"Why do people — even identical twins — differ from one another in personality?"

This question needs to be asked because of the widely held conviction that we already know the answer to it. We don't. Okay, we know half of the answer: one of the reasons why people differ from each other is that they have different genes. That's the easy half.

The hard half is the part that isn't genetic. Even people who have identical genes, like Freeman Dyson's twin grandsons (see his question), differ in personality. I am not asking about the feeling each twin has of being "me": George and Donald could be identical in personality, and yet each could have a sense of me-ness.

But if George and Donald are like most identical twins, they aren't identical in personality. Identical twins are more alike than fraternal twins or ordinary siblings, but less alike than you would expect. One might be more meticulous than the other, or more outgoing, or more emotional. The weird thing is that the degree of similarity is the same, whether twins are reared together or apart. George and Donald, according to their grandfather, "not only have the same genes but also have the same environment and upbringing." And yet they are no more alike in personality than twins reared by two different sets of parents in two different homes.

We know that something other than genes is responsible for some of the variation in human personality, but we are amazingly ignorant about what it is and how it works. Well-designed research has repeatedly failed to confirm commonly held beliefs about which aspects of a child's environment are important. The evidence indicates that neither those aspects of the environment that siblings have in common (such as the presence or absence of a caring father) nor those that supposedly widen the differences between siblings (such as parental favoritism or competition between siblings) can be responsible for the non-genetic variation in personality. Nor can the vague idea of an "interaction" between genes and environment save the day. George and Donald have the same genes, so how can an interaction between genes and environment explain their differences?

Only two hypotheses are compatible with the existing data. One, which I proposed in my book The Nurture Assumption, is that the crucial experiences that shape personality are those that children have outside their home. Unfortunately, there is as yet insufficient evidence to support (or disconfirm) this hypothesis.

The remaining possibility is that the unexplained variation in personality is random. Even for reared-together twins, there are minor, random differences in their experiences. I find it
implausible, however, that minor, random differences in experiences could be so potent, given the ineffectiveness of substantial, systematic differences. If randomness affects personality, the way it probably works is through biological means — not genetic but biological. The human genome is smallish and the human brain is vast; the genome couldn't possibly contain precise specifications for every neuron and synapse. Identical twins don't have identical brains for the same reason that they don't have identical freckles or fingerprints.

If these random physical differences in the brain are responsible for some or all of the personality differences between identical twins, they must also be responsible for some or all of the non-genetic variation in personality among the rest of us. "All" is highly unlikely; "some" is almost certainly true. What remains in doubt is not whether, but how much.

The bottom line is that scientists will probably never be able to predict human behavior with anything close to certainty. Next question: Is this discouraging news or cause for celebration?

Judith Rich Harris is a developmental psychologist and author of The Nurture Assumption: Why Children Turn Out The Way They Do.