The Better Angels of Our Nature: Why Violence Has Declined
This inflation, Flynn realized, is not a kind that one should try to whip, but is telling us something important about recent history and the human mind. Later generations, given the same set of questions as earlier ones, got more of them correct. Later generations must be getting better at whatever skills IQ tests measure. Since IQ tests have been administered in massive numbers all over the world for much of the 20th century, in some countries, down to the last schoolchild and draftee, one can plot a country’s change in measured intelligence over time. Flynn scoured the world for datasets in which the same IQ test was given over many years, or the scoring norms were available to keep the numbers commensurate. The result was the same in every sample: IQ scores increased over time.229 In 1994 Richard Herrnstein and the political scientist Charles Murray christened the phenomenon the Flynn Effect, and the name has stuck.230
The Flynn Effect has been found in thirty countries, including some in the developing world, and it has been going on ever since IQ tests were first given en masse around the time of World War I.231 An even older dataset from Britain suggests that the Flynn Effect may even have begun with the cohort of Britons who were born in 1877 (though of course they were tested as adults).232 The gains are not small: an average of three IQ points (a fifth of a standard deviation) per decade.
The implications are stunning. An average teenager today, if he or she could time-travel back to 1950, would have had an IQ of 118. If the teenager went back to 1910, he or she would have had an IQ of 130, besting 98 percent of his or her contemporaries. Yes, you read that right: if we take the Flynn Effect at face value, a typical person today is smarter than 98 percent of the people in the good old days of 1910. To state it in an even more jarring way, a typical person of 1910, if time-transported forward to the present, would have a mean IQ of 70, which is at the border of mental retardation. With the Raven’s Progressive Matrices, a test that is sometimes considered the purest measure of general intelligence, the rise is even steeper. An ordinary person of 1910 would have an IQ of 50 today, which is smack in the middle of mentally retarded territory, between “moderate” and “mild” retardation.233
Obviously we can’t take the Flynn Effect at face value. The world of 1910 was not populated by people who today we would consider mentally retarded. Commentators have looked for ways to make the Flynn Effect go away, but none of the obvious ones work. Writers on the egalitarian left and on the lift-yourself-up-by-the-bootstraps right have long tried to undermine the very idea of intelligence and the instruments that claim to measure it. But the scientists who study human individual differences are virtually unanimous that intelligence can be measured, that it is fairly stable over a lifetime of an individual, and that it predicts academic and professional success at every level of the scale.234 Perhaps, you might think, children got more quiz-savvy over the decades as schools began to test the living daylights out of them. But as Flynn points out, the gains have been steady over time, while the popularity of testing has waxed and waned.235 Could it be, then, that the content of the test questions, like “Who wrote Romeo and Juliet?” have become common knowledge, or that the words in the vocabulary section have spread into everyday parlance, or that the arithmetic problems have been taught earlier in school? Unfortunately the biggest gains in the IQ tests are found in exactly those items that do not tap knowledge, vocabulary, or arithmetic. 236 They are found in the items that tap abstract reasoning, such as similarities (“What do a pound and an inch have in common?”), analogies (“BIRD is to EGG as TREE is to what?”), and visual matrices (where geometric patterns fill the rows and columns of a grid, and the test-taker has to figure out how to fill a gap at the bottom right: for example, going from left to right in each row, a shape may acquire a border, lose a vertical line, and then have a hollow area blacked in). The subtests on vocabulary and math have shown the smallest rise over time, and others tests that are filled with them, like the SAT, have even shown a bit of a decline in some age groups in some years.237 Figure 9–2 shows the increases on IQ and its various subtests in the United States since the late 1940s.
FIGURE 9–2. The Flynn Effect: Rising IQ scores, 1947–2002
Source: Graph from Flynn, 2007, p. 8.
The Flynn Effect was a scientific bombshell, because if one were to focus only on the improvement in the Matrices and Similarities, one would think that what rose over the decades was general intelligence. These subtests are considered the purest measures of general intelligence because they correlate well with the tendency of people to score well or poorly on a large battery of dissimilar tests. That tendency is called g, and the existence of g is often considered the most important discovery in the science of mental testing.238 If you give people just about any test you can think of that taps the commonsense notion of intelligence—math, vocabulary, geometry, logic, text comprehension, factual knowledge—the people who do well on one also tend to do well on the others. That was not a foregone conclusion. We all know the inarticulate math whiz and the eloquent poet who can’t balance a checkbook, and one might have thought that the different kinds of intelligence trade off resources in the brain, so that the more neural tissue you have for math the less you have for language and vice versa. Not so. Some people are indeed relatively better at math than others, and some relatively better at language, but compared to the population as a whole, the two talents—and every other talent that we associate with the concept of intelligence—tend to go together.
General intelligence, moreover, is highly heritable, and mostly unaffected by the family environment (though it may be affected by the cultural environment). 239 We know this because measures of g in adults are strongly correlated in identical twins separated at birth and are not at all correlated in adopted siblings who have been raised in the same family. General intelligence is also correlated with several measures of neural structure and functioning, including the speed of information processing, the overall size of the brain, the thickness of the gray matter in the cerebral cortex, and the integrity of the white matter connecting one cortical region to another.240 Most likely g represents the summed effects of many genes, each of which affects brain functioning in a small way.
The bombshell is that the Flynn Effect is almost certainly environmental. Natural selection has a speed limit measured in generations, but the Flynn Effect is measurable on the scale of decades and years. Flynn was also able to rule out increases in nutrition, overall health, and outbreeding (marrying outside one’s local community) as explanations for his eponymous effect.241 Whatever propels the Flynn Effect, then, is likely to be in people’s cognitive environments, not in their genes, diets, vaccines, or dating pools.
A breakthrough in the mystery of the Flynn Effect was the realization that the increases are not gains in general intelligence.242 If they were, they would have lifted the scores on all the subtests, including vocabulary, math, and raw memory power, with a rate related to the degree each test correlates with g. In fact the boost was concentrated in subtests like Similarities and Matrices. Whatever the mystery factor in the environment may be, it is highly selective in the components of intelligence it is enhancing: not raw brainpower, but the abilities needed to score well on the subtests of abstract reasoning.
The best guess is that the Flynn Effect has several causes, which may have acted with different strengths at different times in the century. The improvements on visual matrices may have been fueled by an increasingly high-tech and symbol-rich environment that forced people to analyze visual patterns and connect them to arbitrary rules.243 But proficiency with visuals is a sideshow to understanding the gains in intelligence that might be relevant to moral reasoning. Flynn identifies the newly rising ability as postscientific (as opposed to prescientific) thinking.244 Consider a typical question from the Similarities section of an IQ test: “What do dogs and rabbits have in common?” The answer, obvious to us, is that they are both mammals. But an American in 1900 would have been just as likely to say, “You use dogs to hunt rabbits.”
The difference, Flynn notes, is that today we spontaneously classify the world with the categories of science, but not so long ago the “correct” answer would seem abstruse and irrelevant. “ ‘Who cares that they are both mammals?’ ” Flynn imagines the test-taker asking in 1900. “That is the least important thing about them from his point of view. What is important is orientation in space and time, what things are useful, and what things are under one’s control.”245
Flynn was putting words in the mouths of the dead, but that style of reasoning has been documented in studies of premodern peoples by psychologists such as Michael Cole and Alexander Luria. Luria transcribed interviews with Russian peasants in remote parts of the Soviet Union who were given similarities questions like the ones on IQ tests:Q: What do a fish and a crow have in common?
A: A fish—it lives in water. A crow flies. If the fish just lies on top of the water, the crow would peck at it. A crow can eat a fish but a fish can’t eat a crow.
Q: Could you use one word for them both [such as “animals”]?
A: If you call them “animals,” that wouldn’t be right. A fish isn’t an animal and a crow isn’t either.... A person can eat a fish but not a crow.
Luria’s informants also rejected a purely hypothetical mode of thinking—the stage of cognition that Jean Piaget called formal (as opposed to concrete) operations.
Q: All bears are white where there is always snow. In Novaya Zemlya there is always snow. What color are the bears there?
A: I have seen only black bears and I do not talk of what I have not seen.
Q: But what do my words imply?
A: If a person has not been there he cannot say anything on the basis of words. If a man was 60 or 80 and had seen a white bear there and told me about it, he could be believed.246
Flynn remarks, “The peasants are entirely correct. They understand the difference between analytic and synthetic propositions: pure logic cannot tell us anything about facts; only experience can. But this will do them no good on current IQ tests.” That is because current IQ tests tap abstract, formal reasoning: the ability to detach oneself from parochial knowledge of one’s own little world and explore the implications of postulates in purely hypothetical worlds.
If Flynn is right that much of the Flynn Effect is caused by an increasing tendency to see the world through “scientific spectacles,” as he puts it, what are the exogenous causes of the availability of those spectacles? An obvious one is schooling. We know that schooling coaxes adolescents from Piaget’s stage of concrete operation to his stage of formal operations, and that even with schooling, not everyone makes the transition.247 Over the course of the 20th century, and all over the world, children came to spend more time in school. In 1900 an average American adult had seven years of schooling, and a quarter of them had less than four years.248 Only in the 1930s did high school become compulsory.
And during this transition, the nature of the schooling changed. Early in the century reading consisted of standing and reciting aloud from books. As the education researcher Richard Rothstein observed, “Many World War I recruits failed a basic written intelligence test partly because even if they had attended a few years of school and learned how to read aloud, they were being asked by the Army to understand and interpret what they had read, a skill that many of them had never learned.” 249 Another researcher, Jeremy Genovese, documented the changing goals of education in the 20th century by analyzing the content of the high school entrance exams in 1902–13 and comparing them to high-school proficiency tests given to students of the same age in the 1990s.250 As far as factual knowledge is concerned, less is expected of adolescents today. For example, in the geography section of today’s high-stakes test, the students were asked to pick out the United States on a map of the world! Their great-grandparents were required to “name the states you would pass through in traveling on a meridian from Columbus [Ohio] to the Gulf of Mexico and name and locate the capital of each.” On the other hand, a typical test item today requires students to grapple with rates, amounts, multiple contingencies, and basic economics:A community is located in a region where very little drinking water is available. In order to manage their water resources, which of the following should the community NOT do?
A. Increase their water usage.
B. Buy water from another community.
C. Install water-saving devices in homes.
D. Charge higher rates for water.
Anyone who understands the phrase law of supply and demand realizes that option D cannot be the correct answer. But if you simply have an image of a pool of water and people drinking from it, the connection between how much it costs and how quickly it shrinks is not immediately apparent.
Flynn suggests that over the course of the 20th century, scientific reasoning infiltrated from the schoolhouse and other institutions into everyday thinking. More people worked in offices and the professions, where they manipulated symbols rather than crops, animals, and machines. People had more time for leisure, and they spent it in reading, playing combinatorial games, and keeping up with the world. And, Flynn suggests, the mindset of science trickled down to everyday discourse in the form of shorthand abstractions. A shorthand abstraction is a hard-won tool of technical analysis that, once grasped, allows people to effortlessly manipulate abstract relationships. Anyone capable of reading this book, even without training in science or philosophy, has probably assimilated hundreds of these abstractions from casual reading, conversation, and exposure to the media, including proportional, percentage, correlation, causation, control group, placebo, representative sample, false positive, empirical, post hoc, statistical, median, variability, circular argument, tradeoff, and cost-benefit analysis. Yet each of them—even a concept as second-nature to us as percentage—at one time trickled down from the academy and other highbrow sources and increased in popularity in printed usage over the course of the 20th century.251
It isn’t just the chattering classes that have absorbed the shorthand abstractions of the technocracy. The linguist Geoffrey Nunberg has commented on Bruce Springsteen’s lyric in “The River”: “I got a job working construction for the Johnstown Company / But lately there ain’t been much work on account of the economy.” Only in the last forty years, Nunberg notes, would ordinary people have talked about “the economy” as a natural force with causal powers like the weather.252 Earlier they might have said “on account of times are hard.” Or, he might have added, on account of the Jews, the Negroes, or the rich peasants.
We can now put together the two big ideas of this section: the pacifying effects of reason, and the Flynn Effect. We have several grounds for supposing that enhanced powers of reason—specifically, the ability to set aside immediate experience, detach oneself from a parochial vantage point, and frame one’s ideas in abstract, universal terms—would lead to better moral commitments, including an avoidance of violence. And we have just seen that over the course of the 20th century, people’s reasoning abilities—particularly their ability to set aside immediate experience, detach themselves from a parochial vantage point, and think in abstract terms—were steadily enhanced. Can we put these two ideas together to help explain the documented declines of violence in the second half of the 20th century: the Long Peace, New Peace, and Rights Revolutions? Could there be a moral Flynn Effect, in which an accelerating escalator of reason carried us away from impulses that lead to violence?
The idea is not crazy. The cognitive skill that is most enhanced in the Flynn Effect, abstraction from the concrete particulars of immediate experience, is precisely the skill that must be exercised to take the perspectives of others and expand the circle of moral consideration. Flynn himself drew the connection in recounting a conversation he had with his Irish father, who was born in 1884 and was highly intelligent but relatively unschooled.
My father had so much hatred for the English that there was little room left over for prejudice against any other group. But he harbored a bit of racism against blacks, and
my brother and I tried to talk him out of it. “What if you woke up one morning and discovered your skin had turned black? Would that make you any less of a human being?” He shot back, “Now that’s the stupidest thing you’ve ever said. Who ever heard of a man’s skin turning black overnight?”253
Like the Russian peasant considering the color of bears, Flynn’s father was stuck in a concrete, prescientific mode of thinking. He refused to enter a hypothetical world and explore its consequences, which is one of the ways people can rethink their moral commitments, including their tribalism and racism.
Or consider the high school test question about the water usage in a particular town, which requires, among other things, thinking about proportions. Flynn notes that proportionality questions are surprisingly difficult for many adolescents, and are among the skills that rose as part of the Flynn Effect.254 As we have seen, the mindset of proportionality is essential to calibrating the just use of violence, as in criminal punishment and military action. One has only to replace “manage their water resources” with “manage their crime rate” in the test question to see how an increase in intelligence could translate into more humane policies. A recent study by the psychologist Michael Sargent showed that people with a high “need for cognition”—the trait of enjoying mental challenges—have less punitive attitudes toward criminal justice, even after taking into account their age, sex, race, education, income, and political orientation.255
Before we test the idea that the Flynn Effect accelerated an escalator of reason and led to greater moral breadth and less violence, we need a sanity check on the Flynn Effect itself. Could the people of today really be that much smarter than the people of yesterday? Flynn himself, in an early paper, noted incredulously that in some countries, if earlier scoring norms were applied today a quarter of the students should now be classified as “gifted,” and the number of certifiable “geniuses” should have increased sixtyfold. “The result,” he said skeptically, “should be a cultural renaissance too great to be overlooked.”256 But of course there has been an intellectual renaissance in recent decades, perhaps not in culture but certainly in science and technology. Cosmology, particle physics, geology, genetics, molecular biology, evolutionary biology, and neuroscience have made vertiginous leaps in understanding, while technology has given us secular miracles such as replaceable body parts, routine genome scans, stunning photographs of outer planets and distant galaxies, and tiny gadgets that allow one to chat with billions of people, take photographs, pinpoint one’s location on the globe, listen to vast music collections, read from vast libraries, and access the wonders of the World Wide Web. These miracles have come at such a rapid pace that they have left us blasé about the ideas that made them possible. But no historian who takes in the sweep of human history on the scale of centuries could miss the fact that we are now living in a period of extraordinary brainpower.