How much should we know about the underlying structure of a technological artifact in order to understand its history? Quite a bit, according to the authors of The First Computers: History and Architectures. This book, a collection of papers presented at the International Conference on the History of Computing in 1998, is aimed at computer scientists and programmers as well as historians of science and technology. The term “architecture” is used in computing to refer to the structure and capabilities of (...) a computer; it includes both logical design and physical design . Architecture is an abstraction that helps the authors make comparisons between computers from different countries and eras.The First Computers brings together a wealth of information on computers from the 1940s and 1950s, which the editors present in five main sections. Part 1, “History, Reconstructions, Architectures,” addresses general themes in the history of computing. Michael Mahoney's chapter, “The Structures of Computation,” examines the relationship between computer science and “pure” mathematics. Even in its most theoretical aspects, Mahoney notes, computer science has always focused pragmatically on feasible implementations of mathematical ideas, thus “blurring commonly made distinctions among science, engineering, and craft practice” . Robert W. Seidel examines the rationale for reconstructing historic computers, an activity that has engaged many of the book's contributors. Seidel notes that “as the artifacts of modern computing become invisible, older, larger computers supply a symbol of computing to practitioners, the public, and patrons which is not only visible, but comprehensible” .The four remaining parts cover computing developments in America, Germany, Britain, and Japan. The American computers include the Atanasoff‐Berry Computer, Aiken Mark I, ENIAC, and the Institute for Advanced Study computer. The section on Germany contains several chapters on Konrad Zuse as well as on such developments as the Göttingen machines, which may not be that well known. Britain is represented by the Colossus, Manchester Mark I, Atlas, and EDSAC. Perhaps least familiar to historians are the Japanese computers: Seiichi Okoma surveys nine early Japanese computers, and Eiiti Wada describes the PC‐1, which he helped build. Many of the chapters, especially those written by engineers, begin with a brief history of the machine followed by a detailed technical description of its design and operation. The contributions by historians tend to describe the computer's features in more general terms and discuss the intellectual, organizational, and economic context in which the machine was built.Because so many of the authors are technically trained, they are able to offer in‐depth information on the theoretical underpinnings and design of early computers. But often the discussion is difficult to follow, even for historians working in this area. Tony Sales's chapter on the Colossus, a British code‐breaking computer, exemplifies the strengths and weaknesses of accounts written by technical specialists. Sales begins with a clear and engaging narrative of the origins of the Colossus, its method for decoding messages, and its significance to the Allied effort in the Second World War. He then veers into a highly technical discussion that will likely be opaque to readers unfamiliar with terms such as “thyratron rings.” The final section is a personal story of Sales's successful effort to have the machine rebuilt. Like the book as a whole, Sales's chapter can be read for a quick introduction to a milestone in the history of computing or for insight into the meanings embedded in technological artifacts—but readers should be prepared to decipher some formidable technical jargon along the way. (shrink)