Some books make modest claims. Not this one. Steven Pinker says that the process he is tracing is perhaps “the most significant and least appreciated development in the history of our species.” Grand claims like this are often made in the spirit of a jeremiad: We are less civil (or less literate, virtuous, or thoughtful) than we used to be. Or we are in greater danger than we think because of global warming or nuclear terrorism—and my book explains why. But Pinker’s book is about the better angels of our nature. It is that rare creation: a long, sprawling, impassioned blockbuster that tells us how things are getting better, not worse. The world is less violent in more or less every respect—less murder, less war; people are less vicious in their criminality, less savage in their response to crime, and less violent in their religious enthusiasms.


Jean-Christian Bourcart/Getty Images

Steven Pinker, Boston, Massachusetts, October 2005

Better Angels doesn’t just argue that the world is getting less violent. It also explains why: actually it offers layer after layer of understanding. Pinker’s explanations range from the apparently trivial, such as an increase in novel-reading (apparently a great source of interpersonal empathy) and an improvement in table manners (the imposition of a discipline of restraint in daily life where sharp instruments like knives are involved), to momentous causes like the invention of the state and the transformation of morality from ancient taboos to the “razor-sharp” arguments of writers such as Jeremy Bentham or Peter Singer.

If my experience is anything to go by, people respond to Pinker’s claims with some quite searching questions. The world is supposed to be getting less violent. Less violent than what? Less violent than we were a hundred years ago, a thousand years ago, or ten thousand years ago? Secondly, what does Pinker mean by violence? Does it include fear as well as mayhem, disruption as well as death, starvation as well as murder? A third question is the most common and it’s also the most indignant: What about Hitler? What about the world wars and genocides and the murderous tyrannies of the past hundred years?

Let’s take these questions in order. What is Pinker’s time frame? Much of the book is concerned with the ways in which we are less violent now than we were in the Middle Ages. Fourteenth-century records reveal that Oxford used to have a murder rate higher than 100 per 100,000 people every year. If you follow Inspector Morse on television,1 you might think it still does, in order to generate the two or three murders a week that that celebrated detective has to solve. In fact these days a real-life Morse would struggle to find more than a handful of murders every year in the jurisdiction of the Thames Valley Constabulary. For England as a whole the murder rate in 2008–2009 was 1.43 per 100,000—a hundredfold decline in seven hundred years. (Even in the United States, the modern murder rate is only 5.6 per 100,000, down from the medieval Oxford levels that characterized the American West in the 1850s). And it is not just murder; there is less torture and less heretic-burning than there was in fourteenth-century Oxford.

Pinker’s basic claim, however, has more of a longue durée than that. He says that everything we know from anthropology and archaeology (looking, for example, at the proportion of prehistoric human remains marked by evidence of violent death) indicates that our distant ancestors were astonishingly violent—not necessarily in pitched battles, but in ambushes and continual small-scale raiding:

A party of men will slink into an enemy village before dawn, fire arrows into the first men who emerge from their huts in the morning to pee, and then shoot the others as they rush out of their huts to see what the commotion is about…. They can kill a lot of drowsy people before the villagers organize themselves in defense, by which time the attackers have melted back into the forest.

This is how many foraging societies still behave now and apparently it is also characteristic of our simian ancestors’ behavior. Pinker’s book introduced me to the term “chimpicide,” for he entertains the possibility “that the human lineage has been engaged in lethal raiding since the time of its common root with chimpanzees around six million years ago.” Apparently the decline of this sort of violence dates back no earlier than eight or ten thousand years ago. It’s basically a Hobbesian process involving the establishment of states, which held something like a monopoly of coercive force and which had an interest in suppressing this small-scale raiding behavior.

Of course, that monopoly of force also enabled states to organize more effectively for large-scale fighting. But the numbers Pinker assembles here are impressive. In nonstate societies, up to 60 percent of all deaths are due to violent causes; in societies dominated by states, the worst numbers are between 1 and 5 percent. We cannot see this decline if we focus only on battlefield casualties; but if we look at the frequency of ordinary homicide in prehistoric times including small-scale raids and ambushes, we get a much higher number. This determination to add up what we know of all the “ordinary” homicides and the toll exacted week after week by small-scale fighting is key to much of what Pinker argues in this book.



Second question: What is violence? What exactly is it whose decline Pinker claims to be charting? Is he really measuring what is most important in this regard? Violence obviously includes killing in all its terrible varieties—from murder to combat to the immolation of thousands of civilians in a firebombing. It also includes wounding and the deliberate infliction of pain (for example, in torture carried out by state or terrorist or ecclesiastical organizations).

What else does violence include? When he wants to be, Pinker is quite expansive in the statistics he cites. There are fewer abortions now. And we are kinder to animals than we were: Pinker devotes twenty pages to the animal rights revolution, alongside gay rights, women’s rights, and civil rights, in his account of a further and remarkable decline in violence in recent centuries.

Some will say that starvation can be violent, and they have a point. Stalin’s terror famines went hand in hand with his purges and waves of execution. And there are the famines that Amartya Sen writes about, which are the product of reckless economic policies or semideliberate neglect.2 A book called Violence for Equality (1980) by Ted Honderich, who used to teach political philosophy in England, had chapter headings like “Our Omissions and Their Violence.” It raised the possibility that extreme economic deprivation may not be all that different from active homicide either in its morality or in the experience of its victims.

Does violence include the destruction of things, too, like buildings and cities? Maybe, if only because of the way such destruction tears apart ordinary life. However, if we are going to count disruption as well as death, then what about fear of death and the apprehension of destruction? Thomas Hobbes wrote in Leviathan (1651) that “the nature of War, consisteth not in actuall fighting; but in the known disposition thereto, during all the time there is no assurance to the contrary.” Even without actual fighting, the apprehension of it is profoundly disruptive, inasmuch as “contuinuall feare, and danger of violent death” can paralyze and incapacitate us, even before the worst of our fears are realized.

The issues about Hobbesian fear are important for how we think about nuclear weapons. Pinker’s view is that the continuous threat of nuclear destruction does not in itself amount to violence. Thousands of nuclear weapons were poised, armed, and ready for use throughout the cold war, but he reminds us that since they were never used after 1945, the era must be described as peaceful, not violent. That’s too quick, in my opinion. Is a successful threat not an instance of violence if the threat doesn’t have to be carried out? Isn’t the threatened immolation of cities violent in itself, whether it is consummated or not?

Quite apart from anything else, “after 1945” is a bit of cheat, because the nation with the largest nuclear arsenal did make use of it, twice, in August of that year. It wasn’t ever clear that the logic that led to their use—to avoid excessive American casualties in a military campaign that had gone on too long—had somehow ceased to be persuasive after 1945. Not only that, but the fact that they would not be used between 1946 and the present is a product of hindsight that wasn’t available in (say) 1956 or 1962. Ask the American schoolchildren who practiced how to “duck and cover” in response to sirens warning of incoming missiles in the 1950s or the recipients of the Thatcher government’s pamphlet “Protect and Survive” distributed in the 1980s. I am not sure that Pinker is entitled to remove from his calculations the nuclear terror that was felt by ordinary people during this period.


What about World War II? Maybe violence declined from the Dark Ages down to the end of the nineteenth century. But the twentieth century, or at least the first half of it, was appalling. There had been mass murders before; but what of six million people killed in an attempt to extirpate an entire race from the face of the earth? There had been wars before; but what of a five-year war with more than 60 million deaths and the concomitant suffering and destruction that it involved?


After Peter Singer published a favorable notice of The Better Angels of Our Nature in The New York Times Book Review, a reader wrote in to say:

I am grateful for Peter Singer’s enthusiastic review…but certain questions come to mind. For example, what does Pinker make of the fact that the history of the 20th century is dominated by two world wars? How does he respond to the comments of historians and social scientists like Charles Tilly, for whom the 20th century was “the most bellicose in human history,” or Eric Hobsbawm, who called it “the most murderous era so far recorded in history”?3

The correspondent has a point. So how does Pinker respond, in the chapter devoted to this issue, which he calls (without any embarrassment) “The Long Peace”? What he says is: yes, no doubt a lot of people were killed in World War II. But look, the population of the world was much larger in the twentieth century. So if you look on it as a percentage, it doesn’t loom as large as the An Lushan revolt in China in the eighth century or the Mongol conquests in the thirteenth. I exaggerate only slightly the insouciance of this response.

Is it sensible for “atrocitologists” to think about the decline in violence purely as a question of percentages? After all, each life ended is a life, whether it is a tiny or a not-so-tiny percentage of a given population: there is the bloodshed, the pain of dying, and the grieving family, multiplied by 60 million. No manipulation of percentages is going to diminish that grief and suffering.

There is precious little discussion of whether to count absolute or relative numbers—one page, to be exact. In my view, a seven-hundred-page book ought to have room for much more reflection. Unfortunately a lot of that space is crowded with lurid accounts of violence from medieval and early modern times, sometimes bordering on a sort of pornography, all too familiar in studies of torture. I take it these accounts are intended to reinforce the visceral impression that times past were very unpleasant, with their tortures, crucifixions, and breakings on the wheel. But is this necessary? One could equally well stack the book with a detailed account of death and terror in a Nazi death camp, or the agony of a wounded marine at Guadalcanal, or the incineration of this or that resident of Nagasaki, in order to burnish our sense that the twentieth century was quite violent too. Pinker would (rightly) say that such descriptions prove nothing; but then something similar could also be said about the medieval ones.

What little he does say about absolute numbers versus percentages indicates that Pinker wants to judge the quantity of violence—and accordingly the decline of violence—by the risk or justified fear that any given person will meet a violent end. If there are four of us (Adam, Eve, Cain, and Abel) and one person is killed, well that’s a 25 percent prospect, which is pretty scary. If there are six billion of us and 60 million are killed over five years, then that’s a one percent chance, and we shouldn’t be so terrified. After all, one’s chances of dying during the fall of the Roman Empire were considerably higher than that.

It is hard to know what to think about this. Does it make any difference that those 60 million deaths were not evenly spread around the world during World War II, that there was something like a 10 percent death toll in Europe, along with more or less comprehensive destruction in many areas? If we are measuring percentages because we want to chart a decline in the risk or danger of death, then shouldn’t Pinker revisit what he said about the danger we all faced during the years of nuclear terror? Pinker acknowledges that his way of thinking about World War II “seems like monstrous disrespect to the victims.” But he combats that impression with lessons in statistical logic—the toll of World War II might represent nothing more than “an isolated peak in a declining sawtooth.” Anyway “the most destructive event in history had to take place in some century” and he adds that it really was just a random event, all the fault of one man (Hitler) and one “singularly destructive ideology” (Marxism, about which Pinker claims Hitler had read in 1913). A responsible book devoted to the amount of violence in the world should have provided a more reflective account than this about the years between the early Thirties and 1946.


If Pinker is right that there has been a decline of killing—and it is pretty clear from his evidence that he must be, at some level (and with some reservations)—then what he has traced over ten thousand years is a change in human nature.

It is not necessarily a physical change, although Pinker is open to an evolutionary explanation. “One can imagine…that in a society that was undergoing a Pacification or Civilizing Process, a tendency toward impulsive violence would begin to pay off less than it did in the days of Hobbesian anarchy”; there could be a proliferation of genes that fortified self-control. “The neurobiology of violence,” he observes, “is a target-rich area for natural selection.”

In fact, Pinker sees little evidence of this. Mostly he puts his expertise as a student of the brain to work in a ninety-page chapter on “Inner Demons” to show how various and complex are the neural pathways that lead to violence. The pathways for predation and exploitation are different from those for revenge or simple rage, and they all interact in complicated ways with other neural systems, meaning that they exhibit a varied amenability to interference and control by reason and culture.

Does Pinker think we are less violent because we are more moral? Well, yes and no. It is certainly not because we pay more attention to morality. “The world has far too much morality,” says Pinker. His view is that if you add up all the killings carried out in the name of morality, especially religious morality, “they would surely outnumber the fatalities from amoral predation and conquest.”

On the other hand, some of the more recent processes that the book traces are undoubtedly moral in character: the humanitarian reform of the penal system in some countries is a heritage not just of Enlightenment thought, but Enlightenment thought about the application of moral values to public policy. And what Pinker calls in one inspiring chapter “The Rights Revolutions” involves the entrenchment in modern democratic culture of rights-respecting responses to racial diversity, gender differences, and so on. Pinker is not dismissive of modern moral philosophy, but he thinks it has changed its character in recent times:

Modern sensibilities have increasingly conceived moral worth in terms of consciousness, particularly the ability to suffer and flourish, and have identified consciousness with the activity of the brain. The change is part of the turning away from religion and custom and toward science and secular philosophy as a source of moral illumination.

Morality now is less a matter of punitively enforced taboos adopted as commandments or expressed as instincts of disgust, more a matter of thinking in an imaginative and calculative way through all aspects of a given issue and connecting that thinking to broader uses of rationality that were not accessible to our great-grandparents. We are better at putting ourselves in others’ shoes than we used to be. (Pinker thinks this is because we have been reading more novels like Uncle Tom’s Cabin and Oliver Twist. Unfortunately, though, his insistence on furnishing us with vivid accounts of the tortures and abuses we have abandoned doesn’t leave him much room in a chapter on “The Humanitarian Revolution” to explore how such works of fiction have had an effect.)

We also, according to Pinker, have ways of incorporating a sense of “the interchangeability of perspectives, the nonspecialness of our parochial vantage points” into our responses to the conflicts and dilemmas that arise in a crowded world. Confronted with a temptation to attack or exploit, we put ourselves in the shoes of someone who might want to preempt such an attack. Confronted with the fear that by laying down our arms, we will be exposing ourselves to attack, we can put ourselves in the shoes of those who are tempted to attack us and communicate to them that we are aware of this temptation but that we are offering to cooperate in a way that will make us both better off. Better Angels claims that we have solved the Prisoner’s Dilemma (Pinker calls it the Pacifist’s Dilemma) “by blending the payoffs of the two antagonists.”

Once we start thinking in these ways—insisting on explanations, addressing the problems we face in an objective calculative fashion, and taking into account the consequences for all parties, not just ourselves—then we are on a sort of “escalator of reason,” from which it is very difficult to jump off. The image of the escalator comes from Peter Singer: “Once reasoning has got started it is hard to tell where it will stop.”4 In sphere after sphere of human life, we begin to think our way to new solutions, or learn from the thinking of others, dissolving the dangerous pessimism that might tell us that nothing but violence will do.


I said at the outset that Pinker has written a cheerful book: everything is getting better. Is it too cheerful? There is no guarantee that the future will be like the past. The decline in violence, says Pinker, is associated with certain political, economic, and ideological conditions: “If the conditions reverse, violence could go right back up.” Better Angels does have a touch of the Pollyanna spirit, so far as the future is concerned. Pinker seems to think that he has to reassure us that international terrorism is dead or dying, that Iran probably won’t get or use nuclear weapons (quite implausible in light of recent reports), and that global warming will not lead to any significant increase in violence or disorder. Like all of us, he wants it to be the case that the processes that have brought us to “the most peaceable era in our species’ existence” will continue.

Some fear that the new moral thinking he extols—a turning away from the taboos of “religion and custom and toward science and secular philosophy as a source of moral illumination”—may yet make us more brutal in our calculations. The rational calculation of consequences has not always been a force for the better. Pinker is vehemently opposed to the sort of critique that associates Enlightenment thinking with the great crimes of Hitler, Stalin, and Mao. The industrial and bureaucratic character of their holocausts was antirational, he says, not the epitome of secular rationality.

This point applies better to the Nazis than to the Soviet gulag. Pinker is right to insist that Hitler’s vision was predicated on a sort of romantic irrationalism, but even there the abyss of the new dark age the Nazis planned for Europe was, in Churchill’s words, “made more sinister, and perhaps more protracted, by the lights of perverted science.”5 And surely we have to ask: Was it made more likely or less likely by a repudiation of traditional taboos, by a sense that there was no longer anything that was unconditionally forbidden?

There is a lot of ambivalence about this question in The Better Angels of Our Nature. On the one hand, some of the “Rights Revolutions” that Pinker charts in Chapter 7 have generated outright prohibitions and new taboos, on rape and sexual harassment for example. On the other hand, Pinker’s discussion of torture reveals some sympathy for what he says “moral philosophers since Jeremy Bentham have pointed out, [that] in theory torture can even be justifiable, most famously in the ticking-bomb scenario.” The use of the ticking-bomb scenario to justify “enhanced interrogation techniques” is, for Pinker, precisely a step on the escalator of reason: we no longer treat the ban on torture as a moral absolute, but we think about its effects and figure out when its use might secure advantages for us and when it might be dangerous. We renounce the concept of the sacred and the idea that there are certain values that are not to be traded off, whatever the consequences.

Pinker, as I said, seems honestly ambivalent about all this in Better Angels. In other work, however, for example in a 2008 article called “The Stupidity of Dignity,” he has been quite scathing about some of the concepts that have been recently introduced into the world of human rights to underpin new forms of respect for persons and to combat old forms of degradation. He dismisses “human dignity” as “a squishy, subjective notion,” which in bioethics at least is doing more harm than good. There is little attention to human dignity in Pinker’s book. But surely it must be one of the ideas that lies behind our concerns to limit violence and that would, if Pinker is right in his basic thesis, motivate our celebration of the decline in violence. If he is right, more people can now live more dignified lives, free of the misery and degradation associated with war, murder, torture, and destruction. We must hope and pray that dignity is not one of the things that has been tossed over the rail of “the escalator of reason” in the course of the rational reconstruction of morality that has brought us to where we are.