In 1983, archaeologists in southern Germany discovered a mass grave containing 34 skeletons. They included 9 adult males, 7 adult females and 16 children.
All of the skeletons showed signs of fatal trauma, including head wounds. None of them showed any signs of defensive wounds, suggesting they were killed whilst running away.
The "Talheim Death Pit" dates from the Stone Age, around 7,000 years ago. It offers some of the oldest evidence of organised group violence between two communities: that is, of war.
Clearly, humans have been fighting wars for thousands of years, and we may not be the only ones. There is growing evidence that several other species also engage in warfare, including our closest relatives the chimpanzees.
That suggests we have inherited our predilection for warfare from our ape-like ancestors. But not everyone agrees that warfare is inbuilt.
Archaeological evidence can be profoundly misleading. The "killer ape hypothesis", proposed by the anthropologist Raymond Dart in 1953, is a case in point.
It turned out that the marks on the fossils were probably inflicted by the teeth of animal predators
Dart discovered the first fossils of Australopithecines, early hominids that lived in Africa 2-3 million years ago. After examining the marks and holes in the bones, Dart became convinced that Australopithecines used primitive weapons like stones, horns and tusks to hunt and butcher their prey and, crucially, one another.
For Dart, Australopithecines' ability to hunt had helped them become "carnivorous creatures, that seized living quarries by violence, battered them to death, tore apart their broken bodies, dismembered them limb from limb, slaking their ravenous thirst with the hot blood of victims and greedily devouring livid writhing flesh."
However, this idea was eventually discredited when it turned out that the marks on the fossils were probably inflicted by the teeth of animal predators.
Nevertheless, the idea that humans have a natural propensity for warfare, a "killer instinct", remains popular. To find out whether it exists, we can study our closest animal relatives.
Warfare is violence involving groups of animals: either group-on-group or group-on-individual attacks. Such "coalitionary violence" is rare in the animal kingdom, confined to a few social insects like ants, and social mammals such as wolves, hyenas and lions.
These supposedly peaceful vegetarians were adept hunters
By far the most studied and debated example is our closest living relative: the chimpanzee, Pan troglodytes.
When primatologist Jane Goodall set out to study a community of chimpanzees in Gombe Stream National Park in Tanzania in the 1960s, little was known about their behaviour. Her work changed all that, for example revealing that chimps make tools.
But Goodall also discovered that these supposedly peaceful vegetarians were adept hunters, who killed other primates – particularly colobus monkeys – for food.
Then in 1974, Goodall got her first taste of something altogether more chilling: inter-group violence between two communities of adult chimps. The "war", as she called it, went on for four years.
The larger of the two groups began "systematically invading" the territory of the smaller group. If the invaders found a rival chimp, they would attack it and leave it to die of its wounds. "They annihilated an entire community that way."
There can be no doubt that groups of chimps kill one another
Goodall was shocked by the brutality of the attacks. She described the invaders "cupping the victim's head as he lay bleeding with blood pouring from his nose and drinking the blood, twisting a limb, tearing pieces of skin with their teeth…"
These events were even more disturbing because the two groups had been united just a few years before, so the victims "were individuals they had travelled with, fed with, played with, grown up with".
There can be no doubt that groups of chimps kill one another. The question is why. Is this a natural part of chimpanzee behaviour, or is it something rare or accidental, or even a response to human interference?
On one side of the debate is anthropologist Richard Wrangham of Harvard University in Cambridge, Massachusetts.
Attacking your neighbours like this is risky: they might fight back
A long-time observer of another chimpanzee community in Uganda, Wrangham believes that chimps and humans are genetically predisposed towards lethal violence. Working with writer Dale Peterson, he set out his ideas in his 1996 book Demonic Males: Apes and the Origins of Human Violence.
Wrangham argues that coalitionary killing can benefit the killers. By taking out a male from another group, the attackers reduce their neighbours' ability to reproduce and at the same time increase their group's access to territory, food and mates.
Of course, attacking your neighbours like this is risky: they might fight back, and kill or at least wound their attackers. But chimp society makes this unlikely.
Although they live in close-knit groups, individual chimps often wander away from their groups to forage alone during the day. These lone chimps are vulnerable.
Wrangham has estimated that a group of chimps should only kill their rivals when they outnumber them by about 5:1. With this overwhelming advantage, the attackers are unlikely to suffer a serious injury.
Lethal raiding has been a feature of human warfare for centuries
This is exactly what Goodall observed during the Gombe war: groups of chimps targeting lone rivals.
This "lethal raiding", as it's called, need not arise from an earlier conflict. It's not an escalation of existing hostilities. Instead, Wrangham argues that it comes from "an appetite" for hunting and killing rivals, "akin to predation".
For Wrangham, coalitionary killing is a natural behaviour that evolved because it could provide more resources for little risk. It evolved in apes, and has carried over to us: lethal raiding has been a feature of human warfare for centuries.
The suggestion that lethal aggression and warfare are innate to chimps is, to say the least, controversial. Many anthropologists reject Wrangham's arguments.
Two of the leading critics are Robert Sussman and Joshua Marshack of Washington University in St Louis. In 2010 they published an extensive critique of Wrangham's hypothesis.
Chimp "warfare" is not an innate behaviour at all
They don't deny that chimps kill, but they question Wrangham's ideas about why they do so.
Sussman and Marshack point out that most animals don't kill one another. Fights are normally displays of aggression rather than physical assaults, and even then they are rarely to the death.
Even male chimps rarely kill. "Most of it is done by threats rather than direct violence… and when they do have fights, most recover." Most of their days are spent engaging in social activity or foraging for food.
Instead, they argue that chimp "warfare" is not an innate behaviour at all, but instead something shaped by the circumstances in which they live – specifically, by human interference. According to Sussman and Marshack, humans have done two things that make chimps more aggressive.
First, we have destroyed much of the chimps' forest habitat, either for logging or to clear space for farming. That means communities are forced to live closer to each other, creating more competition for resources.
There was a great deal more fighting than ever before
Secondly, at a few study sites the researchers fed the chimps, to get the chimps used them. In Goodall's case, this "provisioning" usually involved fruits like bananas. But she soon realised it was having a negative effect on the chimps.
"They were beginning to move about in large groups more often than they had ever done in the old days. Worst of all, the adult males were becoming increasingly aggressive… there was a great deal more fighting than ever before."
Wherever researchers provisioned, the chimps became more agitated and aggressive as they competed for these high-quality foods.
These points are certainly suggestive, but by itself it's not proof that chimps are naturally peaceful. There were several key questions.
Their aim: record every chimp killing at every study site in Africa
What happened when provisioning stopped, as it did in most sites: did the chimps revert to playing nice? What about sites that had never been provisioned: did the chimps there kill less often?
It also wasn't clear that the habitat effect was real. Were chimps really more aggressive in areas that were more severely deforested?
To tackle these questions, anthropologist Michael Wilson of the University of Minnesota in Minneapolis – a former student of Wrangham – teamed up with several dozen colleagues.
Their aim: record every chimp killing at every study site in Africa. The team analysed data from 18 chimpanzee communities, studied over a total of 426 years.
The results were published in 2014 in Nature. There were 152 deaths in total: 58 directly observed, 41 inferred from mutilated bodies and 53 suspected – because the animals had either disappeared or had injuries from fighting.
This pattern of evidence suggests that chimpanzees just do this naturally
About two-thirds of all deaths were the result of coalitionary killing by males from another group.
The team found no correlation between human impacts and the rates of death. "Some communities that had been fed by researchers had high rates of violence and some communities that had been fed didn't have any killings," says Wilson.
What's more, the size of the protected area did not predict the rates of killing. In Kibale in Uganda, which Wilson described as "a high-quality forest that hasn't been logged", the chimpanzees killed at a higher rate than any other community, including Goodall's chimps at Gombe.
"This pattern of evidence suggests that chimpanzees just do this [killing] naturally," says Wilson.
Jill Pruetz of Iowa State University in Ames contributed to the study. She studies west African chimps in Senegal, which are thought to be less aggressive than the east African chimps studied by Goodall.
The other chimps spent hours attacking and biting the corpse
Pruetz initially supported the human interference hypothesis, but has now cautiously changed her mind.
"It seems that [killing] has allowed males to increase their home range sizes," says Pruetz. "That means more resources for those males, like access to females, and ultimately their reproductive success ties into that."
Pruetz has never seen a killing herself, but the chimps at her research site do behave with startling violence.
"One night we heard a series of vocalisations that were unlike what I call normal chimpanzee aggression," she says.
The next morning, her assistant found the dead body of the alpha male. The other chimps spent hours attacking and biting the corpse.
For some, the debate is over. Others reject the Wilson group's findings.
Sussman criticises the study for combining observed, inferred and suspected killing. He calls it "extremely unscientific".
If inter-group killing is an evolved adaptive strategy, it is a pretty rare occurrence
The data also makes inter-group killing look more common than it really is, says Brian Ferguson of Rutgers University in Newark, New Jersey.
He points out that the two sites with the most killings, Tanzania's Gombe and Uganda's Ngogo, account for nearly 60% of all deaths. If you remove them, the death rate drops to 0.03 chimps per year across 416 years observations.
"If inter-group killing is an evolved adaptive strategy, it is a pretty rare occurrence outside those two clusters," says Ferguson.
Furthermore, there is an elephant in the room: bonobos, a second species of ape that are just as closely related to us as chimpanzees. Bonobos muddy the waters still further.
We know far less about bonobos than we do chimpanzees. But what is clear is that bonobos are less aggressive than chimps.
Humans, chimps and bonobos are all descended from a common ancestor
They are sometimes called "hippy apes" because of their peaceful ways – and their habit of having sex as a way of saying "hello".
Wilson's team also analysed data from bonobos. In 92 years of observations of four bonobo communities, there is only one suspected death, and the data includes one site that was heavily provisioned.
Nobody in this debate, on either side, is clear what we can learn from bonobos about innate lethality. Humans, chimps and bonobos are all descended from a common ancestor, but was that ancestor violent or peaceful? No one knows.
It is difficult to look at the data objectively, because every interpretation tells us something about ourselves, and we all have preconceptions about humanity. Frans de Waal of Emory University in Atlanta, Georgia expressed it best: "what we discover in nature is often what we put into it in the first place".
People are reluctant to accept that chimps are violent, says Pruetz, because we use them to understand how our own behaviours evolved. "If it was a more distantly related primate, I don't think we would have the same issues."
He remains convinced that chimps, and therefore humans, have an innate capacity for violence
Wilson goes further. He says his critics are happy to accept that the minds of other animals were shaped by evolution, but won't accept that the same is true of humans.
"They want to believe that… what happens in our heads, in our minds, is entirely the result of the culture around us," says Wilson.
He remains convinced that chimps, and therefore humans, have an innate capacity for violence, shaped by an evolutionary history in which violence was sometimes advantageous.
Clearly, not everyone agrees. But for now let's accept his data at face value. Suppose we do have "violent genes". Does this mean war is inevitable?
On this question, there is cause to be positive, because everyone concerned seems to agree.
Even if we did inherit a propensity for violence, it's not the only thing we inherited
Humans may well be cursed with "a demonic male temperament", says Wrangham. But "we are also blessed with an intelligence that can, through the acquisition of wisdom, draw us away from the five-million-year stain of our ape past."
"Even if we did inherit a propensity for violence, it's not the only thing we inherited," says Stephen Pinker of Harvard University. "We have self-control, empathy, reason and cognition, we have moral norms."
In his 2011 book The Better Angels of Our Nature, Pinker sets out the ways in which humanity has overcome its violent urges. For instance, murder rates and killing in war have both been falling for centuries, and other forms of violence are also becoming less common.
"There can't be a debate about whether we can overcome our inclinations towards violence, because we obviously have and do," says Pinker. "There have been enormous reductions in the rate of violence over the course of history."
War-like chimps are not showing us our destiny at all
Pinker is unsure how far this trend will continue, in particular whether we will ever live in a zero-violence society. "What we do know is that it can be lower than what it is now," he says. "Both the history and the science tell us that it's possible."
On this view, war-like chimps are not showing us our destiny at all. Instead they are showing us the limits of their own ability to restrain themselves; limits that we don't have.
Maybe we shouldn't be so upset by the idea that we have an innate capacity for warfare. Instead, perhaps we ought to simply accept it – and then commit not to act on it.
You can listen to Rami Tzabar's documentary The Origins of War on the Radio 4 website.