Science, Technology, and the First Amendment

January 1988

NTIS order #PB88-166426 Recommended Citation: U.S. Congress, Office of Technology Assessment, Science, Technology, and the First Amendment, OTA-CIT-369 (Washington, DC: U.S. Government Printing Office, January 1988).

Library of Congress Catalog Card Number 87-619897

For sale by the Superintendent of Documents U.S. Government Printing Office, Washington, DC 20402-9325 (order form can be found in the back of this report) Foreword

In honor of the Bicentennial of the United States Constitution, OTA is con- ducting a study of Science, Technology, and the Constitution. At the request of the Committee on the Judiciary of the House of Representatives, and its Subcom- mittee on Courts, Civil Liberties, and the Administration of Justice, we are examining ways in which continuing scientific advances and new technological developments may influence the scope and meaning of enduring constitutional principles and protections. A background paper, Science, Technology, and the Constitution, was released in September 1987. Freedom of speech and press, embodied in the First Amendment, are among the most cherished of those protections. As the first of a series of special reports this document considers challenges to freedom of speech and press that come from the advance of science and technology. Part I of the report discusses how the meaning of “the press” has expanded from printed material to include a wide range of broadcast and electronic media. Satellites, computers, electronic bulletin boards, teletex, videotext, and other new ways of gathering, editing, and delivering news are blurring legal and regulatory distinctions between common carriers and “the press, ” thus changing arguments about the constitutional rights that they have each enjoyed. Part II addresses freedom of speech and press as they apply to scientific communications and technological know-how. As science and technology become ever more important to our economy and our military strength, the delicate balance between individual rights and the national interest becomes both more important and more difficult to maintain.

Director

,.. Ill .— —— —...—-

ERRATA-- SN 052-003-01090-9

Science, Technology, and the Constitution Project Review Panel William Carey Monroe Price Advisor to the Carnegie Foundation of Benjamin Cardozo Law School New York New York, NY Washington, DC Mark Rothstein James Duggan Director of Health Law Director University of Houston New Hampshire Appellate Defender Houston, TX Program Concord, NH Thomas Smith, Esquire Assistant Director Judith Lichtenberg Criminal Justice Section Center for Philosophy and Public Policy American Bar Association University of Maryland Washington, DC College Park, MD Paul Stephen Peter Low University of Virginia Hardy Cross Dillard Professor of Law and School of Law John V. Ray Research Professor of Law Charlottesville, VA School of Law University of Virginia Laurence R. Tancredi Charlottesville, VA Kraft Eidman Professor of Medicine and the Law The Honorable Pauline Newman University of Texas Health Sciences United States Circuit Judge Center United States Court of Appeals for the Houston, TX Federal Circuit Washington, DC

Contractors Christopher Burns, Christopher Burns, Inc. Harold Green, George Washington University National Law Center

NOTE: OTA appreciates and is grateful for the valuable assistance and thoughtful critiques provided by the reviewers. The reviewers do not, however, necessarily approve, disapprove, or endorse this report. OTA assumes full responsibility for the report and the accuracy of its contents.

Iv Science, Technology, and the First Amendment OTA Project Staff

John Andelin, Assistant Director, OTA Science, Information, and Natural Resources Division

Fred W. Weingarten, Program Manager Communication and Information Technologies Program

Program Staff

Vary T. Coates, Project Director Benjamin C. Amick III, Analyst* Robert Kost, Analyst Mary Ann Madison, Research Analyst

Administrative Staff

Liz Emanuel Sandra Holland Becky Battle

*Until May 1987. Page Summary ...... 1

Part I. Freedom of the Press in the Information Age Chapter 1. New Technologies for Gathering News and Information ...... 5 Newsgathering ...... 9 Computer Databases ...... 10 Media Satellites ...... 11 Implications for Privacy ...... , . . . 12 Implications for National Security ...... 16 Chapter 2. New Technologies for Editing and Selecting News and Information ...... 19 Electronic Publishing...... 19 Editorial Control and Liability ...... 21 Global Networks and the International Press ...... 23 Chapter 3. New Technologies for Publishing and Disseminating News and Information ...... 27 The Convergence of the Media ...... 27 Cable Television ...... 30 Information Services Delivered Over Telephone Lines ...... 32

Part 11. Scientific Communications and the First Amendment Chapter 4. National Security and Scientific Communications ...... 37 Science, Free Speech, and National Security ...... 37 The Executive Branch and Classification of Documents...... 41 Export Controls ...... 49 Chapter 5. The 1980s: Converging Restrictions on Scientific Communications . . 55 Contractual Restrictions on Communications...... 55 Restrictions on Informal Communications ...... 56 Self-Restraint ...... 59 National Security Directives and the Role of the National Security Agency. 60 Chapter 6. Constitutional Issues: An Overview...... 65 Appendix A. Acknowledgments ...... 71 Summary

The First Amendment says: scene, threatens national security, or is an incitement to imminent violence or to the over- Congress shall make no law respecting an establishment of religion, or prohibiting the throw of the government; it can place reason- free exercise thereof; or abridging the freedom able restrictions as to the time, place, and man- of speech, or of the press; or the right of the ner of commercial speech or speech that takes people peaceably to assemble, and to petition place in public; it can force one to compensate the Government for redress of grievances. victims of defamation and other forms of speech injurious to private interests; and it can The provisions of the First Amendment are regulate speech that takes place over the air- designed to protect the most sensitive areas waves. Moreover, when speech or the exercise of personal belief-religion, ethics, political of religion merges into action, government can philosophy-and to act as a bulwark against regulate those forms of expression to protect government intervention in the most basic ele- the public health, safety, and welfare. ments of our democracy: the expression of thought and opinion. As necessary conditions o democratic governance, the rights embod- As necessary conditions to democratic ied in the First Amendment occupy a “pre- erred position” in the hierarchy of constitu- governance, the rights embodied in tional rights and powers. The first clear the First Amendment occupy a “pre- statement of this preferred position doctrine ferred position.” came in Thomas v. Collins, in which Justice Rutledge, speaking for the majority of the Su- preme Court, said: Because freedom of expression is not absolute, it often involves a balancing of govern- This case confronts us again with the duty mental interests against the rights of individ- our system places on this Court to say where uals where the two are in conflict. Balancing the individual’s freedom ends and the State’s power begins. Choice on that border, now as rights against power occurs in the context of always delicate, is perhaps more so where the contemporary values and institutions: eco- usual presumption supporting legislation is nomic, political, ethical, legal, and scientific balanced by the preferred place given in our or technological. scheme to the great, the indispensable democratic freedoms secured by the First Amendment . . . That priority gives these lib- Science and technology may affect the erties a sanctity and a sanction not permitting balance between First Amendment dubious intrusions, and it is the character of the right, not of the limitation, which deter- rights and government interest by mines what standard governs the choice . . . .1 changing power relationships. Notwithstanding the absolute language of the First Amendment and its preferred posi- Science and technology may affect the bal- tion in the constitutional order, however, the ance between First Amendment rights and Supreme Court has never interpreted freedoms government interest by changing power rela- of religion, speech, press, or assembly to be tionships between individuals and between the without limitation. These rights, which are col- individual and the state. As it first did with lectively referred to as freedom of expression, the printing press, technology will give rise to can be limited in a variety of ways. Govern- new ways of communicating, which amplify ment can prohibit entirely speech that is ob- the ways in which individuals and organizations express themselves. Information and ’323 U.S. 516, 530 (1945). communications technologies, such as satellites, computers, and digital transmission lines, country, or to speakers in foreign countries are, like the telegraph, telephone, radio, and when they are heard or read here, will also be television technologies before them, changing raised. the ways in which we communicate ideas, the- The open communication of scientific infor- ories, opinions, and incitements to action— mation—data, hypotheses, conclusions, explana- they affect who can say what, to whom, to how tory theories, technological know-how—is many, and at what cost. special kind of speech or publication. There Taken together, advances in computers and no consensus on the question as to whether telecommunications may change the concept scientific communication enjoys the full proof ‘the press’ from one in which one organiza- tection provided by the First Amendment 1 tion publishes for many to one in which many political communications. In a society in which share information amongst themselves. With science and technology play a central and critical role, this is an issue meriting continuing these changes will come new First Amendment challenges to the power of the government to attention and debate. regulate access to and ownership of commu- It is well established that scientific communications media. New technologies, such as nication can be limited when necessary to pro- electronic publishing, may not fit easily into tect national security. But how severe can and old models of regulation, and First Amend- should that limitation be? As science and tech- ment distinctions between the rights of print nology become ever more important to our publishers, broadcasters, and common carriers economy and our military strength, the deli- will become increasingly difficult to justify. cate balance between individual rights and the national interest becomes both more important and more difficult to maintain. Do the limits imposed by classification, congressionally leg- Distinctions between the First Amend- islated restrictions, and export controls, col- ment rights of print publishers, broad- lectively negate the right of free speech and casters, and common carriers will be- free press in the field of science? come increasingly difficult to justify. In fields such as mathematics, biology, or physics, basic research results in certain areas can have direct and immediate implications for New capabilities for the press to gather, technological development. In those case store, and retrieve information on individuals where the line between basic knowledge (sci- may require that rules of liability for constitu- ence) and its implementation (technology) be- tionally protected speech be reexamined. The comes thin and difficult to discern, a balance potential for technology to decentralize the between the right of expression and interest editorial function may raise questions of edi- of the state in preserving security is very dif- torial control and liability under the First ficult to achieve. There are likely to be many Amendment. And, in an era of global commu- situations in the future in which the govern- nications, the question of whether First Amend- ment will assert compelling reasons for limit- ment rights extend to foreign speakers in this ing basic scientific communications. Part I Freedom of the Press in the Information Age

We should note the force, effect, and consequences of inventions which are nowhere more conspicuous than those three which were unknown to the ancients, namely, printing, gunpowder, and the compass. For these three have changed the appearance and state of the whole world . . . —Francis Bacon, Novum Organum, Aphorism 129*

Networked computers will be the printing presses of the twenty-first century. If they are not free of public control, the continued application of constitutional immunities to the nonelectronic [press] . . . may become no more than a quaint archaism, a sort of Hyde Park Corner where a few eccentrics can gather while the major policy debates take place elsewhere. –Ithiel de Sola Pool, Technologies of Freedom (1983)

*AS ~uo~ ~ E~abet,h L. Eisens~~, Prjntjng ]+rss as an .4gent of Ch.wIge (Cambridge, MA: ~arnbridge University Press, 1979), vol. 1, p. 43. Chapter 1 New Technologies for Gathering News and Information

Despite its origins in the context of print- prohibited from licensing or otherwise control- ing, “freedom of the press” has come to be in- ling the use of the technology.3 terpreted as protecting communication to the In the two centuries that have passed since public generally, regardless of the medium. the ratification of the First Amendment, in- Print media, motion pictures, broadcasting, ca- novations in technology have added to the va- ble television, and even the mails have come riety and power of the press, and have, as a to be considered as the “press” for purposes consequence, changed the social, economic, and of the First Amendment. As the Supreme 4 political impact that the press has had. Court has said, "[press] comprehends every Thanks in large measure to technologies such sort of publication which affords a vehicle of 1 as the communications satellite, for example, information and opinion.” Moreover, while global television is now a reality; over 500 mil- some have argued that freedom of the press lion people watched the moon landing in 1969, was only intended to shield the dissemination and over 2 billion may have seen the Los An- of news and opinion, the protections of the geles Olympics in 1984.5 First Amendment have been extended to protect scientific, literary, and artistic messages Changes to the legal environment in which as well. It is this broad notion of the press as the press operates have accompanied these a vehicle for every kind of public expression technological changes.6 “Freedom of the that is used in this report. ~The importmce of this concept must be understood in the The printing press provided, for the first context of the English law that preceded it, which granted the Stationers Company monopoly rights over printing and required time, a capability for mass communication, government licensing to own and operate a printing press. Even whereby one individual or organization could after the ratification of the First Amendment in America in 1791, inform, entertain, or persuade many others. At however, the exercise of this freedom could be severely curtailed and punished. the time the Constitution was written, pub- ‘The new capability for rapid, accurate, and mass publica- lishing in the United States had not yet be- tion provided by the printing press facilitated the speed with come the “mass medium” it is today.2 A which the Protestant Reformation spread through Europe: “her- alded on all sides as a ‘peaceful art,’ Gutenberg’s invention prob- craftsman printer produced one page at a time, ably contributed more to destroying Christian concord and in- and could produce about 2000 copies of it in flaming religious warfare than any of the so-called arts of war a 10-hour day. The technology was inherently ever did. ” Printing Press as an Agent of Change (Cambridge, MA: Cambridge University Press, 1979), p. 319. Four centu- egalitarian; it took neither political power nor ries later, with live coverage of the civil rights movement in large sums of money for an individual to pub- the American south, and riots in the north, east, and west, ra- lish a work. The “freedom of the press” had dio and television thrust the issue of racial equality before the American public in the 1960s. Television has also been credited a more or less literal meaning; government was with a major role in ending the United States’ involvement in the Vietnam War. Michael Mandelbaum, “Vietnam: The Tele- vision War, ” Daedalus, fall 1982, p. 157. bJoseph Pelton, “The Technological Environment, ” Toward a Law of Global Commum”cations Networks, Anne Branscomb (cd.), by the Science and Technology Section of the American l~ove~] v. Grjffjn, 303 U.S. 444, 452 ( 1938). Bar Association (New York: Longman, 1986), pp. 37, 43. ‘The press as a mass medium awaited the industrial revolu- GThe inherent differences in the tdIIIOIOgY of Print ‘d ‘hat tion; with its steam-driven power (and later rotary) presses, which of broadcast, for example, led the Supreme Court to uphold the increased production tenfold, and its new modes of distribu- FCC’s regulation of “indecent” speech over broadcasting, be- tion, its assembly lme methods, and division of entrepreneurial cause of broadcasting’s “uniquely pervasive presence on the functions. Ithiel de Sola Pool, Technolo&”es of Freedom (Cam- lives of all Americans, ” and the fact that “prior warnings can- bridge, MA: Belknap Press, 1983). (Hereafter referred to as Tech- not completely protect the listener or viewer from unex~cted nolop”es of Freedom) and Encyclopedia Britannica, ‘‘Publish- program content, ” FCC v. Pacifica Foundation, 438 U.S. 726, ing, 1986. reh. denied, 439 U.S. 883 (1978).

5 6

\. T Y 4 +

Communications satellite SBS-3 being deployed from Franklin Press Space Shuttle Columbia

Today communication satellites distribute staggering amounts of information over thousands of miles in a fraction of the time needed for Ben Franklin’s press to print one page. press” has been extended to all forms of publishing, but real or perceived limitations on the technology-most notably, in the case of broad- New technologies will not only aug- casting-have nevertheless resulted in a patch- ment the capabilities of the press they work of exceptions to the freedoms originally may give rise to new forms of press, granted the printing press. alien to the last 200 years of First Amendment jurisprudence. However, even broadcasting, with its national and now global audience, and its compelling use of images from all over the world, did not change the “one-to-many” nature of journalism that has characterized the press since the first era of printing. Although the Observers argue that we are now entering number of broadcast organizations now far ex- an era that is variously referred to as an infor- ceeds that of newspapers, legal and economic mation age, an information society, or a post- barriers to entering and successfully compet- industrial economy. These phrases conjure im- ing as a broadcaster have perpetuated this one- ages of new technologies-high-speed com- to-many character. puters, global communications networks, “in- 7 —.

elligent” machines, and low-cost storage one message may be integrated into a whole media of astonishing capacity. But, more im- and distributed to an audience larger or smaller portant for present purposes is what these than the original. technologies mean for the structure of commu- New forms of publishing will grow up along- nications, and consequently, for the way in side the “mass” communication that we are which the press gathers and publishes infor- familiar with today. Individuals will be able mation. In changing the way in which infor- to select the subject matter of the information mation is produced and disseminated, technol- they receive, and determine its format and ogy may change who and what is considered manner of presentation. This processing may the press. ” New technologies will not only be done by consumers at intelligent terminals augment the capabilities of the press as we in their home or business, or it may be done know it today, they may give rise to new forms further “up the line, ” by the local telephone of press, alien to the last 200 years of First or cable company. Amendment jurisprudence. As a consequence of this shift from a cen- Taken together, current and anticipated ad- tralized press with a uniform product to a de- vances in technology suggest a fundamental centralized network selling diverse services, shift from the concept of “press” to the con- the courts and Congress may face new ques- cept of “network. To some extent, the past tions of constitutional interpretation not pre- mode of one organization publishing for many sented when the press was a more or less dis- may give way to a communications mode in tinct, identifiable institution. Even today, with which many share knowledge among them- the convergence of information processing with selves. One-to-many publication will no doubt telecommunications, questions of categorization continue, but will be joined by new and un- for purposes of the First Amendment–beyond familiar forms. Gathering, editing, and dissem- those of common carrier, broadcaster, and inating news and information, which today is print publisher-are emerging. commonly integrated in one organization, may eventually be fragmented between many specialized entities. The electronic publisher of the Like the underground press that flour- future may act more as a clearinghouse for the ished in the 1960s, the electronic un- exchange of news and information than as a derground press may become the cru- gatherer. Global electronic networks may even- cible of cultural change. tually allow the gathering, writing, filming, editing, and publishing of news to be decen- tralized among many organizations, which Technology will further challenge distinc- may sell one another specialized services. tions between the freedoms of speech and press, “nonmedia” and “media” that were al- One-to-many publishing will also coexist ready difficult to make. Whether there is a with one-to-one publishing, such as electronic difference between First Amendment rights mail, and many-to-many publishing, such as of speech and press is a matter of disagreement computer conferencing. Each of these permu- among scholars,7 but practical consequences tations may merge into the other under given circumstances—what started out as an elec- ‘Scholars disagree over whether the distinction between the tronic mail message may be integrated into a freedom of speech and the freedom of the press is of constitu- broadcast, which individuals may then store tional or legal significance. Former Supreme Court Justice Potter Stewart, for example, adheres to the view that freedom of speech, in a database, to be redistributed in different worship, assembly, and other liberties guaranteed by the Bill forms to different audiences. Specialized, in- of Rights are substantive and individual in nature, while the dividualized reports may be generated by one freedom of the press is 4 ‘structural, and extends freedom of expression to an institution. “The publishing business is, in organization for a few subscribers or patrons, short, the only organized private business that is given explicit and distributed over electronic mail. Pieces of constitutional protection. Potter Stewart, “Or of the Press, ” flow from it. Many of our 20th century assumptions about freedom of the press depend on no- Limitations on the liberty of the press tions of the press as a business, located in a may come from a failure to reconcile community, dependent on the support of readers and viewers, committed to a regular private ownership interests in the cycle of publication and broadcast, managed physical media with public interest in by responsible persons, and striving—to some the message carried over these media. degree-for objectivity, balance, and a fair presentation of the facts. What if those ele- ments change? aboard orbital satellites, or “mediasats,” together with an unprecedented ability to search, The Colonial press was characterized by ir- store, and process large amounts of informa- regular appearance, pseudonymous invective, tion on individuals may give rise to new con- and a boisterous lack of respect for any form cerns over the ongoing conflict between pri- of government. Modern, high-technology in- vacy and freedom of the press. ternational versions of this may flourish in par- The continuing debate over the deleterious allel with the established press. New interactive communications technologies can make effects of barriers to entry on the diversity of producing and disseminating underground news and information available, and the role ‘‘newspapers’ as economical as a phone call, of government in mitigating them, will inten- and as egalitarian as a New England town sify. The assumption of Classical Liberalism that the principal threat to individual liberty meeting. Like the underground press that flourished in the decades before and after the is from large public organizations-such as the First World War, and again in the 1960s, the Federal Government-may no longer hold. In- stead, limitations on the liberty of the press electronic underground press may become the 8 may come from a failure to reconcile private crucible of cultural change. The first simple ownership interests in the physical media with experiments of this sort are already underway on many computer bulletin boards. public interest in the message carried over these media. New questions of liability and potential for Courts and First Amendment scholars have harm will also be raised. Distinctions between developed a variety of ways of classifying and primary publishers, secondary publishers, and organizing the functions of the press for pur- republishers-which served to allocate respon- poses of analyzing First Amendment issues.1 sibility for defamatory publications-will be 9 I°For purposes of assessing the press’ regulatory status an more difficult to make. Locating the “source” shielding from Liability under the First Amendment, a distim of inaccurate or false information that causes tion is commonly made between the press as printed mediurr harm may prove elusive on electronic networks. as broadcaster, and as common carriers (which are not consid ered press). OTA Workshop on “The Future of the Press am The press’ use of remote sensing systems the First Amendment,” Mar, 26, 1987, Washington, DC. % also: Richard Neustadt, The Birth of Electrom”c Publish”n/ (White Plains, NY: Knowledge Industry Publications, Inc. 26 Hastings LtIw Jan-d 631 (1975). Former Chief Justice 1982); and Lynn Becker “Electronic Publishing: First Amend Burger disagrees, arguing that “the First Amendment does not ment Issues in the Twenty First Century, ” 13 Fordhaxn Urbax ‘belong’ to any definable category of persons or entities: it be- Law Journal 801 (1984). With regard to defamation, one text longs to all who exercise its freedoms. ” First IVational Bank book organizes constitutional privileges around notions of th~ of130ston v. Beflotti, 435 U.S. 765, (1978), Burger, J. concurring. press as the originator of speech, as commentator, or as repeater Whe “cultural revolution” of the 1960s found expression and Marc Franklin, Cases and Maten”afs on Mass Media Law momentum in a thriving underground press-’ ’[o]nly on the (Mineola, NY: Foundation Press, Inc., 1987), In analyzing Firs{ pages of the underground press, with its melange,of stories, Amendment tensions between rights of access and exclusion articles, events, hunches, graphics, fantasies, exposes and the another author distinguishes between the press as editor (mak ories can one find the Movement. Laurence Learner, The Pa- ing decisions about the content of messages, which decisions per Revolutionaries: The Rise of the Underground Press (New are protected by the First Amendment) and the press as ownel York: Simon & Shuster, 1972) p. 14. (making decisions about the use of the medium, which decisions 9w Kwton, D. Dobbs, R. Keeton, and D. Owen, ~osser ‘d are protected by the Fifth Amendment). Mark Nadel, “A Uni- Keeton on Torts (St. Paul, MN: West Publishing Company, fied Theory of the First Amendment: Divorcing the Medium 1984), pp. 799-811. from the Message, ” 11 Fordham Urban Law Journal 163 (1983). 9 —

ogies. These capabilities and press functions Familiar constitutional issues—pitting can be most usefully grouped as follows: freedom of the press against privacy Ž New technologies for gathering news and rights and national security interests information: databases and satellites Ž New technologies for editing news and in- —will be placed in unfamiliar contexts. formation: electronic publishing • New technologies for publishing and disseminating news and information: the conver- Because this report is concerned with the im- gence of computers and communications. pact of technological innovation on the press, The rest of this chapter will deal with the first however, press functions will be organized of these broad topics; new technologies for edit- along lines that reflect more or less discrete ing and publishing are covered in chapters 2 types of capabilities offered by new technol- and 3 respectively.

NEWSGATHERING In order to publish news and information, nied to the press, 14 but the government may he press must have an ability to gather it in not close down avenues for gathering and ac- he first place. The Supreme Court has yet to quiring news that are generally available to decide, however, whether newsgathering is it- the public, without a compelling reason. self a protected First Amendment activity, sep- Although the press may not, as a constitu- 11 arate from speaking and publishing. The tional matter, have any greater rights to gather Supreme Court said in Branzburg v. Hayes information than the general public, they may, that “it is not suggested that news gathering as a practical matter, have a greater and more does not qualify for First Amendment Protec- concerted ability to gather information than tion; without some protection for seeking out most individuals. In the coming years, tech- the news, freedom of the press could be eviscer- nology will greatly amplify the information- 12 ated. . . .” However, the Court has declined gathering resources of the press. As a result, to say that government has a positive duty familiar constitutional issues–pitting freedom to allow journalists special access to informa- of the press against privacy rights and national tion. The press has access to government pro- security interests—will be placed in unfamiliar ceedings, records, or other information that is contexts. Technology is likely to blur distinc- available to members of the public generally. 13 tions between gathering information and pub- Presumably, the converse is also true; access lishing it, and the Court will eventually have denied to the general public may also be de- to confront the question of whether the press

U.S. 1 (1978). See also, Rita Ann Reimer, Legal and Constitu- Mediasat would supply a stream of tional Issues Involved in kfd”asat Actiw”ties (Washington, DC: Congressional Research Service, The Library of Congress, 1987), timely information—peering where Report No. 86-823A, pp. 6-8. repressive governments or dangerous 14 When, in 1983, the United States invaded Grenada, the natural environments had formerly government imposed a total news blackout and prohibited members of the public and the press from traveling to Grenada. The kept the press at bay. press sought prospectively to enjoin the Executive from im- posing any such future ban. The case was dismissed as moot, but the court went on to say that “[the] decision whether or not to impose a press ban during military operations and the llBrmzburg v. Hayes, 408 U.S. 665 ~1972~ nature and extent of such a ban if imposed are matters that “Ibid., p. 681. necessarily must be left to the discretion of the commander in “Pen v. Procum”er, 417 U.S. 817 (1974); Saxbe v. Washing- the field. ” Flynt v. Weinberger, 588 F. Supp. 57,61 (D.D.C. 1984), on Post Co., 417 U.S. 843 (1974); cf., Houchins v. KQED, 438 affirmed (on the basis of mootness), 762 F.2d 134 (D.C. Cir. 1985). 10

interests in gathering news merit constitu- ellites. The use of these technologies may raise tional protection under the First Amendment. a number of constitutional questions concer- Among the new tools that the press will have ning, for example, individual privacy rights, at its disposal for gathering information are national security, and the expungement of computer databases and remote sensing sat- records.

COMPUTER DATABASES Databases are records of information stored The database industry is growing at a rate in machine readable form, and are typically of 15 to 18 percent a year. Total revenues were accessible by personal computer over packet- estimated to be $2.2 billion in 1986, and are pro- switched data networks (e.g., Telenet, Tym- jected to be more than $4.2 billion by 1990.16 net, or a variety of privately owned communi- In 1986, there were 3,200 online database serv- cations networks). Information in electronic ices available worldwide compared with 400 databases is usually searched and retrieved by in 1979.17 software provided by the vendor of the data- In principle, anything that can be repre- base service. sented in digital form can be stored and For purposes of this report, the concept of retrieved from a database. Today, this includes computer databases is kept separate from that the full text of newspapers, magazines, jour- of electronic publishing, which is discussed in nals, and publications from any major scien- chapter 3. In reality, database vendors are, by tific or professional discipline. In the future, definition, also electronic publishers, since they computer software, motion or still pictures, publish information in an electronic, machine- and high fidelity music and voice will be stored readable form. However, electronic publishing on computer and shipped over telephone or ca- is a broader concept, which includes transac- ble television lines. Existing database services tional services (i.e., banking and merchandis- are categorized as bibliographic and abstrac- ing) and messaging services that are beyond tive (e.g. the Library of Congress’ SCORPIO); those available from database vendors. Per- full text (e.g. Mead Data Central’s LEXIS); haps the distinction is easiest to make in the or numeric (e.g. Data Resources, Inc. ’s DRI- following way: database vendors provide in- Securities and Exchange, Current Economic formation; electronic publishers provide infor- Indicators, Bank Analysis Service, and Finan- 18 mation services, which may include database cial and Credit Statistics). 15 access. Optical disk storage (principally, Compact Disk Read-Only-Memory, or CD-ROM), with its extremely high capacity, low cost, and dura- bility, is likely to change both the economics 15Terminology for these new forms of publishing is still un- of online databases, and the type of informa- settled. In the “Huber Report, ” the author distinguishes be- tion sought on them. Optical disks offer much tween “information service providers, ” who offer call and net- more powerful searching software, and avoid work management services, timeshared computing, access and retrieval systems, messaging systems, and transactional serv- rising telecommunications costs. In the future, ices, and ‘‘computerized databases and electronic publishers, who offer retrieval of documents, data and text from magnetic or other storage media. The Geodesic Network: 1987 Report on Competition in the Telephone Industry, prepared by Peter ‘GOn/ihe Database Systems Market in the U.S. (#1517), Frost Huber as a consultant to the Department of Justice, Antitrust and Sullivan, Inc. 1986, as quoted in Information Hotline vol. Division, January 1987, chs. 6 and 7. Other authors make the 19, No. 2, February 1987, p. 3. distinction between online databases and videotex. See, e.g., ‘TFrom Information Industry Association estimates. Hugh E. Look, Electrom”c Pubh”shing-A Snapshot of the Early l~Thi9 taxonomy was put forth by Martha Williams, ‘ ‘Elec- 1980s (Oxford, England: Learned Information, 1983). tronic Databases, ” Science, vol. 228, Apr. 26, 1985, pp. 445-456. ——-——

11 — news organizations might make extensive use Databases–whether online or on disk–are of optical disks instead of the online services powerful tools for the press. Researching sto- employed in the past. 19 ries, investigating the background of subjects Today, there are many “gateway” services, and sources, corroborating information, drawing out latent connections between people and in which one database vendor sells the infor- events, and constructing “mosaics” of information services of another. A possible scenario mation from disparate sources, will all become for the future database industry is the emer- more practicable, and in some cases, possible gence of a “meta-library” or “virtual data- for the first time. base, ” which would interconnect many or all of the now separate database services. Such It is safe to say that, by the end of this dec- a system might allow a user to enter a query ade, every recent news story, news picture, wire in everyday English on any given subject, con- service report, and major press release will be duct a search for the location of the informa- commercially searchable from the reporter’s tion, and retrieve it. Achieving this goal de- workstation, subject only to the cost his news- pends on, among other things, considerable paper is willing to incur. The press will have technical progress on computer memory, trans- a long and comprehensive memory. mission, and processing speeds; advances in computer software; and the development of standards controlling how computers and networks ‘‘talk’ to one another. Current technological trends suggest that this scenario is obtainable. 20 Much depends on the establishment of standards.21

190pticaVEJectrom”c Publishing Directory 1986, Learned In- formation. Online databases are likely to remain valuable for tim~sensitive, transient information, while CD-ROM will be favored for archival information. See also, Bradford Dixon, “The Impact of CD-ROM on On-Line Data Bases, CD-ROM Rew”ew, vol. 1, October 1986, p. 52. zoGreat prowess is being made in the ability tO tr~smit large quantities of information very rapidly. INTELSAT satellites can now transmit the equivalent of 20 copies of the En- cyclopedia Britanm”ca every minute. J. Pelton, M. Perras, and A. Sinha, INTELSAT, The Global Telecommum”cations Net- work (Honolulu, HI: Pacific Telecommunications Conference, FCC, in its Computer Inquiry III, Phase I Decision, mandated 1983), p. 17. Depending on bandwith required, fiber optic com- the adoption of an “Open Network Architecture” standard, with munication cables can now transmit the entire contents of a the goal of facilitating a total free market in telecommunica- CD-ROM disk–the equivalent of 200,000 pages of text–in tions and information services. See: A,M. Rutkowski, “Open about 1 second. Science and Technology in Japan, October- Network Architectures: An Introduction, ” Tekwommum”cations, December 1986, p. 8. February 1987, p. 29. Voluntary communications standards, Z! These stmdmds concern the way in which information ~d called Open Systems Interconnect, have already been estab- instructions are communicated to and within a network. The lished by the International Standards Organization.

MEDIA SATELLITES’* The news media are increasing the use of sat- wish to own and operate their own remote sens- ellite imagery in reporting world events. This ing systems dedicated to newsgathering— leads some to believe that the press will soon dmn, OTA-TM-ISC-40 (Washington, DC: United States Gov- ‘2The issues concerning media satellites are examined in ernment Printing Office, May 1987). This section draws solely depth in U.S. Congress, Office of Technology Assessment, Corn- on the analysis of that publication, hereafter referred to as News- mercial Newsgathering From Space—A Technical Memoran- gathering from Space,

80-6/i’) - 87 - p 12 termed “mediasat.” Mediasat would supply To be effective, a mediasat would need more a stream of timely information—peering where than high resolution; it must also be able to repressive governments or dangerous natural sense news wherever and whenever it occurs environments had formerly kept the press at and to transmit the news rapidly to the news bay. agency. A mediasat system would need at least two satellites to ensure same day coverage of events around the globe. In order to receive Many nations will have their own re- data in near-real-time, a mediasat system would mote sensing systems within a decade, need to have access to ground stations all over and the press might purchase data the earth and use on-board tape recorders. from, or invest in, these foreign systems. Satellite and database technologies may change the very meaning of “pub- Today’s press obtains data from two remote lic” and “private.” sensing systems, EOSAT—formerly the U.S. Government Landsat system–and SPOT, a 23 French system. Neither of these systems, however, is particularly well suited to the needs Although the technology is available to cre- of the press; the resolution of these sensors ate such a system, including very high resolu- may be too low (EOSAT yields a maximum tion technology, the high cost and currently of 30 meters resolution, and SPOT a maximum low demand for remotely sensed data will limit of 10)24 and their coverage of a given point on media efforts to own and operate a dedicated the Earth too infrequent for them to be a remote sensing satellite system. Moreover, the timely, valuable, and reliable source of infor- value of satellite imagery to the press is un- mation for the press. In addition, the press’ certain, and is likely to remain so until experi- access to data cannot be assured because the ence and a more robust remote sensing mar- satellites’ owners currently depend on ground ket combine to define a stable demand for these stations owned by other countries to collect data. preliminary calculations indicate that the certain data. It is likely, therefore, that the costs of a mediasat system might exceed its press would require a dedicated mediasat sys- expected revenues. Notwithstanding the considerable barriers tem to-—— meet their needs. - ‘EOSAT (Earth Observation Satellite Co. ) is a private com- to implementing a mediasat system, it may at pany which now handles the operation and marketing of data some point in the future become a viable con- for Landsat, which was formerly owned by NASA. For a his- cept. The press might form a consortium to tory of the transfer of the Landsat system to the private sector, see: U.S. Congress, Office of Technology Assessment, Re- share the cost of a mediasat system, or it might mote Sensing and the Private Sector: Issues for Discussion—A resell the data collected by mediasat to subsi- Technical Memorandum, OTA-TM-ISC-20 (Springfield, VA: Na- dize its own use, or it may wait until technical tional Technical Information Service, March 19841. SPOT (Sys- tem Probatoire d ‘Observation de le Terre) is a French company advances reduce the cost of sensors, satellites, responsible for marketing the data from the satellite owned by and launch vehicles. Moreover, many nations the Government of France. will have their own remote sensing systems ‘iThe greater the resolution of the sensing system, the more detail that can be discerned. The SPOT system, for example, within a decade, and the press might purchase allows one to see individual buildings and highways. data from, or invest in, these foreign systems.

IMPLICATIONS FOR PRIVACY

When the press gathers and publishes infor- tion as source material, there is a potential for mation about a person, or uses such informa- conflict between the individual’s common law 13 right of privacy and the right of the press to privacy interests, the concept of privacy de- gather and publish news and information. veloped as common law is most often asserted Databases store many pieces of personal in- against the press. 29 formation, and permit the creation of larger Satellite and database technologies may mosaic pictures of the individual from these change the very meaning of “public” and “pri- pieces. Satellites allow information to be ob- vate.” As remote sensing satellites become tained without individuals’ knowledge, and more sophisticated, for example, it is possible without physical intrusion or proximity. Be- that the average person’s expectation of cause of the way in which these technologies privacy could be eroded. Satellites are cur- enhance the newsgathering ability of the press, rently capable of spotting certain crimes, such they create a potential for conflict between as violations of environmental control laws. privacy and the First Amendment. Eventually, satellites may be able to perform other functions, such as identifying and locat- Until recently one of the best barri- ing marijuana fields, or determining the inven- tories of manufacturers. In the far future, sat- cades against breach of privacy was ellites may be able to monitor the activities the difficulty and impracticability of of individuals. integrating all of the public data about a person. Under current law, a person is protected against publicity given to facts about his or her private life. Although hard to define, the It was in fact the press, and reactions to the protections afforded by this right to privacy are clearly reduced when a person appears in press, that first precipitated legal recognition 30 of an individual right to privacy. 25 Privacy is public. Mediasat could alter the current un- a word that embraces a number of separate derstanding of what the law regards as “ap- pearing in public. ” Recently in California v. but similar values. It has been variously de- 31 fined as “the right to control information about Ciraolo, the Supreme Court decided that oneself; ”26 “ the claim of individuals, groups aerial reconnaissance was an acceptable law or institutions to determine for themselves enforcement technique and that activities tak- when, how and to what extent information ing place in the defendant backyard were in about them is communicated to others;"27 or “plain view, “ even though they were surrounded simply, “the right to be let alone."28 Although by a 10-foot-high fence. Applying Ciarolo’s State statutes have recognized a panoply of logic broadly, one could argue that citizen’s have no right of privacy for any activity that might be seen from an airplane or satellite. Computer databases may also change the - ““The wminal article hy l,ouis 1). Brandeis and Samuel ktrar- meaning and expectations of privacy, The ren, ‘Arrhe Right to Pri\’acy, ‘‘4 Harvard Law Retriew 193( 18!30), press may take advantage of the storage, re- which structured subsequent debate, litigation, and legislation on pri~’ac~ in the [Jnited States, was a reaction to the editorial practices of the E30ston newspapers: “The press is overstepping in ever-} direction the ob~~ious bounds of propriety and decency ‘qCommon law invasion of privac~’ is subdi~ided into f{jur .,. To occup~r the indolent, column upon column is filled with separately actionable torts: intrusion, disclosure, false light, and idle gossip. ~’hich can only be procured by intrusion into the appropriation, William I.. Presser, ‘‘ Privac~’, 48 C:ilif(jrnia l,a }{’ domestir circle. ., ,” W’herr the Warren/Brandeis theory of l?e~’iew. 383 (1 960). Of particular concern for the presf’nt ciis- prit’acy was r~’jected in the first major case to consider it, f70ber- cussion is the tort of ‘public disclosure of private facts, which son ~, Hoche.qter [’olding l~cj.y [’o., 171 X.Y. 538, 64 N.17. 442 requires that the information made public b~’ the prx)ss be in

( 1902), the VCW Yrork legislature reacted by creating a statu- fact pri~ate, that the disclosure be highly offensi~(’ t () ii r(~as(~n- tory right of prl~a[’}, New York Civil Rights I.aw, $ j 50 and51, able person, and that the subject matter of the diw’losure not “This definition forms the basis for the Privacy Act of 1974, be of legitimate concern to the public. I’ublic I,aw’ 93-579, 5 (J. S. C’, \5fi2{a), “)l]anson, libel and Related Torts, !260 ( 1969). -;Alan Jt’est in, i)ri~art’ and Frwxiom (1’Y’Pw York: Atheneum, ‘1 106 S. Ct. 1809 ( 1986). Ciraolo was a criminal case invol\- 1 967), p. 39. ing a warrantless search. As such, its reasoning ma~’ not be ‘I)ran[lei+ an(i \4’arr[~rl, op cit dirf’ctl~’ applicable to rit’il suits for in~. asion of prilrac~’. 14 trieval, and processing capabilities of modem performed by the reporting of the true con- computer technology to construct comprehen- tents of the records by the media. The free- sive pictures of an individual from a myriad dom of the press to publish that information of transactional details-much as a mosaic appears to us to be of critical importance to painting is constructed from smaller pieces of our type of government in which the citizenry is the final judge of the proper conduct of pub- no artistic significance in and of themselves. lic business.35 This capability was not practical in the world of print, where storing, retrieving and collat- Whether this First Amendment right of the ing a mass of trivial detail was inefficient and press to publish court records also applies to wasteful, even when possible. Computer data- other public records and publicly available bases permit one person to fabricate whole new information in general remains to be deter- bodies of knowledge out of heretofore uncon- mined.36 The problem has not been of wide- nected pieces of information. Much of the in- spread concern, in part because until recently formation about an individual’s life is not, one of the best barricades against breach of when taken in isolation, intimate or confiden- privacy was the difficulty and impracticabil- tial. Purchasing merchandise at a department ity of integrating all of the public data about store, traveling on holiday, visiting the doctor, a person. Records such as arrest and prosecu- joining an association, reading a newspaper— tion data, credit status, purchases, mortgages these activities are often done in the open, and and property records, hospital admissions, are available to anyone who cares to watch. travel information, associational behavior, banking activity, and previous appearances in Although the press may, in certain instances, newspapers or on television were either uneco- be liable for the collection or publication of per- nomical to keep for long periods of time, inac- sonal information, this liability may conflict cessibly “buried” with thousands of other with the freedom of the press, especially when records, geographically dispersed, or not cross- the information collected is available through referenced. 37 public sources. In Cox Broadcasting Corp. v. Cohn, 32 the Supreme Court struck down a The conflict between privacy and press free- Georgia statute that barred publication of the doms may be most acute in cases where the names of rape victims. Although the Court rec- government acts on behalf of the individual’s ognized that “there is a zone of privacy sur- privacy to foreclose the revelation of informa- rounding every individual, a zone within which tion that is normally public. In the case of sat- the State may protect him from intrusion by the press, “ it said that State may not censor “judicial records which are maintained in con- S5420 U.S. 469 (1975) at 32. sGIn Vhgy v. Time, Inc., 527 F.2d 1122 (9th Cir. 1975), cert. nection with a public prosecution and which denied 425 U.S. 998 (1976), the Ninth Circuit Court rejected 33 themselves are open to public inspection.” the notion that the press has a First Amendment right to pub- The Court limited its ruling to court records,34 lish all private facts that are publicly available (in this case, facts drawn from an interview of the plaintiff). Only if the facts but also said that: are newsworthy or of legitimate public concern would the press’ First Amendment defense obtain. Public records by their very nature are of sTThe new power to aggregate information was Mustrated interest to those concerned with the adminis- last year when two prominent businessmen were competing pub tration of government, and a public benefit is licly to buy a major U.S. newspaper company. An enterprising journalist ran a check on both their names in Mead Data Cen- tral’s NEXIS database, and learned that one of them, who lived in Indiana, was married to the sister of the investment banker representing the target newspaper’s interests. He also learned ‘]’420 U.S. 469 (1975). that the Indiana man had just returned from a weekend with 3sIbid. his wife in Mexico City, where the competitor lived. The two 34 The court declined to decide the more gener~ question men announced a few days later that they were joining forces whether “the State may ever define and protect an area of to buy the newspaper together, leading the journalist to report privacy free from unwanted publicity in the press, ” and instead that the businessmen were colluding, rather than competing. focused on the narrower question concerning publicly available Christopher Bums, “Freedom of the Press in the Information judicial records. Age, ” OTA contract report, Apr. 21, 1987, p. 20. —

However, because more and more information is created or accessed in machine-readable State policy concerning the expunge- form, and because the costs of storing it in that ment of records on individuals maybe form are plummeting, newspaper morgues are ineffectual, because it must yield to growing in size and comprehensiveness. The freedom of the press. question therefore arises whether the press will be required to expunge records that were once public and are now kept in its own files. Re- ellite surveillance, for example, government cent Supreme Court decisions suggest that no might forbid press acquisition of private or pro- such requirement could be imposed on the prietary pictures through remote sensing. This press, especially if it carries a criminal pen- would undoubtedly be assailed as a “prior re- alty.40 In some cases, State policy concerning straint” on press freedoms. The doctrine of s the expungement of records on individuals may prior restraint holds that advance limita- be ineffectual, because it must yield to free- tions on protected speech may not be “predi- dom of the press.” cated on surmise or conjecture that untoward consequences may result."39 Constitutional When information is not only stored and ac- issues concerning prior restraint arise most cessed, but also provided by the press elec- often where the government acts to protect na- tronically, further complications ensue. In in- tional security. teractive electronic services which provide information and services to the home, the use The State or Federal Government may act of the service by the consumer/reader also gen- on behalf of individuals’ privacy rights in ways erates information for the provider about read- that conflict with press access to information ing, viewing, and consumption patterns. This held in public or private databases. Many information can in turn be used to target and States, for example, have “expungement stat- tailor information, such as advertising, that utes that apply to criminal or other records, is fed back to the consumer. According to one which typically require that records be de- author: stroyed after a certain period of time, or when a defendant in a criminal case is acquitted. The Every transaction which is executed, and theory behind the statutes is that an offender, every page of information or service which is once rehabilitated, deserves a chance to be free delivered, will generate its own electronic (machine-readable) record. The return channel of his past and start anew. in an interactive system will perform double labor for the interactive services industry; not only will it facilitate consumer requests for If interactive electronic services are services, and thereby stimulate consumption, considered analogous to cable opera- it will also transmit back to industry much tors or newspapers, then they may, relevant information concerning the modali- like cable and newspapers, claim First ties of consumer demand and consumption. 42 Amendment protection. Early concerns about privacy pertaining to the collection and sale of transactional information generated by the electronic press were ad- ‘“A prior restraint is government censorship. Other forms of iability or punishment for speech are imposed after the harm :aused by the speech has occurred. Prior restraint, in contrast, ‘OSee Landmark Communications v. Virginia. 435 U.S. 8’29 s a prohibition of speech or publication before it occurs. “Any (1978); and Smith v. Daily Mad Publishing Co., 443 U.S. 97 ;ystem of prior restraints of expression comes to this Court (1979). )earing a heavy presumption against its constitutional valid- 41 ShiffJet v. Thomson Newspapers, Inc., 431 N.E.2d 1014, 8 ty. ” Bantam Books, Inc. ~. Sullivan, 372 U.S. 58, 70 (1963); Med. L. Rptr. 1199 (1982). ;ee also New York Times Co. v. United States, 403 U.S. 713 ‘iKevin Wilson, “The Videotex Industry: Social Control and 1971 ), and Near \r. Minnesota, 283 U.S. 697 ( 1931), the Cybernetic Commodity of Home Networking, ” Media, Cui- ‘~’]New York Times Co. i’. United States, 403 U.S. 713 (1971), ture, and Socie.t~’, vol. 8, 1986, pp. 7, 25. (Hereafter referred to Justice Brennan concurring at 724. as “Cybernetic Commodity.”) 16 — dressed in the Cable Communications Policy duction and consumption.”46 The difference Act of 1984.43 And, the Privacy Commission, between such cybernetic control and the fami- whose recommendations were in large part liar television or newspaper advertisement may adopted by Congress in the Privacy Act of be simply a matter of the degree of precision 1974, suggested principles for the use of infor- and power that electronic systems provide. mation gained through interactive electronic Moreover, a greater sensitivity to consumer services. However, the Supreme Court has preferences may be generally desirable. Limit- decided that individuals have no inherent le- ing the cybernetic control of consumer prefer- gal interests in personal records owned by third ence is more likely to be a political decision than parties.” a judicial issue. However, a question could also But privacy may only be part of the prob- arise as to whether the collection and feedback lem. The use of interactive information sys- of information through interactive services in tems to provide the press with a precise con- protected speech, and thus whether this “cyber- sumer stimulus/response mechanism suggests netic cycle” is an activity protected by the to some that “improved techniques of social First Amendment. If interactive electronic management are on the technological horizon services are considered analogous to cable oper- . . . creating a truly cybernetic45 cycle of pro- ators or newspapers, then they may, like cable and newspapers, claim First Amendment pro-

] tection. What information to provide to which ‘ 47 U.S.C. j631. “’U.S. v. Miller 425 U.S. 435 (1976). consumers may be a matter of editorial dis- “’The term, “cybernetic, comes from control theory and cretion protected by the First Amendment. 47 refers to systems that are highly adaptive, responding to their environment by sensing changes and responding by altering the environment or their response or both. In this case, a cyber- “’’Cybernetic Commodity” at page 35. See also Deann netic cycle is one in which the electronic media, by virtue of Collingwood-Nash, and John Smith, Interactive Home Med~ its individualized and rapid interaction, not only adapts itself and Privacy, report prepared for the Office of Planning, U. f to individual consumers wants, but also acts to influence those Federal Trade Commission, 1981. wants, ‘7 Miamj I{erdd ], 7’Orni)lo, 418 U.S. 241 (1974).

IMPLICATIONS FOR NATIONAL SECURITY Where the press seeks to gather information concerning national security, whether through The Federal Government may at- satellite surveillance or computer databases, there is a potential for conflict between national tempt to limit access to or use of sat- security policies and the First Amendment. ellite imagery by the press. With satellite images, for example, the press could: ● reveal facts about an unfolding crisis making it more difficult for government disseminate information regarding U.S. leaders to act calmly and responsibly; and military operations, thereby depriving • misinterpret satellite data in such a ways U.S. troops of the critical element of 48 surprise; as to precipitate a crisis. reveal information considered sensitive by In response to these potentials, the Federal foreign governments, thereby prompting Government may attempt to limit access to them to retaliate against U.S. Govern- or use of satellite imagery by the press. In the ment activities, assets, or personnel; provide valuable intelligence to countries currently lacking their own reconnaissance satellites; ‘“Taken from OTA, ,Vew’slathering From Space, op. cit, p. 4. .

case of a dedicated “Mediasat,” it might do immediate, direct, and irreparable harm to the o permanently, through the licensing proce- interests of the United States"54 (emphasis dres established in the 1984 Landsat Act,49 added). or temporarily, during a crisis, by limiting the Although the Progressive case did not in- resolution of the satellite’s sensors, the images volve computer databases, the enhanced abil- the satellite is allowed to collect, or the images ity of computer systems to achieve the same the press is allowed to disseminate. Any of “aggregation and synthesis” of unclassified these options may run afoul of the doctrine of materials was at the heart of a recent public prior restraint. Prior restraints are allowed debate concerning National Security Decision only if necessary to prevent “direct, immedi- 55 Directive 145 (NSDD-145). The efficiency ate, and irreparable damage to our Nation or 50 with which online databases can construct mo- s people." The outcome of such a challenge would turn on the exact nature of the Govern- saic information was, in part, the rationale be- ment limitations and the Supreme Court ulti- hind NSDD-145, which, among other things, mate determination of the status of news- established a‘ sensitive, but unclassified’ category for information in government data- gathering under the Constitution. bases, and perhaps privately owned commer- Similar First Amendment difficulties may cial databases as well.56 NSDD-145's focus on e encountered with attempts to suppress or electronic storage and retrieval systems rec- limit access to information in computerized ognized that databases store information that databases. The ability of electronic informa- ‘‘even if unclassified in isolation, often can re- tion systems to construct revealing mosaic pic- veal highly classified and other sensitive in- tres from many smaller pieces of information formation when taken in aggregate.’’57 has many parallels to the situation underlying the case of United States v. The Progressive, The Federal Government proposed taking Inc.51 In that case, The Progressive magazine measures to protect sensitive information from proposed to publish an article on ‘The H-Bomb hostile governments, including screening data- Secret: How We Got It, Why We’re Telling base entries, precluding the electronic publication of certain databases, providing database It," which was derived entirely from public do- subscriber lists to the government, and/or lim- main, unclassified sources. The Federal Gov- 58 iting foreign subscriber access. Some of ernment sought, and was granted, an injunc- these proposed measures may raise constitution barring publication of the article under 59 the Atomic Energy Act.52 Notwithstanding tional issues. the fact that the injunction constituted prior ‘ 167 F’,supp. 990, tit 996. restraint--the most severe abrogation of First NSI)I)- 145, Sept. 17, 1 9N4. ‘ ,4s of ,J anuar~ 1987, the status of N S1)[1- 145 ~it h rt’sp~’ct Amendment rights–and the fact that most of tG commercial databases had not been clarified. Ipor a m(~rt’ the research for the article was done in freely detailed discussion of NSDD- 145, and related executi~t’ and Accessible government libraries,53 the court 1egislati\e actions, see part 11, ch. 4 of this report, “TNational Polic~ on Protection of Sensiti\e but (Unclassified held that the various sources of information, Information in Federal (;o\’ernment Telecommunicate ions an(i "when drawn together, synthesized and col- Automated Information Systems, NTISSP No. 2, (M, 29, 19F(i lated. . . . acquires the character of presenting (hereafter cited as ‘ ‘National Policy’ ‘). This perception t~ a~ echoed in a C 1 A Report, .%~’iet Acquisition of .Ililitaril.} .>’iL~- nificant Western Technolog-j-: An Update See part 1, ch. :1 of this report, ‘{ 1,; [J. S,C, t }420 1-4292. ‘HNASA has alread~’ implemented an access policy which re- “’,\-ew I’ork ‘Nme.s CO. t. 1‘njfed .State.$, -$():J [ 1.S. 7111 (1971), stricts access with respect to foreigners, and members of the usticws Stewart and \$’hite, concurring Air Force have visited several commercial database ~’endor< ‘46’7 F. Supp. 990 (W’. 11. Wrisc. 19’79), The fb-o~p-essi~re case asking for subscriber lists. See Scientific Informat i~)n. + discussed i n part 1. ch 2, of this report. “’9For example, the Federal Government has a ri~ht to (’rt’- ’68 stat. 919.42 U.s. c. 2011-2296. ate conditions on access to, or publication of, go~’c’rnr]}t’~lt:~ll?r Sfw: Rita Ann R[~imer, Legal and Con.stit utiona] I.s,sues ln- funded information, so long as they do not conflict with the ol~’(d in .Iledia.sat 1 cti~’ities (Yf’ashingt on. 1)(’: (’on~rres+ional k’reedorn of 1 nformation Act., ,5 [1, S.C. \,552(a). hforeot’er. as we {vsear{h Ser\rice. ‘l’he 1,ihrary’ of (’{)n~rt,<+, 19H7 I, l{~p[)rt No. ha~e seen, the Supremt’ (’ourt has not recognized a constitu- { f;-,< 2~~ .+~, p 12 tion] right of the press to ~zther information. so it is unlikel}” 18

proaches to national security problems of data- The Federal Government proposed base mosaics may pose constitutional issues. In the case of NSDD-145, for example, sensi- taking measures to protect sensitive tive but unclassified information is defined as information from hostile governments, “information the disclosure, loss, misuse, al- including screening database entries. teration, or destruction of which could adversely affect national security or other Fed- eral Government interests.”60 But database Prior restraint on commercial database pub- vendors may have little clue as to whether the lishers raises peculiar problems for the tradi- data that they sell could “adversely affect national theories under which prior restraint is tional security” until it is conjoined with other permissible. Individual database entries do not data; the “sensitivity” of information in a data- necessarily pose a threat to national security base will depend on its combination with other that would justify restraining them. It is os- information. Attempts to control disaggregated tensibly the concatenation of individual data- data may therefore run afoul of the “vague- base entries that raises national security con- ness ” doctrine, which is based on the due proc- cerns, but this concatenation may not be ess clause of the 14th Amendment, and which specifiable before a given database search. Yet requires that a statute “neither forbid nor re- judicial precedent with prior restraint has, quire the doing of an act in terms so vague that without exception, concerned a single publica- men of common intelligence must necessarily tion, the contents of which could be known guess at its meaning and differ as to its appli- ahead of time. cation."61 Vague laws may also infringe upon the First Amendment rights of the press by Prior restraint issues normally arise when “chilling” protected expression. the government seeks an injunction prohibiting publication. But, even alternative ap- continued from previous page 61)~rom ‘‘Nat,ion~ po~cy} that the press, as database users, would be able to assert a con- “ Section II—Definition. 61zwej~er v. ~CW~~, 38$) U.S. 241 (1967). stitutional interest in access to sensitive information. Chapter 2 New Technologies for Editing and Selecting News and Information

Far more information is produced than con- It does not follow from this, however, that smed. Measuring information in words, one editing will continue to be a centralized activ- census found that, on average, over 11 million ity, nor one conducted solely by people. Quite words are made available to Americans over the contrary: the deluge of information made print and electronic media each day, and this possible by electronic publishing technology figure has been growing at roughly 8 percent may require technological methods of sorting each year since 1980. Of these words supplied, information for relevance and importance. only about 48,000 were actually read or heard Technologies such as expert systems are emerg- by any one individual (on average). 1 Editors, ing, which when used in conjunction with elec- whose task it is to gather, select, and organize tronic publishing, can disperse editorial control this welter of information for particular pur- to the recipients, rather than the originators, poses or consumers, are more crucial to the of news stories and information. Citizens may information consumer than ever. As one com- come to see “news stories” not as a standard- mentator has observed, “in the Age of Infor- ized, authoritative, and “balanced’ text, but mation, editors assume an even greater impor- as a largely self-selected collection of sources tance; people will pay not to be deluged with to be assayed in context. reedited data.”2 The emerging electronic publishing industry will pose unique questions concerning For the electronic publisher, exercising editorial control and responsibility under the responsibility for identifying, verify- First Amendment. Traditional assumptions ing, and policing sources of informa- about the press’ editorial responsibility for the tion may become difficult or impossi- information it publishes may be drawn into ble, raising questions about standards question. The First Amendment serves to in- sulate responsible press conduct from liabil- of liability. ity in the interest of robust debate on matters of public importance. But for the electronic publisher, exercising responsibility for identifying, verifying, and policing sources of in- IIthiel de Sola Pool, et al., Communications Flows: A Census n the United States and Japan (New York, NY: Elsevier Sci- formation may become difficult or impossible, mce Publishing Co., 1984), p. 16. Survey years 1960-1980. raising questions about standards of liability 2Bruce Owen, “The Role of Print in an Electronic Society, ” for what is carried over phone and coaxial ca- ;ommum”cations for Tomorrow: Poh”cy Perspectives for the ‘980 ‘s, G. Robinson (cd. ) (New York, NY: Praeger Publishers, ble lines. Electronic publishing also challenges 1978); as quoted in Mark Nadel, “Editorial Freedom: Editors, traditional distinctions between publishers and ?etailers, and Access to the Mass Media, ” 9 Comment: Hast- common carriers, further complicating ques- ngs Journzd of Communications and Entertainment Law 213 winter 1987), which provides a useful discussion of the role and tions of liability and First Amendment pro- inctions of editors. tections.

ELECTRONIC PUBLISHING Electronic publishing combines information point communication system in which text, au- access and retrieval capabilities with messag- dio, or video information may be carried by ing and transaction services. It is a point-to- telephone network, microwave transmission,

19 satellite relays, or even coaxial cable television home shopping and home banking.8 Although lines. It is roughly equivalent to what the Jus- there are substantial differences between the tice Department calls “information services” types of electronic publishing (teletext, for and the Federal Communications Commission example, is broadcast, while videotex is tele- calls “enhanced services.”3 Within the ambit phonic), they will be referred to collectively as 4 5 9 of electronic publishing are teletext, videotex, electronic publishing.— electronic mail,6 electronic bulletin boards,7 and electronic transactional services, such as Excluded from present consideration is broadcast radio or television, cable television unenhanced voice telephony (i.e., exclusive of S1n um”t~ States V. A7’&T CO., 552 F.supp. 131, 181 (D-c. Cir. 1982), affirmed 103 S. Ct. 1240 (1983) (Modified Final Judg- voice-mail storage and forwarding systems) ment), Para IV J, “information services” were defined as “the and conventional print publishing. The distinc- offering of a capability for generating, acquiring, storing, trans- tion between electronic publishing and conven- forming, processing, retrieving, utilizing, or making available information which may be conveyed via telecommunications. tional communications, such as TV or radio In Computer 11, 77 F.C.C. 2d 384, at 498, “enhanced services” is somewhat artificial, but suited to present are “services, offered over common carrier transmission facil- analyses. In the long run, the convergence of ities used in interstate communication, which (1) employ computer processing applications that act on the format, content, modes of communication will destroy distinc- code, protocol or similar aspects of the subscriber’s transmitted tions between electronic publishing and other information; (2) provide the subscriber additional, different or forms of publishing; “the one-to-one relation- restricted information; or (3) involve subscriber interaction with stored information. ” ship that used to exist between a medium and ‘Teletext is a one-way broadcast transmission of textual in- its use is eroding. This is what is meant by the formation, using the Vertical Blanking Interval (VBI) of a con- convergence of modes."10 ventional television signal. Data is transmitted to a subscriber’s television from a central computer, and the viewer selects from Electronic publishing may look like a differ- “pages” of information. Teletext requires a decoder, attached to a television set, to be received. The most common form of ent kind of service at different times, depend- teletext transmission in the United States is closed-captioning ing on the use the subscriber makes of it. It for the hearing impaired. See: Efrem Sigel, The Future of may function like conventional mail at one Videotext (White Plains, NY: Knowledge Industry Publications, 1983); Richard Neustadt, The Birth of Electronic Fublish”ng time; book or newspaper publishing at another; White Plains, NY: Knowledge Industry Publications, 1982); a bulletin board with messages pinned to it at Neustadt, Skall & H amrner, “The Regulation of Electronic Pub- another; a conference room at another; or even lishing, ” 33 Federal Communications Law Journal 331 (1981); and Anne Branscomb, “Electronic Publishing: A Global view a bank or retail store at another. of Videotex, ” 36 Federal Commum”cations Law Journal 119 (1984). svi~eotex differs from teletext in SWI?rd ways: ‘t ‘‘inter- In the long run, the convergence of active, “ in that information flows in two directions, from the user to the host computer and back again; information is typi- modes of communication will destroy cally organized by search, rather than by preset pages; there distinctions between electronic pub- is no theoretical limit on the amount of information that can be transmitted and stored; and videotex is normally transmitted lishing and other forms of publishing. via contained media (i.e., telephone or cable lines) in digital, computer-processable form. For present purposes, videotex includes networks accessed by personal computer, such as The Moreover, electronic publishing networks SOURCE, CompuServe, LEXIS, NEXIS, WESTLAW, and AR- PANET, as well as television-adapted networks, such as the can be “nested” within other networks; a serv- experimental Viewtron, Keyfax, T~intex, and “dedicated” network systems such as the Telidon (Canada), Minitel (France), and Prestel (England) systems. “Home banking, for example, allows a customer to handle %lectronic mail is an electronically transmitted message electronic funds transfers from his or her own computer. The from one computer terminal to another, usually via a host com- first system in the United States was Chemical Bank’s “Pronto” puter, which stores and routes the message. Electronic mail is system, which began in September 1983. See: U.S. Congress, commonly available in mostvideotex services. Current electronic Office of Technology Assessment, Efftits of Information Tech. mail services include MCImail, CompuServe, the SOURCE, nology on Finanu”al Sem”ce Systems (Springfield, VA: National GTE Telemail, and Western Union Easy Link. Technical Information Service, September 1984). ‘An electronic bulletin board is a publicly accessible video- ‘See p. 15 of this report for a discussion of the distinction tex computer file, which is typically organized into “conferences’ between electronic publishing and database services. or “topics” that users may create and join, or limit member- IO Ithiel de Sola Poo1, Technolo~”es of Freedom (Cambridge, ship to. See note 5 above. MA: Belknap Press, 1983), p. 280. —

ce for routing messages from one location to ject to centralized control by the electronic pub- another maybe added on top of a network pro- lisher. A paper publishing system which today 13 viding for their storage, which may itself lease looks something like this: mes and switching service from one or more Source(s) + Reporter + Publication + Distribution + networks. Members of these various networks Point of Sale + User can also “talk” to each other between otherwise self-contained networks.11 may one day be joined by an electronic publishing system that looks more like this: Electronic publishing is also interactive; unlike all previous media used by the press, where Source(s) + Interconnection + User he journalistic product was a single package These features of electronic publishing-the whose contents and priorities were controlled decentralization of editorial control, the mul- by editors, many electronic publishing systems tiplicity of versions of news and information permit the style, organization, order, and con- products, the loss of an authoritative text, the tent of the items to be selected by the user, many roles of electronic publishers, the disper- rather than the publisher. The converse is also sion of news and information sources—raise true: writers and publishers may choose their significant questions about liability for false, readers, and differentiate their products across inaccurate, defamatory, indecent, obscene, and classes of readers.12 infringing information. Holding the electronic 14 Interactivity means, moreover, that users publisher liable for injurious information en- can themselves become reporters or publishers. tails conflicts with the “profound national com- This feature is of great significance for the mitment’ made in the First Amendment ‘‘to the principle that debate on public issues should question of “who is press, ” and it also suggests 15 that the sources for news and information may be uninhibited, robust, and wide open . . ." in the future become dispersed, and less sub- ‘ ‘From an OTA staff conversation with Mr. R. Taylor Walsh,

J l+~lectronic Information Services Development, Silver Spring. ] IThe~e ~~metwork lix& are known as ‘‘gateways. % MD. @u-terman and J. Hoskins, “Notable Computer Networks, ” 1 IThe term i ‘injurious information will be used to refer to ;ommum”cations of the ACM, vol. 29, October 1986, pp. 932- defamation, negligence, copyright infringement, fraud, in~asion 171, for a good summary of computer networks and related of privacy, intentional infliction of emotional distress, and other :oncepts. uses of information that cause pecuniary, emotional, reputa- 1 ~The press has already be~n this differentiation prOCeSS tional, or bodily harm. This is in accordance with a recent ap- with, for example, the publication of domestic and international peals court decision that held that “First Amendment protec- /ersions of ~“me and Newsweek, inclusion in newspapers of sup- tions apply to all actions whose gravamen is injurious plements aimed at suburban readers, and the circulation of 7’}’ falsehood. ” Blatt-v \’. New York Times, Inc., 12 Med.1,, Rptr, ;uide to different viewing audiences. Electronic publishing will 1928 (2nd Cir. 1986), cert. denied, No. 86-1803) (19871. mable finer distinctions aJong different classificatory lines, ‘-’New York TirzIes t. Sulli$’an, 376 U.S. 254 (1 964).

EDITORIAL CONTROL AND LIABILITY The editorial decisions of the press go to the whether fair or unfair-constitute the exercise heart of the First Amendment protection of of editorial control and judgment. It has yet its freedom. As the Supreme Court said in Mi- to be demonstrated how governmental regu- ami Herald Publishing Co. v. Tornillo,16 lation of this crucial process can be exercised consistent with the First Amendment guaran- The choice of material to go into a newspaper, tees of a free press as they have evolved to and the decisions made as to the limitations this time. on the size and content of the paper, and treatment of public issues and public officials— A corollary to editorial freedom is editorial responsibility; under common law, the exer- — cise of editorial discretion over whether and 1“418 U.S. 241 (1974). what to publish is the basis for liability for defa- 22 mation and other forms of injurious informa- may use their services.20 They may in fact tion.17 An assumption underlying conven- choose who shall “speak,” but they are less tional theories of press liability is that the often in control over what is said over the net- press—be it newspaper, journal, magazine, ra- work. The choice of who speaks may be made dio, or television–is capable of exercising explicitly-as where a computer network con- editorial discretion over the content of its pub- tracts with news or database services for the lications. Where the press is required by law provision of information to subscribers.21 Or to accept content “as is, ” it is generally ab- it may be made tacitly, and involve nothing solved from responsibility for damages caused more than permitting subscribers to speak, or by false or inaccurate publication.18 Where a allowing the creation of bulletin board con- medium is treated by law as a common car- ferences. rier, and is therefore forbidden from dis- Whether an electronic publishing service criminating against any speaker or otherwise should or can bear liability for information dis- exercising editorial control, it is absolved of tributed on its service is a matter of some con- any legal responsibility for statements made 19 stitutional consequence. If, in lieu of action by over the medium. Congress or the Supreme Court, the Federal This simple symmetry between control and Communications Commission (FCC) were to liability may become hard to apply to electronic decide that electronic publishing is beyond its publishers. On a network, electronic publishers jurisdiction as being neither “ancillary” to may resemble conventional print publishers in- broadcasting, nor common carriers (and thus sofar as they can or do choose which speakers not relieved from publisher liability),22 an argument can be made that it should be treated in the same way as print publishers. The ques- An assumption underlying conven- tion then is whether the electronic publisher can, as a practical matter, exercise the same tional theories of press liability is that 2 control as these other types of publishers, the press is capable of exercising edi- and if requiring it to do so will inhibit the con- torial discretion over the content of tent of or access to the system. In order to its publications.

20 For exmple, Sewices such as ‘1’HE SOURCE, LEX*S NWI, and CompuServe contract with selected newswire serv

- ices to provide information to their subscribers. Some electroni[ 17& with torts gener~y, the basis for liability for ‘ ‘informa publishers, while they do not themselves provide particularnew: tional torts, ” such as defamation, negligent misrepresentation, services, may allow such services to be “ported” from anothel fraud, and even product liability, is that of misfeasance; that online source to their own. National security directives may even is, once one undertakes a discretionary action, such as publica- tually force videotex providers to exclude some potential users tion, one is bound to exercise reasonable care in carrying it out. 21 For exmple, west pubhshing contracts with DOW Jones Although the press is, in certain instances, no longer strictly News Retrieval Service to enable WE STLAW subscribers ac liable for what it publishes, New York Times v. Sullhm, 376 cess to Dow Jones’ financial analyses and news services.— U.S. 254 (1964) and Gertz v. Robert WeJch, Inc., 418 U.S. 323 ZzTeleco~um”cation9 Research and Action Center V. FCC (1974), the basis for liability for defamation, invasion of privacy, 13 Med. L. Rept. 1881 (D.C. Cir. 1986) and negligent misrepresentation remains an intent to publish. 230n networks that provide bulletin bo~d ~d news ad in’ Where the exercise of editorial judgment and control in con- formation services, there may be literally hundreds and thou- veying or publishing information is prohibited by law—as is sands of new messages each day. They may come in faster than the case with common carriers such as telephone and postal any publisher could scan them. Even if such scanning were pos. services—there is no liability for the information. sible, controlling what is posted may be impossible. “To con. lsFor exmple, when a broadcaster is required by law ‘o Pro trol what is posted on a network, it is necessary to control ac- vide “equal time” for the responses of political candidates, it cess to that network. Most existing networks are not strong will not be held liable for defamatory statements made in the on security. The safest policy in using networks is to assume response. Farmers Educational cooperative Union of America that any network can be broken, that any transmission can be V. WDA Y, 360 U.S. 525 (1959). recorded, and that most can be forged. ” J. Quarterman and J. lgNation~ Association of Broadcasters V. FCC, 740 F.2d Hoskins, “Notable Computer Networks, ” Communications oi 1190 (D.C. Cir. 1984). the ACM, vol. 29, October 1986, p. 967. 23 —

avoid civil liability for defamation, negligence, The question whether the privileges extended or infringement of copyright, or criminal lia- the press also apply to “nonmedia” defendants, bility for fraud or theft, the electronic publisher i.e., to individual citizens, and whether a dis- may be forced to censor what is written to its tinction between media and nonmedia defen- host computer. dants can be made at all, is crucial to the extent of First Amendment protections afforded The spectre of “self-censorship’ has caused electronic publishers. The Supreme Court was the Supreme Court, on several occasions, to recently urged to hold nonmedia defendants recognize First Amendment privileges that ef- to a higher standard of conduct than the me- fectively limit the circumstances under which dia, but it avoided the issue entirely, holding newspapers and broadcasters will be held lia- 24 instead that press privilege against defama- ble. Electronic publishing raises the question extended only to “matters of public con- tion whether the limited types of privileges cern” (an incorrect credit report was held not now granted the press will be adequate to avoid 26 to be a matter of public concern). However, self-censorship. Monitoring the accuracy and because conventional print and broadcast me- truthfulness of all of the information posted dia are in great part responsible for defining to a computer conference, for example, may what constitutes a protected “matter of pub- well be beyond the system operator’s abilities. lic concern, ” that question may hinge on the The press, in becoming electronic, may require medium of communication, after all. The role more latitude for error than conventional print 25 of technology in expanding who constitutes publishers. “the press” was recognized by Justice Bren- “Beginning with IVew York Times, 1rIc. v. Sti”van, 376 U.S. nan who wrote in his dissent that: 254 (1964) and cuh-ninatingin Gertz v. Welch,418 U.S. 323 (1974), the Supreme Court fashioned First Amendment-based press “[f]irst amendment difficulties lurk in the privileges with respect to defamatory speech concerning pub- definitional question such an approach would lic officials and public figures. In Rosenblatt v. I%er, 383 U.S. generate. And the distinction would likely be 75 (1966), the Court recognized a similar privilege with respect born an anachronism. (Brennan’s footnote): to speech that is invasive of individuals’ privacy. Owing to transformations in the technologi- ‘sOf course, the sources of information posted on a computer network may also be held liable for injurious information. While cal and economic structure of the communi- straightforward in theory, this may be difficult to implement cations industry, there has been an increasing in practice. Database providers to the network offer thousands convergence of what might be labeled “media” of “pieces” of information, most of which they have no involve- and “nonmedia."27 ment in creating; mediated by artificially intelligent “front ends, ‘‘ information may come to consumers that is a machine- compiled synthesis of many sources, which may themselves not be cited in the allegedly defamatory or negligent report; and the long and complex chains of distribution that are possible in value-added networks may give a totally different and often circuitous meaning to the nature of what counts as a ‘‘fact” ZeDun & Bradstr~t, Inc. V. Greenmoss Builders, Inc., 472 or the “truth’ for purposes of verification of information. See U.S. 749 (1985). In a recent State court decision, Culh”ton v. generally, Ithiel de Sola Pool, Technolo&”es of Freedom, ch. 7 Mize, 403 N.W.2d 853 (1987), the court held that the First (Electronic Publishing) and Anthony Smith, “InformationTech- Amendment privileges recognized in New York Times v. Sul- nology and the Myth of Abundance, ” Daedelus, vol. 111, fall livan apply to nonmedia defendants. 1982, pp. 1-16. 27472 U.S. 749 (1985) at 34.

GLOBAL NETWORKS AND THE INTERNATIONAL PRESS Electronic publishing may blur distinctions of worldwide communications protocols, makes between “domestic” and “foreign” press. In- the concept of a “purely domestic” electronic ternetworking (communicating between two publisher an anachronistic term. Except with or more networks), in combination with global respect to cost, geography is largely irrelevant communications satellites, the emergence of to modern telecommunications systems. “metanetworks,” and the eventual conformance 24

Congress and the courts may eventually Though the right of foreign journalists to have to confront the question of whether First publish in the United States is not yet firmly Amendment rights extend to the foreign press established in constitutional law, the right of publishing in the United States. The question readers, viewers, and listeners to receive news may take many forms—whether standards of and information has been established in other liability will apply to foreign-originated trans- contexts, and may be indirectly available to missions, whether domestic regulations apply, nonresident electronic publishers. Even this whether import/export laws conflict with First rationale, however, is tenuous; relying as it Amendment rights, or whether control, even does on cases involving very specific circum- if permitted under the First Amendment, is stances and court dicta.32 The fact that sec- possible.” tion 310 of the Communications Act requires that broadcast licenses be held only by citizens The First Amendment status of the foreign of the United States, and that a recent Su- press, and foreign speakers in general, is one preme Court decision upholding registration of the least understood areas of First Amend- and labeling requirements for Canadian made ment jurisprudence. A case involving depor- 33 films, suggests that the foreign press may tation of aliens for allegedly subversive speech have lesser First Amendment rights, if they under the McCarran-Walter Immigration Act have any. of 1952 is currently pending before a Federal court.29 In the only case on point, Times Newspapers Ltd. (Of Great Britain) v. McDon- 30 Short of monitoring all international nell Douglas Corp., the court held that The data traffic, there may be no way to Sunday Times of London, an English news- stem news and information from for- paper with offices and a small circulation in the United States, was protected from having eign sources, even if it were desirable. its publication suppressed. The court said that freedom of the press is not limited to what is published in America for American readers, The political and legal impetus for recogniz- but was also “designed to protect the rights ing First Amendment rights for the interna- of readers and distributors of publications no tional press may not necessarily come from ju- less than those of writers or printers.”31 dicial interpretations of the Constitution; it may come from international authority, and from the interconnectedness of nations that 34 Congress and the courts may eventu- has accompanied technological change. For ally have to confront the question of whether First Amendment rights ex-

tend to the foreign press publishing l~In R~ Ljon Broadcast@ C’O. V. FCC, 395 U.S. 367 (1969). in the United States. in the context of the constitutionality of the fairness doctrine, the Court said that “the people as a whole retain their interest in free speech . . . [it] is the right of viewers or listeners, not the right of the broadcasters, which is paramount. ” See also Vir@”nia State Board of Pharmacy v. Vir@”nia Citizens Con- ‘8These issues belong to a class of emerging issues concern- sumers Council, 425 U.S. 748 (1976), Board of Educ. v. Pico, ing “transborder data flows”; the movement of data between 457 U.S. 853 (1982); which are commonly cited as constitutional countries. For an overview of the legal and policy issues sur- foundations for a right of the public to receive information un- rounding transborder data flows, see: A. Branscomb, “Global der the First Amendment. See, e.g., Mark Nadel, “A Unified Governance of Global Networks: A Survey of Transborder Data Theory of the First Amendment: Divorcing the Medium From Flow in Transition, ” 36 Vanderbilt Law Review 985 (1983). the Message, ” 11 Fordham Urban )aw Journai 163, 187 (1982), ‘g’’ Federal Judge Declines To Rule on Free Speech Rights of and Emerson, “Legal Foundations of the Right to Know, ” 1976 Foreigners, ” New York Times, May 21, 1987. The case of Mar- Washington Um”versity Law Quarterly 1, garet Randal is still pending as of December 1987. :n~=se “. Clmkson, 55 U. S.L.W. 4586 (Apr. 28! 1987)” ~0387 F. Suppo 189 (C. D. Cd. 1974). 34This topic will be developed in a forthcoming OTA special ~1 Id Quoting Qu~tjty of Copies of Books v. State of bicentennial report on Science, Technology, and Constitutional Kansas, 378 U.S. 205 (1964). Governance. .

25 example, Article 19 of the Universal Declara- the network.) As Leonard Marks, a former di- tion of Human Rights proclaims that: rector of the U.S. Information Agency, couched the issue, Everyone has the right to freedom of opinion and expression; this right includes freedom Global electronic networks . . . will be effec- to hold opinions without interference and to tively beyond the reach of the traditional seek, receive and impart information through forms of censorship and control. The only way any medium regardless of frontiers.35 to “censor” an electronic network moving. . . [at] 648 million bits per second is literally. to This does not have legal effect, but it is pull the plug.37 widely recognized as the basis for international law. Whether the foreign press may avail it- In other words, the only effective means of con- self of First Amendment rights or not may be- trolling foreign speakers speaking in the come an academic exercise. Short of monitor- United States may be to disable the commu- ing all international data traffic, there may be nications abilities of American citizens,38 an no way to stem news and information from for- action that would directly raise First Amend- eign sources, even if it were desirable. A hoax ment rights. on April Fools Day in 1984 on a computer network named USE NET may be a harbinger of ‘“L. hf-arks, International Conflict and the Free Flour of in- things to come—"kremvax! kgbvax!chernenko” formation in Control of the Direct Broadcast Satellite, I’alues 36 in Conflict (Palo Alto, CA: Aspen Institute Program on Com- joined the network. (The notation is the munications and Society, 1974); as quoted in J. Pelt, on, ‘“I’he name, address, and routing message used on Technological Environment, ” Toward a I.aw of Global Cmn- munications Networks, Science and Technology Section of the American Bar Association, A. Branscomb (cd. ) (New York, N1’: ‘r’(Jniversal Declaration of Iluman Rights, G.A. Res. 217 Longman, 1986), pp. 37, 44-45. (11 I) U.N. Dec. A81O, at 71, art, 19 (1948) {emphasis added); .~~Though not in the context of foreign speech, this stip is not as quoted in John Eger, “The Global Phenomenon of Teleinfor- unprecedented; computers and communications were confiscated ma tics: ,4n Introduction, 14 Cornell international I,aw Jour- from a s}”stem operator in 1984 whose communications facil- nal 203, 20’7 ( 19!31 ), ities were used for telephone credit card fraud. “Free Speech ‘(’” N-otahle Computer Networks, ” (’ornmunications of the Issues Surround Computer Bulletin Board Use, ” A’ew I’ork .4(’.11, \ol. 29, october 1986, p. 967. 7’ime.s, Nov. 12, 1984, p. Al, CO1. 1. Chapter 3 New Technologies for Publishing and Disseminating News and Information

THE CONVERGENCE OF THE MEDIA Of all the First Amendment issues concerning the press and new technology, none is more The seemingly absolute prohibition on contentious than Federal regulation of the means of publishing and disseminating news abridgments of press freedom enun- through electronic media. The seemingly abso- ciated in the First Amendment has lute prohibition on abridgments of press free- nevertheless been found compatible dom enunciated in the First Amendment with a three-tiered system of commu- (“Congress shall make no law. . .“) has never- nications freedom: print media, broad- theless been found compatible with a three- cast media, and common carrier. tiered system of communications freedom: print media, broadcast media, and common carrier. This separation was at first a product Print media4 occupy the first tier of con- of market economics and agreements between stitutionally protected communication, and are key players in the communications industry,1 subject only to laws concerning injurious rather than being mandated by technology.2 speech (like defamation and negligence), con- It was nevertheless embodied in the regula- stitutionally unprotected speech (obscenity tions of the Federal Communications Commis- and “fighting words”), and those laws regu- sion (FCC). 3 Today, technology is ushering in lating the press as a business, without regard a convergence of forms of press publishing that to the press’ communicative functions (e.g., were once partitioned by technology: print pub- corporate, labor, and antitrust laws).5 lishing, mail, broadcasting, and telephone. The broadcasting media occupy the second tier. Under the Communications Act of 1934, the FCC has the task of:

)” In ~926 AT&T abandoned broadcasting in return for . . . regulating interstate and foreign commerce RCA’S commitment to use AT&T’s lines rather than Western in communication by wire and radio so as to Union’s for networking. The agreement provided for exchange make available, so far as possible, to all the of patent licenses, and WEAF was sold to RCA, where it be- people of the United States a rapid, efficient, came WNBC, the flagship station for the National Broadcast- nationwide and worldwide wire and radio coming Co. ” Ithiel de Sola Pool, Technologies of Freedom (Cam- bridge, MA: Belknap Press, 1983), p. 35. See also Lynn Becker, munication service with adequate facilities at ‘‘Electronic Publishing: First Amendment Issues in the Twenty- reasonable charges for the purpose of the na- First Century, ” 1,3 Fordham Urban Law Journal 801, 818-820 tional defense, for the purpose of promoting ( 1984). safety of life and property through the use of “’Other approaches were proposed and tried. For a quarter wire and radio communication.6 of a century from 1893 phones carried music and news bulle- tins to homes in Budapest, Hungary. Thomas Edison thought “’Print media” refers not only to all paper and ink publica- that the main use for the phonograph he had invented would tions; it is coextensive with any “work fixed in a tangible medium be for mailing records as letters. “. . . The fact that different of expression, ” as defined under the Copyright ,Act, 17 (J, S. C’. technologies were consecrated to different uses protected me- \ 102(a), and thus includes motion pictures, paintings, sculpture, dia enterprises from competition from firms using other tech- photographs, computer-processible information (including pro- nologies. ” Pool, Teclmo]o&”es of Freedom, op. cit., p. 27. grams and databases), and sound recordings. ‘Congress provided for three distinct regulatory schemes: for ~~fiam; }{er~d Co. k,. Tornilo, 418 U.S. 241 11974), has bro~dcasters (47 U.S.C. j$ 301-332), common carriers (47 U.S.C. reaffirmed the premier position occupied b~ the print media, ! j201-224), and for non-broadcast users (47 U.S.C. I! 351-362). so far as freedom from goy’ernment interference in editorial con- Print media, which do not travel by ‘I wire or radio, are be>rond trol is concerned. the jurisdiction of the FCC to regulate, 47 U.S.C. !151. “47 IJ.S.C, 1151.

8(1-(,/1’, - }j’{ - -j 28

In carrying out this responsibility, the FCC (H.R. 1934) was recently vetoed by the must conform its actions to those ‘consistent President, and the FCC voted to repeal its with the public interest, convenience [and] ne- fairness regulations on the ground that cessity.”7 Under this authority, and based on they offended the First Amendment.* a rationale that “the electromagnetic spectrum ● The Equal Time Doctrine: “If any licen- is simply not large enough to accommodate see shall permit any person who is a le- everybody, the FCC licenses broadcasters gally qualified candidate for any public of- and conditions the grant and renewal of fice to use a broadcasting station, he shall licenses on compliance with a variety of con- afford equal opportunities to all other such tent and structural regulations. These regula- candidates for that office in the use of such tions include: broadcasting station.”13 ● ● Reasonable Access: Broadcasters must al- Cross-Ownership Restrictions: In the in- low “reasonable access . . . for the use of terest of promoting diversity, the FCC im- a broadcasting station by a legally qual- poses three general types of restrictions ified candidate for Federal elective office on multiple ownership of broadcasting fa- 14 on behalf of his candidacy." cilities: those limiting ownership in a sin- ● 9 Indecency: Under criminal law and FCC gle community, those limiting owner- regulations, broadcasters are held to a ship of broadcast facilities by single 10 higher standard than other publishers, in- entities nationwide, and those forbid- sofar as they are prohibited from broad- ding newspapers from owning television casting not only “obscene, but also “inde- stations in the same community in which 15 11 cent” programming. The rationale for they publish. this higher standard is that the broadcast ● The Fairness Doctrine: Under FCC deci- audience, and particularly children, are sions construing a 1959 amendment to the 16 ‘‘captive." Communications Act of 1934, broadcasters were obligated to “operate in the FCC regulations, and indeed, much of the public interest and to afford reasonable rationale under which the FCC regulates, have opportunist y for the discussion of conflict- come under attack in recent years, largely as ing views on issues of public impor- a result of technological challenges to the no- tance.” 12 The FCC added a further gloss tion of scarcity .17 to this statutory language in the Report *5G u.siL.w. 2112 (Aug. 25, 1987). Concerning General Fairness Doctrine ’347 U.s.c. 5315. Obligations of Broadcast Licencees, 102 “47 U.S.C. j312(a)(7). r F.C.C. 2d 143, 146 (1985): licensees must IsBroadcast of ‘‘obscene, indecent, Or prOf~e’ l~~age ‘ images is a criminal violation, 18 U.S.C. $1464, for which the “provide coverage of vitally important FCC may revoke a license, 47 U.S.C, 312(a)(6). This power was controversial issues of interest in the com- upheld in FCC v. Pacifica Foundation, 438 U.S. 726 (1978). The munity served by licensees . . . [and] pro- FCC recently recommended expanding the operative definition of “indecency” beyond seven particular words to include the vide a reasonable opportunity for the pres- generic definition of broadcast indecency advanced in l%cifica, entation of contrasting viewpoints on such which is: “language or material that depicts or describes, in terms issues. ” The codification of the Fairness patently offensive as measured by contemporary community standards for the broadcast medium, sexual or excretory activ- Doctrine in the Communications Act ities or organs. ” The Commission added that “in certain circumstances, innuendo or double entendre may give rise to ac- 7 47 U.s.c. j307. tionable indecency. ” 52 FederaJ Reg”ster 16386-01 (Tuesday, ‘~~~io~~ ~ro~~c~s~~~g CO. v. Vn.i”ted States, 319 U.S. 190, May 5, 1987). 16* e.g., Cohen v. C~”fo~”a, 403 U.S. 15 (1971) 213 (1943). The scarcity rationale was reaffirmed as the basis - for broad FCC authority to regulate in Red Lion Broadcasting ITFor’exmple, ~ U.S. Congress, 98th Cong., 1st Sess.! sen V. FCC, 395 U.S. 367 (1969). ate Committee on Commerce, Science, and Transportation, Print 9 CFR jj 47 73.35 (AM), 73.240 (FM), ~d 73.636 ‘W)+ and Ekctrom”c M&”a: The Case for First Amendment Parity, ‘“Ibid. printed at the direction of Hon. Bob Packwood, Chairman, for “see Second Report and Order, Docket No. 18110,50 F.C.C. the use of the Committee on Commerce, Science, and Transpor- 2d 1046 (1975). tation, Senate (Washington, DC: U.S. Government Printing Of- ’247 U.S.C. $315(a). fice, 1983). 29

Whether broadcasting is a scarce resource ing and compression techniques may further depends to a great degree on how “scarcity” overcome physical limitations of the elec- is defined. Absolute numerical comparisons of tromagnetic spectrum, the limitations on time te number of media outlets maybe mislead- and space inherent to the broadcasting ing measures of scarcity. If scarcity is meas- medium. ured by the number of organizations or individuals wishing to broadcast as compared with the number of available frequencies, then scar- Whether broadcasting is a scarce re- city is still the rule for broadcasting. There are, source depends to a great degree on for example, no open broadcast television chan- how “scarcity” is defined. nels in the top 50 markets in the United States. If, as one First Amendment scholar suggests, scarcity occurs in situations where ‘one utter- In this context, the FCC has now rejected the scarcity rationale-at least as the basis for ance will necessarily displace another, ” then 20 scarcity takes on yet another meaning: the fairness doctrine. Many in Congress are seeking to legislate the fairness doctrine out . [the] opportunities for speech tend to be of concern over the scarcity and high cost of limited, either by the time or space available broadcast outlets. The FCC has in other infor communicating or by our capacity to di- stances suggested an alternative rationale hav- gest or process information. . . . The decision ing to do with the broadcasting industry’s to fill a prime hour of television with Love Boat necessarily entails a decision not to broadcast “power to communicate ideas through sound a critique of Reagan’s foreign policy . . . dur- and visual images in a manner that is signifi- ing the same hour.18 cantly different from traditional avenues of communication because of the immediacy of 21 The development of cable television (which the medium.” typically carries anywhere from 34 to 120 sta- Yet even these constraints on the broadcast- tions), the direct broadcast satellite, other ing media are being overcome by technologies, microwave communications systems, low the videocassette recorder, for instance, that power television, and other new technologies permit an audience to select among and store cast doubt on scarcity as the premise for gov- programs of its choosing. The element of viewer ernment regulation. The broadcast medium far selection and timing was a principal reason for exceeds the print medium in sheer number of the Supreme Court’s finding that “time-shift- outlets. 19 Broadcasting has become more ing’ television programs constituted ‘fair use’ ubiquitous and far more diverse than news- and is not a violation of copyright law.22 papers in many metropolitan areas. Multiplex- Common carriers, finally, are subject to yet different treatment under Federal law and the 180wen Fiss, “Free Speech and Social Structure, ” 71 Iowa >aw Review 1405, 1412 (July 1986). First Amendment. Under Title II of the Com- lgAt the end of 1985, there were 9,871 radio and 1,220 tele- munications Act, communications common vision stations operating in the United States, compared to 1,676 carriers are subject to franchise, rate, service, hily and 6,600 weekly newspapers, and more than 10,000maga- ~ines. Broadcasting Yearbook 1986, and E&”tor and Fubh”sher and reporting requirements, and “must hold Yearbook 1986, as cited in Christopher Bums, Freedom of the themselves out to all comers” on a nondiscrim- Press in the Information Age, OTA contract report, Apr. 21, inatory basis.23 Thus common carriers are not 1987, p. 20. In 1986, A.C. Nielsen counted 86 million television Zom Ort conce~ng Genel-~ Fairness Doctrine Obli@iom lomes and a total of 157 million television sets in use, compated P ;O 63 million daily newspaper subscribers. Television Bureau of Broadcast Licensees, 102 F.C.C. 2d 143 (1985); see also ]f Advertising, “Trends in Television, 1986, ” as cited in Bums, Meredith Corp. v. FCC, 55 U. S.L.W. 2391 (1987). op. cit. In 1985, the average home had the television on 7 hours zl~port ~d Order, 53 Rad. Reg. 2d (P&F) 1309 at 132A md 10 minutes per day, compared to an average of 34 minutes (1983), as quoted in Tekcornmum”cations Research and Action per day spent reading the newspaper (of course, the time spent Center v. FCC, 13 Med. L. Rptr. 1881, 1883 (D.C. Cir. 1986). watching television was probably for entertainment, rather than 22Sony v. Um”versaJ Stuch”os, Inc. 464 U.S. 417 (1984). Elec- news; the comparison may thus be misleading). NationalReader- tronic publishing, also illustrates the way in which technology ~hip Study conducted in 1971 by Audits & Surveys, Inc., for is overcoming time and space scarcity. the Newspaper Advertiser Bureau, as cited in Burns, op. cit. 2347 USC. 201-224 (1982). 30 generally liable for the content of the messages no interaction between common carrier status they transmit. and First Amendment concerns.28 However, Drawing on the precedent of the postal sys- since the 1982 consent decree between AT&T tem, the telegraph, and the railroads, the FCC and the Justice Department, AT&T has been defined communications common carriers as denied the right to disseminate its own messages over its lines until 1989 because of the ‘‘any person engaged as a common carrier for 29 hire. ’24 This circularity causes a number of potential for anti-competitive behavior. The conceptual problems, particularly in questions of whether and how to regulate new media. The The First Amendment issues concern- Supreme Court has defined common carrier in a less circular fashion, as “one who makes a ing the relationship between media public offering to provide [communications fa- ownership and control over the con- cilities] whereby all members of the public who tent that flows through that medium choose to employ such facilities may commu- have been joined. nicate or transmit intelligence of their own design and choosing . . . "25 but this does not solve the problem of when common carrier First Amendment issues concerning the rela- status may be mandated and imposed by the tionship between media ownership and control government. 26 The decision that a communi- over the content that flows through that cation system is or is not a “common carrier’ medium have been joined. Among these issues is a political rather than a technical decision. are whether common carriers can also be pub- Legislatures tend to decide to regulate a sys- lishers, who–as publishers-have the right to tem as a common carrier if it appears to have exclude other publishers from their fora; at least some of the characteristics of a natu- whether a monopoly common carrier can also ral monopoly .27 Until recently, there has been do some publishing; and whether the government can compel some access over monopoly 2447 USC. 153(h). controlled facilities. The controversies over ca- “’FCC v. Midwest Video Corp., 440 U.S. 689, 701 (1979). ‘hBroadcasters may not be treated as common carriers. 47 ble television and information services that U.S.C. $153(h). Notwithstanding this, Direct Broadcast Satel- might be available over telephone wires are il- lite owners are permitted to choose between broadcast and com- lustrative of this issue, which promises to be mon carrier type regulation. National Association of Broad- casters v, FCC 740 F.2d 1190 (D.C. Cir. 1984). the focal point of much First Amendment liti- ‘7’’When ‘natural’ monopolies are recognized and entry is gation for years to come. prohibited to all but the designated monopolist, the monopolist is normally required to provide universal service as a common carrier, ” Mark Nadel, “A Unified Theory of the First Amendment: Divorcing the Medium From the Message, ” 11 ‘“In the case of the telegraph, for example, the reason for the Forchm Urban Law Journal 163, 193, n. 109 (1982). “The rate “dim perception [of First Amendment concerns] was that the and service requirements imposed on these carriers under the early telegraph carried so few words at such a high cost that Communications Act reflect the view that their natural mon- people thought of it not as a medium of expression but rather opoly status justifies government control of their business activ- as a business machine. The computer suffered the same mis- ities. ” Lynn Becker, “Electronic Publishing, ” op. cit., p. 855. perception a century later. Pool, Technolo~”es of Freedom, op. See also flni’ted States v. RCA, 358 U.S. 334 (1959). Cable tele- cit., p. 91. vision was initially proposed as a candidate for common carrier ‘gConsent decree in U.S. v. Western Electric, 552 F. Supp. regulation by the FCC, but the plan was abandoned. 131, 180-86 (D.C. Cir. 1982).

CABLE TELEVISION When cable television entered the scene in tion area. Since that time, however, cable the 1940s it was called Community Antenna television has multiplied channel capacity Television (CATV), carried only existing broad- many fold. Systems that run two coaxial ca- cast channels, and was intended merely to pro- bles into the home can now provide up to 120 vide better signals to homes in a bad recep- different channels. Initially, the FCC declined 31 — jurisdiction over cable TV, but throughout the [cable] operator,"35 and it permits local fran- 1960s and 1970s, the Commission imposed a chising authorities to reserve public, educa- variety of access, content, and distant signal tional, and government channels.36 But cable importation requirements. The Commission’s operators take the position that because they authority to do so was based on the rationale do not suffer the physical limitations inherent that its regulations were “reasonably ancillary in broadcasting, they are in the position of to . . . the Commission’s various responsibili- other publishers, and ought to have absolute ties for the regulation of television broadcast- editorial discretion. It seems likely, therefore, ing, "30 and broadcast television was “placed that even the limited content regulations set in jeopardy by the unregulated growth of forth in the Cable Act will be challenged on CATV."31 First Amendment grounds. Then, in the late 1970s and early 1980s, a But, as Ithiel de Sola Pool pointed out, “[the series of Federal appeals court rulings struck problem of access] may become the Achilles down a variety of programming content regu- heel of what could otherwise be a medium of lations, based on either the First Amendment communication every bit as free as print."37 or statutory grounds.32 Finding first that the Though many have argued that cable tele- FCC had failed to show that cable systems are vision is not a “natural monopoly, one cable ‘‘public forums’ (i.e., common carriers), the lan- franchise per municipality is nevertheless the guage of one appeals court decision went on rule and not the exception. ” This suggests to to frame the issue thus: some that cable operators ought to be treated in the same way as any other essential facility The First Amendment rights of cable operators rise from the Constitution; the public’s with substantial power to exclude others. Sev- ‘‘right’ to ‘‘get on television’ stems from the eral commissions have come to this conclusion, Commission’s desire to create that “right.”33 but have still accepted the argument that treat- ing cable as a common carrier would not pro- In the Cable Communications Policy Act of 34 vide adequate economic incentives for opera- 1984; Congress created such a right, albeit tors to build cable systems.39 Even if cable in limited fashion. It requires cable operators systems are not treated as total common car- to provide “leased access channels” for com- riers, the question remains whether the gov- mercial use “by persons unaffiliated with the ernment, to promote diversity, can require access to a certain portion of the available “’U.S. f. Southwestern Cab/e Co., 392 U.S. 157 ( 1968) ‘1 Ibid. channels. ‘iFor example, in Home Box Office, Inc. J“. FCC, 567 F.2d 9 (DC. Cir, 1977), overturned, on First Amendment grounds, The delicate equilibrium that exists today rules restricting movies that could be shown over subscription between cable operators and television pub- television. [n Midwest Video Corp. ~’. FCC, 440 U.S. 689 (1979) lishers will likely be disturbed as technology (Midwest Video II), the Supreme Court struck down Federal (though not necessarily state) cable access requirements as be- 1“47 U.S.C. j612 yond FCC jurisdiction; if access rules could not be imposed on ’647 U.S.C. $611 broadcasters, they could not be imposed on cable operators on jTIthiel de Sola Pool, Technologies of Freedom (Cambridge, an ‘‘ancillary to broadcasting’ rationale. In 1985, the Court of MA: Belknap Press, 1983), p, 166, Appeals for the D.C. Circuit decided that “must carry” rules, ‘hCable is often called a “natural monopol~’” because, as a which required cable systems to carry local broadcast signals, practical matter, only one operator ma~~ use rights of way over were unconstitutional under the First Amendment. Quinc.y Ca- poles or through conduits to connect with subscribers’ houses. ble TV v. FCC, 768 F.2d 1434 (D.C. Cir. 1985), cert denied, 106 Where these physical limitations are not present, or where there S. Ct. 2889 (1986). The court suggested, moreover, that even if is sufficient excess physical capacit-y, a cable operator ma~’ raise the “must carry” rules had furthered a significant governmental First Amendment objections. Los Angeies t’. Preferred Com- interest, they might nevertheless be contrary to the principle munications, Inc., 106 S. Ct. 2034, 54 U. S.L. W. 4542 ( 1986). announced in Miami Herald Co. v. Torrdlo, &“cta of some por- ~gTwo ealy stu~es of the cable industry, the Sloan Founda- tent for categorization of cable for First Amendment purposes. tion (On the Cable (New York: McGraw-Hill, 197 1) and the ‘‘The language is from the appeals court ruling in Midwest Whitehead Report (Cabinet Committee on Cable Communica- Video II, Midwest Video Corp. IF. FCC, 571 F.2d 1025, 1054 tions, Cable–Report to the President (Washington, DC: U.S. (D.C. Cir. 1978). Government Printing Office, 1974). The Whitehead report “47 U.S.C. ~ \ 601-639, Public 1,. No. 98-549, 98th Cong. 2d reached the conclusion that when cable reached 50 percent sess., 98 Stat. 2779-2806, 1984. penetration, it should change to carrier status. 32 —

again brings new interests into play. Although cable is today primarily an entertainment The delicate equilibrium that exists medium, it may not necessarily remain so. Be- today between cable operators and cause it is a “broadband” medium (meaning television publishers will likely be dis- that it has capacity for handling high volumes turbed as technology again brings new of all sorts of electronic traffic), cable is a suit- able carrier for computer data, electronic mail, interests into play. videotex, databases, security monitoring, home banking and shopping, teleconferencing, and other interactive services. If cable systems are publishers under the First Amendment, phone connection to the home may make the and allowed to choose the content of what goes telephone company a broadband highway for through their lines, they may well discriminate all information (with cable operators perhaps against content that is competitive to their becoming customers of the telephone com- own, or which do not yield as large a profit as pany), but for now, the tension between cable entertainment products, such as Home Box Of- operators as First Amendment speakers and fice (HBO) and other movie channels. In the as forums for other would-be speakers will future (perhaps the mid-1990 s), fiber optic tele- heighten.

INFORMATION SERVICES DELIVERED OVER TELEPHONE LINES

Technology has also blurred distinctions between computers and communications, between those who create a message (or data) and The issue here is whether telephone those who transmit it. This confounding of companies will be permitted to pro- roles has raised First Amendment issues sim- vide information services, using their ilar in kind to those raised by cable television; own facilities, in competition with in- that is, how to reconcile the First Amendment dependent providers of the same or interests of communications companies as similar services. speakers with the First Amendment interests of those who seek access to these companies’ communications facilities. While the tension Association that telephone companies would in the cable industry concerns whether cable provide electronic yellow pages and draw away operators will be required to grant a limited a substantial chunk of newspapers’ classified form of access to other would-be program pro- ads. Given the evidence of past discriminatory viders, the issue here is whether telephone com- actions by AT&T against competitors, the panies will be permitted to provide informa- modified final judgment (MFJ) settling the tion services, using their own facilities, in AT&T lawsuit barred the telephone company competition with independent providers of the from entering the electronic publishing same or similar services. business.

Companies offering stock quotations, sports It is unclear whether the First Amendment scores, airline schedules, and news retrieval permits such an absolute prohibition. Cable services, among others, are concerned that the operators have been charged with discriminat- telephone companies could offer these same ing by favoring their own affiliated pay net- services themselves. Even more significant is works over those of their competitors, yet they the fear of the American Newspaper Publishers are not barred from carrying any of their own 33 services. One could argue that cable television that of a hierarchical pyramid to that of a geo- systems are different from telephone compa- desic ring, so that the threat of dominance by nies because the former are not natural mo- one or a few industries is no longer possible. nopolies, but in The Geodesic Network,40 a con- sulting report prepared for the Department of Meanwhile, the FCC appears to believe that Justice, Peter Huber argued that local tele- information services will not become widely phone companies are not natural monopolies available until the telephone companies offer either. He concluded that technology has information services. The agency, in its Com- changed the nature of communications from puter III decision, argues that requiring all competitors to grant comparable efficient in- “’[J. S. Department of tJustice, Antitrust Division, The Geo- desic ,N”etwork 1987 Report on Competition in the Telephone terconnection would be sufficient to ensure Industr~, ,January 1987. nondiscriminatory treatment. Part II Scientific Communications and the First Amendment

Reconciling the maintenance of constitutional liberties with the requirements of national security poses an arduous challenge to democracy. Granted that a balance must be struck, where should the line be drawn? That is the puzzle for all who would presume to lead a free people. It implicates perhaps our most cherished contribution to social intercourse: Separation of Powers . . . . It is the undisputed responsibility of Congress and the Courts to maintain and regulate the right balance between measures necessary for the invulnerability of national security and the preservation of free expression. –Martin L.C. Feldman, U.S. District Judge for the Eastern Division of Louisiana, Jan. 14, 1987 —

Chapter 4 National Security and Scientific Communications

SCIENCE, FREE SPEECH, AND NATIONAL SECURITY

Science and technology were recog- Scientific activity or communications nized by the Founding Fathers as in- have not in practice enjoyed the spe- dispensable to the “common Defense cial status given political comment. and general Welfare.”

They understood well the importance of sci- The U.S. Constitution empowers Congress entific freedom, as is shown by their writings. to regulate commerce, to fix the standard of But even with the addition of the first Ten Amendments to the Constitution, in 1791, sci- weights and measures, to establish post offices and post roads, and to secure for authors and entific freedom was not singled out for special inventors the exclusive right to their respec- protection; it was presumed to be included with tive writings and discoveries in order to “pro- other areas-politics, religion, philosophy, eco- mote the Progress of Science and useful Arts.”1 nomics--in the broad protections given to speech, publication, assembly or association, There is nothing else in the Constitution exercise of religion, petition and protest, all directly related to science and technology. It included within the First Amendment.2 was assumed that the States would have primary responsibility for the useful arts such as None of these First Amendment protections agriculture, manufacturing, construction, and is absolute under prevailing constitutional doc- medicine. But the few provisions in Article I trine. There are times and conditions when the are significant, because they clearly indicate interests of the Nation as a whole override the that science and technology were recognized right of the individual to say and do as he or by the Founding Fathers as indispensable to she wishes. Both Congress and the Supreme Court have treated political speech as that the “common Defense and general Welfare” and an appropriate subject of attention and speech most strongly protected by the First Amendment. Commercial speech-that is, ad- support for Congress. vertising—is least protected; and may be reg- The men who wrote the Constitution were ulated as to time, place, and manner.3 well educated in the science of their day and Scientific activity or communications have enthusiastic about advances in technology. not in practice enjoyed the special status given James Madison avidly studied and wrote about natural history and agriculture science. Ben- political comment, although some constitu- jamin Franklin was one of the world’s leading 2Steven Goldberg, “The Constitutional Status of American scientists. Hamilton studied medicine, chemis- Science, ” University of Illinois Law Forum, vol. 1979, No. 1, 1979. pp. l-6ff. try, and mathematics. ~Before Vir~”m”a State Board of Pharmacy v. Vir@”m”a Citi- zens Consumer Council, Inc., in 1976, the Supreme Court did ‘Article 1, Sec. 8 also provides authority for Congress to not treat commercial speech as protected speech; it was fuily ‘‘make all I.aws which shall be necessary and proper for carry- subject to State police power. The extent of, or limits on, pro- ing into Execution’ these and other powers derived from the tection for commercial speech are somewhat uncertain at this Constitution. time.

37 38

Is there a potential constitutional con- One issue addressed in this report is flict between the rights of free speech the extent to which national security and press guaranteed by the First and foreign trade interests have con- Amendment, and government restric- verged and perhaps in some respects tions on the communication of scien- may even have been confused. tific information in the interest of national security? widely disseminated, and thoroughly argued, and the results replicated. In part this is in or- tional scholars argue that this was the clear der to share knowledge with other scientists intent of the First Amendment separation of for the ultimate benefit of people in general. church and state.4 It is not, indeed, clearly More immediately, it provides a test and means established that there is “a right to do re- of validation. search, ” nor have the limits of governmental Science gets at the truth by a continuous authority to restrict speech and press (publica- process of self-examination which remedies tion) in the area of science and technology been omissions and corrects errors. The process re- clearly defined, by either Congress or the quires free disclosure of results, general dis- Courts. semination of findings, interpretations, conclusions, and widespread verification and Representative George Brown, in 1982, warned criticism of results and conclusions.6 Congress: Recent administration actions . . . sharpen The First Amendment the conflict between constitutional protections and the requirements of national security. . . , The Supreme Court has recognized in numer- The issues at stake stem from the conflicting ous cases that Congress (or State legislatures) demands of the most fundamental matters in may make laws that limit freedom of speech national policy: the security of the Nation and or press. Government must, however, sustain its economic well-being, versus the rights of a substantial burden of proof to justify an in- citizens to privacy, assembly, free speech, terference with speech or press. As a general travel, and freedom from unwarranted Gov- rule, expression may not be restricted because 5 ernment interventions. of its content, although some categories of ex- Is there a potential constitutional conflict pression are given less protection than others.’ between the rights of free speech and press Government restrictions on free speech may guaranteed by the First Amendment, and gov- however be valid under a “balancing of inter- ernment restrictions on the communication of ests test when those restrictions are inciden- scientific information in the interest of national tal to legitimate government purposes not 8 security? Is the balance between these two directly related to speech or press. interests—both of critical importance to Amer- The protection of freedom of speech and free- ican constitutional government—being main- dom of press is therefore not absolute. Many tained, or is it endangered? GHuold c. Relyea (~.), “shrouding the Endless FrOntier— A fundamental tenet of scientific methodol- Scientific Communication and National Security: The Search for Balance, Str&”ng a Balance: National security and Scien- ogy is that basic scientific research results or tific Freedom (Washington, DC: American Association for the new scientific theories should be published, Advancement of Science, 1985), p. 76. TPoh.m ~eptiment o~c~”c%o V. hfos(z?~y, 408 U.S. 92 (1972) ‘Goldberg, op. cit., says that the Founding Fathers, as “men “Fighting Words” and obscenity are generally not protected. of the Enlightenment, ” saw established churches as having been ~Adderly v. Florida, 385 U.S. 39 (1966); Brown v. Louisiana, the enemy of free scientific investigation. 383 U.S. 131 (1966); Kovacks v. Cooper, 336 U.S. 77 (1949); ‘Congressional Record, vol. 128, No. 16, Feb. 25, 1982, p. Breard v. Alexandria, 341 U.S. 622 (1951); O’Brien v. United H511. States, 391367 (1968). —.——— —

39 restrictions are placed on scientific communication in the interests of national security. Na- The real question is whether taken to- tional security may be defined as the military, defense, and foreign relations objectives of this gether, the effects of these restrictions Nation. This definition has implicitly been place a limitation on scientific free- broadened over time to include protection of dom so as to contravene the intent of economic and trade objectives. One issue ad- the First Amendment. dressed in this report is the extent to which national security and foreign trade interests have converged and perhaps in some respects What Activities Are Restricted may even have been confused. and How? Where government action involves a “prior restraint, that is, a prohibition prior to rather Scientific information includes not only pub- than a punishment after the communication, lishing in professional journals or the public the constitutional test is much more severe. media, but also: There is a “heavy presumption’ against the presentation of papers or giving of lectures constitutional validity of any prior restraint.9 in professional meetings or other fora, The seminal Supreme Court decision is Near distribution of unpublished papers or v. Minnesota where it was held that with cer- reports, tain limited exceptions prior restraints were participation in workshops or working constitutionally impermissible.’” sessions, discussions among colleagues, and Threats to national security can make it nec- classroom or laboratory instruction. essary to limit free speech. The court ruled in Schenck v. United States, ” that impingement These activities have been restricted in the in- on freedom of speech in certain circumstances terest of national security: ‘‘appears to be a reasonable exercise of sover- Ž under Executive Orders; eign power . . . in the interest of the common Ž by or under legislation, including: defense and security. ” –the Atomic Energy Act, In recent years not only has scientific ex- –the Invention Secrecy Act, and change or publication been limited in the in- –Export Administration and Arms Ex- terest of national security, but it arguably has port Control Acts; also been limited in the interest of national Ž by contractual agreement between re- trade balances. Thus the question: As we en- searchers and government agencies; or ter the third century of constitutional govern- Ž by self-restraint agreements of research- ment, have these restrictions taken together ers and professional societies. burdened the exercise of free speech and press Each of these means is to be considered in to a degree that may violate the First Amend- the context of judicial precedents as to their ment, and by so doing threatened the future constitutional standing. However, as already advancement of science and technology? noted, the real question is whether taken together, the effects of these restrictions place a limitation on scientific freedom so as to contravene the intent of the First Amendment and ‘Bantam 1300ks ~’. SulIi\ran, 372 U.S. 58, 70 ( 1973); Vance ~. [ /nj\er.saj .~musement Co., 445 U.S. 308, 317 ( 1980); Organiza- its declaration of the rights of free speech and tion for a Better ,4 ustin t’. Keefe, 402 U.S. 415, 419 ( 1971). press. 1“293 U.S. 697 (1931). Exceptions for example include ob- struction of military recruiting, publication of the sailing dates The Background of the Issue of military transports or of the number and location of troops; publication of obscene matter, and incitements to violent or for- cible o~erthrow of the go~ernment, At least since World War II, it has been gen- ‘ ] 249 us. 47 ( 191 9). erally accepted that national security requires 40

industry and for national defense was clear. As political tension grew between the West- After the war, the growing centrality ern allies and the Soviet bloc countries, there of science-based technology both for were early signs that traditional assumptions industry and for national defense was about science and the First Amendment would clear. be challenged. In 1948, scientist and states- man Vannevar Bush noted that: that secrecy be imposed on some research. . . . the critical point [where fundamental science gives rise to applications] may well be Government has been able without serious reached far earlier in the process than we are challenge to restrict dissemination or results accustomed to think, and. . . we must be alert of research conducted by Federal employees, to it and ready at once to erect the defenses even in peacetime. Beginning in 1940, a series of protection and security which it demands. 13 of Executive Orders established criteria and classifications for assuring the secrecy of gov- Even during the 1940s other leading scien- ernment documents. The first War Powers tists complained about excessive secrecy. They Act, immediately after Pearl Harbor, also gave argued that restricting access to scientific the President the power to censor all commu- knowledge might do more harm than good for nications with foreign countries.12 Later this America’s continued leadership in science and power to censor direct communications across technology. Determined collection of informa- national borders was expanded to encompass tion by hostile nations can seldom be effec- publication of information that would prej- tively blocked, they said, but internal flows of udice our military/defensive interests or aid an ideas and research results may be unintention- enemy. ally obstructed. There has been only slightly more question- By the mid-1970s, there was strong concern ing of restrictions on the dissemination of re- over international competitiveness in both search paid for by the Federal Government but world markets and domestic markets. The done in universities and other nongovernment United States was no longer the unquestioned institutions. Scientists have also voluntarily leader in all areas of advanced science and tech- withheld scientific information in the interest nology, as it had been in the 1950s and 1960s. of national security. Even before the United Increasingly, scientific leadership translates States entered the Second World War, a spe- directly into military advantage. Thus trade cial coremittee of the National Research Coun- and technology policy clearly overlaps national cil, working cooperatively with editors of security policy. Congressman Don Bonker, professional journals, reviewed papers for pos- chairman of both the House Foreign Affairs sible defense-related information that should Subcommittee on International Economic Pol- not be published. This self-restraint, it should icy and Trade and the House Export Task be emphasized, applied to information with a Force in 1986, says flatly: clear tie to offensive or defensive weapons. Our defense strategy rests on qualitative The Federal Government exercised tight technical superiority over the Warsaw Pact countries, and we must insure that the West- security over scientific research during the war, ern alliance maintains this technological most notably over the Manhattan Project and edge. 14 other activities related to development of atomic weapons. After the war, the growing centrality of science-based technology both for 13u . s. cOnwess, HOuge Gmunit,tee on Government tions, “Availability of Information From Federal Departments ‘2U.S. Congress, Office of Technology Assessment, The Reg- and Agencies, Part 8,” Hearings, 85th Cong., 1st sess., Wash- ulatory Environment for Sa”ence-A Twhm”cal Memorandum, ington, DC, pp. 2159-2160. OTA-TM-SET-34 (Washington, DC: U.S. Government Print- 14 Don Bonker, “Protecting Economic Interests, ” Issues in ing Office, February 1986). Science and Technology, fall 1986, p. 97, 41

that funding comes from the Department of Defense (DoD).15 The boundary between The boundary between traditional cat- traditional categories of “basic” and “applied’” egories of “basic” and “applied” re- research has blurred, making it harder to research has blurred, making it harder strict only the latter. There is interdependency to restrict only the latter. between government research and that done in universities and independent laboratories even without government funding. The tension between scientific free speech DoD statements about national security re- and national security protections has become strictions on scientific communications are increasingly troublesome in this decade. One sometimes misunderstood by the unwary be- reason is that there has been a significant ex- cause of the tendency of many DoD officials pansion in the meaning of national security. to make a sharp, but sometimes unspoken, dis- The term no longer applies merely to direct mil- tinction between “scientific information” and itary threat. It also means the long-term risks ‘‘technical data, ” to associate the former only of change in the military, economic, and polit- with “fundamental research, ” which by DoD ical balance of power between nations. To this definition is unrestricted, and to associate the balance, relative scientific and technological latter with applied research and development, capabilities are deemed critical. or with technology. Thus “scientific informa- Many or even most areas of advanced indus- tion’ is by a truism, unrestricted. With regard trial technology have potential military appli- to most areas of advanced scientific research, cations. Most nations potentially hostile to the and particularly those that have to do with United States and its allies make no real dis- computers and communications technologies, tinctions between government, military, and it is increasingly difficult to understand, to de- scientific institutions. Any data exchanged be- fend, or to make such a distinction between tween U.S. scientists and scientists of those scientific information and technical data. nations has also been communicated to and between government institutions of those l~The Feder~ Government supported about 48 Percent of ~1 countries. R&D in the United States in 1986 (industry support accounts for another 48 percent). The average annual rate of growth in The scope of national security restrictions constant dollars is 5.9 percent for all R&D spending, 5.8 per- has been significantly broadened in the past cent for industry, and 7.3 percent for the Federal Government. DoD’s R&D obligations increased by 13 percent annually from decade. These restrictions apply to a growing 1982 to 1986, and by 20.9 percent for 1986 to 1987. Govern- proportion of scientific activity. The high cost ment provided 63 percent of funding for academic research in of research at the leading edge of science and 1986. Defense R&D in 1986 probably accounted for 73 percent of all Federal R&D support. National Science Foundation, Sci- technology has led to more of it coming from ence and Technology Data Book, NSA 86-311 (Washington, DC: government funding. A growing proportion of National Science Foundation, 1986).

THE EXECUTIVE BRANCH AND CLASSIFICATION OF DOCUMENTS The classification of information, in catego- dential Executive Orders. It is intended to ap- ries ranging from “confidential” up through ply to information that would create or increase increasingly stringent classes of ‘secret” and agency heads (primarily DoD, the intelligence agencies, and the “top secret, ‘’16 is done under a series of Presi- Department of Energy [DOE]) to provide greater protection for certain kinds of information or to conceal the means and chan- ‘GWithin the Secret and Top Secret categories there are many nels through which information is acquired, or for similar rea- “compartments’ or “special access programs” or subcategories sons. See Sec. 4.2 of Executive Order 12356, Apr. 2, 1982. Ac- to further restrict access to information to those with a need cess to restricted information defined by the Atomic Energy to know, defined with appropriately varying degrees of rigor. Act, as described later, is governed by DOE Q Clearance and These special access programs may be set up by designated is also compartmentalized. 42 a military risk to the United States or prejudice its foreign policy objectives. The Presi- Many researchers argue that the pos- dent, as Commander-in-Chief, is considered to sibility of classification has a strong have full authority to classify information generated by the government. “chilling effect” on scientific communications in general. In 1970, a Defense Science Board task force concluded that too many documents were clas- 17 sified. It recommended revision of the secu- policy of several preceding Administration rity classification process to declassify as much had been to lean toward not classifying or as 90 percent of classified information. As a toward a lower classification. For these rea- general rule, the task force said, basic research sons, and because the order had been devel- should never be classified. Even “confidential” oped in a particularly closed and secretive man- or special access limitations were inappropri- ner, it was widely criticized in Congress. 23 ate for basic research and likely to seriously impede technical programs. Security is most The number of decisions made in one year essential for information at the applied end of to classify documents probably hit an all-time the spectrum, that is, close to design and pro- peak in 1982 of over 1 million. By comparison, duction. Throughout the 1960s and most of the there were approximately 900,000 new classifi- 24 1970s, the trend was toward classifying less cations in 1985. In the Department of De- material.18 fense, 2,300 officials now have the authority to make classification decisions.25 The total Executive Order 12356 and volume of classified documents is huge and growing rapidly. the Corson Panel

19 Most of these are not from university-gen- Executive Order 12356, issued in April erated information. A DoD study in 1984 sur- 1982, reversed a 30-year trend toward narrow- ing the scope of classification and the discretion sified pending a determination by an original classification au- of bureaucrats in assigning secrecy classifica- thority, who shall make this determination within thirty (30) 20 days. If there is reasonable doubt about the appropriate level tions. It eliminated an explicit provision, of classification, it shall be safeguarded at the higher level of which appeared in earlier classification orders, classification pending a determination by an original classification authority, who shall make thisdetermin ation within thirty forbidding classification of information in (30) days. which the government has not acquired a prop- WJ.S. Congress, House of Representatives, Committee on erty interest. It lowered a standard, adopted Government Operations, Executive (h-cler on Security Classifi- cation, Hearings, 97th Cong., 2d sess., Mar. 10 and May 5, 1982. in 1978, that required an identifiable harm to Also House Report 97-731. Representative Glenn English, Chair- national security from disclosure. It added two man of a House Subcommittee on Government Information and new categories, for cryptology and for infor- Individual Rights, said, “The basic message to bureaucrats 21 would be: When in doubt, classify. ” Morton Halperin, director mation bearing on systems vulnerability . The of the Center for National Security Studies, said that the order order instructed classifiers, when in doubt, to would eliminate key provisions intended to prevent national err on the side of classification,22 whereas the security concerns from encroaching on scientific research. —- — “According to Steven Garfinkel, the chief of the U.S. Infor- ITU.S. Department of Defense, Report of the Defense Board mation Security Oversight Office, in a talk at the annual meet- Task Force on Secrecy (Washington, DC: July 1970). ing of the American Association for the Advancement of Sci- IHHarold C. Relyea, National Secw”ty Controls and Scientific ence, Feb. 18, 1987, as reported in The Institute, a publication Information, Congressional Research Service Issue Brief of the Institute of Electrical and Electronic Engineers, Inc., B82083, updated June 17, 1986, p. 6. vol. 11, No. 4, April 1987, p. 1. 19A7 Fed. Reg. 14sTA.14w4 (Apr. G, 1982). 25The count was made by the Security Review Commission, ‘i’ Harold C. Relyea, National Security Controls and Scientific a group created by the Secretary of Defense in 1985 and headed Information, Congressional Research Service Issue Brief by General Richard G. Stilwell; according to information pro- B82083, updated June 17, 1986, p. 6. vided to OTA by the Office of the Secretary of Defense, May ~lThese categorie9 were lisbd explicitly in Executive Order 7, 1987, citingS.F.311. For further discussion see “Espionage 12356; some national security experts had argued for some time and Security Leaks: Diagnosis and Therapy, by Harold P. that such information was implicitly subject to classification. Green, of the National Law Center and the Graduate Institute ~~~ .1 1. (c) say9: If there is reasonable doubt about the need for Policy Education and Research, The George Washington to classify information it shall be safeguarded as if it were clas- University, 1986. 43 — veyed over 123,000 DoD research reports. Only 19 percent came from university researchers; The power of government to protect 3.5 percent of these fell under distribution limitations, and only 1.3 percent were classified.26 State and military secrets has always Many universities will not do classified re- been regarded as inherent and funda- search, which explains the low volume. Some, mental. however, do permit classified research if approved by the school’s administration. An unknown number of papers are probably “sani- Information that meets all of these criteria tized” before publication, and many research- should be classified. But since most universi- ers argue that the possibility of classification ties will not do classified research, the Panel has a strong “chilling effect” on scientific com- recommended as an alternative, written agree- munications in general. ments between the university and the government that (a) prohibit participation by nationals A special panel of the National Academy’s of some foreign countries in such research, and Committee on Science, Engineering, and Pub- (b) require pm-publication review of articles by lic Policy was formed in 1982, before Execu- tive Order 12356, at the request of DoD and the Federal agency. intelligence agencies. The panel, chaired by DoD concluded that the Corson panel criteria Dale R. Corson, president emeritus of Cornell were ‘too difficult to translate into operational University, conceded that there had been much considerations’ and decided simply to “retain “involuntary technology transfer” from the its black/white policy towards university United States to potentially hostile countries. research-i. e., if not classified, then no re- But the panel also said that relatively little striction. ’27 of the deleterious leakage came from universi- The present Administration continues to em- ties or from open scientific literature. phasize classification of government docu- The panel recommended that there be no ments. It has, in addition, made increasingly restrictions limiting access to any area of explicit and forceful use of other means of re- university-performed basic or applied research, stricting scientific communication. unless: ● the area is developing rapidly, and the The Authority for Presidential time from basic science to application is Classification of Documents short; The power of government to protect State ● the information has identifiable, direct and military secrets has always been regarded military applications or is dual-use, involv- as inherent and fundamental. Although mili- ing process and production techniques; tary and State secrets have been protected in ● the information would give the U.S.S.R. the United States at least from the time of a significant near-term military benefit; Washington’s Presidency, there is little clear and statutory authority, aside from the atomic ● the United States, or other friendly na- energy area, for classification of informa- tions with secure control systems, is the tion.” The closest approach to a statutory only source of the information. basis is probably Section 161 of Revised Stat-

2TThe quotations are from a review of an OTA draft of this report by officials in the Office of the Secretary of Defense, May 2GMitchel B. Walierstein and Lawrence E. McCray, “Scien- 1987. tific Communication and National Security: Issues in 1984, ” yuconstitution~ 1mph”cations of Federal Restrictions on Sci- NAS News Report (Washington, DC: National Academy of Sci- entific Research and Commum”cation, March 1987, by Harold ences, April 1984). The study did not say what proportion of P. Green, The George Washington University National Law university research is restricted. Center, Washington, DC. 44 utes,29 dating back to the early days of the Republic. As originally written, it authorizes The Atomic Energy Act of 1946 added the head of each government department: a new concept to traditional classifi- to prescribe regulations . . . for the distri- cation procedures—the idea that some bution and performance of its business, and information was “born classified.” the custody, uses, and preservation of the records, papers, and property appertaining to it. Troubling First Amendment issues are pre- The basic authority for classification of in- sented if government attempts to restrict the formation has been the President via a series freedom of scientists to do independent, pri- of Executive Orders—not grounded on explicit vate research or to communicate information statutory authority but on the authority that is privately generated. Except in the area vested in the President by the Constitution and of atomic energy, however, the government has laws of the United States. The first of these not generally attempted to extend classifica- Executive Orders was No. 10290, promulgated tion to scientific endeavors conducted without by President Truman in 1951, which limited government involvement, although recent ac- ‘‘classified information’ to “official informa- 30 tions to be described below have introduced tion,” assumed to be information in which some uncertainty about this policy. the government has some kind of proprietary interest. In President Eisenhower’s Executive Order 10501,3’ use of the word “official” The Congress and Legislated Secrecy again connoted that classification was limited The Atomic Energy Act to information that in some way belonged to The Atomic Energy Act of 1946, modified the government. President Carter’s revision, 35 in Executive Order 12065,32 made this ex- in 1954, added a new concept to traditional plicit by specifying that a product of non- classification procedures —the idea that some government research and development accom- information was “born classified. ” The 1946 Act included provisions, frankly headed “Con- plished without access to classified information 36 was not subject to classification ‘until and un- trol of Information,” which established a less the government acquires a proprietary in- category of information called “Restricted terest. . . " Data, ” defined to mean” . . . all data concerning (1) design, manufacture, or utilization of President Reagan’s Executive Order 12356,33 atomic weapons; (2) the production of special however, includes no such limitation, but nuclear material; or (3) the use of special nu- brings within its ambit any information that clear material in the production of energy . . . ;" “is owned by, produced by, produced for, or except when such data has been expressly is under the control of the United States Gov- declassified. Misuse of Restricted Data was ernment” (emphasis added). This provides subject to heavy criminal penalties that some “color of authority” for classification of paralleled the more universal provisions of the information that is privately developed with- Espionage Act.37 out any involvement or funding by the gov- ——. 34 ‘bThe fi54 Act, 68 Stat. 919; 42 U.S.C. 2011-2296, relaxed ernment. the provisions of the 1946 Act in some regards to allow private sector development of atomic energy. ——— ~G42 USC. 2004 (y). The Atomic Energy commission (AEC) subsequently incorporated into 5 U.S.C. 22; and since 1966 was given explicit authority to promulgate regulations and is- found at Sec. 301 of Title 5. sue orders for the protection of Restricted Data. The Act re- J016 Fed. Reg. 9795, et seq. (Sept. 27! 1951)” quired, or at least was interpreted as requiring, that no person :31 18 Fed, Reg. 7049, et seq. (Nov. 101 1953)” could be given access to Restricted Data without a prescribed ‘Z47 Fed. Reg. 14874, et seq. (Apr. 6, 1982). investigation into his/her character, associations, and loyalty 3s47 Fed. Reg. 14874, Sec. 61(b) . on the basis of which Security clearance was to be granted. ~~These ExWutive Orders all provide that authority under ‘T42 U.S.C. 2275-2277. Unlike the situation with respect to the Atomic Energy Act pertaining to Security regulation of pri- ordinary classified information, there is an explicit nexus be- vate information (discussed below) is not affected by Execu- tween Restricted Data and those special espionage-type pro- tive Orders. visions. 45 —

A provision of the Act authorized the issu- extend to privately developed information; ance of court injunctions to restrain any threat- Senate staff members who played a major role ened violation of any provisions of the Act or in drafting the 1946 Act41 later wrote: its implementing rules or regulations. Since the It does not matter whether these (Re- injunction provision was applicable to the in- stricted) data are discovered or compiled in a formation control provisions, it can be read as government laboratory or in connection with encompassing prior restraints on the commu- the private research of an individual scientist. nication of Restricted Data. On July 21, 1947, the Attorney General wrote These provisions are unique in a number of the Chairman of the Atomic Energy Commis- respects. No other Federal statute has ever sion (AEC) stating that there was “consider- purported to control information in this way. able indication” that Congress meant the in- The definition of Restricted Data is extremely formation control provision of the Act to apply broad and could embrace a great deal of infor- to nongovernment information, but recom- mation contained in conventional textbooks mending a‘ ‘simple amendment “of the Act to on physics and chemistry. Normally, classifi- remove any doubt. 42 cation of information requires an affirmative determination that it needs security protection, Nevertheless, there is no statutory provision but Restricted Data is “born classified”38 if it that explicitly authorizes restrictions on pri- falls within the statutory definition. While or- vately developed information. Some experts dinary classified information (at least until hold that merely inferring such authority from statutory provisions would not pass constitu- issuance of President Reagan’s Executive Or- 43 der) has always been limited to “official” in- tional muster. formation, the definition of Restricted Data The AEC and its successors, the Energy includes ‘‘all’ data, thereby raising the ques- Research and Development Administration tion of whether it includes data generated (ERDA) and later the Department of Energy wholly in the private sector without any gov- (DOE), have proceeded on the assumption that ernment support or involvement.39 privately developed information is subject to 44 The answer to this question is not entirely the full array of controls. The information clear. Legal arguments can be constructed for control provisions have been invoked for reg- both positions. Under the Atomic Energy Act ulatory purposes— such as retarding activities in the United States that could lead to prolifer- of 1954, Congress accepted at least implicitly 45 the proposition that Restricted Data included ation of nuclear weapons in other countries . some data generated wholly outside the government, since one provision of the Act explicitly treats information developed in other . —. 40 4*J. Newman and B. Miller, The CorItrof of Atomic Energ~’. countries as Restricted Data. (New York: Whittlesey House, 1948), p. 15 ff. Newman and Miller were on the staff of the Senate Special Committee on Atomic Moreover, the authors of the 1946 Act seem Energy when the Act was drafted. Newman and Miller wrote, to have intended that the information controls 4, . . . if the Act does not restrict the liberty of scientific thought, it without question, abridges freedom of scientific communication. The controls on information were deliberately designed ‘“Richard Hewlett, “ ‘Born Classified’ in the AEC: A His- to regulate the interchange of scientific ideas . . .” (p, 208). torian’s View, and Harold Green, “A Legal Perspective, Bu.U+ 42 Accor&ng t. professor Harold P. Green, of the ~;eorge tin of Atomic Scientists, December 1981, pp. 20-27 and 28-30, Washington University National Law Center; documentation 3gOne provision, 42 U,S. C. 2201(i), authorizes regulations “to is available in Department of Energy Archives. protect Restricted Data received by any person in connection 43 Greene v. McElroy, 360 U.S. 474, at 506-608 (19591. with any activity authorized pursuant to this Act. ” This is prob- 4JDraft memorandum, “Authority To Control Dissemination ably not sufficiently broad to reach privately generated infor- of Privately-Developed Restricted Data, ” Feb. 28, 1966, from mation. Another provision, 42 U.S.C. 2201(p) authorizes regu- Franklin N, Parks, AEC Assistant General Counsel, to Joseph lations “necessary to carry out the purposes of this Act. ” It F. Hennessey, AEC GeneraI Counsel, available in DOE Archives. is this latter authority that the AEC relied on for the proposed 4hRichard Hewlett, “ ‘Born Classified’ in the AEC: A His- Part 26, to be discussed. torian’s View, ” Bulletin of Atomic Scientists, December 1981, ’042 U.S.C. 2162(e). p. 22. 46

The AEC on several occasions declared Where government action involves a wholly private scientific information to “prior restraint” on communication, be Restricted Data. the constitutional test is most severe.

In congressional hearings in 1955,46 the AEC The Court said that classification did not ab- General Counsel asserted that a scientist work- solutely prohibit the plaintiff use of the con- ing in his own laboratory, with no government cept but only regulated its use, and was thus connection, could be compelled to submit to not a taking. Further, when the purpose of a AEC security requirements, including classifi- regulation is to prevent injury to the public, cation and personal clearance, if he was creat- compensation is generally not constitutionally ing data that would be regarded as Restricted required. Data if created in an AEC facility. The Federal Government bases the restric- The AEC on several occasions declared tion of privately generated information on the wholly private scientific information to be Re- Schenck v. United States49 finding that im- stricted Data.47 In these cases the applicabil- pingement on freedom of speech “would ap- ity of atomic energy information control pro- pear to be a reasonable exercise of sovereign visions has affected primarily profit-seeking power . . . in the interest of the common de- business organizations rather than universi- fense and security.”50 The provisions of the ties. Although in each case the affected com- Atomic Energy Act taken together seem to au- pany sustained real economic injury, there has thorize a Federal court to issue an injunction been only one judicial challenge. The corpora- restraining a defendant’s communication or tions apparently concluded that private busi- publication of even privately developed infor- ness interests would not prevail in court over mation, which would be prior restraint.51 Is government’s national security claim. this authority constitutional? In the only case in which a court has consid- As already noted, where government action ered the constitutional issue,48 the Court of involves a “prior restraint” on communication, Claims rejected a contention that the classifi- the constitutional test is most severe. The gov- cation of an industry process was a “taking ernment sought an injunction to restrain pub- of property” entitling the corporation to just lication of the Pentagon Papers by The New compensation under the Fifth Amendment. York Times and The Washington Post. In The New York Times Co. v. United States,” the Supreme Court held in a 6-3 decision that this d6Hewings on s. J. Res. 21 to Estabh”sh a COm~”SSiOn on burden was not met. Government Security, Subcommittee on Reorganization, Sen- ate Committee on Government Operations, 84th Cong., 1st sess. In 1950, the AEC ordered Scientific Amer- (1955), pp. 240-241, 267-271. ican not to publish, without specified deletions, 47 Richard Hewlett, “ ‘Born Classified’ in the AEC: A His- torian’s View,” and Harold Green, “A Legal Perspective,” Bulk tin of Atom”c Scientists, December 1981, pp. 22,27 and 29. One example involved a corporation attempting to use gas centrifuge 4’249 U.S. 47 (1919). technology for separating uranium isotopes. The most explicit 50 Accor&ng t. Richwd Hewlett, “ ‘Born Classified’ in the assertion by the AEC of the right to control private informa- AEC: A Historian’s View, ” Bulletin of Atomic Scientists, De- tion came in 1967, when the AEC proposed a new regulation, cember 1981, pp. 23-25. Hewlett notes that this reasoning was- Part 26, dealing with “Private Restricted Data, ” whereby in plied to AEC’S proposed “Part 26” (see footnote above) in 1966. effect private research and development in the fields of nuclear The Atomic Energy Committee of the Bar of the City of New explosives and gas centrifuge technology for producing nuclear York sent the AEC a detailed brief challenging the constitu- fuel, was totally prohibited. Eventually, this AEC proposal was tionality of the proposed rule. withdrawn. Proposed 10 CFR Part 26, App. A, 32 Fed. Reg. 6707. blsuch ~ injunction was in fact issued in the ~ro~essive A~Ra&”oPtics, Inc. v. Um”t~ States, 621 F. 2d 1113!223 Ct. case, to be discussed. 467 F. Supp. 990 (W.D. Wise. 1979). Cl. 594 (1980). 52403 Us. 713 (1971). —

47 an article on thermonuclear weapons. The ar- hydrogen bomb was “analogous to publication ticle, which had already gone to press, was by of troop movements or locations in time of Dr. Hans Bethe, an eminent theoretical phys- war, ” thereby falling within the “extremely icist long involved in the nuclear weapons pro- narrow exception to the rule against prior re- gram. The publisher protested that all of the straint. ’56 technical information in the article “was well The decision was never subjected to review known to physicists . . . and had been widely by higher courts, since it was mooted by pub- published. ” The AEC insisted that Bethe’s lication of essentially the same information in authorship confirmed the authenticity y of pre- another journal. According to Congressman viously published information. Scientific 57 George Brown: American capitulated under the AEC’s threat to seek an injunction, and AEC security . . . [because] the Supreme Court had ruled in officers supervised the destruction of the type the Pentagon Papers case that prior restraint and plates and the burning of 3,000 copies of was not tenable, it is not clear what would have that issue.53 happened to the The Progressive case if it had been reviewed by the Supreme Court. Until 1979, 33 years after the enactment of these information control provisions, there was Professor Harold Green of the National Law no litigation challenging the constitutional Center says: validity of prior restraint on publication of Re- [T]he case stands as vivid testimony to the stricted Data. United States v. The Progres- potential impact on scientific freedom of the 54 sive (1979) made a weak case for the chal- information control provisions of the Atomic lenge. The Progressive proposed to publish an Energy Act. article55 in which a journalist purported to describe how hydrogen bombs are made and The Invention Secrecy Act work. The intent was to alert the public to the 58 “false illusion of secrecy” created by the gov- The Invention Secrecy Act passed in 1951 ernment and the necessity for decisive action allows the Federal Government (through the to halt the proliferation of nuclear weapons. Commissioner of Patents) to block the grant- The information was derived entirely from the ing of a patent or the disclosure of technologi- public domain. The government held that: cal information by an inventor, when this disclosure “would be detrimental to national When drawn together, synthesized and col- security, ” even where the government has no lated, such information acquires the charac- property interest or right in the invention.59 ter of presenting immediate, direct, and ir- In the 1970s the National Security Agency reparable harm to the interests of the United (NSA), for example, frequently asked for States. ‘‘secrecy orders” for cryptographic inventions, The court found “no plausible reason why but this decreased sharply in the late 1970s the public needs to know the technical details about hydrogen bomb construction to carry 5’Um”t~ States v. The Progressive, Inc., 467 F. Supp. 990 on an informed debate’ on the issue of prolifer- (W.D. Wise. 1979), 996. ation. With respect to prior restraint, the court s?conpession~ Record, vol. 128, No. 16, Feb. 25, 1982, returned to Near v. Minnesota, in which the H511. “Invention Secrecy Act of 1952 (35 U.S.C. 181-188). Supreme Court had spelled out certain situa- ‘gThere are two categories of inventions subject to this act. tions in which restraints on expression might In the first, government has a property interest, and the inven- be constitutionally permissible; and said that tion is therefore already subject to classification; in the second, government does not have a property interest. Patent officials publication of technical information on the estimate that roughly 4,800 inventions are the subject of invention secrecy orders at any one time; that about 1,000 of them “~Scientific American, May 1950, p. 26. are foreign inventions covered under reciprocal agreements with M467 F. &Ipp. 990 (w.D. Wise. 1979)- various friendly nations; and that of the remainder, approxi- 35 The article, not published, was Howard Morland, “The H- mately 20 percent fall into the second (nongovernmental prop- Bomb Secret: How We Got It, Why We’re Telling It. ” erty interest) category. 48

when the procedures for requesting secrecy orders were changed and made more rigorous. The constitutionality of the Invention Secrecy orders are effective for only one year, Secrecy Act has never been tested. but may be renewed. However, although this limitation was written into the Invention Secrecy Act when it was passed in 1951, for is subject to the conditions established by the the next 27 years it was inapplicable. The law government. Some experts have held that con- contained an “exception” clause to the effect stitutional objections are further obviated by that the yearly renewal requirement was not the statutory requirement that the patent ap- operative for the duration of a war or national plicant receive just compensation for any loss emergency, and for some months thereafter; that might be suffered by reason of the secrecy and the National Emergency proclaimed dur- order. ing the Korean War was not officially terminated until 1978. The annual renewal cycle has However, a House of Representatives Re- 61 been in effect since that time, but has been pro- port in 1980 noted that: tested by national security officials, at least No secrecy order ever underwent judicial re- as it applies to inventions already subject to view for appropriateness. There has been no periodic reexamination for classification down- First Amendment judicial test of the Inven- grading (i.e., those in which government has tion Secrecy Act, and the statutory right of a property interest). an inventor to just compensation for secrecy order damages appears more illusionary than Many secrecy orders are issued in connec- real. tion with already classified patent applications. Sometimes, however, the patent application has The report noted that from 1945 to 1980 been filed by persons who developed the inven- there had been 29 claims for compensation, one tion without any government involvement. claim for every 1,000 secrecy orders; 6 were settled before or during litigation by DoD and In February 1987 the Army requested and 1 by a private relief bill, 10 were terminated the Patents Commissioner granted a secrecy by denial, and the rest were still pending. As order on an application for an American pat- the report also noted, “Agencies have little or ent in the field of ‘zero-knowledge proofs, ” by no incentive to settle a claim, and claimants, an Israeli mathematician, working in an Israeli 60 frustrated, often drop the matter. The House institute. While such proofs are used in Committee on Government Operations, which cryptology, they are generally regarded as ad- prepared the report, found “little judicial guid- vances in theoretical or “pure” science. The ance on First Amendment questions. ” It quoted applicant was ordered to ‘recover and destroy’ an official of the Department of Justice on the all related materials. The secrecy order was question but was told that the Department quickly withdrawn, since the government can- “has thought it wise to follow a rule of self- not classify work done by foreigners in their restraint in expressing public views on con- own country, but not before there were many stitutional questions presented by the statutes protests from mathematicians. This episode we are called upon to enforce."62 was taken to indicate that American achieve- ments of the same kind might be restricted. The constitutionality of the Invention Se- crecy Act has never been tested. Since the grant of a patent secures a property right con- ferred by the government, a patent applicant “The Government Classification of Private Ideas, Thirty- Fourth Report by the (House of Representatives) Committee on Government Operations, House Report No. 96-1540, 96th ‘“’’ Brief U.S. Suppression of Proof Stirs Anger, ” New York Cong., 2d sess., Dec. 22, 1980, p. 2. Times, Feb. 17, 1987, p. C3. “Ibid., p. 27. 49 -.

Restricted Data or under export controls.) Nevertheless, imposition of a secrecy order This authority must be considered as does operate as a restriction on traditional free- part of the total burden on the exer- dom of scientific communication. This author- cise of free speech. ity must therefore be considered as part of the total burden on the exercise of free speech.

Imposition of a secrecy order can be avoided simply by not seeking a patent. (The information may nevertheless be subject to control as

EXPORT CONTROLS Export controls are also a form of legislated secrecy in scientific research, with special at- restraint on scientific communication. They are tention to the problem of “involuntary tech- considered separately because national secu- nology transfer” to hostile or competitive rity is only one objective of these controls. nations. The Bucy panel argued that the knowledge most vital to protect is not embedded in The Arms Control Act military weaponry per se, but knowledge that conveys design and manufacturing know-how. This Act provides authority for restricting The export of technological information con- or prohibiting export of technical data related tained in scientific publications in some areas to defense articles.63 It applies to blueprints, is harmful to the United States. These areas drawings, photographs, plans, instructions, were “arrays of design and manufacturing software, and documentation. This Act is one know how, ” “ keystone” manufacturing proc- of the U.S. Statutes used to restrict unclassi- esses, inspection and test equipment, and prod- fied information. ucts requiring sophisticated operation, appli- Under the Arms Control Act, International cation, and maintenance. Traffic in Arms Regulations (ITAR) are devel- This recommendation was a significant ex- oped and a U.S. Munitions List is maintained pansion of the term “militarily useful. ” The by the Department of State, with the help of transfer of design concepts and manufactur- DoD. The regulations” operate in much the ing processes can relate directly to the manu- same way as regulations under the Export facture of weapons.” But a further, and ulti- Administration Act, discussed below. ITAR, mately more important point, may be the does not specifically exclude ‘fundamental re- recognition that modern concepts of national search, ” but it does exempt general mathe- security depend at least as much on the matical and engineering information that is strength of the Nation’s industrial base as on only indirectly useful to military purposes. It the stock of military weapons. There is a close is not clear whether any information exempted tie between scientific information and indus- from Export regulations is restricted under try strength and competitiveness. ITAR.

The Export Administration Act — ““’A high government official noted to OTA, “The informa- In 1976 a Defense Science Board panel tion used in- manufacturing high-tech products used in weap- headed by Fred Bucy reexamined the need for ons systems diffused into the civilian private sector and could no longer be controlled by DoD, at least to the extent that it —— once was. This recognition, above everything else, forced us to ‘“{22 U.s, c. 2751. redefine what was militarily useful. (Private communication, ‘“A revised version of ITAR became effective Jan. 6, 1985. June 19, 1987. ) 50

Multilateral Export Controls (CoCom), which under multilateral agreements and procedures There is a close tie between scientific provides for cooperative control of exports to information and industry strength the Soviet bloc with restrictions on munitions, and competitiveness. nuclear energy, and some other dual-use technologies. Many but not all of the items on the U.S. Commodity Control List are also on the The Bucy panel recommendations set in mo- International (CoCom) List. tion more vigorous efforts to control dissemi- The early use of export controls emphasized nation of technical knowledge related to mili- products more than information as such. Thus, tarily useful advanced technology, and led to when the Commerce Department halted ship- the strengthening of export control laws in the 66 ment of magnetic computer tapes to Eastern following years. bloc countries in 1982, “the action was taken The Export Administration Act had already to ban the medium, not the message.”69 Al- been passed in 1969, and was amended in 1979, though these were standard IBM computer 1981, and 1985. It controls the export of “goods tapes, DOC said that Eastern European coun- and technology which would make a signifi- tries could not manufacture such high-quality cant contribution to the military potential of tape and should not learn how from U.S. countries which would prove detrimental to the products.70 national security. . . "67 Technology is defined For U.S. exporters, restrictions on trade with to include “information and know how, the Eastern bloc nations may be less impor- whether intangible form. . . or intangible form, tant than the effects of unilateral restrictions that can be used to design, produce, manufac- on trade with other Western nations. The ture, utilize, or reconstruct goods, including 68 Administration insists on such restrictions to computer software and technical data . . " curb the indirect flow of technology to the So- The Export Administration Act not only ap- viet Union, because once information is out- plies to information passing across our borders, side the country it is outside of our control.71 but also limits access of foreign nationals to This imposes costs on U.S. exporters and information in this country. causes friction with our trading partners. Some Regulations for administering and enforcing the Export Administration Act are promul- ‘gRoland W. Schmitt, “Export Controls: Balancing Techno- gated by the Department of Commerce (DOC), logical Innovation and National Security, ” Issues in Science and !l’echnology, vol. VI, fall 1984, p. 117. Schmitt is head of and the products or areas covered are identi- the General Electric Research and Development Center. fied in a Commodity Control List, which is ToIbide Sm ~so cheml”c~ & Engineering News, ‘‘U.S. B~s maintained by DOC with the help of DoD and Tape Exports to East Bloc,” Sept. 20, 1982, p. 6. As a result, Chenu”cal Abstracts, which had supplied its bibliographic tapes DOE. The list specifies the countries to which to Warsaw Technical University since 1974, did not get its ex- each of about 100,000 items (in 1984) cannot port license renewed. The Institute for Scientific Information, be exported without a validated license. All a commercial firm providing bibliographic services, could no longer send its standard tapes to customers in Poland, the export of unpublished technical data to Com- U. S. S. R., and Hungary, but could send the same information munist bloc countries requires a license. The on low-quality tapes. United States is also a founding member of 71 Export Administration Regulations require priOr authorization from the Office of Export Licensing in certain situations the multinational Coordinating Committee on before foreign products that incorporate parts, components, or materials of U.S. origin, can be exported from one foreign coun- 66s~Ph~~ B. GO~d, “secrecy: Its Role in National Scientific try to another. Until recently, all foreign companies needed per- and Technical Information Policy, ” Library Trends, summer mission from DOC before reexporting a product that contains 1986, p. 72. Gould was until rnid-1987 Director of the Project any U.S. components, even a microchip. New regulations effec- on Scientific Communication and National Security, of the Com- tive in 1987 somewhat simplified the procedure for requesting mittee on Scientific Freedom and Responsibility of the Amer- such authorization and exempted foreign-manufacturer prod- ican Association for the Advancement of Science. ucts destined for some countries when the U.S. content is no ’750 U.S.C. 2402(2)(A) App. more than 20 percent by value. (15 CFR Parts 376 and 385; G850 U.S.C. 2415 APP. Federal Register 51, No. 129, July 7, 1986, 24533.) 51

foreign companies have notified their U.S. sup- tecting the ability of scientists and other pliers that they are “designing out” U.S. parts scholars freely to communicate their research to avoid this additional effort; others consider findings in accordance with applicable provi- this additional cost in deciding on suppliers.72 sions of the law, by means of publication, teaching, conferences, and other forms of Many U.S. industry leaders have protested ex- scholarly exchange. 76 port controls.73 Although the House wished to avoid “overly Export controls apply not only to commer- broad interpretation” of the Export Adminis- cial products but to technical data, research tration Act, the Senate Committee on Banking reports, and some other kinds of information added a provision making academic institu- as well. A Defense Science Board Task Force tions subject to requirements of the commer- on University Responsiveness to National cial agreements provision. This meant that Security Requirements, in early 1982, found university agreements with certain foreign that the shift in emphasis in export controls countries must be reported to the Department from products to technological information of Commerce. Some believe that this gives gov- was seriously disturbing the relationship be- 74 ernment a “ready opportunity to regulate, tween the government and universities. The through the application of export controls, the Task Force recommended that “clearer guide- content of lectures, conference presentations, lines,” not overly restrictive, be formulated for teaching, and publications by U.S. academics DoD-funded university research with the help in certain foreign countries. “77 Others insist of universities. This recommendation however that there is no necessary connection between led to new confusion and dispute over the ba- “recording commercial transactions after the sis for applying either classification or export fact, ” and “controlling scientific information controls. In theory, they involve quite differ- before the fact. ” ent objectives, standards, and procedures, but in practice they are often seen as alternatives. The House and Senate failed to agree on re- authorization of the Act in conference. The When the 1979 Export Administration Act President then issued Executive Order 12470 was about to expire in 1983, efforts to rewrite (Nov. 14, 1983), declaring a national emergency the bill revealed sharply divergent concerns. and activating the same powers under the In- Some in Congress sought to strengthen export ternational Emergency Powers Act. controls, while others argued that excessive controls reduced the competitiveness of U.S. The Export Administration Act was finally companies and contributed to our trade defi- reauthorized in April 1985. The new law at- cit.75 There were hearings in both the Senate tempted to streamline its procedures but did and House. The House accepted a provision not reduce the number of items covered. GAO proposed by the Committee on Foreign Affairs, has said that “The government continues to which read: require export licenses for more dual-use items than is necessary to protect national secu- It is the policy of the U.S. to sustain vigor- rity." 78 Dual-use items are devices, systems, ous scientific enterprise. To do so requires pro- or “know-how” that have both military and ‘zDon Bonker, “Protecting Economic Interests, ” Issues in nonmilitary applications. Science and Technology, fall 1986, p. 100. ‘qAccording to officials in the Department of Commerce Ex- Revisions to the implementing regulations port Administration Program Review Staff, “The possibility of reverse engineering [indirect leakage of useful information under the Act in 1986 exempted ‘‘fundamen- as a result of West-West trade] has been discounted by many exporters who claim that their manufacturing techniques cannot be derived by examining the finished products. ” (Direct com- 7’50 U.S.C. 2402 (12) App. munication to OTA project staff, June 19, 1987. ) ‘THarold C. Relyea, National Security Controls and Scientific ‘JU.S. Department of Defense, Report of the Defense Science Information, Congressional Research Service Issue Brief Board Task Force on University Responsiveness to National B82083, updated June 17, 1986, pp. 12-13. Security Requirements (Washington, DC, 1982). “U.S. Congress, General Accounting Office, Export Control ‘; Don Bonker, “Protecting Economic Interests, ” Issues in Regulation Could Be Reduced Without Affecting NationaJ Secu- Science and Technology, fall 1986, p. 100. rituv, May 26, 1982. 52 .

In January 1987 the National Acad- The Export Administration Act has emy of Sciences and the National a legitimate government purpose other Academy of Engineering issued a re- than restricting speech. port sharply critical of the implementation of export controls. tion to their effectiveness and cost”; by try- ing to impose its own export restrictions on tal research” from export controls, using the countries that import U.S. technology, the definition of fundamental research introduced Administration had injured both U.S. competi- in NSDD 189, September 1985, which is dis- tiveness and relations with our allies. The De- cussed below.79 University research is nor- partment of Defense has been given too large mally considered “fundamental’ unless scien- a role in export policy, the panel implied, be- tists have accepted prior restrictions on ing more concerned with national security than publications through contract agreements.80 with the competitive strength of American industry.” The DoD Appropriations Act for fiscal year 1984 gave the Secretary of Defense authority Many question the effectiveness of export to withhold from public disclosure, under the controls. One expert has been quoted as say- Freedom of Information Act (FOIA), any tech- ing that ‘the avenues for transfer open to Rus- nical data with military or space application sia are so broad that they are almost impossi- already under control of DoD, if that data ble to control,” adding that the primary avenue would fall under Export Acts. DOE policy is is probably through Western plant workers to use this authority only if the data has “mil- and junior executives recruited over the past itary centrality, and requires the recipient to 30 years in the United States and Western promise not to reveal the information to the Europe. 83 public. This avoids releasing the data from effective government controls.8l The Constitutionality of In January 1987 the National Academy of Export Controls Sciences and the National Academy of Engi- It is assumed that any impingement on First neering issued a report sharply critical of the Amendment rights under the Arms Export implementation of export controls. It said that Control and Export Administration Acts is in- the Administration had “tended to focus on cidental, because these regulations have a tightening controls while giving little atten- legitimate government purpose other than to 84 —- - — restrict speech and press. The Court might 7915 CFR 379, May 16, 1986. The regulations say that information resulting from fundamental research qualifies for unrestricted export to any destination under a General License ‘zNational Academy of Science, Panel on the Impact of Na- GTDA, that is, a “general license for technical data available tional Security Controls on International Technology Trans- to all destinations. ” fer, Balancing the National Im.%xwst: U.S. National Secun”ty Ex- “)Unclassified contract research funded by DoD under bud- port Controls and Global Econoxm”c Competition, Washington, get category 6.1 is considered fundamental research, and most DC, 1987. The report was prepared by a panel chaired by Lew unclassified research at universities funded under budget cate- Allen, Director of the Jet Propulsion Laboratory. The panel rec- gory 6.2 is also considered fundamental research. Contract re- ommended that the United States rely on, and seek to improve, search done in off-campus university facilities and not supported the Multilateral Export Controls Coordinating Committee (Co- under budget category 6.1 is however generally not automat- Corn) consisting of the NATO countries plus Japan and France. ically considered fundamental research, but may still be free ~Frederick Kempe, “Losing Battle: Keeping Technology Out of restrictions. Research done by business or industry is con- of Soviet Hands Seems To Be Impossible, ” Wall Street Jour- sidered fundamental research only if the researchers are free md, July 24, 1984, p. 1. to make it available without pay or restrictions (i.e., it is not ~Adder~y v. ~]ofl”da, 385 U.S. 39 ( 1966); Bro~ v. Lom”si~a~ proprietary) and it is not classified. 383 U.S. 131 (1966); Kovacks v. Cooper, 336 U.S. 77 (1949); “communication to OTA from officials in the Office of the Breard v. Alexandria, 341 U.S. 622 (1951); O’Brien v. United Secretary of Defense, May 1987. States, 391 U.S. 367 (1968). 53

tend to give them the benefit of the doubt, as data to that which “relates in a significant fash- 85 in the only case so far that directly chal- ion to some item on the Munitions List, ” as lenged the applicability of the Arms Export opposed to being “merely vaguely useful for Control Act to unclassified data. This case in- the manufacture of arms. Moreover, it is nec- volved business rather than scientific interests. essary that the relationship to that item be Edler provided technical assistance to French clear and that the defendant know or have rea- companies on tape-wrapping techniques that son to know that the data was intended for had both commercial and military applications, a prohibited use.87 despite rejection of Edler’s application for an export license. The U.S. Court of Appeals said The court explicit sensitivity y to the neces- that Edler had some First Amendment rights sity for protecting open dissemination of sci- with respect to the transaction, but concluded entific knowledge offers hope that the statute that the Arms Export Control Act was a‘ ‘gen- cannot be applied in a manner that interferes eral regulatory statute, not intended to control with traditional modes of scientific communi- the content of speech but incidentally limiting cation. The Office of Legal Counsel of the De- its unfettered exercise,"86 and might, there- partment of Justice addressed this issue in sep- fore, be constitutionally permissible. arate communications to the Secretaries of Defense and Commerce in July 1981, assert- Observing that a broad interpretation of ing that export control regulations may not “technical data” would “seriously impede sci- impinge on scientific communication unless the entific research and publishing and the inter- ‘‘ speech" is directly related to a business trans- national scientific exchange, the court adopted action.88 Nevertheless, some critics maintain a narrow construction that limited technical that this distinction has not been adhered to in the ensuing 6 years. — “Ibid.. ~D. 520 and 521 ‘HHarol~ ~. Rel~Tea (cd.), “Shrouding the F;ndlms 12rontier - Scientific Communication and National Securit~:: The Search for Balance, Striking a Balance: Nationai Secur]t? and S’ciem tific Freedom (Washington, DC: American Association for the Ad\rancen~ent of Science, 1985), p. 90-92. —.—.

Chapter 5 The 1980s: Converging Restrictions on Scientific Communications

CONTRACTUAL RESTRICTIONS ON COMMUNICATIONS

National Security Decision Directive 84 (NSDD 84),2 issued on March 11,1983, requires Most universities are reluctant to un- all present and future government employees dertake classified research, but many to sign a lifetime nondisclosure agreement as are willing to accept contractual re- a condition for access to classified information, strictions that have the same effects. or to “Sensitive Compartmented Information” (SCI). Federal classifying agencies have the right of pre-review of public statements, lec- Dissemination of scientific information or tures, and speeches. Contacts between media technical data’ can be and often is restricted representatives and agency personnel are con- by the terms of written agreements between trolled. This directive was aimed primarily at government funding agencies and nongovern- Federal employees, but secrecy agreements can ment researchers. Most universities are reluc- also be required of government contract re- 3 tant to undertake classified research, but many searchers under the directive. are willing to accept contractual restrictions According to a recent study supported by that have the same effects. In some cases, in- the Fund for Investigative Journalism, by the deed, refusal to accept such contracts is con- end of 1985 more than 290,000 individuals had sidered by faculty to be an infringement on signed lifetime prepublication review agree- their academic freedom. Some civil Libertar- ments (Non-disclosure Agreements Standard ians, on the other hand, object to such con- Forms 189 and 189A) under NSDD 84, and tracts. While a contract, freely entered into, more than 14,000 speeches and articles had is assumed to benefit both parties to it, this been submitted for review.’ There have been does not provide for consideration of the pub- strong protests against the requirement for lic interest (and investment) in scientific knowl- lifetime agreements. However, the Federal edge, which may not entirely coincide with or Government takes the position that the con- be limited to national security interests. tractual obligation of an employee overrides The government often requires that contrac- First Amendment protections. tors and grantees agree to submit publications In Snepp v. United States5 the Supreme resulting from nonclassified government-spon- Court found no constitutional impediment to sored research for prepublication review. This enforcing such an agreement. This case, how- raises the question of prior restraint. There are, ever, involved CIA employees, and the intelli- however, no reported court decisions involving prepublication review clauses in contracts 2The nature of National Security Decision Directives is dis- and grants to universities. cussed below. 3Secrecy agreements or contracts used within the intelligence community were found to be a proper enforcement device in 1980 by the Supreme Court (Snepp v. U. S., 444 U.S. 54) 4Donna DeMoc, “Sworn to Silence, ” The Progressive, May ‘As already noted, national security officials tend to make 1987, p. 29. The study presumably covered SF 189s: SF 189A a sharp distinction between 4 ‘scientific’” communications and was approved for inclusion in the Industrial Security Manual technical, technological exchanges that is not always either clear in late April 1987. or acceptable to researchers in an area. ’444 Us. 507 (1980).

55 56 gence community is generally conceded to have classified materials. The threatened expansion particularly strong interests in internal secu- of polygraph use brought strong protests. In rity. Two Circuit Court decisions, one before the fall of 1983, Congress temporarily prohib- and one after Snepp,6 indicate that such an ited implementation of the polygraph provi- agreement might not be enforceable if classi- sion, and President Reagan agreed not to pur- fied information is not involved. These cases sue this policy immediately.7 In effect, the did not involve research results or scientific provisions of NSDD 84 dealing with polygraphs information. have been rescinded, except as they pertain to DoD, where polygraph tests are widely used NSDD 84 also sought to expand the use of not only in investigations but for routine screen- “lie detectors” by Federal agencies. Executive ing of recruits, promotions, etc. branch employees can be required to submit to polygraph examinations for access to certain classified information, or in the course of ‘Harold C. Relyea, National Security Controls and Scientifi( investigations of unauthorized disclosure of Information, Congressional Research Service Issue Briel B82083, updated June 17, 1986, p. 69. See also National Acad ‘The first was United States v. Afarchetti, 366 F.2d 1309 (4th emy of Sciences, Committee on Science, Engineering, and Pub. Cir., 1972). cert. den., 409 U.S. 1063 (1972). The second was lic Policy, Panel on Scientific Communication and National Sew McGehee v. Casey, 718 F.2d 1137 (D.C. Cir., 1983). rity, Scientific Communication and National Security, 1982.

RESTRICTIONS ON INFORMAL COMMUNICATIONS

In 1980, the Department of State informed Cornell University that a visiting Hungarian Classification, legislative mandates, engineer would have to be limited to classroom contract agreements, and export con- pursuits and could not participate in certain trols have all been used in the last dec- professional seminars or receive prepublication ade to restrict informal communica- copies of research papers. Rather than abide tions among scientists. by these restrictions, the university canceled his visit. Later that year, the Department of State, acting under export restrictions, asked Classification, legislative mandates, contract universities to prohibit visiting Chinese stu- agreements, and export controls have all been dents from engaging in certain studies.8 There used in the last decade to restrict informal com- were reports of a few instances in which univer- munications among scientists. By informal sities, in ill-considered, hasty responses, listed communications is meant modes of communi- short courses or seminars “for U.S. citizens cation other than formal publishing: speaking, only;” such decisions were apparently few, and classroom teaching, participation in profes- were soon terminated.9 sional meetings and seminars, and similar On February 27, 1981, the presidents of five activities. Restrictions on campus teaching are leading universities (Stanford, the California particularly irksome to many scientists. Since Institute of Technology, the Massachusetts In- the 1970s, there has been a steady influx of stitute of Technology, Cornell, and the Univer- foreign students to U.S. universities. Any restrictions on who may be taught what in a ‘HRelyea, op. cit., p. 4. university are of profound importance, infring- There were references in the press at that time to university ing on academic freedom, on institutional courses or seminars so advertised, but recent discussions with a number of education association officials identified only one responsibility, and on the prestige and eco- specific incident of a course listed ‘‘for U.S. citizens’ at a univer- nomic viability of the institution. sity, and that decision was said to have been quickly overruled. 57

Scientific exchange is a primary pur- Under current DoD directives, unclas- pose and role of professional societies. sified papers containing export-controlled information cannot be presented at professional and academic sity of California) sent a letter of protest to meetings unless dissemination and ac- the Secretaries of Commerce, State, and De- cess controls are “equivalent to those fense. The university presidents said that the used for distributing the data directly government had resorted to measures that by DoD.” could ‘‘irreparably harm university-based research."10 Scientific exchange is a primary purpose and the organizers that other papers were sensi- role of professional societies. These organiza- tive. DoD representatives were present as the tions depend on their members to judge whether meeting began and questioned participants as communication of research results or other in- to whether their papers resulted from work formation violates national security restric- sponsored by DoD and whether they had re- tions, and they can be faced with dilemmas ceived clearance. One hundred papers were when their members either intentionally or in- withdrawn. A SPIE official later said that gov- advertently transgress. In February 1980 there ernment officials had overreacted to the pres- were strong efforts by the Carter Administra- ence of Soviet citizens, and that SPIE mem- tion to regulate the communication of scien- bers themselves had probably panicked at the tific information by professional societies. sudden crackdown. Many of the papers that When the American Vacuum Society orga- had been hurriedly withdrawn were later cleared and published or presented, although others nized an international meeting on magnetic 12 bubble memory devices, the Department of were found to deal with classified research. Commerce notified the society that the ex- Under current DoD directives, unclassified pected presence of nationals of certain foreign papers containing export-controlled informa- countries subjected the proceedings to com- tion cannot be presented at professional and pliance with export licensing. The Association academic meetings unless dissemination and promptly rescinded invitations to scientists access controls are “equivalent to those used from Hungary, Poland, and the U.S.S.R. Chi- for distributing the data directly by DoD"13-- nese scientists, already en route, were allowed so-called unclassified/limited access presenta- to attend after signing an agreement not to tions.14 This rule was applied on an ad hoc ba- "re-export" to nationals from 19 countries 11 sis during another meeting of the Society for what they learned. Photo-Optical Instrumentation Engineers in In August 1982, 4 days before a meeting of April 1985. Two weeks before the meeting, or- the Society of Photo-Optical Instrumentation ganizers were notified by DoD that a special Engineers (S PIE ), DoD learned that four So- session, organized by one of its members at viet scientists were to attend. DoD confiscated a military base, with 43 papers scheduled, of dozen papers from DoD employees who were to present them at the meeting, and notified — 1’This account is based on discussion with S1)l l’: dire(tor Joseph Yaver in May 1987; for a contemporar~. account, see Joel Greenberg, “Science’s New Cold War, ” Science ,VeWrs, ~o]. ] “(; ina Ilara Kolata, “ Attempts To Safeguard Technology 123, Apr. 2, 1983, p. 218. Also see Relyea, op. cit., 1986. )raw K’ire, .Scienc(j, ~ol. 212, No, 4494, hla~’ 1981, p. 523. The ‘‘Statement supplied for (3TA by DoD officials, office of the etter also said: ‘“’[’here is no eas~r separation in any engineer- Secretarv of Defense. ing curriculum. Furthermore, producing graduates with no 11[) .s fiepartment of Defense, C) ffice of the Under Secretary hands on’ experience in these areas would he of little \ralue to of Defense for Research and Engineering, ‘‘Policy and Guide- \merictin high technology industries. lines for the I’resentation of DoD Sponsored Scientific and Tech- Ibid., p, 3. nical l)apers, draft proposal. oct. 24, 19H5. would have to be canceled because the papers Nation of open exchange of scientific informa- could not be given in open sessions. The meet- tion, pointing out that such exchange is nec- ing organizers and an official from the DoD essary to validate findings, to cross-fertilize Office of Research and Advanced Technology scientific knowledge and activity, and to avoid worked out a compromise designed to salvage duplication of effort. The society president as many presentations as possible. After hec- argued that secrecy hurts the national posi- tic review and revisions, 28 of the papers were tion in science, technology, and industrial com- presented in “closed” sessions, at which at- petitiveness more than it strengthens national tendees were screened and required to sign the security. They notified the Secretary of De- agreement used to control distribution of ex- fense that they will not “be responsible for, port-controlled DoD technical data.15 nor will sponsor, closed or restricted-access sessions” in the future.17 SPIE officials emphasize that the compromise was worked out in a friendly spirit and in good DoD again, in April 1986, issued clarifica- faith by both sides, but the society insists that tion of its procedures in screening papers on the compromise was a “one-time necessity” unclassified DoD-sponsored research for pre- and not a precedent. SPIE now dissociates it- sentation at meetings. DoD says it will review self from classified, controlled-access meetings papers under specific time frames (10 to 30 or sessions that may be arranged by its mem- days) to help avoid last minute pressure for bers, but acknowledges that such meetings are withdrawals. However, the statement also held in parallel or immediately following soci- made clear that information deemed classified et y meetings and are regarded by many mem- could be presented only on DoD premises, and bers as desirable and necessary. unclassified information would still be subject to export control laws. Furthermore, the spon- Other professional societies admit uncom- fortably to similar situations. They officially sors of scientific meetings are responsible for limiting access to authorized individuals oppose and disclaim closed or limited access 18 sessions that do not serve all members (espe- (which societies say they cannot do). Some DoD officials, and some “neutral observers," cially when only non-U.S. citizens are excluded, say that these actions have effectively al- since most identify themselves as international leviated professional concerns; a number of societies); yet recognize that such meetings are professional society officials consulted by OTA organized by members in parallel with society meetings— ‘‘a bit of a fiddle, one society offi- report that this issue remains one of active and cial says. In 1984 the American Association strong concern to their societies and to most for the Advancement of Science compiled a list of their members. A nongovernmental science of 12 events in which professional societies policy specialist asserts that DoD, in practice has been far less strict and less restrictive than limited their traditional information dissemination function or activities to comply with their official policy guidelines and directives indicate they will be or should be. On the other security policies. For example, one session at a professional association meeting required hand, a scientist and society administrator ar- participants to bring to the session proof of gues forcefully that DoD policies have an citizenship. 16 However, on September 17, 1985, “extraordinarily chilling effect” on scientific a Joint Communication was sent to the Secre- communication because scientists, fearful of tary of Defense from the elected presidents of prejudicing the essential source of funding for 12 scientific and engineering societies, protest- future research, lean over backwards and prob- ing DoD actions. It stressed the value to the ably restrict themselves more than is abso- lutely required.

‘This account was developed from discussions with both SPIE and DoD officials in May 1987 and differs in some details from accounts in the general and specialist media at the time (1985) “American Association of Engineering Societies, Policy and and soon thereafter. En@”neering Priorities, 1986, Washington, DC, 1986. ‘8 Relyea, op. cit., 1986. “Ibid. 59 ——

informal—within the United States as well as The constitutional issue has not often across its borders, since publication in U.S. journals is tantamount to worldwide publica- been explicitly raised, possibly be- tion. As already discussed, however, their con- cause Congress has appeared to add stitutionality has not really been tested. The its authority to that of the executive constitutional issue has not often been ex- agencies. plicitly raised, possibly because Congress has appeared to add its authority to that of the executive agencies. Scientists instead have The export control statutes appear to pro- tended to try to minimize the opportunist y for vide a legal basis for restricting economic activ- government intervention through a strategy ities and scientific communication-formal and of “self -restraint.”

SELF-RESTRAINT

(NSF) .20 Having failed, he urged the American The strategy of self-censorship could Council on Education to form a public cryptography study committee, which then recom- be a significant limitation on dissemi- mended a voluntary system for prepublication nation of scientific thought and on the review by NSA of manuscripts on cryptog- exercise of constitutional freedoms. raphy. In 1981, NSF adopted a policy of requiring such prepublication review on “potentially classifiable results” coming from its The strategy of self-censorship poses an in- research grants. teresting ethical question. The objectives in In his AAAS speech, Inman included not self-censorship may be the exercise of reason, only cryptography but other areas in his argu- self-control, and patriotism, and the desire to ment for self-monitoring or self-censorship: avoid provoking authoritarian restrictions; but computer hardware and software, electronic the end result could nevertheless be significant gear and techniques, lasers, crop projections, limitation on dissemination of scientific thought and manufacturing processes. He warned bluntly and on the exercise of constitutional freedoms. that if this was not done voluntarily, “public In 1982, at the annual meeting of the Amer- outrage’ would produce laws further restrict- ican Association for the Advancement of Sci- ing publication of scientific work that govern- ence (AAAS), CIA Deputy Director Admiral ment considered sensitive. Inman repeated this Bobby Inman said in a public address: warning 3 months later at a congressional hearing held by two subcommittees of the House A potential balance between national secu- Science and Technology committee,21 and said rity and science may lie in an arrangement to include in the peer review process (prior to the —.— start of research and prior to publication) the ‘°Competing scientific, industry, and national security inter- question of potential harm to the nation. ’g ests in the development of encryption and related techniques and technologies for safeguarding data in computer and telecom- Earlier, as Director of the National Security munications systems has made this scientific technology par- Agency (NSA), Inman had tried to take con- ticularly troubled by the tensions between national securit y restrictions and other societal priorities, as described in this trol over government-funded cryptography re- chapter. search from the National Science Foundation *’U.S. Congress, House of Representatives, Committee on Sci- ence and Technology, Subcommittees on Science, Research, and Technology, and Investigation and Oversight, Impact of Na- tional Secun’ty Considerations on Science and Technology Hear- ‘gltelyea, op. cit., June 1986, p. 2. ings, 97th Cong., 2d sess., Mar. 29, 1982. 60 that unless Soviet access to American science, Protests were immediately raised about Ad- technology, and industrial information was miral Inman’s new call for self-restraint. The voluntarily controlled, there would be a move Executive Director of AAAS said: by the government to further regulation. He has asked that research scientists sub- Inman directly recognized that: mit voluntarily to open-ended censorship by the CIA or face the likelihood of being forced . Science and national security have a sym- to do so by Congress. Even in wartime, such biotic relationship. . . . In the long history of a demand would be an extreme one, and in the that relationship, the suggestion is hollow that absence of national security emergency it is science might (or should somehow) be kept incongruous. It raises troubling questions in- apart from national security concerns, or that volving both scientific freedom and the force national security concerns should not have an of constitutional provisions against arbitrary 22 impact on “scientific freedom." government. 23

221nman, a statement in Aviation Weeh % Space Technology, “ClassifyingS cience: A Government Proposal. . . And a Scien- tist’s Objection, ” Feb. 8, 1982, p. 10. “Ibid., statement by William Carey, Executive Officer, AAAS.

NATIONAL SECURITY DIRECTIVES AND THE ROLE OF THE NATIONAL SECURITY AGENCY National Security Directives NSDD 84, in 1983 (discussed above), marked “It is the policy of this Administra- the first general public knowledge of National tion that, to the maximum extent pos- Security Decision Directives (NSDD) as a form sible, the products of fundamental re- of Presidential directive distinct from Execu- search remain unrestricted.” tive Orders and Proclamations.24 Executive Orders and Proclamations are always published. About 200 NSDDs have been issued of the issue of government secrecy (National by President Reagan since 1981, but only 5 Security Study Directive 14-82). The review have been publicly disclosed, the rest being was itself classified, so that there was no in- classified. Between 1947 and 1981, other presi- put from universities or other nongovernment dents had issued “National Security Action sources. OSTP’s report was due on March 1, Memorandums” and “Presidential Directives” 1983, but did not appear. There were rumors which like NSDDs were kept secret, from both and stories in science policy newsletters that the public and from Congress. They are all the Administration was about to announce new thought to be associated with the National formal controls over “sensitive fundamental Security Council, created within the Office of research, ” i.e., a fourth category of classifica- the President in 1947. tion below that of “confidential.” But such a strategy would involve both monetary and po- NSDD 189 litical costs. Classified information requires In late 1982, after the recommendations of special procedures, controlled facilities, etc. the Corson panel, President Reagan had or- Contractual agreements between researchers and funders are less expensive, legally defen- dered his Office of Science and Technology Pol- 25 icy (OSTP) to coordinate an interagency review sible, and less likely to evoke protests. —— “Harold C. Relyea, Congressional Research Service special- “DoD was in fact considering a fourth security classification ist in American National Government, testifying before the at this time, according to sources in the Pentagon, to protect House Con-u-m“ttee on Government Operations on H.R. 145, Mar. ‘‘military operational data and high tech data—not fundamen- 17, 1987. CRS press release. tal research. ” (Communication to OTA May 1987.) 61

Instead of the long-delayed report from A memorandum written by Under Secretary OSTP, in May 1984, DoD’s Deputy Under Sec- of Defense Richard DeLauer on October 1, retary for Research and Advanced Technology, 1984, to reassure the universities, again speci- Dr. Edith Martin, unexpectedly released a fied that no restriction would be put on pub- draft Decision Directive that was to become lication of fundamental research sponsored by NSDD 189. Signed on September 21, 1985, DoD. It defined fundamental research to in- NSDD 189 said: clude virtually all of that done on university campuses, with rare exceptions “where there It is the policy of this Administration that, is a likelihood of disclosing performance char- to the maximum extent possible, the products of fundamental research remain unrestricted. acteristics of military systems, or of manufac- ... Where the national security requires con- turing technologies unique and critical to de- trol, the mechanism for control of information fense.” In these cases, restrictions must be put generated from federally-funded fundamental into the research contract. research , . . at colleges, universities, and laboratories is classification. . . . No restrictions The Role of the National may be placed on conduct or reporting of [such] research that has not received national secu- Security Agency rity classification, except as provided in appli- The Brooks Act of 1965 gave the National cable U.S. Statutes. Bureau of Standards authority for developing It was not clear what was encompassed in technical standards for computer systems. Pri- the term “fundamental research, ” a term not vate firms were developing an interest in this until then in common use. Since the “applica- market, and commercial security devices meet- ble U.S. Statutes” include the Export Admin- ing NBS standards were developed for com- istration and Arms Export Control Acts, puters. During the mid-1970s, a government- critics said that this directive did not materi- certified cryptographic algorithm and a‘ ‘pub- ally change the existing situation, except that lic-key” algorithm were announced (in the open either more Federal agencies would have power literature), and inexpensive security devices to classify, or DoD’s scope of authority would were developed commercially. broaden. More scientists would become subject Presidential Directive/National Security to NSDD 84 and be required to sign lifetime Council 24 (PD/NSC-24), issued by President nondisclosure contracts. National security offi- Carter in 1977, said that nongovernmental cials on the other hand deny this because “fun- telecommunications information “that would damental research” is excluded from export be useful to an adversary” would be identified, controls and there is no evidence that DoD is and the private sector informed of the prob- overusing classification procedures. lem and encouraged to take appropriate ac- NSDD 189 was interpreted by some as an tions.27 This was a clear sign that the Federal effort to “cool the campus secrecy issue” by Government, DoD in particular, would take a dropping the idea of further controls in “gray stronger hand in telecommunications security. areas’ (between classified and unclassified re- 26 The Secretary of Defense was to be respon- search). But in a memo accompanying NSDD, sible for protecting government communica- the White House stressed that it “preserves tions, both classified and now “unclassified but the ability of the agencies to control unclassi- sensitive” communications. The Secretary of fied information using legislated authority pro- Commerce would be responsible (through vided expressly for that purpose in applicable NTIA) for government-derived unclassified U.S. statutes. ” data not related to national security, and would

“Science and Government Report, “White House Decides To Cool Campus Secrecy Issue, “ vol. XIV, No. 11, June 15, 1984; “National Telecommunications Protection Policy (unclassi- Colin Norman, “Universities Prevail on Secrecy, ” Science, vol. fied), Feb. 9, 1979, unclassified excerpts from Presidential Direc- 226, No. 26, October, 1984, p. 418. tive/NSC-24, Nov. 16, 1977, classified. 62 -.

nology and for making decisions about technical standards, ignoring the role of NSDD 145 worried civil libertarians NBS under the Brooks Act; and because of the broad scope of nonclas- ● established an interagency group to im- sified information to be protected and plement and enforce NSDD 145 policy. because of the central role given to NSA, an agency outside of the usual This announcement worried civil libertarians because of the broad scope of nonclassified in- modes of accountability. formation to be protected and because of the central role given to NSA, an agency outside of the usual modes of accountability. From the deal with the private sector to “enhance their time it was issued, there have been conflict- communications protection and privacy. ” The ing interpretations and official pronounce- Defense and Commerce Departments attempted ments about the details of what it means. At to develop a joint proposal for a national pol- first, “sensitive information” was defined as icy on cryptography but were unable to reach “unclassified but sensitive national-security- agreement, They submitted separate propos- related information. ” In June 1985, Donald als which, however, were never acted on by the Latham, Assistant Secretary of Defense and 28 President’s Science Advisor. Chairman of the National Telecommunications and Information Systems Security Commit- NSDD 145 and HR 145 tee (NTISSC) established under NSDD 145, said that “sensitive” information might in- National Security Decision Directive 145 clude anything from crop forecasts to person- (NSDD 145), September 17, 1984, superseded nel records.29 Three months later, in testifying PD/NSC-24, and made NSA the central agency before the Committee on Government Opera- (“Executive Agent”) for development and tions, September 18, 1985, Mr. Latham said choice of cryptography-based technology for specifically that “non-national-security-related the security of unclassified, but sensitive in- information was not included in the purview formation in telecommunications and com- of NTISSC. But in October 1986, an NSA puter systems of all government agencies.* memorandum, NTISSP No. 2,30 extended this This seems to apply whether the information purview to: concerns national security or not, and regardless of whether the technology is to be used Other government interests . . . related but for security or for authenticating transactions. not limited to the wide range of government or government-derived economic, human, fi- In short, NSDD 145: nancial, industrial, agricultural, technological, and law enforcement information, as well as • broadened NSA protection to encompass the privacy or confidentiality of personal or unclassified information in telecommunica- commercial proprietary information provided tions and automated information systems; to the U.S. government by its citizens (Sec- • assigned to NSA the full responsibility for tion II, Definitions). developing and advising of safeguard tech- ..—.— The NSA policy announcement also said: l~For ~ more det~led treatment of the Government’s role in computer and communications security, see U.S. Congress, Of- The NSDD-145 Systems Security Steering fice of Technology Assessment, Defending Secrets, Sharing Group has established that sensitive, but un- Data: New Locks and Keys for Electronic Information, OTA- CIT-31O (Washington, DC: U.S. Government Printing Office, October 1987). “U. S. House of Representatives, Committee on Science and

*The Nation~ Security Agency was established by Execu- Technology, Subcommittee on Transportation, Aviation, and tive Order in 1952, during the “Cold War, ” to carry out “sig- Materials, Hearings on Computer Security Policies, June 27, nals intelligence’ against hostile countries (i.e., to intercept and 1985, statement of Donald Latham. decode information they were sending through electronic com- ‘“National Policy on Protection of Sensitive but Unclassified munications) and to safeguard U.S. classified information Information in Federal Government Telecommunications and against similar actions by those countries. Automated Information Systems, NTISSP No. 2, Oct. 29, 1986, 63

classified information that could adversely affect national security or other Federal Govern- ment interests shall have system protection There is a strong concern that DoD and safeguards; however the determination of may require private sector database what is sensitive but unclassified information operators, at a minimum, to provide is a responsibility of Agency heads. the government with lists of their sub- It now appears that the definition of "sensi- scribers, to place limits on foreign sub- tive" could be applied to almost any informa- scribers, and to increase the use of tion, or at least a very broad range of informa- password protection and encryption. tion, even if it is already published or available. NSDD 145 applies not only to Federal agencies but also to their contractors who electron- The report went onto say that “One solution ically transfer, store, process, or communicate appears to be to thoroughly screen all candidate sensitive but unclassified information. It gives database entries and keep sensitive govern- NSA the dominant role in developing the tech- ment information out of the public databases nology to be used, or deciding which technol- . . . “ but added, “Unfortunately, this may also ogy will be used, by private sector organiza- inhibit the United States’ own national re- tions affected by the directives. NSA and DoD, search effort by resisting the ready availabil- however, emphasize that “protection’ in this ity of such information. ” context literally means guarding against un- It was clear that NSDD 145 had already authorized access and malicious misuse by ‘hackers. ” given NSA the decisive role in prescribing security measures for government’s automated This new role for NSA would have gone be- databases, as well as telecommunication sys- yond development of technology to involve- tems, but its authority over commercial data- ment in decisions about the content of infor- bases is still in dispute. Diane Fontaine, direc- nation, in order to prescribe what kind of tor of information systems in the office of the security is appropriate. In late 1985 there had Assistant Secretary of Defense, reportedly said been signs of renewed government concern at a meeting of the Information Industry Asso- about the “leakage’ of information from cam- ciation, November 11, 1986: “The question is puses and through commercial databases. The not should there be limits [on information in Pentagon released a CIA report, Soviet Acqui- commercial databases] but instead what those sition of Militarily Significant Western Tech- limits should be.”32 Ms. Fontaine has since nology: An Update, based on Soviet documents. stated that she was misquoted and was refer- It listed 62 American universities “targeted ring only to government databases, but many for scientific and technical espionage” aimed auditors understood the reference as being to at information about applied technology and commercial databases. There is a strong con- engineering but also including “fundamental cern that DoD may require private sector data- research for both Soviet military- and civilian- base operators, at a minimum, to provide the related science developments. ” This scientific government with lists of their subscribers, to and technological data is often in commercial place limits on foreign subscribers, and to in- electronic databases providing services to business and the public: of a declassified Central Intelligence Committee report origi- The individual abstracts or references in nally released in 1982 by the Senate permanent subcommittee government and commercial data bases are un- on Investigations. Dan Morgan, “Stolen U.S. Technology Boosts Soviet Strength, Report Says, ” Washington Post, Nov. 15, 1982. classified, but some of the information, taken “Quoted by Johanna Ambrosio, “Attempts To Restrict Pri- in the aggregate, may reveal sensitive infor- vate Data Bases Vex Industry, Government Computer News, mation concerning U.S. strategic capabilities Dec. 15, 1986. Another reporter’s version was “The question and vulnerabilities. . . .31 is not whether we are going to protect information; the question is where the controls will be applied. ” Irwin Goodwin, “Mak- 1‘Science and Got’ernmen t Report, Oct. 1, 1985. The report ing Waves: Poindexter Sails Into Scientific Data Bases, Ph&Ysics eleased by Secretary of Defense Weinberger was an “update” Today, January 1987, p. 52. 64

crease the use of password protection and en- Technology Utilization is a service to dissem- cryption. inate technical information to the public. Its information is not classified. In 1986 there were 3,200 electronic databases available worldwide through 486 online As major opposition to the thrust of NSDD information services; 70 percent of these data- 145 developed within Congress, a new National bases are produced in the United States, and Security Advisor, Frank Carlucci, who had suc- all but two of the 20 largest database compa- ceeded Mr. Poindexter, decided to review nies are American corporations.33 Throughout NSDD 145. His “key objective [was] finding much of 1986, a U.S. Air Force team visited a mechanism to eliminate any possible ambi- commercial database owners to inquire about guity regarding the role of the National Secu- the extent of foreign access to these data- rity Advisor in connection with the System bases. 34 Security Steering Group, ” and Carlucci instructed his staff “to initiate the procedure Nonclassified government databases, even and prepare the papers necessary to rescind those specifically set up to provide better pub- 36 NTISSP 2. lic access to information that is in the public domain, are in some cases already restricting H.R. 145 access. An internal NASA memo of Septem- ber 29, 1986, labeled “The So-called ‘No-No’ NSA had expanded its role into areas legis- list,”35 provides names of 33 organizations or latively assigned to the National Bureau o individuals who are not to be “provided with Standards (NBS). H.R. 145, the Computer Secu- subscriptions to NASA Tech Briefs, technical rity Act of 1987, was accordingly introduced support packages, or other Technology Utili- It would assign to NBS the responsibility for zation documentation. ” The memo adds that developing and promulgating standards and “all embassies and consulates in the U.S. and guidelines for safeguarding unclassified sen- representatives of foreign companies or orga- sitive information in Federal systems, and for nizations are to be included in this list. ” NASA helping both civilian agencies and the private sector in using computer security safeguards.37 ‘gInformation supplied by Information Industry Association, This bill was passed by the House in July 1987 See Potential Government Restrictions on Um”ted States Un- but had not been taken up by the Senate a classified Commercial Databases, Mar. 9, 1987, available from the association. of early January 1988. 9iIbid., s= ~So: Michael SChrage, “U.S. f%king TO Limit Ac- cess of Soviets to Computer Data, ” Washington Post, May 27, Wetter from Frank C. Carlucci, National Security Advisol 1986; and “Are Data Bases a Threat to National Security, ” Busi- The White House, to the Hon. Jack Brooks, Chairman, Corr ness Week, Dec. 1, 1986, p. 39. But DoD officials describe this mittee on Government Operations, U.S. House of Represent as “data gathering with no purpose related to immediate ac- tives, Mar. 12, 1987. tions. ” (In communications to OTA, May 1987.) “Hearings on HR 145 were held on Feb. 26 by the Subcom 9SSi@~ b Walter M. Hefl~d, Manager, Technology Utili- y mittee on Legislation and National Security, Committee on GOT zation Office, and addressed to “All TU Officers, IAC Direc- ernment Operations, U.S. House of Representatives. The Whit, tors and Other Members of the TU Family, ” and reprinted by House accepted this measure and the bill was passed by th Translational Data and Cornmum”cations Report, February House in July 1987, and is still waiting to be introduced in th 1987, p. 21. Senate as of early January 1988. Chapter 6 Constitutional Issues: An Overview

The U.S. Constitution says little about what s a major force in our society: the development Only a few civil libertarians would dis- and use of science and technology. Yet in spite of its silence about science, the constitutional allow any restriction on the dissemi- framework has proved remarkably hospitable nation of scientific information, but to the flowering of scientific research in this others believe that the burden created century. There has been a strong mutual de- by existing controls is excessive. pendence and respect between government and scientific enterprise that is only rarely threatened. But challenges to the constitutional rity restrictions on scientific communication status of science have occurred in the past, and apply not only to documents, reports, and pub- will surely occur in the future. lications, but to informal modes of communication, such as talks at professional meetings The right to free dissemination or commu- and participation in scientific seminars. Al- nication of scientific research results or other though primarily applied to information in scientific and technical information is not un- which government has a “property-type’ in- limited. It is and has always been limited in terest—which results from research carried out the interests of national security. by, or contracted out by, government–national security restrictions may also apply in some In spite of its silence about science, cases to research done under a government the constitutional framework has grant for support of basic research, or even, at least in the case of nuclear science and engi- proved remarkably hospitable to the neering, to information generated entirely flowering of scientific research in this without government funding or involvement. century. Almost all of the people and groups that have spoken out on this issue acknowledge that free- Existing restrictions on the flow of scientific dom of speech and press are not absolute and information may be administrative, statutory, must be balanced against the competing in- contractual, or voluntary. The nature of infor- terest of national security. They disagree on mation deemed to be sensitive from the stand- where that balance should be struck. point of national security goes far beyond that Some people take the middle position that related to weapons and includes almost any while the current Administration defends a kind of scientific or technological data that strong position on national security protec- may bear on our industrial capability or com- tions and particularly emphasizes the need for 1 petitive position in world trade. National secu- strong technology export controls and a dom- ‘Section 1.3 of Executive Order 12356 says that information inant role for the military in defining those con- lay be classified “if it concerns military plans, weapons, or oper- trols, nevertheless there has been no obvious tions; the vulnerabilities or capabilities of systems, installa- tions, projects, or plans relating to the national security; for- abuse of authority. The Department of Defense ign government information; intelligence activities (including has not pushed dramatically to the theoreti- pecial activities), or intelligence sources or methods; foreign cal limits of its power to restrict communi- elations or foreign activities of the United States; scientific, echno]ogical, or economic matters relating to the national secu- cations. ity; U.S. Government programs for safeguarding nuclear ma- erials or facilities; cryptology; a confidential source; or other Others argue that even stronger controls ategories of information that are related to the national secu- over information are needed to limit the ‘ ‘un- ity . . . .” It also says that such information “shall be classi- wanted transfer’ of American science and tech- ied . . [when] its unauthorized disclosure, either by itself or 1 the context of other information, reasonably could be expected nology to countries that are very ready to use o cause damage to the national security. ” (Emphasis added. ) it against us. The courts have long recognized

65 66 the need of private parties for protection of rather loosely defined, and economic or indus- their trade secrets. Some urge that a compara- trial objectives, acts to confuse and prejudice ble right of protection be granted the govern- the justification of government actions to limit ment to protect sensitive knowledge generated constitutional rights. by government-sponsored research. Thus one Among those who are concerned with pre- can defend a proposition that scientific inter- venting the “chilling’ effects of excessive re- change should, in general, be encouraged; but strictions on freedom of speech and press, there it is inherently limited by the proprietary na- is a further difference of opinion. Some argue ture of scientific knowledge and therefore, un- that science was intended by the Founding like political speech, is not protected by the Fathers to enjoy a special position and spe- First Amendment. cial protection under the First Amendment; Critics of this position point out that the in- and that the courts must recognize and im- formation in which government has a property- plicitly have recognized that special position. type interest is exactly the information that Other scholars deny this special protection and should be freely communicated; because gov- emphasize that the courts have always applied ernment authority to support scientific experi- to First Amendment challenges on behalf of mentation flows from its power to spend in or- scientific communication a “balancing of in- der to promote general welfare, and funding terests.” that produces that property-interest is generated through taxation of the public. Some argue that science was intended Only a few civil libertarians would disallow by the Founding Fathers to enjoy a any restriction on the dissemination of scien- special position and special protection tific information in the interests of national under the First Amendment. security, but others believe that the burden created by existing controls of several kinds, Professor Steven Goldberg of The George- taken together, is excessive. It has become, they argue, counterproductive in its effects on town University Law Center is among those science and industry, and more importantly, who argue that science enjoys, under the Con- an erosion of constitutional rights and liberties. stitution, possibly more protection than even political or literary speech. He argues that There have been only a few direct challenges those who participated in drafting the U.S. to the constitutionality of these limitations on Constitution, particularly Jefferson, Madison, the cherished freedom of speech and press. The Hamilton, and Franklin, were men of the En- President as Commander in Chief is considered lightenment, with broad interests in science, to have full authority to order classification who regarded scientific freedom from con- of government-generated information. Although straint by church and state as essential to the constitutionality of some aspects of the democracy and constitutionalism.2 Goldberg Atomic Energy Act and Invention Security argues that the Constitution contains an ‘im- Act might be questioned-especially their prior plied science clause”: that Congress may leg- restraint provisions and their applicability to islate the establishment of science but not pro- privately generated information—only a few hibit the free exercise of scientific speech. particularly weak challenges have been brought. ‘Steven Goldberg, “The Constitutional Status of American Ironically, or so it may seem to the layman, Science, ” University of Illinois Law Forum, No. 1, 1979, pp. restrictions on speech and press under export 1-6 ff. Unless otherwise attributed, discussion of Goldberg’s controls may be less subject to challenge, views throughout this section is based on this article. Other scholars share the view that scientific communication enjoys rather than more so, because their infringe- an especially protected position under the Constitution. See for ment on freedom of speech and press is con- example, John A. Robertson, “The Scientist’s Right to Research: sidered incidental to another legitimate pur- A Constitutional Analysis, ” 51 Southern CA”fornia Law Re- view 1203 (1977); and Delgado & Millen, “God, Galileo, and Gov- pose. Yet some critics argue that in this area, ernment, Toward Constitutional Protection for Scientific In- the overlapping of national security objectives, quiry, ” 53 Washington Law Review 349 (1978). 67

Under the provisions of Article 1, Section Goldberg maintains that while “The Courts 8, which authorizes Congress to spend money have had little occasion to define precisely the for the general welfare and therefore for scien- limits of the government’s power to classify tific research, Goldberg says, “Science is estab- scientific material, ” a reading of those deci- lished in the sense that religion cannot be sions that exist and of the opinion of constitu- established. ” The First Amendment was in- tional scholars shows that science “is fully pro- tended to prevent “the suppression of enlight- tected by the speech and press clauses. ” He ened science by the Church.”3 The “free exer- says: cise” and establishment clauses are, Goldberg As leading first amendment scholars have says, complementary. Science enjoys a pro- long recognized, suppression of scientific in- tected status. formation is inconsistent with the democratic political process . . . . Even when scientific work is not immediately applicable to politi- “A meaningful questioning of the cal controversies, it plays an important role value to the national security of re- in maintaining a free and informed society. strictions on scientific expression will, Such was the view of the framers, and it has however, come about only if the ques- been the consistent view of the courts (p. 16). tion is pressed by a vigilant scientific Other scholars, such as Professor Harold community that will consider and act Green of the National Law Center, discount on these issues creatively and con- this thesis. Green says, “There have been Su- structively.” preme Court and lower court decisions that have involved or have referred to science in pro- tective terms, but the involvement of science He also argues that modern constitutional in these cases has usually been collateral to decisions support this thesis. In Roth v. United some other issue. ” For example, in the Ar- States,’ concluding that obscenity is outside kansas case concerned with the teaching of evo- of First Amendment protection, the Court said lution, Green maintains that it was the reli- that discussion of “sex in scientific works is gious purpose of the statute rather than the not itself sufficient reason to deny material the restriction on teaching science that was held constitutional protection of freedom and to violate the First Amendment. press. ” In Miller v. California5 the Court said that the “First Amendment protects works which, taken as a whole, have serious literary, Congress, therefore, has an essential artistic, political, or scientific value. ” When role to play in preserving the balance the Court invalidated an Arkansas statute through the formulation of public pol- barring the teaching of evolution in public icy and the oversight of executive schools,6 the Court’s scrutiny of the statute agencies. (according to Goldberg) “was more intense than in the usual establishment case because the competing value at stake was science.’” Green, like Goldberg, points out that there have been few challenges to the constitution- ‘Goldberg points (p. 5) to Madison’s “Memorial and Remon- ality of government restrictions on scientific strance Against Religious Assessments’ and to comments on Galileo by Jefferson, saying, “The Jeffersonian wall between communications. The critical question, he says: church and state was designed in part to protect American . . . is the degree of protection that will be af- Galileos. ” forded against government encroachments in 4354 U.S 476 (1957). “413 Us. 15 ( 1973). the name of national security, [and] the answer ‘Epperson v. Arkansas, 393 U.S. 97 (1968). to this question depends on a case by case ‘The Court said, “The State’s undoubted right to prescribe balancing of the respective interests. the curriculum for its public schools does not carry with it the A meaningful questioning of the value to the right to prohibit . . . the teaching of a scientific theory or doctrine where that prohibition is based upon reasons that \,iolate national security of restrictions on scientific the First Amendment. Ibid. at pp. 1 I 2-113. expression will, however, come about only if 68

the question is pressed by a vigilant scientific congressional action, the executive branch community that will consider and act on these must act. Its daily pressures to assure national issues creatively and constructively. It is not security may make it less careful in preserv- enough merely to proclaim the shibboleth that ing the essential balance. As Justice Marshall science is a sacred preserve entitled, by its very noted in the Pentagon Papers case, unilateral nature, to special constitutional protection. executive efforts to restrain communications It may not be adequate for citizens or Congress without following careful criteria established to rely on scientists to provide the watchful by Congress may run afoul of the constitu- eye and determined protests that will maintain tional rule that the executive is not free to leg- a healthy balancing of interests in safeguard- islate.8 ing First Amendment rights. Scientists, after Gerard Piel, former President of the Amer- all, are often wrapped up in their immediate ican Association for the Advancement of Sci- scientific pursuits and reluctant to involve ence, argues that there is a social contract themselves in policymaking-at least until embedded in the Constitution through which their own activities are threatened. Moreover, our society attempts to foster science. Piel those scientific pursuits may be entirely de- points out that Thomas Jefferson, in justify- pendent on government funding and access to ing the First Amendment, argues that “to pre- government scientific data. serve the freedom of the human mind . . . and Congress, therefore, has an essential role to freedom of the press” is a cause worthy of mar- 9 play in preserving the balance through the for- tyrdom. mulation of public policy and the oversight of executive agencies. The courts nearly always defer to Congress when it presents a consist- %eNew York ~“mes Co. v. Uru”t-ed States, 403 U.S. 713, 74( (Marshall, J., concurring). ent, thoughtful position on the constitutional ‘Gerard Piel, ‘‘Natural Philosophy in the Constitution, Sci rights of citizens. Moreover, in the absence of ence, vol. 233, No. 5, September 1986, p. 1056. Appendix Appendix A Acknowledgments

Science, Technology, and the Bob Smith Constitution in the Information Age: Executive Director Videotext Industry Association The Future of the Press and the First Amendment Workshop–Mar. 26, 1987 David Thurm Counsel New York Times Vivian Arterbery Executive Director Jane Wilson U.S. National Commission on Library and Assistant Director Information Science Telecommunications Affairs American Newspaper Publishers Association Joseph Bremner Database Development, Inc. Lisa Carlson Director Scientific Communications, National Networking Services Security, and the First Amendment Metasystems Design Groups, Inc. Workshop–Mar. 11, 1987 Michael F. Cavanagh Executive Director Allan Adler Electronic Mail Association American Civil Liberties Union Mike Greenly Kenneth Allen Mike Greenly Consultants Information Industry Association Reed Irvine Jerry J. Berman Accuracy in Media, Inc. Director, Project on Privacy and Technology Richard P. Kleeman American Civil Liberties Union Director Daniel DeSimone The First Amendment Center Executive Director Patrick Maines American Association of Engineering Societies President Diane Fountaine The Media Institute Director, Information Systems Phil Moore U.S. Department of Defense President Harlow Freitag Networking & World Information Chairman, Technology Transfer Committee Richard Neustadt Institute of Electrical and Electronics Private Satellite Network, Inc. Engineers Kendrick Noble, Jr. Stephen Gould Payne Webber, Inc. Committee on Scientific Freedom and Responsibility David Peyton Information Industry Association American Association for the Advancement of Science Monroe Price Harold P. Green Benjamin Cardozo Law School National Law Center Judith Roales George Washington University Vice President Richard Kleeman Independent Newspapers, Inc. Association of American Publishers Richard Schmidt, Jr. Jonathan Knight Cohn & Marks American Association of University Professors 71 Harold C. Relyea Part 11 Specialist in American National Government Ronald Calinger Congressional Research Service The Catholic University Frank Sobieszczyk Steven Goldberg Chief, Research Programs Office (DTAG) Georgetown University Law Center U.S. Department of Defense Stephen Gould Paul Stephan American Association for the Advancement of University of Virginia Science School of Law Daniel O. Hill Tom Suttle U.S. Department of Commerce Institute of Electrical and Electronic Engineers Richard Kleeman Mitchell B. Wallerstein Sigma Delta Chi Foundation National Academy of Sciences Jonathan Knight Association of University Professors Additional Contributors and Reviewers Frank Sobieszczyk Part I U.S. Department of Defense Mitchell B. Wallerstein Lee Burdick National Academy of Sciences The Media Institute David Whitman Henry Geller U.S. Department of Defense Telecommunications Research and Action Committee Within OTA: David Littlejohn Richard Dalbello University of California, Berkeley International Security and Commerce Program David Peyton Lisa Heinz Information Industry Association Science, Education and Transportation David Thurm Program New York Times Nancy Naismith David Walsh Science, Education and Transportation Electronic Information Services Development Program Peter Sharfman Within OTA: International Security and Commerce Program Richard Dalbello Chuck Wilk International Security and Commerce Program Communication and Information Technologies David Margulius Program Communications and Information Technologies Program Mark Nadel Communications and Information Technologies Program

72 U.S. GOVERNMENT PRINTING OFFICE: 1987 - 80-465