The Mirage of a Space Between Nature and Nurture [Paperback]

In this brief book, Evelyn Fox Keller points out, in convincing detail, the subtle "chronic ambiguity, uncertainty and slippage in the very language we use to talk about" the tangle of questions called the nature/nurture debate, and how our linguistic confusion makes them so hard to untangle. This book is not an easy read -- easy reads are part of the problem -- but I could not put it down. I only wish it were longer and broader in scope: genetics is not the only field needing Keller's sharp logic.

Everyone seems to know that the nature/nurture debate is over. But still, we all persist in talking about it still. Why? That is what Evelyn Fox Keller is trying to answer in this very short but insightful book. Why do we persist in speaking as if nature and nurture are separable and distinct variables when we all (seem to) know that they aren't? If we all know that questions about how much (what percentage) height is due to nature and nurture is unanswerable, then why do we keep wanting to ask it? And why do studies that only show how much of a population's variance in a trait can be ascribed to genetics seem always to be interpreted as studies showing how much of that trait is actually genetic? (After all, to know that x population varies in height and that 50% of the variance seems to correlate with genetic inheritance is a FAR different question from how much of Suzy's height is attributable to genetics.)

Keller's broad answer is mistakes like this are based on subtle ambiguities in the terms we use like "heritability," "gene," and even "nature" and "nurture." Keller's book is less about convincing us THAT nature and nurture can't be meaningfully separated and that questions about how much any trait can be attributed to nature or nurture (many others have made that point). Rather, she is trying to trace out why, even when we know this, we keep making the mistakes.

The first chapter traces the history of attempts to conceive of "nature" and "nurture" as distinct variables that can be separated. And - surprise, surprise - the chief culprit looks to be Francis Galton. In Galton's zeal to measure the heritability of things like genius, Galton began writing, and designing experiments, as if these two variables can be teased apart. (Of course, something like intelligence involves such an intertwining of genetic and environmental factors that literally there is no 'space' where nature ends and nurture begins.) This also involves a subtle change in the meaning of 'heritability' which went from meaning anything that was passed from generation to generation (regardless of the mechanism) to meaning ONLY those inherited traits that could be accounted for purely by biology.

The second chapter discusses the key difference between traits and differences in traits, suggesting that it is not so much traits, but differences in a trait's expression, that biologists usually study, even as the literature often sounds as if it is discovering things about the former rather than the latter. (To simplify it, genetic studies don't really aim themselves toward answering what causes height or intelligence, but what is responsible for the variation in a population between heights and intelligence. Very subtle, but important, difference.) The next chapter discusses why studies dealing with variance in x trait within a population SHOULD NOT be read as studies concerning how much x's expression in INDIVIDUALS owes to nature or nurture. (Keller often uses the analogy of drummers. If a study shows that 40% of the variation in drummers' sound correlates to the differences in drum used, this DOES NOT mean that a drummer's sound is 40% attributable to the drum.)

And lastly, there is a chapter discussing what to do about all of this. If the question we really want to answer is "How much of each person's intelligence is fixed by nature or malleable by nurture," how can we ask a question like that in a way that is answerable? Keller suggests we can change the question slightly to read more like "[H]ow malleable is a trait, at a specified developmental age?" Rare is the trait that is simply fixed, iron-clad, through the lifespan, and rare is the completely malleable trait. Maybe longitudinal studies should focus less on where nature ends and nurture begins, and stick to the more general question of how fixed certain traits appear to be over time (without reference to whether "nature" or "nurture" is responsible for the fixity or malleability).

If anything, I'd loved this book to have been longer. Keller is a good writer, but not being a statistician myself, it was often hard to grasp some of the very subtle statistical points she was making. Several times in the book, she appeared more concerned with demonstrating (repeatedly) that a mistake is persistent in the literature than with really pointing out WHY it is a mistake. (And the very persistence of these mistakes should indicate indeed how subtle the difference between the correct and incorrect views are! All the more reason to really explain and drive home the difference.)

This book is by no means an easy read, but it is definitely an important one. Keller does what a philosopher/historian of science should do: examine (usually hidden) assumptions that may be problematical, and argue as to why they are problematical and, possibly, why they persist. This set of very interesting essays/chapters do exactly that.