in: O’Donnell, James J. Avatars of the Word. From Papyrus to Cyberspace. Cambridge, MA: Harvard UP, 1998. 64-70.

H y p e r l i n k

THE SHRINE OF NONLINEAR READING

Much has been written and much said about the library past and future. I want here to juxtapose two images that can give some idea whither we come and whither we go.

I should say first that libraries are the great love objects of my life. I remember a narrow Quonset hut with a single aisle of children's books at an army post in the western desert with deep affection, and the next somewhat more permanent building they built there, where I looked in vain for years for the biography of Mickey Mantle they were supposed to have but must have lost (to my delight I found an autographed copy for twenty-five cents at a used book sale twenty years later). We moved to the big city—feebly supplied with books and libraries in those days—but there was a branch library we could walk to, a Carnegie library downtown, and a decent little library at my high school. I suppose I could even say that my first job (volunteer) was typing catalog cards in that high school library.

But it was going to college and finding myself suddenly granted the run of the open stacks of a great research library that well-nigh swamped my consciousness. The riches of that place and its successors during my graduate school years were an ocean that I did my best to drain dry. Since then, I have become agreeably amphibian, drifting in and out of library buildings with armloads of books. I have at this writing five valid library cards and somehow there is always room to bring a few more books home, but harder to find room in my briefcase to bring them back—a curious gravitational effect. But what do I find there? Just a lot of books?

Consider for a moment a shelf in any library you might visit. What strikes you first is the diversity of materials there. The commonality of the codex form connects physical artifacts that may have come from all over the world and range in age from a few weeks to a few centuries. There is an anglophone bias in our libraries in the United States, but even so the books you handle may come from England or America, or possibly Canada or Australia, and will have been printed typically any time in the last century. But this shelf may also have books written and printed in at least a handful of European countries and, depending on subject matter, other continents as well. (I emphasize here the physical origin of the artifacts: there is another cultural marvel in the business of translating and republishing the world's intellectual product, but that act of integration takes place outside the library proper.)

These materials are not haphazardly placed, but rigorously organized, according to cataloging systems of great beauty. The utility of each book is significantly enhanced by that organized placement side by side with the others. For instance, even if you had a library where books were all shelved on some other principle, say by date of acquisition, you would need to know how to find them through some subject-oriented cataloging system to make them useful at all. The limiting factor, to which we will return, is that each book must choose one place to reside, though there may be multiple subjects in a single volume. (As we just saw, the scholarly journal suffers from this particularly: the library's catalog does not index individual articles, and so many subjects are clumped together under single titles and single call numbers.)

The library not only shapes but also creates the resources we see. No one could depend on bookstores for all that we get from libraries. Libraries are the after-market stock managers of the world's publishers. When the publishers have wearied of a book and made all they can from it, the librarian takes it over, cherishes it, rebinds it, lets you read it, chides you to bring it back when you forget, and eventually worries about photocopying it or otherwise preserving it when it grows old and tired. Scholarly research would be crippled without libraries.

One other feature of the library shelf needs mention: it is in constant flux. Most American libraries allow books to circulate, and so there are always books that aren't actually there, but are known to belong there and can be recalled if necessary. The structure of the library's stacks is similarly open—new volumes are always being inserted at their own point in the run of shelves, and the whole collection gradually balloons.

The collection is also chronologically diverse. We are familiar with looking at shelves of books published over decades. I recently had the experience of seeing the shelves in a brand new university where all the books had been bought within the last five years. There was something almost frighteningly clean and regular about the shelves of shiny new bindings, all uniformly tagged. It felt more like a bookstore than a library.

This thought experiment highlights the flexibility, the diversity, and the subtlety of the machine we have constructed. It also invokes a picture not of the future—for I would not pretend to know what the library of the future will look like—but of the present. How has the library changed already?

Nostalgia about our old card catalogs can easily blur the real issues surrounding their eclipse. First, given the cost savings of electronic technology, it would simply be irresponsible for a major library to maintain a traditional card catalog. Second, the power of the new technology vastly transcends the old. But third, the integration of the new with the old already changes the nature of the collection it describes and expands its uses.

The catalog of my university's library is now accessible from any continent on earth through the Internet, and I consult it regularly when I am far from home: that is already a change. But also, this catalog is a tool that describes the library's collection—by virtue of its technology—far more flexibly than the card catalog ever could; and that enriches the collection it describes, even if the collection itself is not changed.

For example, in 1973, in graduate school, I spent half an hour one day trying to find a specific author in the old card catalog. He was the medieval Irish philosopher known variously as John the Scot or as Erigena. I went on a chase through cross-reference cards in file drawers: I bounced from Scotus to Scottus to John to Johannes to Erigena to the variant spelling Eriugena. I found individual items, but I never found the author file for this author, and gradually I began to feel like Pooh and Piglet on the trail of the woozle, going round and round the bush, the tracks getting more numerous as more woozles joined the parade, but no woozle appeared. Finally I stopped, took a deep breath, and went on the chase one more time to discover that the cards had indeed been correctly directing me to the drawer with the Latin form of the writer's first name—but the spelling used was crucially "Joannes" not "Johannes." I had seen this half a dozen times at least on my paper chase, but my eye had not registered it and I kept going to the drawer for "Johannes." With a guffaw, I went to the requisite drawer and went to work.

To compare my memory with the present reality, I performed this same search on our online catalog a few minutes ago. It took three steps and less than a minute to get to the full set of records I sought. The path I took was one of literally dozens that, depending on what I had asked first, would have taken me to the same goal just as efficiently. (When I went back not long ago to the same card catalog in which my original odyssey played out—still open to the public but with no new additions since 1977—I found that my experience had been negated by the instability of catalog subject headings. All the "Joannes" cards have now been moved over to the "Erigena" drawer, a transition made at about the time scholars were firmly coming around to agree that this was not the correct way to spell his name.)

Now when I get to that virtual shelf we were imagining earlier, I have gotten someplace I could never go before. All the library's holdings on my Irish philosopher are displayed before me, including (1) those in the main collection on the Library of Congress shelves, (2) those in special collections around the campus, (3) those in the rare book collection under special care, and (4) those not actually in the building just now but out on loan. (Those in the main collection with the older books on Dewey Decimal shelves are a few keystrokes away.) It can tell me where each book is and when the books that are checked out will be due. Our library has a prototype of what some other places have done more ambitiously, namely a system for placing requests for recalls and even, on some campuses, delivery to an office. To be sure, it would be a pleasure to have the full text of all the books available online behind their bibliographic record, but that is a fantasy for now. My point is that the collection is already a different thing because of new ways I can know and use it.

It is better too for questions I can ask that I could not ask before. Keyword searches let me search for topics of my own devising, without depending on a librarian to have selected it as a proper subject heading, and let me combine terms to refine searches. All the traditional subject headings are there, so if I do not know what I am looking for, I may try my own combination, then inspect a record to see what other ways of asking the question might be there. If at first a search is fruitless, I can submit another within seconds, and I can roam about the electronic catalog with far more flexibility than among rows and rows of catalog drawers. In those days, when I made my way through the rows and found myself in the Ws, a query that would take me back to the Bs might very well go unpursued. Now it too is a few keystrokes away.

Finally, even today our libraries are pointing beyond their physical collections in creative ways. Our collection allows those of us with University of Pennsylvania IDs to search the following resources: Medline (online medical journal abstracts—a resource famously interesting to amateurs and hypochondriacs as much as to physicians); RLIN (a union catalog of research libraries)—if my library does not have the book, I can consult this resource to see where it exists and use the information to order an interlibrary loan; MLA Bibliography (a huge resource of literary scholarship, including journal articles, indexed in great detail); the Oxford English Dictionary online (I gave away my old two-volume compact edition a couple of years ago: I hadn't consulted it in years); Lexis-Nexis (online newspapers, licensed to let us consult it for classroom use); Dow Jones News/ Retrieval; Books in Print; several more specialized indexes of scholarly and scientific journal literature, and then a world beyond all that as well, through the library's pointers to Internet information sources publicly accessible. (My own university's catalog on the Internet is joined by hundreds of others. There is real value to being able to consult, for example, the catalogs of European libraries from the comfort of my study in the United States to search for publications not readily accessible in this country.)

To describe how the Internet has already become a kind of alternate library would require another volume, and dozens of versions of that volume already exist, becoming obsolete hourly. But in one important regard, the Internet is not a library: nobody built it. There is great value in the diversity and abundance of information out there, and one may reasonably expect that diversity and abundance to continue to explode. But the qualities that make the library valuable are not quite there yet. There is no organized cataloging, there is no commitment to preservation, there is no support system to help you find the difficult or missing resource. Finally, there is no filter: that is, there is none of the sense that a user of a great library has that somebody has thought about the possibilities and selected a set of materials to be both comprehensive and yet delimited. On the Internet, you never know what you're missing. That may change. Or we may change. For the moment, the library is still the most powerful paradigm for the organization and management of knowledge ever invented.