The world’s largest sand sea, the Empty Quarter in the Arabian peninsula, was once dotted with freshwater lakes and lush oases – but prehistoric humans appear to have avoided settling around them.

Today, scientists have found out why: the early inhabitants chose to live near useful deposits of rock instead. After gathering data during a 1500-kilometre expedition across the vast desert and using computer modelling to understand the evidence, researchers have concluded that even plentiful fresh water wasn’t enough to attract early humans if they couldn’t also find exposed formations of suitable tool-making stone like flint and quartz.

“People who live by stone tools never stray far from good stone,” says at the University of Algarve in Portugal, who was involved in the analysis. “You just really don’t see Stone Age sites unless you’re within a few kilometres of an outcrop.”

The Sahara is the world’s largest hot desert, with landscapes ranging from rocky highlands to vast gravel plains. But the Rub’ al-Khali – the Arabic name for the Empty Quarter – is the world’s largest continuous expanse of wind-swept dunes, known as a sand sea. It covers 650,000 square kilometres of the southern Arabian peninsula.

Today, the region is overwhelmingly dry, other than for a few permanent lakes in its interior that are fed by groundwater. But over the past century, scientists have uncovered evidence of numerous ancient lakes that formed during wetter periods, when stronger Indian Ocean monsoons helped .

Researchers have long suspected that these greener periods opened routes for early humans spreading out of Africa and across the Arabian peninsula – an idea supported by the . However, all these sites line the desert’s margins, whereas its vast interior is still largely unexplored by researchers.

Subscriber-only newsletter

Sign up to Our Human Story

Sign up to newsletter

In 2023, Rose’s colleague at the University of Vienna and his team spent 35 days crossing the Empty Quarter’s dunes, examining 125 locations for traces of ancient lakes. They focused on lakes that were present between about 122,000 and 67,000 years ago, when early humans were spreading into Arabia.

The team found evidence of ancient lakes at 55 of the locations. This evidence included freshwater snail and clam shells, ostrich eggshell fragments and traces of animals that once burrowed into the muddy lake beds. But despite these signs of a once-thriving environment, they found no stone tools or other evidence that prehistoric humans had settled by any of the 55 lakes.

“If we could spot a snail shell the size of a fingernail, we would have spotted a stone tool,” says Dellmour. “But there were none.”

To test whether the shortage of suitable stone could explain the lack of human settlements, Rose and Dellmour and their colleagues turned to computer modelling. They mapped ancient water sources and the geology of the Arabian peninsula, then compared the results with the distribution of hundreds of archaeological sites found across the region in earlier research.

The analyses revealed that 57 per cent of the peninsula’s terrain within reach of water lacked suitable stone for making tools – and that prehistoric human settlements were concentrated in places offering both resources.

That could be partly due to ancient tool-making being extremely “wasteful”, requiring abundant access to raw material, says Rose. “They weren’t thinking about efficiency,” he says. “So, you’d have a pile of garbage produced for one spear point.”

Neanderthal Europe was as much a wooden age as a stone age

When Neanderthals dominated Europe, it was likely that wood was at least as important to them as stone for working into tools

Groundwater-fed lakes in the Empty Quarter wouldn’t have brought flint, chert, obsidian or quartz into the sand sea like rivers could have, says Rose. And while people could have carried tools into resource-poor areas, this would probably have supported only small groups for short periods.

The findings show that geology was potentially just as important as climate in shaping prehistoric human migrations, says Rose.

The study sheds important light on a long-standing debate among archaeologists about the importance of raw stone in ancient human sites, especially in harsh environments like the desert, says at the University of Tübingen in Germany. “Probably what’s most surprising to me is the lack of archaeological stone tools in otherwise perfectly fine circumstances, and close access to fresh water,” he says. “It really shows that early humans were picky with where they went and that they did not settle for just any spot.”

A huge stash of reddish minerals from a cave in Ethiopia shows how Stone Age people gradually adapted their technologies and practices over a 4500-year period.

“It’s one of the rare sites where we can see a very precise evolution of this cultural feature through thousands of years,” says at the University of Valencia in Spain.

Rosso and her colleagues studied materials from Porc-Epic cave in Ethiopia. The cave first became known to scientists in the 1930s, and was thoroughly excavated in the 1970s. It was used by people in the Middle and Late Stone Age, between about 80,000 and 40,000 years ago, but the bulk of archaeological material dates from a 4500-year-long period about 40,000 years ago.

This material included 4213 pieces of “ochre” – an umbrella term for minerals that are rich in iron and consequently have vivid colours, typically red. Prehistoric people often collected these minerals, but the original excavators of Porc-Epic did not study them. “This is the first time there is a systematic study of ochre use at this site,” says Rosso.

Rosso and her colleagues examined what the various pieces of ochre were made of. This changed over time: ochre from the beginning of the 4500-year period was typically high quality and rich in iron, while ochre from the end of the period was lower quality and had less iron. The later ochre was also coarse-grained, so instead of grinding it to powder the people tended to chip and cut it.

There are several possible explanations for the shift. One is that the people at Porc-Epic may have been using the ochre for different purposes as time went on, and chose different types accordingly.

Subscriber-only newsletter

Sign up to Our Human Story

Sign up to newsletter

The most famous use of ochre is as a pigment for artworks, but Rosso says it was probably sometimes used in utilitarian ways – for making adhesives, or as sunscreen, for example.

However, running counter to the idea that the shift was deliberate is evidence in a 2022 study by at the University of Tübingen in Germany and his colleagues. They reviewed . Dapschauskas says prehistoric people consistently sought out “fine-grained and blood-red materials”, which were the best for pigment as they could be ground to a very fine powder and produced vivid colours. “People really, really preferred those reddish colours,” he says.

So it may be that, as time passed, the people at Porc-Epic simply found it increasingly difficult to source the best-quality ochre. The team examined local geological deposits and found that the available ochres did not match those in the cave: they were often coarser-grained and had less iron. “Probably they had to go further away” to find the best ochre, Rosso says.

Why it became harder to get the high-quality ochre is unclear, says Dapschauskas, but it may be that the social situation changed: for instance, if the people at Porc-Epic relied on trade to secure good-quality ochre, then conflict with neighbouring groups might have led to shortages.

The study adds nuance to our understanding of technological stasis in the Stone Age, says Dapschauskas. “There’s a form of stability,” he says. “The cultural knowledge is transferred from generation to generation to generation.” But at the same time, the people were flexible and changed their practices over time. “They can really trace several thousands of years of behavioural change.”

An analysis of a 39,600-year-old bone containing strange indentations claims it was used as a punch board for making holes in leather, revealing how Homo sapiens in Europe made clothes to help them survive cold climates at that time.

“We do not have much information about clothes because they’re perishable,” says at the University of Bordeaux, France, who led the study. “They are an early technology we’re in the dark about.”

The bone, from the hip of a large mammal such as a horse or bison, was discovered at a site called Terrasses de la Riera dels Canyars near Barcelona, Spain. It has 28 puncture marks on its flat surface, including a linear sequence of 10 holes about 5 millimetres apart from each other, as well as other holes in more random positions.

This pattern was “highly intriguing”, says Doyon, because it didn’t appear to be a decoration or to represent a counting tally – the usual explanations for deliberate patterns of lines or dots on prehistoric objects. Microscopic analysis revealed that the line of 10 indents was made by one tool and the other dots were made at different times by five different tools. “Why do we have different types of arrangements on the same bone?” says Doyon.

The researchers used an approach called experimental archaeology, in which you try out different ancient tools to see how marks were made. “We’re attempting to replicate the gestures that were used by prehistoric people to produce a specific modification on the bone,” says Doyon.

They found that the only way to recreate the type of indents on the Canyars bone was to knock a chisel-like stone tool called a burin through a thick hide, a technique called indirect percussion. The same method is still used by modern-day cobblers and in traditional societies to pierce leather.

Subscriber-only newsletter

Sign up to Our Human Story

Sign up to newsletter

The most likely explanation for the indents is that they were made during the manufacture or repair of leather items, say the researchers. After punching a hole in the animal hide, a thread could be pushed through the material with a pointed tool to make a tight seam, says Doyon.

“It’s a very significant discovery,” says at the University of Sydney, Australia. “We have no direct evidence for clothes in the Pleistocene, so finding any indirect evidence is valuable. The oldest surviving fragments of cloth in the world date from around 10,000 years ago.”

This discovery helps solve a mystery about the emergence of fitted clothing. Homo sapiens reached Europe around 42,000 years ago, yet eyed needles haven’t been found in this region from earlier than around 26,000 years ago and these aren’t strong enough to repeatedly puncture thick leather – raising the question of how these ancient people managed to make garments to fit them.

“The knowledge about making fitting clothing without bone needles is something we didn’t have access to before,” says Doyon.

“The location and date are interesting: southern Europe nearly 40,000 years ago,” says Gilligan. “That’s quite soon after the arrival of Homo sapiens, during some rapid cold swings in the climate. It’s when and where we’d expect our ancestors to need good clothes for protection.”

Doyon and his colleagues argue that this punch board marks a crucial cultural adaptation to climate change that helped modern humans expand to new regions.

The punch board was one of six artefacts found at the Canyars site, they say, and could have been part of a repair kit.

DEEP inside Gargas cave in the Pyrenees mountains of southern France is something that has puzzled every visitor who has made the journey into its dark inner chambers. Among prehistoric paintings and engravings of horses, bison and mammoths are hundreds of stencils made tens of thousands of years ago by people spitting red and black paint over their outstretched hands. Such motifs are found at ancient sites around the world, from Australia to the Americas and from Indonesia to Europe. For years, archaeologists have wondered at their meaning. But those in Gargas are especially mysterious because around half of the hands appear to be injured.

“It’s very obvious that some of the fingers are missing,” says at the National Centre for Scientific Research (CNRS) in Bayonne, France. So-called mutilated hands can be seen at many other prehistoric rock art sites, but Gargas cave is the most striking example of this phenomenon.

It has been suggested that these missing fingers are the result of accidents, frostbite or ritual mutilation. Another possibility is that their creators deliberately folded away their fingers to produce specific patterns. Irurtzun and , also at CNRS, have now found a way to test this idea. What they have discovered convinces them that Gargas’s hand stencils reflect a Stone Age sign language. If so, these patterns add to a growing body of evidence suggesting that Palaeolithic cave paintings may contain a variety of hidden codes. The Gargas stencils could even represent the oldest writing system we know of – by a very long way.

Prehistoric hand stencils have puzzled researchers for more than a century. Only a few – from Europe and Indonesia – have been reliably dated, and they turn out to be among the earliest known artistic motifs. Most are around. One, found in Spain, is more than 66,000 years old, leading researchers to conclude a Neanderthal made it (see “Whose hands?”). “We know that hand stencils were some of the first markings of a visual culture to appear. They go back a lot longer than figurative art,” says at Durham University, UK. “Conceptually, they’re fascinating. It’s an odd thing to do, to create not a positive print, but a negative impression of it.” In his view, they were the inspiration for figurative art, as humans started recognising their ability to create lifelike forms in paint.

While these stencils might look like mere doodles to the untrained eye, they are often found deep in caves in hard-to-reach places, suggesting that they had some special significance. “They’re not just someone accidentally slapping their hand on a wall,” says Pettitt. Irurtzun shares this view. “Going deep into the cave, with the painting material and carrying a torch or lamp… it has to be something really profound for them. The question is, what type of meaning did they have?” he says. at the University of Coimbra in Portugal suggests that, among other things, they could be indicators of danger, orientation signs, group identity symbols or markers of hidden goods in the cave. “It is the big question everyone would like to answer,” he says.

Missing digits

Hand stencils with missing fingers are even more intriguing. They are most common in , where 114 of the 231 hand images are missing at least one finger segment. They also feature prominently in another French cave, Cosquer in Marseille, where 28 of the 49 hand stencils are missing digits.

One idea is that Stone Age people deliberately removed their digits. That might sound brutal, but ritual finger amputation is actually a relatively common practice. A 2018 study found it occurring in 121 recent societies – although it is generally limited to the pinkie finger. Amputation might explain some Stone Age hand stencils. “In Palaeolithic rock art, the most common hidden finger is only the pinkie,” says Collado Giraldo. But the stencils at Gargas and Cosquer exhibit a variety of missing digits. Moreover, the most common pattern observed in Gargas is an extended thumb with all the other digits displayed as stumps – an extreme mutilation that would have been catastrophic for the recipient. Besides, there are no missing fingers on any of the positive handprints in prehistoric European cave art – made by daubing the hand with paint and pressing it against a cave wall. These observations seem to rule out the mutilation idea and also the possibilities that fingers were lost to frostbite and accidents – at least at Gargas and Cosquer.

Instead, many researchers think that prehistoric artists deliberately created these patterns. “The missing fingers are only hidden fingers under the palm of the hand,” says Collado Giraldo. If so, this has intriguing implications. “It’s almost certainly some kind of communication system,” says Pettitt.

One possibility is that this was a way of counting or representing numbers. This explanation seems , where there are five different configurations of missing digits, resembling the pattern that many people use if counting up to five on one hand. But at Gargas, there are more finger patterns in more complex configurations (see “Signs of the times”, above). Another suggestion, proposed by archaeologist André Leroi-Gourhan in the 1960s, is that . He pointed out, for example, that modern-day hunter-gatherers in the Kalahari desert signify a warthog by folding the middle three digits of the hand inwards.

Taking this idea a step further, Irurtzun and Etxepare, who are both linguists, wondered whether Stone Age hand stencils might represent a prehistoric sign language. After all, various lines of research suggest that language originated with hand signs as well as vocalisations. Indeed, many societies continue to use a wide range of symbolic hand gestures during hunting, storytelling and rituals alongside – and sometimes in place of – their spoken language. These “alternate” sign languages can function as a lingua franca between groups that don’t share the same spoken language.

There is even evidence of people representing their sign language in cave symbols. More than a century ago, anthropologist Walter Roth documented an alternate sign language made by Queensland First Nations communities in Australia, which has parallels with depictions of hands in the rock art of the region. For example, a fist with just the little finger outstretched was the sign for a small caterpillar or grub.

To test whether the motifs in Gargas cave might represent a sign language, Irurtzun and Etxepare turned to a system used to analyse the ease with which the gestures employed in alternate sign languages can be made. By considering the physiology of the hand and forearm, they rated each of the patterns of the hand stencils on the cave wall. If these were random, and made with the support of a surface, you would expect 32 different permutations. Instead, there are just 10, all of which can be made in the air, suggesting that they correspond to particular hand gestures. Moreover, shapes that can’t be made in the air but only against a surface, aren’t seen in Gargas – or anywhere else. “We don’t find evidence of hand stencils that would be impossible in sign language,” says Irurtzun.

This isn’t the first time researchers have suggested that Stone Age cave paintings might contain a hidden code. Among stunning depictions of mammoths and bison, there are many graphic marks, ranging from simple lines, dots and triangles to complex configurations, such as ladders and feather shapes called penniforms. Genevieve von Petzinger at the Polytechnic Institute of Tomar, Portugal, has made a comprehensive catalogue of these signs from caves in Europe and has found that the Stone Age people living there had a repertoire of 32 different ones. What’s more, some of these symbols are found in caves throughout the world. Certain signs, including disks and hand stencils, are often found close together, and such combinations are of great interest for understanding the origins of writing, says von Petzinger. After all, combinations of just 26 letters of the Latin alphabet encode the vast amount of information of the English language.

A study published earlier this year even of some of these symbols. Ben Bacon, an independent researcher based in London, worked with Pettitt and others to analyse dots and “Y” shapes found close to depictions of animals and created between 20,000 and 10,000 years ago. They discovered what looks like a sort of hunting calendar to record the behaviour of prey, with the number of marks seeming to record an animal’s mating season in months after the beginning of spring, and a Y denoting the month it gave birth.

It is becoming increasingly clear that Stone Age people were playing around with graphic symbols in surprisingly sophisticated ways. So the idea that hand stencils with missing digits might depict a Stone Age sign language doesn’t seem so far-fetched. Collado Giraldo thinks it is one of the more promising interpretations of their meaning – at least at Gargas and Cosquer. Von Petzinger is also open to the idea. “At a basic level, an outline of a hand is actually a sort of message, perhaps ‘I was here’. Could they have been sending more complex signals in certain times and certain places? I’d say it’s possible for sure,” she says.

But some experts, including Karenleigh Overmann at the University of Colorado, are more sceptical. She isn’t averse to Leroi-Gourhan’s suggestion that the mutilated hands could be a code to coordinate hunting. “But the idea that they encode more language-like messages – I find that a little hard to accept,” she says.

Origins of writing

The jury is still out. But if Etxepare and Irurtzun are correct, their research has an extraordinary implication: the Gargas stencils could be an early form of writing. “If writing is a graphic representation of language – of linguistic expression – and the hand stencils represent an alternate sign language, then yes, that will be a sort of writing,” says Irurtzun. If so, it could be the oldest example of proto-writing that we know of.

Exactly how old remains uncertain because establishing the age of prehistoric cave art is notoriously challenging. It had long been assumed that the Gargas stencils were made in the Gravettian period, between 33,000 and 21,000 years ago. This seems to be backed up by from a crack in the cave wall, which was found to be around 27,000 years old.

In Cosquer cave, some hand stencils have been dated directly and they, too, came in at 27,000 years old – , which is missing its fourth and fifth digits. It is possible that the Gargas hands are much older still, given that many stencils on European cave walls were made 40,000 years ago or more. But even the most conservative date makes them mind-bogglingly older than the first formal writing system. Known as cuneiform, it arose in Mesopotamia just 5500 years ago, although it is thought to have developed from proto-writing with its roots in clay counting tokens that first appeared around 10,000 years ago.

Of course, it will require more evidence to back up this extraordinary claim. Irurtzun is already working on that. Late last year, he joined a team of archaeologists who descended into the deep, dark chambers of Gargas and neighbouring caves with 3D-imaging cameras and special lighting to detect paint marks that can’t be seen with the naked eye.

Their findings haven’t been published yet, but we should soon know more. If they find similar hand gestures throughout the region, that would indicate a more widespread communication system, bolstering the idea that writing has its origins in the Stone Age.