ABSTRACT
If we felt the need to give the history of museum automation a beginning, then perhaps the date might be 1963 and the place Washington, DC. Certainly, other forays in automation were also beginning around this time.1 However, it was at this particular date and in this particular location that the then Director of the Museum of Natural History, Smithsonian Institution, appointed a scientifi c staff committee, chaired by Donald Squires, to develop through a series of meetings, consultations and training courses a ‘general understanding of the potential of data processing for the museum community’ (Smithsonian Institution 1967: 6; Squires 1969a: 2). Subsequently, in 1965 the Management Services Division of Peat, Marwick, Mitchell & Co. undertook a major study into the potential utilisation of computers by the Smithsonian Institution. The resulting report made specifi c recommendations for the development of a data processing system in the Museum of Natural History. Following this, in July 1967, the museum embarked upon a programme of research into the automation of documentation.2 The work was carried out in conjunction with the newly founded Information Systems Division of the Smithsonian Institution and through the support of the Offi ce of Education Department of Health, Education, and Welfare (Squires 1969a: 14). For the fi rst 18-month phase of development the Smithsonian (1967) had requested $292,927 of federal funds. Essentially, the Smithsonian was attempting to understand whether it could apply computer technology to the management of museum collections and what the benefi ts might be. In its endeavour, the project used as its sample data, specimen records from the national collections of sea birds, marine crustaceans, and rocks. By January 1969 a data processing system had been developed and implemented, which included the standardisation of input procedures for cataloguing using punched paper tape units, as well as the creation of data entry and query software. By 1970, as the second of the two grants from the Offi ce of Education was drawing to a close, personnel involved in the research and development were quick to ensure the legacy of the project. ‘Your accomplishments’, wrote Donald Squires to his team, ‘are part of a task which is increasingly being recognised as a major accomplishment in the fi eld of information storage and retrieval [. . .] there is now every reason to believe that our collective labors
will bear fruit and that a major renovation is in store for the museum community’ (1970). These not only proved, in the long term, to be somewhat prophetic words, but, more practically, this evidence (and no doubt this rhetoric) also had an impact on the museum’s directorate.3 For in that same year the National Museum of Natural History created its new Automatic Data Processing program, demonstrating a clear commitment by the museum to take this research and development further (Bridge 2005). Shaped by Reginald Creighton, the next more fl exible iteration of the Smithsonian’s system – SELGEM4 – described itself as a collection of general purpose programs developed for information management. ‘SELGEM can handle any type of information about anything’, so one user manual explained, ‘provided that it is organised according to some basic rules’.5 By the mid-1970s the system consisted of about twenty-fi ve computer programs written in COBOL. To run, it needed a minimum of four tape drives, 20,000 characters of core and a COBOL compiler. The Smith sonian was using a Honeywell 2015 computer, though other institutions running SELGEM used platforms such as IBM 360, CDC 3100, CDC 6400, UNIVAC 1110 and GE635. Once data had been submitted to the system (either through keypunch worksheets, or punched paper tape or typed according to conventions for machine-optic recognition of characters) the data could then be retrieved via indexes, occurrence tallies or even labels. From the USDA Forest Service to the University of Alaska, and from Florida State University to the University of Cape Town, by May 1972 the SELGEM system was being used in almost a dozen institutions across the US and beyond. By November 1974 this number had risen to around fi fty – with the Museum of Natural History itself, by then, boasting over 2 million specimens referenced through the system (Creighton and King 1974). A year later the number of institutions using SELGEM was over sixty, and included institutions as varied and as dispersed as Arizona State Museum, the Field Museum of Natural History, Maine State Museum and the Museum of New Mexico. With its early start, innovation, thoroughness, productivity and substantial federal investment, the Smithsonian was probably justifi ed, therefore, in claiming to government that it had ‘assumed leadership in this area’ (Smithsonian Institution 1967).
