AI & SOCIETY, VOL. 2, 245-255 (1988) © 1988 Springer-Verlag London Limited

The Closed World: Systems Discourse, Military Strategy and Post WWII American Historical Consciousness

Paul N Edwards Silicon Valley Research Group, Stevenson College, University of California, USA

ABSTRACT

This essay proposes a cultural and historical explanation for the American Military's fascination with computing. Three key elements of post-WWlI US political culture - apocalyptic struggle with the USSR, subsuming all other conflicts: a long history of antimilitarist sentiment in American polities; and the rise of science-based military power -- contributed to a sense of the world as a closed system accessible to American technological control. A developing scientific systems discourse, centrally including computer science and AI, was adopted for strategic thinking and military technology. The Strategic Computing and Strategic Defense Initiatives are discussed as contemporary examples of this conjunction.

KEYWORDS

artificial intelligence, computer systems, military systems, SAGE, SCI, SDI, strategy, discourse, systems

1.0 INTRODUCTION

Until the early 1960s, the military forces of the United States were the single most important influence on the development of digital computers. Their research organiza- tions, both public and private, sponsored technical progress. Military use of many kinds of computer systems provided both a proving ground and a market-place.1

After the emergence of mature commercial computer markets by about 1960, US military agencies (and military-related agencies such as the Atomic Energy Commission and NASA) continued to invest heavily in advanced computer research, equipment and software. Conceived in the mid-1950s, artificial intelligence was nurtured by the Advanced Research Projects Agency (ARPA, now called DARPA) throughout the 1960s and 1970s, as were such other important innovations as timesharing and large-scale computer networks. During those decades, however, the Department of Defense (DOD) allowed commercial manufacturers and independent university researchers to

246 Paul N Edwards

take the lead in most areas of computer research and development. In the early 1980s even relatively esoteric AI began to find major backing in the commercial world. Starting in 1983, with the Strategic Computing Initiative (SCI) and the Strategic Defense Initiative (SDI), the Pentagon began a conscious attempt to regain control over certain critical fields of leading-edge computer research in America, and to guide them in particular directions favorable to its own larger goals. 2

Why has the American military maintained such a deep and intimate involvement with computer research? The most obvious answer is that computers are simply the right tools for important military tasks. The speed and complexity of high-technology warfare have created control, communications and information analysis demands that completely defy the capacities of unassisted human beings. Computers improve military systems by 'getting man out of the loop' of these critical tasks -- tracking the enemy, analyzing data from remote electronic sensors, piloting missiles, handling complicated command- control-communications (C3I) networks and many others.

There are tradeoffs between complexity and efficacy in all technological systems, and computer-based military technology is no exception. For example, the US BMEWS (ballistic missile early warning system) has produced a number of relatively serieus false alerts due to hardware, software and operator errors. Because of its complexity, detection of these errors can be quite difficult, thus raising the level of uncertainty about the correct interpretation of an alert -- and the stakes, obviously, are very high. Accidents traceable to computing problems are by no means uncommon and they have had real, and occasionally lethal, effects in international incidentsi 'In a 1977 test of the Worldwide Military Command and Control System (WWMCCS), attempts to send messages failed 62 per cent of the time. One part of the network, the Readiness Command, broke down 85 per cent of the time. Such communications breakdowns have been implicated in the capture of the USS Pueblo by North Korea in 1968 and in the Israeli bombing of the US intelligence ship Liberty during the 1967 Six Day War'. 3 Many other examples could be cited.

But on the whole computerized systems have been highly beneficial in military terms. They act as 'force multipliers', reducing the numbers of necessary combat personnel. In general computers have limited rather than expanded the chances for human error, made possible more sophisticated and accurate intelligence information and increased military capacities for speed and coordination. An explanation in terms of automation, com- munication and coordination partly accounts, then, for theUS military's historical drive to computerize many elements of its forces.

But utility alone explains neither the urgency nor the magnitude of the US effort in computing. From the end of World War II until the early 1950s Great Britain's sophistication in computing at least equalled that of the US. Yet the British military failed to pursue the field with the intensity of the United States and subsequently fell far behind. Nor does it explain the pervasive fascination with computers that may be detected in many elements of the US armed forces. Though usually restrained in the event, this fascination has repeatedly led Defense Department planners to overestimate the capacities of computers and to contemplate for them entirely inappropriate roles, for example as an ultimate decision system for nuclear defenses or as natural language interfaces to computerized weapons systems. 4 Beyond the relationship of tool to task, there is a peculiarly American history here, one that bears telling because of the light it sheds on certain contemporary developments in AI on the American political scene.

The historical and cultural analysis sketched in this article thus seems to me to provide

The Closed World: Military Strategy and Post WWII American Historical Consciousness 247

an accessory, but indispensable explanation of US military involvement with computers for two reasons. First, in the high technology world of the 20th century the technologies available to a military force both influence and reflect its internal sense of its own capabilities. As the foundation of force, military technology has profound political effects, since it largely defines the ultimate limits and parameters of national power. For instance, the origins of the concept of political 'superpowers' are directly linked to the technology of atomic weapons. Thus the worldview of military institutions and their technological choices are reciprocally bound up. In other words, the tasks assigned to the US military by the political process determine the types and quantities of technology it develops and deploys. But, at the same time, the available technologies also affect which assignments it believes itself ready to accept. US foreign policy, military strategy and military technology have evolved in a tightly linked partnership.

The second reason is that it is possible to trace a generalized influence of computers on US strategic thinking through what I will call 'systems discourses'. A 'discourse' is the social product of a cultural subgroup. It involves a language, a set of shared values, a worldview and a political or institutional structure. Systems discourse, then, names the language, worldview and institutional basis of the broad postwar scientific movements toward mathematical formalization of complex interactive systems. Computerized control systems crystallized the power of systems thinking, and computers themselves became an icon of the approach to the world as a formal machine subject to manipulation and control. My thesis is that post-WWlI US strategic discourse had -- and still has -- deep affinities with systems discourse that drew the two together. Their confluence was most prominent in 1950s strategic planning and during the Vietnam War. 5 But with SDI and SCI the computer influence on strategic thought is undergoing a renaissance.

2.0 TRANSFERENCE AND APOCALYPSE

World War II was a 'good war', justified in the eyes of most US citizens as a war not only against greedy aggressors, but against an antidemocratic ideology as well. This nearly universal sentiment was vindicated, after the war, by the revelations of Nazi atrocities. But the aura of Biblical struggle that surrounded the fight against Nazism and Italian fascism did not simply fade away after the war. Soviet maneuvering in Eastern Europe, as well as the openly expansionist Soviet ideology, provided a means for a transition into a Cold War. Stalin was equated with Hitler; communism replaced fascism as a kind of total enemy. It was seen not just as another human order with its own goals and interests, but as a completely inverse and implacable system, opposed to the US on military, political, ideological, religious and economic grounds.

This transference also carried with it the WWlI sense of a global and total scale of conflict. The partitioning of Europe, revolutionary upheavals across the post-colonial world and the contest for political and ideological alliances throughout Europe, Asia and the Middle East encouraged American perceptions that the world's future balanced on a knife edge between the US and the USSR. The Truman Doctrine of worldwide American military support for 'free peoples who are resisting attempted subjugation by armed minorities or by outside pressures '6 codified the continuation of global conflict on a permanent basis and elevated it to the level of 'a struggle between light and darkness. . a universal conflict between freedom and slavery'. 7

As the only combatant nation to come out of the war more or less unscathed, the US

248 Paul N Edwards

also emerged from the economic depression and political isolationism of the 1930s into a sudden and unaccustomed role as world leader. The only immediately available models for action in this new role were those of the war itself. The key events of the war thus became basic icons in the organization of American foreign policy and military strategy, from Munich (the danger of appeasement) to Pearl Harbor (always be prepared for surprise attack) to Hiroshima (victory through technologies of overwhelming force). The unfolding political crises of the Cold War were always interpreted in those terms: Berlin as another Munich, USA forward basing of nuclear weapons as the risk of another Pearl Harbor, the growing nuclear arsenal as a reminder of Hiroshima, horror and symbol of ultimate power. Thus, in many respects, the Cold War was not a new conflict with communism but the continuation of WWII, the transference of the apocalyptic struggle onto a different enemy.

3.0 AMERICAN ANTIMILITARISM

Another element of the postwar cultural situation was the long history of antim~litarist sentiment in US politics. The authors of the American Constitution drew a distinction of principle between national self-defense and the unruly European standing armies of the 17th and 18th centuries. Victors in a war fought by future citizens, but anxious not to recreate the social abuses of the British armies long quartered on American soil, they understood both the importance of military power in international conflict and the dangers it posed in domestic political life. Later generations carried with them this ambivalence toward military force -- a necessary evil, but also a source of patriotic pride.

American antimilitarism, then, is not the same thing as pacifism, or principled objection to armed force itself. Instead it derives from essentially political motives having to do with the location of power in a democratic government. It is an instance of what the conservative political scientist Samuel Huntington calls the 'anti-power ethic' in American society, the enormous value it has always placed on political limits to power, hierarchy and authority, s Historically, prior to World War II, what America had strenuously avoided was not so much war itself as the permanent presence of a powerful national military institution.

The disappointing aftermath of World War I, when dreams of world unity through a final defeat of nationalist aggression were shatterd, added force to the antimilitarist impulse. Americans responded to the failed efforts at world political cooperation with withdrawal and wishful thinking. Between the wars 'most Americans believed in a natural harmony of interests between nations, assumed that there was a common commitment to peace, and argued that no nation or people could profit from a war'. 9 In 1941 Japan and Germany forced a reversal of these ways of thinking. The Allied military success, the smooth transition from World War II into the Cold War, the occupations of Germany and Japan, the relative insignificance of American suffering and America's full-blown emergence as a world power all contributed to a rapidly changing national perception of the need for a large armed force. Technological factors, such as the bomb and the maturation of air warfare, also made it possible to conceive of a new military role for the United States in world affairs outside its traditional North American sphere of influence. Thus the Cold War marked the first time in its history that America maintained a large standing army in peacetime. But the profound historical roots of

The Closed World: Military Strategy and Post WWII American Historical Consciousness 249

American antimilitarism ensured that the institutional form taken by a more vigorous American military presence would differ from the more traditional European and Soviet approaches of large numbers of men under arms.

4.0 SCIENCE AND MILITARY POWER

A third component of the postwar political/cultural situation was the rise of 'big science' and engineering, linked to the increasing prestige, presence and ambitions of the military. Science and engineering were, it was widely believed, largely if not entirely responsible for the ultimate Allied victory. The defeat of the Luftwaffe at the Battle of Britain was credited to the crash development of radar. British computers cracked the ciphers of the German high command. American radar and anti-aircraft technology played crucial roles in the Pacific war and in the defeat of German V-1 attacks on London. Finally, of course, the atomic bomb represented the military apotheosis of science. In the eyes of many these weapons -- and their creators -- became part of the war's pantheon of heroes.1° This coincided with a more general increase in the number, prestige and activity of

engineers in America through the science-based industries that rose to prominence just before World War I, such as the chemical, electronics and telephone industries. Engineering academies like MIT and Cal Tech played major roles in the war effort, vastly increasing their base of political power and prestige in the process. American culture had always valued pragmatic enterprise. During the war, the engineers got a chance to show what they could do given virtually unlimited resources, and it was indeed impressive.

The postwar consequences were the emergence of a powerful and self-conscious science and engineering lobby and a permanent governmental association with science, largely mediated through the military services. The wartime science/engineering com- munity, organized around military problems, had been an exhilarating experience for many of those involved. It had provided invigorating interdisciplinary contacts, vast amounts of money, equipment and research time, and the euphoria of rapid success. It had also opened up a multitude of new and intriguing research problems. The postwar scientific community therefore enjoyed an uprecedented s e n s e of community, and its wartime miracles had opened the ears of political and military leaders to its voice.

5.0 THE COMPUTER AGE

All of this occurred simultaneously with the dawn of the computer age. Though described and actually invented just before WWlI, it was the war effort which provided the real impetus to build the first digital computers. In Great Britain the Colossus computers assisted in crucial code-breaking efforts, while the US effort centered around the ENIAC, built to automate the calculation of ballistic firing tables. The ENIAC was not completed until after the war. Nevertheless, the first problem posed to it was a physics equation from the Los Alamos atomic bomb laboratories. The hydrogen bomb could not have been built without ENIAC's successors.

Military and other government support for computers, then seen primarily as scientific

250 Paul N Edwards

research devices, was the main driver of ongoing computer development until the early 1960's, when a fully viable commercial computer industry first emerged. When Vannevar Bush's proposed civilian National Science Foundation ran into political stumbling blocks after the war, the Office of Naval Research filled the gap, funding upwards of 40 per cent of all US scientific research by 1948. Military funding backed the development of the transistor. The armed services provided the exclusive market for integrated circuits for several years after their invention in 1959. Through the SAGE air defense project, the military was responsible for several generations of advanced computer development during the 1950s.11

6.0 POST-WWII HISTORICAL CONSCIOUSNESS: THE WORLD AS A CLOSED SYSTEM

Perhaps the most significant feature of the bilateralized world that rose from the ashes of the war was its truly global character. World War I, despite its worldwide staging ground, was essentially the last in a long series of local European wars. But the import of the partitioning of Europe after World War II was the sense, in America, that the world was now closed, fully occupied by the apocalyptic struggle between the American and Eurasian super powers. This perception was soon reinforced by the US intervention in Korea and by the other Asian upheavals of the forties and fifties. Under the Truman Doctrine and the Marshall Plan, the world had become a system to be protected -- and manipulated -- by the United States government. In the quasi-religious American mythos, no ideological space remained for other conflicts. Bilateralism created a systematic vision of the world by making all third-world conflicts parts of a coherent whole, surrogates for the real death struggle between the Free World and its communist enemies. 12

Television and other technological changes, of course, added to the sense of the world as a 'global village'. High-speed air transportation annihilated distance -- and high-speed distance weapons made the threat of conflict more immediate. But it was the atom bomb, with the vast scale of conflict it implied, and space flight -- also born of the events of World War II -- that completed the sense of closure. The first photographs of the globe returned from space gave a new sense of reality to the vision of the world as a unit, a closed system, a spaceship. As the last of the terrestrial frontier vanished, America began to experience, perhaps for the first time, a sense of limits.

6,1 High Technology: Armed Forces for Global Conflict Military preparations in the post-war world both responded to and magnified the sense of closure and system. If Americans gained anything from World War II, it was the dual sense of awesome national power and compelling national responsibility. Yet in the relief that followed the war's end, US military force, the instrument of that power, was rapidly and thoroughly demobilized. By 1947 the US Army stood at one million men (down from eight million in 1945). At the same time the Red Army of the Soviet Union encompassed over 260 divisions. What remained of the decisive US role in WWII was technology -- especially the bomb.

The scientific and technological successes of the war, and the great distances involved in conflict with the Soviet Union, served both to justify and to exaggerate the focus on

The Closed World: Military Strategy and Post WWII American Historical Consciousness 251

high technology. The Los Alamos experience held out the promise of unlimited military power through American know-how. At the same time, the policy of containment 13 required an ability to intervene with force anywhere in the world. The conjunction of high technology, containment policy and demobilization led to the result, in the years immediately after the war, that American 'foreign policy and military policy were moving in opposite directions. Truman and his advisers wanted to meet the Communist challenge wherever it appeared, but except for the atomic bomb they had nothing with which to meet it . . . . Truman wanted a balanced budget and was enough of a politician to realize that the public would not support higher taxes for a larger military establishment.'~4

Antimilitarism, too, helped to focus strategic planning on technological solutions. The Strategic Air Command rose to prominence in the post-war defense establishment, despite bitter interservice rivalries, largely because it possessed and controlled the technological means for intercontinental nuclear war. It was the primary threat America could wield against the Soviets -- yet it required mainly money and equipment, not large numbers of troops. The Army's massive manpower seemed less impressive, less necessary and more of a political liability in the face of minimally manned or even automated forces.

But the technological threat of atomic war -- total, global war, without front lines, in which millions of civilians would perish -- brought with it demands for a total technological defense. In the 1950s, the Air Force developed the first in a series of responses to these demands, the SAGE air defense system. SAGE was a vast computer-integrated linkage of radars and tactical air defense squadrons designed to detect and overwhelm attacking Soviet intercontinental bombers before they could strike the continental US. It was the first large-scale demonstration of digital computers in a real-time military command and control operation. The system, finally completed in 1963, involved 23 air defense sectors, each with a central computing center capable of analyzing early warning radar data and assigning interceptor aircraft and anti-aircraft missiles to targets. 15

SAGE was necessitated by two factors: the sheer numbers of airplanes and quantities of data to be coordinated, and the very high speed of the anticipated attacks. But SAGE was also a response to the sense of global closure: it was a true system, an all- encompassing, totally integrated management of the entirety of continental US airspace. Its direct descendants, WWMCCS (World Wide Military Command and Control System) and the proposed Strategic Defense Initiative space-based ballistic missile defense network, carry on the SAGE vision of total defense and global oversight.

7.0 MILITARY STRATEGY AND SYSTEMS DISCOURSE

Thus the rapidly evolving geopolitical concerns of the US as a nuclear power shaped a strategic discourse involving a closed but complicated system, accessible to technological control. Very soon after the war computers were integrated into that discourse, both as tools in the solution of strategic problems and as models of the strategic situation itself. Computer scientists and those in related fields were developing their own, scientific discourse about systems, communication and control. It became widely known that digital computers were universal machines, theoretically able to solve any problem that

252 Paul N Edwards

could be formulated precisely enough -- i.e. that could be systematized, mathematized, modelled, reduced to an algorithm.

At the same time, disciplines such as cybernetics, information theory, game theory, systems analysis, operations research and linear programming succeeded in isolating algorithms for difficult problems in communication, social interaction and the control of highly complex systems. This extension of mathematical formalization into the realm of social problems brought with it a sense of new-found power, the hope of technical control of social processes to equal that achieved in mechanical and electronic systems. What I am calling 'systems discourse' included not only the formal techniques and tools of the systems sciences, but also this ideology of technical control.

Strategy, especially nuclear strategy, seemed to some military planners to fit this mould. At the RAND Corporation, the Air Force think-tank, game theory and systems analysis became central paradigms for strategic thought. Fred Kaplan has argued that RAND studies of the vulnerability of the Strategic Air Command's forward bases, carried out with the mathematical techniques of systems analysis, were the goad behind the US nuclear build up of the 1950s and 1960s. ~6 Time constraints in high technology war are so intense that pre-programmed responses are almost a necessity, and RAND claimed to be able to optimize war plans through its formal methods. RAND analysis also lay behind flexible-response strategy, the no-cities doctrine of counter-force, and other key elements of US nuclear strategy. Similarly, game theory and RAND systems analysis were largely responsible for the strategic thinking of Robert S. McNamara and his 'Whiz Kids' in the early years of the Vietnam War, when each military maneuver was conceived as a political message about the costs of continued war to the other side. Not by chance, RAND was also a center of advanced computer research throughout the 1950s and 1960s.

Within computer science artificial intelligence, notorious for its active imagination with respect to the potential of computers in human affairs, participates perhaps most fully in the characteristic technical and ideological elements of systems discourse. The 'microworld' approach to AI illustrates both. Microworlds are internally consistent, complete environments with their own private ontologies, simpler than the open universe of which they are a partial model. They are true closed worlds, in the sense that they cannot recognize or address entities or processes beyond those that are given or taught to them. The assumption of many AI workers is that ~the limitations of microworlds simply reflect 'inadequate complexity. Given enough processing power and a sufficiently sophisticated model, they believe that the microworld could simply be expanded to reflect reality itself.

Yet there are good reasons to believe that this is not the case. For most complex social interactions, it may be fundamentally impossible to achieve the kind of complete specification of entities and processes required by algorithmic problem-solving techni- ques. As Winograd and Fiores observe, 'the essence of intelligence is to act appropriately when there is no simple pre-definition of the problem or the space of states in which to search for a solution'.17 One of the primary criticisms of computer control of military systems is precisely this ontological closure. The performance of computer systems, unlike that of human experts, degrades very rapidly once entities or processes outside the programmed field of expertise are introduced.~8 Thus, for example, battle management computers based on expert systems are likely to suffer from the same inability to recognize unfamiliar objects that caused nuclear early warning systems to raise alerts at

The Closed World: Military Strategy and Post WWlI American Historical Consciousness 253

the rising moon and at flocks of geese. Their much greater sophistication may prevent gross errors like these -- but at the same time it opens such systems to failures much more subtle and difficult to detect.s 9

This objection to computerized weapons systems applies on a larger scale to all strategic thinking based on a systems approach. Part of the lesson of Vietnam was precisely that their 'simple pre-definition of the problem' had obscured the realities of the situation to US military and political leaders. This realization partially rectified the tendency to rely on formal analysis. 'Vietnam . . . exposed something seamy and disturbing about the very enterprise of the defense intellectuals. It revealed that the concept of force underlying all their formulations and scenarios was an abstraction, practically useless as a guide to action. '2°

The frustrations of the world's most highly computerized and technology-based war brought home -- at least to some -- the hard lesson that the closed world of military force could not prevail alone in the open-ended world of political struggle. Even Secretary of Defense McNamara 'regretted the "almost ineradicable tendency to think of our security problem as being exclusively a military p r o b l e m . . . We are haunted by this concept of military hardware"'. 21

Yet for others, the lesson was in fact the opposite: that US technology simply had not yet matured. Drawing on his Vietnam experience, in 1969 General William Westmore- land proposed the concept of the 'automated battlefield', a war zone 'under 24 hour real or near real time surveillance of all types . . . an Army built into and around an integrated area control system that exploits the advanced technology of communications, sensors, fire direction and the required automatic data processing'. 22 Computers were the heart of Westmoreland's vision, which survives in the SCI and the SDI. DARPA's Second Annual Report on the SCI stresses that 'commanders must have automated systems that can deal consistently and accurately with the complex interrelationships of the many variables inherent to battle management during combat and crisis action . . . . Machine intelligence t echno logy . . , will provide capabilities to dramatically reduce the time required for planning and monitoring operations, support the development of early offensive postures, identify sensitivities in key strategic and tactical decision, and demonstrate the implications of complex combinations of events and decisions' .23 As for the SDI, the original proposal for the SCI noted that in ' . . . the projected defense against strategic nuclear mi s s i l e s . . , systems must react so rapidly that it is likely that almost complete reliance will have to be placed on automated systems'. 24

The parallel between the microworlds of artificial intelligence programs and the blinkered strategic thinking of the Vietnam War points to the fundamental comparison I am trying to make in this essay. Computers are the basic tools of high technology war, without which command, control and communications under modern military conditions would be difficult or impossible. They have made it possible to link other technologies together into a smoothly functioning, coordinated system. But computers also represent a particular way of understanding reality through formal techniques. This style of thinking, crystallized in proposals for total defense such as the SDI, assumes the possibility not only of complete and consistent surveillance and comprehension of inherently unstable and dangerous situations, but of controlling them by technical means.

Seen in this light, terrorism and guerilla war, the major forms of conflict the US has had to face since WWII, are a logical response to systemic closure. They are a refusal to

254 Paul N Edwards

fight by the 'rules of engagement' , to accept the fixed set of entities and actions of a systems discourse. Their randomness, their failure to distinguish 'civilian' from 'military' targets, the ability of their perpetrators to fade into the general population, and the vast frustrations they have caused US armed forces across the globe are all marks of the realization that the military microworld is a closed system.

Systems discourse created the preconditions for the ascendancy of the rationalistic, bureaucratic, formal-analysis tradition of military strategic planning in the post-war years. In consequence an older, competing tradition of individual leadership, soldierly creativity in strategy, and heroism went into decline -- though it is far from disappearing. 25 This happened, not because of something inherently militaristic in systems discourse, but because systems thinking fits so neatly into the cultural niche formed by post-war American bilateralism, antimilitarism, technology-based military power and computerization. Strategic thought in the nuclear age has a suitability to formal modelling, a kind of cultural 'fit', that makes military fascination with computers more likely. Thus military interest in advanced computing stems not just from practical demands, but from the larger political discourse in which the military participates.

SDI and SCI, then, appear as the latest steps in a long post-war history of global system-building. The SDI has its ancestry in the SAGE system, the first to contemplate a high-technology, electronically based continental defense against nuclear attack. It is perhaps the ultimate example of the closed-world philosophy of defense, enclosing the entire globe inside a cloud of satellite sensors and battle stations and purporting to defend not only the US but potentially Europe and the Soviet Union as well. Here, naturally, attacks that escape the system's closure are the obvious response, and so many of these have already been proposed that the continued (but dwindling) Congressional support for the initiative is a puzzle.

The moral of this story is that real-world political and military situations are not microworlds. The lesson of history is that closed systems always have unforeseen and unforeseeable flaws: in the case of technology, Chernobyl and the Challenger are our latest reminders of this obvious but too often disregarded fact. A closed-world strategy, military or political, ignores the open-ended reality of human politics and the endless creativity of human beings, both in fighting and resolving conflicts. Systems thinking and artificial intelligence c a n be useful tools. But the sheer convenience of these limited modes of thinking should not continue to blind us to the as yet unformalized complexity of human affairs.

The Closed World: Military Strategy and Post WWlI American Historical Consciousness 255

REFERENCES

1 Flamm, K. (1987). Creating the Computer and Targeting Technology. Brooking Institution. 2 See Edwards, P. (1985). Border Wars: The Science and Politics of Artificial Intelligence.

Radical America. 19. Also see Edwards, P. and R. Gordon (forthcoming). Defense Research and High Technology,

in Edwards and Gordon (eds.) Strategic Computing: Defense Research and High Technology. Columbia University Press.

3 Jacky, J. (forthcoming). Software Engineers and Hackers: Programming and Military Comput- ing, in Edwards and Gordon op. cit.

4 See the many articles on this topic in the CPSR Newsletter, Computer Professionals for Social Responsibility, Palo Alto, California. Also see Pullum, G. (forthcoming). Natural Language Interfaces and'Strategic Computing, in

Edwards and Gordon, op. cit., available in a somewhat different form in AI & Society. 1,1. 47-58.

5 See Chapman, G. The New generation of High Tech Weapons, and Edwards, P., Computers and Weapons systems: An Historical Overview. Both in Bellin and Chapman (eds.) Computers in Battle. Harcourt, Brace, Jovanovitch. Also see Gibson, J. (1986). The Perfect War: Technowar in Vietnam. Atlantic Monthly Press,

and van Creveld, M. (1985). Command in War. Harvard University Press. 6 Harry S. Truman, cited in Ambrose, S. (1985). Rise to Globalism. Penguin. 86. 7 Ambrose, ibid. 87. 8 Huntington, S. (1981). American Politics: The Promise of Disharmony. Belknap Press. 9 Ambrose, op. cit. xiv.

10 See, for example, the lionizing of science and scientists in James Phinney Baxter's history of the Office of Scientific Research and Development, Scientists Against Time. (1946). Little, Brown.

11 See Flamm, op. tit., and Edwards, Computers and Military systems. Op. cit. 12 Compare Baritz, L. (1985). Backfire. Ballantine. 13 On containment doctrine see Gaddis. J.L. (1982). Strategies of Containment. Oxford University

Press. 14 Ambrose, op. cit. 80. 15 Jacobs, J.F.(1983). SAGE Overview. Annals of the History of Computing. 5. 16 Kaplan. op. cir. 17 Winograd, T. and F. Flores (1986). Understanding Computers and Cognition. Ablex. 98. 18 For other aspects of the discussion in this section see Edwards, P. (1986). Artificial Intelligence

and High Technology War: The Perspective of the Formal Machine. Silicon Valley Research Group Working Paper 6. Edwards, P. (forthcoming). The Army and the Microworld: Computers and the Militarized Politics of Gender Signs, also published as Silicon Valley Research Group working Paper 8. Both papers are available from the Silicon Valley Research Group, University of California, Santa Cruz.

19 See Jacky, op. cit. 20 Kaplan, op. cit. p336. 21 McNamara, cited in ibid. 336. 22 Westmoreland, cited in Baritz, op. cit. 35. 23 DARPA (1987). Strategic Computing: Second Annual Report. DARPA. 24 DARPA (1983). Strategic Computing-- New Generation Computing Technology: A Strategic

Plan for its Development and Application to Critical Problems in Defense. DARPA. 3. 25 See Gray, C.H. (unpublished ms. 1987). Artificial Intelligence and Real War. History of

Consciousness Program, University of California, Santa Cruz.