chapter seven - AAAS · CHAPTER 7: HOW IS INTELLECT MOLDED BY GENES AND ENVIRONMENTS? 99 ......

chapter seven

HOW IS INTELLECT MOLDED BY GENES AND ENVIRONMENTS? 77

■ ■ ■ Mr. Huang, a puzzled patriarch Last night, Mr. and Mrs. Huang were introduced to their future daughter-in-law, Vivian

Lee. They had been eager to meet the woman whom their son Frank had met in med-

ical school. Frank had told them that Vivian graduated first in her class at a private

high school and held a biology degree from Harvard. He also had said that Vivian was

from a well-to-do family: her father is a banker and her mother a university professor.

The Huangs liked what they heard about Vivian because they both come from high-

achieving families. However, when they met her they had a bit of a shock.

The Huangs, who are of Chinese descent, had assumed that Vivian was Chinese, too.

This had been a realistic assumption because Frank had always dated Chinese girls

before and he knew that his parents wanted him to marry within their culture.

Besides, Vivian’s surname is Lee.

Last night, one look at the girl told them she was different from what they had expected. It turns out her father

is a descendent of Irish immigrants. Her mother is a descendent of Kikiyu Kenyans. The Huangs quickly overcame

their surprise and greeted Vivian warmly. They enjoyed their evening with Frank and his fiancée, and of course

have no intention of undermining the couple’s engagement.

But reading the newspaper at breakfast this morning, Mr. Huang wonders about the direction in which his family

tree is bending. The paper contains an article about research into intelligence. Among other things, the article

reported that both White Americans and Black Americans fare less well on IQ tests, on average, than Asian Americans.

For a brief moment Mr. Huang wonders if his future grandchildren will be less intelligent than any children

Frank would have with a Chinese woman. He immediately dismisses this notion as ridiculous speculation and

feels bad for even thinking it.

Focusing back on the facts, Mr. Huang reminds himself that average IQ scores by race do not tell much about

how individuals achieve, particularly individuals like his grandchildren who aren’t even born yet. What might be

more indicative, Mr. Huang ponders, is the fact that Vivian herself is highly intelligent. She has had a good

upbringing and the best possible education. She and Frank are in an excellent position to raise their own

children into intelligent, successful adults. Those children will have a mix of European, African, and Asian

ancestry, but so what? What possible effect could that have on inborn brainpower?

Defining intelligence Smart. Quick-witted. Sharp. Clever.Astute. The English language has manywords that describe intelligence, whichbriefly defined is the ability to absorb,process, recollect, and apply information.Cognition is the word used by scientiststo specifically refer to the process ofthinking.

Certainly one reason why there are somany ways to describe intelligence is thatthe trait involves so many applications ofthe mind. Intelligence has to do with theability to reason, to think abstractly, todraw conclusions, to solve problems, tolearn from experience, and to rememberwhat has been learned. It also has to dowith how the mind makes use of sensory

information and how it takes cues fromemotions (our own and others’). Thusintelligence is a general capacity of thebrain.

Intelligence stands on its own as a trait,but it also permeates all other aspects ofpersonality. Sense of humor, risk-takingtendencies, and sexual attractiveness arejust three examples of the many humantraits that relate to intelligence.

Intelligence very clearly correlates withsuccesses in school and work settings.Somewhat surprisingly, intelligence alsocorrelates with overall success in living.Less delinquency, greater wealth, bettercoping skills, higher income, betterhealth, fewer accidental deaths, andlongevity are all associated with higherintelligence, according to research.“Bright people have a tail wind in virtu-ally all aspects of life,” claims oneresearcher. 1

History of intelligence testing Intelligence has been more intenselystudied than any other behavioral trait.Intelligence tests, first developed in the19th century, have been so heavilyresearched, refined, and applied that theones used today are fair predictors(though by no means complete predic-tors) of academic and professionalachievement. Many scientists believethem to be “among the most accurate (in technical terms, reliable and valid) ofall psychological tests and assessments.” 2

9 8 BEHAVIORAL GENETICS

Some researchers believe that“intelligence” acts like a wind in the sails, pushing one towardsuccess in all aspects of life.Whether intelligence actually has that effect, and why, requiresfurther research.

CHAPTER 7: HOW IS INTELLECT MOLDED BY GENES AND ENVIRONMENTS? 9 9

Francis Galton is the father of thegenetically based research into intelli-gence. His 1869 book Hereditary Geniusis hailed as “the first quantitative analysisof human mental ability.” 3 Galton pro-posed that human mental ability is aninherited trait and that individual“genius” (as he called it) varies along acontinuum from very dull to very bright;the percentage of the population at anypoint on the continuum decreases as theextremes are approached. Galton wasdescribing a normal curve of distribution,popularly known as the bell curve. Hisidea about a range for human intelligencewas novel at a time when it was gener-ally believed most people had the sameintelligence, with just a few being excep-tionally slow- or quick-witted.

To explore the connection betweenintelligence and inheritance, Galtonselected for study men of high stature in their respective profession – the law,literature, science, the military, and soforth. He assumed that high achievementin a field correlated with high mentalability. He then observed how frequentlythe first, second, and third-degree rela-tives of his selected subjects alsoachieved high standing in their fields. He compared these frequencies to thatwhich could be expected based on anormal curve of distribution for intelli-gence. Galton discovered that highachievers did in fact cluster in certainfamilies, and concluded that genius wastherefore hereditary.

Inspired by Galton’s theories, other scientists began to devise tests for intelli-gence. One of these was the French psy-chologist, Alfred Binet. By trial-and-errorexperiments with schoolchildren, bothnormal and mentally handicapped, Binetdeduced what the average child couldaccomplish at each age level. Binetteamed up with another psychologist,Theodore Simon, to create a test used byFrench school administrators to identifychildren less able to profit from instruc-tion in regular school classrooms.

The Binet-Simon scale included 30tasks such as shaking hands, findingrhymes for words, repeating back asequence of random digits, andanswering logic questions. Performanceon the tests was evaluated to determinemental age. According to the Binet-Simon scale, a child had the mental age

An early researcher into intelligence, Francis Galton, proposed that eminence in a profession was an indication ofmental ability; when he was ableto show that certain families hadmore than a normal number ofeminent members, he concludedthat mental abilities run in families. At the time (1869), this was a radical idea.

1 0 0 BEHAVIORAL GENETICS

of 12 if he or she could accomplish thetasks that matched the abilities of theaverage 12-year-old.

In the early 20th century, a Germanpsychologist named Wilhelm Stern cameup with the idea of stating intelligence interms of the ratio of mental to chronolog-ical age. The popular label for this ratio,intelligence quotient or IQ, was coinedlater by Lewis Terman, an American psychologist. Terman was a member of acommittee assigned with the task ofdeveloping intelligence tests that could beused to help sort the large number ofWorld War I inductees into appropriateassignments.

Intelligence tests gained in popularityafter that war. The military continued touse them for recruitment and placementpurposes, and workplaces and schoolsstarted to use them, too. Social scientists

conducted IQ tests to support theoriesabout which types of people were moreor less intelligent. At this point inAmerican history, slavery had been abol-ished less than a hundred years earlierand waves of new immigrants werecoming into the country. Much was madeover the weak IQ performance ofAmerican Blacks and of immigrants fromEastern Europe.

In the latter part of the 20th century,controversy arose over whether intelli-gence tests were biased. Some questionedGalton’s original work by pointing outthat high professional achievement mayhave less to do with inborn intelligencethan with having the means for educationand social connections. Other criticsnoted that performance on IQ tests mostaccurately reflects whether you can readthe language of the test and are familiar

In the early twentieth century,laws restricted immigration fromEastern Europe, based in part onthe false belief that such peoplewould dilute the intelligence ofthe American population.

with facts from the culture of those whocreate these tests. Critics also claimedthat scientists were interpreting data inways that supported their pre-existingracial and cultural prejudices.

Advocates took some of these concernsto heart and sought to create less biased,more accurate tests. Many tests havebeen developed, each using a differentselection of tasks to measure intelligencein its various manifestations: mastery ofvocabulary; text comprehension; mathskills; memory; visual acumen; under-standing of universal concepts such asup/down and in/out; grasp of generalknowledge; speed of response; applica-tion of logic; and more. These tests continue to have an enormous role inmodern society. This is particularly true ineducation, as attested to by the wide-spread use of the SAT and other tests thatmeasure scholastic aptitude.

Today, performance on IQ tests is con-sidered a reasonable predictor of success.IQ scores correlate reasonably well,though imperfectly, with academic andprofessional achievement and they corre-late to a lesser extent, though more thanrandomly, with success in social and personal aspects of life.

Measuring Many intelligence researchers accept thetheory that there is some underlyingability to learn that feeds into all cognitiveperformance. This ability is called g.The term stands for general cognitive

ability. The concept of g was first articu-lated more than 100 years ago by CharlesSpearman, a British psychologist, and itrefers to a basic, core ability to make useof information. Some researchers refer tog by two other eponymous terms, “globalfactor” and “general intelligence factor.”Scientists debate whether g is an actualproperty of the brain such as some neuro-logical mechanism at work (the minorityview) or merely an abstract property like“horsepower” (the majority view).

The concept of g derives from theobservation that any given individualtends to perform at about the same levelon a diverse range of intelligence tests.Whether you give someone a math test ora vocabulary test, an oral test or a pen-and-paper test, a test with words or a testwith pictures, a timed test or an untimedtest: across all types of tests, most people

1 0 1CHAPTER 7: HOW IS INTELLECT MOLDED BY GENES AND ENVIRONMENTS?

Employers, the armed forces, colleges, and other importantsocial institutions make use ofintelligence and aptitude tests to screen and place applicants.

g

score somewhat consistently.A statistical procedure called factor

analysis is applied to scores on a variety oftests, weighting each differentlyaccording to the complexity of the test’scognitive demands, to arrive at a numberfor this pattern of consistency. Thatnumber is called g, and it is roughlyequivalent to IQ, but is consideredslightly more pure. This is because thepooling of data from many types of testsdilutes the possibility that what is beingmeasured is a specific set of learned skillsrather than a basic aptitude.

Across IQ tests, g accounts for about 40 percent of the variance in perform-ance. Within any single test, most of thevariance is independent of g. The signifi-cance of that anomaly is that g is not theonly factor involved in cognition.

Some researchers have added a refine-

ment to the theory of g. They suggest thatg is comprised of distinct subsets of cognitive ability including fluid intelli-gence (abstract reasoning), crystallizedintelligence (vocabulary and generalknowledge), visual-spatial ability,memory, and speed of processing. Thistheory is also based on factor analysis: anindividual’s scores on certain types ofintelligence tests tend to correlate morehighly compared to the correlation of thatindividual’s scores on all types of tests.

An alternative to the notion of g is theidea that intelligence spans many different human abilities. One suchtheory of multiple intelligences (proposedby Howard Gardner in 1983) holds thathumans have eight forms of intelligence:linguistic, logical-mathematical, spatial,kinesthetic, musical, intrapersonal, inter-personal, and naturalist. Other similar


Researchers have observed that some people apply mathe-matical calculations and otherproblem-solving skills in real-lifesituations much better than theydo on IQ tests.

theories propose different cognitivedomains. The multiple intelligencestheory raises the possibility that measure-ments of g capture some, but not all,aspects of intelligence, and that it may notbe possible to comprehensively measurethe trait in full.

Other researchers are highly critical ofthe whole concept. Some argue that theconstruct of g is of no value in the searchfor the underlying neurobiologicalprocesses that contribute to cognition.One scientist has decried “the unreality ofg and the fallacy of regarding intelligenceas a single-scaled, innate thing in thehead.” 4 Critics point out that if you alterthe underlying assumptions used in factoranalysis, no single number g emerges.They believe that the positive correlationsin an individual’s performance on variousIQ tests do not have to do with anyunderlying factor, but rather simply reflectthe individual’s life circumstances. They assert that someone who is well fed,adequately sheltered, educated, andexpected by family and peers to achieve,will do well on most tasks set before himor her. Someone in opposite circum-stances will not do as well, across the board.

Some suggest that g only reflects test-taking skills, not intelligence as it is usedin real life. For example, the social anthro-pologist Jean Lave has observed that poorfavela children in Rio de Janeiro easily doarithmetic when selling wares on thestreet but perform miserably when asked

to add and subtract in a classroom. By thesame token, Liberian tailors are able tomake precise arithmetic calculationswhen measuring customers but cannotdo the same when tested.

This idea of multiple intelligences wasfirst raised in the 1950s, about half a cen-tury after g was proposed. It has not beenas extensively applied and researched asg. Indeed, among many scientistsresearching intelligence today, g is consid-ered beyond dispute. g is the core piece ofdata used in analyses that compare theintelligence of individuals and groups. If the critics are correct, however — if there is no g — then many of the con-clusions drawn from intelligence researchwill have to be revisited. The reader isadvised to keep this in mind whilereading the next sections of this chapter.

Quantitative research intointelligenceA great many quantitative studies haveshown that IQ correlation increases asgenetic similarity increases. Quantitativestudies also provide evidence of environ-mental influences on intelligence:• In studies of identical twins raised

together, IQ scores do not correlatecompletely. You would expect near-per-fect correlation between genetic equalsif IQ were solely hereditary.

• In studies comparing blood relativesraised together to those raised sepa-rately, the IQ scores of the former tend


is the ability to learn thatunderlies all cognitive perform-ance. Some researchers areseeking the genetic contributorsto , while others questionwhether is actually a real oruseful concept.

g

g,g

to be more similar than those of thelatter.

• In studies of family members notrelated by blood (for example, parentsand their adopted children), there is modest, but statistically significant,correlations of IQ scores.Most quantitative studies of intelli-

gence measuring modern Western popu-lations have produced heritability

estimates ranging broadly around 0.50.Taken together, they suggest that geneticsand environment are roughly equivalentin their influence on the variation in IQ.

The relative contributions of the sharedand the non-shared environment alsohave been partitioned out in heritabilityestimates, and this had led to one sur-prising bit of data. Compared to mostother behavioral traits, intelligence inchildhood seems to be more significantlyshaped by the shared environment com-pared to the nonshared environment.This could possibly mean that the homeenvironment has a greater influence on achild’s cognitive development than onother characteristics, such as personality.

Researchers are interested in how thecontributions of genes and environmentchange over time. Do inherited character-istics become more important as one getsolder, or does the influence of the environment become more important?Most studies of IQ kinship correlationsshow that genetic influence grows andshared environmental influence fallsaway almost completely.

One possibility is that heritabilityincreases with age because of genotype-environment correlation. That is, aspeople grow up they are able to exertmore control over their experiences. Forexample, children with higher innateintelligence might choose to engage inactivities that stimulate their intellectsuch as pursuing formal education,reading, spending time in stimulating


Children with strong early intelli-gence may tend to select environ-ments that nurture and supportthe development of their intelli-gence. Thus their genetic tendency would be reinforced by their environment throughdevelopment.

conversation, and engaging in work thatchallenges the mind. By contrast, children with lower innate intelligencemight gravitate towards non-intellectualactivities. Thus, genetic tendencies wouldbe reinforced over time.

The opposite possibility is that theeffects of the environment becomestronger as one gets older. Under this sce-nario, children who grow up in stimu-lating homes and who attend qualitypreschools would follow a higher intellec-tual trajectory than children whose earlyyears are less supportive. They would getbetter grades in school, have a betterchance of making it to college and havemore motivation to enroll there, have theacademic credentials to obtain intellectu-ally-challenging jobs, etc. Thus, the effectsof early environment would become mag-nified over time.

Another possibility is that the relativecontributions of genes and environmentremain stable over the lifetime. In 1997,data supporting this hypothesis emergedfrom a study of 240 pairs of Swedishtwins in their eighties. In this study, theheritability of general cognitive abilitywas computed at 0.62.5 This is relativelyclose to the heritability estimates fromstudies of younger subjects and thus sup-ports the proposition that the heritabilityof IQ does not shift significantly with age.More research is needed before the ques-tion of how intelligence is affected bygenes relative to environment over timecan be answered.

Molecular research into intelligenceAs the careful reader should realize bynow, twin studies tell us nothing specificabout the genetic and environmental fac-tors underlying intelligence. Molecularresearch seeks to bridge this gap, at leastin terms of identifying genetic contributorsto intelligence; in turn this may aid in thesearch for environmental contributors.

The operating assumption is that cogni-tion is a complex trait with many genesinvolved, each of relatively small effect.There are exceptions to this general rule.As we described in earlier chapters, somepeople acquire damage to their brains, significantly affecting their intelligence,from disorders such as PKU, Fragile X,and early-onset Alzheimer’s. Each of these


Well-nurtured and stimulated children tend to perform better on IQ tests than children who lacksuch benefits. This is evidence forthe environmental contribution tointelligence.

disorders stem from problem alleles in asingle gene.

The general rule, however, is that cog-nitive abilities are shaped by a multitudeof genetic and environmental factors thatinterweave throughout development. It isestimated that any single gene accountsfor at best 1 to 2 percent of the variationin intelligence; discrete environmentalfactors likely make similarly modest con-tributions. This means that highly intelli-gent people are blessed not with onebrainy gene, but rather with an abun-dance of “positive alleles for high g,” assome researchers refer to them.6 Theymay also have experienced favorablenutritional and nurturing conditionsduring their development and/or beenraised in an environment conducive tointellectual growth. Persons with lowerintellect may possess an inordinateamount of “negative alleles for high g,”may have experienced unfavorable condi-tions during their development, and/ormay have been raised in intellectuallyunsupportive environments.

Genes and environment interactingthrough processes of development form atriple whammy: anyone who comes upfavorably in all three categories has astrong likelihood of high g. Anyone whocomes up short across the board has astrong likelihood of low g. For those witha middling assortment, the likely rangefor g is much broader.

To focus back on just the genetic con-tributors to intelligence: one way to

imagine the variety of genes involved in atrait like intelligence is to picture the different products involved in the struc-ture of a house. Concrete, plaster, wood,plastic, nails, glass, metal — each serves adifferent purpose, yet contributes to theoverall structure. Endophenotypes are theequivalents of the products that make ahouse. They are the intermediate traits,the underlying processes that contributeto a phenotype. For cognition, theseendophenotypes would include suchthings as speed of information processing,capacity of working memory, synapsedexterity, receptor functionality, and agreat many more elements of brainactivity. (A diagram illustrating endophe-notypes appears on page 86.)

To use another metaphor, just as tinyrivulets running down a mountain jointogether into a mighty stream, each rele-vant gene contributes to the causal

pathway by which intelligence is formed.For genetic researchers, the concept ofendophenotypes narrows and makesmore manageable the search for relevantgenes. Using genomic scans or associationstudies, they can look for correlationsbetween tiny but discrete aspects of brainfunction and specific alleles of singlegenes. This is how several candidates for“IQ genes” have been identified.

The quest to find the individual genesthat contribute in some small but measur-able way to intelligence bring to mind anold Chinese proverb: “In vain do menseek the source of great rivers.” Despite


Endophenotypes are the underlying processes to a trait. For example, for the trait of intelligence, endophenotypesmight include speed of informa-tion processing, capacity ofworking memory, and other traits.A useful analogy can be found inthe component parts of home construction such as bricks,lumber, and tools.

formidable obstacles, some researchers arepursuing this quest. Hundreds of molec-ular studies have been conducted, andseveral genes contributing to cognitivefunction have been tentatively identifiedin humans, mice, and even in fruit flies.

One such gene achieved some promi-nence in a 1998 study. Named IGF2R,the gene is on Chromosome 6 and is anacronym for “insulin-like growth factor-2receptor.” Insulin-like growth factors arehormones that affect the work of thenervous system. In this particular study,researchers conducted scans of a segmentof Chromosome 6 in high-IQ children anda control group of children. A tiny but sig-nificant proportion of the high-IQ chil-dren were found to have a particularallele for IGF2R, Allele 5, compared tothe control group children. Allele 5 alsowas found in slightly greater proportion intests on children with extremely highIQs, high mathematical ability, and highverbal ability. 7

A follow-up study by the sameresearchers failed to confirm the hypoth-esis that the IGF2R gene is associatedwith cognition.8 Other genes proposed ascandidate “IQ genes” have not been confirmed, either. But the point of thisresearch is not only to discover the par-ticular genes that help shape intelligence.It also is to understand how intelligence isaffected by an overall biological/environ-mental/developmental process in whichgenes play a part. For example, there isspeculation that insulin may play a role in

learning and memory. Investigations intoIGF2R, whether or not they confirm thisparticular gene as having anything to dowith intelligence, may shed light on howinsulin figures in cognition. Following thetrail of other implicated genes may offernew insights into how intelligenceemerges in human beings.

Predicting individual intelligenceMolecular research sheds light on geneticinputs, but not on specific individuals.IGF2R gene researchers made this verypoint in their published results. Theyspecifically noted that more than half ofthe high-IQ children did not have Allele 5for IGF2R, while almost a quarter of thecontrol-group children did. When and ifthe gene IGF2R is confirmed as having arole in IQ, it will not be a signpost forwho’s bright and who’s not. It may, how-ever, be recognized as an intelligence sus-ceptibility gene.

Suppose molecular researcherssomeday identify the scores of genes thatweave together into cognition. They fur-ther identify the “positive” and “nega-tive” alleles for each gene. Then they doa genome scan of an individual to countthe number of “positive alleles” present.They still will not be able to state withcertitude anything about that person’sintellectual abilities. Scientists might beable to offer a likely range for that indi-vidual’s intelligence, but it would be a


Several genes have been investigated as being potentiallyinvolved in intelligence, includingone called IGF2R, located onChromosome 6. None of the can-didate genes has been confirmed.In any event, there is certain tobe a great number of genesinvolved.

guess all the same. This is because genesare probabilistic, not deterministic, for thereason that should be obvious by now tothe reader — they operate within envi-ronments over time.

Eyesight makes a useful example here.Presume that nearsightedness is a genetictrait. If you inherit your father’s nearsight-edness, a pair of eyeglasses can correctyour vision to 20:20. In the same way,

supportive environments may be able tobuild intellect even in individualsbequeathed with a relatively poor set ofgenetic material. Some researchers specu-late that the environment during prenataland early childhood development are particularly important factors for intelli-gence. A child who receives adequatenutrition in the womb, is nurtured andstimulated as a baby, and is exposed toearly education can obtain importantboosts to cognition.

One psychologist asserts, “We havedemonstrated that intellectual skills oftenbelieved to be innate are extremely sensi-tive to the environment.” 9 This psycholo-gist has conducted research thatproduced correlations between the verbalstimulation provided by parents and thevocabulary and grammatical dexterityacquired by children. Other researchershave found correlations between manyother specific environmental conditionsand intelligence.10 But the key word hereis correlation. Researchers have not yetfound a way to move beyond correlationto identify discrete environmental equiva-lents of genes that mold behavior.

Even if every genetic and environ-mental input were identified, predictionwill never be perfectible. To understandwhy, think about weather forecasting.Meteorologists today make use of a greatdeal of data concerning atmosphere, terrain, precipitation, temperature, oceancurrents, and all the other discrete vari-ables that coalesce into rain on this side


Even if you are born with weakeyesight, your vision can get close to perfect using correctiveprocedures and lenses. In thesame way, even if you are bornwith “weak” intelligence, environ-mental interventions can boostyour brainpower dramatically.

of the mountain and sunshine on theother. As a result, weather prediction ismore accurate than it used to be, butbecause of the nonsystematic way thesevariables interact it may never be exact.Similarly, as scientists learn more aboutthe processes by which genes and envi-ronment interact over time, they mightget better at predicting individual intelli-gence. Yet, because the inputs are nonsystematic, such predictions willnever be precise and sometimes they willbe really wrong.

Mr. Huang’s speculations Many studies have analyzed academicachievement by race. This research maybe of interest to Mr. Huang, who is won-dering about the cognitive prospects offuture children born to his son, who is ofAsian descent, and his future daughter-in-law, who has European and Africanancestry.

There is reason to believe that race andintelligence are not related biologically.Important data on this point comes fromThe National Collaborative PerinatalProject, a study that began in 1959. Some48,000 pregnant women were enrolledfrom several major U.S. cities. A roughlyequal number of Black and White partici-pants were involved. Importantly, thesocioeconomic status (SES) of the subjectscovered a broad span. SES refers to one’sposition within a hierarchical social struc-ture, and it is determined by several vari-

ables including occupation, education,income, wealth, and place of residence.The original intention of this study was tolearn about cerebral palsy and otherhealth complications from childbirth, notintelligence. However, many thousands ofpieces of data were collected on the sub-jects, including IQ scores for the morethan 59,000 children born to thesemothers, obtained at ages four and seven.According to this data, when sorted bySES there is very little difference in IQbetween Black and White children.11

Despite the significance of this finding,it has rarely been cited in scholarly paperson intelligence. What has gained moreprominence is other IQ testing data col-lected before and since (mostly from sub-jects in the upper tiers of SES) and pooledtogether to draw bell curves for the popu-lation as a whole as well as for varioussubgroups.


Black children given IQ tests scoresimilarly to White children of thesame socioeconomic status.

Looking at these bell curves, Mr. Huang might be relieved to learn thatpeople of all races and ethnicities can befound at every IQ level, including thehigh end of the curve. On the other hand,Mr. Huang might be disturbed to learnthat the bell curve for White Americans iscentered somewhat lower compared toEast Asians and Jews. The bell curve forBlack Americans and Hispanics sits evenlower. Mr. Huang also might be troubledby U.S. Census Bureau data indicatingthat, compared to Asian Americans, asmaller percentage of White and BlackAmericans complete high school, havecollege degrees, and are in professional ormanagerial jobs.

Whites do better than Asians on somemeasures; for example, their unemploy-ment rate is lower and their home own-ership rate is higher. But Blacks do worsethan both Asians and Whites on severalmeasures. For example, even whenBlacks and Whites achieve equally wellon standardized tests such as the SAT, the former are outperformed by the latteracademically. Grade point average islower, they drop out of school at higherrates, and the time to graduation is longer.This has been called the over-predictionphenomenon: for Blacks, performance onstandardized tests over-predicts their aca-demic achievement compared to Whiteswith the same score. In other words,even when Blacks are as well preparedacademically as Whites, they do not perform as well academically.

Accounting for disparities in population IQsThe discrepancies described above havebeen used by some people to argue thatthere is an intrinsic, biological differencebetween people of different racial andethnic groups. “Most experts believe thatenvironment is important in pushing thebell curves apart, but that genetics couldbe involved too,” according to an essayon intelligence signed by a large numberof researchers in the field and publishedin the Wall Street Journal in 1994.12

There are many possible explanationsfor how the environment pushes the bellcurves apart. It is speculated that theAsian culture’s emphasis on achievementcould account, in part, for the compara-tively higher I.Q. performance of AsianAmericans. It should be noted here that“Asian American” is a very broad termthat encompasses Asian Indian,Bangladeshi, Cambodian, Chinese,Filipino, Hmong, Indonesian, Japanese,Korean, Laotian, Pakistani, Thai,Taiwanese, Vietnamese, and others.


Education is highly revered inChinese and Japanese cultures.This could explain in part whyChinese-American and Japanese-American children, as groups,tend to do well in school.

Of these groups, the “model minority”label best fits Chinese and JapaneseAmericans because they rank so highly onvarious measures of achievement.Different immigration experiences couldexplain the much lower levels of achieve-ment for other Asian American popula-tions. Similarly, the immigration strugglesof Hispanic Americans might account fortheir lagging IQ scores relative to Whitesand Asians.

The poverty, segregation, and discrimi-nation that African Americans have his-torically experienced since slavery daysand continue to experience couldaccount for their relatively lower achieve-ment, as a group, on IQ. Another recenttheory is that African Americans performless well academically because they suc-cumb to stereotypes about themselves. A number of studies have demonstratedthat people’s performances suffer whenthey know they are being watched, whenthey know that their performance has aconsequence (for example, that theymight lose a prize), or when they areaware that they are different from othersin the test group (for example, if they arethe only female being tested in a group ofmales). Therefore, it could be that themere fact of being African American in aculture that stereotypes Blacks as inferiormay be sufficiently disturbing to impairperformance.

Indeed, this is the finding of a ground-breaking study published in 1995. In thisexperiment, groups of college students

were given a section of the verbal part ofthe GRE test. Some of the students weretold that the test was to measure theirability, while others were told that thetest was simply a problem-solving task.Interestingly, Black students underper-formed White students only in the groupsthat had been told their ability was beingmeasured. The researchers ascribed thisphenomenon to “stereotype threat.” In their words,

…whenever African American students perform an explicitlyscholastic or intellectual task, theyface the threat of confirming orbeing judged by a negative societalstereotype — a suspicion — abouttheir group’s intellectual ability and


Research suggests that academicperformance suffers when peoplefeel they are being watched orjudged or when they feel some-thing important rides on theresults.

competence. This threat is notborne by people not stereotyped inthis way. And the self-threat it causes– through a variety of mechanisms— may interfere with the intellectualfunctioning of these students, partic-ularly during standardized tests.13

The researchers also argue that long-termeffects on Blacks of stereotype threat maybe for them to reject the ideal of academicachievement, to lose interest in achieve-ment, to put less effort into it, and as aresult, to achieve less.

Class (which closely parallels race inthis country) could be another environ-mental factor that pushes the bell curvesapart. A 2003 study made fresh use oftwin data from the National CollaborativePerinatal Project (which, as we men-tioned previously, had been underappreci-ated). The NCPP included so manymothers and their children that it con-tained data on more than 600 twin pairs.NCCP researchers were able to locateabout half of those pairs for IQ testing at

age seven. The new study used the NCCPtwin and SES data along with advancedmodel fitting techniques to parse out therelative contributions of genetics andenvironment. The researchers concludedthat genetic influence varies dependingon socioeconomic status.

According to the abstract (summary) ofthe study, “In impoverished families, 60 percent of the variance in IQ isaccounted for by the shared environment,and the contribution of genes is close tozero; in affluent families, the result isalmost exactly the reverse.” The implica-tion is that the quality of the environmenthas a swamping effect; impoverished con-ditions tend to suppress genetic potential,while enriched conditions tend to allowgenetic potential to flourish. The studyconcluded, “Although there is much thatremains to be understood, our study andthe ones that have preceded it havebegun to converge on the hypothesis thatthe developmental forces at work in poorenvironments are qualitatively differentfrom those at work in adequate ones.”14

How would genetics push the bell curves apart? It would mean that“positive alleles for g” are more prepon-derant in one racial group than another.Such differences among geographicallydispersed human populations could con-ceivably have occurred through the forcesof natural selection.

Since environments and genetics worktogether over development, the humangroup that achieves best on intellectual


Based on twin studies,researchers have suggested that the environment may have a greater impact on the develop-ment of intelligence for childrenraised in impoverished environ-ments compared to those raisedin rich environments.

scales would be the one that has advan-tage in both environmental and geneticdepartments. And so while one couldspeculate that Asian Americans have agreater assortment of “positive alleles forg” than any other racial or ethnic group,alternate hypotheses also could be pro-posed. For example, it could be that AsianAmericans are not more gifted geneticallythan other groups, but they are advan-taged environmentally. Or it could be thatAfrican Americans are so disadvantagedenvironmentally that their equivalent oreven superior genetic gifts cannot makeup the difference.

But this is all speculation. The difficultyin researching racial differences in cogni-tion lies in the fact that racial identifica-tions are less a biological fact than a socialconstruct. Population genetics is thebranch of science concerned with genefrequencies — the prevalence of alleles inpopulations. As any population geneticistwill tell you, there is more genetic diver-sity between individuals of one popula-tion than there is between populationgroups. A White person from Denmarkmay be more similar genetically to a Blackperson from Sudan than to another WhiteDane. Your “race” (as determined by yourown choice or as assigned by others)yields less data useful to behaviorresearch than specific information aboutyour particular biological ancestors.

Yet considered on a statistical basis,some alleles do occur more frequently insome racial and ethnic populations.

As a well-known example, the alleles thatlead to the blood disorders of sickle cell

disease and thalassemia appear more fre-quently in African and Mediterraneanpopulations, respectively. Such facts jus-tify the quest pursued by someresearchers to discover the “IQ genes”that would explain why some racial andethnic groups dominate in intelligencetesting and achievement.

Researchers might be able to obtainusable data on the race and intelligencequestion through molecular techniques.For example, it may be possible to deter-mine ancestry of subjects through DNAanalysis, give them intelligence tests (thatare not culturally biased), and see howthose whose lineage traces back to Asiaperform on these tests compared to thosewhose ancestors come from Europe orAfrica.

In the final analysis, though, theanswers will have to do with populations,not individuals. Mr. Huang cares notabout how Asian Americans, WhiteAmericans, or Black Americans fare asgroups in terms of intelligence. He caresabout the potential of his individual,mixed-race American grandchildren.


It is possible for two people of different races to be more relatedgenetically to each other thanthey are to some people fromtheir own respective race.

Perhaps some day, when scores of sus-ceptibility genes for IQ have been identi-fied, microarray analysis could determinethe number of alleles associated with highcognitive capabilities that are possessedby Mr. Huang’s descendents. That is notyet possible. Even if it were, the datawould produce only a measure of prob-able potential. What happens over timewith that potential — whether it is maxi-mized or wasted — remains outsideanyone’s full control.

Eugenic concernsIn an ideal world, Mr. Huang wouldnever even entertain the notion that agrandchild with a superficial racial differ-ence might have inferior intelligence. Butthe world in which Mr. Huang lives is notideal. There is a long history of humansseeking to find distinctions betweengroups in order to declare that one isbetter than another.

Francis Galton, cited at the beginningof this chapter as the father of moderngenetics, is also the fellow who coinedthe term eugenics. It comes from theGreek for “well born,” and stands for theidea that human races can be improvedthrough selective breeding. Tied to thisphilosophy is the concern that humanitywill degenerate without deliberate inter-vention. Conceived in noble ideas aboutthe betterment of mankind, in practicetwentieth century eugenics was base andshameless.

Eugenics, nurtured in America andother western countries, led to the aggres-sive promotion of childbearing amongeducated White women. This is anexample of positive eugenics. It also led tothe institutionalization of the “unfit” toprevent their procreation and to theextensive use of sterilization on thosedeclared feeble-minded or insane. Theseare two examples of negative eugenics.Sterilization victims often were poorand/or non-White and in most cases con-sent was not obtained. In the U.S., morethan 40,000 eugenic sterilizations wereconducted in thirty states between 1907and 1944.

Eugenic ideas were adopted to hideouseffect by the Nazi government inGermany, which sterilized hundreds ofthousands and killed millions on thegrounds that they were diseased, of lowintelligence, or politically troublesome.They targeted enemies, primarily theJews, and used eugenic arguments to justify their policies.

After World War II, eugenics fell intodisfavor, though some policies such assterilization remained in effect in a fewcountries for many more years — in theU.S., through the 1970s. Yet the basicphilosophy of eugenics endures. In 1994,controversy over eugenics re-emergedwith the publication of The Bell Curve. In this book the authors described a “dysgenic trend”: the poorly educatedwere reproducing more rapidly comparedto the well educated. The authors


claimed that race and class differences inIQ are primarily genetic and immutable.15

Many people from the fields of genetics,sociology, and statistics have presentedstrong scientific arguments against theseclaims. Nonetheless, the debate itselfgave fresh sustenance to the perceptionthat race matters.

Genetic discoveries open the door to new ways to practice eugenics. For example, it is now possible to selectchildren based on genetics. Selectiontechniques fall into two categories: selec-tions for therapeutic reasons — to avoidserious medical disorder — and selec-tions for enhancement — to obtain a desired trait.

DNA can be obtained from fetusesthrough amniocentesis and other tech-niques used for prenatal diagnosis, andsome parents now use this information todecide whether or not to continue a preg-nancy. DNA also can be obtained fromembryos created through in vitro fertiliza-tion, and parents sometimes use thisinformation to select which embryo orembryos to implant in the mother.

At the present time, prenatal diagnosisis used primarily to determine the pres-ence of problem alleles linked to single-gene disorders, so as to avoid giving birthto a child with a fatal illness. Forexample, in the U.S. standard prenatalcare now includes a DNA screening testfor cystic fibrosis. Tests for many otherdiseases will probably become routine inthe coming years. It should be mentioned

that prenatal screening today is imper-fectly implemented; Cystic fibrosistesting in the United States has resultedin some terminations of healthy fetusesdue to misinterpretation of results,according to medical geneticists.16

Prenatal diagnosis also can be used toselect for sex, either for medical reasonssuch as if the family has a history of an X-linked medical disorder (in which casethey would select for a girl), or if the parents simply prefer to have a child of aparticular sex (in which case they mostoften select for a boy).

It is not possible to genetically select forsuch behavioral traits as pleasant disposi-


The Nazis used pseudo-scientificeugenic theories to justify theoppression and mass murder of Jewish people and other non-Aryans. Today, some peoplefear that behavioral genetic theories could be distorted toadvance the idea that somepeople are better than others.

tion, athleticism, altruism, or high intelli-gence. One reason it is not possible is thatwe do not yet have solid proof about anyof the genes that play any part in thesebehavioral traits. Even if and when all relevant genes are identified, geneticselection for a behavioral trait will becomplicated by the fact that many genesare involved.

Yet as research advances, it maybecome feasible to identify the manygenes associated with a behavioral traitand to engineer embryos to carry all theknown “positive alleles” for that trait.Would such engineering actually create achild who grows up to have the desiredtrait? Possibly yes, but possibly no,because the effects of genes on behaviorare determined by their environmentalcontext over the course of development.

A more likely possibility is thatenhancement opportunities will be mar-keted by genetic testing companies beforethere is any solid scientific basis to theirproducts. This, of course, would be ethically problematic. Many other ethicalconcerns also attach to the possibility ofgenetic enhancement for behavioraltraits: • When and if enhancement technology

becomes available, who will haveaccess to it – only the rich, or everyoneequally?

• Who decides what enhancements aredesirable – individuals or society?

• Will parents feel obliged to select forenhancement so as not to disadvantagetheir child?

• How will selecting for particular traitsaffect the parent-child relationship?

• What are the psychological, legal, andsocial ramifications when children donot meet the potential expected bytheir enhancements?

In the end all such questions may bemoot, because complex behavioral traitssuch as intelligence may be foreverbeyond our ability to engineer. The morewe learn about behavioral genetics, themore likely it seems that we will never beable to completely predict or control howour children think, how well they think,and what they do with their thoughts.

Notes

1 Linda Gottfredson, Professor of Education, University of

Delaware, quoted in Holden, C. (2003, pg. 192).

2 Gottfredson, L., et al. (1994, pg. 13).

3 Wozniak, R. H. (1999).

4 See Gould, S. J. (1995, pg. 2).

5 See McClearn, G. et al. (1997).

6 See, for example, Chorney, M. J. et al. (1998).

7 Chorney, M. J. (1998).

8 Hill, L. et al. (2002).

9 University of Chicago psychologist Janellen Huttenlocher,

quoted in Wickelgren, I. (1999. pg. 1832).

10 See Wickelgren, I. Ibid.

11 Nichols, P., and V. Elving Anderson (1973).

12 Gottfredson, L., et al. (1997, pg. 15).

13 See Steele, C. M. and J. Aronson (1995, pg. 797).

14 Turkheimer et. al. (2003). Abstract quotation from pg. 623.

Study quotation from pg. 628.

15 Herrnstein, R. J., and C. Murray (1994).

16 See Concar, D. (2003).


Genetic research could make itpossible for parents to pre-selectthe behavior characteristics oftheir children — or at least to try.Even if such genetic selectioncould be done, the unforeseeableeffects of the environment wouldmake the results uncertain.

R E S O U R C E S F O R C H A P T E R 7

“Alfred Binet.” 2003. (Accessed 26 March); available at

http://www.indiana.edu/~intell/binet.html.

Asian-Nation. 2003. “The model minority image.” (Accessed 30

June); available at http://www.asian-nation.org/model-

minority.shtml.

Chapman, A.R. and M. S. Frankel (eds.), 2003. Designing Our

Descendants: The Promises and Perils of Genetic Modifications.

Baltimore: The Johns Hopkins University Press.

Chorney, M. J., K. Chorney, N. Seese, M. J. Owen, J. Daniels, P.

McGuffin, L. A. Thompson, D. K. Detterman, C. Benbow, D.

Luninski, T. Eley, and R. Plomin. 1998. “A quantitative trait locus

associated with cognitive ability in children.” Psychological

Science 9, no. 3: 159-166.

Concar, D. 3 May 2003. “Test blunders risk needless abortions.”

New Scientist 178: 4.

Gottesmann, I. I. 1997. “Twins: En route to QTLs for cognition.”

Science 276: 1522-1523.

-----. 2003. “The endophenotype concept in psychiatry: etymology

and strategic intentions.” American Journal of Psychiatry 160:

636-645.

Gottfredson, L. 1998. “The general intelligence factor.”

Scientific American (Accessed 31 March 2003); available at

http://www.psych.utoronto.ca/.

---. “Mainstream science on intelligence.” 1997. Intelligence 24 (1):

13-23. Reprinted from Wall Street Journal 13 December 1994

(51 additional signatories).

Gould, S. J. 1995. “Curveball.” In The bell curve wars: Race, intelli-

gence, and the future of America. Steven Fraser, ed. New York:

Basic Books.

Herrnstein, R. J., and C. Murray. 1994. The bell curve: Intelligence

and class structure in American life. New York: Free Press.

Hill, L., M. J. Chorney, D. Lubinski, L. A. Thompson, and R. Plomin.

2002. “A quantitative trait locus not associated with cognitive

ability in children: A failure to replicate.” Psychological Science

13: 561-562.

Holden, C. 2003. “The practical benefits of general intelligence.”

Science 299: 192-193.

Lave. Jean. 1988. Cognition in practice: Mind, mathematics, and

culture in everyday life. New York: Cambridge University Press.

Lubinski. 2000. “Intelligence, success and fitness.” Novartis

Foundation Symposium 233: The nature of intelligence. Gregory E.

Bock and Jamie A. Goode, eds. Chichester, U.K: John Wiley.

McGue, M., T. J. Bouchard, Jr., W. G. Iacono, and D. T. Lykken. 1993.

“Behavioral genetics of cognitive ability: A life-span perspective.”

In R. Plomin and G.E. McClearn, eds. Nature, Nurture, and

Psychology. Washington D.C.: American Psychological Association

Press, 59-76

McLearn, G. E., B. Johansson, S. Berg, N. L. Pedersen, F. Ahern,

S. A. Petrill, R. Plomin. 1997. “Substantial genetic influence on

cognitive abilities in twins 80 or more years old.” Science 276:

1560-1563.

National Center for Education Statistics. 2003. “Dropout rates in the

United States: 2000.” (Accessed 30 June); available at

http://nces.ed.gov/pubs2002/droppub_2001/.

Nichols, P., and V. Elving Anderson. 1973. “Intellectual performance,

race, and socioeconomic status.” Social Biology 20: 367-374.

Nuffield Report. 2003. Chapter 7: “Review of the evidence:

Intelligence” and Chapter 13: “Selecting and changing behav-

ioural traits.” In Genetics and human behaviour. London:

Nuffield Council on Bioethics.

Plomin, R. 2003. “Genetics, genes, genomics, and g.” Molecular

Psychiatry 8: 1-5.

Sofair, A. N. and L. C Kaldjian. 2002. “Eugenic sterilization and a

qualified Nazi analogy: The United States and Germany,

1930-1945.” Annals of Internal Medicine: 132: 312-319.

Steele, C. M. and J. Aronson. 1995. “Stereotype threat and the intel-

lectual test performance of African Americans.” Journal of

Personality and Social Psychology 69: 797-811.

Turkheimer, E., A. Haley, M. Waldron, B. D’Onofrio, I. I. Gottesman.

2003. “Socioeconomic status modifies heritability of IQ in young

children.” Psychological Science: 14: 623-628.

Wickelgren, I. 1999. “Nurture helps mold able minds.”

Science 283: 1832-1834.

Wozniak, R. H. 1999. “Francis Galton: Hereditary genius (1869).”

Extracted from Classics in Psychology, 1855–1914: Historical

Essays. Bristol, UK: Thoemmes Press. (Accessed 26 March 2003);

available at http://195.12.26.123/psych/galton.htm.


chapter seven - AAAS · CHAPTER 7: HOW IS INTELLECT MOLDED BY GENES AND ENVIRONMENTS? 99 ......

Documents

Transcript of chapter seven - AAAS · CHAPTER 7: HOW IS INTELLECT MOLDED BY GENES AND ENVIRONMENTS? 99 ......