with its infamous QWERTY arrangement (the name comes from the first six letters of the upper row). Devised to separate keys that jammed easily on early machines, it is far from ideal.

The Sholes and Glidden machine of 1874 was the first practical model for office use, but the typewriter could not achieve acceptance until its keyboard was standardized.

The QWERTY layout is a still much-debated case of acceptable, widely adopted technology triumphing over superior technology. The Microsoft trial focused attention on similar standards and technological “lock in.” VHS versus Betamax, Windows versus Macintosh—the fault lines are familiar from our own day.

Once “locked in,” standards have curious effects. No less shrewd an observer than the French architect Le Corbusier remarked how the typewriter had helped bring the adoption of 8.5 by 11 inches as the standard American business-letter size. Without the typewriter, we might have a neater number—8 by 12, say. The stationery industry standardized its sizes by the 1920s, and without any of us having a vote. Lawyers got their own size pads and files to fit older courthouse systems.

That in turn led to standard dimensions for file folders to hold documents and file cases to hold file folders. Desk drawers took on consistent dimensions as the desk became standardized. Modern clean-surfaced, drawered desks replaced the more private hutchlike environments of the nineteenth century, where papers had been rolled into scrolls, bound in the red tape that became the classic symbol for bureaucratic sloth, and stored in pigeonholes.

Then there is the case of’carbon paper. Around 1806 Ralph Wedgwood, of London, produced a “writing press” using paper imbued with printer’s ink. His system was taken up by Thomas Jefferson, that inveterate tinkerer with home-office equipment, and other leading political and business figures of the time. By 1823, Cyrus P. Dakin, of Concord, Massachusetts, was making similar carbon paper, and it was quickly adopted by newspaper reporters.

One Lebbeus H. Rogers came across the Dakin paper in use by reporters in the late 1860s, perceived its wider commercial utility, worked out a licensing arrangement, and founded a firm to exploit it. The first key sale came in 1870: $1,500 worth of paper to the United States War Department. Then in 1873, the “killer application” for carbon paper arrived: the typewriter, one of whose advantages was the ease with which it could hammer out multiple copies. The paper was of course 8 V2 by 11 inches—the same size as the bins on today’s laser printers.

The basic typewriter mechanism could be adapted to other tasks too. Beginning in the 1880s, one was modified to punch holes in cards that recorded data, such as rail schedules, freight shipments, payrolls, and worker records. This technology eventually became the focus of last fall’s election controversy, for the butterfly ballot and the Votamatic voting system are refinements of the old Hollerith punch card, another

staple of the American office. It made possible the modern census and the Social Security system. (For years, government checks were punch cards.)

The roots of this key office technology lay in a specific provision of the Constitution itself, the one mandating the U.S. census. When the first census took place, in 1790, tabulating its data took nine months. The results of the 1880 census took seven years to compile. With the 1890 survey looming, Herman Hollerith, a former employee of the Patent Office, built a tabulating machine using punch cards to record and sort census information. He made the cards the same size as dollar bills, so that existing storage cabinets could be used. His machines compiled the results of the 1890 census in six months.

It was the first of many instances in which the demands of government bureaucracy would profoundly change practices in private as well as public offices. Punch cards had been used in looms and employee time clocks, but Hollerith pointed to another application as the initial inspiration for his cards. He had encountered a railroad conductor punching holes in a railroad ticket to record information about the passenger. The ticket was called a “punch photograph.” Hollerith said he simply took the punch photograph idea one step further. He compared the census itself to a huge photograph of the American population—”full of life and vigor,” he said—made up of the punch photographs of individuals.

His cards found immediate applications by the railroads that had inspired him. They turned to them for tracking cargoes, employees, and passengers and for monitoring other vital information. More large corporations with information to record and manipulate followed suit. The New York Central hired Hollerith in 1896; the Marshall Field department store in 1903.

In 1911, he sold his tabulating-machine company to the combine that would become International Business Machines. The future of data processing was launched, and IBM assumed the central place in the unfolding of office technology it would occupy for the rest of the century. The punch card also put office data processing firmly on the road to the binary system. Even if vacuum tubes, not to mention transistors and chips, still lay far in the future, the card was already storing information in essentially the same way a computer does, as 1’s and O’s. A hole in the card is either punched through (a one) or not (a zero)—except, of course, in the case of hanging chads. The cards remained largely unchanged as the machinery to manipulate them evolved from mechanical counters feeding glorified adding machines to optical readers feeding ENIACS and Honeywells. By the time they were replaced with tape and discs, the pattern was set.

The inventor James Smathers had pioneered the electric typewriter before World War I, but during the twenties its sales amounted to only a few thousand machines. Refined and improved by

the Electromatic corporation, his technology attracted the attention of IBM, which bought out Electromatic in 1933. The first electric model with the IBM nameplate came on the market in 1935, launched into the face of the Depression. To promote the machine, IBM hired a champion typist, Margaret Hamma, who gave demonstrations in which she reached speeds up to 150 words per minute with cups of water balanced on the backs of her hands to show how little effort the typewriter took.

The advantages of the electric, however, lay as much in its ability to deal with multiple-copy forms as in its sheer speed. And it was a platform from which typing could evolve toward more sophisticated manipulation of text. It was the gateway to, yes, word processing, and the personal computer beyond.

In 1961, IBM launched its Selectric typewriter, based on a radically new mechanical system. The Selectric’s alphabet was housed on a golf-ball-size unit instead of the individual mechanical keys of the past. This type element could be changed, allowing one machine to employ multiple typefaces.

The outer shell of the machine was designed by Eliot Noyes to reflect the aesthetic of the ball print unit while softening its geometry into a more natural shape inspired by concepts of biomorphism. Noyes wanted the Selectric to suggest a sculpture by lsamu Noguchi, who provided art for IBM’s headquarters, and to radiate a soothing presence in the rigid geometry of the office. And the machine was made available in many different colors. Companies could even order custom colors; the University of Kentucky, for instance, located near the Selectric factory in Lexington, patriotically put in an order for a number of typewriters in the shade of blue worn by the school’s teams.

The Selectric established the idea that a piece of office equipment should be attractive. It was a long time before personal computers were as pleasant to look at and be around. But the keyboard on the Selectric remained the QWERTY keyboard. So does the keyboard on the glowing Apple iMac. And, even though there are now staplers whose translucent hues and blobby shapes imitate the look of the iMac, they still use the same standard staples as my old Swingline.