SIGCIS 2012 Workshop, Traditional Papers II: Information Societies

Name: Gustav Sjöblom
Institutional Affiliation: Chalmers University of Technology
E-mail address:
Paper Type: Traditional
Paper Title: The Soundtrack of Early Computing: Sound and Human-Computer Interaction in Sweden in the Early Mainframe Age
Paper Abstract: In this paper I investigate the role of sound in early Human-Computer Interaction in Sweden. The main source is oral history evidence gathered in the project “From Computing Machines to IT”, complemented by an extensive range of archival and printed sources. When the first Swedish digital electronic computer BESK was operational in November 1953, its constructors reported that they had ‘connected a loudspeaker, with which one can follow acoustically the running of the machine’. They had also produced ‘music codes suitable for demonstration’. These statements indicate that there was an aural dimension to human interaction with computers from the very beginning. From the 1960s the reliance on sound decreased as a result of the development of operating systems and high-level programming languages. This state of affairs has been recognized for many of the early mainframe computer installations worldwide, but has not been given extended scholarly treatment. In this paper I investigate in the Swedish context the following questions: Why did computer constructors attach loudspeakers to early mainframes? How was the aural dimension of computing utilized by early computer people? How important was it to the development of computing?

The findings are that the sound had two main applications: as an aid in debugging and monitoring and as a means of expressing national and professional identities. The debugging of early computer programmes was extremely time-consuming and programmers mobilized sound in order to facilitate this tedious task. The use of sound was widespread and can be divided into four categories: detecting non-convergent loops, testing the installation of new hardware, enable passive monitoring, and more advanced troubleshooting. The second use of the mainframe loudspeaker was turning the computer’s chirping into music. While the rise of computer music as an art form has been covered by previous historians, the evidence presented here instead points to the importance of music-based rituals in the construction of national and professional identities related to computing. The tunes played by early Swedish computers invariably carried national connotations (the national anthem Du gamla du fria, songs expressing holidays such as Christmas and Lucia Day (13th December), songs with strong national connotations like folk songs and tunes by the songwriter Evert Taube, and the drinking song Helan går). I argue that these musical practices were important in establishing a public image of the computer as a versatile multi-purpose machine in the service of the nation. But computer music was also important in shaping professional cultures relating to computing. All evidence I have found relating to early computer music derive from the Swedish computer BESK and its offshoots FACIT EDB, SARA, SAAB D21 rather than from installations by IBM and other American manufacturers. This implies that computer music was correlated to computer cultures geared towards tinkering rather than business computing. Summing up, the early use of sound in computing corresponds to two main function of sound in contemporary HCI: monitoring and music. The use of sound for monitoring was widespread and valuable although probably not indispensable. The use of sound for expressing national and professional identities is an important and previously understudied aspect of the shaping of modern computing.