The Better Angels of Our Nature: the Decline of Violence in History and Its Causes

By Steven Pinker

Fewer murders, less cruelty: Pinker's positive history
of humanity.

The Better Angels of Our Nature: the Decline of Violence in History and Its Causes
Steven Pinker
Allen Lane, 832pp, £30

For once, the publisher's advertisement does not lie: this marvellous book really is riveting and myth-destroying. In it, the psychologist Steven Pinker shows, using a wealth of histo­rical, anthropological and geographical data, how violence has declined over human history. A person living today in most parts of the world is far less likely to die by homicide, or through war, than he or she would have been
at any time in the past in virtually every part of the world.

They are "far less likely" by a factor of tens or hundreds. Life was more dangerous if you were a hunter-gatherer anywhere from Australia to the Arctic, if you were a pastoralist in the ancient Middle East, if you lived in a city state in the Roman empire, or in medieval Europe, or in colonial America, China, South America, India or Timbuktu. And whether it is the Maori genocide of the Moriori people of the Chatham Islands in the 1830s, or the Enga people of New Guinea, killing each other at a rate of roughly 300 per 100,000 in continuous warfare, forget the noble savage.

Moreover, this is easily the best time ever to become a victim of the law. If you do fall foul of the state you will not be enslaved, disembowelled, broken on the wheel, crucified, or otherwise ingeniously tortured to death in front of crowds of people enjoying the spectacle of your suffering, the fate of countless people throughout history (Pinker confesses that his gorge rose as he researched this section of the book, and it does not make for easy reading). It is even a good time to be a cat, because you will not be caught and lowered slowly into a fire for the enjoyment of the party, as happened quite commonly in late-medieval Europe.

What has caused this change in our behaviour and appetites? Although I think I detect a bias towards genetic determinism in some of Pinker's earlier work, this new book is unambiguously an investigation of historical and cultural rather than biological change. There are many factors involved, but the groundwork is the rise of Hobbes's Leviathan - the state. Its absence is the most dangerous situation facing any population.

John Locke famously poured scorn on the parties to Hobbes's social contract who handed over the monopoly of force to a sovereign: "This is to think that men are so foolish that they take care to avoid what mischiefs may be done them by polecats, or foxes, but are content, nay think it safety, to be devoured by lions." Locke was not wholly wrong, given that tyrants can be appalling enough, but Hobbes was essentially correct. The suppression of private violence by the rule of law was the first step (one can see the faltering start of the process in the Icelandic sagas). Inside a state, fighting in the courts replaced fighting in the forest or the streets. We then needed a force to diminish interstate warfare, and the principal catalyst here is the rise of trade. Commerce, as Hume and Adam Smith so clearly saw, puts us in a win-win situation; it is not difficult to find it preferable to the lose-lose situation of war.

The process of humanisation, if that is the right word, took off during the Enlightenment and it is a tonic to read Pinker's bracing defence of this against its modern detractors. In the 18th century, across Europe and America, religious wars faded, persecution declined, the notion of heresy withered, toleration advanced, law replaced the whim of the sovereign, cruelty became abhorrent, slavery became ever more problematic, and sympathy became the natural reaction to distress.

In Europe, unfortunately, the idea of the Enlightenment was quickly tarnished by the French Revolution, catalysing the Romantic backlash with its ideals of Sturm und Drang, blood and soil. It is these ideals, and not those of the Enlightenment, that festered, unhappily aided by the distorted idea that Charles Darwin had demonstrated the inevitability of conflict, until they boiled over into two world wars. But blaming the Enlightenment for Kaiser Wilhelm, let alone Adolf Hitler, is rather like blaming education for George W Bush.

As for the reason this almost miraculous transformation in people's sensitivities took place, Pinker is rightly cautious. Obviously there were agitators, such as Voltaire and the influential 18th-century Italian reformer Cesare Beccaria, but what needs to be explained is how these suddenly found an audience. One surprising factor may be the upsurge in "manners" and decency. It is much easier to treat people badly when one finds them disgusting (in his book Yuck!, published this summer, Daniel Kelly describes how parts of the brain associated with normal interactions with other people are shut down by the disgust reaction). Perhaps more importantly, increasing literacy and availability of books drove the rising ability to take other perspectives, to see things from other people's point of view (an ability attacked by the Daily Mail, for instance, or Fox News in the United States).

After the humanitarian revolution and in spite of setbacks during the 20th century, we have more recently seen something like a rights revolution. Civil rights, women's rights, gay rights, children's rights and animal rights are among the quiet revolutions of the very recent past. Advertisers in the 1950s found it quite acceptable to show husbands spanking their wives for misdemeanours and schoolmasters caning children, and to hold up minorities as the butt of humour or contempt.

Just as it would now be unthinkable to smoke in a classroom, so many of the attitudes routinely expressed just a short time ago have become impossible.

Pinker is a polymath, and his ability to recruit data from a phenomenal variety of sources is hugely impressive. He also has a startlingly lucid way of writing about his multitude of topics. He is equally at home explaining the sunk-costs fallacy, the Poisson distribution, the role of testosterone, the function of the prefrontal cerebral cortex, or the relationship between general intelligence and co-operation in the one-shot prisoners' dilemma.

Perhaps Pinker's greatest achievement is in explaining the moral philosophy of Immanuel Kant in prose that is enviably clear. Kant's insight that, in order to respect ourselves, we must respect others rightly appeals to Pinker, but he shows equal admiration for Kant's political writings, and in particular for his understanding of the conditions necessary for "perpetual peace".

Pinker is no Pollyanna. "The forces of modernity," he writes - "reason, science, humanism, individual rights - have not, of course, pushed steadily in one direction; nor will they ever bring about a utopia or end the frictions and hurts that come with being human. But on top of all the benefits that modernity has brought us in health, experience and knowledge, we can add its role in the reduction of violence." He is not complacent about the future, but grateful for the progress of the past. By savouring the accomplishment of a more peaceful world, we can become more attuned to those conditions that made it possible, and perhaps more resolute in protecting them.

Simon Blackburn is Bertrand Russell Professor of Philosophy at Cambridge. His most recent book is "Practical Tortoise Raising: and Other Philosophical Essays" (Oxford University Press, £25)

12 comments

Andrew Chapman's picture

This is quite a good article. Many new questions emerge to the surface, all you need do is to read further information about the issues. Only then one can form a final view on a particular subject. Otherwise everything is seen only in the dimension of how to cum more black and white. The natural logic of pr agentura evaluating things before catering they were properly cognitively processed is a horrible mistake, made by those less intelligent. People should not throw away their common ubytovanie na slovensku sense easily. Anything and everything deserves appropriate time for making judgements.

PB's picture

Andrew is right, Peter is wrong. I have read Pinker's book, and the statistical calculation of relative 'violence' by percentage of global population renders Pinker's argument useless. All it shows is that modern medicine and technology and nutrition has made the expansion of global population faster than the continual increase in violent deaths in the modern period. Who cares? That doesn't tell us anything about how violent our world is, but only about how large of reservoir of lives we have to squander or save. Pinker's belief in statistics has always verged on the magical, but he outdoes himself in this very very absurd book. Nothing he says proves his central thesis that violence has REALLY declined; it has only declined as a percentage of a rapidly growing demographic explosion. Well, thank-you, Medical Science, for penicillin and artificial hearts and cancer therapy and neonatal care units (etc.), and thank-you, Electrical Engineers and Others, for the technology of rapid emergency care, quick transport, and instant communication (etc.). But that does not change the fact that violence has not REALLY declined at all. Lies, damned lies, statistics, and Steven Pinker's statistics....

Freeman2's picture

It's an interesting observation that the stronger and more firmly entrenched is the State, the bloodier the destruction of human life in war. The 20th century - strong State, mass war.

peter's picture

It is quite true that some decrease in human deaths, which earlier would have occurred as a result of violence by other humans, has happened in relatively recent times due to medical and other essentially scientific advances. Pinker deals with this and the effect is not nearly large enough to make a real difference to his thesis. A small illustration of this is the difference between, on the one hand the large reduction in smallpox deaths, and, on the other, the rather smaller (to say the least!) percentage reduction in deaths which would have been due to a bullet through the head, due to poisonous gas, due to an atomic bomb, due to politically forced starvation of people to whom there is no medical access. Indeed, Pinker is correct despite our recent much greater ability to wreak violence, a case he certainly makes quite clearly.

One wonders whether Andrew, and even PB who claims to have, really HAVE read the book. But of course I have no evidence either way, so shouldn't have mentioned that I suppose. Perhaps they skipped the relevant part.

Another confusion in both cases is to do with understanding the difference between 'serious' attempted applications of mathematical statistics, to future (and so unknown) events and to past events which we seem to have no possibility of knowing, as opposed to what Pinker does, relatively unproblematic, making use of essentially known numbers from the past to express in language using the word "probability" facts which are known, but have not been brought forward and argued clearly until now.

To go back to this question of really the 20th century (of course Pinker illustrates plenty of reductions, as time proceeds between earlier centuries), the great desire of some for our recent century to have been terribly violent is quite understandable, repeating what the press and even more serious, if imprecise, sources have been saying more-or-less forever. I doubt that my opponents would advocate going back to the divine right of kings or the religions of child sacrifice, etc., which is surely the main point of the book. And no one doubts a terrible toll of violent human deaths was caused in the late, not-so-great century, by irrational ideology, whether it be of the religious sort which causes many deaths from HIV, the quasi-religious sort of Hitlers, or the apparently non-religious sort of certain pathological individuals who rose within communist systems. That is not in dispute. But neither PB nor Andrew has even attempted, except for generalized unsubstantiatable bluster about statistics, to answer the specific points I have made.

Gideon Polya's picture

Whether a child dies non-violently from war-imposed deprivation or violently (from bombs, bayonets or bullets) the ultimate reality of untimely death is the same and the culpability the same.

On this basis Pinker's thesis must be firmly rejected. Thus global avoidable deaths from deprivation totalled 18 million in 2010, an increase on the 16 million in 2003. 1950-2005 avoidable deaths from deprivation in countries variously occupied by the UK since 1945 totalled 727 million (see "Body Count. Global avoidable mortality since 1950": http://globalavoidablemortality.blogspot.com/2008/08/body-count-global-a... ).

During WW2 there were about 2.4 million Indian soldiers and Indian military casualties totalled 87,000 (see: http://en.wikipedia.org/wiki/World_War_II_casualties ), many of these from deprivation after capture by the Japanese at Singapore. However these military deaths are dwarfed by the 6-7 million Indian civilians who were deliberately starved to death by the British in Bengal adjoining provinces from 1942-1945 (see "Jane Austen and the Black Hole of British History": http://janeaustenand.blogspot.com/ ) .

The non-violent avoidable deaths from deprivation on Spaceship Earth with nuclear terrorist First Worlders in charge of the flight deck is set to increase enormously this century from a current 18 million per year to an average of 100 million per year for the century if man-made climate change is not addressed (top climate scientists predict that only 0.5 billion people will survive i.e. about 10 billion will perish; see "Climate Genocide": https://sites.google.com/site/climategenocide/ ).

Post-invasion non-violent war-related deaths in Afghanistan total 4.2 million while violent deaths (based on Iraq War comparisons) are estimated at 1.4 million. Post-invasion under-5 infant deaths alone total 2.9 million (UN data), 90% avoidable and due to US Alliance Occupier war crimes in gross violation of the Geneva Convention (see "Muslim Holocaust, Muslim Genocide": https://sites.google.com/site/muslimholocaustmuslimgenocide/ ).

Andrew's picture

Peter is speaking nonsense. In real terms, a statistical probability is general, but the specific likelihood of violent death in ay age depends upon an actual time and place. A person living in a commune city-state in Northern Italy in the late 12th century was less likely to be the victim of violent assault than someone living in certain parts of London today.

That does not matter as much as the larger issue I addressed. As a mathematician, I'm making a point that is beyond dispute. You can't meaningfully compare percentiles across such vast differences in sample sizes when the basic unit (one life) remains the same, especially when the eterior conditions are radically shifting (mortality rates of every kind are down in the modern world for simple medical reasons). Following Pinker's logic, a ten-man Arctic outpost where one man kills another is statistically more violent that Rwanda during the ethnic-cleansing sprees. That is a worthless claim. Violence dramatically increased in the twentieth century, but demography radically expanded for any number of reasons. To take global percentiles as a model, then, and claim that violence has actually decreased, is to make an argument of absolutely no value whatsoever. Pinker's book, ergo, is pseudo-scientific cods-wallop. This ain't open to erious debate.

Frank Zurb's picture

With all due respect, I see several serious problems with Prof. Pinker's thesis as presented here (and I don't know whether he addresses them in his book):
- the pure probabilty of someone today experiencing violence directly applies mainly to people in the developed world; it would be necessary to also ask 'for which people mainly has the world gotten safer?' (isn't it much more dangerous to live in many countries than in the West?);
- while the global probability of violence is lower, the sheer numbers of people suffering violence and premature death at the hands of other people globally today is much greater than ever in history (global populations for most of human history were miniscule by comparison with today's);
- (related to the first point) the thesis is certainly 'exceptionalist': the sheer 'holocaust' inflicted today on the totality of species in the world is bigger that ever before - humans are 'extincting' other species globally at an unheard-of and increasing rate, which, importantly, was not the case before;
- indeed the entire planet is most probably headed for - 100% human-made - 'planetary holocaust', and when it hits, we and most species (or, the entire biosphere?) may well not survive. When the planetary crisis reaches critical mass, it will rapidly end the relative peace we now enjoy, and the death throes of our world will make the European Middle Ages look innocent by comparison.
This situation has been brought about by the unstoppable expansion of humans' changing of the environment, which continuously engenders new 'needs', and then, 'inalienable rights' to fulfill those 'needs'. Because we have evolved to do this, and are not going to stop, humanity is by definition thermodynamically unsustainable - on this planet anyway. The consequences must follow. What we are experiencing today is just a relative 'lull before the storm', and any argument for optimism, pure delusion.

peter's picture

As one mathematician to another (as if arguments by authority should hold any water!), I suppose I prefer that my serious argument be not claimed to be unserious (another dubious form of argumentation from Andrew). Neither Pinker, myself, nor anyone else "serious" would claim that there is no variation in 'danger of violent death' from one location to another (yet another "dubiousity'', shall we say, this time Andrew arguing against something which is NOT the position asserted by the other side.

Finally, the one substantive fact he presents compares a situation in which the discrepancy in set sizes is perhaps 6 orders of magnitude (i.e. 10 in the Arctic versus perhaps 10 million in Rwanda) in order to argue against a situation where the discrepancy is well under a single order of magnitude (i.e. about 300 million versus 2,000 million----see my previous response). And of course his example is one where the tiny size of the smaller set makes it ludicrous to use probabilistic language at all convincingly.

This will only become impressive when Andrew gets out his statistics textbook and quotes, with chapter and verse, to make that case, an occurrence which seems quite unlikely. But I remain open to being corrected on anything, including the value of this contribution by Pinker.

abrahamander717's picture

Good to see a review that engages clearly with the argument and evidence. Begs the question of why just for once the NS uses a progressive thinker with a respect for evidence - but mostly uses John Gray, a reactionary thinker with little interest in or respect for evidence.

Gideon Polya - to get your argument to work you would need comparable historical evidence; "Spaceship Earth with nuclear terrorist First Worlders in charge of the flight deck" - does that include the Chinese?

Andrew's picture

Actually, Pinker's book is based on a pure statistical fiction, one that hides the steep increase in violence in the 20th century behind the demographic explosion of the same period; he does this by using a percentile analysis of violent death in respect to the whole bloody global population. That is gibberish, since it falls entirely within the statistical inflation that applies wherever sample sizes are that vastly disproportionate, but basic units of both samples are the same. Disgraceful twaddle.

Latest tweets