Walt Crawford says some very sensible things in his recent meditation, "The Card Catalog and Other Digital Controversies," (AL, Jan., p. 52-58). I was flattered to discover myself likened to a "shaggy beast from the forgotten past." At the risk of further beastliness, I would like to add a couple of codicils to his article.

Crawford says that I made "questionable claims about the dereliction to the book collection" of the San Francisco Public Library. My claim was that during the tenure of Kenneth Dowlin (city librarian from 1987 until 1997), the library discarded over 200,000 books. If anything, that number is overly cautious; it understates the severity of the damage the library sustained. It is based on the best estimates of librarians and other staff members who worked at the library and who saw city DPW trucks full of books regularly drive off to the dump. There were two periods of extremely heavy and indiscriminate weeding: one just after the earthquake in 1987, and one in 1995-1996, prompted by the panic-stricken, eleventh-hour revelation that the new building didn't have the capacity to hold the existing collection.

My article in the New Yorker offered estimates and not hard numbers, because unfortunately estimates were what I had to work with. The library's automation services department deleted all but the last "Purge of Items Declared Withdrawn" report from its file server, and apparently kept no backups. Still, that last 1996 purge report (which I had to sue the library in order to download) is plenty chilling on its own. In it are approximately 140,000 books, and 19,000 of them represent the last copy of an edition in the library system at the time it was removed. No "routine library procedure" sort of justification can possibly explain why my random scroll through the last-copy discard list, incomplete as it is, would turn up titles like Bennett Cerfs Reading for Pleasure, Hugo Grotius's Freedom of the Seas in a 1916 Oxford University Press edition, Victor Hugo's Actes et Paroles in a 19th-century multivolume Quantin edition, The Natives of Northern India (1907), and A Child's Garden of Verses for the Revolution (Grove Press, 1970). All these books are now gone from the library. The 164th numbered copy (out of 400) of a Chiswick Press edition of the works of Thomas Campion, published in 1889, is in the Purge Report, because the book was processed for discard, but it was one of thousands saved at the last minute by the rearguard action of several heroic librarians.

A small fortune squandered

The library has still not complied with a legal request to furnish memos and reports relating to the 1995--96 weeding. Staff members forwarded me e-mail from the then-chief of the Main Library, Kathy Page, to her department heads, saying that she was preparing to supply the relevant "documentation [they] have on file"--but it was never sent. Perhaps it was destroyed. Had the library not thrown it away, The Dynamics of Bureaucracy by Peter Blau might have offered some insight into the ongoing effort to cover up what went on.

Dowlin himself is gone, but Page, who was responsible for the worst of the weeding directives, is still consulting for the library. Steven Coulter, who, as head of the (non-elective) Library Commission, oversaw the whole building debacle, is still determining library policy. Coulter is also a senior executive at Pacific Bell, a major vendor to the library. He supported Dowlin's spending excesses and destructive bibliophobic reorganizations, I believe, because Dowlin was buying a whole lot of what he was selling. The larger the percentage of the pie-chart allotted to cables, data-switches, and leased telecommunications links--to all the chronically costly forms of remote access---and the smaller the slice devoted to on-site books, book-repair, shelves, and to the all-important cataloging and reference staff necessary to make sense of the library's holdings to the general public, the more the phone company stands to gain.

A bystander can celebrate or condemn the library's unpublicized decision to abandon its former role as a regional research institution, but the fact that it squandered a small fortune in communally owned books along the way is beyond contention.

That's one codicil; here's the second. Crawford refers to the "impossible economics of keeping card catalogs up to date" and says that librarians should tell patrons that the reason the card catalog was dumped was that "it was too expensive to maintain." There's a straw man for you. It isn't a question of maintaining card catalogs in the active, labor-intensive sense of filing cards into them and keeping them up to date. Most big card catalogs had been frozen for years by the time they were thrown out. As I wrote in the New Yorker piece (reprinted in a 1996 collection called The Size of Thoughts): "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." I nowhere suggest that we should abandon online catalogs and revert to manual filing and card revision. The reasons for keeping an out-of-date, frozen card catalog are the same as the reasons for keeping the National Union Catalog pre-1956 Imprints or a manuscript catalog for a monastic library whose books and buildings were scrapped by iconoclasts centuries ago: They hold interesting things in store.

Any book catalog is a panoramic tableau of a lost world--an itemization of what was available to a group of readers as of a certain time in the past. It tells us how those people thought books ought to be described and ordered and interlinked. It is often the only record (in the case of San Francisco, for example) that certain discarded books or serials were ever in the library. And card catalogs hold local information--analytics, book prices, color coding, guide cars--that is frequently lost when cards are matched with records on national databases. A card catalog is flawed, of course, as all finding aids are, and it serves up the vastness of recorded knowledge differently from any other referential innovation. Just as it helps to own several dictionaries, each with respective oddities and anachronisms and characteristic strengths, so several bibliographic search engines, even out-of-date ones with wrong call numbers and obsolete subject headings, used together may turn up items or remind us of things that, if we consulted them individually, we might miss. Finally, card catalogs hold the accumulated intelligence and hard work of generations of librarians--for that reason alone they deserve preservation.

~~~~~~~~

By Nicholson Baker

NICHOLSON BAKER, novelist and essayist, lives in Maine and is currently working on another piece about libraries for the New Yorker.