In sixteenth-century Paris, a popular form of entertainment was cat-burning, in which a cat was hoisted in a sling on a stage and slowlylowered into a fire. According to historian Norman Davies, "[T]hespectators, including kings and queens, shrieked with laughter asthe animals, howling with pain, were singed, roasted, and finallycarbonized." Today, such sadism would be unthinkable in most of theworld. This change in sensibilities is just one example of perhapsthe most important and most underappreciated trend in the humansaga: Violence has been in decline over long stretches of history,and today we are probably living in the most peaceful moment of ourspecies' time on earth.

In the decade of Darfur and Iraq, and shortly after the century ofStalin, Hitler, and Mao, the claim that violence has beendiminishing may seem somewhere between hallucinatory and obscene.Yet recent studies that seek to quantify the historical ebb andflow of violence point to exactly that conclusion.

Some of the evidence has been under our nose all along. Conventionalhistory has long shown that, in many ways, we have been gettingkinder and gentler. Cruelty as entertainment, human sacrifice toindulge superstition, slavery as a labor-saving device, conquest asthe mission statement of government, genocide as a means ofacquiring real estate, torture and mutilation as routinepunishment, the death penalty for misdemeanors and differences ofopinion, assassination as the mechanism of political succession,rape as the spoils of war, pogroms as outlets for frustration,homicide as the major form of conflict resolution--all wereunexceptionable features of life for most of human history. But,today, they are rare to nonexistent in the West, far less commonelsewhere than they used to be, concealed when they do occur, andwidely condemned when they are brought to light.

At one time, these facts were widely appreciated. They were thesource of notions like progress, civilization, and man's rise fromsavagery and barbarism. Recently, however, those ideas have come tosound corny, even dangerous. They seem to demonize people in othertimes and places, license colonial conquest and other foreignadventures, and conceal the crimes of our own societies. Thedoctrine of the noble savage--the idea that humans are peaceable bynature and corrupted by modern institutions--pops up frequently inthe writing of public intellectuals like Jose Ortega y Gasset ("Waris not an instinct but an invention"), Stephen Jay Gould ("Homosapiens is not an evil or destructive species"), and Ashley Montagu("Biological studies lend support to the ethic of universalbrotherhood"). But, now that social scientists have started tocount bodies in different historical periods, they have discoveredthat the romantic theory gets it backward: Far from causing us tobecome more violent, something in modernity and its culturalinstitutions has made us nobler.

To be sure, any attempt to document changes in violence must besoaked in uncertainty. In much of the world, the distant past was atree falling in the forest with no one to hear it, and, even forevents in the historical record, statistics are spotty until recentperiods. Long-term trends can be discerned only by smoothing outzigzags and spikes of horrific bloodletting. And the choice tofocus on relative rather than absolute numbers brings up the moralimponderable of whether it is worse for 50 percent of a populationof 100 to be killed or 1 percent in a population of one billion.

Yet, despite these caveats, a picture is taking shape. The declineof violence is a fractal phenomenon, visible at the scale ofmillennia, centuries, decades, and years. It applies over severalorders of magnitude of violence, from genocide to war to rioting tohomicide to the treatment of children and animals. And it appearsto be a worldwide trend, though not a homogeneous one. The leadingedge has been in Western societies, especially England and Holland,and there seems to have been a tipping point at the onset of the Ageof Reason in the early seventeenth century.

At the widest-angle view, one can see a whopping difference acrossthe millennia that separate us from our pre-state ancestors. Contraleftist anthropologists who celebrate the noble savage,quantitative body-counts--such as the proportion of prehistoricskeletons with axemarks and embedded arrowheads or the proportionof men in a contemporary foraging tribe who die at the hands ofother men--suggest that pre-state societies were far more violentthan our own. It is true that raids and battles killed a tinypercentage of the numbers that die in modern warfare. But, intribal violence, the clashes are more frequent, the percentage ofmen in the population who fight is greater, and the rates of deathper battle are higher. According to anthropologists like LawrenceKeeley, Stephen LeBlanc, Phillip Walker, and Bruce Knauft, thesefactors combine to yield population-wide rates of death in tribalwarfare that dwarf those of modern times. If the wars of thetwentieth century had killed the same proportion of the populationthat die in the wars of a typical tribal society, there would havebeen two billion deaths, not 100 million.

Political correctness from the other end of the ideological spectrumhas also distorted many people's conception of violence in earlycivilizations-- namely, those featured in the Bible. This supposedsource of moral values contains many celebrations of genocide, inwhich the Hebrews, egged on by God, slaughter every last residentof an invaded city. The Bible also prescribes death by stoning asthe penalty for a long list of nonviolent infractions, includingidolatry, blasphemy, homosexuality, adultery, disrespecting one'sparents, and picking up sticks on the Sabbath. The Hebrews, ofcourse, were no more murderous than other tribes; one also findsfrequent boasts of torture and genocide in the early histories ofthe Hindus, Christians, Muslims, and Chinese.

At the century scale, it is hard to find quantitative studies ofdeaths in warfare spanning medieval and modern times. Severalhistorians have suggested that there has been an increase in thenumber of recorded wars across the centuries to the present, but,as political scientist James Payne has noted, this may show onlythat "the Associated Press is a more comprehensive source ofinformation about battles around the world than weresixteenth-century monks." Social histories of the West provideevidence of numerous barbaric practices that became obsolete in thelast five centuries, such as slavery, amputation, blinding,branding, flaying, disembowelment, burning at the stake, breakingon the wheel, and so on. Meanwhile, for another kind ofviolence--homicide--the data are abundant and striking. Thecriminologist Manuel Eisner has assembled hundreds of homicideestimates from Western European localities that kept records atsome point between 1200 and the mid-1990s. In every country heanalyzed, murder rates declined steeply--for example, from 24homicides per 100, 000 Englishmen in the fourteenth century to 0.6per 100,000 by the early 1960s.

On the scale of decades, comprehensive data again paint a shockinglyhappy picture: Global violence has fallen steadily since the middleof the twentieth century. According to the Human Security Brief2006, the number of battle deaths in interstate wars has declinedfrom more than 65,000 per year in the 1950s to less than 2,000 peryear in this decade. In Western Europe and the Americas, the secondhalf of the century saw a steep decline in the number of wars,military coups, and deadly ethnic riots.

Zooming in by a further power of ten exposes yet another reduction.After the cold war, every part of the world saw a steep drop-off instate-based conflicts, and those that do occur are more likely toend in negotiated settlements rather than being fought to thebitter end. Meanwhile, according to political scientist BarbaraHarff, between 1989 and 2005 the number of campaigns of masskilling of civilians decreased by 90 percent.

The decline of killing and cruelty poses several challenges to ourability to make sense of the world. To begin with, how could somany people be so wrong about something so important? Partly, it'sbecause of a cognitive illusion: We estimate the probability of anevent from how easy it is to recall examples. Scenes of carnage aremore likely to be relayed to our living rooms and burned into ourmemories than footage of people dying of old age. Partly, it's anintellectual culture that is loath to admit that there could beanything good about the institutions of civilization and Westernsociety. Partly, it's the incentive structure of the activism andopinion markets: No one ever attracted followers and donations byannouncing that things keep getting better. And part of theexplanation lies in the phenomenon itself. The decline of violentbehavior has been paralleled by a decline in attitudes that tolerateor glorify violence, and often the attitudes are in the lead. Asdeplorable as they are, the abuses at Abu Ghraib and the lethalinjections of a few murderers in Texas are mild by the standards ofatrocities in human history. But, from a contemporary vantagepoint, we see them as signs of how low our behavior can sink, notof how high our standards have risen.

The other major challenge posed by the decline of violence is how toexplain it. A force that pushes in the same direction across manyepochs, continents, and scales of social organization mocks ourstandard tools of causal explanation. The usual suspects--guns,drugs, the press, American culture-- aren't nearly up to the job.Nor could it possibly be explained by evolution in the biologist'ssense: Even if the meek could inherit the earth, natural selectioncould not favor the genes for meekness quickly enough. In any case,human nature has not changed so much as to have lost its taste forviolence. Social psychologists find that at least 80 percent ofpeople have fantasized about killing someone they don't like. Andmodern humans still take pleasure in viewing violence, if we are tojudge by the popularity of murder mysteries, Shakespearean dramas,Mel Gibson movies, video games, and hockey.

What has changed, of course, is people's willingness to act on thesefantasies. The sociologist Norbert Elias suggested that Europeanmodernity accelerated a "civilizing process" marked by increases inself-control, long- term planning, and sensitivity to the thoughtsand feelings of others. These are precisely the functions thattoday's cognitive neuroscientists attribute to the prefrontalcortex. But this only raises the question of why humans haveincreasingly exercised that part of their brains. No one knows whyour behavior has come under the control of the better angels of ournature, but there are four plausible suggestions.

The first is that Hobbes got it right. Life in a state of nature isnasty, brutish, and short, not because of a primal thirst for bloodbut because of the inescapable logic of anarchy. Any beings with amodicum of self-interest may be tempted to invade their neighborsto steal their resources. The resulting fear of attack will temptthe neighbors to strike first in preemptive self-defense, whichwill in turn tempt the first group to strike against thempreemptively, and so on. This danger can be defused by a policy ofdeterrence--don't strike first, retaliate if struck--but, toguarantee its credibility, parties must avenge all insults andsettle all scores, leading to cycles of bloody vendetta. Thesetragedies can be averted by a state with a monopoly on violence,because it can inflict disinterested penalties that eliminate theincentives for aggression, thereby defusing anxieties aboutpreemptive attack and obviating the need to maintain a hair-triggerpropensity for retaliation. Indeed, Eisner and Elias attribute thedecline in European homicide to the transition from knightlywarrior societies to the centralized governments of earlymodernity. And, today, violence continues to fester in zones ofanarchy, such as frontier regions, failed states, collapsedempires, and territories contested by mafias, gangs, and otherdealers of contraband.

Payne suggests another possibility: that the critical variable inthe indulgence of violence is an overarching sense that life ischeap. When pain and early death are everyday features of one's ownlife, one feels fewer compunctions about inflicting them on others.As technology and economic efficiency lengthen and improve ourlives, we place a higher value on life in general.

A third theory, championed by Robert Wright, invokes the logic ofnon-zero- sum games: scenarios in which two agents can each comeout ahead if they cooperate, such as trading goods, dividing uplabor, or sharing the peace dividend that comes from laying downtheir arms. As people acquire know-how that they can share cheaplywith others and develop technologies that allow them to spreadtheir goods and ideas over larger territories at lower cost, theirincentive to cooperate steadily increases, because other peoplebecome more valuable alive than dead.

Then there is the scenario sketched by philosopher Peter Singer.Evolution, he suggests, bequeathed people a small kernel ofempathy, which by default they apply only within a narrow circle offriends and relations. Over the millennia, people's moral circleshave expanded to encompass larger and larger polities: the clan,the tribe, the nation, both sexes, other races, and even animals.The circle may have been pushed outward by expanding networks ofreciprocity, a la Wright, but it might also be inflated by theinexorable logic of the golden rule: The more one knows and thinksabout other living things, the harder it is to privilege one's owninterests over theirs. The empathy escalator may also be powered bycosmopolitanism, in which journalism, memoir, and realistic fictionmake the inner lives of other people, and the contingent nature ofone's own station, more palpable--the feeling that "there but forfortune go I."

Whatever its causes, the decline of violence has profoundimplications. It is not a license for complacency: We enjoy thepeace we find today because people in past generations wereappalled by the violence in their time and worked to end it, and sowe should work to end the appalling violence in our time. Nor is itnecessarily grounds for optimism about the immediate future, sincethe world has never before had national leaders who combinepre-modern sensibilities with modern weapons.

But the phenomenon does force us to rethink our understanding ofviolence. Man's inhumanity to man has long been a subject formoralization. With the knowledge that something has driven itdramatically down, we can also treat it as a matter of cause andeffect. Instead of asking, "Why is there war?" we might ask, "Whyis there peace?" From the likelihood that states will commitgenocide to the way that people treat cats, we must have been doingsomething right. And it would be nice to know what, exactly, itis.

By Steven Pinker; Steven Pinker is Johnstone Family Professor atHarvard University. His latest book, The Stuff of Thought: Languageas a Window Into Human Nature, will be published by Viking thisyear.