Thursday, June 15, 2006

Before the Dawn slammed in Nature   posted by Razib @ 6/15/2006 07:20:00 PM

Readers know that I don't have a totally positive opinion of Before the Dawn, but this hatchet job in Nature is out of control, it stinks of "The Sociobiology Study Group." Full text below the fold, I'll let readers point out the issues....
In Your Own Image

Care must be taken when looking for natural selection to explain the evolution of human behaviour.

Kenneth M. Weiss and Anne V. Buchanan review Before the Dawn: Recovering the Lost History of Our Ancestors by Nicholas Wade

In Before the Dawn, journalist Nicholas Wade explores the "lost history of our ancestors" from a genetic viewpoint. He presents a compilation of scenarios that are meant to explain the evolutionary origins of human behaviour and social structure during the past 50,000 years, discussing archaeological findings but mainly focusing on how genes can reveal the 'hidden' past that can't be inferred from fossils or archaeology. He uses examples ranging from the first wearing of clothes and the origin of hairdressing to the evolution of language, race and intelligence.

These inherently interesting and smoothly presented tales of progress towards the human present will doubtless captivate many readers. However, what could have been a tempered and timely treatment of an important subject is, in our view, regularly undermined by Wade's determination to find simplistic natural selection behind every trait, and by a lack of attention to issues that are known to inhibit a credible understanding of complex traits, never mind their evolution.

Wade's explanations commit various well-known errors, such as equating correlation with causation and extrapolating from individual traits to group characteristics. Often his arguments and trait choices are laden with Western-oriented value judgements. The following are a few examples of the kinds of problematic scenario that can be found in the book. Wade suggests that until recently racial variation has not been the subject of scholarly pursuit, and confidently asserts that there are five "major" races, identified statistically by neutral markers (from a genome that he assures us is 97% "filler"), yet "the alleles involved in differentiating the human population are likely to be of the selected kind not the neutral kind". He also states that stressing genetic racial differences as he does is "objective" and scientific, but stressing human similarities is "political". Wade argues that Europeans resist 'mad cow disease' because their ancestors were selected for cannibalism. He also says that Jews were selected for higher intelligence than other peoples because of the calculational demands of money-lending. He suggests that high intellectual skills are a genetic adaptation that occurred only after the origin of settled societies in places such as Europe. And he says that the Chinese as a "race or ethnic group" excel at ping-pong, which should encourage researchers to look for a genetic explanation.

Extraordinary claims demand extraordinary evidence. In The New York Times on 15 January 2006, Wade warned against journalists being too ready to accept "overstated or wrong" claims from the science literature, but in too many places where it makes a difference he has ignored his own advice. A journalist doesn't create facts, but he does select what to repeat and how to colour it, and Wade is long on speculating about what "is reasonable to assume", and short on circumspection of his own, or anthropologists', yarn-spinning. Most of the scenarios he reports have not been rigorously tested, nor is it clear how they could be. The book has many internal inconsistencies, and one can easily find contrary evidence or readily construct alternative 'just so' stories that invoke the same genetic scenario and the same kind of reasoning.

How could this subject be better treated, without denying the importance of genes in human traits? For a start, evolutionary arguments should be based on sufficiently credible, consistent and compelling scientific evidence. It is easy to claim that a trait is due to natural selection, but responsible selection-based arguments should have substantial experimental mechanistic support, at least for the fact of selection. That's not the state of most current evidence. Indeed, after 50 years of investigation, we can't convincingly demonstrate selection for most of the red-blood-cell diseases, other than sickle-cell anaemia, that are probably coevolving with the strong selective force of malaria. Other best-case scenarios for human genetic adaptation, such as adult lactase persistence and skin colour, are also incomplete. Explaining selection is particularly problematic for behavioural traits because of the powerful role of culture and facultative ability, which is probably what human evolution really favoured. Human phenotypic changes can far outpace genetic ones, making it challenging to know whether such traits are even genetic, much less what they 'evolved for' millennia ago.

In addition, assertions of genetic causation should be built on what is already known about the difficulties of explaining complex traits, including behaviour or intelligence. The extensive literature documenting the subtleties of such traits undermines simplistic 'evolved for' scenarios, but Wade largely ignores it. The aetiology of complex traits is influenced by environmental factors as well as variation at multiple genes, greatly attenuating the causal impact of individual genes. We are far from understanding either the genetic architecture or the evolution of complex biological traits, even in the best data from experimental organisms unaffected by the blur of culture. Intensive gene mapping has typically failed to identify more than a fraction of even the genetic variation, much less all the variation, in such traits. The effects of experimental genetic manipulation in laboratory animals routinely vary significantly even among the few strains tested, and the life experiences of litter-mates, twins, inbred animals and clones are far from identical. Despite this sobering knowledge, Wade claims example after example of 'genes for' traits.

But why not just enjoy the sport of fanciful speculation, even if the arguments leak like sieves? Because it's not just sport. Positions on genetic determinism often correlate with social politics, and few of us are neutral or even changeable on the issues. Wade recognizes that his ideas may not be acceptable to everyone but warns that "to falter in scientific inquiry would be a retreat into darkness". He seems to be warning, appropriately enough, against benighted political correctness. But we should never become casual about how comparable 'slopular' science and very similar speculative evolutionary reasoning by leading scientists fed a venomous kind of darkness not too many decades ago. Wade's post-hoc tales often put him in step with a long march of social darwinists who, with comfortable detachment from the (currently) dominant culture, insist that we look starkly at life in the raw and not blink at what we see. But given today's limited understanding of complex traits, too often what one sees is oneself.