One often-heard explanation, Kahan said, is lack of information: policy makers and the public do not know the scientifically solid information that research has provided or the information has been wilfully distorted.

Another explanation cited by Kahan focuses on the source of information: perhaps scientists lack credibility. "Maybe they are arrogant or can't communicate properly, or maybe there's some segment of American society that doesn't like scientists," he said, before adding that neither explanation suffices and going on to offer an alternative based on his own research.

That alternative explanation is "cultural cognition", a term that Kahan and colleagues use to describe "the tendency of people to fit their perceptions of risk and related kinds of facts to the values that bind their identities". That dynamic, he said, "is an important part of the political conflict that we see over issues that admit of scientific investigation". Studies show that when people receive new information about, for example, climate-change research, they do not update their prior beliefs but rather assess the new information in the light of those beliefs.

"You all know what nanotechnology is," Kahan told the science writers, "but that's because you're freaks". When the laughter died down, he said that 80% of the subjects in a study he conducted said they had never heard of nanotechnology or had heard only a little bit about it. "But," he added, "90% had an opinion about it." The point of the exercise was to determine which values subjects held dear and which might influence the formation of their beliefs.

The subjects' world-views were assessed, using a scale developed by anthropologist Mary Douglas. One axis measures preference for individualist versus communitarian solutions, the other measures preferences for hierarchical versus egalitarian social ordering.

Douglas concluded that people's perceptions of risk cohere with one or another set of those values, Kahan said. Subjects high in hierarchical and individualist values would believe that if climate change is a high risk, it would lead to restrictions on commerce and industry, activities that are important to their world-view. Those ranking high on the communitarian and egalitarian scales are, in contrast, more ambivalent about commerce and industry and more amenable to the notion that climate change posed a real risk to society.

Both groups believe that the scientific consensus tends to support their view, said Kahan. Presented with a benchmark National Academy of Sciences report on climate change, the egalitarian-communitarians in his study "were very likely to perceive that there is a scientific consensus that climate change is happening and humans are causing it" while hierarchical-individualists "were very unlikely to think there is scientific consensus on that issue; they're more likely to believe there is division among experts on these facts". Kahan added that some hierarchical-individualists even think that the scientific consensus goes the other way.

So what is to be done? "We can't get rid of cultural predispositions," said Kahan. One path might be to widen sources of information, he suggested, especially if new information can come from various experts who individuals believe are congenial to their own views. It will take time, though, he acknowledged.