Restrictions on scientific communication are nothing new—recall that Galileo was sentenced to house arrest by Church inquisitors in 1633 for supporting the Copernican theory. But serious threats to the freedom of scientific discourse have come to the forefront recently, after suspicions were raised of Chinese spying at U.S. national laboratories. These allegations triggered the imposition of new restraints and revived some old ones. So it is a good time for the scientific community to reflect on the evolution of such government controls in our time, to express support for regulations that are justified and to protest those that do harm.
After the outbreak of World War II, many physicists voluntarily censored themselves, with the result that papers about nuclear fission disappeared from the pages of the Physical Review starting in 1941. The government soon issued explicit restrictions on scientists and engineers privy to military secrets in the form of a classification system. Under that arrangement, information was labeled "Special Compartmentalized Information," "Top Secret," "Secret" or "Confidential," depending on the severity of damage to the nation should the material be compromised.
To accompany this system, government agencies introduced various levels of clearance. People seeking employment in organizations dealing with classified information or equipment surrendered some of their privacy by permitting a field investigator to talk about their past conduct with their neighbors and associates. In addition, inquiries into the subject's vulnerability to blackmail could be made, on the premise that such a liability might result from alcoholism, a drug habit or even money troubles. What is slightly more controversial—because it smacks of guilt by association—is the extent to which government investigators were allowed to inquire into people's political activities and, during the Cold War, their contacts with Communists or former Communists.
After the Cold War, it became much more difficult for government investigators to decide which associations were good and which were bad. I was once asked, for instance, as part of the renewal of my clearance, to submit copies of all my correspondence with my native country, Germany, which I left 61 years ago. The request turned out to apply to such a large pile that I was able to talk the investigators out of it. But even if investigators clear someone quickly and without hesitation, in the end, they have to make a leap of faith that the person will be loyal and responsible tomorrow as well as today.
Indeed, the clearance system is fundamentally based on trust, because, among other things, the cleared worker carries classified information in his or her head, something that cannot be censored. So it is imperative that a person given clearance play by the rules. Although certain security measures lessen the opportunity for misconduct, the system has had past failures: Aldrich Ames, John Walker and Ronald Pelton, for example, were all cleared. Yet they all succumbed to a desire for money and revealed military secrets to potential adversaries.
So classification and the accompanying clearance process have obvious shortcomings. But classification has worked reasonably well and is generally accepted as necessary and appropriate by the scientific community. In particular, it draws a well defined line between communications that are approved and those that are forbidden. But recent allegations of Chinese spying have brought on a worrisome shift in the nature of these governmental controls.
A New Inquisition
Curiously, the charge of spying itself remains extremely badly documented, the widely publicized Cox report notwithstanding. That report greatly oversimplifies the problem by broadly alleging "stolen" secrets without identifying what was widely known beforehand. The report does not specify the source of the losses, even though the wording implies that a Los Alamos scientist of Taiwanese birth was responsible. It blames no particular person for the leaks, and neither the Cox report or any other government report alleges that foreign visits to American labs, or vice versa, were culpable in any way. Yet a direct consequence of these and earlier allegations of spying has been an extension of the classification and clearance processes to enmesh broader areas of science and to restrict sweeping categories of "sensitive" information.
Concerns of this nature were already voiced in 1982, when Admiral Bobby Inman, then deputy director of the Central Intelligence Agency, charged that scientific communication had led to "hemorrhage" of military data to Cold War opponents. A committee set up by the National Academy of Sciences and chaired by Dave Corson, then President of Cornell University, effectively rebutted the charge. Subsequent Academy committees have re-examined the issues, and all of them have offered the same basic recommendation: Build high fences around truly critical information, such as detailed designs for nuclear weapons, specific war plans, negotiating positions and similar items, but let the rest of scientific communication proceed unfettered. Passage of the Freedom of Information Act in 1966 expands the prescription for openness by requiring that information generated with government support be made publicly available, unless it falls into certain specified categories. But this relatively benign framework is now collapsing, or is at least being severely shaken.
In an almost hysterical reaction, the executive and legislative branches of government seem in competition to appear toughest in preventing the release of "sensitive information." There has, for example, been a stream of regulations issued by the Department of Energy that add to the already overly broad reach of its security regulations. These rules restrict the participation and conduct of foreign visitors, the dissemination of fully unclassified information and the foreign travel of cleared individuals.
For a limited period, the DOE also introduced the new category of Sensitive Unclassified Technical Information (SUTI). Its description included such broad terms such as nuclear reactor systems, flash radiography, inertial-confinement fusion and many other general topics in science and technology. The DOE has also begun to apply export controls to technical information, an exercise that poses similar problems because the material being regulated remains ill-defined. These regulations apply to "deemed exports," which include even speech in the presence of possible non-nationals!
The Pendulum Swings
The only guideline thus far for deemed-export controls is that a person in the presence of a foreign national may not give "gratuitous information" or allow the "premature release" of sensitive information. One problem is that the regulations do not give a clear description of what is or is not permissible to talk about in front of foreigners. And as a professor, I might add that giving gratuitous information in which the audience may or may not be interested is our bread and butter. Moreover, such restrictions are clearly ineffective, because they do not prevent the U.S. citizens who are listening from passing the information on to foreigners later. Thus, in respect to export control, U.S. scientists are poorly informed as to what is forbidden and yet may feel they are putting their jobs in jeopardy if they even approach the indistinctly drawn line. As a result, these rules have discouraged many scientists from attending foreign conferences or making foreign contacts.
So the new regulations erode some researchers' ability to do good science. They also constitute a direct threat to true national security, for several reasons. On the most basic level, they burden the officials who are now charged with enforcing the existing classification and clearance processes. These people, who in the DOE are part of the Information Security Oversight Office, are woefully over-burdened. A backlog of applicants for clearance numbering in the tens of thousands has accumulated throughout the government. Administering the recent, more restrictive changes in classification rules, which had been altered substantially not so long ago by the "Openness Initiative" of former Secretary of Energy Hazel O'Leary, puts additional demands on these officials. Ill-defined restrictions on the freedom of speech of scientists tax these people even more and thus compromise their ability to do their jobs in a timely fashion.
The new regulations create a further risk to true national security by hindering the many members of U.S. weapons labs who have engaged in a successful program with their Russian and Chinese counterparts to improve the protection, control and accounting of fissile materials and nuclear weapons. Under the new restrictions, contacts between foreign scientists and America scientists having access to classified information are subject to checks on whether the foreign participants have connections with spy organizations. This new requirement may appear sensible, but it creates delays measured in months rather than days and bogs down activities that in a real sense make our country safer.
The DOE has also proposed rules limiting the access of foreigners to unclassified computers, even those at national laboratories and universities doing no military work. The managers of those systems rightfully maintain that no cyber-security beyond "good business practice" is needed. By "good business practice" they mean the use of access codes and other tools of industry for protecting the privacy of information about personnel and finances. Further restrictions would do a great deal of damage to the worldwide collaborations that now share in the design, execution and interpretation of experiments at facilities in the U.S. and abroad. Some of the suggested regulations would make such joint work entirely impossible.
American scientists should not have to work under such unnecessary and counterproductive controls. Indeed, the scientists at the national laboratories should feel reassurance and encouragement, not general suspicion based on unproven charges, which will only erode their morale. Over the longer term, this sort of treatment will tend to dissuade students from taking jobs at national laboratories, perhaps even from choosing science for their careers. In this respect, increasing the restrictions on the freedom of speech for scientists is the worst move that could be made in the interest of the national security.