Programming and Planning

Some 15 years ago, I argued in the Annals that the word program entered the vocabulary of computer developers from the ENIAC project. Within the ENIAC, the term referred to the control signal that synchronized and directed the actions of the machine's individual units. "It is convenient in discussing the ENIAC to distinguish between the numerical circuits (which operate on signals representing numbers) and the programming circuits (which recognize when and how a unit is to operate and which then stimulate the numerical circuits to operate)," wrote Adele and Her mann Goldstine in their initial published paper on the ENIAC. Eight years would pass before the field would shift the usage of the word program away from the electronic circuits that controlled the machine toward the symbolic instructions that describe a set of opera tions to be set in motion by those circuits. In the intervening time, engineers more commonly used the word planning to describe the process of preparing a list of instructions for a computer. Konrad Zuse called his programming language PlanKalcul. The Mathematical Tables Project, which operated on the cusp of the electronic computing era, was gov erned by a planning committee. The initial articles on programming techniques by John von Neumann and Herman Goldstine were called "Planning and Coding Problems for an Electronic Computer." The term plan derived from the production engineer ing field, a discipline that developed in the 1920s in England in response to the production demands of World War I. The story of its usage shows how the con cept of programming started to develop and suggests, as authors such as Michael Mahoney have argued, that programming owes much to production engineering and the fields that derived from it, notably systems engineering.