Cognitive Science: An Introduction/The Nature-Nurture Debate

Psychologists and cognitive scientists have long debated about the extent to which psychological traits (personality, intelligence, and so on) are due to our genetic makeup or our experiences.

As described in a previous chapter, randomness during early brain development generates variation in innate psychological traits that are not genetic in origin.

When we try to understand these issues, it often comes down to what scientists call explaining variance. Let's suppose we have a group of people. We'll call this group a population. They have some variation. Maybe some are smarter than others. We also have variation in factors we believe might be related to intelligence, like genetic makeup, nutrition, parenting style, and so on. Scientists can measure all of these things and use statistics to determine what percentage of the variation is explained by these factors.

For each population, we have whatever variation. It might be a little variation, or a lot, but whatever it is it adds up to 100%. Then the job of scientists is to find what parts of that 100% total are explained by each of the candidate factors.

These findings are notoriously difficult to think about properly. It's important to note that when, say, we say that a factor explains 10% of the variation, we are not saying that the factor causes this variation. What many people are interested in is the amount that genes versus environment contributes to some trait. But this is not what explaining variance is all about. To help understand this, suppose we were to know that if somebody had 100 IQ points, 50 of them would be determined by genes, and the other 50% were to be determined by environment and the randomness associated with development. This would not mean that, if we studied a population of people, we'd necessarily find that genes explained 50% of the variance.

Here's why: if the population studied had very similar genes, then the analysis of variance would show a reduced effect of genes. This is because, if everybody has similar genes, they will not explain what variance is there in some trait. Similarly, if everybody has the same culture, then the effect of environment will be less, and genes will be more. So even if genes contributed the same amount to a trait, the amount of variance explained by genes could be very different if you studied different populations that had different variability in environment and genetics.

Similarly, if everyone in the population has the same IQ, say, then there is little variance to be explained anyway. The fact that they can all remember things well does not enter into it, and the results give no information about what the genetic contribution to memory is.

Measurement Error

Right off the bat we know that around 10-20% of the variation is explained by measurement error. The tests we have for psychological traits are not perfect. If we give someone an IQ test, and then test them again, they will not always get the same exact score. The closer the score is to the previous score is the measure's test-retest reliability. The reliability of IQ tests is about 0.9 (which means that measurement error explains 10% of the variance we see in a population), and is about 0.7 for personality traits (measurement error of about 30%).

When we measure a trait across a population, then, we expect 10-20% of the variation we will find is just because our measurements aren't perfect. So what about the other 80-90%?

Genes

Depending on the psychological measure you're looking at, genes account for about 50% of the variance we see.

Each of us has a number of mutations, some more than others. The average is around 150, depending on how you count them. And some of these mutations matter more than others. A redundant gene might make up for a deficiency in another that got mutated. How well an individual manages with their mutational load can be thought of as robustness, and this, too, varies among individuals. This robustness can be seen in how symmetrical people's bodies are. Even if you look at identical twins, the ones that are more symmetrical (for example, have more symmetrical faces) look more like each other. This might explain a part of why better looking people are also smarter: the same robustness that makes their faces more symmetrical also makes their brains work better.

For intelligence, studies of adopted siblings and separated twins suggest that genes explain about 50% of the variance in IQ in childhood, but by the time these children are adults it has grown to 80% or more, and the effect of family environment goes to zero. It appears that the effect of family environment gets smaller over time. Over time, the become more and more who they are genetically predisposed to be.

Experiences and Environmental Factors

What an individual experiences has an effect on who they are. These factors are sometimes referred to as the environment, or nurture. This includes parenting styles and family environment, interactions with friends and culture, as well as interactions with pathogens, and aspects of the natural environment, such as weather and terrain. There have been hundreds of studies looking at influence of these factors, and the surprising result is that the effects are small, inconsistent, or absent. In other words, your experiences account for very little of the variance in personality or cognitive ability.

Generally people tend to overestimate the strength of the effect of experience. The family environment, for example, only explains about 10%-15% of the variance in traits, and sometimes none at all. Twins raised apart are nearly as similar as twins raised together.

It's also important to note that genetic and developmental differences can cause the individual's environment to change. Genes can affect the environment in three ways. Passively, a child with a genetic predisposition for working hard at school also probably has parents with the same genes, causing a home environment more conducive to study. There can also be an evocative effect: a smart child who is interested in learning will be given more attention by teachers. A violent child is more likely to be punished or become the victim of violence. Finally, there is an active effect, in that that genetic predisposition will make the child seek out learning experiences.

Randomness in Brain Development

Your genes provide the basic instructions for how to build a body and brain, but during early brain development there is a great deal of randomness in the system. Even identical twins, who have the same genome (minus a few copying errors) will not have identical brains. This explains variance too, even though it is not typically thought of as either nature or nurture, though it is, in some sense, innate. See the chapter on Brain Development for more details.

Through experience, the brain strengthens some connections between neurons are weakens or removes others, according to statistical regularities in the world. It makes the brain efficient for processing the world it finds itself in. After a time, this flexibility is turned off, at different times for different brain systems, and the ability for the brain to change, after that, is severely reduced. The brain is substantially developed by the time it is born, though new features are expanded until the early 20s.