The Size of Thoughts

by Nicholson Baker

  The Widener Library cards, once they are processed by OCLC, are packed up on Wednesdays and sent back to Harvard. There are no lost cards. (“In seventeen years, we’ve never lost a card,” Maureen Finn said, with justifiable pride.) At Harvard, Dale Flecker’s staff takes random samples from the boxes of returnees, to see whether OCLC is staying within its contractual quality-control tolerances. If the spot-checkers find unacceptable inconsistencies between a card and its computer version, they retain the card for reference and correct the record online.

  And what of the rest? What of the cards that resided in their drawers for fifty or a hundred years, some of which perhaps caught the eye of Charles Cutter or Fred Kilgour or one of the other great names in librarianship who trained or worked at Harvard? What of the handwritten ones? (“Library hand” was a special kind of backward-slanting penmanship meant specifically for card catalogs, and taught in library school through the 1920s.) Where are they all going?

  There have been a few requests for particular sets of cards as souvenirs. Someone, for example, wants all the cards for the Gutenberg Bible. And some undetermined but large fraction of the totality is being sent to an artist named Thomas Johnston, at Western Washington University. The rest are being “discarded.” “At the end of this project,” Dale Flecker told me, “there won’t be card catalogs left in the university.” I asked him if there were any card catalogs, anywhere in the world, that he thought worthy of preservation. “In general, they’re being discarded,” he said. “I’m not sure I know of anybody who’s decided to preserve them as physical objects.” Maureen Finn said much the same thing to me: “The institutions still want the cards back, and then I think they’re storing them. But most library managers that I talk to will say, ‘We are storing them because it makes the staff feel good, and we will be getting rid of them.’ ”

  Online catalogs are wonderful things in principle—things of efficacy, and even, occasionally (as in the case of OCLC’s Online Union Catalog), of grandeur. As a result of the publicly financed expansion of higher education in the sixties, and the boom in academic publishing that followed it, the holdings of the typical university library have doubled over the last twenty years. Without online catalogs, and the circulation and acquisition modules of software with which online catalogs are linked, libraries would simply not have been able to process all the books and journals that were arriving on their loading docks. By the early seventies, there was an ominous arrearage of uncataloged material waiting on herds of rolling carts near the overtaxed cataloging departments of most large libraries. Cataloging had reached a state of crisis similar to the one that forced the financial markets to close early during some of the high-volume weeks of 1969, when Wall Street’s paper-based methods of order processing couldn’t keep up with each day’s orders to buy and sell stock. Computerized stock-execution saved Wall Street (and made program trading and Black Monday possible, too), and computerized cataloging procedures saved libraries.

  So card catalogs had to be closed and “frozen.” Nobody can expect a library to maintain sequences of alphabetized cardboard for a collection that is growing, as some currently are, at a rate of five hundred items a day. Even at the beginning of this century, a writer for the Boston Evening Transcript could playfully fret about the unwieldiness of card catalogs:

  As these cabinets of drawers increase in number until it seems as if the old joke about the catalogs of the Boston Public Library and Harvard University meeting on Harvard Bridge might become literally true, the mental distress and physical exhaustion suffered by those consulting one of them becomes too important to be disregarded.

  And online catalogs, despite their neolithic screen displays and excruciatingly slow retrieval rates, offer many amenities. They do not grow mold, as the card catalog of the Engineering Library of the University of Toronto once did, following water damage. And they are harder to vandalize. Radical students destroyed roughly a hundred thousand cards from the catalog at the University of Illinois in the sixties. Berkeley’s library staff was told to keep watch over the university’s card catalogs during the antiwar turmoil there. Someone reportedly poured ink on the Henry Cabot Lodge cards at Stanford. The huge frozen card catalog of the Library of Congress currently suffers from alarming levels of public trauma: like the movie trope in which the private eye tears a page from a phone book at a public phone rather than bothering to copy down an address and a phone number, library visitors—the heedless, the crazy—have, especially since the late eighties, been increasingly capable of tearing out the card referring to a book they want. Indeed, one of the reasons that the New York Public Library had to close its public catalog was that the public was destroying it. The Hetty Green cards disappeared. Someone calling himself Cosmos was periodically making off with all the cards for Mein Kampf. Cards for two Dante manuscripts were stolen: not the manuscripts, the cards for the manuscripts. Card catalogs attract vandals because they are expressive of needful social trust and communal achievement, as are other common targets, such as subway cars, railroad bridges, mailboxes, and traffic signs. To the extent that a cluster of computer terminals linked to an online catalog protects a room full of older card cabinets by marginalizing them, they are performing a great service.

  And of course an online catalog is, when viewed in just the right flattering light, extremely convenient. “It’s wonderful,” said Julie Miran, of Swarthmore’s circulation desk, enamored of keyword searching: “It’s like watching the war on CNN.” I began the research for this article by dialing up an online university catalog from my computer and printing out a sixty-eight-page list of volumes relating to LIBRARIES—AUTOMATION—CASE STUDIES, LIBRARIES—AUTOMATION—CONGRESSES, and so on. Though it took ten times as long to view and reject a given irrelevant book or subject heading onscreen as it would have if I had a drawer of cards in front of me, and was able to riffle through it at medium thumb-and-finger speed, I didn’t have to drive to campus and find a place to park and walk to that particular drawer in order to arrive at a useful preliminary idea of the domain of locally available knowledge concerning the demise of card catalogs. Online catalogs are wheelchair-accessible in the best possible way, and now—through Gopher software on the Internet—I can poke around in the catalogs from hundreds of libraries all over the world, despite the fact that I am not affiliated with a single one of them.

  Why would I want to take a look at the online catalogs of libraries I may never actually visit? To find out what clever-clever names their administrators have given them, for one thing—possibly out of their belief that people get exasperated at CRT terminals in libraries, and curse at them, simply because people suffer from a subcortical fear of technology, and that an evocative human or animal or mythopoeic name, hallowed by the vernacular, will make everything better. Thus, the New York Public Library has CATNYP. There is BEARCAT (Kutztown University) and ALLECAT (Allegheny) and BOBCAT (NYU’s Bobst Library) and CATS (Cambridge). There is VIRGO (the University of Virginia), FRANCIS (Williams College), LUCY (Skidmore), CLIO (Columbia), CHESTER (the University of Rochester), SHERLOCK (Buffalo State College), ARLO (the University of Colorado at Colorado Springs), FRANKLIN (the University of Pennsylvania), and Harvard’s appropriately Eustace Tilleyish HOLLIS. There is BISON (SUNY Buffalo), OASIS (the University of Iowa), ORION (UCLA), SOCRATES (Stanford), ILIAD (Butler), EUCLIDPLUS (Case Western), LUMINA (the University of Minnesota), and THE CONNELLY EXPLORER (La Salle). MELVYL (the University of California system) is named after Melvil Dewey; the misspelling was reportedly intentional, meant to emphasize the difference between Dewey’s cataloging universe and our own. SUNY Brockport’s Drake Memorial Library greets its users with a typographically generated image of a card catalog:

  Your automated catalog, by DYNIX.

  Copyright (c) 1992 by DYNIX, Incorporated.

  Meanwhile, the cards from Drake’s actual card catalogs are, according to a reference librarian I talked to in October of 1993, being used as scrap paper. The reference man gave
a small, embarrassed laugh when he passed this piece of news on to me. And he is quite right to be embarrassed. Drake’s physical catalog has been replaced by an eager-beaver screenful of exclamation points and bracketed equal signs, as if to insist on equivalency, when in fact there is no equivalency. The unfortunate truth is that, in practice, existing frozen card catalogs, which just sit there, doing no harm to anyone, are typically being replaced by local databases that are full of new errors (an early OCLC study found 1.4 errors per record input), are much harder to browse efficiently, are less rich in cross-references and subject headings, lack local character, do not group related titles and authors together particularly well, and are in many cases stripped of whole classes of specific historical information (e.g., the original price of the book, its acquisition date, its original cataloging date, its accession number, the original cataloger’s own initials, the record of any copies that have been withdrawn, and whether it was a gift or a purchase) that existed free, using up no disk space or computer-room electricity, requiring no pricey software updates or daily backups or hardware service calls, right in the original Remington or Brodart wooden cabinets.

  Think of yourself as a successful literary agent, with a big Ferris wheel of a Rolodex on your desk. You and your Rolodex go back fifteen years. It holds hundreds of names and numbers, many of which you have updated by hand when a writer or an editor has moved or got married or had a child or hired a new underling. Fond though you are of your Rolodex, it is hardly portable, and you are doing a lot of business on the other coast now. So you decide to get one of those electronic calendar-spreadsheet-address books—something along the lines of a Psion Series 3a, say. After careful planning, you freeze the Rolodex, and then you assign one of your interns the exacting task of keying into the Psion all the up-to-date information that your Rolodex contains. It takes the intern a solid week—there are four hundred and sixteen names. You spot-check the project as it progresses. The intern has made a few screwups here and there, reversed some numbers, made some typos in foreign addresses, but in general it’s surprisingly clean work.

  When this retrospective conversion is complete, however, a question arises. What do you do with the big frozen Rolodex? Burn it, pulp it, shoot it? It does take up a lot of desk space. Do you have a big party, and invite all the people in your Rolodex to join you on the roof of your building to drink champagne and tie their address cards to helium balloons and release them over West Fifty-seventh? No, because you are a literary agent, after all, not a publicist. And you quickly see that it would be an error to throw away your Rolodex cards right now, since you are going to want to refer to them from time to time in the months ahead, when there is a question of an electronic address’s correctness. At one point, looking up someone’s name in your Psion, you find that it isn’t there: an undetected typo made by your assistant has displaced the record somewhere, hiding it from you. You find the address easily on the Rolodex. Not only that: the démodé Rolodex, you discover, groups things in a way that is at times more useful to you than Psion’s rote technique. Rather than alphabetized solely by name, for instance, the old paper-based system offers you all your friends at Simon & Schuster together in one clump, a form of what librarians call “collocation.”

  Also, in a more reflective moment it occurs to you that there is considerable information on the Rolodex cards that didn’t make its way into your new toy: old, crossed-out addresses, old phone numbers, old spouses, old editorial assistants who are now publishing titans in their own right. The very degree of wornness of certain cards that you once flipped to daily but now perhaps do not—since that author is drunk and forgotten or that magazine editor has been fired and now makes high-end apple chutneys in Binghamton—constitutes significant information about what parts of the Rolodex were of importance to you over the years. Your new Psion can’t begin to tell you that: its addresses are ageless, as fresh and yellowy-gray as the current in a diode. Your Rolodex is a piece of literary history, in a way. It is also the record of some of the most cherished connections you have formed with the world. Would it be stretching things too much, you suddenly wonder, to call your Rolodex a form of autobiography—a manuscript that you have been writing these fifteen years, tinkering with, revising? Perhaps it is the only manuscript you will ever write. Throw it out? No, you will donate your Rolodex to your alma mater’s library, valuing it at several thousand dollars in order to get a tax write-off, and the librarians there will recognize its importance to future scholars of late-twentieth-century publishing practices and will lovingly catalog it online, assigning it a Library of Congress call number and an appropriate list of subject headings.

  Now imagine something just a little larger than your Rolodex. Think of an unbound manuscript, the only one like it, composed of a great many leaves of three-by-five-inch cardboard—a million of them, in fact—each leaf covered recto and sometimes verso with detailed descriptions of certain objects that the world has deemed worthy of organized preservation. The authors of this manuscript have worked on it every day for a hundred and twenty years. It is, then, the accreted autobiography of an institution whose job it is to store and retrieve books and book-like materials. Many of its authors were smart and careful people—perfectionists, wide readers, though by predilection keeping themselves as anonymous in their authorship as medieval cathedral builders. Some of them had specialized knowledge and idiosyncratic enthusiasms, which they worked into the pages of their creation by employing thousands upon thousands of “See” and “See also” pointers to other pages. Together, over the years, they achieved what one of their early masters, Charles Ammi Cutter, called a “syndetic” structure—that is, a system of referential links—of remarkable coherency and resolution.

  The authors made one serious mistake, however. Although they had taken great pains to be sure that within their massive work every book and manuscript stored in their building was represented by a three-by-five page, and often by several pages, describing it, they had forgotten to devote any page, anywhere, to the very book that they had themselves been writing all those years. Their card catalog was nowhere mentioned in their card catalog. Dutifully, they had assigned call numbers to large-type Tom Clancy novels, to magnetic tapes of statistical data, to diskettes full of archaic software, to old Montgomery Ward catalogs, to spools of professional-wrestling magazines on microfilm, to blueprints, wills, contracts, and the archives of electronic bulletin boards, to pop-up books and annual reports and diaries and forgeries and treaties and realia of every description, but they left their own beloved manuscript unclassified and undescribed, and thus it never attained the status of a holding, which it so obviously deserved, and was instead tacitly understood to be merely a “finding aid,” a piece of furniture, wholly vulnerable to passing predators, subject to janitorial, rather than curatorial, jurisdiction—even though this catalog was, in truth, the one holding that people who entered the building would be likely to have in common, to know how to use from childhood, even to love. A new administrator came by one morning and noticed that there was some old furniture taking up space that could be devoted to bound volumes of Technicalities, The Electronic Library, and the Journal of Library Automation. The card catalog, for want of having been cataloged itself, was thrown into a dumpster.

  Now some history. In 1791, in Paris, the Revolutionary government, having confiscated a number of private and monastic libraries throughout France, became curious to know what interesting books it suddenly possessed. The Imprimerie Nationale issued an Instruction pour procéder à la confection du catalogue to those charged with watching over new state property in outlying departments. Inventorists were told to number every book in a library, and then to write down, on ordinary playing cards, each book’s number, its author, its title, a brief physical description, and the name of the library where it could be found. (Aces and deuces, it was suggested, might be pulled from the deck and set aside for books with wordy titles.) These cards were to be alphabetized by author, strung toget
her, and sent on to Paris.

  In 1848, Anthony Panizzi, Keeper of Printed Books for the British Museum, had a similar notion:

  By an alphabetical catalogue it is understood that the titles be entered in it under some “headings” alphabetically arranged. Now, inasmuch as in a large library no one can know beforehand the juxtaposition of these headings, and it would be impossible to arrange them in the requisite order, if they cannot be easily shifted, each title is therefore written on separate “slips” of paper … which are frequently changed from one place to another as required. It is self-evident that if these “slips” … be not uniform, both in size or substance, their arrangement will cause mechanical difficulties which take time and trouble to overcome.

  Slips of paper and decks of playing cards eventually gave way to drawers of annotated cardboard; these were employed, through the 1860s, not as ends in themselves, to be browsed by patrons interested in finding books, but as a convenient means for the staff to keep track of what it had, or to prepare for the publication of a formal catalog. For the ornate, expensively produced catalog in book form was the traditional way a library presented itself to the public—the way it entered, as it were, the library of libraries. And, as it happens, a more than perfunctory catalog of a library’s holdings is an exceedingly difficult book to edit and publish. Charles Coffin Jewett, the librarian of the Smithsonian Institution, in his Smithsonian Report on the Construction of Catalogues of Libraries (1853), wrote: