Building IBM: Shaping an Industry and Its Technology by Emerson W. Pugh

Abstract

TECHNOLOGY AND CULTURE Book Reviews 385 is no attempt at theoretical reflection; the work remains empirical throughout. A more complex study on the subject, promised by the title—“science between university and industry” in the Nether­ lands—has yet to appear. But this book nevertheless provides a good starting point. Karel Davids Dr. Davids, professor of economic and social history at the Vrije Universiteit Am­ sterdam, is the author of Zeewezen en wetenschap: De wetenschap en de ontmkkeling van de navigatietechniek in Nederland tussen 1585 en 1815 (Amsterdam, 1986). Building IBM: Shaping an Industry and Its Technology. By Emerson W. Pugh. Cambridge, Mass.: MIT Press, 1995. Pp. xvi+405; illustra­ tions, tables, notes, appendixes, index. $29.95 (cloth). The first third of this excellent corporate history offers a detailed, if familiar, account of IBM before World War II. In the mid-1880s, Herman Hollerith invented electromechanical punched-card tabu­ lating equipment to aid in tallying the U.S. census. His Tabulating Machine Company then marketed this equipment to businesses. Hollerith’s strong patents—he had been a patent examiner—forced competitors either to use purely mechanical techniques or to license from Hollerith. These patents, his policy of renting rather than sell­ ing equipment, and his requirement that customers purchase punch cards only from Tabulating Machine made for an extraordinarily robust business. Emerson Pugh argues convincingly that IBM’s be­ ginnings should be dated directly to Hollerith’s enterprise. A 1911 merger created the Computing-Tabulating-Recording Company (CTR) and made Hollerith a millionaire. By 1915 Thomas J. Watson, a former National Cash Register (NCR) salesman, had become CTR’s president; he changed its name to International Busi­ ness Machines in 1924. No engineer, Watson nevertheless had an unmatched eye for a saleable product and proved a highly effective (if paternalistic) executive. During the Depression, the rental-only policy dating from Hollerith proved a godsend, providing steady, relatively insulated income. The 1935 creation of the Social Security Administration, with its gigantic data-processing burden, brought IBM vast new revenues. IBM equipment also found myriad uses dur­ ing World War II, ranging from record keeping to codebreaking. Watson’s chief engineer, James Bryce, was a prodigious inventor who personally acquired over 500 patents. Bryce and his colleagues patented new devices and techniques as a matter of routine, not just when they seemed marketable, thereby generating a reserve of technological resources and simultaneously erecting ever-higher hurdles for IBM’s competitors. In the postwar era, IBM set the pat­ tern for other high-technology enterprises by reinvesting nearly half its net profits in R&D. 386 Book Reviews TECHNOLOGY AND CULTURE The remainder of Building IBM treats IBM’s entry into the com­ puter era, and here Pugh makes significant new contributions. Pugh argues that IBM—including Watson, generally pictured as a hide­ bound proponent of aging technologies—moved smoothly and rap­ idly from electromechanical to electronic equipment after the war. Watson opposed the anthropomorphic word “computer” because it implied that the new machines might stealjobs from human com­ puters. Therefore, IBM’s early computers were all called “calcula­ tors,” and Pugh suggests that this terminology is partly responsible for the mistaken view that IBM was slow to develop true computers. He claims that inJanuary 1948 IBM’s Selective Sequence Electronic Calculator—a bizarre hybrid incorporating vacuum tubes, relays, and paper-tape reader-punches—became “the first operational computer to satisfy the modern definition of a stored-program com­ puter” (p. 136), months before the Manchester University Mark I, usually so honored. While most historians designate the 1952 IBM 701, backed by Tom Watson,Jr., as IBM’s first commercial computer, Pugh views IBM’s 1949 Card-Programmed Electronic Calculator as functionally comparable to the 1951 UNIVAC. If true, then rather than being “scooped” by its competitor, IBM computers far outsold UNIVACs in the early 1950s. However, Pugh apparently defines “electronic” computing rather broadly—admitting some compo­ nents, such as relays, often classed as electromechanical—in order to claim an early lead for IBM. Subsequent chapters hold fewer surprises. Topics covered are IBM’s early efforts in computer languages, the SAGE air defense sys­ tem (IBM’s largest contract of the 1950s), its Stretch supercomputer...