The social and behavioral sciences have in recent years played host to what has come to be called the “nature versus nurture debate.” This is the question of to what degree our genes (“nature”) explain human differences in certain phenotypic variables such as intelligence, personality predispositions, criminal tendencies, and so on, versus the degree to which the social environment (“nurture,” which includes many social structural variables such as schooling, one’s social class status, one’s parent’s or parents’ social class status, one’s family socialization practices, nutrition, peer group influences, and so on) also explain intelligence, personality, and criminal tendencies.
Comprising part of this debate is the question of whether and to what extent there is interaction between genes and environment upon intelligence or personality predispositions. This is the issue of whether and to what extent the effect of genes depends upon what social environment(s) the individual is raised in, and also whether the effect of social environment itself depends upon a person’s genes. Thus gene-environment interaction would be present if both genes and environment combine in particular ways to account for a person’s intelligence and/or personality.
The whole issue is a debate, because while some researchers argue that the effect of genes is the greatest (in the neighborhood of 70 percent, if expressed as a percentage, thus leaving only 30 percent to social environment and gene-environment interaction), other researchers argue that the effect of the social environment is the greatest, comprising 50 percent or more of the effect on intelligence and personality. Then there are researchers who argue that gene-environment interaction is the largest component, though they are reluctant to place a percentage estimate on its effects.
The evidence and the methodology of the genetic effect as well as the effect of social environment lead to the conclusions (a) that the effect of genes on intelligence (and personality), though estimated by some researchers in the past to be “high” at around 70 percent, is instead more in the neighborhood of 50 percent, and very possibly less, in the neighborhood of 30 to 40 percent, if appropriate new methodological considerations are introduced; (b) that the effect of the many variables in the social environment (assuming for the moment that gene-environment interaction is zero) is at least 50 percent, and very probably more; and, further, that (c) while gene-environment interaction is no doubt a reality, it is not empirically estimable at the present time. So leaving aside the question of gene-environment interaction, the nature-nurture debate boils down to the question of whether the effects of genes on intelligence and personality predisposition are greater than or approximately equal to the effects of the social environment, or whether the effects of the social environment are greater.
Measurement of Intelligence
To estimate the effects of genes versus social environment upon something like intelligence, intelligence must first be measured accurately—that is, measured with little or no measurement error. Considerable debate exists about this issue—even apart from the question of the relative effect of genes versus social environment. Some researchers argue that intelligence and academic potential are quite accurately measured today, with little measurement error. At the same time, other researchers argue with equal verve that standardized ability tests today are not very accurate or fair measures of intelligence or academic potential of minorities as opposed to whites, of women as opposed to men, and of working-class persons as opposed to upper-middle- and upper-class persons. Both sides note, however, that achievement (specific subject) tests, as opposed to “ability” tests, are somewhat more valid across these groups.
Intelligence is an abstract concept; individuals are presumed to differ or vary from each other. “Intelligence” is itself not directly observable or measurable, but it is inferred to exist, to greater or lesser degrees, from actual results obtained on a test or indicator, such as an IQ test or the SAT (Scholastic Assessment Test). Thus “intelligence” (concept) is distinct from “intelligence test” (presumed indicator). The extent to which the indicator accurately measures how individuals differ on the concept is called the validity of the indicator or test. If the test lacks validity, then by definition some measurement error exists in the test. For example, what is called the “predictive validity” of a test, such as an IQ test (like the Stanford Binet or the Wechsler test) or the SAT, is the extent to which it predicts some later measured criterion, such as freshman-year college grades earned after taking the test. In other words, does the test do (predict) what it is supposed to? Hence, if for a sample of individuals, the higher one’s SAT score and the higher one’s freshman grade average (that is, if the test and grades had a high correlation), then the test would by definition have high predictive validity. If one’s SAT score bore no such relationship (no correlation) to freshman grades, then the test would have low or no predictive validity.
The differential predictive validity of a test is the degree to which it predicts the same or differently, test-to-criterion, for different racial/ethnic groups, different social classes, or the two different genders. Thus if an ability test taken in high school predicts freshman grades in college to the same degree (with the same correlation and/or slope) for whites as for blacks, then the test would be considered to be of equal predictive validity for both racial groups. By definition, the test would then be considered racially unbiased—even if whites scored on the average higher on the test than blacks—as is the case with the SAT, on which blacks (and Hispanics) score about 100 points less on average than whites (on the 200-800 points scale per section). The lower average score for blacks would then be attributed to two possible causes. First, the genetic (nature) explanation, which states that, assuming the test is not biased, the average difference in test score is due to fewer “intelligence” genes on the part of blacks and Hispanics. (More on this follows, under “Heritability.”) Second, the social environmental (nurture) explanation states that, again assuming that the test is not biased, the average black-white (and Hispanic-white) difference in test score occurs because of fewer educational opportunities and more environmental deprivation for blacks and Hispanics, rather than different genes.
But all this assumes zero measurement error; that is, the same predictive validity between racial groups. The fact of the matter is that, as shown in more than 100 studies of differential predictive validity, considerably more studies show different (lower) validities for blacks and Hispanics than for whites. A small number of studies do indeed show the same predictive validities for all three racial groups, but those showing different validities—thus measurement bias (measurement error) in the test—are far more numerous and also more recent. This means that the tests are significantly more biased—with significantly more measurement error—in the case of blacks and Hispanics than in the case of whites. (Asians tend to fall between whites on one hand and blacks and Hispanics on the other on predictive validity, even though Asians score on average higher than whites on the SAT math sections and somewhat lower on the verbal sections.)
The same issue applies to comparisons of females to males and to comparisons across social classes, as upper (or middle) class versus working class. While females and males score on average about the same on the SAT verbal (and writing) sections, men have consistently outperformed women on the mathematics (quantitative) sections. Some argue that this is because of “inherent” (genetic?) intellectual differences between men and women—a position once taken by the president of Harvard University, Lawrence Summers. But the results of quite a few studies on test bias strongly suggest otherwise: differential predictive validity, predicting college grades from SAT math scores, is higher (in correlation) for men than for women. This means that by definition, the test is biased (in the prediction sense) against women. And women tend to get higher grades in the sciences in college than their SATs would predict. In this sense women’s grades are underpre-dicted. This finding is buttressed by two kinds of evidence. First, the SAT, as well as intelligence tests, contains a number of sex-typed (thus biased) questions on the word-math problems (questions more relevant to traditional male socialization practices, like calculating the volume of an automobile crankcase, than to traditional female socialization practices, such as calculating the volume of a pot on the stove). Second, evidence, especially recent evidence, shows that overall socialization practices in general differ between boys and girls as children and as teens, thus leading to different cultures for men and women. Thus cultural differences between female and male show up as test bias on tests such as the SAT, as is the case also with racial/ethnic comparisons.
Finally, there is the case of social class differences. In general, the higher one’s family (parents’) income (and education), then the higher one’s test score. This is a consistent, well-documented, clear relationship found in literally all studies. This relationship remains the same even when one compares people of different social classes who are of the same race. The richer your parents are, the more likely you are to score on average higher on the SAT (and the ACT as well). The poorer your parents, the lower your test score. In addition, the lower the social class of the test-taker, the less the predictive validity of the SAT test. So the SAT has some class bias in it, just as it has bias on the basis of race and of gender. Two arguments for all this are, first, families of higher social class (higher parental income, education, and occupational status) are more able to give their children better educations, wider experiences, and more cultural advantages, which all translate into higher test scores and also higher predictive validity of the test. Second, families of higher social class are more likely able to foot the bill for a test-taking seminar (such as the well-known Kaplan Seminars or Princeton Review series), which are expensive and currently cost $1,000 or more for one sequence of test prep sessions. Thus, the often heard argument that the lower a person’s social class, the less intellectual genetic endowment the person has (the argument advanced a few years back by the best-selling book entitled The Bell Curve ) is untenable given the evidence of both less test validity for lower-class persons and fewer economic and cultural advantages for them.
The question of whether human differences in intelligence and personality are caused more by their genetic differences or by their differences in social environment has been addressed by studies in several behavioral science disciplines using what is called the “heritability coefficient.” The heritability coefficient is the proportion of the total differences (total variance) in a phenotype (phenotypic variable) that is causally attributable to genetic factors. Typical phenotypic variables studied are intelligence, personality predispositions like aggressiveness, criminal tendency, and self-reported criminal acts. Equivalently, the proportion of variance in the phenotypic variable is accounted for by genetic similarity between pairs of biological relatives. Consequently, if heritability of a phenotypic variable (say, intelligence) is 70 percent, then, statistically speaking, 70 percent of the differences in intelligence (assuming no measurement error) in that population are due to genetic factors. That would mean that 30 percent of the differences in intelligence would be due to environment and to gene-environment interaction. A 50 percent heritability would mean that the differences in intelligence are equally caused by genetics and by environment (including gene-environment interaction). The heritability estimate applies only to the population studied (e.g., whites) and cannot be generalized to another population (e.g., blacks) or between populations (e.g., comparing blacks to whites), although many researchers throughout history have tried. (Those who have tried, via such faulty between-group comparisons, to say that heritability calculated on whites applies as well to other groups, such as blacks, have been, for example, Sir Francis Galton in the mid-1800s; Sir Cyril Burt in the 1920s through the 1960s; Arthur R. Jensen and Richard Herrnstein in the late 1960s and early 1970s; Charles Murray—focusing on between-class differences—in the 1990s; and J. Phillipe Rushton in the 1990s and since 2000.) The same applies to between-class comparisons and to between-gender comparisons.
How can one estimate heritability on a single population? There are generally two methods. One is to note that as the biological overlap between pairs of relatives is greater, then so is the similarity between them in the phenotype of interest. Thus, identical twins raised together (such twins are by definition exact genetic duplicates, or clones, of each other) are more similar in measured intelligence to each other than are fraternal twins raised together similar to each other. Estimates of heritability based on such comparisons tend to be in the neighborhood of 70 percent for intelligence.
The problem with this method is that as biological overlap between pairs is greater, then so is their intra-pair similarity in social environment. Identical twins raised together are not only high in genetic similarity, but high in environmental similarity also. Thus genetic causation is confounded with social environmental causation, and the relative effects of the two cannot be viewed separately. A way around this is the second methodology: the study of identical twin pairs separated early in life and raised in separate environments. If this is true, then any similarity between them in intelligence is due to genes alone, and not social environment, since their environments are separate and uncorrelated. Estimates of the heritability of intelligence (and of certain personality predispositions) using the study of separated identical twins center around 70 percent.
This separated-twins method makes considerable sense and is methodologically sound. The problem with it is that for the actual studies of twins separated at birth because of the early death of a parent, family problems, and so on, some twin pairs are more separated than others. Some pairs were indeed separated at birth and raised apart, never meeting prior to being tested for intelligence; they were raised in completely separate families in different towns and attended different schools. Yet other pairs—said by the researchers to be separated—were in fact raised next door to each other, attended the same schools in the same grades, and had other similarities in their familial social class environments. Some were separated considerably after their birth or were reunited for a time after an initial separation. The heritability estimate for these not-so-separated twins is around 80 percent, but the heritability estimate for the remaining quite separated twins is around 40 percent—considerably distant from an estimate of 70 percent. In general, the more separated identical twin pairs are, the less similar they are in such phenotypic variables as intelligence and certain personality predispositions. This is evidence that, although genes as well as the social environment have strong effects, the effects of the social environment are greater.
In conclusion, while both genetics (nature) and the many variables comprising the social environment (nurture) affect phenotypic variables such as intelligence, personality predispositions, criminality, and others, it appears from the relevant data that the effects of the social environment are somewhat greater than the effect of genetic heritability. This assumes no measurement error for the measurement of intelligence—an implausible assumption, given the evidence of lack of predictive validity, plus evidence of racial, social class, and gender bias in current measures of intelligence and academic ability. The most convincing evidence concerning the genetic heritability of intelligence comes from reanalysis of data gathered on samples of identical twin pairs separated at birth and raised in different environments, showing that some pairs were more separated than others, and further, that the more separated the pairs are, then the less similar they are in intelligence, and thus the less the size of the heritability estimate characterizing them. No recent estimates of the effects of gene-environment interaction are available.
- Bouchard, Thomas J., Jr. 1998. “Genetic and Environmental Influences on Adult Intelligence and Special Mental Abilities.” Human Biology 70(April):257-79.
- Bouchard, Thomas J., Jr. and Matt McGue. 2003. “Genetic and Environmental Influences on Human Psychological Differences.” Journal of Neurobiology 54(1):4—45.
- Chipuer, Heather M., Michael J. Rovine, and Robert Plomin. 1990. “LISREL Modeling: Genetic and Environmental Influences on IQ Revisited.” Intelligence 14(1):11-29.
- Gardner, Howard. 1998. “AMultiplicity of Intelligences.” Scientific American Presents 9(Winter):19-23.
- Goldberger, Arthur S. 1977. “Twins Methods: A Skeptical View.” Pp. 299-324 in Kinometrics: Determinants of Socioeconomic Success Within and Between Families, edited by P. Taubman. New York: North-Holland.
- Goldberger, Arthur S. 1979. “Heritability.” Econometrica 46(November): 327-47.
- Herrnstein, Richard J. and Charles Murray. 1994. The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.
- Jensen, Arthur R. “How Much Can We Boost IQ and Scholastic Achievement?” Harvard Educational Review 39:1-123.
- Kamin, Leon J. 1974. The Science and Politics of I.Q. Potomac, MD: Erlbaum.
- Manning, Winton H. 1968. “The Measurement of Intellectual Capacity and Performance.” Journal of Negro Education 37:258-67.
- Scarr, Sandra. 1981. Race, Social Class and Individual Differences in IQ. Hillsdale, NJ: Erlbaum.
- Steele, Claude M. 1999. “Stereotyping and Its Threat Are Real.” American Psychologist 53(June):690-91.
- Taylor, Howard F. 1980. The IQ Game: A Methodological Inquiry Into the Heredity-Environment Controversy. New Brunswick, NJ: Rutgers University Press.
- Taylor, Howard F. 1995. “Review Essay: The Bell Curve.” Contemporary Sociology 24(March):153-58.
- Taylor, Howard F. 2002. “Deconstructing the Bell Curve.” Pp. 60-76 in 2001: Race Odyssey, edited by B. R. Hare. Syracuse, NY: Syracuse University Press.
- Young, John W. 2004. “Differential Validity and Prediction: Race and Sex Differences in College Admissions Testing.” Pp. 289-302 in Rethinking the SAT, edited by R. Zwick. New York: Routledge.
This example Nature-Nurture Debate Essay is published for educational and informational purposes only. If you need a custom essay or research paper on this topic please use our writing services. EssayEmpire.com offers reliable custom essay writing services that can help you to receive high grades and impress your professors with the quality of each essay or research paper you hand in.