Secrets and Open Societies

I grew up in a small East Tennessee town where many people spoke with a Northern accent and held jobs that they would not describe in detail. Oak Ridge was a plant town that had been built by the government during World War II to help build the bomb. By the 1970s, many people were still involved in various ways in weapons development, and there was an entrenched culture of secrecy about the place. We knew who worked at which plant—there were three of them: X-10, K-25, and Y-12—but not what they did there, and it wasn't something you asked about. During the war, most people did not fully know what they were doing themselves. Only a few people understood the overall project.

The scientific culture in Oak Ridge was quite different from the one Michael Selgelid addresses in his article in this issue. Selgelid considers the possibility that on some occasions, openly published biological research both "alerts terrorists and 'rogue' nations to possible ways of making biological weapons and provides them with explicit instructions for doing so." Selgelid describes two recent studies that show this possibility to be more than theoretical and argues that, even if these two studies warranted publication, studies can be imagined that would not. At least some biologists agree. A 2003 report from the National Research Council, which Selgelid criticizes, recommended a system of regulation combining federally organized review with self-censorship by scientists and publishers.

But biologists put a very high premium on openness. "The norm of open communication is one of the most powerful in science," says the NRC report. "To limit the information available in the methods section of journal articles would violate the norm that all experimental results should be open to challenge by others." This norm is not just a quirk of this scientific community. By making contest and confirmation possible, open communication within science promotes scientific progress, and because contest and confirmation is supposed to be the method of the "marketplace of ideas," open communication within science can be seen as just a special form of a more general openness favored in a democratic society. Still, these linkages are not airtight. Medical progress would not grind to a halt if some studies never saw the light of day, and even open societies can proscribe some forms of communication, particularly forms that might lead to violence. Thus the scientific norm of openness can probably not be explained entirely in terms of more general considerations. Though not just a local quirk, it seems to be of very special interest to the locals. The contrast between biology and nuclear physics, as Selgelid notes, brings this out.

Selgelid's beef with the NRC recommendations is the reliance on self-censorship. Scientists are not trained to be national security analysts, he observes, and their own interests will bias them against censorship. Besides these straightforward points about scientists' training, however, Selgelid's discussion underscores that the decision about whether to publish a piece of scientific information is itself a matter of values, and that scientists may bring a special set of values to that decision. The norms accepted within a field, however, are not necessarily those that should govern it, nor are the people within a field necessarily the best placed to select and apply the norms that should.