Historical Reconstructions

During the last two decades an increasing number of initiatives has sought to construct or reconstruct historic computing devices, and restore original specimens to working order. The feature articles in this issue describe six such projects written by those who delivered the successful outcomes or who were instrumental in their realization. In each of the four reconstruction projects described the original historic machine was not available as a reference datum: It either did not survive or it never existed. Thomas Fowler’s ternary calculator (1840), Konrad Zuse’s Z3, the Manchester Baby, and the Bletchley Park Colossus were broken up for spares, destroyed, or lost. In the case of Charles Babbage’s Difference Engine No. 2, designed in the late 1840s, there was no contemporary original. The first parts were not made until 1986. These new/old machines pose intriguing questions. To what category of object do these machines belong? Are they fictitious antiques or new primary sources? What is the historical purpose of such constructions? Is there anything we can learn from them that we could not otherwise learn? In the case of restoring a surviving machine to working order, what are the limits of legitimate intervention in modifying or replacing parts? What status does a restored original have as an evidentiary source? Without exception, the sources that the projects used to inform the constructions are incomplete. In the case of Fowler’s ternary calculator described by Mark Glusker, David Hogan, and Pamela Vass, there are no drawings or original parts. The only visual clues are found in a stained glass window in a church. The modern realization of the machine is based on partial descriptions found in contemporary (19th century) textual accounts. The Fowler project represents a remarkable blend of archaeology, history, detective work, and engineering. In reconstructions, constructions, replication, and restoration there is often the unquestioned assumption that the new machine should resemble the original in all respects. Raul Rojas, in his article about the reconstruction of Konrad Zuse’s Z3, challenges this wisdom. In seeking to make the workings, operation, and design of the machine intelligible to modern generations, Rojas deliberately untethers the new machine from the traditional constraints of physical fidelity to the original. Dag Spicer chronicles the restoration of an IBM 1620 at the Computer History Museum in Mountain View, California, and addresses museological issues related to restoration. He prompts us to ponder the question: What serves history better, a nonworking machine intact in electrical and physical detail, or a restored working machine but with nonoriginal parts or internal modification? Christopher Burton’s account of the reconstruction of the Manchester Baby, which first ran in 1948, offers us a master class in systematic reconstruction with meticulous attention to sources and issues of authenticity. His account of the role of the original engineers and practitioners in authenticating aspects of the reconstruction, and the power of the new machine to prompt recovered memories in the original pioneers, is both sobering and touching. Tony Sale’s account of his efforts, against formidable odds, to create a functional reconstruction of Colossus is a case study of detective work, stamina, and ingenuity. More than anything, it is testimony to an unshakeable determination to succeed. My own article describes the construction, completed in 2002, of Babbage’s Difference Engine No. 2. This issue serves, in part, as a tribute to the astonishing levels of commitment, technical expertise, political activity, and detailed historical excavation by those responsible for producing anew machines that memorialize the efforts of past generations of pioneers.