Close this search box.
Close this search box.

The Ultimate Internet Café: Reflections of a Practicing Digital Humanist about Designing a Future for the Research Library in the Digital Age

Bernard Frischer

I wish to thank Abby Smith for the invitation to speak and CLIR staff for their help in researching and writing this article. Special thanks to Amy Harbur for gathering statistics about book production in the United States and to Kathlin Smith for invaluable editorial advice. A number of scholars read and commented on earlier drafts of this article. I fear that the traditional disclaimer about how they do not all agree with everything I have written is even more necessary than usual! For taking the time to provide thoughtful comments and to suggest improvements, I wish to thank Jon Bonnett, Andrew Dyck, Tom Martin, Franco Niccolucci, Charles Rhyne, Nick Ryan, and Donald Sanders.

[Editor’s note: This paper was originally presented on April 26, 2002, as the keynote address at CLIR’s annual Sponsors’ Symposium. It was rewritten in the summer of 2002 for publication and is published here with minor changes.]

A Vision of the Year 2012

Let me start this essay with a vision. The year is 2012. Chris Borgman’s predictions about the global information infrastructure have been realized, and a vast amount of high-quality information is easily accessible online (Borgman 2000). But our great, historic research libraries, far from disappearing or even shrinking, are as alive and vital as ever. How can this be?

Instead of fighting a hopeless rearguard action against digital technologies in the early years of the twenty-first century, research librarians decided to embrace them. As a result, most research libraries are now outfitted with a real-time, immersive theater seating at least 50 people; some libraries even have several such theaters. Each theater features three highly luminous projectors with edge blending. A powerful supercomputer pumps out 60 frames per second of imagery onto the screen while generating appropriate sounds and even permitting users to move virtual objects around in three-dimensional (3-D) space. Users of the theater feel as if they are right in the middle of the subject of their study—be it ancient Rome, the three stable members of the C2H4O group of isomers, the interacting galaxy NGC 4038/9 in Corvus, or the geological stratigraphy of Mars. At will, users can fly over Earth and, moving a time bar, set themselves down at any one of several hundred sites of great importance to humanity’s cultural history.

The information contained in the computer models projected in the theater, though in part speculative, is by no means fanciful. Published by university presses, laboratories, and professional organizations, the models are readily available and reasonably priced. They have undergone the same academic peer-review process that has long been applied to print publications. As visualization tools, the models are a powerful resource in instruction, but—since they represent the state of our knowledge and ignorance—they are equally effective as midwives of new ideas and discoveries in pure research. Since the theater is the only place where users can work with this information in groups and in a totally immersive environment, the research library has become, more than ever before, the center of learning and research on campus. The theater is booked all day long by classes, research groups, and individual scholars. In the evening, community groups use it to catch up on the latest medical, astronomical, or archaeological discoveries. The theater is but one of many ways in which the research library has adapted itself to digital technologies, which, far from undercutting its raison d’être, have been greeted as tools that help the library achieve its goal of supporting research and teaching.

Now, let’s cut back to 2002. The University of California, Los Angeles (UCLA) is the only American university with such a theater—and it is buried in the bowels of the central computing facility, not prominently displayed in our research library. At UCLA, my little team at the Cultural Virtual Reality Laboratory (CVRLab) has made a handful of models of cultural-heritage sites, and a few hardy souls at other universities have made a couple of models of the other things I mentioned—brains, molecules, planets. But there is no easy way to license a model, or even to find out whether it exists, if you need one. There are no technical standards to ensure that my model of a building in ancient Rome will interoperate with another scholar’s model of the neighboring building or with another building on the same site at a different period of time. Nor are there any standards about the documentation, or metadata, that should be published along with the raw model so that users can quickly understand who made the model, on the basis of what hard data, and using what process of reasoning.

How do we get from 2002 to 2012? The central feature of the vision is a new activity—collaborative, interactive demonstrations of virtual reality models in the context of teaching and research—housed in a new space, the immersive theater. In this essay, I argue that putting that activity and space into the research library is both appropriate, in view of the research library’s mission, and desirable, if we wish to see the research library flourish well into the new century. I also argue that this is just one way in which the research library might embrace the new opportunities presented by the digital age, which always entail incorporating new user activities and services while developing suitable architectural designs to give them tangible form and support. My message is thus an optimistic one: The research library will survive because of the introduction of ever more and newer digital technologies, not in spite of them. If managed well and if understood strategically in terms of the evolution of our educational system and culture, the transformation of the library from the old analog technologies to the new digital technologies can occur with a minimum of pain and a maximum of gain.

The Future of the Book and the Future of the Research Library

Before writing about the future of anything these days, it is well to begin with a caveat: Things are changing so fast that we can at best speak of the short term (the next 5 to 10 years) and the medium term (the next 10 to 20 years). Beyond that, we quickly get into a realm better left to futurologists of the stature of a Ray Kurzweil (1999). So by “the future of the book and of the research library,” I mean what I think will happen in the next 5 to 20 years.

When we think of research libraries, we think, first of all, of books—lots of them. The fate of the research library, then, is an epiphenomenon of the fate of the book itself. And the good news is that the traditional printed book is doing better than ever. The same digital technology that might seem to threaten the book’s very existence is also giving us “print on demand,” making it easier and cheaper to produce books, reissue them, and publish new editions, all in relatively small print runs.1 The real problem that librarians may soon face is not the death of the printed book but the profusion of new titles, reissues of old titles, and new editions of scholarly books by living authors—all made more economical and practical by print on demand.2

Moreover, even power users of devices such as personal digital assistant (PDAs) overwhelmingly say that they prefer printed books to books online.3Not surprisingly, a recent survey of professional humanists found much the same result (Brockman et al. 2001, 3–4). This is just as well, since we are unlikely to have digital versions of every last obscure text and document for a long time, if ever. Digital-conversion projects, like their microfilm and microfiche predecessors, bump up against the realities of economic constraints that force us to set priorities for what is converted (Smith 2001) and to confront the ever-dreaded roadblock of copyright protection.4 So, the transition from the printed book to the book online is going to occur slowly, if relentlessly. Even the massive digitization project announced in December 2004 by Google will not include books under copyright or the bulk of the world’s collections of unpublished manuscripts.

But even if we imagine that, with time, more and more readers will be habituated to the online book—in part because they become accustomed to the technology and in part because the technology platform of the online book is more ergonomically designed—we can still safely predict that research libraries will continue to be needed because they are our repositories of precious documents: manuscripts, rare books, and similar materials. Humanists are known to prefer original documents to facsimiles (Brockman, et al. 2001, 2, 4), and there is no reason to think that this will, or should, change in this century. Even if these materials are put online (a massive task requiring many years), scholars will still find that nothing can replace autopsy of the original document. As a practitioner of statistical stylometry for the analysis, attribution, and dating of literary works (Frischer et al. 1996; Frischer et al. 1999), I will gladly stipulate that digital technologies can offer as much new support to the autopsy of texts and manuscripts as they offer, for example, to the medical autopsies of pathologists.

Three Consequences of Digital Technology for the Research Library

If research libraries continue to exist as the repositories of manuscripts, rare books, and printed books not yet available in digital format, then they will also face new opportunities and responsibilities in the digital age. I see three consequences for librarians, creators of digital products, and library designers.

  • First, in the digital age, the research library will be special not so much because of the quantity of information it can offer the user but because of the quality of the experience in which that information is presented.
  • Second, producers of digital content will need research libraries every bit as much as print authors needed them in the age of Gutenberg.
  • Finally, in the age of cyberspace, real space and compelling architecture will matter more than ever.

First Consequence: The Quality of Experience

The research library has always been what could be called the “high-end” place where information has been stored, cataloged, and delivered. Some research libraries have also been places where information was produced; however, production has always been considered a secondary part of the library’s mission. For example, the UCLA Young Research Library has generally received a very high ranking from the Association of Research Libraries (ARL); however, because of its lack of reading rooms, carrels, and meeting rooms, it has generally not been the place where UCLA students and scholars actually got their work done.

In the predigital age, “high-end” referred mainly to the quantity of information. We measured the importance of a library primarily by the number of books on its shelves and the quantity of journals to which it subscribed. In the digital age, information online will soon far outweigh information stored on a particular site (if it hasn’t already). Hence, it is not surprising to read the ARL report of a drop in total circulation between 1991 and 2003 (Kyrillidou and Young 2004, 10). What is astonishing is that the drop was only 7 percent: This trend will surely pick up strength in the next decade.

This will not necessarily consign the research library to the rubbish heap of history, because the research library can be a place where users find it convenient and even preferable to access a great deal of the online resources that they use. In the digital age, what makes a library high-end will pertain more to the quality of information management and presentation than to the mere quantity of information stored locally.

Users of digital content may not know it, but they need research libraries more than ever. It is true that we can access such content in our offices and homes. But as Friedrich Nietzsche, who started out as a professor of classics, once observed, a good philologist needs to consult 200 books a day. This may be an exaggeration, but humanists do need to read or browse through many books in a day, and often many books are open on their desks at the same time, as they compare one passage to another.

In the age of the digital library, this is still the case. While we can open many windows on one PC, wouldn’t it be nice if we could go somewhere on campus where we could find special digital work environments with multiple screens and multiple log-ins so that you could have the equivalent of 10 books open before you at the same time? And shouldn’t such a space be designed with printed matter in mind, too? Most of us live in a hybrid world in which the information we need comes both from traditional and from new media. Wouldn’t it be appropriate for the research library to be the one place on campus that offered such a workspace? Of course, this will require new space or a reconfiguration of existing space. But the fact that more and more books are being converted to digital format does offer the possibility that librarians can, in good conscience, consign the print versions of those books to long-term storage, thereby freeing library space for other uses. One high-priority use could well be the provision of the new, high-tech workplaces for which I am arguing here.

If so, the library, not the home, could become the preferred place for scholars to work.5 This would especially be the case if, as the number of books on the shelves declines in tandem with the rise in the number of online texts, libraries change their status from circulating to noncirculating. This would mean that scholars could count on finding the books they need on the shelves (I don’t propose closing library stacks). If books are fitted with inexpensive radio frequency identification devices, scholars could even locate books that are not in the right places because they have been misshelved or are being used by someone else in the building.

And couldn’t the Internet itself further serve the research library if, for example, the electronic catalogs of our libraries, which are now increasingly available on the Internet, were broadcast inside the library so that users could use wireless PDAs wherever they were in the building to discover where a book is shelved and whether it has been checked out? If readers were required to swipe each book they took off the shelf for use at their workplaces so that the central catalog could keep track of the position of each item in the collection, then systems of collaborative filtering could be used to convey that information by e-mail to readers with similar interests who are in the building at the same time (Sarwar et al. 2001). If such readers then chose to meet to discuss their work, their scholarship would be enriched and the library would have taken on a new role that is consistent with its original mission to further collaborative research.

Second Consequence: Why Creators of Digital Content Need Research Libraries

Once again, the need for the research library can be justified simply on the basis of its traditional role. Digital products need to be preserved just as much as books do. Digital products, moreover, may be more fragile than printed publications not only because of the vagaries of the storage medium but also because of the ephemeral nature of the hardware and software that supports them. Someone needs to preserve high-quality digital products. Why not the research library? Several forward-thinking librarians and information scientists have already begun to recognize this responsibility.6

But once again, there is a new role for the research library to play. If providing state-of-the-art, hybrid workstations will be a boon to a library’s users, it will also help digital producers who deliver their content over the Internet, encouraging them to produce versions of their sites that require the highest-possible bandwidth. But not all digital products are best delivered over the Internet; indeed, some were never intended for the Internet in the first place. They are best seen in theater-like spaces and in social settings I described in my vision of the year 2012. In 2002, the only such theater in an American university was the Visualization Portal located in UCLA’s Academic Technology Services. Since UCLA’s CVRLab is a major content producer for the portal, let me digress a bit to describe its projects and mission.

The Cultural Virtual Reality Laboratory: Producing New Tools for Teaching and Research in the Digital Age

The CVRLab was established at UCLA in 1997 to create scientifically authenticated, 3-D computer models of the world’s cultural-heritage sites. The hard part comes in defining and implementing what is meant by “scientific authentication.” Of course, in using the computer to re-create a building that was destroyed long ago, it is impossible to know whether you have achieved total accuracy. Our idea is that a computer model is scientific if it is transparent. We must publish not only the 3-D data about an archaeological site but also the footnotes, or metadata, that tell users everything they might like to know about the reconstruction, from who made it to why one kind of marble or plant material was used instead of another. By publishing the metadata along with 3-D data, the CVRLab wants to enable users to distinguish the securely known from the hypothetically reconstructed, to be aware of current scholarly controversies, and even to empower users to tear apart a model and put it back together in a way that seems more cogent. In developing metadata standards, the CVRLab and similar laboratories around the world are taking advantage of groundwork laid by librarians for the Dublin Core Metadata Initiative and the Visual Resources Association. On its own metadata committees, the CVRLab is seeking the active participation of librarians and information scientists.

Equally important for authentication is to use a scientific method for producing a model. This starts with something very simple—but something often missing in a commercial model of a cultural heritage site: A scientific model must have an author. The CVRLab has developed the notion of collaborative authorship involving, ideally, the cultural authority responsible for the site, a scholar who has written a technical monograph about how the building on the site was constructed, and a cultural historian who can put the site into a broader context. We include experts such as these on the team so that we can base the model on large-scale, measured drawings and on high-resolution photographs of the actual surfaces that remain of the monument. We also want our models to reflect up-to-date thinking and theories and to include all necessary permissions and blessings from the cultural authorities in charge of preserving the places that we re-create. As an example of one of our authorship teams, I would cite the group that directed our modeling of the early-Christian Basilica of Santa Maria Maggiore in Rome. The church was built in the first decades of the fifth century A.D. and has undergone many changes and transformations since then. Our goal was to strip away the accretions of later ages and to restore the building to its original appearance, which was dominated by a fine cycle of polychrome mosaics illustrating the lives of Abraham, Isaac, Moses, and Jesus. The team of authors consisted of a professor at the Dutch School in Rome who had recently published a highly regarded monograph reconstructing the original phase of the building and two curators at the Vatican museums—one in charge of the excavations underneath the church revealing the pre-Christian phase, the other responsible for the maintenance of the present building. It was thanks to our Vatican scholars that we were able to use the state plans of the basilica for our model as well as excellent photographs of the mosaics.7

The CVRLab is doing similar projects for the Roman Forum, the House of Augustus, and the Colosseum in Rome; the Villa of the Mysteries in Pompeii; the cathedral of Santiago de Campostela in Spain; the Second Temple in Jerusalem; the English colonial town of Port Royal in Jamaica; and the Inca sanctuaries on the Island of the Sun on Lake Titicaca. We hope to continue this work indefinitely, creating models of significant cultural-heritage sites around the world and showing the main phases in the development of each site (including, when pertinent, the destruction phase). In other words, working with colleagues at similar labs around the world, we hope to create a virtual time machine that will permit students and scholars to visit the very places they are studying.

It is one thing to create a real-time, interactive model and quite another to deliver it to our users. We do this in a variety of ways, including print, video, and the Internet. The way in which we deliver a given model is determined by our users’ pocketbooks and specific needs. The computer model is a flexible digital asset that can be used in a variety of ways. At the low end, in terms of interactivity, immersivity, and price, is a 2-D image that can be used to illustrate a publication or a sign in a museum. A bit higher up the scale is the video documentary. We can output fly-throughs of our models to DigiBeta, edit the segments, add music, voice-over, and other visuals. The result is a documentary presenting an archaeological site to the public. We have produced such videos for a number of exhibitions, including the London Science Museum, the Jubilee Year show on Christian art in Rome, and the new museum of the Basilica of Santa Maria Maggiore in Rome.

Then there is, of course, the Internet. We can post 2-D pictures on the Internet, and we can stream our video documentaries. But we can also put actual interactive models online. These versions do, of course, have less detail, and Internet users do not experience them with the degree of immersivity that is possible on other delivery platforms.

At the top end of the scale is the Reality Center or CAVE,8 two special kinds of spaces where users can come together in groups of typically 10 to 50 people to enjoy a fully immersive, real-time experience. A Reality Center has one screen that is curved 166° to 180° around the room. In the CAVE, there are screens on at least the three front walls of the space, and, ideally, also on the floor, ceiling, and back wall (Cruz-Neira, et al. 1993; Basu 2001). Since CAVEs and Reality Centers are expensive and available only in a handful of American universities, it is no surprise that the top end of our scale is also the least-frequent way of using our models. This is a shame, and it is where research libraries might be able to help.

Desirability of Displaying 3-D Computer Models in the Research Library

The Reality Center and the CAVE are two examples of spaces suitable for the display of digital products never intended for dissemination on the Internet. Today, they tend to be found not in a university’s libraries but in its central computing facility or department of computer science. This is not surprising, because both the technologies and content are still in the research-and-development phase. But this is quickly changing, and some would even argue that viable commercial solutions already exist. Be that as it may, all would agree that we are on the threshold of a period in which the 3-D computer model of a mathematical equation, complex molecule, distant galaxy, or ancient city will be as commonly used in university research and teaching as 2-D slides were throughout the twentieth century.

But before 3-D technology catches on, it must overcome the famous paradox of the chicken and the egg. Until CAVEs and Reality Centers are common on our campuses, no audience for models will exist. If there is no audience, there will be little funding and little incentive to carry forward this kind of work. You don’t have to be a constructivist to intuit that, all things being equal, students will learn more about the Roman Forum by visiting it than by reading about it, and that scholars are more likely to have new insights about the data they study if they immerse themselves in detailed, photorealistic representations of it than make doodles of it on their whiteboards. Research libraries could fill the void in our universities, at least in the first stage of the growth of computer modeling. As there are more and more 3-D models and more users, other venues will naturally develop. The price will drop as demand grows. But at first, the research library may be best equipped, in terms of its mission and skills, to host visualization theaters. It could then become the physical equivalent of the virtual communities that have been springing up with increasing frequency since the advent of the Internet and the concomitant growth of collaborative research in the humanities (Brockman et al. 2001, 13). In economic terms, research libraries could do for the digital publication of scientific 3-D models what they have long been doing for the print publication of scholarly books and journals: through standing orders, give publishers the courage and incentive to take the risks inherent in developing and marketing any new product.

I like to think that by embracing this particular digital technology, the research library will also, in effect, return to its roots, for the first research library—the great library of Alexandria—not only housed a great collection of books but also had botanical and zoological gardens, an astronomical observatory, and an anatomical theater (Schaer 1996, 12). In the modern period, such features have been spun off the library, which has come to offer the representation of our objects of study, but not the objects themselves. Given the infinite increase in the number of objects tracked by the modern university, it would be unrealistic to attempt to build an updated version of the Alexandrian library. But by admitting the new form of 3-D representation into the sacred precinct of the modern research library, we can eventually re-create something of the richness of the first great museum-library with the help of virtual reality technology.

Third Consequence: The New Importance of Architecture and Design

In attracting people to the new library of the digital age, digital theaters, high-end equipment, and digital services such as the wireless transmission of the catalog throughout the library will be magnets, but let’s not forget the important role of architectural design in creating spaces that are functional and, even more important, inspirational. In a sense, this, too, represents a return to (modern) origins: The first treatise on library organization, written in the mid-seventeenth century by Gabriel Naudé, placed great emphasis on the siting, orientation, design, and decoration of the library. Likewise, in the age of cyberspace, real space, made of bricks and mortar, still matters. It matters, I would argue, even more than it did in the last century, when the measurement of a library’s excellence was mainly quantitative. Those elaborate work spaces with many screens and multiple log-ins that I hope to find someday soon in my local research library—not to mention the virtual theaters I called for—will all take talented architects to design.

But beyond the needs occasioned by these new features of my ideal library in the year 2012, librarians need to think more about architectural design because in the digital age, users of physical libraries will want to experience something in a library that cannot be had in the office or home, and that something is the drama of community. Library buildings that communicate and foster a sense of that awe will be a centripetal force on our increasingly silo-ridden campuses, drawing people in and facilitating contact between faculty and students and between colleagues in different fields.

Research suggests that if you build it (or, at least remodel it), they will come. Just as power users of PDAs still surprisingly prefer printed to online books, so, too, typical owners of a PC unexpectedly often choose to work not in the splendid isolation of their homes or offices but in a bustling, 24/7 Internet café.9 Or, perhaps that isn’t surprising. After all, in Berlin or Vienna, the fact that you own a coffeepot doesn’t keep you from becoming an habitué of your local Konditerei.

The ARL statistics mentioned earlier offer some support for this in terms of the research library: Whereas total circulation fell between 1991 and 2003, the number of group presentations held in research libraries soared by 61 percent in the same period (Kyrillidou and Young 2004, 10).

Configured in the right way for work in the digital age and offering facilities such as reality theaters that can never exist in the home, the research library can become the ultimate Internet café where we find it convenient and congenial to connect to remote places. With this in mind, I think that the kind of research libraries that will encounter difficulty in making the transition to the digital age are those modernist structures with no inspiring communal working spaces that are more book warehouses than libraries. As a classic example, I must, alas, cite the Young Research Library at UCLA, where I taught for 28 years. For all the excellence of its collections and staff, this library has no grand entrance to lift the user out of the humdrum routine of everyday life, nor even a main reading room. Instead, it isolates readers in individual desks lined up along the perimeter of each floor. This is exactly what will no longer work when people can get from the Internet their fill of disintermediated rationality and Sherry Turkle’s pluralistic self, or, as Internet critic Hubert Dreyfus would more pessimistically characterize it, plain old-fashioned alienation (Turkle 1995; Dreyfus 2001).

I like to imagine the ideal new research library as following the lead of Pei Cobb Freed & Partners’ San Francisco Main Public Library, whose uplifting foyer was used to represent heaven in the film, “City of Angels.” More down to earth is the new Middlebury College library, which was completed in 2004. Architect Bob Siegel describes the project as driven by a concern for making the library a “social gathering center” on campus. It not only has meeting rooms, classrooms, and faculty offices but also provides enough book storage for an anticipated 50 years of service while at the same time accommodating other media, including digital media. It will be interesting to see how well it meets the goal of creating hybrid and high-end digital workspaces.10

What will happen to architecturally outdated buildings such as the Young Research Library? Assuming there is no money available to tear them down and start over, there are other, less expensive solutions. Important missing pieces—for example, a suitable entrance and reading rooms with state-of-the-art digital workstations—could be added. If, as I suspect, the aversion to reading books online diminishes as the e-book becomes more familiar, the graph line trajectories of new books printed and old books digitized may cross at some point 10 to 20 years from now. At that point, collections can actually start to shrink each year, as the newly digitized books are transferred to long-term storage facilities. This will free space in existing buildings for retrofitting along the lines suggested here.


The UCLA research library was one of the first in the world to complete a retrospective digital catalog of its collection, to make that catalog available online, and to remove the card catalog from the library. The space occupied by the card catalog is now devoted to current periodicals and to computer workstations that give access to the online catalog and other finding aids. The next logical step might seem to be the removal of all journals and books from the library, their replacement by an online digital library, and the closure of the library itself. Some observers have predicted such an evolution (Basili 2001, 35-46).

In this essay, I have argued against this scenario for a variety of reasons—some empirical (e.g., readers’ resistance to reading books online, the greater ease of publishing and reprinting physical books in the digital age) and some logical (e.g., the need for a place to store and access important digital and nondigital documents, new digital products that are not intended for delivery over the Internet). The essence of my argument is that, even in the digital age, some activities can take place in the research library more appropriately than anywhere else on campus and that there is a positive interaction between those activities and the design of the spaces provided to house them. As the activities change to take greater advantage of digital technologies and products that help the library realize its basic mission of promoting research and learning, so, too, must the physical design of the library.

The experience with UCLA’s retrospective digital-catalog project encapsulates some of the key features of this interaction: As the activity of book finding evolved from shuffling through note cards in hundreds of drawers in scores of cabinets to searching online, the cabinets could be removed and the computers put in their place. But this was not a zero-sum game. First, much more searching activity could take place both inside the library and, via the Internet, outside. Moreover, not all the freed space was devoted to the searching activity; some of it was allocated to the display of current periodicals.

But this example does not capture my entire thesis because it omits three important subsidiary points. First, there should be new space in our libraries for products made possible by digital technologies that are immersive and interactive, and that are not primarily intended for dissemination over the Internet. Second, the library needs to be the place for the production, not simply the distribution and consumption, of knowledge. It can do this by using technology to facilitate information gathering and by creating hybrid workstations where students and scholars can work and interact as individuals and as parts of larger collaborative work groups. Third, the architectural space of the library itself must be reconceptualized to express and leverage its main advantage over the Internet: the centripetal, community-building power of real physical presence over the alienating, community-rending effects of mere virtual presence. And let’s not forget the great cappuccino!


Basili, Carla. 2001. La biblioteca in rete. Strategie e servizi nella Società dell’informazione second edition. Milan: Editrice Bibliografica.

Basu, Paroma. 2001. The Virtual Voyager. Technology (Sept. 5). Available at

Borgman, Christine L. 2000. From Gutenberg to the Global Information Infrastructure: Access to Information in the Networked World. (Digital Libraries and Electronic Publishing Series). Cambridge, Mass.: MIT Press.

Brockman, William S., Laura Neumann, Carole L. Palmer, Tonyia J. Tidline. 2001. Scholarly Work in the Humanities and the Evolving Information Environment. Washington, D.C.: Council on Library and Information Resources. Available at

Cesana, Roberto. 2002. Editori e librai nell’era digitale. Dalla distribuzione tradizionale al commercio elettronico. Milan: Editore Franco Angeli.

Chodorow, Stanley. 2001. Scholarship, Information, and Libraries in the Electronic Age. In D. Marcum, ed. Development of Digital Libraries. An American Perspective. Westport, Conn.: Greenwood Press.

Cruz-Neira, C., D. J. Sandin, and T. A. DeFanti. 1993. Surround-Screen Projection-Based Virtual Reality: The Design and Implementation of the CAVE. ACM Computer Graphics 27(2): 135–142.

Dreyfus, Hubert L. 2001. On the Internet: Thinking in Action. New York: Routledge.

Frischer, Bernard, Donald Guthrie, Emily Tse, and Fiona Tweedie. 1996. ‘Sentence’ Length and Word-type at ‘Sentence’ Beginning and End: Reliable Authorship Discriminators for Latin Prose? New Studies on the Authorship of the Historia Augusta. Research in Humanities Computing 5; 110–142. Oxford: Oxford University Press.

Frischer, Bernard, R. Andersen, S. Burnstein, J. Crawford, H. Dik, R. Gallucci, A. Gowing, D. Guthrie, M. Haslam, D. I. Holmes, V. Rudich, R. K. Sherk, A. Taylor, F. J. Tweedie, and B. Vine. 1999. Word-Order Transference between Latin and Greek: The Relative Position of the Accusative Direct Object and the Governing Verb in Cassius Dio and Other Greek and Roman Prose Authors. Harvard Studies in Classical Philology 99: 373–406. Cambridge, Mass.: Harvard University Press.

Frischer, Bernard, F. Niccolucci, N. Ryan, J. Barcelò. 2002. From CVR to CVRO. The Past, Present, and Future of Cultural Virtual Reality. In F. Niccolucci, ed. Proceedings of VAST 2000. British Archaeological Reports 834: 7–18. Oxford: ArcheoPress. Available at:

Kirkpatrick, David D. 2002. Publishers and Libraries Square Off over Free Online Access to Books. The New York Times (June 17): C7.

Kyrillidou, Martha, and Mark Young. 2004. ARL Statistics 2002–03. A Compilation of Statistics from the One Hundred and Twenty-Three Members of the Association of Research Libraries. Washington, D.C.: Association of Research Libraries. Available at

Kurzweil, Ray. 1999. The Age of Spiritual Machines: When Computers Exceed Human Intelligence. New York: Viking.

Marcum, Deanna B. 2001. Digital Preservation: An Update. In D. Marcum, ed. Development of Digital Libraries. An American Perspective. Westport, Conn.: Greenwood Press.

Naudé, Gabriel. [1644] 2000. Advis pour dresser une bibliothèque: présenté à monseigneur le president de Mesme. Reproduction of 1644 edition. Paris: Klincksieck.

Sarwar, Badrul, George Karypis, Joseph Konstan, and John Riedl. 2001. Item-Based Collaborative Filtering Recommendation Algorithms. In Proceedings of the 10th International World Wide Web Conference (WWW10), Hong Kong, May 1–5, 2001. Available at

Schaer, Roland. 1996. Il museo. Tempio della memoria. Translated by Silvia Marzocchi. Trieste: Electa Gallimard.

Smith, Abby. 2001. Strategies for Building Digitized Collections. Washington, D.C.: Council on Library and Information Resources. Available at

Task Force on the Artifact in Library Collections. 2001. The Evidence in Hand: Report of the Task Force on the Artifact in Library Collections. Washington, D.C.: Council on Library and Information Resources. Available at

Turkle, Sherry. 1995. Life on the Screen: Identity in the Age of the Internet. New York: Simon & Schuster.

Waters, Donald J. 2001. The Uses of Digital Libraries: Some Technological, Political, and Economic Considerations. In D. Marcum, ed. Development of Digital Libraries. An American Perspective. Westport, Conn.: Greenwood Press.

Web Sites Referenced

Dublin Core Metadata Initiative:

Reality Center:

San Francisco Library:

UCLA Academic Technology Services.

UCLA Cultural VR Lab:

Visual Resources Association.


1 Cesana 2002, 179–189. Cesana mentions the Ingram Book Company, which is collaborating with IBM Printing System Company, which handles printing, and Danka Service International, which manages distribution. The books are printed on IBM’s InfoPrint 4000 High-Resolution Printer (which prints up to 666 pages per minute in high resolution—up to 600 dpi). Michael Lovett is quoted (Cesana 2002, 186) as saying, “This is a win-win situation for everyone involved in the book industry. The publishers win insofar as they sell books that otherwise would go out of print; distributors win since they can sell more books to a larger number of customers; consumers win because they have a larger selection of titles; and authors win because they continue to keep the copyright on their work” (my translation). IBM is in a similar partnership in Europe with Chevrillon Philippe Industrie, one of the biggest French publishers, and there is a similar operation in Italy in Trento at the firm Editrice Bibliografica.

2 For example, the Library & Information Statistics Tables report an increase of 27 percent in the number of books published in the United Kingdom between 1997 (98,477) and 2002 (125,390). Source:

3 Cesana 2002, 179.

4 For the state of the battle between publishers and librarians, see Kirkpatrick 2002.

5 See Brockman et al. 2001, 8, for evidence that most scholars prefer to do their most intensive reading at home; and see the same report, page 31, for a suggestion, complementary to the one I make in the text, namely, that the research library should facilitate scholars’ use of computers and online resources.

6 See, for example, Chodorow 2001, 12f.; Waters 2001; Marcum 2001; Task Force on the Artifact in Library Collections 2001, 41–54. See also the DSpace project at the Massachusetts Institute of Technology, funded by The Andrew W. Mellon Foundation, which is creating an archive in the MIT library for digital documents (

7 For more information about the project, see

8 “CAVE” is both a recursive acronym (Cave Automatic Virtual Environment) and a reference to “The Simile of the Cave” found in Plato’s Republic, in which the philosopher explores the ideas of perception, reality, and illusion. Plato used the analogy of a person facing the back of a cave alive with shadows that are his/her only basis for ideas of what real objects are.

9 An excellent example is the Easy Everything chain of very large Internet cafés. See

10 More information on this library is available at

Skip to content