This section is courtesy Frank Fujita
One of the big debates that occupies a lot of many people's time is the NatureNurture controversy. It is also sometimes known as the geneticenvironmental controversy. A lot of this controversy has been recently sparked by The Bell Curve, but in some form or another goes back to antiquity. In this controversy, we ask if the behavior of people is due to their Nature (or genetics) or to their Nurture (or environment). This is a philosophical question which science can only answer at the sidelines.
Let me spend some time discussing some trivially true examples, so that we can limit the future discussion. If we take a baby human, and a baby monkey and give them both the best environment that we can imagine, the child will be radically different from the monkey and the differences will be almost totally caused by genetic differences. At the extreme, the proponents of the Nature side of the controversy would have us believe that the differences in human behavior are like the difference between human behavior and monkey behavior, just less extremely.
Similarly, if we take identical twins, and give one the best environment possible, and put the other one in closet for eighteen years the differences will be profound, and caused totally by environmental differences between the two children. At the extreme the proponents of the Nurture side of the controversy would have us believe that the differences in human behavior are like the differences in the twin's behavior, just less extremely.
Thus, we can create a scientific environment that will produce an almost totally genetic answer to our question, or a scientific environment that will produce an almost totally environmental answer to our question. The answer we get will be determined by the situation that we set up. This point cannot be overemphasized, any scientist can create the answer to the NatureNurture question that she desires.
Some scientists produce results, and other scientists say that the scientific environmental environment was not a good one (too much like the monkey or twin environment above). Usually, it is the scientists that favor the Nurture side of the argument that actually set up the studies, and the scientists on the Nature side of the argument that criticize the studies. This leads to name calling. The Nurture scientists say in effect, produce data or be quiet. The Nature scientists say in effect, if you can't produce quality data, be quiet.
The heritability coefficient sounds like it answers the NatureNurture question, but it only does so at the edges.
Family, adoption, and twin studies and combinations of these designs can be used to estimate the magnitude of genetic effects as well as their statistical significance. This is the descriptive statistic called heritability. Heritability is an estimate of effect size given an particular mix of existing genetic and environmental factors in a particular population at a particular time. It is a descriptive statistic that estimates the proportion of phenotypic variance (i.e., individual difference in a population, not behavior of a single individual) that can be accounted for by genetic variance. It describes «what is» rather tan predicting «what could be» or «what should be.» Heritability does not imply genetic determinism-it refers to probabilistic propensities, not predetermined programming.Consider height. Correlations for first-degree relatives are about .45 on average, whether relatives are reared together or adopted apart. Identical and fraternal twin correlations are .90 and .45, respectively, regardless of whether they are reared together or adopted apart. These results indicate significant genetic effects. For these height data, heritability is estimated as 90%. This estimate of effect size indicates that, of the difference among individuals in height in the population sampled, most of the differences are due to genetic rather than environmental differences among individuals. (Robert Plomin, Genetics and Experience: The Interplay Between Nature and Nurture, p 43-44)
Certainly we can change height by adding growth hormone, restricting vitamin intake, or any of many other environmental interventions. However, in the sample of people measured, at the time measured, the individual differences in the sample were mostly genetically caused.
There are some questions about the heritability coefficient. First, in some studies, the computed heritability coefficient is greater than 1.0, that is in some samples the identical twins are more than twice as similar as fraternal twins. Secondly, adoption studies show that siblings are even less alike than we would expect (based on the heritability coefficient). This may be because of nonadditive genetic variance. That is that a configuration of genes may produce a certain trait (say Extraversion) rather than a simple additive scenario where the more genes you have to be extraverted, the more extraverted you are. One estimate of variance in personality traits (Dunn & Plomin, Separate Lives: Why Siblings are so Different, p. 50) is that 40% is genetic, 5% is shared environmental, 35% is non-shared environmental, and 20% is error variance.
Shared environmental differences are differences that can be given to everybody, say living in a big home is a shared environmental difference, as is a nice school, a good library, kind parents, etc. Non-shared environmental differences are differences that are specific to a child (within a family). Environmental influences in individual development are specific to each child rather than general to an entire family.
One myth that I would like to dispel is that the Nature- genetic differences are difficult to change, but Nurture- environmental differences are easy to change. Many genetic defects are very easy to correct, and many environmental deficiencies are impossible to change. For example, poor eyesight is genetic, and yet there are many ways to correct it, eyeglasses, contact lenses, and surgery. On the other hand, a childhood with little protein makes one short, and once grown, we cannot raise the individual's height to what we might want to think of as her genetic potential.
In short, the answer to the question «Is X caused by the environment or is it genetic?» is usually «Both.» Even heritability coefficients of 1.0 do not mean that the environment cannot (under different circumstances than ones in which the study was performed) affect the trait in question, and even coefficients of 0.0 do not mean that in some other sample of people that there will not be a genetic influence on the trait. The genetic influence may be additive (as in height) or non-additive (as in extraversion). The environmental influences may be things that the parents (or government) can change, or they may be non-shared environmental differences that cannot be fine-tuned with our present, crude understanding of the influences of the environment on behavior. Lastly, whether the influences are genetic or environmental do not inform us as to whether the trait will be easy to change or not - some genetic traits will be easy to fix (like eyesight) some environmental traits will be hard to fix.