杏吧原创

SCIENTISTS are accustomed to investigating the natural world and to drawing conclusions about it. In Model Behavior, historian of science Nicole Nelson turns the spotlight on them for a change.

Like an anthropologist facing a new tribe, Nelson focuses her keen analytical mind on a US team researching the genetics of alcoholism in mice. The result is a stimulating and challenging exploration of science, and of how social scientists think about science. Everybody wins.

Nelson was allowed not only to observe lab meetings and experiments, but to be involved in data collection. At every stage, she recorded her impressions and interviewed lab members of all ranks. Her study has an unmatched richness and depth, thanks to this qualitative data.

Nelson assumed she would be highly critical of the lab鈥檚 practices. She viewed them as too reductionist: surely human alcoholism could not be usefully understood by probing the effects of genes in mice? But she was soon disarmed by the complex views of the researchers and their acceptance of many of her criticisms, which they had worked through years earlier.

Instead of studying a group of gung-ho gene jockeys who boasted about having found the gene for alcoholism, Nelson discovered a set of sophisticated scientists who thought hard about their work and made ultra-cautious claims. Above all, they worried about the effects of what is termed uncontrolled environmental variation, such as a change in the identity of the researcher engaged in a particular observation, or the fact that a female experimenter adopted a dog halfway through the study 鈥� so potentially altering the rodents鈥� responses to her smell.

This obsession meant that much of the lab鈥檚 focus, as Nelson describes it, was paradoxically not on genes but on apparently minor yet potentially confounding factors. The genetic component in alcoholism 鈥� in both mice and humans 鈥� is produced by many genes, each with a tiny effect. The team鈥檚 focus on a particular gene involved in one small aspect of alcoholism meant that the work was vulnerable to being knocked off course by environmental effects, requiring them to control for these factors.

This painstaking approach will come as no surprise to anyone who knows the field, but it is gratifying to see it documented in such detail, and to read of Nelson鈥檚 surprise at discovering it.

鈥淚nstead of gung-ho gene jockeys, Nelson discovered scientists who thought hard about their work鈥�

If her investigation undermines anything, it is not behavioural genetics but her own interpretative framework, derived from a field known as science and technology studies. For example, Nelson describes how the researchers, along with two rival labs, tried to publish conflicting results in a leading journal. The editor asked them to produce a single coherent story, but the scientists wanted to retain the contradictory aspects of their findings, considering that they might all be partly true.

Nelson鈥檚 theoretical training suggested that the scientists should compete and try to undermine each other鈥檚 data. Instead, they bent over backwards to try and criticise their own results. On this aspect of how science (and scientists) work, her framework is mistaken. We are not simply competitors in some market; we are trying to discover things that are true. There is no point in 鈥渨inning鈥� by publishing something that is not correct.

Nelson writes with passion and describes the link between genes and behaviour in a complex and nuanced way, one that scientists will recognise. Some jargon does creep in: scientists teeter on an epistemic void, we witness a status degradation ceremony, and are confronted with ineffective ontological assertions. These are no more difficult than the terms scientists use, and often there is a simple solution: a dictionary. 杏吧原创s, historians of science and the public will all learn a great deal from this excellent account.

This article appeared in print under the headline 鈥淎 culture of complexity鈥�