The Politics of Early Programming Languages

Abstract

There probably has never been such a controversial programming language as Algol. In the early 1960s the disciplinary success of the so-called Algol project in helping to forge the discipline of computer science was not matched by a significant adoption of the Algol language, in any of its three versions. This contrast is even more striking when considering the contemporary success of IBM’s Fortran, a language that, like Algol, was also conceived for scientific computation, but unlike Algol, initially only available for IBM computers. Through extensive archival research, this article shows how the relentless pursuit of a still better language that came to dominate the agenda of the Algol project brought to the fore the tension between the research-driven dimension of the project and the goal of developing a reliable programming language. Such a strong research-oriented agenda increased IBM’s doubts about a project that the firm already felt little urge to support. Yet IBM did not want to appear as obstructing the development of either Algol or Cobol, even if these “common languages” posed a clear risk to the firm’s marketing model. The US Department of Defense’s endorsement of Cobol and the rising popularity of Algol in Europe convinced IBM to push for the use of Fortran in Western Europe in order to protect the domestic market. IBM’s action in support of Fortran reminds us of the power imbalances that have shaped computer science.