Comments on Pinker’s History of Violence

In Steven Pinker’s “The Better Angels of our Nature: Why Violence Has Declined” he argues that we have experienced a radical decrease in violence from pre-state societies, to state level societies, and then again in the transition to modern societies. The importance of the transition to state level society, according to Pinker, was that the state ended the perpetual feuding and and small scale warfare that exists in the state of anarchy.

I’m most interested in the claims about levels of violence in pre-state societies. Pinker’s evidence comes basically from two source: archaological sites, where skeletons have been analyzed for signs of violence, and ethnographic data from recent times, where anthropologists have calculated death rates due to homicide. The below graph, taken from the book, summarizes the data from Pinker’s sources:

Pinker arrives at an average of 15% deaths due to homicide in Hunter-gatherer societes, 24.5% in Hunter-horticulturalists & other tribal groups, and for state societies the numbers range from less than 1% to 5%.

Pinker takes the Hobbesian view that life in the state of nature was “solitary, poor, nasty, brutish, and short.” He doesn’t say much about how far back this tendency goes, but given that he doesn’t never qualifies his statements, and his general perspective on the role of states in pacifying violence, it seems he views these numbers as representing long term patterns of human behaviour. Futher, in an interview, when presented with the view that evidence for extensive violence mostly arises in the last 10,000 years, Pinker seems to think this is largely a matter of the deficient skeletal record, which supports the view that he takes the high homicide figures to be representative of the distant past as well.

The point I’m trying to get to is that, in reading Pinker’s book one gets a definite sense that extreme levels of violence (by today’s standards) have been with us throughout human history, and is only diminishing in turns by increasing state intervention and the discovery of more civilized norms.

Never referenced are, for example, anthropologists like Brian Ferguson or Douglas Fry, who take a very different view of the evolution of violence and war. Actually, there is overlap between the two views, but the fact that Pinker never even mentions them or their evidence in a history on the decline of violence is pretty remarkable. At least one would think that, since Ferguson and Fry (and others) see strong evidence for a drastic increase in violence beginning around 10,000 years ago, this would be worth a note in a history on the rates of inter-personal violence.

Pinker suggests that there simply isn’t enough skeletal evidence from before 10,000 years ago – but I find this a bit weak. Brian Ferguson in Archaeology, Cultural Anthropology, and the Origins and Intensification of War (2006), goes over the worldwide evidence of violence, both before 10,000 years ago, as well as tracing the emerging evidence of war and violence in different regions.

While there is certainly evidence of violence, it is not very extensive before 10,000 years ago. For example Brennan (1991) studies all skeletons available from the 100,000 – 10,000 BP from southwestern France – fragments of 203 individuals – and only 5 (2.5%) showed any signs of fractures, but even these were compatible with accidental injury. In the Middle East, 370 skeletons from the Natufian (10,800 – 8,500 BC) were analyzed, and only 2 showed signs of any kind of trauma. There are outliers of course – Nile Site 117 from around 12,000 BC, has 24 out of 59 skeletons are associated with projectiles. As Ferguson notes: “This is a true outlier without continuation…” But, naturally, it’s been included as one of the sites used as archaeological evidence for high rates of violence in Hunter-gatherer societies in Pinker’s book.

A general problem with the kinds of data sources used in Pinker’s book is noted by Ferguson  – that is, people often measure the populations, whether archaeological or ethnographic, because something about the violence catches their eye. Then someone else compiles a list of groups / sites with definite numbers. This is not exactly a random sample.

Ferguson also makes a strong argument for the very misleading nature of ethnographic data, in that it is collected in a world already highly distorted by the impact of Europeans or other more complex societies. Though there is pretty good evidence for a decided lack of peacefulness in the pre-colonial period, we still can’t take numbers from the 20th century and extrapolate them backwards indefinitely, or even numbers from the pre-contact period.

It is also worthwhile to distinguish Simple Hunter-gatherers from Complex Hunter-gatherers if we want to extrapolate very far back in time to when all populations would have been fairly simple. For example, today the less violent hunter-gatherer societies also tend to be the more simple ones. Looking back up towards the graph at the beginning of this post, one can see the huge variability in violence rates even among these perhaps not entirely representative samples. This suggests just how important it is that we not skew the data with unrepresentative or inappropriate data.

I don’t doubt there has been a general decrease in violence. I suspect however the picture is something more like that presented by Ferguson and Fry than the impression one gets from ‘The Better Angels of Our Nature’ – at least in relation to the ‘origins and intensification of war.’

Ferguson presents a picture of initially low-level of violence (relatively), and then, in the last 10,000 years war emerges and violence intensifies – then, with colonialism there is a dramatic dislocation of culture which in many cases increased violence, for a time, before pacification. Where Pinker has a point is that, despite increasing ‘warfare’ that went along with state development, there seems to have been a drop in actual interpersonal violence. Take the study of thousands of skeletons from North & South America over the last several thousand years. They found a 13.4% violent trauma rate among hunter-gatherers, and a 2.7% rate among pre-Columbian city and village dwellers. They also found an increase of violent trauma in the colonial period, which supports Ferguson’s thesis (the highest rates of violent trauma were actually found on black males, 25-34, at rates of 18.53%). However, and this is important, the 13.4% hunter-gatherer rate is within a  relatively recent past (6,000 years), and can’t be extrapolated far beyond that. In fact, data from North and South America from the earliest periods supports the hypothesis that violence was not originally nearly as high. For example, in the Great Plains region, of 173 individuals from all periods up through Woodland, there was only 1 with signs of violence, whereas after 500 AD, 74 of 447 skeletons show signs of violence.

This kind of pattern of violence coming into the archaeological record occurs over and over again – but, strangely, despite his whole book tracing the historical development of interpersonal violence, Pinker does not treat pre-state societies historically, instead lumping them all together. I think that the reason he does this is probably because of his Hobbesian view of man in the state of nature and the connected belief in the decline of violence as being heavily tied to the development of the state and social complexity. Whereas, if this alternative view is the case, there were fairly low levels of interpersonal violence and warfare in a pre-state, anarchic condition – until increasing complexity, and whatever other conditions pertain to the last thousands of years, caused violence to increase (before state level complexity reverses the trend). When stated like this, it is clear why Pinker might simply sidestep the whole issue.

– See also the review by Douglas Fry

– Others have brought up issues with the medieval data as well

Online videos touching on these topics:

If you want to watch lectures/interviews of the ideas mentioned here, watch Steven Pinker summarize his book; watch an interview with Brian Ferguson in which he discusses his issues with attempts to paint human history as extremely violent. Also of interest, a conference on on ‘coalitionary violenence and warfare‘ (2 videos).

1 thought on “Comments on Pinker’s History of Violence

Leave a Reply to rubenmathisen Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s