IQs are rising all over the world and have been doing so for at least a hundred years—a phenomenon known as the Flynn effect. This fact alone is extraordinary, but what follows is a question that is equally compelling: What is the cause? Once the Flynn effect was accepted as authentic, psychologists became fascinated with possible explanations for the effect. They knew that the effect did not result from genetic shifts in the population—100 years is far too short a timeline to produce such significant genetic change. The underlying cause had to be environmental. Soon a consensus began to form: The Flynn effect was not the result of a single cause, but of multiple converging forces working together, their independent effects aggregating to produce a truly remarkable and unprecedented phenomenon.
This chapter is not strictly about the causes of the Flynn effect, but rather something much broader. We need to understand the range of environmental variables that influence IQ. Some of these are likely to be causes of the Flynn effect; others are important by virtue of their potency to predict IQ gains in small experimental studies rather than entire nations. Our ambition must be broad—to understand the scope of factors that raise cognitive functioning. We need to know how environmental factors contribute to IQ gains, but not only for the sake of our understanding of the Flynn effect. Our agenda includes the revolutionary possibility of deliberately enhancing human intelligence.
Almost certainly, one cause of rising IQs is the greater global availability of nutritious foods. The world is not yet cured of hunger and malnutrition, but the typical diet even in the poorest countries is much improved over past generations. More productive varieties of food crops and improvements in agricultural methods mean that severe deficiencies in protein, vitamins, and minerals have become rarer. One manifestation of improved nutrition is the steady increase in height from one generation to the next. Another consequence is longer life spans, although improved medical care also contributes to people living longer.
The effects of good nutrition are not mysterious: Research on child development shows that nutritional factors affect brain function. Various research projects have examined the effects of nutritional enhancements on children’s cognitive development. Some of those studies entailed nutritional improvements to the diets of expectant mothers. One classic experiment showed that prenatal vitamin supplementation to poor mothers boosted their children’s IQ by 4 to 8 points.1 Other studies involved making nutritional improvements to children’s diets, sometimes through vitamin and mineral supplements. Among school-age children, vitamin supplementation has been associated with IQ gains on the order of 4 to 9 points.2 Across this body of research, two conclusions seem clear. First, dietary improvements, whether to expectant mothers or to children, can have measurable effects on the cognitive ability of young children. Second, those effects are strongest among families at risk for dietary deficiencies associated with poverty. Good nutrition is important for everyone, but programmatic improvements are most effective among poor families at risk for nutritional deficiencies.3
One class of nutrients deserves special attention—the category of fats known as long-chain fatty acids. The best known of these fats are the omega-3s, which are commonly found in coldwater fish. These fats confer protective effects for heart disease and other chronic illnesses. The “3” in omega-3 refers to the link on the fatty acid chain that is “unsaturated,” that is, not filled with hydrogen atoms. In the same general category are omega-6s and omega-9s, which also indicate the points where unsaturated links are found. The unsaturated feature is crucial, because it introduces a crook in the molecular chain, which pushes the fatty acids apart and makes them fluid rather than sticky. In fact, you can place fish oil in the freezer and it will remain fluid rather than harden because of the curved structure of the fatty acid chain. The unsaturated portion of omega-3s and similar fats conveys health benefits, but also makes the fat delicate and vulnerable to spoilage. Fish are especially vulnerable to oxidation, which can produce an unpleasant smell and, ultimately, rancidity. Nuts also contain unsaturated fats, which explains why they, too, are prone to spoilage. The health-promoting properties of unsaturated fats are related to their role as anti-inflammatory agents. Many chronic diseases, such as heart disease and arthritis, are linked to physiological inflammation.
Omega-3s and other fatty acids are significant components in the body’s nervous system, including the brain. Introductory courses in brain anatomy teach that the brain is substantially composed of fat, both in the myelin sheaths that insulate neurons and in the glial cells that provide nourishment to hard-working and nutrient-demanding brain neurons. Efficient and healthy brains need fat. The brain’s structural need for fat would not be an issue except for one important fact: Omega-3s and related fats are essential nutrients. This means both that they are required for life and that the body cannot synthesize them. Like vitamins, minerals, and essential amino acids, polyunsaturated fats must be consumed in the diet. If they are consumed in insufficient quantities, then the body will operate suboptimally. In the case of vitamins and minerals, severe deficiencies can eventually lead to disease or even death. The same is true for amino acids, the building blocks of proteins. Deficient consumption of essential fats might be manifest in poor health, but it’s also possible that it can lead to suboptimal cognitive development.
Nutrients are substances whose consumption promotes positive development. Toxins, by contrast, are harmful. In trying to understand the forces that affect the developing mind, we must consider both. Both nutrients and toxins and are important through the life span including prenatally. Fetuses differ significantly in their exposure both to nutritive environments and to toxins. Prenatal exposure matters greatly: Maternal effects on IQ are large, accounting for 40% of all nongenetic effects.
Cigarette smoke, drugs, alcohol, pesticides, heavy metals, and disease can all affect the developing fetus.4 Many of these toxins are also associated with IQ deficits during childhood. For example, maternal use of the barbiturate phenobarbital has been associated with a later cognitive decrement of about 7 IQ points.5 In high concentrations, the environmental toxins mercury and polychlorinated biphenyls (PCBs) exert harmful effects.6 Maternal infection during pregnancy, such as with cytomegalovirus, can also interfere with an infant’s later cognitive development.7 Exposure to lead has been liked to lower IQ scores. Maternal blood lead concentrations above 10 micrograms per deciliter are predictive of suppressed cognitive development in children.8 Evidence is accumulating that even maternal stress can have lasting negative effects on infants’ later development.9
Among the many toxins that can disrupt a child’s cognitive development, none causes more widespread destruction than alcohol. Regular and significant exposure to alcohol in utero is associated with the pattern of abnormal development known as fetal alcohol syndrome. The syndrome is identified by a child’s small stature, facial abnormalities, and very significant cognitive impairment. Fetal alcohol syndrome is associated with a decrement of 35 IQ points, a horrendous penalty.10 Through prenatal exposure to alcohol, a child who would otherwise have an average IQ of 100 would instead have an IQ of 65—well into the range of mental retardation. As devastating as these effects are objectively, another class of children—12 times as many—do not exhibit the physical traits of fetal alcohol syndrome yet suffer cognitive deficiencies that are almost as severe. The IQ decrement for a child with significant prenatal alcohol exposure, but who lacks the defining characteristics of fetal alcohol syndrome, is about 25 IQ points—still devastating and greatly worrisome because of the vast numbers of affected children.
Nutritional factors matter a great deal in infant development. Breastfeeding is especially important for ensuring that a baby has an unrivaled source of nutrients. Breast milk is much more chemically complex than infant formula, with thousands of identifiable chemical components. These include vitamins, minerals, peptides, immunoprotective factors, and growth hormones, as well as omega-3s and other polyunsaturated fats—nutrients that are not always present in infant formula. These fats are essential to the optimal functioning of the central nervous system and are associated with superior problem-solving ability.11 Deficiencies of essential fats, in turn, are linked to suboptimal cognitive development.12 Consistent with the nutritional profile of breast milk, evidence has accumulated for positive associations between breastfeeding and infants’ cognitive development.13 Studies have shown that breastfeeding can confer a cognitive advantage of between 2 and 5 IQ points for full-term infants. Among low birth-weight infants, the advantage is higher still—about 8 IQ points.14
The cognitive benefits of breastfeeding were confirmed in a large-scale study in which 31 hospitals in Belarus were randomly assigned to one of two conditions—a program that encouraged and supported mothers to breastfeed their infants or continuation of normal hospital practices regarding breastfeeding.15 Infants in the experimental group were breastfed exclusively for longer periods of time. At age 6.5 years, children in the experimental group registered higher scores on an intelligence test, ranging from 2 to 6 IQ points, depending on the subscale. Additionally, teachers rated the academic performance of children in the experimental group to be higher in reading, writing, mathematics, and other subjects. The teachers’ ratings were highest for those children who were breastfed exclusively for three to six months.
The role of breastfeeding in infants’ cognitive development is more far reaching than we might assume. That is because the likelihood of breastfeeding is associated with social class: Middle-class women more often breastfeed their babies, while poorer mothers more often rely on infant formula.16 As we will see in greater detail, IQ correlates significantly with social class: Average IQs are higher in the middle class than they are in working-class and poor families. Unsurprisingly, poverty is associated with decreased prospects for children’s intellectual growth. It’s fair to wonder whether the lower incidence of breastfeeding among poorer families is a contributing factor. In any case, parents facing a choice between breastfeeding and formula ought to consider the many nutritive advantages of breast milk, including an innumerable variety of nutrients, abundant essential fats vital to the development of the central nervous system, and immune protection against disease.
There is no question that nutrition can affect a child’s IQ score, but can dietary quality also affect a child’s ability to learn in school? Here, the data are thinner, but we have at least suggestive evidence that when children have access to nutritious food they become more effective learners. This was the conclusion to a dietary intervention study in New York City schools.17 Over a period of three years, several changes were made to school meals, namely, the successive elimination of refined sugars, food colorings, and preservatives. One nutritional change was phased in each year. Test scores climbed in parallel, suggesting that the nutritional improvements were at least partly responsible.
The investigators did not believe that the rising test scores resulted directly from removing harmful additives to food. Instead, they conjectured that when processed foods were removed from the school meal menu, children were forced to choose among more healthy alternatives. Foods with fewer preservatives and with less artificial color and sugar were presumably more nutrient-rich and increased the aggregate quality of children’s diets. We can’t say with full certainty that dietary changes in the New York City schools were responsible for the increases in students’ tests scores. The study was not a controlled experiment in which some children received dietary improvements and others did not. Pinning down a cause-and-effect relationship is therefore not possible, but we have at least correlational evidence that nutritious meals can lead to more effective learning on a large scale.
Very few parents or teachers would be surprised at research showing an association between good nutrition and effective learning. Much more surprising would be the absence of a connection. And, in fact, other research has reinforced a functional association between reliable nutrition and a child’s ability to learn. At a very basic level, energy in the form of glucose—the brain’s essential fuel—can help assure that cognitive processing is optimal.18 The brain is known to consume enormous amounts of caloric energy—about 10 times the rate of the average human body cell. Although the brain accounts for just 2% of the body’s total mass, it consumes roughly 20% of its metabolic energy.19 It’s no surprise, then, that simply supplying the brain with its necessary operating fuel in the form of glucose makes a difference in cognitive performance. Beyond basic calories, children are advantaged if they have dependable access to a richly nutritious diet; children who lack such access are at risk for suboptimal learning.20
As we continue our search for possible causes of rising IQs around the world—the Flynn effect—we must consider the rapid expansion of mass media and information technologies. During the last century, the introduction of radio into remote villages of developing nations may have had widespread effects on people’s thinking and communication patterns.21 Likewise, newspapers, magazines, and television were a source of intellectual enrichment for cultures that previously lacked access to these media. This does not mean that more informational media is always better for cognitive enrichment. Any intellectual benefit associated with mass media is really an optimization function in which an ideal mix involves moderate exposure. Excessive exposure to television reverses the correlation: Viewing television for several hours each day is associated with decreased learning.22 This risk of overexposure does not nullify the more general pattern, however: The global spread of mass media helped to spur intellectual development around the world.
Television, radio, newspapers, and magazines are all old-style media. We also want to know whether the Internet is having effects similar to previous information technologies. It’s a credible possibility: Through the Internet, users not only access mass media instantaneously, they can also query vast databases to gain specific knowledge on politics, history, science, events, and consumer products—indeed, every imaginable knowledge domain. Also, online and PC-based games are potential hothouses for specialized forms of intellectual growth. Many games stretch a user’s ability to manage complex information, and so may have the unintended but positive effect of expanding cognitive powers. Greater intellectual fitness may in turn generalize to school subjects and everyday life.
To see how this is possible we must appreciate that, in order to become more proficient in computer-based games, users often must build skill in mental visualization and the ability to manipulate images in the mind. Games can stretch the game player’s ability to multitask—to hold and coordinate the dynamic information needed to solve a problem. Computer-based games, sometimes feared for their possible anti-intellectual effects, may prove to be closer to mental workouts than to mind-numbing time wasters. Some research shows that computer-based games can actually enhance a player’s ability to represent and transform spatial information, as well as to allocate attention on complex tasks.23 Yet any positive effects of computer-based games will at some point yield diminishing returns—more cannot always be better—and in excess may cause a decrement in other valued forms of intellectual growth, such as the kinds of learning supported by books.
Compared to previous centuries, life in modern society is vastly more complex. We live in a world far more complicated than the one inhabited by our grandparents. The rapidity of societal change is vividly displayed in the information technologies we use to communicate and network. The shelf life of a laptop computer, mobile phone, or software release is a few years at most. Soon the technology becomes unsupportable, incompatible, and obsolete. In a thousand other ways, too, we are forced to adapt to rapid change and expanding choices. Decisional complexity even applies to mundane activities. The grocery store offers aisle after aisle of bewildering food options, each potential purchase conditioned by nutritional and ethical trade-offs. Food choices are just one example. It is incumbent upon each person to keep up with innumerable options and obligations or else risk inundation in a heap of unanswered e-mails and past-due notices; likewise, “must-read” books, articles, and business advisories pile up unheeded. A lot is expected of citizens of the 21st century—arguably too much.
It’s possible that the mind has upgraded its own “operating parameters” in response to rising demands. The Flynn effect might have arisen, in part, from pressures on the mind to adapt to greater complexity in the environment. In response to the need to manage the flow of information intelligently, perhaps the mind has obliged by becoming a more capable instrument. Rates of change in societies seem to link to IQ gains over time. Early accounts of the Flynn effect were largely reported in developed nations, while later gains reflected more recent modernization in other regions. Indeed, as IQ gains in developing nations advance steadily, the rise of cognitive abilities in more advanced nations, such as Denmark, seems to have leveled off.24 The important point is that a particular society’s IQ gains are apparently a function of changes within the culture, including modernization and its correlates—improvements to education, information, health care, and nutrition. Psychologist Ulric Neisser expressed the possibility this way: “Complexity of life” has produced “complexity of mind.”25
Another likely cause of the Flynn effect may be the most surprising of all—education. It’s surprising because we normally think of education as conveying knowledge to young learners, but not intelligence. Yet rising IQs around the world may be quite directly linked to the expansion of schooling to wider populations. The expansion of schooling is a wonderful accomplishment. During the 20th century, emerging conceptions of literacy as a human right meant that, minimally, a primary school education should be made available to everyone, everywhere. Progress toward this vision, although imperfect, has been remarkable.
Alongside gains in universal education is a parallel trend toward longer periods of education. In the United States, the average number of completed grades rose substantially between 1900 and 2000. At the start of the 20th century, a high school education was a rare accomplishment, but by 1950 it had become the norm in the United States.26 By the end of the 20th century, most young Americans had completed at least some college. In the sections that follow, we will see evidence showing that more years of schooling make a difference to students’ ability to think. Stated more directly: Education has effects that go beyond teaching specific knowledge and skills—education also increases intelligence.
The relationship between education and intelligence is both enlightening and pragmatically relevant. We will see that, through education, intelligence has increased in the past. We will also see how this effect can be accomplished purposefully and intentionally in the future—schooling can systematically introduce complex and abstract information to exercise the cognitive skills that collectively define intelligence. During the 20th century, IQs increased, as did the reach of formal education in schools. Because both increased substantially over the course of a century, we can easily discern a correlation between them. But before placing too much confidence in education as a means to increase intelligence, we need firmer evidence that the connection between the two is causal, not simply correlational. Let’s have a look at a few examples.
Evidence that education can enhance intelligence was discovered as early as the 1920s. During that era, one important means of trade in England and northern Europe consisted of transporting goods on barges through extensive systems of canals. Nuclear families often traveled together from town to town on their barges, buying and selling. Families could make a decent living this way, and to all appearances the children raised on canal boats were healthy and well behaved. Even so, English educational authorities knew that something was wrong: The canal boat children were experiencing serious delays in their cognitive development.
The canal boat children lived very different lives from their peers who were not involved in the canal trade. One crucial difference was that the children who lived on barges rarely attended school because their families were itinerant. The British government responded by building special schools that were proximal to city ports. Even with this accommodation, the children attended school on average less than 10 days each year.27 A second major difference between canal boat children and their geographically stable peers was that the IQ scores of the itinerant children declined steeply over time. Among five-year-olds, IQs were in the normal range, but by the age of 10 the average IQ drifted down into the 60s and 70s, a region of the IQ scale associated with mental retardation. The canal boat children’s meager exposure to schooling seemed to have direct and potent negative effects on their IQ scores. With each year of schooling missed, those effects became more devastating.
The case of the canal boat children does not prove a connection between amount of schooling and measured intelligence. It does, however, offer a clue about the relationship between the two. To be convinced that education really does enhance IQ, we need more than a single historical case study. Fortunately, several other studies likewise show that IQ is correlated with the number of years of schooling, but focus less on deprivation and more on the effects of opportunity for further study. Let’s now consider evidence that links changes in IQ with the extent of education experienced.
Research on young men recruited into the Swedish military provides a second strand of evidence that IQ is related to educational level. By the mid-20th century, several decades’ worth of data showed that IQ was correlated with the number of years of schooling the men completed. Again, by itself, a positive correlation between IQ and educational achievement is unimpressive. After all, we expect bright students to have higher scores on selective admissions tests that control which students are admitted to universities, professional schools, and graduate degree programs. It’s quite easy to believe that higher test scores led to more years of education rather than the other way around.
With some care in interpretation, though, it’s possible to see that the reverse holds, too: More education results in higher test scores. This is clearest when the data are analyzed in a way that compares the Swedish males only with peers who received similar levels of education. The outcome variable of interest was IQ change between the age of 10, when all the males were still in school, and 20, after they had joined the military. Depending on the category of education level, the average change in IQ ranged from −1 to +11, with larger upward shifts associated with more formal education.28 When the data were analyzed yet another way, the conclusion was essentially identical. Statistically, the children were subdivided into “blocks” with similar IQs at age 10. The analysis showed that, within blocks, children who had more education had higher IQ scores at age 20. In other words, when comparisons were made only among 10-year-olds with similar IQs, those who completed more years of schooling had measurably higher IQs as young adults. This pattern held even when statistical controls were used to adjust for variation in family wealth.29 These analyses suggest that the IQ–education link is direct, not merely a function of selective admissions to advanced education.
A similar pattern was confirmed, years later, on a separate data set of Swedish youth.30 The parameters differed slightly. This time, IQ was measured at the ages of 13 and 18. Again for this analysis, the children were blocked by their initial IQ scores. Once more, IQ changes within blocks were associated with the number of years of schooling completed.31 Upward IQ drifts of 8 to 10 points were associated with higher levels of education. The Swedish data begin to present a convincing case that education is causal in shifting IQ upward. Taken together, the data are consistent with the hypothesis that formal education does not function simply to convey more knowledge, but also to boost intelligence as measured by IQ tests.
In the United States, investigators obtained results similar to those found in Sweden: IQ was significantly linked to the degree of schooling completed. The American data set differed slightly. The IQ scores of American males were recorded at two age points, 14 and 34. The key predictor variable was the level of education each man achieved. The objective was to determine the pattern of connections between education level and IQ scores at the two time points.32
One correlation was fully expected: IQ scores taken at age 14 predicted subsequent educational achievement with a moderate correlation of r = .36. The correlation is unsurprising, though, because cognitive ability paves the way for educational achievement in at least two ways. One is that college admissions tests function as gatekeepers to more advanced educational levels; the second is that high IQ scores represent greater intellectual readiness to succeed in education. We can easily confirm the standard way of viewing the relationship between intelligence and education—that IQ is a good predictor of educational achievement because it represents the intellectual potential to learn effectively. What we want to know is whether the relationship works the other way: Does education affect IQ?
Because the American data included IQ at age 34, it was possible to test whether educational attainment predicted future IQ. The interesting finding here was that the men’s educational attainment quite strongly predicted their IQ at age 34 with a value of r = .68. This value was much higher than the correlation of IQ at age 14 with later educational achievement. In fact, the effect of education on adult IQ was more than three times the size of the effect of child IQ on education. As with the Swedish data, the findings fall short of being definitive, but the American data add further evidence that IQ scores are affected positively by formal education. The pattern is recurrent: Education can shift IQ scores upward.
Assume for a moment that the foregoing argument is valid: Higher levels of education result in higher levels of measured intelligence. As a practical matter, does a higher IQ really make a difference? Does a change in a test score really show up in more intelligent behavior? Some evidence suggests that it does. To answer, psychologist Lloyd Humphries examined performance differences between American soldiers who served in World War I and those who served in World War II. Using data gathered by military psychologists, Humphries was able to compare how efficiently soldiers learned their jobs. This was possible because many job skills and responsibilities remained much the same between the two wars. The comparison would be telling, because between 1917 and 1942 the average IQ of military recruits rose substantially—about 15 IQ points.33 Examining military records on job performance, Humphries concluded that the IQ shift made a real difference: World War II soldiers learned their jobs more efficiently and effectively than had their counterparts a generation earlier.34
World War II soldiers were advantaged by being more educated. Between World War I and World War II, the average level of schooling in the United States rose from a 9th grade education to the completion of high school at 12th grade. Three additional years of schooling may not seem very significant: Why should three years matter? Simply because those three years are particularly challenging in their demands for mastery of abstract and complex subject matter in comparison to the earlier grades. The ability to learn complex material is crucial. It seems to have made the soldiers in World War II more intelligent as well as more effective learners on the job.
Statistically, more extensive schooling predicts higher IQ scores. The same connection between education and IQ can be discerned from another data source—historical disruptions of schooling. When historical crises result in the suspension of formal education through school closures, the affected children suffer measurable decrements on IQ tests. One example derives from the battlefields of World War II. Between 1941 and 1944, Eindhoven, Holland, was occupied by Nazi soldiers; the town was held by Allied troops starting in 1944. Through the wartime years and for a short time afterward, schooling in Eindhoven was spotty at best. Before the war’s end, many male teachers went into hiding, presumably because they were perceived as a threat to the occupying forces; children attended school irregularly, if at all.35
The effects of school closure on cognitive abilities can be estimated because of a convenient fact: Near Eindhoven was the Philips Academy. Philips routinely tested its applicants, and the resulting performances were readily converted into IQ scores. The scores show a definite pattern during the affected period: They first declined and then returned to baseline levels. Before 1943, the IQs of applicants averaged about 100, exactly as expected. Between 1944 and 1948, IQs dropped approximately 5 points, arguably the result of school disruption, although emotional trauma may well have contributed to the decline. Although the war ended in 1945, average IQs did not return to prewar levels until 1949. Thus, school closures in wartime Eindhoven coincided with a dip and resurgence of IQs among school-age children.
Another historical episode presents a similar pattern—a drop in IQ that coincided with the disruption of formal education. The public schools of Prince Edward County, Virginia, were shut down from the spring of 1959 to the fall of 1963. The closure was a response to enforcement of the 1954 Brown decision by the Supreme Court. In the Brown v. Board of Education ruling, the precedent of “separate but equal” facilities for Black and White students was declared unconstitutional, making segregation illegal. The Prince Edward County School Board resisted the mandatory desegregation by taking a dramatic step: They shut down the public schools. To meet the educational needs of the county’s Caucasian students, an all-White private academy was established in September 1959. In one of the most bizarre episodes of recent American history, Black students had no official provision for their education. Hastily constructed arrangements were made for some of the county’s Black students to receive “compensatory education,” partly through the activities of volunteers. Black students who did not receive compensatory services suffered measurably: Comparisons between the students who received compensatory education and others who did not showed a 15-point IQ advantage for those whose education continued in some form.36
These historical case studies are vivid, but the power of formal education to influence cognitive ability is also evident in more normal and typical events. Consider what happens to children who are almost exactly the same age, but whose exposure to formal education differs significantly. Owing to the arbitrary cutoff dates set by school boards, children can begin kindergarten or first grade with significant age variation—some school starters begin their formal education nearly a year younger than their peers. This difference carries forward such that children at any given age, say 10 years old, can differ by a whole year of formal education. Children of roughly the same age might differ considerably in exposure to more advanced material. What we want to know is whether an additional year of education affects the developing mind. Do children who start school earlier have a cognitive advantage compared to their same-age peers who are “delayed”? Alternatively, are the delayed children better off because they are neurologically more mature and therefore better prepared to learn challenging material and to experience academic success?
Not too surprisingly, researchers have found that intellectual growth is a function of both schooling and maturation. Through statistics, it is possible to separate the effect of schooling from that of biological maturation. Both effects are positive, meaning that maturation and schooling factors are each associated with intellectual growth. What may be surprising, however, is that schooling is the more powerful influence: On most tests, an additional year of schooling has twice the benefit of an additional year of maturation.37 This means that children on the cusp of cutoff dates for grades are cognitively advantaged by going to the higher grade rather than the lower grade. The positive effects of schooling on cognitive growth are especially noticeable on tests of crystallized intelligence. Though somewhat weaker in magnitude, the same effects of schooling also show up on tests of fluid intelligence. Such information may be relevant to parents and teachers. Of course, when facing a decision about whether to place a particular child into a higher or lower grade, this research finding should not be the only factor. Nonetheless, as parents and teachers work toward a decision, the cognitive effects of education should be taken into account. Otherwise, there is a risk of regarding the child’s maturation as the dominant, or even sole, consideration.
Other interesting effects can also be detected from the normal ebb and flow of schooling. For example, the typical cycle of the school year means that most children experience time away from the classroom during the summer months. We might assume that summer vacation has the same effect on all children, but research shows that this is not the case. In general, the summer experiences of middle-class children result in continuity of learning, whether through exposure to books, summer enrichment programs, or travel.38 Poorer children are likely to have less intellectually enriching experiences—at least this seems to be the case based on patterns of cognitive change during the summer months. Differences in summer experiences have consequences: Middle-class children show continuous growth in academic achievement throughout the calendar year. During the summer months, their reading proficiency levels continue to climb, just as they do during the school year. Poorer children show a different pattern: Rather than continuity, their achievement scores display a saw-tooth pattern of climbing during the regular school year followed by a partial downward regression during the summertime.
These two patterns—steady progress versus saw-tooth—lead to a divergence of trend lines over time. In New York City, children from wealthier schools displayed reading levels two years ahead of children from poorer schools by the end of sixth grade. Half of the two-year gap could be accounted for by changes during the school year, but the other half had built up during the summer months. In other words, 50% of the achievement gap between rich and poor schools resulted from divergent summer experiences.39 By the end of the eighth grade, the effects of summer were even more dramatic. Average achievement now differed by 2.5 years, and summers accounted for two years of that gap—a full 80%.
At this point, we must remember to acknowledge two complementary truths about how children’s experience affects their development. First, schools promote intellectual growth. If it were not for schools, the intellectual trajectories of children would almost certainly be more divergent than they currently are. Seen in this light, schooling acts as a buffer to reduce differences in the slopes of achievement rates.40 The second effect, though, is that nonschool experiences also have educational value. Much of the variation in students’ academic achievement is a product of nonschool factors. We can generalize this way: All experiences have potential value for learning, regardless of whether those experiences occur in schools or away from the classroom.
Our span of interest concerns all forms of experience on IQ, including negative experience. That said, we note that a life of delinquency also appears to have significant effects on IQ and intelligence profiles.41 Crime and aggression are associated with an IQ decrement of about 8 points, other factors being equal. The delinquent’s profile of cognitive abilities is fairly consistent—scores on crystallized intelligence are lower than scores on tests of fluid intelligence. The reasons for the lower crystallized scores are not completely clear. The low scores could result simply from poor participation in schooling or greater impulsivity when taking school-like tests of knowledge. Some scholars have speculated that lower crystallized scores and lower IQ might be the result of brain injuries from life patterns of violence or drug abuse. Another possibility is that both lower IQ and a predisposal to delinquency are related to social factors, such as poverty.
It’s a little surprising that nonschool experiences can produce such potent differences in learning outcomes, but what about school experiences? News stories often report that differences in educational quality distinguish schools in wealthy neighborhoods from those in poor communities. Not only are such differences in educational quality real, they also have effects on cognitive development, measured as IQ change. In one study, poor educational quality was associated with a decline of about 1.5 IQ points each year. This translated to the loss of a full standard deviation between the ages of 6 and 16.42 The detrimental effects of poor educational quality held even when controlling for students’ home background and were confirmed with a separate data set. IQ differences of this magnitude are consequential: They have significant effects on college admissions tests, readiness to engage in the study of complex material, and preparedness to enter intellectually challenging careers.
We can speculate about which aspects of the school curriculum might hold special importance in producing the worldwide increase in IQ scores called the Flynn effect. In broad terms, the school curriculum introduces subject matter and literacies that are progressively more abstract and complex. It’s reasonable to suppose that the school curriculum as a whole cultivates a rising capacity to deal with abstract subject matter and to solve complex problems. School subjects might differ, however, in their promotion of the cognitive proficiencies that underlie intelligence—fluid intelligence in particular. Recall that the Flynn effect was manifest on fluid intelligence at least equally as much as on crystallized intelligence, and in some data sets the magnitude of the effect of rising IQs on fluid intelligence was double that of crystallized. One domain of study might have been pivotal in the cultivation of fluid intelligence—mathematics.
Studies of the mathematics curriculum during the span of the 20th century show a clear evolution toward increasing sophistication. The curricular transformation ran in tandem with increasing access to formal schooling in general. The cognitive demands of the mathematics curriculum changed progressively to align with the cognitive characteristics of fluid intelligence.43 In the early 1900s, mathematics was rarely offered as a subject of study to children at the early grades. The math curriculum of the early 20th century was fixated on basic counting skills and rote memorization; later curricula focused much more on spurring students’ ability to think mathematically. By the 1960s, a complex and articulated mathematics curriculum had been introduced through the span of grades in elementary school education, with a rising proportion of challenging mathematics skills, including geometry. Other problem types required students to engage in categorization and multistep problem solving. Some of these exercises called upon students to understand patterns, infer rules, and apply those rules in ways that are very similar to the inductive reasoning skills characterizing fluid intelligence.
Changes to the mathematics curriculum imply that younger students learned progressively more complex mathematics as the decades progressed. Is this the case? A common assumption is that mathematics achievement in the United States is declining, not rising. The data tell a different story. National trends in achievement show that students around the world have become more proficient in mathematics over the decades. The trend holds in other countries as well. Many teachers, particularly those in developing countries that have experienced rapid modernization, rate current students as more intelligent than students of a generation prior.44 In between those two time points the world experienced dramatic change, including change to the typical learning experience in schools.
The mathematics curriculum appears to be a reasonable candidate for spurring the rising IQ scores that constitute the Flynn effect. Like IQ gains, the introduction of complex mathematics spread rapidly and evolved significantly over the course of a century. Mathematics may have yet another vitally important role in the school curriculum. Early mathematics achievement now appears to have surprising power to predict student academic achievement in high school—both in mathematics and in reading.45 We do not understand fully why math seems to have such sweeping importance to the developing intellect. Research that shows this effect presents an intriguing mystery, indeed.
The research we have examined challenges the way we normally think about intelligence and education. The typical way to understand their relationship is that intelligence is an input—a raw material—and an educated mind is the output. Through education, an intellectually able learner is transformed into a knowledgeable citizen and a technically skilled worker. There is nothing incorrect about this conceptualization; it is merely incomplete. The counterpart truth can be expressed this way: Intelligence is not simply a raw material for education; it is also a product of education. We can even quantify the impact of education on IQ: For every year of education, the counterpart gain in IQ is about ½ point.46
To see education as causing IQ gains can help us to interpret the purpose of education, as well as the role of teachers, in entirely new ways. Parents also play a revised role as cultivators of intelligence in their children. We know from research that parental education is a very good predictor of a child’s cognitive development. The mother’s level of education is particularly important in this regard. Well-educated mothers tend to interact with their children in ways that are linguistically rich. Rich language experiences, in turn, help to prepare the child for future academic success. The social and intellectual benefits of education are therefore not limited to a single generation, but can propagate across generations. Cross-generational intellectual richness is exhibited not only in the extent of parent–child interactions, but also in the tone. Well-educated, middle-class parents more often engage with their children in conversations that encourage intellectual exploration and critical thinking.
In the past, education’s role in cultivating intelligence has been largely unrecognized. Now we have plentiful data showing that education has contributed to the world’s supply of intelligence—but not in any planned or deliberate way. What if that role became better known and more intentional? A redefined mission of education might shift the intelligence of students even more significantly, honing their minds into more capable instruments ready to engage the problems and opportunities of the 21st century. A more intelligent population could, in turn, elevate individual and collective human effectiveness. Higher levels of intelligence would equip humanity with a greater readiness to take on the world’s most pressing problems. Intelligence, combined with compassion, conscientiousness, and goodwill, could enhance our prospects for a brighter future.
Among the most important environments for growing young minds is an intellectually rich and supportive home. Research tells us that a child’s home life matters greatly to his or her intellectual development. The expression of every child’s genetic potential relies on environmental circumstances that are information-rich and socially supportive. Impoverished or abusive home environments, by contrast, result in abnormal development. The point is proven by the extreme case of Genie, a girl who was socially isolated by her parents until she was rescued by authorities at the age of 13. Often forced to sit on a potty chair during the day and confined to a sleeping bag at night, Genie’s intellectual and social skills were closer to those of a toddler than a typical teenage girl. When a medical team intervened, Genie was eager to learn. She rapidly developed an expanded vocabulary but never acquired the ability to speak in complex fluent sentences.47
The case of Genie demonstrates the necessity of normal social and linguistic interactions to cognitive function. One of the most potent predictors of cognitive development is the linguistic richness of the home, especially the type and amount of spoken interaction between children and caring adults. Some children grow up in homes that are language-rich, and this is a monumental advantage. Their spoken interactions with adults are frequent and entail a broad vocabulary. Other children’s home experience involves much less interaction using language. As a generalization, middle-class children are exposed to richer verbal interactions than children raised in impoverished homes. Middle-class children hear, and learn to use, a much wider range of vocabulary to understand the world and communicate with others.48
Spoken language is crucial, but so are other aspects of home life. Children’s minds are enriched by the presence of books and other media in the home environment. Material advantages such as good nutrition and health care also contribute. In the child’s direct experience, the availability of toys and play materials, access to a public library, and even family vacations confer detectable benefits on the child’s intellectual development. These experiences are not strictly determined by family wealth and educational background, but there is a connection. Specifically, the characteristics of home environments associated with higher IQ scores are also correlated with a child’s social class, or what sociologists call socioeconomic status.
Socioeconomic status, abbreviated SES, is a complex and abstract variable, but when measured by sociologists it is rather simply calculated. SES can be measured by gathering data on parents’ education, jobs, and family income. When these variables are combined, the composite number is SES. As sociologists know very well, SES is a powerful predictor of a broad range of population characteristics. Among those, SES is positively correlated with both a child’s cognitive development and school learning. Perhaps the most important point here is that the home conditions that lead to healthy, normal children are not exactly the same as the conditions that lead to optimal cognitive development.49 Except in cases of abuse or of significant poverty, the conditions for raising children with normal levels of physical, social, and cognitive development are less demanding than the array of experiences that prime children to reach toward ideal development of intellectual readiness and success in the academic and professional realms. To some extent, these differences in home conditions are what distinguish middle-class from working-class families.
Perhaps more than we would like to contemplate, social status is linked to all manner of cognitive variables, including IQ. Among adults, the correlation of SES with IQ is high, about r = .50. This difference is manifest as an IQ gap of 10 to 20 points between professional-technical workers and blue-collar workers.50 Links between job status, education, and IQ are not too surprising: IQ predicts education, and education in turn opens opportunities for high-status jobs.51 Among children, IQ and SES are correlated about r = .30. As a result, the average IQ gap between children from high-SES families and those from low-SES families is approximately one standard deviation, or about 15 IQ points.
Could there be a genetic explanation for these IQ differences? It’s possible. Evidence for a genetic influence is derived not from the stability of social classes, but rather from social mobility data. To some extent, nuclear families are socially resifted every generation, and an important sifting element is the intelligence of children. IQ can vary substantially within families. This fact, when combined with the predictive function of adult IQ, translates into significant social mobility from one generation to the next.52 To be specific, sons whose IQs are higher than their fathers’ IQs tend to be upwardly mobile; sons whose IQs are lower than their fathers’ tend to drift downward in socioeconomic status. Genetic factors are the most intuitive explanation for brothers’ IQ differences because their home experiences are unlikely to vary much—at least not in comparison to the experiences of children raised in different families. Material and monetary inheritance may help preserve social standing from parents to children, but not completely. To some extent, at least, IQ seems to reassemble the composition of socioeconomic status afresh with each new generation.
SES is an abstract variable. Although it may correlate mathematically with children’s intellectual growth, the abstraction of SES itself cannot be responsible. Rather, it must be other variables—patterns of parent–child interaction or the presence of books, for example—that are the proximal forces on developing IQ.53 Indeed, children’s IQ scores are more highly correlated with specific qualities of the home environment (r = .5) than they are with SES as an abstract variable (r = .3). Positive qualities of the home environment include the availability of play materials and the amount and variety of spoken language between parent and child. Such features of the home not only predict children’s current IQ, they also predict future IQ levels. So-called lag correlations indicate whether qualities of the home at time 1 predict IQ at time 2, years later. Research has shown that these particular forward lag correlations are stronger than the counterpart lag correlations in which IQ at time 1 predicts home qualities at time 2.54 Lag correlations are evidence that the home environment has a causal role in shaping the child’s intellect.
Parents’ attitudes are also crucial, more so than actual socioeconomic status or income. Scholars have found that parents’ positive value on academic achievement, language, and the provision for children’s learning inside and outside the home had a combined predictive effect of .76 on their child’s IQ.55 When we consider the potency of attitudes, we start to understand more fully that SES is only a first approximation of factors that really matter in shaping a child’s intellectual development. These factors affect impoverished families greatly. We know this because data showing the heritability of IQ—roughly .50 over all populations—seems to differ markedly between middle-class and poor families. The heritability of IQ among wealthier middle-class populations is quite high, around .70, presumably because consistent environmental advantages reduce the role of nurture in shaping intelligence. The opposite is true among poorer families, where the heritability of IQ is an astonishingly low .10.56 This tells us that environmental factors are supremely important among less wealthy families. Among the economically disadvantaged, nurture is a powerful force, overshadowing nature in its ability to influence future variation in intelligence and all its repercussive effects. With this realization comes a very significant responsibility, and opportunity, for all of society.
Among the many variables that compose family background, one has particular potency—the educational level of the mother. A mother’s education correlates impressively with her child’s IQ at approximately r = .50. Because a mother often has significant linguistic and emotional interaction with her children, the developing child is strongly affected by the extent and nature of spoken language with her. When compared to poorer or less-educated mothers, college-educated middle-class mothers not only interact with their children using a great deal more language, but also using a different style of interaction. In particular, middle-class mothers tend to use emotionally supportive tones that are nondirective and that encourage free exploration of the environment.57 Poorer and less-educated moms instead tend to use language to warn their children not to behave in certain ways.58 A stern tone may be understandable if the child’s environment is unsafe, but the developmental effects can be detrimental if exploration is discouraged.
Patterns of linguistic interaction employed by many mothers, especially middle-class moms, include “known-answer” questions. When a mother asks her child a question such as “What does a doggie say?” the mother already knows the answer, of course. The purpose of asking the question is to help the child practice what he or she already knows. It’s a teaching device. The known-answer give-and-take is a staple of school discourse: The teacher poses a question, and the student responds. Posing known-answer questions is also the discourse pattern assumed in all forms of testing. The answer to the question is not valued in itself, as it would be in everyday conversation. The ability to provide the correct answer is only a means to an end—as a way of generalizing about the person who answers. For all its artificial qualities, the known-answer discourse pattern is fundamental to the enterprise of Western education. By engaging her child in known-answer dialogue, the middle-class mother begins to prepare the young mind for successful participation in the discourse of formal education.
Parent–child interactions are not characterized simply by the amount and variety of language, but also what values are communicated. Values are important to cognitive development: Correlations between IQ and measured values and aspirations range as high as r = .75. Such values are communicated in parenting styles. Although limited data are available, the conclusion is consistent with common expectations or stereotypes: Middle-class parents tend to be democratic in their parenting style, whereas poorer or working-class parents gravitate toward authoritarian styles.59 Different ways of parenting are manifest in how control is gained over children’s behavior. A democratic but authoritative style lends itself to discussions and joint decisions between parents and children. Reasons are given for rules, but those rules are not necessarily ironclad; exceptions are possible when there are reasons to deviate from them. Authoritarian parenting styles, by contrast, brook no challenge: Parents’ decisions are absolute and nonnegotiable.
Up to this point, we have examined patterns of correlation between SES and child’s cognitive growth measured as IQ. We are left to wonder whether SES directly causes IQ change. Fortunately, a limited amount of direct evidence is available on the question. The cognitive effects of SES were examined in a rare experiment conducted in France. In this study, the influence of genes and social environment were pitted against each other in what became known as the French adoption study.60 In a study design that was both ingenious and daring, birth families were categorized into high-SES and low-SES clusters. The adopting families, also, were categorized as either high-SES or low-SES. The study is daring in that its designers did not employ the typical practice of matching the SES of a child’s birth family and adoption family. Instead, they systematically made placements on the basis of a “cross-adoption” design. This meant that half the babies born to high-SES mothers went to wealthy families and half went to poor families. The same logic applied to babies born to low-SES mothers: Half were adopted into wealthier families and half to poorer families. Through the study design, the classic developmental forces of nature and nurture were deliberately separated, with birth family SES representing genetic influences and adopting family SES representing the influence of the environment.
The findings from the adoption study were illuminating. Focusing on IQ as the main variable of interest, the layout of outcomes is shown in Figure 6.1. The standout data point is that children with the highest measured IQs were born to high-SES mothers and were adopted into high-SES families. Those “high-high” children had an average IQ of about 120. Next highest in IQ were the true cross-adoption children. These were the “high-low” children, those born to high-SES mothers and adopted into low-SES families, whose average IQ was 108. Slightly lower were the “low-high” children—those born to low-SES mothers and adopted into high-SES families—whose IQ averaged about 104. Least advantaged were the children born to low-SES mothers and adopted into low-SES families. Their IQs averaged about 94.
Figure 6.1
Neural Cross-SES Adoption Study: IQ Is Influenced by Biology and SES. Adapted with permission from Capron, C., & Duyme, M. (1989). Assessment of effects of socio-economic status on IQ in a full cross-fostering study. Nature, 340, p. 553.
Notice the IQ span separating the “high-high” and the “low-low” children, approximately 120 to 94, roughly a full 25 IQ points. One standard deviation on the IQ scale is 15 points, which means that the combined effects of birth SES and adopting SES are almost two standard deviations. The cross-adoption categories, “high-low” and “low-high,” yielded measured IQs close to the center point of 100. This pattern shows that, in structural terms, the SES of the birth family and the SES of the adopting family both matter. We have reasons to be cautious in the interpretation of these findings, yet if one condition (birth family) can roughly be equated to nature and the other (adopting family) to nurture, we see in this experiment a parity of power to shape the young mind. Nature and nurture operate jointly. Both are important; neither is dominant by necessity, and neither is to be dismissed.
The effects of families on IQ are manifest in yet other ways. Family size matters, for example. Children’s IQ is negatively related to family size, meaning that with the addition of each child the average IQ drops. The trend is quite robust, but it is a generalization over a large number of families and never a necessary fact about any particular family. Still, interpretations for the trend are easy to generate: With the addition of each subsequent child, the amount and perhaps the quality of parent–child interaction may drop off—again not inevitably, but as an average over hundreds or thousands of families. The effect of family structure on children’s cognitive development seems to generalize to school learning: A small negative effect has been found for increasing family size on mathematics and reading achievement, for example.61
Like family size, birth order displays a negative relationship with IQ. Firstborns are typically advantaged, with average IQ dropping slightly with each subsequent child. Several mechanisms may contribute to this effect. One example is fairly obvious: birth order and family size are not independent. To be a fourth-born child requires that one’s family have at least four children. Therefore, the decline in average IQ with family size is probably related at least partly to birth order effects. With the birth of each child, the average level of intelligence in the family can be seen as decreasing in absolute terms. Spacing of children also factors in: The birth of a new sibling does not affect a teenager as much as it does a preschooler.62
Whatever the explanation for the decline in average IQ with birth order, the trend has a notable exception: The last child is often the beneficiary of an IQ boost. The youngest sibling may be the recipient of special attention from parents or may benefit from older siblings who may relate to the child in a quasi-parental style. Older siblings might also benefit from this quasi-parental role: Birth order effects that associate higher IQ with older siblings may result in part from their opportunities to teach younger siblings. Finally, we should note that the effects of family environment on cognitive development become weaker as children grow older. Their effects are strongest during childhood.
Among adults, too, cognitive trajectories are sensitive to the quality of experience along the life path. We have seen that a more extensive education generally translates to a higher IQ score, and that this effect is detectable at all levels—elementary, secondary, and postsecondary. The effect of a university education on IQ is in the region of 5 to 15 points, an appreciable boost and not surprising given the complexity of material studied in higher education.63 As we noted earlier in the context of family effects, the benefits associated with a college or advanced degree education extend beyond the student’s immediate benefit. Education affects parenting styles, the quality of parent–child discourse, and, inevitably, the values and goals propagated to subsequent generations.
Jobs, too, have an influence on cognitive development. Their effects depend on the nature of the work performed. Jobs that require initiative, judgment, and independent thought increase a worker’s ability to think flexibly. Such jobs encourage a personal value system that prizes self-direction over conformity.64 The cognitive nature of work can, over time, shape the intellectual profiles of workers. Technical jobs that are primarily scientific or mechanical in nature enhance spatial and technical abilities; jobs that are verbally oriented enhance verbal abilities.65 Of course, cognitive ability profiles initially predict career selection; a student who has an aptitude for spatial ability might gravitate toward engineering, for example. Yet research tells us that that the relationship between cognitive abilities and work profession is reciprocal. Cognitive abilities both influence and are influenced by subsequent experience, including experience on the job.
We know that IQ predicts job performance, and that the predictive power of IQ is somewhat higher in more complex jobs. The elevated predictive function for cognitively demanding jobs suggests that those jobs draw more heavily on the resources of intelligence. Jobs that regularly demand the application of intelligence may, in turn, nudge the trend line of IQ upward over the life span. An intellectually challenging job might generate a steady rise in crystallized intelligence—the specialized knowledge that supports highly skilled performance—as well as the flexible thinking and capacity for innovation that marks fluid intelligence. One application of this pattern is rather straightforward: Adults ought to think of their career pathway at least partly in terms of its ability to nourish and develop the intellectual resources of intelligence.
In old age, too, a person’s quality of experience can shift the trajectory of cognitive growth and prevent its decline. One serious threat to cognitive performance in advanced years is degenerative disease, notably Alzheimer’s. Pharmaceuticals can slow the progression of Alzheimer’s disease, though at this time there is no cure. Yet even in the absence of identifiable disease, old age is statistically associated with cognitive decline. The effects of aging are most marked on fluid intelligence, which entails the capacity to think flexibly, to identify patterns in complex information, and to apply those patterns to solve new problems. Declines in fluid intelligence might be explained by decrements among older adults in processing speed as well as by reductions in working memory capacity.66 In less technical language, older adults often cannot think as fast or cope with as much informational complexity as they did when they were younger.
No one really knows if the trend lines that associate cognitive decline with aging are inevitable. We do know, however, that individuals’ life choices can affect the slopes of those trend lines. The principle “use it or lose it” inspires elderly people to stay intellectually active by reading, doing crossword puzzles, traveling, taking courses, and enjoying the company of interesting friends. There is reason to believe that intellectual activity influences brain processes in a positive way. Activity in the brain’s neurons releases brain chemicals known as neurotrophins, which function to stimulate the formation of new synapses between neurons.67 Thinking rewires the brain. Another potent stimulator of neurotrophins is physical exercise.68 For reasons that are not entirely understood, exercise has surprisingly large effects on the formation of new brain connections. Not only that, exercise increases blood flow to the brain by promoting enhancements to the vascular system. Through both cognitive and physical exercise—and perhaps with the aid of pharmaceuticals or supplements—elderly people have powerful tools to maintain and perhaps extend their cognitive powers.
One personal trait that has enhanced potential for growth in the later years of life is wisdom. Wisdom entails the ability to make good decisions, particularly when the conditions underlying good judgment involve competing interests.69 Superficially, wisdom and intelligence might seem to be quite independent personal traits, but it’s possible to see a connection between the two. In weighing competing interests and finding solutions that balance many different considerations and project consequences of decisions, a high capacity to entertain complex information can help. The connection is implicit, in fact, in the Chinese language. The Chinese conceptualizations of intelligence and wisdom are quite close—linguistically, the two are not distinguished.70
A large body of research now shows that human intelligence is a function of experience. The quality of that experience varies from person to person, and the effects of experience on cognitive abilities extend through the life span. Salient factors include variation in the degree and quality of nutrition, mass media, information technology, formal education, and work. The child’s family experience, including exposure to rich linguistic environments, holds special importance in setting the initial trajectories of cognitive development. In adulthood, the cognitive demands of work matter, as do engagement in physical exercise and ongoing intellectual self-challenge in the twilight years. The binding concept among these many forces for cognitive growth is that a person’s level of intelligence is neither an accident nor fully determined by DNA. Instead, the potent resources that compose human intelligence are shifted up or down by specific qualities of experience.
Intelligence is malleable. One particularly engaging follow-up question is whether the human intellect responds favorably when experience is specifically designed to boost cognitive power. In other words, can we deliberately and intentionally enhance human intelligence? The notion that human intelligence can be intentionally enhanced has a strange ring. Visions of deliberate attempts to make people smarter can seem far-fetched, even audacious. Yet the conclusion from research is plain: IQ is sensitive to qualities of experience throughout the life span. Variation in experience has measurable effects on a person’s cognitive profile, including measured IQ. Earlier sections of this chapter focused on aspects of the environment that varied naturally, not programmatically. In the pages that follow, we will explore the possibility that if unplanned variation in life experience can affect IQ, then planned interventions might also produce such effects.
Knowing that experience strongly affects intellectual change makes us aware of the potential for significant cognitive enhancement. Through deliberate attempts to raise intelligence, we can capitalize fully on that potential. In theory, it should be possible to design interventions that will improve cognitive ability markedly. If the results of such interventions are positive, then theory has translated into practice in a very consequential way. The question we must now address is whether specific programs designed to increase intelligence have been successful.
Fortunately, we have answers: Programs designed to raise intelligent functioning have been successful among participants ranging from infancy to old age. Those programs varied widely. Some programs targeted low functioning or at-risk children, while others were designed for gifted students. Some programs had financial sponsorship from national governments; others were carried out in remote corners of the world or run on shoestring budgets. In the remaining pages of this chapter, we will examine a sample of programs that were intended to improve the broad thinking capability of participants. To judge how well these programs worked, we will concentrate on IQ gains without being exclusive of other measures of intelligence. The focus on IQ is favored for quite natural reasons: For many readers, the IQ criterion lends credibility to claims that intelligence per se is enhanced. Also, IQ scores, for all their simplistic reductionism, are nonetheless an acceptable proxy for the wildly complex collection of intellectual competencies we call intelligence.
In most cases, neither parents nor health professionals would consider trying to enhance the cognitive functioning of newborn babies. One exception might be newborns most at risk for suboptimal cognitive growth. Such was the case in a yearlong program designed for premature and low-birth-weight Black infants in urban Philadelphia.71 The intervention began shortly after birth: Visual stimulation was provided by suspending “nursery bird” mobiles above the infants’ cribs. Hospital nurses gave extra verbal and physical stimulation to the babies above the typical and expected standard of care in the hospital. When the babies were sent home, their families were given rattles and picture books. Mothers were also trained in how to enhance their interactions with the newborns to provide extra cognitive stimulation.
The enhanced cognitive stimulation had measurable positive effects. After just a single week of intervention, the babies displayed greater alertness compared to similar infants in a control group. At four weeks of age, infants in the experimental group exhibited superior development and weight gain. One year after birth, the children in the treatment group showed a 10-point advantage in sensorimotor IQ compared to infants in the control group. Sensorimotor IQ is much more basic than cognitive IQ manifestations among older children and adults, but it does indicate whether neurological development is proceeding along a normal and healthy pathway.
There have been many programs that have successfully raised intelligence among preschool children. One of the first programs to attempt direct cognitive enhancement was carried out in the late 1930s by psychologist Helen Dawe. Dawe met regularly with preschool children who resided in the Iowa Soldiers’ Orphans’ Home. Dawe’s “curriculum” was not a sequence of set lessons, but rather a set of guiding principles for interacting with the children. Those interactions were language-intensive, an emphasis based on Dawe’s view that “linguistic symbols [can] serve as intellectual tools.”72 She believed that as children become familiar with the forms of discourse that support asking questions, positing ideas, and presenting explanations, those children can use language to support their own growing capacity to think.
As a scholar of child development, Dawe was careful to document the effects of her program at the Orphans’ Home. She matched the 11 participating children with 11 control children of comparable ages and intellectual characteristics. The IQs of both groups averaged around 80. Dawe noted that children who resided at the Orphans’ Home were cognitively at risk for a simple reason: Their interactions with adults were much less frequent than in a typical family, and far below ideal. While engaged in meaningful conversations with the children, Dawe stressed the proper use of words, encouraged an attitude of curiosity, and tried to replace careless thinking with critical thinking. The entire intervention spanned a two-year period and entailed 92 days of interaction. This translated to about 50 hours of contact time with each child, either one-on-one or in groups.
Children who participated drew benefits from Dawe’s program. When Dawe compared children from the treatment and control groups, she found that they differed in the quality of their spontaneous speech. Children in the treatment group asked more “intellectual” questions per hour and made analytical comments more frequently. When the two groups were compared on the Stanford-Binet Intelligence scale, children in the treatment group scored 16 points higher than did the controls. The gains of participating children were impressive; even so, Dawe did not think that they had reached their limits. Given more time, Helen Dawe believed that the children of the Iowa Soldiers’ Orphans’ Home would have made even greater gains.
A second program designed for preschool children was the Early Training Project, conducted in the 1960s.73 This intervention was intended to forestall declines in cognitive ability among preschool children living in poverty. The project design assumed that such children normally have no discernible neurological abnormality, but that conditions associated with poverty can impede their intellectual development. The cumulative effects of cognitive deprivation and possible trauma may lead poor children to enter school seriously underprepared. Compared to their peers in middle-class neighborhoods, children from poor families often begin school with an intellectual disadvantage that is extremely difficult to overcome.
The Early Training Project involved poor African-American preschoolers and their families. The goal of the project was not to increase the intelligence of the children, but, more modestly, to prevent its decline. Nineteen children participated. The program (offered free of charge to families) was carried out over three summers and lasted 10 weeks during each summer. Like Helen Dawe’s program, the Early Training Project focused heavily on spoken language between children and teachers. The intent was to enhance the children’s use of language, as in asking thoughtful questions and forming precise verbal expressions. Self-regulatory functions were exercised by helping children to develop persistence and the ability to delay gratification. Mothers were trained in the same strategies that the teachers used so that children’s exposure to the precise use of language extended to the home. The Early Training Project had measurable positive effects: Participating children gained a 9-point IQ advantage over matched controls. The 9-point IQ advantage was maintained for several years, even though the average for both groups declined over time.
These results are not uncommon. Other cognitively intensive preschool programs have likewise produced a cognitive advantage that lasted long after the program ended. More typically, though, the advantage of participating children gradually faded such that, eventually, the IQ profiles of participants were no different from those of the controls. Overall, then, program findings tended to be mixed. The most effective preschool programs showed ongoing cognitive benefits: When participating children were compared against matched controls, the IQ advantage could reach 15 to 20 points. Differences of this magnitude are remarkable and would be very consequential if they could be maintained during the school years. Yet even a very good preschool experience might not be sufficient to set a child on a permanent trajectory toward academic success. The possibility that program effects will fade is not too surprising: If the child’s experience is important during the preschool years, then it will continue to be important once schooling begins. To repeat a theme expressed earlier, the quality of experience is relevant and consequential to intellectual development through every stage of life.
Among the best known of the small, intensive preschool programs is the Milwaukee Project. So important is the project that it has been called the “high water mark of environmentalist accomplishment.”74 The Milwaukee Project spanned the range of preschool years, from infancy to primary school entry. Poverty had left a mark on participating families: Mothers of participating children had IQs of 75 or less.75 Like the other programs that we have considered, the Milwaukee Project focused on children’s language development and concentrated on maintaining high levels of intellectual interaction between children and their teachers. In advancing children’s thinking through language, teachers commonly posed open-ended questions as well as helped children to engage in problem solving and logical thinking. The main goal was not to raise the intelligence of children, but simply to prevent its decline—that is, to help children maintain pace with their more advantaged peers.
At first, the Milwaukee Project did not produce evidence of differences between treatment and control groups. After the first few years of the project, however, the cognitive trajectories of treatment and control children diverged dramatically, eventually indicating a 25-IQ point advantage for participating children.76 Some intriguing data showed that the children enrolled in the project were not the only beneficiaries. Positive cognitive effects were also detected among siblings and neighborhood children. This finding was remarkable—somehow cognitive enrichment had spread to nonparticipants through some unidentified means, whether through peer interactions or through the mediation of parents or other adults.
The Milwaukee Project has been controversial, no doubt resulting in part from the startling gains made by participating children.77 Some criticism was leveled against the researchers for the lack of detail provided about the research design and for the inordinate delays in providing specifics about the treatment to the research community. A more pointed critique concerned whether the treatment and control children were comparable at the start of the project. The question of comparability focused on rather arcane details about the measured body lengths of the infants at birth. What bothered some critics was not that children in the two groups differed in their average body lengths, but rather in the variability of lengths. Control children displayed somewhat greater variability, which in the opinion of some critics called into question whether the summary findings could be trusted.
Later, when Milwaukee Project children were tracked through their school years, the children continued to manifest positive effects. As a group, children who participated in the Milwaukee Project as preschoolers achieved higher scores on tests of cognitive ability during their school years. In comparison to controls, they were more academically successful even years after the program ended. Despite possible flaws in its study design, the Milwaukee Project still represents to many readers a very potent example of how an intensive cognitive intervention can have lasting effects on at-risk children. The Milwaukee Project produced such dramatic results that it attracted attention far beyond the community of academic psychologists and early childhood educators. Federal policymakers, including members of Congress, also considered the relevance of the project to government programs. Ultimately, the project prompted massive federal investment into the largest of all compensatory preschool programs, Head Start.
Head Start is a program that directs federal funds for the purpose of establishing preschool programs for children living in impoverished neighborhoods. Like the small-scale preschool interventions we have already considered, Head Start sprang from rising optimism during the 1960s that the government could combat poverty through targeted policies and programs. Head Start was a key element in President Lyndon Johnson’s vision of a Great Society in which poverty was battled with the financial backing of the US government and with the intellectual resources of American scientists and citizens. Since that time, despite a somewhat mixed reputation, Head Start has reached millions of children.
Head Start was offered to preschoolers during the year just prior to school entry. The hope was that enrichment at this crucial juncture would translate to more effective learning when participating children entered school. Researchers made comparisons with matched control children and particularly monitored their comparative progress during the elementary school grades. Initially, Head Start seemed to be working extraordinarily well. Participating children showed significant cognitive growth by the end of the project year, displaying a 9-point IQ advantage over matched controls as well as a better understanding of elementary math concepts and foundational reading skills.78 Unfortunately, the effects of Head Start faded rapidly during the first few years of school. While in first grade, Head Start children still displayed a measurable advantage over control children in their ability to learn, but by second and third grades those advantages were no longer detectable. The early gains of Head Start children had dissipated rapidly, an all-too-common pattern referred to as “fade-out.”79
Head Start’s disappointing findings of rapidly fading gains led to further analyses in which enduring boosts in cognitive functioning were documented. Field researchers found that the cognitive effects of Head Start varied somewhat by geography: Effects were comparatively large in urban communities in the southern states.80 Even so, the overall pattern held: Initial cognitive gains were modest and tended to taper off during the first few years of elementary school. For this reason, Head Start disappointed many observers who had hoped that participants would experience long-term benefits. At its inception, Head Start presented what seemed like a marvelous possibility: If at-risk preschool children were simply provided with high-quality environments—rich in linguistic interaction, nutritious food, and medical care—they might be launched on a stable upward trajectory through the school years. Instead, Head Start children were advantaged only temporarily. Differences between participants and nonparticipants reduced to zero after only a few years. The “fade-out” phenomenon quenched what was probably unwarranted optimism that a well-funded, large-scale program could redress the cognitive effects of poverty with one masterstroke.81 On the other hand, Head Start also taught a valuable lesson to its proponents: The preschool years are not necessarily more important than later years in the child’s development.
Part of the difficulty in judging the results from Head Start is that the program was not focused solely on children’s intellectual development. The goal of cognitive growth was only one of many program components. During the 1960s, other programs for preschoolers focused more directly on children’s intellectual growth and produced much more impressive results. We examined a few of these in the preceding paragraphs. When compared to Head Start, these so-called boutique programs were tiny in scale and entailed extraordinary levels of effort and commitment from all project participants—children, teachers, and parents. Several such programs resulted in sizeable and durable cognitive benefits to participating children.82
The earliest programs for enhancing intellectual skills trace back to roughly 1900 and to Alfred Binet himself. When Binet was tasked by the education leaders of belle époque Paris with finding a way to measure students’ intellectual capabilities, he was not deterministic: He did not believe that those capabilities were fixed, even for those children whose intellectual development was very significantly delayed. In a strong sign of faith in the plasticity of the human intellect, Binet designed a program of exercises that he called “mental orthopedics.” The comparison with physical orthopedics was direct. Binet proposed that “just as physical orthopedics correct a curvature of the thoracic spine, mental orthopedics straighten, cultivate, fortify such mental abilities as attention, memory, perception, judgment, the will.”83
Binet’s program of mental orthopedics consisted of a year of training exercises that were not tied to any particular academic subject; they could be practiced in the school or at home. The exercises included a memory development activity involving a board with objects glued to the surface. After viewing the board for five seconds, children were asked to name as many objects as they could remember. Another activity was intended to develop self-regulation. When “playing statue,” the teacher signaled the children to stop all activity and freeze. Periods of immobility were gradually increased from just a few seconds to one minute over the course of the training year. In addition to training memory and self-regulation, Binet’s program of mental orthopedics taught children observation skills, perception, imagination, analysis, and judgment. Binet reported that his program was effective: By his own account, after one year of participation the children’s school performance had advanced by two years.84
In recent decades, one of the most impressive intelligence-enhancing projects was a short-lived initiative called Project Intelligence.85 Conducted in Venezuela in the early 1980s, Project Intelligence was the brainchild of the government’s designated minister of intelligence, Luis Machado. Machado’s energies as a cabinet minister were focused on achieving an astonishing vision—raising the intelligence of Venezuelan youth. Machado had a plan. He secured the financial support of his government along with the assistance of some of the top cognitive scientists and intelligence theorists in the world, including scholars at Harvard and the Boston-based think tank BBN.86
Project Intelligence was based on curriculum units designed for seventh grade students. The intervention was rather modest because its design had significant practical constraints. In particular, it had to be accommodated within the normal school agenda of ongoing instruction in reading, mathematics, and the other school subjects. The curriculum units developed for Project Intelligence were studied three times per week for a total of 56 lessons over the course of the year. The lessons emphasized thinking skills, including reasoning, the analysis of language, problem solving, decision making, hypothesis generation, and inventive thinking.
The 460 participating seventh grade students made measurable IQ gains—as much to 6 IQ points depending on the test—by the end of the school year.87 In comparison to a control group of peers, they also demonstrated superior design and problem-solving skills. One project investigator, Richard Herrnstein, later a coauthor of The Bell Curve, summarized the effects of Project Intelligence by noting that “the change in behavior that resulted from this really rather modest intervention … was profound and dramatic.”88, 89 Unfortunately, Project Intelligence did not last long. A change in the Venezuelan government spelled the end of Mr. Machado’s term as minister of intelligence, along with his bold plan. Nonetheless, Project Intelligence remains a vivid historical precedent: It showed that a government can take the idea of learnable intelligence seriously and can allocate its resources to mount a high quality program to pursue a unique vision for its citizens: the enhancement of intelligence.
Several other programs have been designed to boost cognitive development among school-age children. One such program is Feuerstein’s Instrumental Enrichment. Conceived by Israeli psychologist Reuven Feuerstein, Instrumental Enrichment was born in the aftermath of World War II. Feuerstein, then a young psychologist working in Geneva, was recruited to test the children of Holocaust victims. The purpose was to understand the specific reasons traumatized children performed lower than expected on cognitive tests and what could be done to remediate them. When Feuerstein tested the children, he formed two powerful and lasting impressions that guided his subsequent work.90 The first was what he called dynamic assessment. Feuerstein noted that at the beginning of testing sessions the performance of children was lackluster but that during the course of testing performance began to improve. In fact, the children’s thinking seemed to become sharper while testing was under way. This implied that the object of testing, the psychological construct measured by the test, was not a static entity as traditionally assumed, but was instead a moving target that changed as assessment took place. This view of testing, which acknowledges that the child’s ability can shift while testing is in progress, is what Feuerstein called dynamic assessment.
Feuerstein’s second discovery while testing the Holocaust children was related to the first principle of dynamic assessment. Feuerstein knew that any learning that occurred during testing was not encouraged by the standardized testing conditions in which a child works alone in silence. Rather, dynamic gains arose most pointedly when Feuerstein interacted with the child during the testing session. Of course, Feuerstein did not tell the child the answers to the questions. The interaction was instead akin to sensitive coaching—focusing the child’s attention, encouraging persistence, and providing feedback. Feuerstein called this form of interaction “mediated learning.” The concept rejects learning as a strictly solo enterprise. Instead, it recognizes the powerful role of social interaction in learning, and in particular the tremendous importance of a skillful teacher in promoting the intellectual growth of a pupil. Though the idea is strongly reminiscent of Vygotsky’s theory, Feuerstein maintained that he developed the principle of mediated learning independently.
These two principles—dynamic assessment and mediated learning—are the cornerstones of Feuerstein’s program. Together they affirm that a child’s broad cognitive ability—his or her intelligence—is not a static entity but a malleable skill set, especially in the context of skilled social interactions with a capable teacher or parent. Even when children appear not to exhibit much growth in their intellectual powers, over time they can gain the capacity for growth. In the language of Feuerstein’s unmitigated optimism, even a child’s modifiability is modifiable.
In Feuerstein’s program, the full potential of cognitive modifiability cannot be realized with the use of assessments alone. The child must also be trained to develop intellectual skills, especially those skills that initially are weakest and those that pose barriers to further progress. Hence, another component of Feuerstein’s program is a curriculum of cognitive exercises called Instrumental Enrichment, which is incidentally also the name of his program. The exercises utilize a series of paper-and-pencil tasks to help students sharpen specific aspects of the intellect. The activities force participants to attend to detail, focus their attention over an extended period of time, and detect patterns in abstract materials. When they face complex problems, children are taught not to attempt random or haphazard solutions, but to seek solutions through systematic exploration. The exercises, engaged under the supervision of a skilled mediator, also teach children to persist when they encounter difficulty.
Reuven Feuerstein’s Instrumental Enrichment has been applied largely, though not exclusively, to students who have learning difficulties. Its primary purpose has not been to enhance intelligence in mainstream populations. Nonetheless, when applied to normative populations of teenagers and military recruits, Instrumental Enrichment has produced cognitive gains corresponding to about 9 IQ points.91 Some data also support Feuerstein’s hypothesis of divergent effects, which refers to the intriguing possibility that cognitive enhancements establish a new intellectual trajectory.92 The idea is that a powerful intellectual experience can send the learner on a long-term track of growth over time. If the hypothesis of divergent effects is correct, a cognitive advantage actually grows in successive years such that the trend lines of participants and nonparticipants continue to diverge even after the program ends.
Unlike Instrumental Enrichment, a project carried out in Croatia focused instead on gifted students. The curriculum, devised by a high school teacher and researcher named Radivoy Kvashchev, was designed to enhance creative problem-solving skills among high-ability high school students.93 Kvashchev found that his curriculum raised the measured intelligence of high school students the equivalent of 6 to 8 IQ points, with gains in both fluid and crystallized intelligence. The enhancements were found not only on tests; gains transferred to solving practical problems, such as how to feed a growing population.
A significant body of research on enhancing the intelligence of adults has been conducted among the elderly. This population focus is quite sensible: Older people are most at risk for declines in cognitive functioning and typically have more discretionary time to devote to activities that may enhance their intellectual abilities. One such intervention was the Cognitive Training Study, a component of the Seattle Longitudinal Study of cognitive development in adulthood.94 Investigators trained 228 participants, whose average age was 73 years, on spatial ability and inductive reasoning. The result was substantial improvement in intellectual functioning: Targeted abilities were restored to their cognitive performance levels of approximately a decade earlier.
The Cognitive Training Study produced improvements even among participants who earlier had experienced cognitive decline. The gains were long lasting, extending at least seven years out from the study period.95 In cases where the benefits of cognitive training were later lost, follow-up training sessions showed that, even 14 years after the original training, gains could be reestablished.96 Those retraining effects were found for both men and women, as well as in participants whose cognitive levels remained stable in the intervening years. The Training Study was subsequently replicated twice, with each replication involving close to 200 participants. The training studies showed that cognitive gains transferred to new measures of the targeted abilities—spatial ability and inductive reasoning—but not to cognitive abilities that were not trained.
A wide variety of projects designed to raise intelligence have been conceptualized and carried out, and many have obtained measurable success. Organized to promote cognitive functioning over the entire life span—from infancy to old age—these programs have produced positive effects that translated to real gains in IQ. The typical range of effects is on the order of 10 to 20 IQ points. Gains of this magnitude have significant consequences, especially at the so-called tails of a distribution. Imagine a bell-shaped curve that shifts upward one standard deviation—15 IQ points. With a shift this large, the number of people with an IQ of 120 or more would be multiplied several times. Equally significant, representation at the low end of the distribution, those with IQs of 80 or less, would also be reduced dramatically. A shift of 15 points, when applied to a large population, would greatly expand its collective capacity for complex, abstract, and creative thought. At the same time, it would reduce the numbers of people at risk for poor decision-making or inadequate coping skills that might predispose them to poverty, crime, or other unfavorable life circumstances.
The audacious and inspiring goal of raising human intelligence is not yet widely appreciated as achievable. With the exception of Head Start, cognitive enhancement programs remain obscure. Programs that demonstrate the feasibility of raising intelligence are generally known only by a small and somewhat insular community of psychologists. Because data from these programs are tucked away in technical journals and scholarly books, the imagination of the broader public has not yet been captured. For most people, the term “learnable intelligence” remains more nonsensical than inspiring—an oxymoron, a contradiction in terms. If the projects we reviewed were less obscure, a greater awareness of their successes could build momentum toward more effective interventions in the future. They could inspire a collective vision for a more intelligent society.
What are the limits to this vision? The documented effects of cognitive enhancement programs might suggest an upper bound on their power to increase intelligence. One way to interpret the results is that gains of 10 or 20 IQ points show the potential of early-stage programs. Future programs, with deeper theoretical grounding, could be more powerful. Historically, attempts to increase intelligence have been based primarily on educated guesses about the most important building blocks of intelligence. For example, many preschool programs were language-focused—a good bet for enhancing intelligent functioning. The centerpiece of the child’s experience was rich and extended linguistic interaction with adults as well as with other children. Other programs, such as Instrumental Enrichment, exercised children’s ability to stay focused even when tasks became difficult or confusing.
Although many programs produced sizeable cognitive gains, most were not based on a strong theory of which elements of intellectual life are both tied to measured intelligence and responsive to treatment. Programs built firmly around recent research could be much more effective than those of the past. That is because recent research has revealed more clearly the key components of intelligent cognition. Not only are the building blocks of human intelligence better understood than ever before, research has shown that many of those components can improve substantially with practice. In the language of psychologists, the components of intelligence are trainable.
The historical legacy of cognitive interventions is that the term “learnable intelligence” is meaningful. The human intellect can be enhanced through well-designed programs, and the resulting heightened capability applies broadly across intellectual pursuits. Intelligence truly can be upgraded, and not just within a narrow collection of skills or for the short term. The effects of interventions have sometimes extended to intellectual achievements even years after the program officially ended. Recent research adds to the legacy of successful programs by giving us even better ideas on how to raise measured intelligence. Findings include the exciting discovery that key parameters of the mind—especially working memory capacity—once presumed to be unalterable are, in fact, not. We now know, as psychologist Richard Mayer puts it, that “intellectual ability can be taught.”97
Learnable intelligence is not a baseless fantasy. Cognitive enhancement programs constitute credible proof that the deliberate enhancement of intelligence is not just possible: It has the potential to transform society with wide-open consequences for individual well-being and collective prosperity. Where does that put us? The first systematic attempts to raise intelligence were carried out in the 1960s. Now, half a century later, we have the basis for much more powerful interventions. Given the extended record of data, theory, and experience on how to teach intelligence, we face an important choice. Will we use that base of experience to launch a sustained effort, guided by research, to raise the intelligence of the human population? Do we have the imagination and the will to pursue higher levels of intelligence in ourselves and among our fellow citizens?