Look at a sixteenth-century merchant or tax collector as he appears in the paintings of artists like Quentin Massys and Marinus van Roemerswaele. Not only the man’s costume but also the tools of his trade are unfamiliar.

He works with two kinds of information: loose documents on shelves, or perhaps hanging from bands, and bound volumes of entered transactions. Absent are filing cabinets, loose-leaf binders, card indexes, and rotary card files. We take these and other commonplace devices for granted—or rather we consider them survivals of the world before word processing.

Even after double-entry bookkeeping was introduced in fourteenth-century Italy, each receipt or disbursement was recorded only in a permanent, sequential form, with the date, nature of payment, and amount. Likewise in the medieval and early modern library, a catalog was a bound volume with books systematically listed by subject. Periodically a revision was published.

Between the medieval and Renaissance countinghouse or library and the late-twentieth-century data base, the world of shufflable paper was not a detour but a milestone. It marked new ways of gathering, storing, using, and even discarding information. These changes, which historians are only beginning to study, constituted a conceptual information revolution, in which flexible, changeable arrangements of facts took the place of fixed listings.

The new information order appeared first in scholarship, not in business. Librarians were already using slips of paper to compile lists for their printed catalogs in the eighteenth century. By the early 1790s the National Convention, in France, was trying to create a national union catalog—on playing-card stock—of the estimated twelve million books in public possession.

It was in the United States that both library and commercial card files first flourished. Our book collections were small by European standards, but they excelled in access. In 1850, when a proper catalog was still a printed one, fully half of America’s largest collections kept printed catalogs current to within ten years; only a few of Europe’s did. As the cost of catalog printing grew, American libraries began to let readers consult the card files that were kept to prepare future printed catalogs. By 1858 the Boston and Philadelphia public libraries had opened their card files to all patrons. Academic libraries followed, led by Harvard in 1861.

Commercial card files appeared around the same time. The first attested use of the phrase card index, according to the Oxford English Dictionary, was in an 1849 patent application. The patent refers to an “Improvement in keeping Ledger Accounts.” The triumph of the card file in the office is not easy to trace, but it does not seem to have occurred widely before the late nineteenth century.

Introduced pragmatically to save time and money, the public card catalog and the commercial card index replaced the static inventory of objects or set of account records needing occasional overhauling. These new open-ended systems permitted unlimited and immediate interfiling, reorganization, and expansion. By the turn of the century the Library of Congress was printing and distributing catalog cards on a national scale.

Meanwhile, habits of scholarship and commerce alike were changing. As the sociologist Wolf Lepenies has pointed out, the English social reformers Sidney and Beatrice Webb pioneered the use of cards in compiling original research. In the preface to their book Industrial Democracy, published in 1897, they wrote that they had relied on “separate sheets of paper, uniform in shape and size, each of which is devoted to a single observation, with exact particulars of authority, locality, and date.” By the early twentieth century the new information mentality was visible in everyday language on both sides of the Atlantic. One writer referred to the “automatic card-index system, known as his memory,” and another to “that marvelous brain of his, ready to be called on almost as if his mind were card-indexed.”

As revolutionary as the electrically sorted punch card and its electronic successors of the twentieth century have been, they embody ideas of organizing knowledge that have changed slowly over decades and that find much of their origin in the homely file card. The file card introduced the notion of the completely flexible collection of units of information; the computer has not changed the concept but only automated it. If we can learn more about the earlier innovation and how it developed, we may come to understand better the rapid changes in our own uses of information.

Edward Tenner is science editor at Princeton University Press and is the author of Tech Speak (Crown, 1986).