c. 1750 – c. 1970 · observatories & flight labs · supplanted by the digital computer
before the word meant a machine, it meant a person. mostly a woman. she sat at a long desk in a hall full of long desks and computed — logarithms, planetary tables, ballistic trajectories, wind tunnel data — by hand, on ruled paper, with a slide rule and a sharp pencil and another woman at the next desk doing the same calculation as a check. errors caught by the second pair of eyes were marked in red. errors caught by neither were sometimes built into bridges.
the harvard college observatory in 1881 hired a room of women to classify quarter of a million stars. they were paid less than the men who measured them and discovered more. annie jump cannon classified three hundred thousand stellar spectra. henrietta swan leavitt found the relation between cepheid variable luminosity and period — the law by which we measure the distance to galaxies. she was paid twenty-five cents an hour. she did not see her own name on the law until after she had died.
at nasa langley in the 1950s, the computing pool was segregated. the west computers were black women; the east, white. katherine johnson calculated the trajectories of project mercury and apollo 11 by hand, on graph paper, and was asked by john glenn to verify the electronic machines' answers before he would fly. by 1970 the room was empty. the word "computer" moved over, intact, to the machine. the people who had been it kept their desks and worked on, as programmers, for as long as they were welcome.
field note · they wrote the result in pencil. then they wrote it again in pen. the pencil version was sometimes wrong. the pen version was almost never.