On March 20, 1981, the New York Times Op-Ed page carried a brief item musing that historians and biographers might soon be in for “slim pickings.” What precipitated this pronouncement was a report that Jimmy Carter, just months out of office and hard at work on his memoirs for Bantam, had lost several pages of text after hitting the wrong keys on his brand new $12,000 Lanier word processor. Word processing, the Times speculated, threatened to put historians—“those bloodhounds of the paper trail”—out of business. “Archivists,” meanwhile, “will be deprived of words scratched out, penciled in and transposed with wandering arrows. They will have to make do with electronically perfect texts.”

Perfect. No other word so encapsulates the aspirations as well as the anxieties that s accrued around word processing. We remember WordPerfect, of course, the software that rose to dominate the market after WordStar (from the mid-1980s through the early 1990s), but there were also now largely forgotten programs like LetterPerfect and Perfect Writer. Etymologically, “perfect” comes to us from the Latin perfectus, by way of Old French; the original Latin meaning encompassed the idea of completeness, the state of fullness or of being finished. The familiar hand gesture, index finger bent to touch tip of thumb, other fingers stiffened in accent, is a way of visualizing that condensed state of ripeness, the attainment of closure and completion. In word processing’s parlance, “perfect” was meant to connote a finished document, flawlessly formatted and printed, cleanly and clearly expressed. The sense of completion, meanwhile, also spoke to the desire for efficiency and productivity, a zeal for getting the job done.

Could a single invention both improve the quality of the work being performed and accelerate the pace at which it was completed? It seemed too good to be true, yet this dual ideal—the quest for flawless efficiency and effectively flawless results—was present from the very inception of word processing as a concept and technology. In 1974 a consultant named Walter A. Kleinschrod produced a report on word processing for the American Management Association. “The touch of a button,” he wrote, “triggers a perfect final copy.” A second AMA text echoed and embellished this language: “The typist can push a button and the machine, tirelessly and flawlessly, chugs out a perfect draft—once, twice, a hundred times if required, without fatigue or the errors fatigue can bring.” Whatever else it was then, word processing was also a form of futurism that, as Thomas Haigh deftly notes, was of a piece with a push-button lifestyle and, indeed, food processing, which was introduced to consumers by Cuisinart at just about the same time. As word processing continued to mature and also transition from dedicated stand-alone systems to early personal computers, this original ideal lost none of its appeal. “With a word processing system,” wrote computer scientist Ivan Flores a decade later in 1983, “you can actually produce a perfect document.”

Error was the archenemy of perfection: typos, misprints, mistakes, caked-up layers of correction fluid, and patches of paper rubbed raw by erasures—all of these disrupted the smooth, homogeneous surface of an impeccably presented text. So obsessed were early office managers with the pernicious influence of error on efficiency that entire studies were commissioned; the habits of typists were scrutinized at the most minute levels. It was known, for example, that a secretary making a mistake in the first few lines of the page was likely to remove it and feed in a fresh sheet and start anew; a mistake introduced farther down the page, however, would be attacked by the eraser. It was known how long the rubbing and retyping would take, along with the erasure and correction of any carbons associated with the original page. And, of course, it was known how much this all cost—in paper, in carbons, in erasers, and above all, in time.

Word processing promised to change all that. “Errors,” proclaimed the Kleinschrod report, “are no longer sins of incompetence, destroyers of confidence.” Indeed, they were so “easily rectified” that “first-time perfection” was no longer demanded, marking an astonishing shift in attitude. “Recording on magnetic tape is just as easy as typing at your fastest speed,” assured a roughly contemporary manual prepared by IBM’s Office Products Division in support of its most advanced word processing product at the time, the Magnetic Tape Selectric Typewriter (see Chapter 8). “Typographical errors are corrected merely by backspacing and typing over the incorrect character. A perfect tape is created which gives you a perfect copy . . . and no erasing!”