Computer Science & White-Collar Work

A review of Post-Industrial Engineering: Computer Science and the Organization of White-Collar Work, 1945-1975, by Andrew Mamo.

How might historians contextualize the way that the differently constituted groups of society conceive of computers and computing, without falling into the trap of understanding the development of computing machines as the culmination of our search for pure rational thought from Plato to Descartes to Leibnitz to Babbage — as several computing pioneers themselves did? One way might be to simply follow the money, akin to Paul Forman’s work on post-War physics, namely showing that the money for most computing developments in the post-War world had one main source: the military. Another might be what one might call the Paul Edwards way: to point out how the metaphors embedded deep inside computer systems are similar to the discourses of the Cold War in the United States. (The Closed World: Computers and the Politics of Discourse in Cold War America. Cambridge: The MIT Press, 1997.)

In his dissertation, entitled Post-Industrial Engineering: Computer Science and the Organization of White-Collar Work, 1945-1975, Andrew Mamo takes a different track: pointing to the relationship between computing systems, “management science” and ideas about automation and expertise in the workplace. This may not seem like a startling vantage point today, given that computers are ubiquitous in every workplace. But Mamo’s point is that these relationships were present during the post-War birth of computing too, despite the fact that computers were primarily viewed as calculating devices. In line with Simon Schaffer’s work, which shows that Charles Babbage’s nineteenth-century calculating machines were modeled on the organization of the factory (“Babbage’s Intelligence: Calculating Engines and the Factory System.” Critical Inquiry 21: 203–227), Mamo suggests that post-War management science and computer science were preoccupied with the same question: how do we understand human decision-making processes (in organizations or otherwise) and then express these processes in terms of rules that everyone may then blindly follow?

The personification of this intertwining of computer and management science was Herbert Simon, the Nobel Prize-winning social scientist, and one of the pioneers of Artificial Intelligence (AI). Simon invented the General Problem Solver (GPS) system in the 1950s, one of the first instantiations of the Planning hypothesis in AI — that human behavior is the “implementation” of hierarchically structured “plans” inside people’s heads. Yet, Simon’s background was in management science; his first book, published in 1947, was Administrative Behavior, in which Simon’s key insight was that the rationality of human beings is constrained by a number of factors, and therefore, institutional structures and scaffoldings must be in place to aid human decision-making in organizations. Simon was a professor at Carnegie’s new business school called the Graduate School of Industrial Administration (GSIA), one of the first “New Look” management schools of the 1950s (the topic of Mamo’s Chapter 1). These schools fused the study of management practices with rigorous engineering methods and techniques. Simon’s move to computer science came about when he visited the RAND Corporation in 1952 to study individual decision-making in aerial defense warning stations and became fascinated with the computers being used. Mamo suggests that this happened because he saw computing systems as analogous to rationalized bureaucracies (and not to individual human beings), and therefore a potential tool for studying information processing and decision-making in organizations. How Simon could be simultaneously a proponent of bounded rationality of human beings and a proponent of strong AI and Planning as a model for human thinking (in his later years) is a paradox worth pondering over (p. 139).

In Chapter 2, Mamo surveys the different histories of computing. He highlights the role of Edmund Berkeley who saw the possibility of using computing devices in the insurance industry (where he worked) and became a tireless promoter of machines and automation. Berkeley went on to become the founding secretary of the Association of Computing Machinery (ACM), the key body in the computer sciences today. Mamo also explores, suggestively, the link between cybernetics and managerialism (pp. 64-77): once the idea of information traversing around feedback loops became central, the back-office could be increasingly seen as the “brain” of the organization.

Chapter 3, the first of three case-studies, looks at the development of time-sharing systems at MIT. Mamo suggests that there were three ideas about how to interact with computers: batch-processing programs where only technicians (usually women) had access to the computers; time-sharing systems that allowed two-way communication between many users and a single computer by optimizing the use of computing power; and “personal computing,” a radically, decentralized approach. All three were also ways of conceiving bureaucracies. Chapter 4 looks at the fights over the notion of Planning in AI. Whether “plans” were the right category for describing individual human actions was a topic that was hotly debated between cognitivists like Herbert Simon and humanist critics like the philosopher Hubert Dreyfus. But surprisingly, management guru Peter Drucker detested AI and Planning as well; he saw these ideas of the computer scientists as too formalist in what he considered their (normative) descriptions of managerial practice. Finally, in Chapter 5, Mamo looks at the fate of the Cambridge Project which sought to standardize the use of computing systems for the behavioral social sciences. The emerging New Left protested against the creation of the centralized database that the project envisioned. The era of the acceptance of rationalized bureaucracies (aided by computing technology) as useful mid-points between laissez-faire capitalism and communism, Mamo suggests, was coming to an end.

The connection between computing systems and managerial practices in the workplace is an extremely rich one. Mamo’s dissertation engages usefully with scholarship in the history and anthropology of computing. For those interested in the relationships between business, computing and the business of computing (i.e., IT or Information Technology), this dissertation serves as an excellent starting point for further analysis.

Shreeharsh Kelkar
Massachusetts Institute of Technology

Primary Sources

Charles Babbage Institute Archives
MIT Archives and Special Collections
Carnegie Mellon University Archives
Library of Congress Archives

Dissertation Information

University of California, Berkeley. 2011. 222pp. Primary Advisor: Cathryn Carson.

Leave a Reply