Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
PRINCIPLES AND PRACTICE 148 4 Principles and Practice Discussions of a national information infrastructure (NII) and discussions of research and education networking tend to embrace several tenets or principles for the design and implementation of information infrastructure.1 There are different classes of principles: some relate to the availability of information infrastructure (e.g., ubiquitous and equitable access); some relate to the terms and conditions of use (e.g., protections for privacy and security). In some instances, desired attributes of the structure of the infrastructure are also presented as principles. In this report, those technical issues, including security protections, are discussed separately as part of a comprehensive consideration of open information networking (Chapter 2). In the past year, the administration, industry, and professional and other advocacy groups have presented sets of principles intended to guide the development of the NII (see Appendix B for samples of such principle sets). These principles reflect value judgments and suggest the directions in which many individuals and groups will try to direct the development of architecture and, in particular, the deployment (including investment, pricing, and financing) of the NII. On the surface there is significant commonality among the lists that have begun to circulate, reflecting the universal quality of several principles, the prominence of public interest advocates in their formulation, and the fact that it is relatively easy to achieve agreement on high-level statements. For example, the presence of privacy, security, and intellectual property protection across various lists of principles shows that these issues not only reflect shared mores and norms, but also are seen as broad enablers of information infrastructure. At the same time, there also are
PRINCIPLES AND PRACTICE 149 differences among these lists in terms of emphasis as well as content. Further, the meaning or interpretation of these principles may differ for different groups. The research, education, and library communities have valuable experience with the implementation, or lack of it, of various principles in the Internet context. Given the nature of their missions and reliance on financial support, their perspectives and concerns will differ from those of commercial or consumer users of information infrastructure (and there will be differences between segments of each of these communities). In addition, communities may differ regarding their preferences for when and/or how a principle may be translated into action, especially where trade-offs must be made. To provide some flavor of research, education, and library perspectives on these dimensions, this chapter provides a brief overview of some specific principles. The issues raised in this discussion can be addressed by technical, market, and/or policy mechanisms (respectively addressed in Chapters 2, 5, and 6). Although the discussion below highlights research, education, and library perspectives, resolving key issues will require finding common ground between those perspectives and others relating to industry and/or the general public. The discussions launched by the administration's Information Industry Task Force (IITF); by various interest, professional, and trade organizations; through various electronic discussion groups; and by congressional hearings are all part of the process of weighing and balancing different perspectives.2 Achieving consensus on fundamental values is an essential element of the process of developing the policy framework for an NII. EQUITABLE ACCESS Access to networks and services should be equitable, affordable, and ubiquitous. This principle, expressed in various ways, is perhaps the most universally expressed but also the most elusive in implementation. Success may be difficult to gauge, given the lack of precision or consensus on what constitutes equity, affordability, or ubiquity.3 The equitable access principle, its components, and its variants refer to an electronic parallel to telephony's universal service expectation.4 Some people, for example, are located in areas that are more expensive to serve with any kind of wireline communications. This includes sparsely settled or remote areas, mountainous regions, and other hard-to-reach places. Basic telephone service costs more to provide in these locations than in the cities and suburbs. As a matter of policy, basic telephone service is made available at about the same price regardless of the conve
PRINCIPLES AND PRACTICE 150 nience of location, and the committee believes that the same should be true for the NII. The experiences in research, education, and library uses of the Internet demonstrate that each community brings new resources and will influence the future direction of the network. Consequently, efforts and strategies that seek to separate out the communities they serve, especially those that are perceived as most profitable, will serve the public interest less than more inclusive strategies for involvement in information infrastructure. Moreover, serving the public interest through broad access should eventually serve private interests, by expanding markets and values associated with the infrastructure.5 This is a key reason for avoiding a situation in which the Internet continues to exist but diverges from the rest of the information infrastructure. At the most basic level, access to information implies access to infrastructureâaccess to a network (including associated software) and interconnection of that network to other networks and to resources that can deliver desired information. The nature and degree of access will depend on technical factors (e.g., standards), economic factors (e.g., charges, subsidies, incentives, and so on), facilities, human resource requirements (e.g., training and expertise), motivation, choice of applications and system platforms, and policies (which will influence all of the above). Cost, always a pacing factor in broadening access to new technology, is considered the key barrier to broader access today. At the next level, access to information raises questions about the mix and range of information services available over the network. Today's research and education communities rely on networks primarily for messaging (electronic mail), accessing information resources and services (e.g., databases), and file- transfer services. In this environment, the users of the infrastructure are the primary producers of the information transferred over it. Tomorrow it is clear that the mix of services will change in two ways. First, the format will change; there will be more video, audio, multimedia, and real-time services, in contrast to the flat text and binary data files that dominate Internet traffic today. Second, there will be more third-party providers of information, which will be offered on a free or at-cost basis by government and on a revenue-generating/profit- making basis by commercial providers.6 Third-party sources of information (databases, numerical, full-text, bibliographic, weather, news, and other kinds of information) have been in use for many years in libraries, originally on a dial- up basis but most recently via the Internet, and demand has been growing. Of course, they have also been used by various businesses, often on a dial-up basis. For the research and education communities in particular, an increasing emphasis on network-based access to information raises questions
PRINCIPLES AND PRACTICE 151 about whether there is a need for a minimum level of accessâand what that level is, how basic it is in terms of the quality of service (speed, support for video or multimedia, and so on), the locus of access (on each desk or in each classroom or in such communal and institutional facilities as libraries), and the breadth of reach (within a state or region, national and/or international, connections to which information provider gateways, and so on). Because some types of research and education are inherently more demanding than others in terms of their bandwidth requirements, equal access to relatively low bandwidth service (e.g., for textual electronic mail) alone would not meet the needs of many users. Although it may be possible, drawing on grant-award and other records, to measure the number of researchers currently funded to undertake high-performance computation and communication, it is much harder to assess how many would do such work were the infrastructure more affordable to them. Similarly, the low-bandwidth/ low-budget applications typical of the K-12 experience suggest that the primary and secondary schools have needs that are typically quite modest. However, presented with better infrastructure options, the K-12 demand might well look differentâand if given adequate local access capabilities, it could grow dramatically. In the 1988 report Toward a National Research Network, the Computer Science and Telecommunications Board (CSTB) made it clear that there is a basic need for communication that is shared by all and should be available at a price that all can afford:7 To encourage the development and use of a widely accessible network service, the committee suggests that the NRN [National Research Network] might provide (1) a universal basic service at some low to moderate speed for electronic mail at a low cost to users and (2) higher levels of service with an appropriate charging structure. A low charge for basic service would lessen the initial resistance that might otherwise be felt by first-time users and yet discourage the overloading that occurs with free networks. (p. 2) Although the above statement was formulated with an emphasis on the research community, the essential principles hold more broadly. The present committee endorses the concept of a minimum set of services, as discussed in Chapter 2, that is available to all. Differences in outlook on equity and ubiquity derive from differences in financial status and ability to payâhence, the many concerns voiced in the research, education, and library communities about affordability. How network and information access are priced will affect the ability of new users and communities to engage with the information infrastructure.
PRINCIPLES AND PRACTICE 152 Linking access and service to ability to pay would be inherently exclusionary. This problem lies behind consumer group and state regulatory efforts to moderate the pace of infrastructure advances and tariff changes. On the other hand, industry efforts to contribute to educational access, in particular, show both awareness in industry of the political appeal of enhancing education and the fact that there are unpredictable alternatives to public financing alone. Although the popular debate over the NII has acknowledged a risk of polarization into information "haves" and "have nots," a more varied spectrum involving "have mores" and "have lesses'' may be more realistic. How ubiquitous is the access will determine, in part, the nature of the benefits: people with desktop or other immediate forms of access will be able to engage in more spontaneous, intermittent, and possibly more interactive uses of the infrastructure than people who must go to a communal point of access and wait their turn. For example, one access point per school represents a far different kind of "ubiquity" than one per teacher or student. Access points for students, in turn, beg the issue of access for parents or guardians, which can increase the impetus for home-based access.8 As the issue of home-based access underscores, ubiquity assumes that there is a way of scaling services. To date, there has been a trade-off between effective speed of service and extent of service reachâslowerspeed links on the Internet already have become associated with service to less affluent users. While performance on the Internet could be enhanced if access were limited to rates of 56 kbps or above, this might disenfranchise groups (such as schools) that can barely afford to be connected. From a practical perspective, the key variable is timing. Current administration and congressional activities aim at stimulating or accelerating the achievement of universal access to likely NII services. At the present time there is no compelling argument for a government program to accomplish universal network access to all individuals immediately. Costs would be prohibitive, and the absence of a universal culture of use means that benefits would be swamped by costs. Rather, what is required is assurance that those individuals who desire access can obtain it in a cost-effective and straightforward manner.9 The 1988 CSTB report pointed out that the equity issue could be divisive:10 If access is granted on the basis of ability to pay, the network may foster a "rich-get-richer" outcome. â¦ [A] situation where large or "rich" users provide their own individual solutions to networking would prevent an NR[E]N from achieving the benefits of scale or access promised by the concept of a truly national network. (p. 24)
PRINCIPLES AND PRACTICE 153 This committee concurs with that assessment. Today, with the move to commercial service, there is the prospect of even greater differentiation than there was in 1988. The committee is pleased to see that the equity principle is captured in the structure and activities of the IITF; the articulation of administration goals for connecting schools, hospitals, libraries, and clinics; and related programs at the National Science Foundation (NSF) and the National Telecommunications and Information Administration (NTIA), including the NTIA Telecommunications and Information Infrastructure Assistance Program that was launched in 1994. FLOW OF INFORMATION Access to information through the infrastructure should be governed by rights and responsibilities relating to freedom of expression, intellectual property rights, individual privacy, and data and system security. Commercial providers of networks and information services and the research, education, and library communities differ in their outlook on the flow of information. While commercial providers focus on the business potential of informationâincluding associated revenue streamsâthe research, education, and library communities emphasize free sharing of information in the course of their activities, and there is a concomitant emphasis on broad sharing of information.11 Creation, discovery, and analysis of information are fundamental aspects of research and learning; organization, storage, location, and provision of information are fundamental purposes of libraries; and discovery, communication, analysis, and creation of information are fundamental aspects of education. In the research community in particular, much more information is available and used than is formally published. Rapid communication of data and results over a network are becoming the norm, confounding conventional publishing cycles.12 The research, education, and library communities have pioneered in the development of electronic library card catalogs, digital library collections, and electronic publications aimed at these communities. For decades, publishers and commercial information services providers also have been exploring ways to create, produce, and deliver information over networks; their activities have included the development of abstracting services, provision of rapid access to news relevant to businesses, and value-added delivery of legal and financial information aimed at paying clienteles from industry as well as academic markets.13 More recently, they have perceived a widening popular market.14 Additional commercial markets relating to network- based delivery of information resources to the educational community are also being explored.15
PRINCIPLES AND PRACTICE 154 Because of the need to recoup their investments and generate additional capital for service improvements and innovations, controls on access and use are very important to these businesses, as is the ability to charge for access and use. Funding for the digitization and electronic supply of noncommercial materials is an ongoing concern facing both publishers and consumers of those materials in the research and education communities.16 Government Information The treatment of government-generated information offered through commercial services illustrates various tensionsâthere has been an ongoing debate in the library, publishing, and consumer advocacy communities about how and by whom information collected or produced by the government should be made available. Some of this debate occurs in conjunction with a broader movement to realize electronic libraries.17 The research and education communities may have special claims to easy, affordable access to government information, a large amount of which consists of data of interest toâand/or generated byâvarious kinds of researchers and some of which, in turn, has proven, through activities associated with the National Research and Education Network (NREN) and High Performance Computing and Communications (HPCC) programs, to have a variety of educational applications (Box 4.1). Such companies as Mead Data Central and WestLaw have made a profitable business of packaging and enhancing legal and other information originating in the government for resale. They have maintained that adding value to information should be done in the private sector, rather than by government, because of the opportunities for promoting competition, the concern that government should not be in the value-adding business (e.g., data interpretation, analysis, and other enhancements), and the concern that government control of the information it generates could result in a monopoly over information made available to the public.18 Note that government supply does not mean free information; some states, for example, are considering charging for data they supply electronically, in the process assessing the balance between straight cost-recovery and generation of additional revenues and the balance between serving the general public and serving commercial customers.19 Public libraries and public interest groups, with the agreement of the information industry, have argued that basic government-generated information should be made available directly, without adding value, to the user, either for free or at the cost of documentation.20 Given that the government has already incurred the costs of creating and processing the information for governmental purposes, the economic benefit to society
PRINCIPLES AND PRACTICE 155 BOX 4.1 GOVERNMENT INFORMATION Government programs can stimulate the development of electronic databases of government information that support research and education, and they can support the development of sophisticated data management and searching software that will make the software accessible and usable by the research and education communities.* Much of the research data available is federally owned and controlled, making the data ideal material for experimental, Internet projects. Furthermore, since the federal data are largely in the public domain, intellectual property problems will not interfere with or confound the research results. Finally, with the increasing use of and reliance on information technologies by federal agencies, it will be important to provide an additional dissemination channel for effective, equitable, and timely access to federal information resources. Such channels are contemplated by: â¢ The Clinton-Gore technology initiative, "Technology for America's Economic Growth, a New Direction to Build Economic Strength," which states, "Government information is a public asset. The government will promote the timely and equitable access to government information via a diverse array of sources, both public and private, including state and local government and libraries." â¢ Recent proposed legislation seeking to stimulate and spur network- based applications of libraries, education, health care, and government information. â¢ Revised OMB Circular A-130, "The Management of Federal Information Resources," providing guidance to agencies in their collection, maintenance, and dissemination activities and articulating a clear and strong statement in support of public access to government information. â¢ The Government Printing Office Electronic Information Access Enhancement Act, S. 564 (P.L. 103-40), which moves the Government Printing Office forward into the delivery of on-line federal data resources, including the Federal Register and the Congressional Record. It calls for the development of an electronic directory of federal public information. These and other on-line data files will be free to depository libraries and available at the incremental cost of dissemination to other users. * This objection is derived in part from discussions among task force members of a joint initiative in the area of intellectual property rights by the Association of American Universities and the Association of Research Libraries.
PRINCIPLES AND PRACTICE 156 is maximized when government information is publicly disseminated at the cost of dissemination to both end users and value-added redistributors.21 Under the administration's National Performance Review, the NII initiative, and legislative proposals, a number of steps to enhance the flow of government information over the Internet are already being explored. These efforts should be monitored and evaluated in terms of their costs and impact.22 Internet delivery of government information, like other channels, reaches only a portion of the desired clientele due to various accessibility and affordability barriers, but it provides a laboratory for beginning to assess what works and what does not. In particular, Internet dissemination may help to explore opportunities for reconceptualizing the kind as well as the form and delivery of information of broad public interest, allowing the government, commercial providers, and the public to move beyond the form and content of "government information" based on traditional printed packages.23 PRIVACY Contrasts in outlook between research, education, and libraries and industry have a different flavor in the area of privacy (Box 4.2). On the one hand, there is broad agreement about the need for confidentiality for sensitive information (independent of infrastructure use, per se) about individuals (e.g., information about personnel or personal matters) or organizations (e.g., trade secrets or other proprietary information).24 Storing and communicating sensitive personal information lead to the possibility that information will be divulged, or that information might be modified, sometimes with life-or welfare-threatening consequences. On the other hand, information about what consumers do is routinely of interest to businesses, as reflected in the proliferation of frequent-buyer programs, point-of-sale information gathering among retailers, and so on. The collection, analysis, and redistribution of information about individual behavior all have commercial value and have become the basis for a number of commercial database services. For example, market research and demographic data can be used to better target products and produce better or customized products; such data are becoming increasingly valuable. Moreover, individuals will sometimes give up privacy as a consideration for obtaining a service (e.g., medical insurance, credit), although they may not understand the ramifications of that decision.25 A principal point of contention appears to be information about patterns of network and information service use, information that is likely to be collected routinely through accounting mechanisms associated with use (and especially purchase) of network and information services. Such
PRINCIPLES AND PRACTICE 157 information collection has been an issue in the deregulation of telephony, where the debates have revolved around "customer proprietary network information" and "automatic number identification" information; these debates have begun to be carried forward into considerations of the larger NII policy.26 Electronic networking makes possible more frequent and more extensive collection of user behavior data as an automatic by-product of use, and it facilitates the amassing of data from multiple sources accessible over networks. As a result, privacy implications of such data gathering and analysis are attracting broader attention.27 The treatment of information about infrastructure useâin particular, who owns and accesses BOX 4.2 CONFIDENTIALITY IN CONTEXT: ACCESS CONTROL NEEDS Community: Key Rationale: Research â¢ Avoid being scooped â¢ Avoid mischief and preserve integrity of databases and research writeups â¢ Protect individual researcher privacy â¢ Avoid giving away market research â¢ Prevent competitive advantage in grant-seeking derived from knowledge of researcher's patterns of information use Education â¢ Protect privacy of student records â¢ Protect individual educator privacy, privacy of educator interactions â¢ Prevent unauthorized disclosure of test and homework answers â¢ Avoid mischief and preserve integrity of educational materials â¢ Prevent unauthorized access to sensitive material (e.g., pornography) while preserving First Amendment rights â¢ Ensure that intellectual property rights requirements are met Libraries â¢ Protect privacy of library user records (e.g., who borrowed or accessed what) Hospitals/health â¢ Protect privacy of patient records â¢ Protect privacy of physician consultations, records â¢ Avoid mischief and preserve integrity of medical records, test results, and so on â¢ Protect eventual flows of funds
PRINCIPLES AND PRACTICE 159 protect public interests in content). A related issue is placement on the network and in address directories, which ideally should be provided without discrimination or favoritism. It is in the public interest to have a variety of information sources. An important lesson from the Internet experience is that users should have the ability to place information on the networkâa capability that has differentiated the Internet, whose users have emerged as both authors and publishers. A corollary is a right of users to offer services over the network.33 How to extend this feature into a larger, commercial environment raises many questions, since the First Amendment provides for freedom to publish by media owners, but it does not necessarily provide a right of access to media. What the Internet experience shows, however, is that the convergence of digital media enables an effective separation of content from the carrier. People are increasingly able to create, deliver, and express in text, images, and sound regardless of the mediumâone can define an information product by what it does rather than by the medium to which it is bound; there may not be a physical medium.34 The Internet experience also illustrates the value of broad-based, many-to- many expression, in contrast to telephony, which is generally point-to-point communication, which inherently limits the scope of one's audience. The Internet provides individuals with access to mass communication channels, a feature particularly evident in its use during popular uprisings in the People's Republic of China and the former Soviet Union. However, open interfaces and nondiscriminatory access to media can produce outcomes that trouble some for various reasons (such as 900-number pornography services in the case of telephony); some of these are subject to legal protections, and others are expected to be addressed by current information policy efforts. In K-12 education, content restrictions are driven by the presence of minors and varying sensitivities about how much access children should have to "adult" material of various kinds. To date, school networking projects have acted to restrict access to pornographic and other controversial information available via the Internet (which is not necessarily possible) and also to educate children, teachers, and parents about appropriate materials.35 Exploration continues. It is worth noting that libraries have a large base of experience in this domain, having foughtâusually successfullyâ censorship battles in every information format introduced to date. Librarians have developed both policies and tools and put them into regular practice. Within the NREN environment, the scope of expression has been explicitly constrained. The most notable attempt to do so is through acceptable use policies that have arisen in part to control how government-funded networks are used. For example, NSF has imposed an honor-code acceptable use policy (AUP) that has restricted the content of
PRINCIPLES AND PRACTICE 160 communication over the NSFNET and militated against the conduct of commercial transactions over the Internet; similarly, research networks established by the Department of Energy and the National Aeronautics and Space Administration (NASA) limit traffic to mission-related transactions. Although AUPs have caused controversy (from allegations of censorship to observations that they are impossible to police), it is presumed that they will diminish with the decline in federal support for network infrastructure.36 That is, the federally supported networks with AUPs, such as the anticipated NSF very high speed backbone network service (vBNS), will constitute a shrinking portion of the Internet (and NII). The current NSF cooperative agreement establishing network access points will allow the AUP-constrained vBNS to be connected to networks the rest of whose traffic is free of the NSF AUP. During times of technological change, the interpretation of constitutional protections, such as the First Amendment, is normally developed by a long litigation process with case law being developed and further interpreted. Over the past 200 years this model has served us well primarily due to the relative slowness of technological change. The environment to be created by the NII suggests that the models implicit in case law may need to be revisited. A number of questions need to be addressed, for example, What is a bulletin board operatorâa publisher, a book store owner, or a new type of enterprise? and, What are its liabilities and responsibilities with respect to the information it is custodian of? What are the responsibilities of the new unregulated data carriers, including e-mail service providers, with respect to the material they carry? The speed with which these issues are faced will affect the growth of the NII culture. Decisions that are excessively restrictive may slow it down or change its nature. A balance must be found between the rights of the operators and those of the users.37 INTELLECTUAL PROPERTY PROTECTION There is a broad public consensus that there should continue to be protection of copyrighted materials, and that the protection of copyrights of materials distributed in the networked environment should reflect the rights of both creators and users.38 There is also broad appreciation that a robust market for networked information and resources is fundamental to the success of the evolving NII. Much less certain is how intellectual property protection can or should evolve to fit the networked environment. Observed legal scholar Thomas Dreier,39 How do digital technology and networks affect [the] background against which copyright was previously applied? Obviously, digital technology alone doesn't, since a work stored on a material carrier in digital form
PRINCIPLES AND PRACTICE 161 also travels the route from private via public to private. Rather, the problem caused by digital technology as such is one concerning the ease of reproduction. However, it is the network which brings about the substantive change. It links the private sphere of the authorâor of the person or entity offering the work in its marketable formâdirectly to the private sphere of the person who enjoys or re-uses the work. Thus, the public sphere on which copyright relies to such a great extent is eliminated, and little more is left than the umbilical cord of the connecting net-line which runs through what used to be the now-eliminated former public sphere. It also follows from the disappearance of the public sphere that any person enjoying or re-using a protected work via a network, reaches from his or her own private sphere directly into the private sphere of the author who makes the protected work available. (p. 193) Intellectual property in the networked environment is bound up with how intellectual property protection should evolve to cover digital media generally. For example, copyright law applies to the multimedia environment as it applies to the print world, but some have raised questions about the application of intellectual property protections to the hybrid products made possible by digital convergence. How these questions are resolved will affect business development and process planning. The area of intellectual property rights represents one of the greatest areas of difference among the research and education and commercial communities. To begin with, it is worth noting that there is broad agreement on one aspect of intellectual property protection: the importance of protecting data integrity. It is critical that creators know that what they produce is what network users get, and for users to be assured that what they are getting is what they think it is. The research, education, and library perspective was articulated by the University of California's Clifford Lynch as follows:40 In discussing issues of access and integrity in networked information today, there is a very strong bias towards issues specific to scholarly information; this is to be expected, given that the academic and research communities have up until now been the primary constituencies on the Internet. These are relatively sophisticated communities, and communities with an ethos that is strongly oriented towards citation, attribution, and preservation of the scholarly record. Indeed, â¦ this ethos ties these communities to the system of print publication, and emphasizes that networked information must offer the same integrity and access if it is to become an acceptable replacement for print. â¦ (p. 16) Concerns relating to integrity suggest a need to protect against the dissemination of an altered original work without authorization, or if this is not possible, a means for "certifying" the authenticity of the data. 41 (Mechanisms to do this may be developed under the security umbrella; see Chapter 2.)
PRINCIPLES AND PRACTICE 162 Where there is fundamental disagreement about intellectual property protection is in the area of charging for both acquisition of the item itself and then for each use, since property rightsâa form of ownershipâcarry with them an inherent ability to reap value through sales. Charging for copyrighted materials delivered over the Internet was contemplated in the High-Performance Computing Act of 1991.42 The opportunity for financial compensation is the reason that intellectual property protection, especially copyright, is broadly recognized as nurturing the quantity and diversity of information creation generally in the commercial marketplace. In an electronic environment, the potential for violation of the law (and for related theories in the law such as misappropriation) is greatly exacerbated. It is increasingly possible for virtually anyone to make a perfect copy of an information product. Indeed, the ability to make multiple perfect digital copies, accessible instantly to millions of users worldwide, is an exciting prospect, but it can also be a license to steal intellectual property cheaply, easily, and in a way that destroys incentives for future creativity. This battle has already been fought by the software industry, which has persuaded at least large organizations to buy or license appropriate numbers of copiesâtheir educational and legal efforts attest to the possibilities for (re)educating the public. Recognizing that the Internet has grown up within a culture that thrives on sharing information and research, commercial information providers have approached the network with great caution. Commercial authors and publishers are concerned that their works will be accessed, easily downloaded, and redistributed without appropriate authorization and remuneration.43 The scale of a network such as the Internet makes this prospect a substantial deterrent to making information available. Publishers and commercial information service providers believe that if this issue is not adequately addressed, not only will many information providers be unwilling to trust their assets to these networks, but they will also, in fact, have a strong incentive to keep their property from ever being included. This obviously runs counter to the most central goals of the NREN and NII efforts. In the research community, the relatively low emphasis on remuneration is one factor behind a greater willingness to publish electronically. Others include a need for rapid dissemination of information and broad circulation of new ideas, "food for thought," and research results.44 These factors have given rise to a substantial amount of experimentation with and demand for electronic publishing capabilities, services, and support.45 Universities and their faculties, students, and libraries play multiple roles in relation to intellectual property: as creators, users, dissemina
PRINCIPLES AND PRACTICE 163 tors, and archivers. U.S. research and education have a special place in copyright law through provisions for fair use, libraries, archives, classroom use, and some performances and displays. While universities eagerly lay claim to patents (revenue-producing), they often forego any claim on copyrights (cost- saving), and faculty often readily transfer copyrights to publishers in exchange for achieving publication and the chance for published knowledge to be attributed to them. The end result is that universities sometimes find themselves in a situation where they cannot afford to buy backâthrough serials and other acquisitionsâ the research information their faculty have generated in their service. In addition, the campus environment is one of a general lack of knowledge about and understanding of the current copyright law, its fair use provisions, and the ramifications of electronic publishing. Where these issues are being addressed, several options are being considered (Box 4.3). In contrast to the academic situation, there are more formal transactions involving rights and remuneration in the commercial publishing arena. However, electronic publication has raised new questions in that arena that may affect all kinds of authors. Specifically, authors may be less knowledgeable than publishers about options for republication and associated revenues through electronic databases and other electronic vehicles.46 Thus, the author may focus on the original publication, and not a far broader publishing process through electronic databases. Overall, however, authors as well as publishers are experimenting in the commercialization of these paths and in the charging for and control of the intellectual material flowing and stored in electronic form. Publishers are experimenting with new ways to distribute information and manage copyright over electronic networks; they are experimenting with developing copyright management mechanisms using digital technology and aimed at detecting, monitoring, and inhibiting improper use.47 Some of those experiments revolve around the academic community. For example, the Elsevier TULIP project, involving several U.S. universities, is exploring technical and economic issues relating to the use of campus networks and the Internet to distribute journals. Springer-Verlag has an experimental project (Red Sage) based on a Bell Laboratories system for delivery of content from 24 journals over the Internet. It can manage journal page image content on a network and alert users by electronic mail when subjects that interest them are covered in journals they do not ordinarily read.48 The possibilities for republication raise questions about shifts in the costs and benefits of publishing.49 Efficient mechanisms for negotiating and paying for partial or complete copying, for "reproduction" rights, and so on are still in development (they exist in certain closed electronic
PRINCIPLES AND PRACTICE 164 delivery systems, such as for the NEXIS and Dialog services), even though technology provides ultralow cost mechanisms for reproduction and copying. Early in 1994, for example, Dialog Information Systems announced its intention to adopt a charging system based on the number of authorized users, effectively selling the right to share electronically the documents obtained through its system. Although such an approach would, if enforceable, reduce loss of revenue from unauthorized copying, it adds to the growing body of questions about redistribution and reprint rights.50 BOX 4.3 COPYRIGHT MANAGEMENT OPTIONS UNDER UNIVERSITY POLICY AND SUPPORT Faculty Ownership. The faculty member, as the author under the copyright law, retains copyright. The faculty member is then free to license uses within the university (and higher education) or assigns or licenses to the publisher the right to reproduce and distribute the work as necessary. Joint Faculty/University Ownership. The faculty member and the university share jointly rights in copyrighted works. Such joint ownership would be limited to that work generated within the scope of the faculty member's employment. University Ownership. The university, as copyright holder, makes all decisions relative to publication, licensing, royalty agreements, and so on. Rights are transferred to the university for works generated within the scope of the faculty member's employment. Ownership by a Consortial Body. The faculty member assigns rights to a consortial body, e.g., the National Association of State Universities and Land Grant Colleges or the American Association of Universities, which then acts as a collective rights society to manage copyrighted works in the best interests of universities and their faculties. Experimentation within the research and education communities suggests that in the future, the modes of information consumption may change or there may be greater variety in modes. For example, there is a tendency toward more casual publishing growing out of liberal electronic mail, public file repository, Gopher server, and Mosaic/World-Wide Web experiences. (See Box 2.2 in Chapter 2.) There also appear to be more "synthetic" applications that involve using small amounts of material from multiple sources, and more short-lived uses of information (in some cases because the information is frequently updated, in other cases as interactive databases or other informa
PRINCIPLES AND PRACTICE 165 tion resources proliferate). So-called "hot links" allow users to reach out from one kind of document into other, related material through the network; the Mosaic interface to the World-Wide Web illustrates this phenomenon. In addition, markets are demanding products by the piece (e.g., single articles) rather than the conventional publishing package, and pieces are being integrated by users. The combination of new applications and a broadening user base challenge conventional approaches to pricing for information services. For example, current expectations about amount of royalty per sale or use may be inappropriate in an environment where a far larger number of people may seek to use the information and/or where the uses may be more short-lived than has been customary with paper or other fixed-medium products. Overall, all players in the NII publishing area must reevaluate their economic expectations and adapt them to the realities of the electronic world. BROADER CONSIDERATION OF ETHICS The committee anticipates that the emerging NII will dramatically increase confrontations among conflicting rights and needs of society. These conflicts will not necessarily be about new issues, but rather about a change in the magnitude and urgency of those that already exist. Core issues will revolve around ability to access information, the cost of that access, and respect for intellectual property, privacy, security, and information integrity. Triggering conditions may include the increased ability to collect and collate data on all aspects of our lives and the increased ability to cause harm by altering public and private records. Technological fixes to these problems are not in sight. The ethical, legal, and social implications of this technology are paralleled by those arising from discoveries in molecular biology. Some of these issues will probably be worked out in the marketplace, some in the courts. As in other domains, government monitoring and possibly more active involvement may become necessary. More important, the courts unquestionably will need guidance in issues that demand a high level of technological knowledge, but are not simply technological problems with technological fixes. And there will certainly be a need to increase public awareness of many complex issues. Building from the IITF information policy efforts and depending on their results, consideration should be given to a program to fund research and public education in this area to provide the guidance that will be needed. A reasonable model would be the Ethical, Legal, and Social Implications (ELSI) program of the Human Genome Project.51 The goals of an ELSI component to the NII would be as follows:52
PRINCIPLES AND PRACTICE 166 1. Address and anticipate the implications for individuals and society of the NII; 2. Develop policy options to assure that the information is used for the benefit of the individuals and society. 3. Enhance and expand public and professional education that is sensitive to individual rights and responsibilities associated with the NII. Examples of such efforts would be workshops, panels to frame policy options, training grants for those with a technological background to acquire training in ethics and law (and vice versa), studies about how current opportunities for abuse are being exploited, public education programs, and pilot programs for introducing the study of such issues in the schools. Within NSF's Computer and Information Science and Engineering Directorate, there are programs (e.g., within the networking and cross-disciplinary activities divisions) that could house such a research program, although the scope of the problem may dictate a more broadly based institutional home, such as a national commission. NOTES 1. This development has been anticipated by the incorporation of activities relating to information policy under the administration's Information Infrastructure Task Force as well as by the corresponding broadening of technical activities under the High-Performance Computing and Communications umbrella. 2. See, for example, U.S. Department of Commerce. 1994. 20/20 Vision: The Development of a National Information Infrastructure. Government Printing Office, Washington, D.C., March. This volume contains papers presented at a special conference that was requested by the administration to stimulate debate over the broad principles and goals advanced in the fall 1993 policy statement on the NII (Information Infrastructure Task Force. 1993. The National Information Infrastructure: Agenda for Action. Washington, D.C., September 15). 3. The issue of metrics and evaluation is recognized by the administration, which posed the problem to speakers at its March 1994 20/20 Vision conference. 4. Although the analogy is frequently made today to universal telephone service, it should also be recognized that that entitlement, to the extent it is one, does not extend to universal access to 900- number or other special services, nor are newspaper publishers obligated to distribute free or discounted copies to the indigentâwho can and do read the paper in public libraries. Also, recognizing that today's universal access for voice telephony took time to be implemented, it is likely that in some transition period there will be inequalities. 5. For example, this tack has been taken by the electronic text archive developers at the University of Virginia, who hope that their apparent success will stimulate commercial publishers. See Seaman, David. 1993. "Gate-keeping a Garden of Etext Delights: Electronic Texts and the Humanities at the University of Virginia Library." Gateways, Gatekeepers, and Roles in the Information Omniverse: Proceedings from the Third Symposium, November 13-15, Washington, D.C., pp. 63-67. 6. Of course, continued "publishing" by information-producing individuals is still expected and desirable. 7. Computer Science and Technology Board (CSTB), National Research Council. 1988.
PRINCIPLES AND PRACTICE 167 Toward a National Research Network. National Academy Press, Washington, D.C. (The Computer Science and Technology Board became the Computer Science and Telecommunications Board in 1990.) 8. A significant percentage of the population is now network literate and has a computer at home. According to the Department of Commerce, over 31 million U.S. households owned personal computers in 1993; two-thirds of home PCs are believed to support people who work from their homes. See U.S. Department of Commerce. 1994. U.S. Industrial Outlook 1994. Government Printing Office, Washington, D.C. While the home-working, network-literate sector of the population will probably never be sufficient to motivate the universal, high-bandwidth networking of the country (which might, however, be motivated by entertainment and consumer activities), it is a sector that must be assured reasonable access to the Open Data Network, for reasons of public policy such as balanced access to network information. Today, it is very difficult for individuals to procure this service, to some extent for reasons of cost but more basically because there is no standard way of purchasing the service. General consumer access does not appear to justify any special subsidy. The consumer should be asked to pay for attachment, assuming that the cost will be reasonable. What is required is a program to develop a well-defined means to provide this service. For example, if a specific initiative were undertaken by the government to drive the deployment of lowcost networking into schools and public libraries, it could promote the development of technology suited for the network attachment of individuals (assuming the small number of schools per central office did not diminish the impact and appeal). 9. Existing and likely demand levels present a natural pacing factor, but it should be recognized that there may be an exaggerated lag in the development of desire for access. In particular, educators, for example, in schools in unaffluent areas and educators without exposure and training relating to the potential of networks and information resources may not know enough to desire access. 10. CSTB, 1988, Toward a National Research Network. 11. Traditionally, some sharing has been limited as a function of competition for tenure and credit, but this resistance has been diminishing in the face of pressures promoting collaboration. 12. Electronic publishing will change many dynamics affecting the scholarly record, the dissemination of research results, and the evaluation of the work of researchers. Some of the confusion now arising is reflected in the following comments (made at a Modern Language Association Convention in Toronto on December 29, 1993) of Bill Readings of the University of Montreal: "if an article is never technically 'in print,' how do you know it was actually 'published'? If revisions and new footnotes and retractions and formal responses become parts of the electronic archive, doesn't 'publication' cease to mean a finished artifact and become instead an unending dialogue of readers and authors? What then becomes of journal editors and their screening panels, the so-called 'gatekeepers' of scholarship, when space and cost considerations vanish and you only need one reader to legitimate publication on the Internet? Many accuse academics of only writing for themselves. . . . The virtual university gives new meaning to the quip." See Trueheart, Charles. 1993. "The Writing on the Cyberspace Chalkboard," Washington Post, December 30, pp. C1-2. 13. A fundamental issue in relation to acquiring, storing, and sharing information resources in the electronic environment is ownershipâwho owns the information has had an enormous impact on the prices charged to the customer. A number of cost-per-unit studies of various kinds of scientific, technical, and medical (STM) journal pricing suggest that the not-for-profit producers sell information significantly less than the for-profit publishers, specifically the large, offshore European STM publishers. See Okerson, Ann. 1993. "Networked Information, Scholarly Publishing, and Electronic Resource Sharing in Academic
PRINCIPLES AND PRACTICE 168 Libraries: a Dilemma of Ownership," paper distributed by Association of Research Libraries, Washington, D.C., June. 14. Within one week alone in early 1994, there were several announcements of new services, including a Los Angeles Times-Pacific Telesis electronic shopping service; the launch of a TV evening newscast by the Philadelphia Inquirer, and the formation of an alliance to develop an electronic publishing service by the Washington Post and Oracle. Glaberson, William. 1994. "Newspapers Race for Outlets in Electronic Marketplace," New York Times, n.d., pp. D1 and D6. 15. For example, Encyclopedia Britannica Inc. recently announced plans for electronic distribution to universities and some public libraries over the Internet. Of the several encyclopedias already available on-line, Encyclopedia Britannica is the largest. The Britannica Online service being tested at the University of California at San Diego features hypertext links that cannot be matched by traditional text versions; links among the macropedia, micropedia, index, and propedia broad outline of world knowledge; and WAIS and Mosaic access. At the time of the announcement, the pricing was uncertainâoptions for pricing on a subscription basis and reference by reference (enabled by WAIS and Mosaic) were being considered. See Markoff, John. 1994. "Britannica's 44 Million Words Are Going On Line," New York Times, February 8, pp. D1-2. 16. Participants in a 1993 workshop noted that there are many questions about the division of labor between public and private sectors in creating, storing, organizing, marketing, and distributing noncommercial information. See Library of Congress. n.d. "Delivering Electronic Information in a Knowledge-based Democracy," Summary of Conference Proceedings, July 14, 1993. 17. The High Performance Computing Act of 1991 and other pieces of legislation have increased support for and exploration of electronic libraries. 18. Some of these arguments are reflected in the Paperwork Reduction Act (PL 96-511) and Office of Management and Budget (OMB) Circular A-130 (revised in 1993) on federal information resources management. The Information Industries Association has published a pamphlet entitled "Serving Citizens in the Information Age: Access Principles for State and Local Government Information," which provides guidelines in support of "three fundamental tenets": a broad public right of access, a right of nondiscriminatory access, and a prohibition on government control of information access and use. 19. Rohter, Larry. 1994. "Florida Weighs Fees for Its Computer Data: Some See Profits; Others, Too High a Price," New York Times, March 31. 20. See, for example, Bass, Brad. 1993. "EDGAR/Internet Effort Leads Public Access to Federal Data," Federal Computer Week, November 1, p. 9. 21. One relevant interagency proposal is for a Government Information Locator System (GILS) describing each agency system containing publicly accessible information plus other cataloging information. The proposal contemplated a "wholesale" service, as opposed to direct end-user access. Power, Kevin. 1993. "How Will We Navigate the Federal Data Maze?" Government Computer News, November 22, p. 47. The OMB plans to issue a directive establishing the GILS and outlining what agencies have to do to comply with it, including creation of agency inventories of information dissemination products and automated information systems. Sikorovsky, Elizabeth. 1994. "OMB to Issue Guide to Establishing a Government-wide Locator System," Federal Computer Week, April 25, p. 6. 22. For an assessment of options, programs to date, and dimensions on which evaluations might be conducted, see U.S. Congress, Office of Technology Assessment. 1993. Making Government Work: Electronic Delivery of Federal Services. OTA-TCT-578. Government Printing Office, Washington, D.C., September. 23. Fisher, Francis Dummer. 1994. "Open Sesame! How to Get to the Treasure of Electronic Information." In 20/20 Vision: The Development of a National Information Infrastructure.
PRINCIPLES AND PRACTICE 169 U.S. Department of Commerce, National Telecommunications and Information Administration. 24. According to a 1993 Louis Harris survey, 53 percent of Americans are very concerned about their privacy. That survey marked the first time such concern was registered by a majority in 23 years. Privacy concerns were greatest for dealings with banks, health insurers, and hospitals among various types of businesses. See "Mind Your Own Business,'' Wall Street Journal, October 5, 1993. 25. A significant feature of the draft "Principles for Providing and Using Personal Information'' and their "Commentary," distributed for comments by the Information Infrastructure Task Force Information Policy Committee Working Group on Privacy in the NII on May 4, 1994, is the "protection principle," which states that "Users of personal information must take reasonable steps to prevent the information they have from being disclosed or altered improperly. Such users should: (1) Use appropriate managerial and technical controls to protect the confidentiality and integrity of personal information" (p. 3). Thus, although the IITF also provides an education principle, recognizing that individuals may not understand the impacts of networked information, it argues that individuals should take an active responsibility for information concerning themselves. See Appendix B. 26. See, for example, "Privacy Comments Urge Industry Self-Regulation." 1994. Telecommunications Reports, April 4, pp. 32-3, which documents comments submitted in NTIA proceedings relating to privacy and the NII. Telephone companies have argued against additional regulation; the National Cable Television Association and others have called for technology-neutral approaches; distinctions have been posed between actions relating to data misuse versus actions relating to collection, transmission, and analysis of data, per se. 27. Various frequent buyer programs, including the :Original frequent flyer programs started by airlines and more recently grocery store-based systems, are examples. 28. There is also the prospect of legislation to prohibit the association of the user's name with the specific user data themselves. 29. Noll, Roger G. 1993. "The Economics of Information: A User's Guide." Pp. 25-52 in The Knowledge Economy: The Nature of Information in the 21st Century. Institute for Information Studies, Nashville, Tenn. 30. For example, the computing and communications community has organized an electronic forum, RISKS, devoted to sharing information about and discussing a variety of risks relating to privacy, security, safety, and the reliability of computer-based systems. There are other electronic discussion groups specifically focused on privacy. 31. Institute of Medicine. 1991. The Computer-Based Patient Record. National Academy Press, Washington, D.C.; and Institute of Medicine. 1994. Health Data in the Information Age: Use, Disclosure, and Privacy. National Academy Press, Washington, D.C. 32. The Taxpayer Assets Project, for example, sent a letter to Congressman Markey expressing concern about increasing vertical integration in service provision, inasmuch as telephone companies, cable companies, and wireless carriers can own rights to programming content available through their facilities (November 10, 1993, e-mail). 33. However, most users who want to offer information will not necessarily want to make the commitment involved in offering a service (e.g., running a computer in their homes with the reliability expected from a service provider). New services providing disk storage and reliable access might emerge as information intermediaries. 34. Baron, David. 1993. "Digital Technology and the Implications for Intellectual Property." Pp. 29-36 in WIPO Worldwide Symposium on the Impact of Digital Technology on Copyright and Neighboring Rights. WIPO, Geneva. 35. Schwartz, Jolm. 1993. "Caution: Children at Play on Information Highway," Washington Post, pp. A1 and A26. 36. This trend is evident in the anticipated proliferation of network access points in the new NSFNET architecture.
PRINCIPLES AND PRACTICE 170 37. The Computer Science and Telecommunications Board is addressing some of these issues through its project "Rights and Responsibilities for Participants in Networked Communities,'' for which a report is anticipated in mid-1994. 38. The Copyright Act strikes such a balance and in particular, section 107 stipulates fair use provisions that should be considered in a networked environment. It is too early to consider or to impose restrictive agreements that limit access, limit creativity, or undermine fair use provisions as these could be detrimental to long-standing policies that promote public access. 39. Dreier, Thomas K. 1993. "Copyright Digitized: Philosophical Impacts and Practical implications for Information Exchange in Digital Networks." Pp. 187-212 in WIPO Worldwide Symposium on the Impact of Digital Technology on Copyright and Neighboring Rights. WIPO, Geneva. 40. Lynch, Clifford A. 1993. Accessibility and Integrity of Networked Information Collections. Background Report/Contractor Report prepared for the Office of Technology Assessment, July 5. 41. Imagine the disasters that could occur if, for example, chemistry laboratory manuals were tampered with. Consider the liability issues. Consumers should have reason to believe that the procedures for conducting a potentially explosive experiment, for example, are accurate to the best of the chemist/author's knowledge and have not been altered by someone using a home chemistry set. Users must be able to choose sources of information with a reasonable degree of confidence. 42. In the interest of promoting the development of information services, such as federal scientific data, directories, and commercial reformation services in support of research and education, "[t]he Network shall provide access, to the extent practicable, to electronic information resources maintained by libraries, research facilities, publishers, and affiliated organizations . . . [and] have accounting mechanisms which allow users or groups of users to be charged for their usage of copyrighted materials available over the Network and, where appropriate and technically feasible, for their usage of the Network" (PL 102-194, section 102). 43. This issue arises with other modes of electronic dissemination. For example, the availability of CD-ROM publications at the Library of Congress was achieved only after two years of negotiations with publishing and library entities; physical security and access restrictions were part of the agreement. See Minahan, Tim. 1992. "With Copyright Protections in Place, Library Receives First CD-ROM Publications," Government Computer News, November 22, p. 10. 44. For example, James Noblitt, head of the institute for Academic Technology at the University of North Carolina, was recently quoted as saying, "These days the world changes every six months and textbooks come out every three or four years. The technology is now in place to keep up with change, but textbooks can't do it." Cox, Meg. 1993. "Technology Threatens to Shatter the World of College Textbooks," Wall Street Journal, June 1, pp. A1 and A5. 45. Perritt, Henry H., Jr. 1992. "Market Structures for Electronic Publishing and Electronic Contracting on a National Research and Education Network: Defining Added Value.'' Building Information Infrastructure: The NREN and the Market. Harvard University Press, Cambridge, Mass. 46. Bald, Margaret. 1993. "The Case of the Disappearing Author," Serials Review 19 (3, Fall):7-14. 47. There are a variety of joint university-industry projects aimed at developing technological means of providing on-line authorization, tracking, and/or collection of royalties. However, the results of these efforts are treated as proprietary information. In addition, the Corporation for National Research Initiatives has been working with the Library of Congress/Copyright Office and ARPA on network-based copyright management.
PRINCIPLES AND PRACTICE 171 See also Oman, Ralph. 1993. "Reflections on Digital Technology: 'The Shape of Things to Come.'" Pp. 21-24 in WIPO Worldwide Symposium on the Impact of Digital Technology on Copyright and Neighboring Rights. WIPO, Geneva. 48. "Electronic Publishing Case Studies: Implications for Periodicals Publishing," PSP Bulletin 7(3, Winter 1993):3-5. 49. See, for example, "Writer Sues Publishers Over Articles in Databases," Wall Street Journal, Friday, December 17, 1993, p. B5(w). 50. Riordan, Teresa. 1994. "Fee Plan to Share On-Line Data," New York Times, April 6, pp. D1 and D6. 51. The idea was first suggested to the committee in a June 1993 briefing by Richard Civille, Center for Civic Networking Groups, and later explored and expanded by biologist and committee member Charles Taylor. 52. The Human Genome Project initially budgeted 3 percent of its budget for its ELSI component and has increased this to 5 percent for the next 5-year period. The committee specifically refrains from making recommendations about funding amounts at this time, recognizing that in its inherent focus on the nature of life the Human Gertome Project presents more uncomfortable ethical and policy challenges than the NII, but the portion chosen should be a correspondingly significant part of the NII budget.