Skip to content

Has Violence Declined? John Gray on Steven Pinker

Steven Pinker, the well-known Harvard evolutionary psychologist, has a new book just out, The Better Angels of Our Nature: Why Violence has Declined, published by Viking. The claim which Pinker defends, that violence is declining and has declined over the course of human history will come as a surprise to many readers who are used to seeing a constant stream of media reports of contemporary wars, riots, homicides, rapes and so on and may have drawn the conclusion that we live in an exceptionally violent age. However, the claim that violence has declined is backed up by a mass of data. Unless Pinker has forged a lot of this data, or ignored significant countervailing evidence, it seems clear that those of us living in the early 21st Century are much less likely to die a violent death than humans living at pretty much any previous stage in history, and much less likely to be subjected to various non-lethal forms of violence too. Pinker offers a rather complicated explanation for the overall decline in human violence. The gist of it is that the spread of various ‘civilising’ cultural influences have operated to reduce violence as states have become successively more powerful and enlightenment ideas have spread. This is a very whiggish explanation of the decline in violence and is very unlikely to be accepted by those who do not share Pinker’s optimistic depiction of the trajectory of human history. 

One very hostile critic of Pinker is the political philosopher John Gray who has penned a review of the book entitled ‘Delusions of Peace’: ( According to Gray, Pinker’s book ‘testifies to our enduring need for faith’, but ‘the result is no more credible than the efforts of Marxists to show the scientific necessity of socialism’. Gray also dismisses Pinker as being ‘like so many contemporary evangelists for humanism’ and he seems to think that Pinker’s views are incompatible with a commitment to Darwinism. This is not because Gray particularly wishes to champion Darwinism but because he appears to thinks it is contradictory for a committed Darwinist such as Pinker to believe that human violence has declined. Gray’s views about the conceptual limits of Darwinist are likely to be rejected by the slew of Darwinists who argue that human morality has evolved and who argue that this evolutionary process is related to the management of interpersonal conflicts. But I don’t really want to berate Gray here for making careless pronouncements about Darwinism. I want to berate him for confusing his attack on Pinker’s explanation of the decline of violence with an attack on Pinker’s case for thinking that violence has declined.

Gray provides reasons to doubt Pinker’s explanation of the decline of violence is correct, however he has not provided reasons to doubt that violence is declining. Unfortunately Gray doesn’t seem to appreciate this rather obvious distinction. According to him:

While Pinker makes a great show of relying on evidence—the 700-odd pages of this bulky treatise are stuffed with impressive-looking graphs and statistics—his argument that violence is on the way out does not, in the end, rest on scientific investigation. He cites numerous reasons for the change, including increasing wealth and the spread of democracy. For him, none is as important as the adoption of a particular view of the world: “The reason so many violent institutions succumbed within so short a span of time was that the arguments that slew them belong to a coherent philosophy that emerged during the Age of Reason and the Enlightenment.

Sorry John, but if you want to dispute Pinker’s case for the conclusion that violence is declining, you’re going to have to engage with the ‘impressive-looking graphs and statistics’. Pinker’s argument is a result of empirical investigation and those graphs and statistics are the heart of its evidential basis. The reasoning about the Age of Reason, the Enlightenment and so on, which Gray does engage with, constitutes a core part of Pinker’s explanation as to why violence is on the decline. This can all be knocked away and the case for the conclusion that violence is declining still stands.

The world we live in a violent place, however available evidence suggest that it was significantly more violent in the not-too-distant past and much more violent in the distant past. It is not entirely clear why violence is declining and it is not clear that it will continue to decline, but it seems hard to deny that it is declining – particularly when you do not engage with the relevant evidence.

Share on

7 Comment on this post

  1. Gray's review has the same logical validity as someone who denies global warming by observing that it's really cold today in his back garden.

  2. I confess I haven't read the book and thus do not know how this data was collected. But I dare to say this graph is utterly wrong:

    Yanomamis are native to my country, so I can give you an additional criticism to what was already said. Data on Yanomami population is hardly exact. Data on Yanomami wars are even less reliable. Since they *are* part of the country, and killing is a crime, a lot of "homicide" tags should pop up when we look at their region in the national violence database.
    Therefore, if not even Brazilian ethnologists can give a precise report on their deaths and births, I wonder how he managed to find them.

    Another question is "why use the Yanomami, at all"? Throughout history they were known for their particular aggressive behaviour, but they are a relatively small tribe. Much larger and peaceful are (were) the Tupi-Guarani, who ruled the whole Brazilian coast down to Uruguay. Scientifically, they would be a better choice.

    Also, they are by no means hunter-gatherers.

    I am sure an ethnologist would be able to give more precise information on that.

  3. Gray is right that Pinker's thesis is rubbish even if his own argument needs a bit more support. The graph that Theo points to is, of course, complete rubbish. These tribes are special cases. Thsi is not a representative selection. The Yanomamo are an example of this. In fact, there is an American anthropologist who married one of their women, a kind of princess, who finds them and their society to possess advantages the West does not possess. They take their children back to her tribe every year. She finds Western society incredibly superficial and shallow. I don't think we can say she is wrong.

    And the graph that Pinker has which "corrects" for deaths over the centuries by adjusting for the total world population is deeply flawed. Another instance of this kind of crap from scientists with no real understanding of the social science ieterature is an example provided by E O Wilson in his Sociobiology. There he discusses an instance where Moses, in conquering another competing group, orders the killing of all pregnant women and young boys &c. Wilson then claims that this is analogous to gibbon behavior, which confers a selective genetic advantage on the predatory group. These are not at all comparable. Moses is not haveing these people killed for any genetic advantage, unconsciously or otherwise. He is having them killed in order to prevent an internecine civil war in the near future when the conquered group might organize resistance to the rule of his group. His slaughter is an attempt to preempt this.

    There is a lot of violence that does not take place on a battlefield. Domestic violence, for example. This has been underreported for years, and unnoticed before that. And what about the psychopaths in our midst, not all of whom are violent? The ordered intelligent psychopath doesn't need to engage in physical violence. He (the majority are men) are able to achieve their ends without the need for physical violence, but the emotional wreckage they leave behind is extensive. This is itself a kind of violence, though not the kind Pinker is concerned with.

    And then there is the distinct possibility that physical violence by groups of th esort Pinker concentrates on is in the slow process of being replaced by more insidious kinds as a consequence of technological innovation, even if it is, as yet, rarely used.

    The conclusion it seems to me has to be that Pinker has entered an arena where he needs to better inform himself.

  4. Pinker provides evidence for a decline in violence but gives no reason to explain this apparent decline. This makes it hard to accept Pinker’s claim. Perhaps increases in rationality might somehow explain this decrease. Gray believes “the idea that humans can shape their lives by the use of reason is an inheritance from rationalist philosophy that does not sit easily with what we know of the evolution of our mammalian brain”. I would agree with Gray rationality does not shape their lives directly but nonetheless I would suggest that our increasing ability to use reason effectively might support Pinker’s claim.

    I certainly don’t believe I am more rational than Plato or that people in general are today more rational than for instance the population of ancient Athens. However my argument does not depend on our increased rationality per se. My argument depends on our increased ability to use reason effectively. This increase is not due to any increase in our brain capacity. Simply increasing a computer’s capacity to compute will not make it intelligent. My increased ability is based on an increase in knowledge, on what I take to be true. I am not naturally more rational than Plato however I live in more knowledgeable times enabling me to use the rationality I possess more effectively. I simply know more than Plato.

    Might our increasing when reasoning effectively support Pinker’s thesis? I suggest any intelligent creature must feel some empathy. I further suggest increasing knowledge naturally leads to increasing empathy because this knowledge forces us to see other people or intelligent creatures as entities that care about things in the same way as we do. This increasing empathy might increase the domain of our empathic concern. It follows our ability to reason effectively might well reduce violence because it increases the domain of our empathic concern, a more detailed argument may be found at . A second reason might be given to support Pinker’s claim even if it is accepted that Hume was correct and that our goals are determined by our emotions. However our increasing knowledge might tell us some goals are unattainable and that we would not appreciate some other goals even if we attained them. It might then be even if our goals are solely determined by our emotions that nevertheless reason determines which of our emotions determine goals. It follows that our increasing ability to reason effectively might rule some emotions, which would produce violence, as reasons to act.

  5. It seems a lot of the previous commentors are missing the point of the article, which is that Gray failed to provide evidential support refuting Pinker's claim of diminishing violence and that in order to do so, he would have to contend with Pinker's statistical data and methods. He does allow that Gray's refuting of Pinker's opinion on 'why' has (or could have) validity.

    If you want to refute Pinker's claim that violence is declining you have to either provide specific evidence that his source data is incorrect, his methods of statistical analysis are subjectively skewed, or his sample population or sample size don't represent the greater populations of the world. But in order to do this, you would have to employ your own statistical analysis by qualifying source data and quantifying trends within different levels of confidence. Or you could get access to Pinker's data and methods and look for 'holes' in his data or potential biases in his choice of methods. Theo gives a good empirical starting point for discounting one graph, but to prove the graph's lack of validity would require a comparative analysis with the greater neighboring populations. And keep in mind that's one graph, the others would have to be addressed as well.

  6. From what I can tell, Gray's stock line is that the modern world is a terrible place and getting worse. It must be anathema to him to be confronted with "700-odd pages […] stuffed with impressive-looking graphs and statistics" showing this to be empirically false. Just as folks hostile to the conclusions implicit in climate change are moved to irrationally reject the data, I expect people of a catastrophist bent are inclined to disregard the enormous body of evidence that suggests that modern life is actually an improvement on the poverty and brevity of pre-industrial life.

    PS – is it just me, or has Prospect turned into a feeble "Lifestyle" mag? A wine column, FFS…

Comments are closed.