University Librarian, The University of Southern California
The second presentation of The Follett Lecture series,
London, 9th June 1994
If, as has been said, we live in an "information society," what is the future of the library, the institution which has traditionally managed recorded knowledge? A library is not only a collection, it is a place ---the centre of a town, a company or a university --- which defines the shared knowledge of a communityand conserves its historical memory. How will the birth of an information society, and the digital communications technology which made it possible, change the library? The library, information technology, and our sense of intellectual community are changing together, and the future of the library must be designed within an understanding of network technology and a vision of the future of education and scholarship.
I envision the electronic library as a global reference room. The global reference room is a renewal and continuation of the enlightenment vision of the library as the organised totality of recorded knowledge. But the global reference room is not only an inventory and index of the great writings, it will be the first global culture and social institution for shared knowledge, and a counterweight to global economic competition. This definition of the electronic library as a medium for global co-operation is not utopian, it is a recognition of the structure of the technology of networked information, and of the emergence of global scientific and scholarly communities.
The scale of the electronic library must be global, because digital networks are global, and increasingly our problems, social visions and economy are global. Networked digital information is an established currency in international trade and global financial markets; as individuals we are accustomed to an environment of ubiquitous computer mediated transactions, such as credit cards and ATM machines. The network is the new geography of trade, thus countries like Singapore and Australia have made significant national efforts to develop national information policy in order to define strategic roles in the global network trade routes. But, as yet we have not thought sufficiently about global markets for higher education services, for teaching and research, or about shared information resources as a tool for solving global problems.
While the electronic library may be very useful for the management and exchange of printed texts in a global collection, two other characteristics of the electronic library are likely to be as important. First, information technologies have become a new medium for the creation and representation of knowledge in formats which cannot be reproduced in print; this is evident in the growth of digital arts such as animation and music, and in scientific and medical imaging. Second, the global reference room will be a new medium for "virtual" social communications and relationships, in which people around the globe will participate in entirely new forms of collective work and relationships; this is suggested by electronic mail and electronic journals, "distance learning" experiments, and even networked computer games.
To me, the Internet may be the most interesting social organisation of our time. It is global in scope; it's almost impossible to know how many people use the Internet, although the last estimate was 30 million people. Its is a global communications network designed and built by the researchers who create information and consume technical and scientific information. It has no governance authority or structure; the Internet is organised around a shared culture, which originated among its inventors --- the scientific and technical community which received funding from our Defence Department and National Science Foundation. Information is "free," although this really means that its use was subsidised. Its user interface reflects its military cultural origins, as reflected in the "control," "break," and "escape" keys and the "command" interface (clearly the electronic library can provide more useful metaphors for global information than war or trade). And, the net is a broadcast medium, invented to share knowledge; experts, using software like WAIS and now World Wide Web and Mosaic, can provide world-wide access to what they know. The net is an enormous collection of information, but it is a random, disorganised and occasionally bizarre collection. To create an electronic library from the Internet, our task as librarians is to invent a way to perform the functions for the network that publishers perform for print, namely quality control, organisation, marketing, pricing.
Although initially designed to provide remote access to super computers, the Internet is now used primarily as a medium for scientific and technical communication. Recent studies indicate that two- thirds of the use of the Internet by research scientists is for information management (such as electronic mail and document exchange) not computation. The values of the library have guided technical innovation on the Internet; client/server technology and applications software (such as Gopher and Mosaic) are designed for the sharing of information, balancing decentralised local control with international standards.
My argument, then, is that the global reference room is already in existence; it is something we must recognise and improve, not something we must invent. It exists, although imperfectly, in the Internet, a global medium for building the electronic library. The Internet is as important as an example of a virtual community and a vision of shared information as it is a technical achievement. The electronic library is not a logical extension of print libraries, it is the recreation of the library in a new medium.
In contract to the idea of the electronic library as a global social organisation, the Clinton administration envisions a National Information Infrastructure (NII), defined as "a seamless web of communication networks, computers, databases, and consumer electronics." Just as the interstate highways created national markets for commodities, the telephone companies will build an "information highway," as they call it, to give universal access to information. Commerce in information commodities is intended to train the work force and promote economic development for global competition. Technological innovation will focus upon universal access through connectivity, and encryption and billing technology will turn ideas into intellectual property and protect it.
It is telling that the American NII is founded upon an anachronistic industrial metaphor like the "information highway." If the computer age has taught us anything, it is that technical innovation must begin with a vision of the communities and institutions which create and use information. Ultimately the idea that information is a commodity is only a fiction useful for understanding markets, but information may only properly be understood within the context of human communication and activity.
The Internet, not the telephone company, is the best test bed for the technologies of the NII, and equally important, a laboratory for understanding how information technology may reshape communication, and thereby make possible new organisational strategies. The Internet is global, not national. It is governed by participation and shared values, not a national government. It is tolerant of a great diversity of user activities. Communication is uncensored, and use is subsidised. Most important, the Internet has been designed by those who create and use knowledge and information within a gift economy; information is given away freely in the expectation that it can be used freely. These characteristics reflect the values of academic and research cultures. Thus one of the most important questions for the NII has been posed by the Computer Professionals for Social Responsibility: "Are the conditions for intellectual freedom possible on the network?"
There are, of course, defects with the Internet model. The Internet is designed by and for technicians and is often difficult to use. While information on the net is vast it is often of poor quality; subsidy is not a viable economic model for a national information infrastructure; and information property rights are often abused. It serves higher education and defence industries well, but secondary education and the public interest poorly. And, thus far it is not clear how academic gift exchange cultures and commercial information markets can coexist on the network, although history suggests that greater synergy is possible by coexistence than the domination of either alone.
Our social vision should begin by asking, whose needs should be served by the information highway? The NII design is being driven by commercial interests in a market for information -- multimedia entertainment, government documents, copyrighted published works -- that doesn't exist yet. But innovation must begin with a vision of the knowledge communities which will use information: local and global; producers and consumers; commercial and non commercial; present and future. Information is part of a process of communication, used within the context of institutions -- work, education, libraries, government --- which are very likely to change in dramatic ways. Technology should anticipate the new kinds of organisations and social relationships which the network will make possible, not reproduce the power structure of an industrial society.
If the global reference room must be discovered rather than invented, it is also the case that the Internet is far from a library, and must be refined through a systematic process of design guided by librarians. There are three design problems which the library community might begin to solve.
First, if, as postulated earlier, a library is the centre of a community, what is our vision of the community for which the electronic library is a defining resource? The global reference room will be based upon a communications medium, thus we must design the library as a social institution, building upon computer mediated social relationships. Cybernetic space by no means replaces the physical library, but will certainly change its boundaries in unexpected ways.
Second, reference in the global reference room will require the design of user interfaces and tools to navigate the network, and will teach people how to use them. Today's computer environments were designed by and for a relatively homogeneous technical and scientific community to manage data, while the electronic library will be used by a wide variety of people for both information management and communications. But even more fundamentally, we all feel saturated with information, and our greatest need is for navigation tools to manage our information environment, and the skills to select and evaluate appropriate information.
Third, the collections of the global reference room will not only require more efficient access to information created for print formats, but will be a medium for shared collections, and require the design of entirely new kinds of multimedia reading environments. Since information in networked environments is more like a utility than a commodity, we do not yet know how networked information would be organised as a market (how prices would be set, how information subsidised or paid for, how copyright enforced).
While thinking about the future of the library, I began with the long term implications of our campus wide information systems which we have implemented using bibliographic databases and network technology. Gopher, WAIS and Mosaic have raised a series of questions that I think are very striking for the library, for we have become a publisher of information which may be read by anyone on the Internet anywhere in the world. When a campus department wishes to place information on the network, the library provides a work station and network connection, and places them on the library's network interface. In return, the library defines standards for the ease of access and the organisation of information, but unlike publishers does not exercise editorial control or control content (except in choosing which information providers it will support). With this publishing strategy, the library has adopted a new role as a partner in the creation and dissemination of information of all kinds, not just authenticated scholarly information. And, just as a collection development policy for a library defines the interests of the college, so too does choices we make in constructing a networked information system (with its Gopher and Mosaic links to information sites all over the world).
Networked library information also raises new questions about our definition of the limits and boundaries of the library as a social institution. There are two specific examples I find thought provoking: Maricopa Community College district, an example of how a college comes to mirror an entire community, when "lifelong" education is taken as a mission; and the San Francisco Public Library, which is being designed to provide a new kind of relationship between citizens and government.
Maricopa Community College is located in Phoenix, Arizona; our community college systems provide the first two years of college instruction. In Maricopa Community Colleges the campus centre is often a "computer commons," a room with several hundred computer workstations with instructional and productivity software, and networked information. The commons is filled with students doing homework, or taking computer mediated courses on an independent basis. In the computer commons, the faculty act as coaches, the students as apprentices; the students will be using software to learn new skills and when they have a question they raise their hands, and the faculty will go and consult with them. One cannot see the commons without asking how the electronic library will change our understanding of teaching and learning, of the classroom? My tentative answer is that the library is the classroom of the future, a future with students learning independently, often at great distances from the college, and often with new kinds of learners. Thus on my last visit to Maricopa I came across something I did not expect, among the students in the computer commons were retired people participating in lifelong education, and unemployed people, for the city of Phoenix had contracted with Maricopa Community College to provide training to the unemployed.
What, then, are the social and geographical boundaries of the electronic library, and who are its users? The work of Ken Dowlin, the librarian at San Francisco public library, suggests that the electronic library is a fundamentally new form of political community. California has a very complex population with over 120 different language groups. Many of them are immigrants who are eligible for social services but do not understand how our social services are organised, and are often afraid of the government. Ken Dowlin is creating the library information system to be the point of access for information about all city services, for all cultural groups in the city of San Francisco. Some of the most interesting visions of the future of the National Information Infrastructure are by social action groups, who envision the network as a dimension of government, not only for information about social services, but as an environment for political participation.
The idea of political participation in the electronic library reminds us that the network is not only a broadcast information resource, it is based upon a two-way communications medium. We do not know enough about the social relationships that emerge from networked information but we know that they are experienced as genuine social relationships. Shirley Turtle's book The Second Self points out that interactive digital environments bring forth both intellectual and emotional responses. But it is an imperfect social relation, or one we do not yet understand. Sara Kiesler and Lee Sproull studied electronic mail as a medium for group decision making in a hundred corporations, and discovered that group decision-making tended to be more radical on electronic mail than it would be in a face-to-face meeting. They theorised that this had to do with the lack of the non-verbal feedback which governs conversation. Moreover, in electronic mail it is hard to place emphasis or nuance, so there is a tendency to exaggerate the message. Electronic mail is best for brief exchanges of information, not deliberation, unless the group is well trained in managing conflict. While electronic mail is a social relationship, it is a volatile one which needs a new software design if it is to be a medium for political participation.
Recently I came across a second case that was described in the December 21st issue of The Village Voice in an article called "A Rape in Cyberspace." Xerox PARC (Palo Alto Research Centre), our most imaginative digital 'think tank,' set up an interactive adventure game on the Internet as an experiment to test object oriented programming. If you logged on, it would appear as if you were walking through an old house. Every member of this conversation designed a private room in the house, symbolising their membership in a community. The social life was conducted in the living room and parlour, where rich conversations occurred every night. Apparently the graphic and interactive quality of this was so successful that people imagined that they were members of a community, although they had never met each other face to face and knew each other only through the characters which they constructed. One night one of the characters raped another character, which shattered this sense of community. They went through absolutely every debate that a real face to face human community might have gone through had a rape occurred. After identifying the rapist, they began a long discussion on the appropriate punishment for a rape in cyberspace. Is it sufficient to erase the user ID? No, that was capital punishment in cyberspace! and many felt that would be wrong. Although this is a kind of negative example of the possibility of community on the network, it suggests that network communication is a genuine form of communication, with all of the intimacy and patterns of domination of everyday life in any other setting.
Clearly we are only at the beginning of a long period of experimentation, which will require building prototype software to experiment with virtual classrooms. And, as each example makes clear, in designing cyberspace classrooms we need to invent both technical environments and new kinds of cultures to regulate networked communication.
The first step in designing the electronic library is the creation of the global reference room. Libraries will be global simply because networks and scholarly communities are global, thus we must think globally although we act locally. Reference will be the core of the library of the future, a library which will be focused on information services and access to information as much as an inventory of print collections. But this does not imply an end to libraries as architectural places. USC has just built a new Teaching Library designed to manage electronic and digital information, to help faculty design information and to use it in their teaching and research, and to teach students how to access it and to create new information resources. It is built to develop and support a digitalcollection, but it does not replace our print collections; new technology does not replace print, it does things that print can not do, and our intent in the new building is to design and build the electronic library.
In calling the electronic library the global reference room to place emphasis upon the problem of navigation, finding the right information, and information literacy, knowing how to evaluate information and use it wisely. The idea of navigation reflects a fundamental change in the ecology of knowledge. While knowledge has been scarce, and as such required centralisation and conservation, today we live in a world saturated in information, whether print, television or computer generated data and records. Our distinctive information problem is selection, as much as conservation. When I ask faculty researchers which aspect of the electronic library they like the most, they say it is the abstracts of articles in their field; when I ask why abstracts are so valuable, they tell me that abstracts help them make judgments about what they don't have to read.
If information literacy implies both selectivity, finding and making judgments about the quality of information, it also implies knowing how to make use of information. Digital environments are not only read (although reading is the essence of a computer interface), they are softwar environments for analysing information, for changing the representation of information (e.g., numerical data into graphic form), and ultimately for publishing information on the network. In other words, in digital environments there is no distinction between reading, writing, and publishing. For this reason, copyright will be fundamentally changed by digital networks.
In thinking about the design of digital formats, it would be a mistake to focus solely on print and multimedia, for the network is at heart a communications medium, and part of the design of digital collections will enable communication among scholars and students, even if they are geographically separated by vast distances. For example, I've just seen a workstation with TV camera built into it, and the screen interface contains a window with full motion video; using this technology I imagine shared televised reference services among libraries across the Pacific, or to Great Britain, or wherever there is subject expertise. Technologically there is no reason we cannot we begin using the network for document delivery, for access to catalogues, for digital reference; we can begin to act as if we are one library. That strategic thought is the essence of the global reference room.
The network is a technology for shared collections: print materials can be scanned, placed in a database, sent to any destination in the world, and printed on demand. The social and economic implications of this technology, however, have not been demonstrated: we do not know how to manage copyright well, or to charge appropriate prices; digital environments have not proved stable for long term conservation of information, for technical standards and technologies evolve quickly. And, the network is far from a ubiquitous environment, connected to schools, libraries and businesses.
But technology also suggests new kinds of collections as well. New digital arts, technologies for scientific visualisation, and electronic journals each represent new kinds of digital collections. Thus far there has been little work on the design of digital books, other than the generic (and redundant) term "multimedia," and the example provided by "hypertext" links. The book and the journal formats have been developed through centuries of experimentation. Appropriate rhetorical forms have been developed for different domains of knowledge and print formats, but there has been little attention to the design of formats for "multimedia" knowledge. Today it seems as if the entertainment industry, not education or science, will create multimedia designs. As a distinguished computer scientist recently told me, "the best minds in the country are working on the design of games."
The network has already changed scholarly communication, understood as the full process of research and writing which leads up to publication, in ways that inevitably will change print publication. In many scientific and technical fields scholarly publishing is less important than access to preprints on the network. If you have to wait for something to be published in a physics journal you are not on the leading edge. The same thing is true in genetics, and many of the electronic journals, and this is a very far reaching trend. Thus scientists and scholarly associations in the US are not thinking about document delivery as much as replacing print with networked forms of scholarly communication. Almost every disciplinary society I know of is thinking about electronic journals, ranging from physics and chemistry to political science, sociology and psychology; the most imaginative electronic journals are often in new interdisciplinary fields. When I ask the faculty if the printed journal is still important, they tell me it's important, but not as a means of research or teaching, it's important for tenure and promotion.
Moreover, it is not clear how print on demand technology will affect printed knowledge. At USC we did a study with McGraw Hill putting some of their textbooks in a database and letting faculty and students select material and print textbooks on demand, that is, edit their own textbooks ("the Primus" project). We discovered two things that were very surprising. First we could not predict the information people would use on the basis of their discipline; everyone reads in an interdisciplinary manner. Secondly we assumed that the unit of knowledge would be the article or the chapter, but the unit of knowledge turned out to be the paragraph. This horrified the McGraw Hill writers, because they had thought they were writing in a narrative form in which knowledge would be cumulative; if you are going to read Chapter Three then you have read Chapters One and Two. But users treated narrative argument as if it were a database, selecting paragraphs without context. Several authors were not willing to have their writing taken out of context. In the online text, the reader designs custom information, selecting it, editing it. We can no longer describe them as readers because in an important sense they are authors. But this raises a serious issue: higher education and the scholarly disciplines are organised around the concept of a literature, which is to say in an authoritative, cumulative version of knowledge; but that is not the way people are using knowledge online.
In a second experiment with the Chronicle of Higher Education, a weekly newspaper on higher education was provided in both print and online forms. We discovered, by the way, that the network version did not "replace print." Everyone read both, but for different purposes: from a print journal they wanted to get a cross section of important news in the field; the online environment was a reference environment to answer specific questions.
I suspect that multimedia will lead to entirely new conceptions of the book. Gary Seaman, a professor of anthropology at USC is a curator of a wonderful archive created by Jane Goodall, of films, photographs, and field notes on the nature of chimpanzees. Gary has written a multimedia software program called The Virtual Chimp, in which students can observe chimps just the way scientist would do in the field, by watching video and interpreting how chimps communicate with each other. It is an aesthetic education, in the sense that it is training the eye to recognise and interpret communication. Thus it has a certain resemblance to a game, and is more successful in teaching students about primate communication than the printed book. Most important, the game is just the front end of a database that includes in principle all of the information, it is not just simulation or substitute for real data, it is a doorway through which a student, having mastered the preliminary material, can go on to explore the real archive.
If The Virtual Chimp is a new model of a textbook, Who Made America? is an enhancement of a print textbook in the multimedia version (published by Voyager). It is American labour history in which a photograph in the print textbook is interactive in the multimedia version. If you see a picture of William Jennings Brian you can click on it and hear the Cross of Gold speech. I am a political scientist and I have been reading about the Cross of Gold speech and the American progressives most of my life, and I was thrilled to actually hear it. Even for some of the teachers in the field multimedia has brought a new way of experiencing the material.
Finally, I hope it is obvious that thinking about the electronic library as the global reference room is a tacit invitation to think about collaboration among the libraries in Great Britain and North America, and beyond.