nological data. Emphasizing restriction over dissemination leads to other problems.

The cost of classification is probably great; no known evaluation with empirical or statistical validity exists. Until a financial impact is identified, top management support for classification management programs will not be forthcoming. Information control must be measurable in dollars and cents.

Executive Order 10501 designates to a number of departments the authority to classify information in the interest of national defense. Within the Department of Defense, each of the component services has specific regulations relating to the protection of scientific and technological information. Even though all of these regulations conform to the Executive Order, their procedures are often quite different from one department to the other.

But these problems are rather pragmatic. Restricting information can often become an end in itself. C. P. Snow writes:

"The euphoria of secrecy goes to the head very much like the euphoria of gadgets. I have known men, prudent in other respects, who became drunk with it. It induces an unbalancing sense of power. It is not of consequence whether one is hugging to oneself a secret about one's own side or about the other. It is not uncommon to run across men, superficially commonplace and unextravagant who are letting their judgment run wild because they are hoarding a secret about the other side-quite forgetting that someone on the other side, almost indistinguishable from themselves, is hoarding a precisely similar secret about them. It takes a very strong head to keep secrets for years, and not go slightly mad. It isn't wise to be advised by anyone slightly mad." What is not yet realized is the effect of today's technologies on yesterday's theory and policies.

PROPOSALS

There are really three considerations in light of the previous analysis. The first is an introspection of the present system as outlined in Executive Order 10501. This requires that: (1) the reader accept the need for restricting scentific and technological information, and (2) the reader accept present classification "theory." Then recommendations for the short-run can be proposed. The second consideration is the need for an extended analysis. A series of questions outline the long-range topics that require examination. Once these have been answered, the above assumptions can be relaxed to consider a third point-an alternative to the present classification system.

SHORTRUN PROPOSALS

A short-range proposal suggested frequently is the elimination of the confidential category. Little qualification is required by a U.S. citizen to be granted a confidential clearance. A very creditable case can be made for eliminating such clearances entirely. Unfortunately, the end result of this proposal might be a dramatic increase in the amount of secret information. Control of dissemination offers a more positive alternative.

Central control of information is not intended. But, a central clearinghouse of scientific and technological information is proposed. Researchers working in a specific scientific or technological area could subscribe to unclassified abstracts of all known published information regarding their particular area. Any references would note applicable defense access restrictions and, on classified projects, unclassified materials would be identified. Some work like this has already been done, but it is not extensive nor does it include very much restricted information. The Federal Council on Science and Technology (10) has offered some rather specific recommendations regarding current classification policy:

1. All government reports should be freely available to the U.S. scientific and technical community unless specifically exempted by the "Freedom of Information Act."

2. The transfer of information is an inseparable part of research and development, and the control and dissemination of information resulting from such research and development is a vital element of the agencies' responsibility. This position has been advocated by the President's Science Advisory Committee (Weinberg Panel).

3. The Federal Government is responsible for insuring that there exists in the United States at least one accessible copy of each significant publication of the worldwide scientific and technical literature.

4. The technical information of government-produced or government-sponsored technical documents must be made available within a reasonable time and without unusual procurement effort by the potential user.

Another policy recommendation of this committee was the development of subject category lists to aid classification of government-wide scientific and technological information. Armed with these recommendations, we can better fit a set of proposals to the needs-information dissemination and a refinement of management policies.

Although Executive Order 10501 is specific in its emphasis on dissemination of information, access to the scientific and technological information is still limited. The AAAS has encouraged the widest dissemination possible:

"Protection measures against the extension of regulations limiting disclosure are necessary, but they are not enough. It is equally essential to develop methods of dissemination and critical appraisal appropriate to the exponential growth of science and technology."

The object should not be to merely protect the data, but also to inform potential users of its availability and provide easy accessibility. (In a sense, dissemination mechanisms exist in the Defense Documentation Center and the Aerospace Research and Application Center.) Additional mechanisms must be provided for restricted information so that classified projects do not suffer from suffocation. Scientific and technological information needs all of the exposure it can get.

At the same time, management policies must be adapted to work with expanded information requirements. The first step would be for security planners to conduct a formal evaluation of information dissemination policies as they actually exist. A balance between cost and benefits must then be determined. Within the Department of Defense, some action is being taken; Government Executive (34) reports that security planners will, in the near future:

1. Review and update all security classification guides which should result in downgrading many documents.

2. Update standard practice procedures within defense industrial facilities. 3. Increase classified document retention within industry to allow industry to maintain a given scientific or technical expertise which otherwise might be lost.

In a recent memo from the Assistant Secretary of Defense (Administration) (52) the following policy was outlined:

1. Only those aspects of development programs involving (technical) information should be classified which are really sensitive and need to be protected in the interests of national defense.

2. Classification should be held to the minimum length of time necessary after considering the degree of sensitivity, securing costs, and probability of compromise.

3. Classification decisions should be reached only after careful consideration of the advantages of open circulation, after public release approval, of the information against the advantages to a potential enemy.

4. Appropriate downgrading and declassification should be forecast and provided for, with dates where practicable.

But defense activities must be coordinated with the activities of other agencies in working toward a single government-wide set of guidelines. A total program is summed up by Palfrey. (29)

"It has often been stated that secrecy has been overrated by the government and has not accomplished what was expected of it. Particularly, it has been said that the cost of secrecy has outweighed its contributions, and that a more balanced evaluation of both could reduce the cost without reducing the actual contributions to our society."

Thomas (44) presents a good interpretation of this analysis:

"An effective classification policy must include considerations of the effect that possible restriction of information will have on other technological developments. Such restrictions will necessarily have some adverse effects on the development of your own systems for national defense and national security. Such restrictions will also necessarily have an adverse effect on the growth of the economy as a whole and national security is not unrelated to this growth." Policy at the central level is posited as a synthesis of all Department requirements. As for basic research, all information should continue to be unrestricted. In initial applied stages, the data could be left unrestricted unless the information provides a definite solution to a military requirement or the application of the information would provide a definite national defense advantage (to

include elements of technological lead time). As research turns to development and application, national security requirements should be more specific. Schloss (34) sees current defense programs forming to:

1. Allow a greater flow of information to news media and the public regarding defense postures.

2. Increase the industrial base because of the availability of such information to small business.

3. Facilitate international export and trade by American industry.

4. Permit wider exchange of know-how among the scientific and technical communities, including colleges and universities, domestic and international.

5. Provide for state-of-the-art technology availability to commercial and civilian firms.

6. Reduce costs associated with safeguarding classified material.

Items 1 and 4 will be greatly facilitated with the new distribution procedures contained in the recently issued DoD Directive 5200.20, "Distribution Statements on Technical Documents." (53)

The retention of classified materials by industry following contract completion has already been mentioned. This is an important area because often a company develops a particular expertise in an area that could well have some transfer value to the civilian sector. However, if the core information is withdrawn and the classified information destroyed, the expertise is nullified to a great extent. Retention of needed classified inventories would assist technological transfer.

As pointed out previously, cost studies are needed. This need has been known for many years, yet no funds have been allocated to conduct in-depth research of the "cost" of classification. We cannot attempt to measure social costs until the monetary results are known.

In general, a complete reorientation of security procedures is needed at all levels. Schloss quotes a Defense planner as saying:

"[A good security man] should take into consideration that effective classification management can also make available technology and techniques for domestic nonmilitary programs. . . . He needs extensive seasoning, exposure to and understanding of national policies and programs."

Training in information science is of potential value to security policy personnel. Such training can assist in developing the needed attitude, which has been described by the Board of Directors of the AAAS: (41)

"A positive program of security can be developed. It requires boldness; it demands continued beliefs in the fundamental loyalty of American scientists, engineers, and industrialists and in their ability to keep the United States ahead of potential enemies."

Once these short-range solutions are in effect, some very important questions regarding the long-range considerations will have to be asked.

Immediate solutions will have only a temporary effect, leaving some important questions unanswered. One of the critical problems encountered in this research was that neither side (classification proponents or advocates of free dissemination of scientific and technological information) has tried to identify the questions to be asked. Too much time has been spent searching for answers.

The basic question is, "How successful have security regulations been in protecting sensitive scientific and technological information?" Such a question suggests that some good, hard research be done in assessing technological lead times. It also requires that some effort be given to measuring the impact of classified restrictions on information flow. And, as noted before, we need to ascertain the true costs of classification. General models are being developed in information theory that might uncover some answers in this area.

A second question is, "Can Executive Order 10501 and its applicable implementing instructions cope with the new advances in communications technology?" Despite the fact that the emphasis is now on dissemination, information flow is still restricted. One critical example can be found in computer data usage. Taylor and Feingold (42) note that classification procedures are being designed for second-generation usage while users are buying fourth-generation equipment. The fifth-generation of computers has been described as an era of "invisible information." Visual display systems pose complex problems in that there is no paper output to classify in the traditional sense. If present policies cannot handle the rapid growth of technology, then a third question is in order.

Negative replies to the two previous questions infer a third one. "Can a strict distinction be made between strategic or tactical military information and scientific or technological information?” The former can be linked easily to

national defense while the latter becomes less easily defined as it interacts with military needs and general social needs.

If these questions cannot be answered, we must take a closer look at the present classification system. More specific questions must then be asked about procedures. Those answers will help in adapting current programs to current needs. But if the questions can be answered, we must rid ourselves of the constraints imposed by Executive Order 10501 and look to other possible alternatives.

ONE ALTERNATIVE

Part of the reason why spokesmen against secrecy restrictions on scientific and technological information have been able to advance such a strong argument is that no one has publicly challenged the "philosophy" behind Executive Order 10501. The present system of classification has not been proved to be the only alternative possible, let alone the best one. Is there not a better alternative? At this point, we really do not know. Not only must the previous questions be answered, but some fresh insight must be applied to the data resulting from the answers.

One possibility is to limit defense classified information to two categories. The first would be strategic and tactical information of a military nature (plans, intelligence, tactics, and so on). It is probably safe to assume that information restrictions such as these would find a broad base of support among those in science and technology. The second type of restrictions would cover information in science and technology in support of the first category-weaponry and weapons system development.

Scientific and technical information could come under a different set of rules. Instead of establishing categories of protection, security policy planners would, along with technical advisors, assign numerical values to "bits" of information. denoting the time (in months) that the information is to be protected. Only one protective category would be applied to scientific and technological information. Thus, confusion over category assignment would be eliminated while downgrading and declassification would be defined very specifically. An index and abstract service by a single agency could serve three purposes. Close control could be maintained over protected information. Also, users would be able to exchange information and ideas with known sources of protected information. Finally. upon expiration of the time restriction, the abstract could be immediately released to public subscribers. Number identifiers would also allow computer technology to assist in documentation, accountability, and revision or removal from protective restrictions.

Despite the fact that this alternative is rather ill-defined at this point; it does point to ideas that could be applied effectively to information dissemination. This presentation is not intended to support this one proposal but attempts to foster some original thinking by the reader in protecting scientific and technological info in the interests of national defense.

(Author's Note: This paper could not have been written without the contributions from those in the fields of research and classification management. Space limitations do not allow me to acknowledge them individually. However, two persons merit singular recognition for the time and effort they devoted to assisting and advising me in this project. Professor James Utterback at Indiana University provided invaluable aid in organizing the research design. Mr. Bob Donovan of United Technology Center, Sunnyvale, California reviewed each of the many drafts providing ideas and criticisms from the security practitioner's point of view. Of course, I accept full responsibility for the final product.)

BIBLIOGRAPHY

1. Aines, Col. Andrew A. "Committee on Scientific and Technical Information." Keynote Address, Fifth National Seminar, National Classification Management Society. Washington, D.C., July 22, 1969.

2. Boffey, Philip M. "Biological Warfare: Is the Smithsonian Really a 'Cover'?” Science (Feb. 21. 1969), pp. 791-96.

3. Borchert, John R. "Remote Sensors and Geographical Science," The Professional Geographer (November, 1968), pp. 371-75. (Addendum by E. Risley.) 4. Carey, R. J. P. Finding and Using Technical Information (London: E. Arnold, 1966).

5. Church. T. S. "Science and Technology and Classification Management." Journal of the National Classification Management Society, III (1966), pp. 39-43.

6. Coblans, H. "Science, Government, and Information," Journal of Documentation, XX (1964), pp. 1-4.

7. Condon, E. U. "Science and Security," Science (June 25, 1948), pp. 659–65. 8. Cross, Harold L. The People's Right to Know (New York: Columbia University Press, 1953).

9. Dickson, P. "Freedom of Information Act," Writer's Digest, XLIX (November, 1969), pp. 44-47.

10. "Dissemination and Technological Progress," panel discussion at the Fifth National Seminar of the National Classification Management Society, Washington, D.C., July 22-24, 1969.

11. Donovan, Robert D. "Science, Security, and Information," presented before the Research Security Administrators, summer, 1966.

12. Fernbach, Sidney, Note in the Journal of the National Classification Management Society, III (1966), pp. 48–53.

13. Garrett, C. Donald, "Airborne Passive Scanning Infrared Imaging Systems," reprint from Defense Industry Bulletin (September, 1967), pp. 10-13, and Suits, Gwynn H. "Declassification of Infrared Devices," reprint from Photogrammetric Engineering.

14.

"The Role of 'Need-to-Know' in Releasing Classified Information," reprint from Defense Industry Bulletin (February, 1969).

15. Gellhorn, Walter, Security, Loyalty, and Science (Ithaca: Cornell University Press, 1950).

16. Greene, Richard M., Jr., editor, Business Intelligence and Espionage (Homewood, Ill.: Dow-Jones-Irwin, Inc., 1966).

17. Guetzkow, H. "Communications in Organizations," in March, James G., ed., Handbook of Organizations (Chicago : Rand McNally, 1965).

18. Hotz, Robert, "The Cost of Secrecy," Aviation Week and Space Technology (Oct. 27, 1969), lead editorial.

19. "Industrial Security and the Scientist,” Industrial Security, II (January 1958), p. 10 and (October 1958), p. 5.

20. "Is Classification Confusion Really Necessary?" Industrial Security, XII (June 1968), p. 39.

21. Kahn, Herman and Anthony J. Weiner, "Technological Innovations and the Future of Strategic Warfare," Astronautics and Aeronautics (December 1967), pp. 28-50.

22. Law week, 37 LW 2489, Epstein v. Resor: 37 LW 2497, American Mail Line v. Gulick.

23. Lear, John, "Predicting the Consequences of Technology," Saturday Review (January 10, 1970), pp. 44-46.

24. McGuire, Martin C. Secrecy and the Arms Race (Cambridge, Mass.: Harvard University Press, 1965).

25. Mitchell, John J. "Government Secrecy in Theory and Practice," Columbia Law Review (February 1958), pp. 199-210.

26. Nieberg, Harold L. In the Name of Science (Chicago: Quadrangle Books, 1970). (Revised edition.)

27.

Press, 1964).

Nuclear Secrecy and Foreign Policy (Washington: Public Affairs

28. Petru, W. C. "U.S. Controls on the Exportations of Unclassified Technical Data: A Fabled Account," Special Libraries, LX (November 1969), pp. 596-600. 29. Palfrey, J. C. "The Problem of Secrecy," Annals of the American Academy of Political and Social Science (November 1953), pp. 90-99.

30. Parks, Wallace, "Secrecy and the Public Interest in Military Affairs," George Washington Law Review, XXXVI (October 1957), pp. 1–22.

31. Price, Ed. "AIAA Sounding Board," Astronautics and Aeronautics, V (May 1969), p. 26+.

32. Rogers, Everett M. Diffusion of Innovations (New York: The Free Press, 1962).

33. Rosenbloom, Richard J. "Technology Transfer-Process and Policy," National Planning Association Special Report Number 62 (July 1965).

34. Schloss, Leon. "DoD Security: The New Look," Government Executive (October 1969), pp. 44-47.

35. "Scientific and Technical Communication: A Pressing National Problem and Recommendations for Its Solution," Washington, D.C., National Academy of Sciences Publication 1707, 1969.

36. "Security and Constitutional Rights," Hearings before the Subcommittee on Constitutional Rights, S/Res. 94, 84th Congress, Second Session, November 1965.