The Computer in the Garbage Can: Air-Defense Systems in the Organization of US Nuclear Command and Control, 1940-1960

Abstract

During the late 1950s, the United States Air Force initiated development on nearly two-dozen military “command and control systems.” What they shared in common was a novel application of digital electronics to the problem of nuclear warfare. Most of these systems descended, in some fashion, from a program called “SAGE,” the Semiautomatic Ground Environment, which gathered data from a network of radar stations for processing at large Air Defense Direction Centers, where digital computers assisted human operators in tracking, identifying, and, potentially, intercepting and destroying hostile aircraft. Although histories of SAGE have been written before, they have tended to stress digital computing as a rationalist response to the threat of mass raids by nuclear-armed Soviet bombers. Nevertheless, organizational sociology suggests that large bureaucratic organizations, such as the United States Air Force, often defy our intuition that decisions, technological or otherwise, must follow a perceived problem to its potential solution. According to the so-called “garbage-can model of organizational choice,” problems and solutions may, in certain circumstances, arise independently and join together unpredictably, because the basic social phenomena do not conform to bureaucratic ideals. This dissertation argues that SAGE, and indeed, the entire Cold War project of nuclear-and-command, can be understood as a sequence of “garbage-can-like” decisions, resulting in a conglomeration of independent systems whose behavior appeared reasonable from the perspective of the using organization, but which nonetheless failed to cohere against the far greater danger of a global thermonuclear exchange. They did, however, succeed at satisfying the government’s need to act by projecting uncomfortable questions of political organization onto popular technology programs.