The scientific community and the federal government are at odds over how to balance national security and scientific openness. At issue is the restriction of scientific publication, the barring of foreign students from certain areas of study, and the labeling of information as "sensitive but unclassified." We will be safer, some argue, if we wall off scientific knowledge of potential use to those who would do us harm. Yet nothing could be further from the truth. Disrupting the process of scientific discovery will put our economic vitality at risk, because it depends on the rapid advance of science and technology. Knowledge creation thrives in openness and suffers in isolation. How we now strike a balance between security and openness matters enormously. I offer three suggestions:

• First, continuous consultation by the federal government with the scientific community is essential. All too often security professionals and scientists neither understand nor trust each other. Scientists may be quite unaware of some of the real risks associated with their work. This has been a major problem within the nuclear weapons arena; it will be even more complex as we worry about basic research in the diffuse, little-understood context of terrorist threats.

• Second, distinct boundaries must be drawn. The use of ill-defined terms like "sensitive but unclassified information" will invite bad decisions, create new dangers for the nation and threaten scientific enterprise. The presidents of the National Academies have rightly warned that "vague criteria of this kind generate deep uncertainties among both scientists and officials responsible for enforcing regulations. The inevitable effect is to stifle scientific creativity and to weaken national security." President Reagan confronted these questions. In 1985, he issued a directive, which stated that federally funded basic research should either be classified or unclassified. Funding agencies retained responsibility for making that vital determination. The Reagan policy should be reaffirmed, and its application extended to other domains.

• Third, we should use voluntary agreements within the scientific community. The most effective thing scientists can do is to create a framework of forums in which scientists can determine the need for new mechanisms appropriate to our security needs and the requirements of science. Simplistic lists of forbidden topics cannot work in fast-paced fields like biology, where unanticipated discoveries are made every day. New modes of collaboration are needed to bring together those in government and science who, now immersed in these vital questions, are working disconnected from one another.

The debate about openness and security is not new. But the unpredictability of terrorism has changed the dynamics. Government and research universities need to work harder now than ever before to find the balance that will better protect us and our future. Our ability to prosper through research must remain undiminished by fear.

This article was originally published in the Wall Street Journal on November 29, 2002.