Human Computer

For more than two centuries, a computer was a person. The word named a job, not a machine: someone who computed, who sat with paper, pen, tables of logarithms, and later a mechanical calculator, and ground through the arithmetic that astronomy, navigation, ballistics, and engineering demanded. The job was old by the 1700s, when nations needed star tables to find longitude at sea, and it was still being advertised in the 1950s, when the work was firing artillery and launching rockets.

The great projects ran on human computers organized like factories. Nevil Maskelyne’s Nautical Almanac, first issued for 1767, was calculated by a dispersed network of computers and comparers working from their homes across England. After the French Revolution, Gaspard de Prony organized hundreds of workers into a three-tier hierarchy to grind out vast logarithmic and trigonometric tables, boasting that he could manufacture logarithms as one manufactures pins. The model was division of labor applied to arithmetic.

Over the nineteenth and twentieth centuries the role was steadily feminized. At the Harvard College Observatory from the 1880s, Edward Pickering hired a corps of women, the Harvard Computers, to classify stars from photographic plates; among them Williamina Fleming, Annie Jump Cannon, and Henrietta Swan Leavitt made discoveries that reshaped astronomy. During the Second World War, women computed artillery firing tables at Aberdeen, and six of them went on to program ENIAC. At NACA and NASA, the West Area Computers, including Katherine Johnson, Dorothy Vaughan, and Mary Jackson, did the mathematics of flight and spaceflight.

The electronic computer, built in part to do exactly this work, took both the work and the name. ENIAC was completed in 1945; through the late 1950s and 1960s, machines like IBM’s mainframes absorbed the calculation. By around 1970 the human computer as a job had effectively ended, but unlike most vanished trades the people did not simply disappear. Many became programmers, mathematicians, and engineers; the title they had held migrated to the machine that replaced them, which is why we still call it a computer.

Linotype Operator

For nearly a century the daily newspaper was built, line by line, out of molten lead by a worker sitting at a strange ninety-key keyboard beside a pot of metal kept at around 280 degrees Celsius. The Linotype, invented by Ottmar Mergenthaler and first installed commercially at the New York Tribune in July 1886, cast an entire justified line of type as a single slug of hot metal. It ended four centuries of setting type one letter at a time by hand, and it made the large, cheap, fast modern newspaper possible.

The Linotype operator was a high-skill craft worker. From the keyboard, pressing a key released a brass matrix — a mold for one letter — that slid into a line; between words the operator dropped in wedge-shaped “spacebands.” When the line was full, the spacebands were driven up to push the words apart and justify the line to an even width, molten lead was injected to cast the slug, and the matrices were automatically sorted back to their channels. One machine did the work of several hand compositors.

The job left its fingerprints on the language. When an operator made an error mid-line, the fastest fix was often to finish the slug with nonsense and discard it; running a finger down the first two vertical columns of the keyboard produced the famous gibberish “ETAOIN SHRDLU,” which slipped into print often enough to become printing folklore. The hot-metal composing room itself became a stronghold of skilled, well-paid labor, organized in the United States by the International Typographical Union.

Then the technology that had displaced the hand compositor was itself displaced. Phototypesetting in the 1950s and 1960s, then computerized and CRT typesetting paired with offset printing, made hot metal slow, dirty, and obsolete. The most-cited symbolic ending is the New York Times’s final hot-metal night of 1 July 1978, captured in the film “Farewell, Etaoin Shrdlu.”