/ The Electronic Librarian Is a Verb / The Electronic Library Is Not a Sentence

A Lecture Delivered at the New York Public Library The Gilbert A. Cam Memorial Lecture Series October 14, 1994 by Kenneth Arnold

The computer game MYST begins with a figure falling through space. During the fall, this person loses a book, which disintegrates into multiple pages that become lost on an island below. The player of the game subsequently becomes, it seems, the fallen figure, who searches the island for the pages of the book. Unlike other computer games, the player is also the point of view: we see but are not seen. We manipulate a pointing finger on the screen to move. We can turn to the left or right—turn around, if we wish—to see a new landscape. We can climb stairs, enter buildings. The graphics are excellent, the environment eerie. Our search is accompanied by sounds and strange music.

The game comes with no instructions and only three "hints," none of which tells the player that the object of the game is to recover the pages of the book—to reconstitute the lost information the book contained. Playing the game requires that we know how to work a computer and a mouse, that we understand what it means to enter a computer scape—or perhaps we should say cyberspace. In this digital land, we can reach out and grasp the page of a book and turn it. We can pick up things. The game differs from other computer games in not relying on warfare—dodging evil creatures or eluding monster traps. There are no weapons. No one dies. And yet young people who have been reared on fantasy computer games—the children of Nintendo—like MYST.

I bought the game for my two children, who are nearly fifteen. I had never played it myself. After they had experimented with it for a couple of weeks, I asked them what they thought. My son said he loved it. (He is a fan of some of the more gruesome combat games, by the way.) My daughter said she liked it but was confused. "What am I supposed to do," she asked. My son replied, "Find the pages of the book." "What book," she demanded. There ensued a long discussion about what my son knew and how he had learned it.

He had learned what to do in two ways. First, he was experienced with computer games and understood intuitively that he was on a quest. The first building encountered in the quest is a library. In the library were, not surprisingly, books. He learned that he could remove books from the shelves and open them. Inside was information, some of it useful. He also experimented, learning what to do by trial and error. The second way he learned was by asking questions of friends who had played the game.

I asked him what interested him about MYST since it lacked combat and danger. He said, "It's a puzzle. I like puzzles." I then loaded the game myself, figuring I would be able to manage pretty well, since I had been given more information to start than most. I was wrong. I could not navigate the landscape, even though I am proficient with computers and a mouse. I couldn't even find the library. I got as frustrated as my daughter did.

Unfortunately, this is the way many people feel about the electronic environment. They do not know where they are, where they are going, or why they are there. It's a puzzle whose puzzle master is part of the puzzle.

We learn how to navigate any world by trial and error, by asking questions, by imitating others. For the last five hundred years or so, we have also relied on printed information as essential navigational tools. But not all information is of equal value. Not all information can be trusted. The invention of printing led Angelo Poliziano, a fifteenth-century Florentine humanist, to scoff: "The most stupid ideas can now in a moment be transferred into a thousand volumes and spread abroad." (Mellon, p.103) And indeed many multiple thousands of volumes later we find ourselves swamped with information. Libraries are filled to overflowing with the fruits of the last information revolution.

It seems that there has always been too much information. Even before Gutenberg printed his first Bible, there were concerns about the problem. As early as 1439 the "chancellor of the University of Paris, Jean Gerson, complained that the boom in book production was dangerous. It was giving rise to theological confusion and shaking the solidity of the church's traditional teaching." (O'Donnell, Symp 93, p. 23) As James O'Donnell, scholar and editor of an online journal in classics, notes, "fear and fascination flourish in the presence of abundance of information." Libraries were established to manage and help readers navigate this abundance easily. The dream of organizing all information in what we would now call a virtual library begins with the library at Alexandria, which was intended to gather together all of the books in the world in one collection. Then, "books" were rolled scrolls stored in pigeonholes.

The codex book, which is our standard form of presenting information, has proved to be an effective and durable technology. The shape lends itself naturally to linear organization on shelves. And because most books are about one thing it is natural to group them according to subject similarities. More elaborate classification systems developed by librarians have helped create a user-friendly way of presenting books. The storage of objects in the traditional library is experiential. We expect to be able to lay hands on the book we want. We expect it to be there.

The book itself is a linear operating system. We read it in one direction, from beginning to end, unless it is a book of reference. But even reference books are organized in a linear, usually alphabetic, order. We use numerical systems to classify books. We impose order on collections and tell people what that order is so that they can find what they want. This all makes good sense. In light of the current excitement about the digital library, one might well ask why anyone would want to mess with the existing arrangement. Many in fact ask exactly that question. There may be too much information, but the system is not broken. Why fix it.

There are problems with the present arrangement. The infrastructure for the production and storage of books and journals that has evolved over time is cumbersome and expensive. Finding information is not as easy as we might hope. If one does not live near a good library—and most people in the world do not—then the problem is extremely serious. Even if one does go to a library likely to have what one wants, the book may be in the hands of someone else. One book can only be used by one person. (That is not true of a digital object, by the way.) Libraries are not always open; in fact, with budget reductions many libraries are infrequently open. The publication system itself is wasteful and haphazard. I know. I've been there. The buying habits of libraries are not exactly rational either.

The problem with the present system, or any system, is access, exacerbated by the sheer quantity of stuff. A friend of mine, a librarian, works in his spare time these days as a consultant to libraries that are trying to decide what to get rid of. Is it really sensible for institutions to build new buildings, or expand existing structures, to house more books when such a large percentage of what is in the library is not being used by anyone? The "just in case" model for the purchase of materials—that is, a good library should have on hand anything some reader might one day want—has created a problem in the simple storage of information. The so-called serials crisis of the 1980s, which led libraries to buy fewer books, added another dimension to the problem. Resources have been reallocated from books to journals, making it increasingly difficult for libraries to maintain representative collections. This situation has also created a crisis in the publishing industry, especially among academic publishers.

Universities and the present system of managing scholarly communication can be faulted for encouraging the production of journals, books, or monographs with limited markets designed primarily to confer academic preferment—tenure and promotion. The system feeds itself. If anything the trade publishing market is worse, spewing out thousands of books that exist only to produce income for publisher and author.

The question is: Will digitized information and the computer make the problems go away? The answer is: yes and no.

There is always too much information, and people have always (so far as we know) worried about it. Think for a moment about filing cabinets. We all use them. They provide a sensible way of storing papers in a retrievable order, often alphabetically and hierarchically. Have there always been filing cabinets? No. "It is difficult today to realize that the common, everyday office filing cabinet represented a major revolution in handling information when it was first introduced in the early twentieth century. It allowed large amounts of material to be organized in an efficient, well-structured manner." The vertical filing cabinet wasn't practical, however, until the problem of copying documents was solved. (Norman, Things) Remember carbon paper?

The fact that we can digitize information does not necessarily mean that it is useful to do so. What is the functional equivalent of the problem of copying documents in today's digital environment? Well, the situation is reversed. We have the materials, but we don't know where to put them. The problem then is creating what I call knowledge structures that will allow us to use digital information. The elements are in place to create such structures—computers, searching tools, and librarians.

Do you remember your first Personal Computer? I bought an IBM PC back in 1983. It had two floppy disk drives and 256 K of memory. It arrived in my office. I could think of nothing to do with it. I had no word processing software yet and was unable to program. In the operating manual I discovered that I could create a simple program that would produce a rising musical tone. I did it, and everyone in the office came running in to exclaim over my cleverness. We stood there looking in awe at this machine, it seemed, for days. I used it for awhile as a typewriter and then, when I had a spreadsheet program, as a calculator. I did what we usually do when we acquire new tools: we do what we have always done. It takes a while to figure out what it is they help us to do that is new.

It took Lisa, Apple's first computer, to begin to show us what might be possible. John Sculley, who came from Pepsico to run Apple, called it a "third wave" company, one that was "flexibly networked" and focused on interdependency, individual entrepreneurship, and growth. The differences between such a company and Pepsico, "transposed into terms humanists would find familiar, might be thought of as the differences between 'modernism' and 'postmodernism.'" (The Electronic Word, p. 681) This approach to the corporate environment was translated into the product, which (unlike the IBM) presented itself as a tool for doing something different—and fun. It is no accident that there were from the beginning more games and educational programs for the Apple than there were for the IBM, which was marketed primarily as a serious business tool.

The personal computer in fact was far more than a typewriter with memory or a super-efficient calculator (although VisiCalc was for many the killer app that made them believe the computer could do something useful). It began to show us how we might reimagine the world of information. Coupled with the development of network communications, the personal computer expands the mind's reach immeasurably. How is this so?

Benjamin Woolley, in an interesting book called Virtual Worlds, describes a computer as a "virtual" machine, "an abstract entity or process that has found physical expression, that has been 'realized.' It is a simulation, only not necessarily a simulation of anything actual." Woolley goes on:

Personal computer users generally become comfortable with the idea of the system being at once a word processor, a calculator, a drawing pad, a reference library, a spelling checker. If they pulled their system apart, or the disks that contain the software, they would find no sign of any of these things....They are purely abstract entities, in being independent of any particular physical embodiment, but real nonetheless. (Woolley, 1992, p.69)

The computer produces a representational reality, as opposed to the experiential reality of the codex book. It is the smile of the Cheshire cat. But this is not magic. Computers are machines that can do only one thing well: mathematical computation. As Woolley points out, however, that is quite a lot, since "there seems to be an underlying mathematical structure to everything...." (p. 70)

Practically speaking, we think of computers as being about the management of information. But what do we mean by that word, "information"? When publishers talk about the work they do, as opposed to the work librarians do, they insist on relegating to librarians the management of information and retaining to themselves the creation of works that are bigger, grander than information. Publishers, and others, worry that the computer will reduce everything to information and eliminate the quality of judgment and analysis that makes knowledge. Information is certainly not equivalent to thinking. In Mind Tools, a book about the mathematics of information, Rudy Rucker writes: "The concept of information currently resists any really precise definition. Relative to information we are in a condition something like the condition of 17th century scientists regarding energy. We know there is an important concept here, a concept with many manifestations, but we do not yet know how to talk about it in exactly the right way." (Rucker, 1988, p. 26)

What we worry about was expressed by Martin Heidegger in 1967:

"Maybe history and tradition will fit smoothly into the information retrieval systems which will serve as resource for the inevitable planning needs of a cybernetically organized mankind. The question is whether thinking too will end in the business of information processing." (Heim, 1003, p.64)

Doing more than processing information is the challenge of creating appropriate knowledge structures in the digital world. The personal computer, which is central to that world, can do more than store and produce information. Certainly, my work on a word processor is more complex than that. Word processing is an example of what I mean by a knowledge structure in this environment, doing a familiar task in a new way. The philosopher Michael Heim has written provocatively about word processing. He says this:

Word processing makes thoughts flow more directly....The eye is wired to the brain via the computer, making a feedback loop between the mind and the written word. Word processing is not just a quantitative improvement in getting a job done more efficiently.... We are now interfacing with our own thoughts. Software transforms knowledge beyond the limits of printed writing. Word processing leads to the more fundamental activity of thought processing. (pp. 42-43)

From this perspective, technology can fundamentally transform the way we know and think.

As a Research Grants brochure issued by the Council on Library Resources notes, however: "Despite many claims and assertions, the information structure of the future has not yet taken shape." The danger is that "the pace of change is such that it is imperative that 'architects' of great skill, who are concerned with the well-being of universities, scholarship, and libraries, go to work with some sense of coordination before a structure is imposed by default." (The Electronic Word, pp. 489-490)

Enter the librarians, who seem precisely suited to shape this information or knowledge structure—and, not surprisingly, many of them are addressing these questions, often in league with publishers and computer programmers. I have described their role in the title of this lecture in grammatical terms—the librarian as a verb—because what we are talking about is in fact creating a syntax of digital knowledge. That is certainly more palatable than "information processing." Given the computer's capability to connect us to a seemingly unlimited array of subjects and objects in the digital language, how should we manage that capability most effectively?

Donald Norman, in his book, Things That Make Us Smart, has this to say:

The new-fashioned information artifacts take on arbitrary shape and form. There is no natural mapping, no natural principles of operation. The critical operations all take place invisibly through internal representations. If we are to be able to use these artifacts easily and efficiently, the designers have to provide us with assistance, with an understandable, coherent structure. We are in the hands of the designers, who have the power to make the artifact meaningful, to provide substance and richness, and to make its use support the activities of interest. The best of the artifacts will become invisible, fitting the task so perfectly that they merge with it. They will be a delight to use. (Things That Make Us Smart, p. 230)

Grammar imposes order, but it develops within and organically from the use of language. No one forces grammer on communication but the users. We speak naturally and fluidly, seldom stopping to think about the agreement of subject and verb. We communicate highly complex ideas or instructions according to a commonly accepted set of rules that help us make sense of our otherwise chaotic thought processes.

Reflect for a moment on the way you think. Your mind rambles. Your attention is wandering even now. You remember something you were supposed to do this morning and wonder if you will have time after I finish to do it. Before you know it, you are thinking about someone you knew in high school whom you haven't seen in years. But, amazingly, you are probably still taking in some or a lot of what I say. Your thoughts are not linear.

Nor are thought processes in cyberspace.

How will librarians work in a complex nonlinear environment? Let's take advantage of the electronic library residing on my powerbook and consult Richard Lanham, who in his book The Electronic Word, has interesting things to say about the organizing of knowledge in the digital landscape. For example:

The library world feels depaysé today, and rightly so. Both of its physical entities, the buildings and the books they contain, can no longer form the basis for planning. And the curatorial function has metamorphosed, to borrow a phrase from an archivist acquaintance, "from curatorial to interpretive." Librarians of electronic information find their job now a radically rhetorical one—they must consciously construct human attention-structures rather than assemble a collection of books according to commonly accepted rules. They have, perhaps unwillingly, found themselves transported from the ancillary margin of the human sciences to their center. (The Electronic Word, pp. 485-86)

What captures my attention in this paragraph is that phrase, "attention-structures." It is one Lanham uses frequently in his book, as in, "The construction and allocation of attention-structures will be a vital activity in our information society," (p. 792) The use of the phrase needs to be heard in the context of his general argument about the emerging information system as one not unlike the ancient Greek paideia, which emphasized the primacy of rhetoric as a mode of education and a syntax for communicating ideas.

We may say that rhetoric—considered as an information system that functions economically, that allocates emphasis and attention—resembles what is now called, in many fields, a nonlinear system. That is, it is dynamic rather than static, a constantly changing emergence rather than a fixed entity; global rather than specialized into disciplines and constituent parts; a system that seeks to describe the confusion of everyday experience rather than narrowing it into a delimited, predictable field of study. (Lanham, p. 235-36)

The new syntax of knowledge that operates in the digital environment is not linear, proceeding from subject to verb to object, with pleasant stops along the way to admire the adjectives or calm the adverbs. It is interactive, composed of feedback loops. This new syntax is driven by verbs, actions taken between attention-structures. There may be no subject or object. One may be working the adjectives only, or manipulating dependent but isolated clauses. Digression is the normal mode of discourse rather than the order of the diagrammed sentence I remember from my days in what was once called grammar school. In a dyamic system, one does not diagram sentences. James Joyce wrote in this new mode in Finnegans Wake, which begins and ends in the middle a sentence—the same sentence—rendering the entire book a mammoth feedback loop encompassing all knowledge. His novel is a virtual library and a radical syntax that anticipates the mind of the computer and encodes the omniscience of a diety. The utopic Xanadu, a hypertextual world of completely accessible knowledge, embodies a similar but unrealized concept. (It was imagined, incidentally, by Ted Nelson, who created hypertext.)

Late in June, I visited the British Library to meet with Robin Alston, a lecturer in library science at University College London and advisor to the British Library on matters electronic. Knowledge, he told me, is not amenable to mechanical transmission, although information certainly is.

Technology can promote the formation of new knowledge...but only with the help of considerable interpretive skills, skills which librarians must learn to develop....The survival of the institutions we call libraries will depend, as always, on enlightened and imaginative librarians able to develop within a hostile political environment a model which can adapt to the evolving needs of research in all disciplines.... The notion that knowledge can prosper by creating vast knowledge warehouses based on the hypermarket model—you can buy it if you can find it—is sheer fantasy as well as being intellectually suspect. (Alston, pp. 6, 8, 13)

What concerns Alston and many of us is the prospect of anarchy resulting from the destruction of one syntax before the new one is understood by anyone. Xanadu cannot be created yet because it would be nothing more than a warehouse.

What was wrong with the IBM PC? It was not designed for the people who were going to use it. Technology fails when it is designed for the makers. We also know that the initial purpose of an effective new technology may quickly be altered in use. It will outgrow its creators. A new technology should be transformative. As we think about knowledge structures for the library, we need to keep in mind that nothing will work that is not designed for the users. The focus of library information systems has to be on people. How do people do their work? How do they look for information? What do they want from a library? How can librarians help them do what they want to do? Norman points out that "Where we expect to find something may depend upon what use we intend to make of it. Intend one use, think of it in one way; intend another, think of it in quite a different way. How we think about something determines how we look for it." (Norman, 1994, p. 389) This is not good news for people who want to find ways to impose order on the emerging chaotic universe of cyberspace. The problem gets worse. Norman again:

The way we think of something today is apt to be very different from the way we think of it later, for as we learn new things, new organizational structures will arise. Couple this to the fact that organizational structures for a society must last decades or centuries, and you see that whole new classifications of knowledge may arise....One need only examine the classification schemes used by modern libraries to see these problems, to see how badly the scheme works for rapidly changing fields. (Things That Make Us Smart, p. 389-90)

What's a librarian to do?

I want to talk about what some libraries are already doing, including this library, but first, I want to say that the situation is much worse for publishers, most of whom lack the technological resources of libraries and are faced with the serious problem of controlling information for economic gain in a system that yearns to be free. I have worked exclusively as a publisher and realized some time ago that the roles of the players in the knowledge system are merging. In the not-too-distant future there will be a reconfiguration of the production and dissemination of knowledge. One aspect of this paradigm shift will be the development of cooperative structures and a shift in the economic model for the production of information that will change the way we view the management of what we now call intellectual property. The concept of ownership itself will change. Librarians are already deeply involved in attempting to understand what this means, partly through the Association of Research Libraries, and partly through the work of specific university libraries around the country. These are questions in part of institutional relations, resource allocation, and academic politics. As it always has, the system will need producers of knowledge, professionals who know how to format and add value to what authors create, and experts in the organization and delivery of attention-structures. The problem is that in some instances one person or institution may now perform all of these functions at once. The individual may write, publish, archive, and disseminate from a single personal computer.

I have a lot to say about these evolving relationships but not enough time today to pursue that thread. In focusing now on the librarian, I am acknowledging a fact of electronic life. Because the libraries created the bibliographic systems that provide information about information, and because the libraries first saw the research value of computer-organized attention-structures, the librarians have been first to identify the problems and address them publicly. Fortunately, they have also learned that what is wanted in the end is not a Librocracy, as one of my friends put it, although some key leaders still believe—incorrectly, I think—that we might well manage in the future without publishers, copyright, and revenue. I do believe, or have come to believe, that we will have to learn how to manage information systems before we can figure out how to publish in them, and that means that the libraries need to hold the lead in setting the standards. Some of my publishing friends will call this heresy. But in fact libraries have always set the standards.

There has been much talk about the digital or virtual library, creating extensive resources in machine-readable form accessible from multiple locations. A lot of progress has been made in developing the tools for managing such a transformation. Money has been allocated for the digitizing of texts, and this work is going forward rapidly. Most of you know about these efforts. I will describe some of them briefly as a way of highlighting emerging knowledge structures.

The development of online journals has been a primary focus of many experiments because it has been assumed that journals can more easily be managed in the online environment—and because they precipitated the crisis in print media. Johns Hopkins University Press and the Milton Eisenhower Library have produced a demonstration model of an online journal delivery system, called MUSE, that is extremely attractive. Although the journals being converted to digital form were published by the press, the library has taken the technological lead in developing the software presentation of the texts. I was shown the system by a Todd Kelley, a programmer/librarian, who spoke about other digital products he hoped to create. The purpose of this work, he told me, was to extend the educational reach of the library beyond the walls of the institution. In this instance, the library is functioning not as a publisher but, in cooperation with a publisher, as an educator and disseminator of information in a new form. Todd represents a new kind of librarian.

When any new technology comes along, we tend to convert old forms to the new system, attempting to reinvent the old in the guise of the new. This effort usually fails. For the most part experiments in digitizing and delivering electronic text are focusing on the relatively straight-forward transformation of print text into digitized text. Large-scale digitizing projects, such as those being funded by the National Science Foundation, will focus initially on such direct translations. This effort is in itself worthwhile: abundance in the electronic world adds value, just as scarcity does in the experiential world. If I capture a digital text, it is still there for someone else to use.

Recently, the Library of Congress announced a massive digitalization project that will convert the most important materials in its collection into a virtual library. "Our goal," announced Susanne Thorin, chief of staff at the library, "is to bring these resources out to people across the country and not just to the people who can come to use the library in Washington." (Lewis, 9/12/94, B11) This project will cost millions of dollars for equipment and conversion operations, money which has yet to be raised. The library hopes to convert its most important materials (whatever that means) by the year 2000. The library plans to take the lead in coordinating technologies and policies for all digital libraries so that they can be connected to the same networks. It is not clear what such an effort means for libraries such as this one, nor is it clear what such massive digitalization will mean for publishers and the management of intellectual property. It is still an important step toward creating the necessary basic resource.

The University of Michigan digital library initiative is one with which the New York Public Library is associated. Michigan has received one of the NSF grants. How these digitized resources will be managed, however, is still an open question. Douglas Van Houweling, Vice Provost for Information Technology, describes the emerging digital university as an information node, functioning almost as an imprint to manage and ensure the quality of scholarship in a dynamic mode. Organizing information in itself adds value. The university does not lose control of its intellectual property in this knowledge structure, as it does when scholarship is distributed in static media (ie, books and journals). Clearly, the distinctions between library and publisher and author begin to blur in the model Van Houweling describes.

The University of Minnesota developed the Gopher search technology—actually computer science students did—and continues to see its mission as the provision of free information. This vision differs from that of Michigan in emphasizing the service role of a public university. Theoretically, and in fact, individuals on the campus can set up their own Gophers and make any information available to anyone. Issues of archiving and ensuring the quality of information supplied by the university or under university auspices still have to be addressed. There is no Minnesota plan to digitize information on a large scale. The librarians I spoke with see ways in which they might package information services for specific users, some of which may have to be developed on a fee basis, but in practice right now the university and the library offer a tool, the Gopher, and stand aside while information freely flows.

The JANUS project, located at Columbia University, on the other hand, is conceived as a large-scale digitizing initiative that will create a university-based foundation for new ways of teaching and doing research. Willem Scholten, who manages JANUS, views the university's body of information, its library and research resources, as a living entity. The library's responsibility is the long-term archival storage of electronic data, but Scholten also envisages collaborations between libraries and publishers in which the libraries distribute and even reconfigure information for specific needs. The digital library in this environment is a twenty-four-hour a day resource that becomes the heart of the university research and education enterprise. JANUS is also working with the Seattle Public Library, which offers internet access to Columbia's resources for its patrons. This project is still very much in progress but on the face of it promises to become an important model for unique knowledge structures whose architecture is flexible, interactive, and open.

Other universities, such as Tennessee, Case Western Reserve, Southern Illinois, Nebraska, and Virginia are developing text-based digital services, often in cooperation with their university presses and computing centers. The University of Virginia's Electronic Academic Village is one of the more advanced environments in which scholars have remote access to digital texts and images. The library has taken the lead in creating this multi-faceted digital environment.

The University of California's SCAN (Scholarship from California on the Net) is a pilot project that will disseminate electronic journals and monographs. The California initiative is also collaborative, involving the library and the university press. Library service organizations such as the Research Libraries Group and the Online Computer Library Center have developed sophisticated bibliographic search engines and document delivery systems, many of which are available here. And both organizations are exploring ways to enhance their online services.

This quick tour of the electronic landscape leaves out many important experiments, including those in other parts of the world. The creation of digital resources is going forward on a grand scale, and it seems clear that the process will continue, particularly as the US Government and private foundations invest in digitalization projects.

Libraries are already working in an electronic environment. If we look at the New York Public Library, we see similar initiatives and the development of enhanced electronic resources. In August and September I spent some time with members of the staff of the Humanities and Research Divisions discussing the work they are doing in the electronic environment—and how that environment is changing the way they think. The Digital Imaging Project, which is being undertaken with the Research Libraries Group and eight other libraries, is producing an interactive visual resource that will allow researchers to locate images and preview them on screen, searching easily by subject or photographer. This pilot project demonstrates one important way in which computers can be used by libraries to make resources available quickly and conveniently across collections, while protecting the physical objects themselves. It also demonstrates how librarians are rethinking their own roles, designing, for example, attractive interfaces that anticipate the needs of potential users.

Rodney Phillips and his staff are developing for the Center for the Humanities an exciting array of electronic options, including a "democratic lab" for electronic scholarship and discovery, an electronic text center, enhanced catalog research tools, Internet access for library patrons, electronic access to a wide range of online resources and texts, and access to a variety of hypertext and interactive resources on fixed electronic media. Many of these projects are still in the planning or early development stages. All of them are based on the premise that the library should supply patrons with expanded access to resources beyond the library itself. The library also plans to support these electronic resources with on-site educational and navigational tools for patrons who need to learn how to use them effectively. These electronic resources may also become the basis for the creation of products based in the library's collections—publications, if you will— that further extend the reach of the institution into the community. This is the work of creating new knowledge structures, and it requires that librarians be activists.

We can also recognize in these various projects the traditional librarian at work, managing and organizing information resources. In most cases, as I have noted, this work involves the relatively straight-forward conversion of resources from print to electronic form. A digital object is like a physical artifact in many ways. But there is an obvious shift in emphasis from the object containing information—that is, the book—to the content itself. Digitized objects will be linked to other objects, as they are now in the World Wide Web, allowing users to jump from one resource to another across the traditional boundaries of the book. The library's Gopher server and developing Web Server will enable the creation of such hyperlinks. The medium allows us to do more, and that more, through the creative application of these new tools, can be the source of enhanced understanding, encouraging interactive learning, conversations with others across space and time, and dynamic encounters with information.

In my conversations with members of the staff here, we talked about the role of librarians as navigators, professionals who, as Ewa Jankowska remarked, "help define for readers what they want." Librarians still need to provide context for patrons. Perhaps most importantly, libraries need to use these new resources to expand access. The public library performs a democratizing function, and electronic resources can actually support that role by opening up avenues of information formerly closed. Connections between branch libraries and central collections are being enriched electronically. Public workstations can open a world of virtual resources to anyone—digital maps at the University of Minnesota's exemplary collection, medieval texts at the University of Virginia, photographs in the Public Library's excellent collection. The librarian's role is to structure access to these resources using the new electronic toolkit—and then teach library patrons how to use those toolkits themselves to navigate freely and privately. As someone else provocatively noted, the public library creates an invaluable private space, which now paradoxically has no boundaries.

The public library faces special challenges in this new environment because it is not part of a university. The library will have to market its services to a diverse clientele, providing both research resources for scholars and public service to the average citizen. Public relations and education lie uneasily together in this environment, and money is always a problem. How are institutions to fund conversion to electronic formats? What kinds of choices must be made? Who will guide the decisions? And for whom are the decisions being made?

The library of the future is an interactive nexus of global information. What is clear from the projects I have described today is that collaborative enterprises will be essential to the invention of the new library. No one institution can manage to become the virtual library. The very concept implies cooperation. Librarians will assist in navigating across disciplinary bounaries. To do that they will need to break out of rigid organizational structures and begin to look across existing barriers. Internally, that may mean a reconfiguration of management. If one department manages to secure the financing to develop new resources, what does that mean for the department that cannot find funding? If one library or collection is digitized but another is not, what does that mean for the integrity of the entire system? Some information will not be available as easily. Should the decision about access be determined by economics? How might existing print media be incorporated into digital search mechanisms?

It is likely that libraries will have to rethink collection strategies. Acquisition budgets may decrease in favor of access technologies. Librarians schooled in acquisitions may become more skilled in creating dynamic links to resources in other locations. Such a shift requires reallocation of personnel as well as financial resources. At the same time, the preservation of existing resources will become important as some libraries develop expertise in managing certain collections not only for their traditional constituency but for the larger digital environment. New delivery options will have to be devised.

Librarians in this digital universe may well become what one publisher has termed links editors, people who specialize in identifying and establishing linkages with other information structures, creating pathways through the virtual library. They will need to become experts in predicting search patterns. How will they do this?

The library we are describing will focus its attention on the needs of its patrons. Librarians are information retailers and in this new consumer market will have to find new ways of providing individualized service. The organization of information for the patron requires the deployment of tools that teach people to navigate resources seamlessly. Increasingly, public support for the library will be based on a perception of the value added by the librarian to the information in the entire digital world. Publishers have always known they have to package information. The publisher finds ways to make a manuscript more attractive. The librarian will find ways to make services more attractive. The librarian will have to create demand for new services in the electronic environment.

The computer is a tool. The electronic networks connect computers to one another. Various search mechanisms have been created to help us find our way through the networks to information stored in computers or managed by them. But these are only tools. As we think of ways to use these new systems, we need to remember that we may use these tools in any way we want. It is up to us to configure information and provide pointers that people can use. It is not necessary for every library to invest millions of dollars in converting to electronic texts. It may be more important to invest in developing sophisticated access systems and new services. We should not allow the technology to drive our decisions.

The librarian as a verb is a useful image, I think, because it implies movement. In the past the librarian, unlike most professionals, has been associated with a place, the library, a building. In the future, the librarian will be a vector, searching for and establishing connections. The library in which this librarian works is more a state of mind than a location. It is a set of neural connectors. And that is why the place called library is not the sentence in which the electronic librarian dwells.

What such a vision means for the organization of knowledge is for some people theatening. A collapse of traditional roles always creates uneasiness. People lose control of their domains and get nervous. Why should this be a problem in the library world? Knowledge is indeed power. And at the moment librarians have the tools to reconfigure the knowledge environment, which inevitably will affect the ways in which scholars, publishers, and even politicians do their work. Right now, the revolution in electronic media is going on in two realms. There is the realm of the public corporations, mostly broadcast media, where the development of systems for the delivery of entertainment is paramount (the pun is intended). The other realm is this one, the environment in which information and education can come together to form a new kind of knowing. The truth is that educators are in a better position right now to take advantage of the emerging technologies. The research and library community can seize the initiative—and in many ways has seized it—to establish an arena for the responsible use of this powerful new technology.

Public libraries and other public institutions will need to lobby vigorously for the funds and public support necessary to hold on to this competitive edge. Information may be free, but delivery systems are expensive. They always have been. Look at this building, for example. It will not be long before the electronic revolution is corrupted by the broadcast media. Libraries and education institutions are not in the broadcast business. They are narrowcasters, focusing their talents on the delivery of services to small groups, to individuals. This is an empowering function, and even in a democracy to empower individuals is to create uneasiness. Finding the money to manage the paradigm shift in the delivery of knowledge in a broad, democratically based system will require political skill. It will be difficult for institutions such as the Public Library to take advantage of the revolution now under way. Changing and then managing information and knowledge systems inevitably challenges existing power structures. Make no mistake about that.

What always strikes me forcefully when I come into this institution is the openness of it. There are people everywhere. It seems chaotic, random, gritty. Oddly enough the new electronic environment allows us to continue to be chaotic and random and gritty. We can design systems that work the way people work, that think as they think. We can hold open the doors to understanding. Librarians are in a unique position to articulate a vision of open access and creative thinking. As advocates for change within the research and library community, they have been effective. Gradually, they are building a new kind of knowledge structure and creating a new kind of institution. This is rare in history, to make something new that will change fundamentally the way people behave.

The challenge to the library is to reflect but, more urgently, shape the expressive culture of our time.

Kenneth Arnold is founder and president of New Century Communications, a publishing and communications consulting firm in New York City. He has worked for twenty-eight years in academic publishing, most recently as Director of Rutgers University Press. Comments and responses are welcome and may be forwarded to the author by email at century@panix.com or by U.S. mail:

NewCentury Communications
275 W 96 Street #27C
New York, NY 10025


Alston, R.C. "The Battle of Books: Inaugural Lecture," London: University College, 1993.

Cummings, Anthony M., Marcia L. Witte, William G. Bowen, Laura O. Lazarus, and Richard H. Ekman, University Libraries and Scholarly Communication: A Study Prepared for The Andrew W. Mellon Foundation. Washington, DC: The Assocation of Research Libraries, 1992.

Heim, Michael. The Metaphysics of Virtual Reality. New York: Oxford University Press, 1993.

Lanham, Richard A. The Electronic Word: Democracy, Technology, and the Arts. An Expanded Book. Chicago: The University of Chicago Press, 1993.

Lewis, Peter H. "Library of Congress Offers to Feed the Data Highway," in New York Times, September 12, 1994, p. B11.

Norman, Donald A. Things That Make Us Smart: Defending Human Attributes in the Age of the Machine. An Expanded Book. New York: The Voyager Company, 1994.

O'Donnell, James J. "The Virtual Library: An Idea Whose Time Has Passed," in Okerson and Mogge, eds., Gateways, Gatekeepers, and Roles in the Information Omniverse. Washington, DC: Association of Research Libraries, 1994, pp. 19-31.

Okerson, Ann, and Dru Mogge, eds. Gateways, Gatekeepers, and Roles in the Information Omniverse: Proceedings of the Third Symposium. Washington, DC: Association of Research Libraries, 1994.

Rucker, Rudy. Mind Tools. London: Penguin Books, 1988.

Woolley, Benjamin. Virtual Worlds: A Journey in Hype and Hyperreality. London: Penguin Books, 1992.

"The Electronic Librarian is a Verb," forthcoming in Biblion, was delivered as part of the Gilbert A. Cam Memorial Lecture Series at the New York Public Library and is copyrighted by Kenneth Arnold. All rights, including redistribution, are reserved to the author. This essay may be downloaded for personal use only.