If you want to know what a card catalogue was, you have to know about the hole.
Each card was three inches by five — roughly seven and a half centimetres by twelve and a half. Down at the bottom edge, in the centre, a small hole was punched through the card. A metal rod ran through the holes, the whole length of the drawer, so that the cards could not be lifted out or spilled if the drawer was pulled too far.
The hole is the detail that makes it a real object, and it is the first thing lost when people describe the catalogue as a search tool. It was not an interface. It was a piece of furniture, with a runner through it.
One
The cards were typed. Not printed, in most libraries — typed, one at a time, by people whose job was to type them. A catalogue department might type thousands of cards a day. Each card carried a book's author, title and subject headings, and the call numbers that told you where it stood on a shelf. A typical drawer held about a thousand of them.
Then the cards were filed, by hand, in alphabetical order. And here is a thing worth holding onto: when a library's collection grew, the catalogue grew with it, and the size of the catalogue became visible from across the room. You could see how much had been written on a subject by how full the drawer was. You could open it and feel the weight of the field.
Two
The card catalogue was, in other words, the memory of the institution, and it existed in exactly one place.
Then it stopped.
The shift began with a woman named Henriette Avram. A mathematician, she was hired by the Library of Congress in 1965 to work out how a catalogue record could be put into a machine-readable form. The resulting format is called MARC — MAchine-Readable Cataloging.
The timeline is precise and it matters. A pilot project ran from 1966 to 1968; test tapes went out in October 1966, and weekly distribution to sixteen libraries began the following month. By June 1967 about 16,000 records had been distributed; by June 1968, around 50,000. The MARC II format was released in January 1968, and in March 1969 the Library of Congress launched the formal MARC Distribution Service, shipping weekly tapes of roughly a thousand records per reel.
The catalogue was becoming data. And data, unlike furniture, could be in more than one building.
Three
Libraries did not empty their cabinets all at once. They did something stranger first.
They stopped adding to the catalogue while leaving it standing. New books were catalogued only online; the drawers stayed where they were, still holding every card typed since the nineteenth century, now growing permanently out of date. This is what librarians called a frozen catalogue, and it was the normal state of a great many libraries from the mid-1970s through the 1980s. The reasons varied: cataloguing backlogs, cards physically deteriorating, a major revision of the cataloguing rules in 1978, and simply running out of floor space.
So for years, in libraries across the English-speaking world, you could stand in front of the authority for the entire collection and know that it had a cut-off date. Everything after that was elsewhere.
Four
Then the cabinets went.
The New York Public Library closed its card catalogue on 1 January 1972, and between 1977 and 1980 microfilmed about ten million cards — and then discarded them. The Library of Congress closed its own card catalogue on 2 January 1981, with more than fifteen million cards in it; the main catalogue was kept, but the catalogue as a working instrument was over. The University of Virginia's Alderman Library converted about four million cards in 1989, out of sixty-five cabinets holding some four thousand drawers. MIT's were hauled away that same year by a recycling company. The University of Michigan removed twelve and a half million cards in March 2010 — among the last of the big ones to go.
Five
Something was genuinely gained, and it is not the point of this piece to pretend otherwise. Online catalogues are searchable, copyable, shareable and impossible to misfile. A record can now exist in fifty libraries at once.
But something specific was thrown out, and it was not only the cardboard.
The drawer told you the shape of a field. The cards carried the marks of their own history — corrections, added subject headings, call numbers struck through and rewritten when a book was reclassified, sometimes a note in pencil from a librarian forty years dead. A frozen catalogue was also a record of when the library gave up on it.
None of that fit into a MARC field. The information was migrated. The object was not.
Six
This is the ordinary fate of a preserved thing when the format changes. What was a shared, physical, slightly worn object becomes a clean record — and the clean record is genuinely better at everything the object was for, except one thing: it no longer shows you that a person made it.
The catalogue was typed by someone. Filed by someone. Corrected by someone. Twelve and a half million cards, and behind each of them a pair of hands.
There is no field for that. So it went out with the cabinets.
https://3026alive.com