Biomedical Computing: Digitizing Life in the United States

3 The LINC Revolution: The Forgotten Biomedical Origins of Personal Computing

3 The LINC Revolution The Forgotten Biomedical Origins of Personal Computing I n 1961, while the efforts of the National Institutes of Health to mathematize and reform the life sciences faltered, computer designer Wesley Clark was spending time he had taken off from his job at MIT’s Lincoln Laboratory to devise ways to turn the problem of biologists’ apparent incompatibility with digital computers on its head. Instead of transforming biologists into “computerniks ,” Clark hoped to transform computers to meet the particular needs of the study of life. In accommodating biologists, Clark would build a machine, the LINC, that defied and subverted the dominant conventions of computer architecture. Clark’s MIT colleagues were pushing for ever larger computers to be shared by ever larger numbers of users, while the small LINC put an entire computer at the disposal of a single user. As the first small, programmable computer to combine visual presentation with the ability to manipulate images in real time, and as the first system to employ small, pocketable storage media, the LINC was a progenitor of both the minicomputer and the personal microcomputer.1 To understand how efforts to computerize biology and medicine lay at the The LINC Revolution 125 heart of the development of small, interactive computers, and how those efforts precipitated one of the most important developments in the history of computer design, we must follow Clark from his days as a University of California, Berkeley, dropout to his career as the architect of the enormous TX computers and finally to his break from the world of large-scale computing.2 During the 1950s and 1960s, biologists’ priorities informed Clark’s design by providing him insight into the problem of how computers could interface with science. Their work also served to test many of the key concepts underlying real-time interactive computing. Furthermore, biologists and their NIH patrons provided Clark with crucial institutional support when his ambitions to provide researchers complete control over small interactive computers ran counter to the prevailing interest in time-sharing. A Whirlwind Tour of Early Computing at MIT The journey of Wesley Allison Clark (b. 1927) to biomedical computing began in 1949, when the young Californian was struggling to find new direction during a leave of absence from the UC Berkeley Department of Physics after “a bruising experience in a seminar with [ J. Robert] Oppenheimer.” Further impetus for departure from graduate school came when Clark realized that “I really wasn’t interested in how the physical world works and how we understand it.”3 Instead of studying for preliminary examinations, he found himself whiling away his days in the department’s model shop, where he had developed an interest in machining metal and had even gone so far as to construct a small lathe for himself.4 During his leave, Clark used his background in physics to obtain a job as a technician at the Nucleonics Department of the General Electric Company, which ran the Atomic Energy Commission’s Hanford Works weapons facility in the desert of eastern Washington state. While waiting for his security clearance, Clark was allowed access only to the site’s paltry collection of unclassified reading material, which included a detailed description of Howard Aiken’s Mark II (a five-ton electromechanical digital computer built in 1947–48) and Edmund Berkeley’s Scientific American article on Simon, an extremely simple computer that performed binary additions—it could add 0 + 0, 0 + 1, 1 + 0, and 1 + 1— using magnetic relays.5 Although Clark devoted most of his spare reading time to the Mark II description, he “became thoughtful” about the tiny Simon, which 126 Biomedical Computing despite its very limited capacity as calculating machine, broached “serious possibilities ” that could be embodied by a machine that was “of a scale significantly smaller than the Mark II.”6 Upon receiving clearance, Clark found his job at Hanford less than optimal. His work was “concerned principally with the design and analysis of neutron experiments using the Hanford Test Reactor and other critical assemblies.”7 This included nerve-wracking tasks such as monitoring nuclear reactors, and physically dangerous work such as handling enriched uranium. Moreover, after reading about the Mark II and Simon, he was disappointed by the facility’s punch card machines.8 Resolving to learn “computerology,” Clark approached the mathematician and numerical analyst Derrick Henry Lehmer to see if he could return to Berkeley to work on...