said.

‘Perhaps surprisingly, given what has been said about Neanderthals, we’re also finding smaller animals. Nearly 90 per cent of all the mammal bones are rabbit.’

I did find this surprising. It didn’t fit with the traditional view of Neanderthals as big-game hunters.

‘And it’s not just land resources that they’re using,’ said Clive, rather proudly. ‘They’re eating limpets and mussels. And just look at this,’ he picked up a jaw with sharp teeth. ‘It’s a monk seal. And this isn’t an isolated case. A lot of the bones have cut marks. I think there’s too many of them to be stranded animals; I think they’re hunting them. And it gets even more complicated when you look at things like this …’

He handed me a dolphin’s vertebra. I could just make out some cut marks on one of the bony levers – the transverse processes – sticking out from the body of the bone. These weren’t trowel marks from a clumsy modern archaeologist: they were thin cuts from a flint tool. It appeared that we were looking at the remains of a Neanderthal’s dinner.⁵ So it looked like Neanderthals were just as capable of adapting to a coastal way of life as our ancestors had been. There were bird bones as well – but could Clive be sure that these had been eaten?

‘There’s very little evidence for other predator action. We don’t have cut marks on these bones, but it seems that there may be tooth marks – from Neanderthals. The dominant species are partridge, quail, ducks: the kinds of birds that you or I might find palatable. Clearly, birds are on the menu.

‘We’ve looked so much at Neanderthals in the north where large mammals were available, and it’s created a slightly biased impression,’ suggested Clive. ‘Here, big game was in the diet,’ he continued, ‘but I see the Neanderthals more as beachcombers, collectors of plants, hunters of birds and rabbits, who occasionally caught a goat, more occasionally a deer, and even more occasionally got a large, dangerous animal. Like human societies today, I’m sure the Neanderthals had cultural and geographic diversity. They were exploiting whatever was available on their doorstep.’

Other sites also challenge the idea of the Neanderthals as behaviourally inflexible. Arcy-sur-Cure in south-west France has produced stone tools that are rather weird – almost sitting on the definitional fence between Middle and Upper Palaeolithic – combining Mousterian-like flake-based tools with ‘Upper Palaeolithic’ blades and bone tools. While it appeared to have developed out of the Mousterian tradition, it was seen as marking the beginning of the Upper Palaeolithic in Europe, and used to be called the ‘Early Aurignacian’. Along with the ‘transitional’ toolkit, there were also pierced fox canines, apparently evidence of personal ornamentation. The site is not unique: the tools from Arcy are classified as Chatelperronian, after the cave at Chatelperron in south-west France where this technology was first described. Examples of this technology have been discovered throughout central and southern France, and in northern Spain. But who made it? Fossils had been found in association with Chatelperronian tools at Arcy-sur-Cure, but they were so fragmentary that it seemed impossible to tell if they came from modern humans or Neanderthals. They dated to 34,000 years ago – when both populations were in Europe.

But then, in 1996, an international team including French archaeologist Jean-Jacques Hublin and Fred Spoor from UCL showed that the anatomy inside the temporal bone – the bone at the side of the skull that contains the workings of the ear including the semicircular canals – was characteristically Neanderthal.⁶ So this suggests that it was Neanderthals who made the Chatelperronian. But did they invent it independently or copy it from modern humans, or is this a change in technology that shows behavioural adaptation to a changing environment? Clive put it to me that it represented a change from ambush hunting to projectile technology. It seems pertinent that the Chatelperronian appears only after modern humans have come into close proximity with Neanderthals – but this was also a time of major climate upheavals. In fact, both possibilities – copying or innovation – suggest a certain behavioural flexibility and intelligence that we have not always credited the Neanderthals with in the past. ‘I think we may have underestimated the Neanderthals,’ said Clive.

The Chatelperronian can be seen as rather problematic as it blurs the distinction between what archaeologists would consider to be a Neanderthal and modern human technologies. It challenges our pre-conceptions. Archaeological evidence like this makes it seem less inevitable that modern humans survived where Neanderthals didn’t – and shows that the Neanderthals were much closer to us than we’ve perhaps liked to think in the past. But they did disappear. On a continental scale, climate changes and the presence of a competitor in the landscape probably played their roles in the demise of our sister species. Their population gradually contracted, until, it seems, just a few of them were left in Gibraltar, happily living a coastal lifestyle in this idyllic corner of Europe, and probably completely unaware that they were the last of a long European lineage. So what happened to the last Neanderthals?

From Clive and Gerry’s work on Gibraltar there doesn’t seem to have been a ‘last stand’ with modern humans taking the Rock from their distant cousins. There’s a distinct gap between the last evidence of Neanderthals, at 24,000 years ago, and the first evidence of modern humans – at about 18,000 years ago. ‘There’s a gap of five thousand years when there’s nobody living in these caves,’ said Clive. So it seems that competition between the two populations could not have played a role in this particular location.

Clive suggested that, in the end, it may have just been a numbers game. If the Neanderthal population was very small, then they could have easily died out – just think of threatened species today. ‘These last Neanderthal populations would have been very small and very vulnerable,’ he said. Inbreeding and increased rates of congenital disease could also have contributed. But Clive also thought that climate played a crucial role in the demise of those late-surviving Neanderthals. ‘We’ve looked at deep-sea cores offshore here, and the most severe climatic conditions of the previous quarter of a million years hit precisely at the point that Neanderthals disappear.’

Although Gibraltar had enjoyed a mild climate up to 24,000 years ago, there then seemed to be a sudden and severe downturn in conditions. Evidence from marine cores shows a drop in temperatures at this time, known to palaeoclimatologists as the ‘Heinrich 2 Event’. These Heinrich Events are moments of ‘ice-rafting’: icebergs breaking off the northern ice sheets then drifting south in the Atlantic caused a distinct chill in the ocean. The Neanderthals had survived cold before, but during this ‘event’, the sea surface temperature was the coldest it had been for a quarter of a million years. For Clive, this sudden cold and dry period could explain the demise of the last Neanderthals. ‘So it could be that climate change was the final nail in the coffin for these last Neanderthals.’

For Clive, the final extinction of the Neanderthals and the expansion of modern humans are separate events. Though he accepts that contact may have happened elsewhere, the evidence on Gibraltar shows that those late-surviving Neanderthals disappeared thousands of years before modern humans arrived on their patch.³

‘But I think it’s a long, drawn-out process [across Europe]; there’s not one event that causes their extinction.’

In contrast to the traditional view of Neanderthals as cold-adapted people of Ice Age Europe, Clive Finlayson viewed them as warm-loving humans, who survived late in Gibraltar because Mediterranean conditions lasted longer there. Even in winter, days were long, and the range of environments around the coast of Gibraltar allowed people with a diversified subsistence to survive in rough times – at least until the Heinrich 2 Event. In contrast, Clive saw the modern humans as the more cold-adapted bunch of people. He may be right. Although Neanderthals are stocky and short-limbed – suggesting they were biologically better adapted to cold climates than rangy, long-limbed modern humans, they may not have had the cultural adaptations to beat the chill of the Ice Age. And their physical adaptations may even have held them back. Contrastingly, if modern humans were physiologically more vulnerable to cold, this may have spurred them on to develop advanced clothing – which later meant that they had the wherewithal to survive through the LGM.⁷

Certainly, at around 25,000 years ago, we see a new culture emerging in Europe, apparently brought in by a second wave of modern humans from the north-east: people who appeared to have been quite comfortable in temperatures that are more reminiscent of Siberia today.

A Cultural Revolution: Dolní Vìstonice, Czech Republic

The period between 30,000 and 20,000 years ago was one of global climatic instability, leading up to the LGM.¹ During this time, a new culture and technology spread across Europe – the Gravettian, named after the site of La Gravette in the Dordogne, where the characteristic stone points of this technology were first recognised.

This culture appears to have started in north-eastern Europe about 33,000 years ago, with sites like Kostenki, on the Don River. It comes along as the world was cooling down, in the run-up to the peak of the last Ice Age, and many archaeologists see the Gravettian as an adaptation to colder, periglacial environments. Perhaps not surprisingly, then, although this culture is a European phenomenon, there are similarities with the middle Upper Palaeolithic culture of Siberia. The Gravettian is the ‘technology of the steppes’: these people were reindeer and mammoth hunters, who thrived in cold climates, and spread into Europe as the temperature dropped off.

Innovations in the Gravettian include better shelters – such as the 25,000-year-old semi-subterranean dwelling at Gagarino on the Don. The hearth in this Ice Age house contained burnt bone: it appears that the Gravettians were turning to alternative sources of fuel. Stone lamps appear in the archaeological record, and, at Kostenki, lamps seem to have been made from the heads of mammoth thigh bones. Just as in Siberia at this time, eyed needles indicate that people were capable of making clothes. The Gravettians also seem to have invented cold storage – digging pits (using mammoth-tusk mattocks) which may have been used to store meat or bones for fuel. The hunting technology of the Gravettians included a range of innovations: bevelled stone points, ivory boomerangs and even woven nets – perhaps for hunting small game. Compared with the Aurignacian, stone blades were narrower and lighter, often retouched into very sharp points. Some assemblages include tanged or shouldered points.² There also appear to have been changes in society. Massive, complex occupation sites have been found that suggest people were ‘getting together’ on a large scale. Whether these are gatherings for communal hunts, feasts or other social occasions is unclear, but this certainly implies that social networks were enlarging and society was getting more complex for these Ice Age hunter-gatherers. 3 Perhaps the most intriguing material evidence passed down to us from the Gravettian, though, are objects that have no obvious function: the so-called ‘Venus figurines’.

I headed to the Czech Republic, taking the train to Brno, then heading south to the small town of Dolní Vìstonice (pronounced ‘dolny vyestonitseh’), close to where one of these mysterious female figures was found. There I met up with Jiri (pronounced ‘Yurjy’) Svoboda at the museum in Dolní Vìstonice. Upstairs in the museum, he brought out shallow boxes full of the finds from the archaeological site. There were numerous animals carved from mammoth ivory, including a tiny, beautifully observed head of a lion. There were also strange bone spatulas, about the length and shape of shoehorns, but flat.

‘What on earth’, I asked Jiri, ‘are these?’

‘If you go to ethnography, you can find a whole variety of uses for pieces like that. You might use it for cutting snow, for example. Eskimos have tools like that. But I have also seen a similar tool in Tierra del Fuego for taking bark from trees. And the Maoris of New Zealand have similar pieces as prestigious, symbolic weapons. Unfortunately it’s very difficult to find any use-wear on the edges, so we can only guess.’

And then Jiri brought out the Dolní Vìstonice ‘Venus’. She was a strange little figure, just over 10cm tall, made from fired clay. She was very stylised, with an odd, almost neckless head; the face was a mere suggestion, with two slanting grooves for ‘eyes’. She had globular, pendulous breasts, and wide hips. Her legs were separated by a groove, and another groove ran around her hips, as though indicating some kind of girdle. On her back, pairs of diagonal grooves marked out her lower ribs. Compared with the stick-thin and often clothed female figurines I had seen in the Hermitage, from the Siberian site of Mal’ta, this ‘Venus’ was more rounded and buxom, and splendidly naked.

I asked Jiri what he thought she represented. He was cautious: the real meaning of these mysterious prehistoric objects is lost, and we can really only guess what they symbolised. Was she a deity? Did she represent some kind of archetypal female, or did she perhaps combine male and female sexuality in one? Jiri covered up her upper half, and the legs of the figurine, with that deep groove between them, could certainly be taken to represent a vulva. Then he covered up the legs, and her head and breasts were transformed into male genitalia.

‘So perhaps she is a combination of the two symbols: male and female.’

‘Do you think she could be a deity, a goddess?’ I asked.

Jiri laughed. ‘Well, maybe a deity. But it depends on how you define that. I think there is some kind of personification, some symbolism here, that is depicted in the shape of a female body. There is certainly some meaning, but it’s very difficult to know exactly what.’

There were other objects in the collection from Dolní Vìstonice that seemed to represent sexuality in some way, including a small ivory stick with a pair of protruding bulges – which could be seen as breasts or testicles. Jiri certainly saw some connection between the archaeological signs of widening social networks and the development of symbolism represented by objects like the ‘Venus’.¹

Whatever the Dolní Vìstonice ‘Venus’ meant to the society who made her, she was special in another way – because she was made of clay. She is among the first ceramic objects in the world – and one of some 10,000 pieces from Dolní Vìstonice and the nearby site of Pavlov. At 26,000 years old, these ceramic objects pre-date any evidence of utilitarian pottery, i.e. ceramic containers, by some 14,000 years.⁴ Many of the fired clay pieces were just irregular pellets, but among them were works of art – more than seventy, nearly complete, clay animals, and the Venus figurine. But there were also thousands of fragments of figurines, many of which had been found in what seemed to be purpose-built kilns up the hill from the occupation area. Analysis of the kilns suggested that they produced temperatures of up to 700 degrees C. The preponderance of ceramic fragments has led some archaeologists to formulate a rather bizarre theory: that the figurine-makers were pyromaniacs, deliberately exploding their creations in kilns – and the smashed figurines were the relics of a strange prehistoric form of performance art. Pottery specialists argued that the pattern of breaks in the clay fragments was commensurate with heat-fracturing, but I remain quite sceptical about the deliberate explosion theory. We were looking at the earliest ceramics in the world – presumably there was a fair amount of experimenting going on – so was it really that surprising that many of the pottery creations had exploded? And it seems reasonable to imagine that successfully fired figurines would have been removed from the kilns, while leaving behind fragments of exploded pieces. Not only that, but the exploded fragments could feasibly have been incorporated into the structure of the kilns; this is something that has been seen, admittedly thousands and thousands of years later, in clay pipe kilns, where broken pipes are included in the kiln.⁵

Dolní Vìstonice is also famous for a strange burial, where three individuals were placed in the ground at the same time. There were two male skeletons on either side of a skeleton whose sex was difficult to determine, but which was definitely pathological. The skeletons lay in unusual positions – the individual on the left lay with his arms stretched down and out towards the person in the middle, his hands lying over the pelvis of his grave mate, while the male skeleton on the right was buried face-down. Red ochre covered the heads of all three, as well as the pubic area of the middle skeleton. Wolf and fox canines and ivory beads were found around the heads of the three skeletons.⁶

It seems that these three individuals were indeed buried together, at the same time, and this in itself is very unusual. It is possible that they were related: certainly, they shared unusual anatomical traits, including absence of the right frontal sinus (the space in the skull bone above the eyes), and impacted wisdom teeth.⁷ But why did they end up in the same grave? They are all quite young: one was a teenager, the other two were in their early twenties. Some archaeologists have suggested their youth indicates a particular adverse circumstance befalling the Gravettian population of Dolní Vìstonice, but, from the other human remains at the site, death in early adulthood does not seem to have been particularly unusual.⁸

The deformed leg bones and spine of the middle skeleton have been variously ascribed to rickets, paralysis or congenital anomalies, but it is difficult to be certain about what caused the bony abnormalities in this individual. Although only in their early twenties, he (or she) had already developed osteoarthritis in the right shoulder. Despite the fact that this individual was young when he or she died, there is no indication that the deformities caused this person’s death. What is certain, though, is that this individual would have had a very obvious pathology; some archaeologists have argued that this could have been part of the reason that he (or she) was accorded respect, and selected for what seems to have been a special burial.⁶

There are certainly unique things about the Dolní Vìstonice burial – but there also aspects of it, in particular the use of ochre and the ivory ornaments, that indicate it was part of a culture that stretched across Europe – and I mean right across Europe – in the Gravettian. Around the same time as the Dolní Vìstonice burial – about 27,000 years ago – a man was buried at Paviland Cave on the Gower in South Wales, with ochre and ivory rods in his grave. And at Sunghir, some 200km north-east of Moscow, about 24,000 years ago, a man and two children were buried along with ochre, fox pendants and thousands of ivory beads, apparently sewn on to clothing.⁹

Jiri Svoboda and I left the museum and drove up the road to the archaeological site of Dolní Vìstonice – which was now, like most of undulating lower slopes of the Pavlov Hills, covered in grapevines. We climbed the hill to the top of the vineyard, where Jiri pointed out the locations of the two main sites, on slightly raised ridges perpendicular to the slope.

‘To our left is the first site that was excavated in this area, in the twenties,’ Jiri explained. ‘There was a priest going to Dolní Vìstonice church, from Pavlov, and he noticed, in the cut of the road, bones and charcoal coming out. When it was excavated, that is where the Venus and the other clay figurines were found.’

The second site had been discovered during commercial quarrying in the mid-eighties.¹⁰

Jiri pointed along the ridges to our right. ‘The triple burial was on the next one – at the site of Dolní Vìstonice 2. It looks like people may have been burying their dead in the settlement, inside a hut. And probably other people didn’t go in any more, and the hut collapsed, and that would be the burial.’

Jiri explained that Dolní Vìstonice was just one of a series of settlements along this escarpment, and he also placed it in its wider geographical – and chronological – context. Where we were standing, on that hillside in Moravia, formed part of a corridor that also led through southern Poland and lower Austria, a low-lying passage between the Carpathian Mountains in the east and the Bohemian Massif in the west. It allowed fauna – including humans – to move from south-west to north-east on the European plain.

At the end of the Pleistocene, between 30,000 and 20,000 years ago, the Moravian landscape would at times have been partly wooded, mainly with conifers, but also with oak, beech and yew. These species make it sound as if the climate was rather more pleasant than it actually was; even when there were trees around, snails indicate very cold temperatures, more like subarctic tundra. The climate was fluctuating, and there were even colder, drier periods when the landscape would have been transformed into a treeless steppe.¹

‘It’s difficult to find a present-day analogy,’ said Jiri. ‘Siberia has a zonality of its own today – it’s different in the south and the north – but it’s possible to imagine what it would have been like here. Mean annual temperatures were very low. But although the winters were much colder than today, some of the summers could have been quite hot.’

Whereas Aurignacian sites often occupied higher ground, the Gravettian sites of Austria, Moravia and southern Poland are clustered along the mid-slopes of river valleys. The hilltops would have been too cold. The large mammals hunted by the Gravettians would have passed through the valleys, and so the camps were well positioned to intercept the herds.¹

‘We can imagine a forested landscape below, and the slopes covered by steppe, but again with some conifers. The sites control the valley, and the game was in that valley. This is the place where we can imagine mammoth herds.’

These Gravettian sites seemed to represent more than just temporary hunting camps: they appear to have been occupied through the year. The intensity of the occupation layers, the richness of artefacts – including things which were both delicate and time-consuming to make, the stability of the house structures – all point to a less nomadic and more sedentary existence.

‘These big sites were almost long-term settlements,’ explained Jiri, as we stood looking down over the vineyard and the gently rolling ridges. ‘But at the same time people were quite mobile. So probably they combined the two: some people stayed at the campsite – here – and others went off to find raw materials and to hunt.’

It reminded me of the Evenki, with their villages and satellite hunting camps.

‘It always depends on local conditions,’ said Jiri. ‘Because normally, of course, hunter-gatherers are mobile people, but there were time periods and specific environments and strategies that enabled more sedentary ways of life.’

The new culture that swept across Europe seems indeed to have been a movement of people and genes – and not just ideas (see map on page 207). Analysis of European mtDNA has revealed two lineages: the haplogroup H (the most common in Europe) and pre-V, which appear to have originated in the east, around the Caucasus Mountains, between the Black and the Caspian seas, and spread across Europe between 30,000 and 20,000 years ago.^11,12,13

But just as this second wave of Europeans spread from east to west, Europe was becoming colder. As the LGM approached and the ice sheets descended, northern Europe – as well as northern Siberia – was all but abandoned. Even the cold-adapted, fur-clad, reindeer-hunting Gravettians couldn’t survive in those truly Arctic conditions. Archaeology and genetics – European mtDNA and Y chromosome lineages – record the contraction of the population into refugia in the south-west corner of Europe.

Sheltering from the Cold: Abri Castanet, France

So it was that I made my way to south-west Europe – to the Périgord region, practically synonymous with the Dordogne département, whose caves and rockshelters contain an incredible record of Ice Age life.

Heading north from Toulouse, I drove through up through wooded gorges so typical of the Dordogne, where the large rivers running west from the Massif Central to the Atlantic have etched the limestone bedrock. I eventually hit the Vézère Valley and followed it west, reaching the town of Les Eyzies, famous for its many rockshelters. The valley was wide and plunging, framed by limestone cliffs which were incised with a deep, horizontal groove. These grooves in the cliffs – high enough to stand up in – were used as rockshelters by modern humans, or, as these ancestors are known in France, in honour of the first fossil found, Cro-Magnon man.

I continued along the Vézère Valley, through the small hamlet of Le Moustier – famous as the place where Mousterian tools were first discovered, though in my journey I had now left the Neanderthals behind. Turning off the main river valley and following a road leading up one of its tributaries, through the village of Sergeac, and into a narrow wooded valley, the Vallon de Castel-Merle, I eventually reached the site of Abri (rockshelter) Castanet. As I pulled up, American archaeologist Randall White emerged to greet me.

‘When does the occupation here date to?’ I asked him.

‘We have good radiocarbon dates of 33,000 years ago,’ Randall replied.

As an early Aurignacian site, Abri Castanet represents evidence of that first wave of modern human colonisers into western France. At that time, during the Würm interstadial of 40,000 to 30,000 years ago, the climate would certainly have been cold, though not fully glacial.¹ During early Aurignacian times, the valleys of the Vézère and its tributaries would have been covered in grassy steppe, with woodland on south-facing slopes and in sheltered valleys.

‘We know the people were living here at Castanet in mid-winter,’ said Randall, ‘when it was probably about 35 degrees below zero outside.’

That sounded cold enough – almost as cold as it had been in Olenek – but Randall thought that people would have been able to keep warm in the deep rockshelters.

‘There are holes in the rock where we think they were running lines to drop animal skins down to close off the interior space,’ explained Randall. ‘And then are fireplaces inside. I think they’re making their world a fairly comfortable place.’

A team of American students were busy trowelling away in the rockshelter. They had come down into the Ice Age floor surface, where, among natural stones pressed into the ground, there were stone tools and flakes lying around. There was a dense black layer in one part of the rockshelter – the remains of an ancient hearth. As sediment was excavated, it was bagged up and carefully wet-sieved. The sievings were retained and dried, then Randall and his team painstakingly picked through the material, looking for minute fragments of flint and bone. There were clues here as to how the Aurignacians had survived the long, hard winters.

‘The bottoms of our sieves are absolutely full of burned animal bone. A large percentage of the animal bone they’re bringing in from hunting is being consumed as fuel,’ explained Randall. But bone fires would not have been straightforward. ‘Bone is a terribly difficult thing to burn. There’s relatively little wood charcoal here, but they seem to be adding wood to bone fires to keep the temperature above a certain threshold. It must have been a constant focus of their attention, keeping these fires going. They’re collecting dung for fuel as well. We take fire and heat for granted, but I don’t think they did at all.

‘This was a very diverse environment,’ Randall told me. ‘Reindeer dominate at Castanet, but we have the remains of nine large herbivore species as well as a lot of birds and some fish. This was a pretty good environment for hunters and gatherers, even though it was cold.’

The fauna roaming this landscape would have included reindeer, horse, bison, ibex, as well as animals more suited to forest environments, like boar, roe and red deer.²

Excavations at Abri Castanet had also turned up hundreds of stone beads. Most of the beads from Castanet were quite tiny, the majority less than half a centimetre across. They were shaped like tiny baskets, and had been carved out of soapstone. Without wet-sieving of the sediments from the rockshelter, to wash out the dirt from the holes in the beads, they would just have looked like tiny stones and could easily have been disregarded.

‘It’s interesting to imagine people here, in winter, making all these thousands of beads. Probably done around these fireplaces, like the way we do embroidery or knitting, like a craft. It occupied time on those long nights,’ mused Randall.

Like Nick Conard, Randall White was interested in what we could learn about Ice Age society from its art and ornaments – in fact, particularly from personal ornaments, which he believed were much more than just trinkets: to him they were important representations of belief, values and social identity.³ While many facets of personal adornment – clothing, body painting and organic ornaments – don’t usually survive in the archaeological record, those little stone beads had stood the test of time at Abri Castanet.

Randall and his team had discovered beads in various stages of manufacture, so it had been possible to work out how the beads had been made: starting with a rod of stone, which was then scratched round and round until a bead blank could be snapped off. The blank would be thinned and flattened at one end, then pierced by gouging on both sides, and finally trimmed down into the classic basket shape. Although made of soapstone, the method – of taking a baton and dividing it into blanks, then gouging out perforations – was similar to that in German Aurignacian sites, like Geissenklösterle, where ivory beads had been made.^4,3 The beads had also been polished to a high lustre. Thousands of years before metals were discovered, these people were clearly after the same qualities we enjoy in jewellery today.

There are no burials from the Aurignacian (did they bury their dead at all, or perhaps leave them out – like the Siberian sky burials?) so it is difficult for archaeologists to determine how these beads were used. However, experimental work with beads combined with electron microscopy suggests that the Aurignacian basket beads were sewn on to something – presumably clothing.⁴ But why was Randall so fascinated by Stone Age beads?

‘Twenty years ago everybody laughed at me when I started working on beads. But it’s about what beads say about the societies that these people were making for themselves. The moment you can begin to construct identities by ornamenting yourself differently, by clothing yourself differently, you can provide a more complicated and effective organisation within a group, but you can also create identities across landscapes,’ he explained. ‘It may well be that the people in the Basque country felt themselves to be part of the same cultural entity as the people here in France – it’s a very large area. Not many people would say that Neanderthals had those sorts of societies. I think the ability to organise large numbers of people across large territories would have been an enormous advantage, and I personally think that’s part of the reason why Neanderthals wilted away.’

As the Last Glacial Maximum approached, northern Europe became virtually uninhabited, with ice sheets and permafrost blighting the ground. But in south-west Europe, modern humans clung on. The seasons there were moderated – as they are today – by proximity to the Atlantic: summers were cooler and, more significantly, winters were warmer than in central Europe. But although Iberia and southern France were south of the permafrost zone, the ground was still often frozen. In the Vézère Valley, the frozen uplands would have become uninhabitable, but the protected valleys still supported the hunters of the steppe. The Vallon de Castel-Merle and other valleys were not abandoned. It seemed remarkable.

But despite these harsh climatic conditions, the steppe-tundra grasslands of south-west Europe were fairly teeming with game. Mellars describes south-western France as being almost like a ‘last-glacial Serengeti game reserve’.⁵ Reindeer, horse and bison – all migratory, herd animals – remained common in southern France, along with ibex, chamois, red and roe deer, saiga antelope, and the occasional mammoth and woolly rhino.⁶ But perhaps this makes it sound too idyllic … ‘At the Last Glacial Maximum, the reindeer here reduce in size. Even for reindeer this was an incredibly cold environment: even the animals were stressed,’ explained Randall. ‘I know that when I was growing up in Canada, we had some winters in which we had five or six weeks when the temperature never got above zero. It starts to act on your head. I can’t imagine what it must have been like spending three months in a rockshelter in those kinds of conditions.’

Those hunters were under considerable stress, but, necessity being the mother of invention, they changed their subsistence patterns and broadened their food base: they were still hunting large animals like horse, reindeer and red deer, but there was an increasing reliance on smaller mammals, fish and birds. This intensified subsistence was accompanied by a change in technology: the hunters started to make finely chipped points with concave bases, characterising a whole new industry, called the Solutrean.^1,7

As the grip of the Ice Age tightened, we might perhaps expect to see arts and crafts gradually diminishing, as life became harder. More Solutrean points to make, less time for art. But, rather interestingly, we see exactly the opposite. In the Vézère Valley the rockshelters not only continued to be occupied into the LGM, but the hardy Ice Age people of south-west France made their way deep into the caves that riddle the limestone hills – to paint.

Visiting the Painted Caves: Lascaux, Pech Merle and Cougnac, France

Ornaments, portable art and cave art are classic features of the Upper Palaeolithic in Europe, but they did not suddenly appear everywhere; rather, they popped up at different times in different places. As I’d seen, early examples of portable art appeared in Germany more than 30,000 years ago, as part of the Swabian Aurignacian. The ceramic models of animals and people from Moravia were much later, dating to 26,000 years ago. Pendants and beads – like those at Abri Castanet – appeared in the early Aurignacian, and even in the Chatelperronian, in France – but were not found in other parts of Europe until much later.

Cave art is concentrated in western Europe, in south-west France and northern Spain. The limestone formations in this area certainly provided the perfect canvas, but there are plenty of limestone caves in other parts of Europe and the world. To find out why cave art happened in south-west Europe in particular, we need to look at the environmental and social context of the paintings. To start with, we need some dates.

There are some very early dates for cave art in France and Spain, going back to perhaps 30,000 years ago, although some of these dates need to be treated with caution. Although many of these painted caves have been known since the nineteenth and early twentieth centuries, it is only very recently that archaeologists have been able to obtain meaningful dates and to fit the rock art into the wider picture, alongside other archaeological evidence. In France, archaeologist and cave art expert Michel Lorblanchet welcomed the opportunity to place the paintings in time, and wrote of the ‘post-stylistic era’, where cave art could be directly dated, rather than relying on style alone to establish chronologies. Where paintings included charcoal, radiocarbon dating could be used to provide a precise timing. Unfortunately, the majority of paintings do not contain charcoal or other organic remains, so they are not amenable to direct dating, in which case indirect dates from excavations within caves, as well as the style of the paintings themselves, remain important clues as to their date of inception.¹

Sometimes it is very hard to get a decent date, even when charcoal exists in a painting. Organic residues from micro-organisms and carbonate from the cave wall itself can affect the date. Radiocarbon dates of black dots from the painted cave of Candamo in northern Spain range widely, and have been reported as being as old as 33,000 years ago – or as young as 15,000. It is difficult to decide which of these is the true age: the dots may be 33,000 years old, and may have been painted over again 15,000 years ago. Or perhaps the younger date is true – and the older date was obtained as a result of contamination with more ancient carbon.¹ Similarly, dating of a black horse from Chauvet Cave in France has produced estimations of about 21,000 years ago (Magdalenian), and around 30,000 years ago (Aurignacian); the style of the paintings in Chauvet suggest that the younger date is more likely. Dating specialists certainly hope that advances in radiocarbon dating, and the use of different labs to test the reproducibility of results, will help to clear up such discrepancies in the future. But for now, the potential Aurignacian dates for Candamo and Chauvet have to be treated with caution.¹

Experts tentatively agree, however, that most cave art, including paintings and engravings, seems to be part of the Solutrean and subsequent Magdalenian cultures of the late Upper Palaeolithic – in other words, it was created around the time of the LGM.²

I wanted to see some of this cave art for myself and so I made my way to Lascaux, near the town of Montignac in the Vézère Valley – the most famous of the Franch painted caves. Unfortunately – but entirely understandably – I was not to see the original cave paintings: the cave has been closed up while conservators attempt to eradicate the mould that has been threatening to destroy the precious paintings. Instead, I visited Lascaux II, the reproduction of the cave that is open to visitors. I had almost visited Lascaux some years before, but then decided against when I found that only the replica cave was open to viewing. This time, however, I had to admit that Lascaux II was worth a visit. I walked down a passageway lined with photographs of the original excavations, and into the ‘cave’. It was great – I was completely taken in. The cool air, the shapes and texture of the walls, and the paintings themselves seemed quite authentic.

But this was a reproduction – by a single, modern artist. The colours were true to the original – manganese black, ochre yellows and reds, a similar palette to the rock paintings I had seen in Australia. Lascaux II is a re-creation of the splendid ‘Hall of the Bulls’ and the passageway known as the ‘Axial Gallery’.

The original Lascaux was discovered by four teenagers exploring the hills above Montignac in 1940. A pine tree had fallen, and where its roots had torn up the earth the boys found a hole in the ground. The sinkhole led straight down to what would become known as the Hall of the Bulls. The young discoverers went straight through this chamber – presumably not looking up, or they would have seen that the walls curving in above them were emblazoned with huge bulls – and on into the Axial Gallery, where they first noticed the cave paintings. I stood in the replica Hall of the Bulls, gazing up at the beasts – the so-called ‘Unicorn’ (strange, as he patently has two horns), above me and on my left, followed by great black-outlined bulls, facing each other, with smaller, antlered deer filling the space between them. On the right-hand part of the ceiling were more bulls in black, and red ochre. Making my way down into the narrower, keyhole-shaped Axial Gallery, I could see a procession of animals on the ceiling above me: a beautiful black deer or reindeer with branching antlers, more bulls, and pot-bellied horses.

With my appetite whetted, I went off in search of an original painted cave: Pech Merle, in the Lot département. A flight of stone stairs led down to a somewhat incongruous white-painted door, through which I passed to emerge into a limestone cave deep within the hillside. I walked through magnificent chambers with huge flowstone creations, enormous stalagmites and stalactites, some of which had met between ceiling and floor to form massive pillars. The cave opened into a great chamber, high and wide and elaborately adorned with speleothem creations. It was like walking into a gothic cathedral. What on earth would it have been like for Ice Age hunter-gatherers? For someone who had never been in a church, let alone a cathedral? If we still wonder at the natural beauty of these caves today, just imagine what it would have been like for our ancestors. It must have seemed magical, otherworldly, and sacred.

I was so distracted by the natural splendour of the cave that I almost missed the cave art. But there, on one rare, smooth part of the cave wall to my left, were two beautiful horses outlined in black, facing away from each other, their hindquarters partly superimposed. They were covered in black spots which also flowed on to the background around them, as though they were somehow camouflaged. There were red ochre spots, too, on the belly of the horse on the left, and on the flanks of the other. I noticed that the flat wall of rock had a strange contour where it ended on the left – almost like a horse’s head. It was as though the artist had taken this suggestion from the natural shape of the rocky canvas and allowed it to direct his hand as he (or she) created these wonderful beasts.

The horses were stylised rather than naturalistic representations. They had great curving necks and small heads, rounded bodies and slender legs. Were they artistic representations of real horses or mythical beasts?

I imagined the artist painting them, in the darkness of the cave, with a tallow lamp flickering and lighting up small areas of the wall as they applied the black and red pigments. There were six negative handprints, stencilled on to the wall around the horses, some left, some right hands, but all matching. Were these the signature of the original artist or additions by a later cave painter?

Further along in the same chamber there was another hand stencil, this time in red ochre. I found these hands incredibly moving – it was amazing to think of that Ice Age artist, so many thousands of years ago, placing a hand on the wall and recording that moment. I felt very privileged to be seeing those images. It was like a message that had been passed down from ancient times to the present. What did it say? I know the real meaning is lost for ever, but for me those hands say ‘we are people, just like you’.

Pech Merle is incredibly rich in cave art; it contains more than seven hundred images, including many black-lined pictures of mammoths, bison and horses.

My next subterranean stop-off was the cave of Cougnac, and, outside the cave, on the wooded hillside, I met Michel Lorblanchet, the man who had spent so many years studying – and re-creating – the ancient art of the French caves. I had many questions for him, but first Monsieur Lorblanchet wanted to show me how the hand stencils were made. He had studied the pigments used and how they might have been delivered on to the wall, and had concluded that the ghostly handprints had been created by spitting colour on to and around a hand laid against the stone.

He demonstrated the technique for me outside Cougnac, on an exposed limestone cliff. First, he donned overalls and an artistic-looking black beret, then fetched a collection of things from the boot of his car: stones, charcoal and a bottle of water. Then he ground up some charcoal, using a pebble to crush the pieces to dust on a large flat stone. He explained that the black pigments in the cave paintings were usually manganese oxide: ‘It’s also black, but it could be dangerous for the experimenter. I spoke to a toxicologist in Paris and this man told me: don’t use manganese oxide, you’ll get poisoned. So, I prefer to use charcoal.’

It became clear why, because the next thing he did was to take a generous pinch of ground charcoal, and transfer it into his mouth. I could hear him crunching it up into an even finer consistency.

‘I grind the pigment between the teeth,’ he said, through blackened, gritted teeth, then he chewed some more, laid his hand against the stone and started spitting the charcoal around it. He spat quickly – ‘pup pup pup pup pup’ – and a fine black spray landed on the wall and on his hand with each spit. He had transformed himself into a human airbrush.

After five minutes there was a slight misting of charcoal around his hand on the wall. He would pause to put more charcoal in his mouth, chew it up, then resume the spitting, an action so fast that I worried he might hyperventilate. But Monsieur Lorblanchet was well practised in this technique. As a piece of experimental and experiential archaeology, he had re-created the entire Pech Merle spotted-horse frieze, in a cave, using his spitting technique – it had taken a whole week to execute. While I suspected there might be a quicker and more efficient way of transferring the paint on to the wall, (through a hollow tube, perhaps?) I admired his dedication. His technique was also based on ethnographic studies of Aboriginal painting in Australia, where handprints also exist. The pigments were similar, too. ‘Red ochre and charcoal, and manganese oxide have been used everywhere in different parts of the world,’ he said. ‘There are not many solutions. Some vegetables can provide pigment, but Palaeolithic people mainly use charcoal, manganese oxide and red ochre. Not vegetable pigment.’

After half an hour, Monsieur Lorblanchet stopped and stood back from the wall. His lips were black with wet, powdered charcoal, and his white beard had assumed a black streak in the centre. He took a swig from the water bottle to clean out his mouth. There on the wall was a modern hand stencil.

‘So I spit the pigment on the wall. It is the same method that was used during the Ice Age because I get a foggy image, a foggy hand stencil. And this sort of form is exactly the same as in Pech Merle.’

‘How did you hit on this method?’

‘Oh, it is difficult. It’s necessary to do it several times to get the right methods, yes. To have a good result it’s necessary to have some experimentation.’

Stencilling wasn’t the only method used by the ancient rock artists. Monsieur Lorblanchet described how lines would be drawn using fingers or chewed sticks to form paintbrushes. But stencilling was an excellent solution to applying paint to a rough or crumbly wall surface: ‘Cave walls are often full concretions, stalagmite and stalactite, so it’s impossible to draw with a finger. And if the wall is soft sandstone, and you paint with a brush or with a finger, you destroy the surface,’ he explained. ‘But with this spitting technique you can paint without touching the wall.

‘Pech Merle is very interesting. In this case, there are six hand stencils around the horses, sometimes right and sometimes left hands. But from the same individual, the same man. And pigment analysis showed that the pigment of the horses is exactly the same as the pigment of the hand stencils. So probably the same man did the horses and the six hand stencils, at the same time.’

‘So you think those stencils are the artist’s signature, a maker’s mark?’ I asked.

‘Yes, I think the main meaning is really a signature. Like in Australia. Australians visiting burial sites, they left their hand stencils during the visit, just to say I went there; I visited my uncle or my grandmother and I left a mark of my visit near the burial.’

‘But the French caves are not burial sites,’ I volunteered.

‘No. It is very exceptional to find a burial in the European caves. But it could also be a way to say, I went to this church and left a mark of my visit.’

I was intrigued to hear Monsieur Lorblanchet talk about the caves in religious terms.

After the hand-stencil demonstration, Monsieur Lorblanchet took me into the cave of Cougnac itself. We walked down steps again, to a door in a wall and into a dank room displaying various stone curiosities, including pieces of medieval masonry and an elaborately carved stone sarcophagus lid. Then we descended more steps into the cave itself. Just as in Pech Merle, I was taken aback by the natural beauty of the cave. Cougnac was smaller in size, and the ceiling was much lower, but it was absolutely crammed with slender stalactites. It felt as if we were entering a temple, and I asked Lorblanchet if he thought that the caves had indeed been sacred places for their Ice Age decorators. He thought that they had: he saw the caves as sanctuaries, special places to which artists kept returning.

‘Yes, it is a natural temple. About 10 per cent of the caves in this area have painting in them, and, usually, it is the largest caves which have been chosen. So these paintings have been made for religious reasons, if you like. These are sacred sites, and they are not painting here just for fun.’

We turned a corner and the walls were covered – wherever there was a flattish space – with line drawings of animals: elk, horses and ibex.

Lorblanchet gestured around him.

‘Animals everywhere … because these people were, of course, hunters and gatherers, so they painted their world around them, and the animal world. And for them, the animals were not only game, but also spirits.’

At the back of the cave there was a small image of a man – lying stretched out, with what looked like spears stuck into him. This was an unusual theme. I felt as though I was looking at an illustration of a myth that has long since lost its meaning. There were also abstract shapes that could be seen as the head and shoulders of humans, or perhaps as vulvas. These were ancient drawings, dating to before the LGM.

‘We know now, after studying this cave for years and years, that Cougnac was used intensively during the Gravettian,’ said Monsieur Lorblanchet, ‘then the cave was forgotten. But then it was rediscovered by Magdalenian people, and the cave again was a sanctuary. There is a gap of 10,000 years between the oldest paintings and the most recent painting in Cougnac.’

Around the earlier paintings there were finger-daubed spots of paint, often double, and around some a mist of applied red ochre. And these had been dated much later, to around 20,000 years ago, the Magdalenian period. Monsieur Lorblanchet mused about the original meaning of these paintings, and the meaning to the later artists. Did they imagine these images had been created by their ancestors or by ancient spirits? From archaeology carried out in the cave, there seemed to have been a large patch of red ochre on the floor at the entrance to the main chamber – Lorblanchet imagined the later, Magdalenian visitors dipping their fingers in the pigment and touching the walls around the ancient symbols.

For Lorblanchet, the next, modern rediscovery of the paintings meant that they were once again being incorporated into a belief system, as we tried to understand what they meant to our ancestors, and gazed at, analysed, replicated and reproduced the ancient images. It is interesting to think about these artistic creations as still conveying information, communicating messages within complex social networks. How many people around the world today have seen those images and gone home with the postcards?

It is also rather wonderful to think of those hunter-gatherers, those ancient Europeans, who, even as the climate chilled around them, carried on making art. During the icy grip of the LGM, it was somehow still important to decorate spear-throwers, carve mammoth-ivory animals – and paint caves. The Solutrean is marked out not only by a change in hunting technology, but by a flowering of art and ornamentation. Although making art doesn’t seem to be immediately relevant to survival in an increasingly hostile environment, many archaeologists believe that the painted caves tell us something very important about Ice Age society: that the proliferation of art indicates an increasing complexity of social networks. So – and perhaps even more tellingly than a new style of stone point – the cave art represents a social and cultural adaptation to survival in extreme environments.²

Perhaps these ‘cave art sanctuaries’ were landmarks in the Ice Age landscape, marking out territories for particular groups, and maybe they were places where people aggregated and which confirmed a feeling of group identity.⁴

‘These people were nomads, of course,’ said Lorblanchet, ‘but they had a territory. And by painting a cave it is a way for them to say, here is our sacred place, here are our gods, our belief, we are here. Like the church today is in the middle of the village, the painted cave was in the centre of the tribal territory.’

Meeting up of local bands – to exchange materials, plan collective hunts or hold ceremonies, perhaps – would have provided opportunities for the exchange of information important to long-term survival. Information could also be passed down the generations. Many archaeologists view the cave art as part of an ‘information system’ – very similar to the place of rock art in Australian Aboriginal culture. ‘Information system’ is little more than a very dry term for what is, essentially, storytelling. Stories could contain useful information – about the landscape, animals or society – extending human experience beyond a single lifetime. Maybe images, like those spotted horses at Pech Merle, were used to illustrate tales. For the hunter-gatherers packed into that corner of Ice Age Europe, on the edge, art and storytelling may have been crucial to survival.^3,4

Lorblanchet also thought that paintings contained an expression of identity that we could still understand.

‘I also believe, by their painting, they show that they were exactly the same as us, you know,’ he said. ‘They are good artists, excellent. They had the feeling, the sense of artistic beauty. And by this act of painting in caves they also expressed themselves as different from the neighbours. The neighbours who were the Neanderthal people.’

We can’t leave our ancestors, struggling to survive the wintry chill of the LGM, without looking at them a little closer. Compared with the brown-skinned ancestors of all the Out-of-Africa lineages, it seems that Europeans were getting paler, with a handful of mutated genes reducing melanin production in skin cells. One gene, imaginatively called SLC24A5, appears to be responsible for around 30 per cent of the skin colour difference between indigenous Europeans and Africans.⁵

Northern and eastern Europeans are incredibly diverse in appearance, in the variety of hair and eye colours in particular. Hair can be black, brown, pale blond, yellow-blond, or red. Eyes may be brown, hazel, blue or green. Some have suggested that this great variability in coloration is random: a product of genetic drift, perhaps, or the result of relaxed reduced selection pressure for dark skin (along with hair and eyes), as populations spread north, leaving other ‘colour genes’ free to vary. But there is an interesting theory that all this diversity is due to sexual selection. What if the harshest years of the LGM meant that many young men died while hunting, leaving women outnumbering men? Polygyny might be one solution, but it would have been difficult to provide food for a harem. So competition for men may have been fierce. In that competition, something that made a woman stand out from the crowd – a strikingly different hair or eye colour, perhaps – could make the difference to whether or not she managed to pass her genes on.⁶ It’s a fascinating theory, but ultimately impossible to test, and doesn’t explain why coloration should have become so diverse in Europe and not elsewhere.

Archaeology and genes tell us how, after the LGM, humans re-expanded across Europe, from their principal Ice Age refugium in Iberia.^7,8,9,10 The changing climate was once again matched by a change in technology. The new European industries – the Magdalenian and Epigravettian – were hugely variable, but there was a general explosion in the use of antler as a material for producing tools, and the harpoon was invented.² Tiny bone points were fitted into antler and wooden spear points.

At 16,000 years ago, populations are re-established north of the Loire, and by 13,000 years ago Britain was reoccupied. The human recolonisation of Europe was part of a general faunal expansion northwards, but some animals were missing from the post-glacial landscape: whether because of climate, hunting or both, there were no more mammoths or woolly rhinos.

Technologies continued to evolve in the late Pleistocene, and by 11,000 years ago the first definite evidence of the bow and arrow appears. The replacement of steppe-tundra with woodland meant that the large herds of horse, bison, saiga, and that ‘Ice Age larder-on-the-hoof’, reindeer, disappeared.² It seems quite counter-intuitive that the warming of Europe actually created a significant challenge to its human (and animal) inhabitants, but the altered environment required the invention of new ways of living off the landscape. The trend towards intensified subsistence that began in the run-up to the LGM continued, with even more hunting and trapping of small animals and birds, fishing, and gathering of molluscs. Ice Age technology was gradually replaced by Mesolithic tools: bows for hunting, reaping knives and axes for felling trees, as the Europeans adapted to newly wooded landscapes, as well as to estuarine and coastal environments. Europe became populated more widely and more densely than ever before, and quite soon an innovation from the Middle East would allow even greater population numbers to be sustained on the landscape.²

New Age Mesopotamia: Göbekli Tepe, Turkey

The end of my European journey would take me back to where I had started, to Turkey, and to the most spectacular archaeological site I have ever seen.

I travelled down the south-east of Turkey, about thirty miles north of the Syrian border, to the ancient town of Sanli Urfa. I was in Mesopotamia, between the Tigris and Euphrates rivers. In Urfa, modern buildings clustered around Roman ruins on the slopes, but I was after much more ancient archaeology.

I travelled for around an hour, west of Urfa, then turned off the main road on to a dusty track, which wound up through a rocky valley and on to a limestone escarpment. Eventually, the track ran out and I found myself at the bottom of a conical hill, with a site hut and an empty tent. As I walked up the hill, German archaeologist Klaus Schmidt came halfway down to meet me.

‘This hill is not made by nature,’ Klaus explained as we walked up it. ‘It’s a tell, a man-made mound created by the ruins of these Stone Age structures. It reaches up about fifteen metres above the natural limestone plateau. I was suspicious when I first saw this site: no force of nature could make such a mound of earth in this location.’

Klaus had discovered this site while surveying the area for potential Palaeolithic sites in 1994. Local farmers had been turning up masses of stone tools in the fields on and around the hill, and occasionally hitting very large stones with their ploughs. Archaeologists had got wind of this before, but assumed that the stones were the remains of a medieval cemetery. But when Klaus investigated, he found finely worked blades, and large, rectangular stones buried in the ground – and so large that they could not be moved or lifted. When he started excavating the site in 1995, he found that these stones were something truly remarkable. They were just the tops of great, T-shaped standing stones. Some were more than two metres high, and, as the archaeologists dug deeper, they found that the stones were arranged in a circle, with two larger standing stones in the centre of the ring. But there wasn’t just one stone circle at Göbekli Tepe: Klaus had excavated four so far, and from geophysical surveys of the hill he supposed that there might be twenty to twenty-five of them, still buried in the rubble of the hill.

Klaus led me to the top of the hill and showed me a stone circle. I was bowled over by it. Archaeologists were busy with a small crane, moving a large fragment of one standing stone from where it had fallen. There was an engraved figure of a person on the narrow outer side of one of the standing stones, at the edge of the circle. There were other stone walls at the top of the hill, which Klaus thought might have enclosed smaller sanctuaries. Rather strangely, there was no sign of habitation, such as hearths, up on the hill: it appeared to have been exclusively a sacred site rather than a settlement.

As we walked down the other side of the hill, I suddenly saw more stone circles, even more impressive than the first one at the top of the hill. The circles were wider and the standing stones taller – and more elaborately decorated. On the sides of the stones there were beautiful, low-relief carvings of foxes, boar, birds, scorpions and spiders. On the inner edge of one stone there was a full-relief, 3D carving of an animal, perhaps a dog or a wolf. It had been carved from the stone in one piece. While I was there, Turkish archaeologists uncovered an animal head sticking out, gargoyle-like, with formidable fangs, from the wall enclosing one of the stone circles.

There seemed to be several phases of building at Göbekli Tepe. The lower, older and more impressive stone circles appeared originally to have been put up as just that: a ring of large standing stones. Then stone walls had been added, creating inner and outer circles. In some places, there were slabs laid on the standing stones as though the ring had originally been covered with a corbelled roof.

The architecture and stone sculpture at Göbekli Tepe is remarkable. But what made it even more arresting was the date. ‘It’s been here, buried in this hill, for 12,000 years,’ said Klaus.

So Göbekli Tepe appeared to be a temple site – built by hunter-gatherers. It challenged paradigms of the origins of the Neolithic. Based on previous evidence, archaeologists have suggested that the emergence of the Neolithic involved a sequence of developments that goes something like this: population pressure led to increased need for food, led to the adoption of agriculture, led to stratified societies and new power structures, led to organised religion. What Göbekli Tepe seemed to show, though, was a complex, hierarchical society – where stonemasons could be tasked with building temples – and organised religion, in the context of a hunter-gatherer society.

It’s difficult to know how to categorise Göbekli Tepe. As a site representing something somewhere between the Upper Palaeolithic and the Neolithic, Klaus thought that perhaps the best label would be ‘Mesolithic’, but then this would have been a very different Mesolithic from the nomadic hunter-gatherer lifestyle, further north in Europe. Having said that, the stone toolkit at Göbekli Tepe was similar to some of the tanged point cultures of central Europe, in the late Palaeolithic and Mesolithic. But other archaeologists, too, talk about a transition straight from the hunter-gathering Palaeolithic to the agricultural Neolithic in the Levant, missing out the Mesolithic. The first step towards this new way of life began around 14,500 years ago, with the appearance of the Natufian culture in the Levant.¹ Hunter-gatherers began to do a new thing: they settled down in villages, in which they stayed all year round. At this point, the development of agriculture seems almost inevitable. The appearance of grinding stones, mortars and pestles at archaeological sites suggests that wild cereals were important in the Natufian diet. There are also dog burials from this time: it seems that man’s best friend had arrived.¹

Klaus suggested that this may have been because gazelle hunters in Turkey required similar technology to the reindeer hunters further north, but also that perhaps there was some kind of connection or communication between the societies of Turkey and those around the Black Sea and the Crimea. However, the complex society and ritual suggested by the stone-circle temples was something else entirely. There was nothing even vaguely like it in Europe – with this scale of monumental architecture – until well into the Neolithic. So, instead, then, Göbekli Tepe is called ‘early Neolithic’, with the implicit understanding that many of the classic features of the Neolithic, like pottery – and in particular, farming – were yet to come.

What sort of rituals were taking place at Göbekli Tepe? The symbolism there seemed to be dominated by animals: snakes were the most common, sometimes appearing singly, or stacked up like waves. Wild boar and foxes are also common motifs, along with leopard-like creatures and stylised aurochs heads. Birds were also depicted: perhaps geese or ducks. The images didn’t seem to relate directly to animals that were being hunted, as boar and snake bones are very rare in the rubbly backfill at Göbekli Tepe. Gazelles were an important food animal, but only one carving of a gazelle has been found at the site so far. This is a bit like the Ice Age caves of France, where one of the most commonly hunted animals, reindeer, is rarely depicted.² Perhaps the various animals represented different ‘clans’ that came to the temples?

Weirdly, the large T-shaped standing stones appeared to have arms, bent at the elbows, with hands clasped at the front. They had no faces, no eyes, nose or mouth, but Klaus thought the stones represented immense, abstract human figures.

‘Who are these beings made of stone?’ he asked, rhetorically. ‘They are the first deities depicted in history.’

He suggested that the animal motifs carved into the sides of these giant ‘figures’ might be guardians or protectors of the megaliths, but in some cases carefully grouped animals suggested that their meaning went further: perhaps they were representing stories or myths.²

‘Maybe these are pre-hieroglyphic messages,’ Klaus suggested.

Where the sex of the animals could be discerned, they always appeared to be male, and a small ‘ithyphallic’ figure with an out-of-proportion semi-erect penis had also been discovered at the site. This was very different from the famous, much later site of Çatal Hüyük in Turkey, where female imagery was common – although inside houses rather than temples. Çatal Hüyük also contained many depictions of vultures, which seemed to be associated with death. There were no pictures of vultures at Göbekli Tepe – but there were many snakes, and objects decorated with snakes have been found among grave goods at other sites of almost the same age.² To Klaus, the predominance of ‘aggressive’ animals, snakes, and absence of female/fertility symbols suggested that Göbekli Tepe might be a burial site – or at least a place dedicated to a cult of the dead.

The monumental architecture and the clustering of the stone circles certainly seemed similar to complexes of burial mounds found in the (much later) megalithic cultures of Neolithic Brittany and the UK. But Klaus hadn’t, as yet, found any evidence of burials at the site. He was just about to excavate part of one of the higher stone circles where he thought that a large, flat stone might just be covering a burial.

Klaus’s team had also found large quantities of stone tools in various stages of production, including nodules, half- and fully-prepared cores, and blades. Evidence of flint-working usually occurs in occupation sites, but here it was in what seemed to be a purely ritual space. Klaus thought that the flint-knapping was somehow part of the rituals that had been carried out there. But it was difficult to draw conclusions about how Göbekli Tepe fitted into the lives of the people who had constructed it, given that no settlements or any kind of camps of the same age have yet been found in the area around the hill.

However, other archaeological sites in the area give us an insight into how life was changing for the hunter-gatherer societies of the Levant. Archaeologists now see the Neolithic as emerging in stages: first, hunter-gatherers became more settled, with more complex social structures; then plant cultivation appeared, and then we see the appearance of large villages and intensive food production.³ So it appears that social change came first – followed by agriculture. Göbekli Tepe fits into the early stage of this transition, as a pre-agricultural, pre-pottery site, demonstrating the existence of a complex society. For Klaus, the social changes may have been the impetus for the development of farming – perhaps to provide feasts to honour the gods. ‘For them it would have been a logical step, to manage Nature and get more food for themselves,’ he said. ‘Religion created the pressure to invent agriculture.

‘It’s very clear we must change our ideas,’ he continued. ‘Hunter-gatherers don’t usually work in the way we understand work.’ But, from the scale of construction at Göbekli Tepe, it was clear that people here had jobs: specific things to do which weren’t immediately about obtaining food, water or shelter, but were nevertheless important to society.

‘They started to work in quarries. They started to have engineers to work out how to transport and erect the stones. There were specialists in stone-working, whose job was to produce sculptures and pillars from stone,’ said Klaus. It appeared that the society that produced Göbekli Tepe could support both a workforce and professional artists.

Klaus also believed that the transition to agriculture might have led to the abandonment of Göbekli Tepe and its gods. His team of archaeologists were digging down through rubble that appeared to have been deliberately piled on top of the stone circles to cover them up. So, social changes may have driven the development of farming, but subsequent changes in society, and the emergence of new religions, had then been disastrous for Göbekli Tepe.

‘The hunter-gatherer societies were on the threshold of inventing a new way of life, of becoming farming communities. And by the ninth millennium bc this process was successful and this new way of life was developing in this region. The old spiritual world of the hunters was without any use, and so this site was abandoned. This world was completely forgotten, completely lost, never repeated.’

I found this story, of the rise and fall of the old hunter-gatherer gods, very compelling. But in fact it is very difficult to rule out farming entirely at Göbekli Tepe, as the earliest farmers would have been planting wild foods. It’s also very difficult to pinpoint that transition from collecting wild-growing plants to intentional planting. And, of course, the first planted crops would have been wild varieties: domesticated varieties would have emerged later as farmers selected particular characteristics. For hunter-gatherers for whom wild varieties of cereals and legumes were already staple foods, it may only have been a short step from gathering them to intentional planting and cultivation.⁴ Klaus thought that feasting, or at least the formation of more settled communities, may have prompted this step, but there may also have been climatic reasons for the adoption of agriculture. Around 14,600 years ago, and again around 11,600 years ago, there was a period of increased temperature and rainfall, each lasting only one or two decades. During each of these warm, wet phases, cereals and legumes would have flourished – providing humans with plenty of easily exploitable resources.³ Between these two wet, warm periods was the cold, dry snap of the Younger Dryas. The effects of this may have been over-emphasised, but for people getting used to living on plentiful cereals and legumes, a period of worsening climate may have encouraged them to start cultivating crops. This climate explanation echoes what I had discovered from reading about the origins of agriculture in China, and I find it very persuasive. Because, at almost exactly the same time, on opposite sides of the globe, people were inventing agriculture. It may not be such a coincidence: both populations were living in environments affected by global climate change.

The date of Göbekli Tepe places it slightly earlier than other archaeological sites where the transition to agriculture is clearly documented. It now seems clear that this area, between the upper reaches of the Tigris and Euphrates, was indeed the place where farming got started in the West.⁵ Early farming communities – still without pottery – became established in Turkey and northern Syria between about 11,600 and 10,500 years ago.⁴ At these very early sites, archaeologists have found burnt remains of wild cereals (like einkorn wheat, rye and barley) and legumes (peas, vetch and lentils). Slightly later, from around 9500 years ago (7500 bc), evidence of domesticates such as emmer wheat and barley appears – sometimes in higher (more recent) layers at the same sites.⁵ Later still, herd animals are domesticated.³ People whose recent ancestors had hunted wild animals in the foothills of the Taurus and Zagros Mountains started corralling, tending and breeding them. Along with plant and animal domestication came a new toolkit, with sickle-knives for cutting and querns for grinding cereals.⁵

Botanical and genetic studies also point to this area as the ‘cradle of agriculture’. Wild varieties of the important Neolithic crops (einkorn and emmer wheat, barley, lentil, pea, bitter vetch, chickpea and flax) are found growing together in this area. The limited genetic variability of domesticated crops also lends support to the idea of a single, core area of plant domestication.⁵

Once agriculture began, it allowed populations to expand further. Food resources became more reliable. People settled down and large villages started to appear. It all sounds wonderful – until you take a closer look at what was happening to the health of individual people. Because, although much of the risk of inadequate food supply may have been shed by growing crops and keeping domesticated animals, the quality and range of the farmers’ diets was not particularly good. Archaeologists have long supposed that the transition to agriculture was positive in all sorts of ways, bringing better health and nutrition, increased longevity, and more leisure time. But the truth is a little harder to bear and somewhat counter-intuitive. From the study of human skeletons from this crucial period of changeover, biological anthropologists have found that the switch from foraging to farming brought with it a general decline in health.

Compared with hunter-gatherers, farmers had more tooth loss and more dental caries, had restricted growth and shorter stature, and reduced life expectancy. Skeletal evidence of trauma becomes more common, indicating an increase in violence and conflict. Neolithic people also suffered more from infectious diseases than previous groups, probably because of the combined effects of a poor diet and more crowded living conditions. Anaemia was also more common.^6,7 Traditionally, archaeologists have argued that the Neolithic brought with it improved health and reduced mortality, and so populations could expand rapidly, but the bones of our ancestors show us that this was not the case at all. The dawn of agriculture and permanent settlement brought about worse health and reduced life expectancy. But in spite of all these disadvantages for individuals, agriculture brought with it an increased birth rate that outstripped the reduced life expectancy – so the populations expanded.^6,8 It is difficult to work out the sizes of past populations, but researchers seem to agree that population growth was extremely slow during the Palaeolithic, and that at 10,000 years the world contained around eight million humans. By ad 1800, the global population was a stonking thousand million people.⁸

Agriculture then spread out of the Levant: into central Anatolia by 8000 to 9000 years ago, then eastwards into the Zagros foothills and Indus Valley, and westwards, along the Danube and along the Mediterranean coast.⁴ By 7500 years ago the first farmers appeared in Hungary, and by around 6000 years ago the culture had spread all the way to northern Spain, where the Neolithic seemed to have been adopted as a ‘package’: evidence of domesticated crops and livestock, pottery and megalithic monuments suddenly appears in the archaeological record.⁹ So, the spread of agriculture into Europe, between 10,000 and 6000 years ago, followed similar routes to the original, Upper Palaeolithic colonisation of Europe.⁴ But was this a movement of people (demic diffusion) or ideas (cultural diffusion) – or both?

When I started considering this question I thought that genetic studies might provide the answer. But in fact, over the years, various studies have turned up somewhat conflicting results. Y chromosome studies have revealed what appears to be a cluster of lineages spreading from the Levant across Europe: in other words, a movement of people.^10,11 Some analyses of mtDNA data have also uncovered a potential Neolithic contribution to the European population,¹² but others have produced very little evidence for demic diffusion. Extraction of ancient DNA from some Neolithic skeletons from central Europe revealed that a quarter of them had a type of mtDNA that is now extremely rare in Europe, suggesting that the genetic contribution to the European gene pool from Neolithic incomers was probably minor, at least within maternal lineages.¹³ The discrepancy between the maternal mtDNA and paternal Y chromosome patterns has been explained by some as indicating a spread of male farmers out from the Near East, intermarrying with indigenous women. Other researchers have suggested that the contradictory results actually warn us against coming up with too simple a model. The spread of farming across Europe would have been a complicated matter, with the degree of demic and cultural diffusion varying in different regions.¹⁰ Broadly speaking, it seems that there was a movement of some people out of the Near East in the Neolithic, but these people mingled with the ancient populations already established in western Europe, rather than replacing them. At the moment, the picture is far from clear, but, as more work is done and more samples collected, particularly from ancient skeletons, the way that the Neolithic made its mark on Europe should become clearer.

The beginning of food production was a revolutionary event in prehistory. It paved the way for large-scale settlement – for civilisation. After hundreds of thousands of years of being nomadic hunter-gatherers, humans began to settle down and farm. It seems like a long time ago, but set against the scale of human prehistory it’s a relatively recent development.

Göbekli Tepe had lain hidden under rocky fields at the top of a hill for 12,000 years, but when the ancient gods had been rediscovered they held their discoverer firmly in their grasp. When Klaus found the place, he was exhilarated, but he also knew at that moment that he had two options: either to walk away right there and then, or to spend the rest of his life there, excavating this remote hill where human society had started to change.

‘I was very excited. But it was clear from the first minute that I had a choice, too: to turn back and not tell anyone about this discovery, and never come back … or to stay and work here for the rest of my life.

‘It is a dream,’ he said, ‘but it is also difficult to have such a site. You are captured by it. You belong to it.’

I left Göbekli Tepe very sure that I would hear more about this amazing place in the future. But, for now, my journey would take me to the lands of my final destination, to the Americas, the last continents to be reached by humans.