A Computer Was a Person
For a long stretch of history, the word “computer” meant a person who performed calculations. Observatories, navy offices, insurance companies and government agencies employed people to work through long arithmetic tasks, usually by hand, with pen, paper and printed reference tables. When electronic machines arrived in the 1940s, they inherited the name because they took over that job.
The work was demanding and repetitive. A calculation might be divided into many small steps, each carried out by a different person and then checked by another. This division of labor is worth noticing, because it resembles how later machines were organized: break a large problem into simple operations, run them in order, and verify the results.
Why Tables Mattered
Much of the demand came from tables. Navigators needed tables of positions of the sun, moon and stars to find their location at sea. Surveyors and engineers relied on tables of logarithms and trigonometric functions. Insurers used tables of mortality and interest. A good table saved every user from repeating a hard calculation.
The trouble was accuracy. Tables were calculated by hand, copied by hand, and typeset by hand, and errors could creep in at every stage. Mistakes in a widely used table could mislead many users. Table makers sometimes checked their work by having separate teams repeat the calculations and comparing the results.
This is the context for Charles Babbage’s early work in the 1820s. His Difference Engine was designed to calculate and print tables mechanically, in part to remove human error. The project was never finished in his lifetime, but it illustrates the motivation: an accurate, repeatable process that did not tire or make slips.
Human Computers at Scale
By the 19th and early 20th centuries, organized groups of human computers were common in science. At some observatories, staff examined and measured photographs of the sky, recording and classifying their findings. Women were often hired for such work, sometimes at lower pay than men, and several made important contributions to astronomy despite limited recognition. This pattern appears again in the story of women in computing, where technical labor was frequently undervalued.
During the Second World War, military needs created huge demand for numerical results, such as firing tables that told gunners how to aim under varying conditions. Teams of people, many of them women with mathematics training, produced these tables with desk calculators. The same kind of work later led to some of the first jobs for early programmers, as the first large electronic machines needed people who understood both the mathematics and the equipment.
Mechanical Aids
People did not work with their hands alone. A range of tools extended what a trained operator could do:
- Abacus and counting boards: physical aids that hold partial results and speed up addition and subtraction.
- Logarithm tables and slide rules: methods that turn multiplication into addition, trading some precision for speed. Slide rules remained a common engineering tool well into the 20th century.
- Mechanical calculators: geared desk machines for addition, and in some designs multiplication, that became common in offices and laboratories by the early 1900s.
- Punched-card tabulators: machines that counted and sorted cards, developed in the late 19th century for the census and then adopted by businesses and governments. Their use of punched cards for storing data influenced the first electronic machines.
None of these devices could follow a long program on their own, and each still depended on a person to set up the problem, move results between steps and interpret the output.
Analog Approaches
Another thread was analog calculation, in which a physical quantity stands in for a number. A well-known example is the differential analyzer, a large mechanical device built at MIT in the early 1930s to solve differential equations, with similar machines built elsewhere. It could handle problems that were tedious by hand, but setting it up meant physically connecting shafts and gears, and its accuracy was limited by mechanical precision. The trade-offs between this style and digital methods are discussed more fully in the comparison of analog computers and digital ones, but for early computing the key point is that analog machines were useful and widely used before digital electronics matured.
What Changed Afterward
Electronic machines did not appear from nowhere. They were built to solve problems that people had been tackling for generations: faster tables, fewer errors and the ability to take on calculations too large to organize by hand. Many of the people who worked on the first machines came from this world of human computing and tabulation, and they brought its habits with them, including careful checking and step-by-step documentation.
The arrival of electronics did not end the older methods overnight. Slide rules, desk calculators and card systems stayed in daily use for years, and for many tasks they were simply more convenient. Only gradually, as machines became cheaper and more reliable, did the profession of the human computer fade, and the name passed from people to equipment.
Understanding this earlier world helps explain why so many ideas in computing look familiar from the start: tables as stored knowledge, cards as input, division of labor, and the constant concern with accuracy. They began as human practices well before they became circuits, and the later computer algorithms behind scientific computing are in many ways formal versions of recipes that human teams once followed with pencils.