The Transistor, an Emerging Invention: Bell Labs as a Systems Integrator Rather Than a ‘House of Magic’

Abstract

The transistor is one of the most consequential human inventions with dissemination of the eventual MOS-FET design estimated to exceed one quintillion devices. However, the transistor’s genesis remains poorly understood. Many received accounts associate transistor invention closely with a small group of Bell Labs scientists during the 1947-1948 period. This paper argues that such a view is too narrow. Rather, the transistor – as a solid-state amplifier – emerged over a period of several decades starting with early observations of anomalous amplification in semiconductor crystals and early device designs during the 1910s and 1920s. Other types of relevant knowledge evolved in the form of advances in solid-state physics and materials processing techniques during the 1930s and early 1940s. Bell Labs identified, absorbed, evaluated, and integrated such diverse but interrelated knowledge streams – making Bell Labs appear much more like a systems integrator than the prototypical closed innovation organization it is often portrayed as. The Bell Labs transistor effort was both mission- and device-oriented with the specific goal of turning existing but imperfect solid-state amplifier designs into reliable substitutes for vacuum tubes – as such the research program is better described as applied industrial research rather than basic research. Through its systems integration activities, Bell Labs catalyzed a qualitative shift in the hitherto fragmented semiconductor field, enabling greater resource allocation and intensified research activity, as reflected in a hike in publication growth rates and the later introduction of marketed products. Thus expanded research activity eventually led to the 1959 MOS-FET design as the transistor’s dominant design used in large-scale dissemination such as in modern computer chips. Consequently, I propose to view the transistor as an emerging invention – in contrast to a discrete or singular one – with an emergence period spanning several decades. I propose to distinguish between an exploration phase (~1920-1945), a consolidation phase (1945-1950), and a maturation phase (1950-) whereas the intermediate consolidation phase represents a topological transition as is characteristic of emerging fields. Another emphasis of this article lies on the role of informal knowledge in the invention process. In the transistor case, such informal knowledge included patent specifications with proposed device designs, amateur radio magazine articles with reported anomalies, and oral anecdotes in practitioner circles describing experimental configurations of interest. This research asserts that such kinds of informal knowledge played an important role early on in the invention process as they guided both early research campaigns and managerial decisions, including at Bell Labs.