Two Daughters and the Question Genetics Doesn't Answer

My older daughter, Alice, was ten when we moved into this house. My younger daughter, Nora, was six. For years I described that year to people the same way, as a hard year for both girls, as if the move were one fact that had happened to each of them in equal measure, the way a closing date applies to every account on a balance sheet. It took me an embarrassingly long time to see that this wasn’t true. Alice cried over the friends she’d left for about a month, then made new ones and seldom mentioned the old ones again. Nora, four years younger and with hardly any friendships yet that were hers to lose, spent the next two years refusing to unpack the last box in her room, as though keeping it taped shut kept the move from being finished. Same move, same parents, same rented truck. Two different events would be closer to the truth, and once I started looking, “the same household” turned out to be a much weaker premise than I had assumed.

Behavioral genetics explains a good part of what I was missing, and it begins by giving genes more credit than the usual shrug, “they’re just different people,” allows. Twin and adoption studies put the heritability of most personality traits at roughly 0.4 to 0.6 (Bouchard and McGue, 2003). In plain terms, a substantial share of the variation in temperament across a population tracks genetic variation. That figure explains a lot about why two unrelated children differ. It explains much less about my two in particular, who share about half their genes, and it says nothing about the “same house” half of the puzzle.

The more useful answer is in a 1987 paper by Robert Plomin and Denise Daniels whose title is the question every parent of two children eventually asks out loud: “Why are children in the same family so different from one another?” Their answer is that the family environment is far less shared than it looks from outside. Birth order alone changes it. The first child gets parents raising an infant, a ten-year-old, a teenager, for the first time; the second gets parents who have done it once already. Treatment that looks like a cause of a child’s temperament is at least as often a response to it. A child who is already anxious gets handled more carefully than her sister was at the same age, and the care reinforces the anxiety it was answering. And an older sister who occupies one role in the family, in our house the easy, adaptable one, can push a younger one to define herself against it. Eric Turkheimer, summing up the field in 2000 as three informal “laws,” put the surprise bluntly in the second of them: the effect of being raised in the same family is smaller than the effect of genes. The move did not happen to “the household.” It happened to a ten-year-old with years of friendships to lose and a six-year-old with almost none, and that is most of why it produced two different experiences rather than one experience wearing two faces.

There is a third, smaller piece. Even identical twins, with the same genome and usually most of the same early environment, still diverge, because some of how a brain wires itself is developmental noise, random variation in how neurons migrate and are pruned that neither genes nor surroundings account for (Finlay, 2007). Alice and Nora were never going to be that alike. But it is worth knowing that even the closest possible case, same genome, same womb, same crib, leaves something over.

Put heritable temperament, an environment that was never really shared, and some developmental noise together, and you can build a complete causal account of why Alice is expressive and Nora is not: why one narrates her feelings before anyone asks, and the other has never once told me she was scared, though I have watched her be scared many times. I call this the easy story. The science isn’t easy, and I have left out nearly all of the mechanism. I call it easy because of what it leaves untouched, a separate question I had no words for until I read David Chalmers’s 1995 paper on consciousness, written about something else entirely, and recognized my own kitchen table in it.

Chalmers separates the “easy problems” of consciousness, explaining how a brain discriminates, integrates information and reports on its own states, from the hard problem: why any of that functioning comes with experience at all, why there is something it is like to be the system doing it, rather than the same operations running in the dark. The easy problems are hard science but not, he argues, mysterious in kind. His point, which took me longer than it should have to absorb, is that solving all of them would not touch the hard one. Thomas Nagel had asked a version of the question in 1974, about a bat. Joseph Levine, in 1983, called the space between a physical explanation and the felt fact an “explanatory gap.”

What I noticed, reading all three, is that the gap does not stay confined to consciousness in general. It reopens, at the same width, when the question changes from why Alice is more expressive than Nora, which the easy story answers, to why there is a particular someone being Alice at all: one center of experience with this specific cast to it, rather than the same profile of traits running unattended, the way a thermostat does its job without anyone inside it feeling the cold.

The obvious objection is that no parent wonders about that. We take it for granted that both our children are conscious; what puzzles us is the traits. I don’t think that is quite right about the wonder parents report. “How can two children from the same two parents be such different people?” doesn’t go away even for a parent who has read and accepts the genetics. Underneath it, I think, is astonishment that there are two of them there to be different, two separate places where it is like something to be someone, rather than one causal system producing two sets of behavior with no one in particular behind either.

Philosophy has two standard answers to a question like this, and both cost something. The broader one is illusionism, defended by Daniel Dennett (1991) and, more explicitly, by Keith Frankish (2016), which treats the apparent gap as a trick of how the mind represents itself rather than a real remainder. The sharper one, for my question, comes from a book about personal identity: Derek Parfit’s Reasons and Persons (1984). Parfit argues that there is no further fact about identity beyond the psychological and physical facts already on the table, nothing extra that settles whether tomorrow’s person is “the same” as today’s. By the same reasoning, nothing extra is needed to explain why there are two subjects here rather than one. There is Alice’s particular pattern of memory, temperament and continuity, and there is Nora’s, both already specified by the easy story. The sense that something beyond the pattern is required, an actual someone underneath each, is the very intuition his reductionism is built to talk me out of.

The competing view takes that intuition at its word. Chalmers’s own position, that experience involves a property physical facts alone do not fix, would treat it as evidence of something real; an older vocabulary calls it a soul. Either way the claim is that a pattern, however fully described, is not enough. Something further makes this subject the one who has this history, and the easy story, however complete, is not the kind of fact that could supply it.

Parfit would tell me I am making the mistake his book was written to correct: inventing a further fact where none is needed, when all I have evidence for is two continuous psychological patterns with two different childhoods, and a feeling of someone home in each that his view predicts and explains away. I have read the argument. I can’t refute it. I don’t believe it, and my reason is professional rather than philosophical. Every year since that move I have run the same check, and it has never resolved the way a misread entry resolves once you find it. An accountant learns, eventually, the difference between a mistake she hasn’t found yet and one that was never there. This is the second kind. I can’t say what the further fact is, a property, a substance, or something nobody has named yet. But I no longer think it is nothing, and I don’t think the genetics, however complete it becomes, was ever the kind of evidence that could tell me otherwise.