From Works in Progress, March 13, 2025:
Life in the state of nature was less violent than you might think. Most of our ancestors avoided conflict. But this made them vulnerable to a few psychopaths.
We are naturally a highly violent species with a thin veneer of civilization that masks a brutal proclivity for violence – or so many people think. In the seventeenth century, Thomas Hobbes said that human life without government is ‘solitary, poore, nasty, brutish, and short’. William Golding’s novel, The Lord of the Flies, which helped him win the Nobel Prize for literature in 1983 and many of us read in school, suggests that boys will rapidly descend into mob violence and brutal cruelty without oversight from authority. To know whether this is true, we need to understand the rates of violence among our ancestors.
There is longstanding disagreement on this issue among scholars: many hold the cultural assumption that humans are by nature bellicose, but there is also a ‘noble savage’ camp that believe the opposite. Steven Pinker’s influential 2011 book The Better Angels of Our Nature tipped the scales by using a data-oriented approach to demonstrate that prehistoric people tended towards extremely high violent death rates, with average rates of violence higher than during the peak years of World War Two.
However, Pinker’s data also showed that prehistoric hunter gatherers seem to have been less violent than prehistoric agriculturalists. This is of critical importance in understanding human history because for 96 percent of our evolutionary history, we were hunter gatherers.
Comprehensive new research has emerged with much more archaeological data on violence in prehistory. Analysis indicates that prehistoric hunter gatherers were considerably less violent than is commonly believed. This finding also seems to be borne out by ethnographic data on modern hunter gatherers with lifestyles relatively similar to their prehistoric ancestors.
Hunter gatherers were not non-violent noble savages by any stretch of the imagination. They were relatively violent when compared with modern standards and even when compared with rates of violence experienced by other primates and mammals in general. However, we think this is primarily because human conflict is so lethal, not because it happens so often. On the contrary, hunter gatherers typically exhibit non-violent norms, with amoral and atypical sociopaths accounting for a disproportionate share of violence, just as in our own societies today.
Understanding this matters. Our extraordinary capacity to inflict lethal violence on each other is normally held in restraint by the natural aversion most people have to violence. If we fail to cooperate, we are vulnerable to falling into vicious cycles of violence that don’t benefit anyone. But we should be more optimistic about our capacity for peacemaking; our ‘better angels’ as Pinker puts it. Despite living in states of political anarchy, hunter gatherers were normally able to cooperate and exist peacefully together.
The debate over hunter gatherer violence
For the first 290,000 years of our species’ approximately 300,000 year history, everyone was a hunter gatherer. In The Better Angels of Our Nature, Steven Pinker argued that hunter gatherers suffered from extremely high rates of violence. Better Angels claims that at least 14 percent of prehistoric hunter gatherers died violently. This equates to a violent death rate of at least 420 per 100,000 people per year, using data on typical hunter gatherer mortality rates.This is a much higher rate of violence than almost anywhere in the modern world in the twentieth and twenty-first centuries. To put it in perspective, global deaths from all types of violence between 2004–21 were around 8 per 100,000 people per year. Even the most violent cities in the world today, in Northern Brazil, South Africa, and on the Mexican side of the US-Mexico border, have murder rates of only around 100 per 100,000 per year.
The implication in Better Angels is that the human mind evolved and developed in a world plagued by constant, endemic violence.
Our 2022 study examined both the ethnographic data – contemporary studies of groups that existed until some modern contact – and archeological data on hunter gatherer violence, much of which comes from data gathered after the publication of Better Angels. We reviewed quantitative estimates of rates of violence in ethnographies, filtering for groups that are most representative of our pre-agricultural ancestors. Our archeological estimates are based on reanalyzing a dataset developed by Gomez et al. (which was released after Better Angels was published and has dozens of extra samples), which attempts to measure rates of violent death by looking for evidence of trauma to skeletal remains. Our study produced estimates for lethal violence around four times lower than Pinker’s figures.
Prehistoric hunter gatherers seem to have been somewhat more violent than the twentieth-century average, but not dramatically so. And this is despite these societies lacking any of the modern state’s apparatus for managing violence: no code of law, no judges, no police, and no sophisticated healthcare.
Building on Better Angels
We think that our data on prehistoric hunter gatherer violence is an improvement on previous estimates for several reasons. Pinker, whose work we greatly admire, generously reviewed our study, writing that it ‘will surely be the standard reference for this issue for years to come.’Firstly, our dataset is more comprehensive. Drawing on work published after Better Angels, our archaeological dataset includes around 150 prehistoric hunter gatherer sites, compared to the 21 sites in Pinker’s source, and our ethnographic dataset draws on data from several modern hunter gatherer studies that Pinker’s source didn’t include.
Secondly, Pinker’s main focus was to compare levels of violence in state and non-state societies. He does not specifically target the question of violence among ancestral hunter gatherers, despite the fact that his data is regularly used in this way by others.
For this reason, Pinker’s archaeological dataset included a number of sites from non-state agricultural societies, which we excluded from our dataset.
Moreover, an important challenge when interpreting evidence from ethnographies of modern hunter gatherers is determining which groups are likely to be representative of hunter gatherers who would have lived before the invention of agriculture. For instance, Pinker includes data on the Ache, Amazonian hunter gatherers from Paraguay. Pinker’s source suggests that 30 percent of Ache deaths were due to warfare: an enormous number. Ethnographies of Ache violence appear to lend weight to this. Consider the following from Pierre Clastres, an ethnographer who studied the Ache Gatu. Clastres describes the Gatu’s struggles against the ‘white men’ and rival groups of Ache, such as the Kyravwa, who ate human fat. On one expedition, the Gatu surrounded a Kyravwa band at dawn.
Almost all the Kyravwa were killed and their women captured. There was a large feast and the . . . Gatu divided up the wives of the conquered men. [The chief] took three young ones for himself.
However, we excluded data on the Ache because they are not deemed to be a good analog for pre-agricultural hunter gatherers by anthropologists. In Clastres’ account, the Ache Gatu are said to be fighting ‘the white men’. In fact, most of the violent Ache deaths in the source were inflicted by loggers with guns – not something pre-agricultural hunter gatherers would have had to contend with....
....MUCH MORE
