Dissertation

Information Friction: Information Technology and Military Performance

Massachusetts Institute of Technology, Department of Political Science, February 2011

Committee: Barry Posen (chair), Kenneth A. Oye, Wanda Orlikowski, Merritt Roe Smith

How does increasing dependence on information technology (IT) affect military power and bureaucratic behavior? I address this broadly important question by focusing on organizational control problems at the operational level of war (i.e., campaign management as distinguished from strategy or tactics).

For several decades, many defense intellectuals have argued that IT has caused a “revolution in military affairs” (RMA) by which reconnaissance networks and precision weapons radically improve the projection of power. RMA skeptics, however, can point to interoperability failures, overbearing micromanagement, and the political problems of counterinsurgency. I argue, by contrast, that the whole RMA debate is a manifestation of the long-term historical growth in the complexity of socio-technical control. Little progress has been made toward understanding the conditions under which RMA expectations succeed or fail, I believe, because there has been insufficient attention to how military practitioners actually use IT on a day-to-day basis.

In order to address the empirical gap regarding military information practice, I carried out a participant-observer study of a special operations task force in Iraq’s Anbar Province, 2007-2008. In my participant role I oversaw information operations and tribal engagement during a critical time when the province was transitioning from civil war to relative stability as a result of the "Anbar Awakening." In my observer role I found that mundane information processing consumed substantial effort and was shaped by endemic coordination problems and insular organizational preferences. Detailed ethnographic observations, in the context of broader theory from the sociology of science and technology, form the empirical basis for information friction theory.

I introduce the notion of information friction as an aggregate measure of organizational difficulties coordinating IT with the operational environment. Organizations must be able to control themselves in order to control the battlefield, and they must use IT to do it. Unfortunately the technology used to reduce uncertainty becomes of source of uncertainty itself. I argue that control can be achieved only when structural information problems align with political agreement about how to solve them, and then only with ongoing redesign (or "user innovation") of information systems in the course of using them. Other accounts that treat technology as an exogenous force, like any sort of technological determinism, cannot explain the endogenous complexity and variable performances we observe in organizational information phenomena.

To test the external validity of these ideas and to show a detailed example of lower friction, I also present a study of the 1940 Battle of Britain, which is often held up as a paradigm case of successful command and control. In this case, the stability of the air defense information problem, British doctrinal preparation to solve it well in advance (the conventional wisdom about a British bomber bias is misleading on this point), and the ongoing adaptation of radar systems by operations researchers, all combined to improve combat performance. These paired cases show that high information friction, and thus degraded performance, can arise with sophisticated IT, while lower friction and impressive performance can occur with far less sophisticated networks. The social context, not just the quality of technology, makes all the difference. Many shorter examples from recent military history are included to illustrate concepts.

This project should be of broad interest to students of military effectiveness, organizational behavior, and information system design.

CHAPTER 1: INTRODUCTION

1.1 HAS THE INFORMATION REVOLUTION IMPROVED MILITARY PERFORMANCE?

1.2 CONSEQUENCES OF TECHNOLOGICAL MISPERCEPTION

1.3 OVERVIEW OF INFORMATION FRICTION THEORY

1.4 EMPIRICAL METHODOLOGY AND FINDINGS

1.5 INTERDISCIPLINARY CONTRIBUTIONS

1.6 ROADMAP

CHAPTER 2: THE REVOLUTION IN MILITARY AFFAIRS

2.1 INCREASING KNOWLEDGE INTENSIVENESS

2.2 PROPHETS OF INFORMATION DOMINANCE

2.3 CRITICISMS AND REALITY CHECKS

CHAPTER 3: INFORMATION FRICTION AND ITS EFFECTS

3.1 DEFINING INFORMATION FRICTION

3.2 VARIETIES OF HUMAN-COMPUTER INTERACTION

3.3 BREAKDOWNS IN DISTRIBUTED COGNITION

3.4 SUMMARY OF EMPIRICAL MANIFESTATIONS

3.5 THE EFFECTS INFORMATION FRICTION ON MILITARY PERFORMANCE

CHAPTER 4: CAUSES OF INFORMATION FRICTION

4.1 DEFINING THE CAUSES

4.2 EXTERNAL STABILITY ON THE BATTLEFIELD

4.3 INTERNAL CONSENSUS IN THE BUREAUCRACY

4.4 EXPEDIENT ADAPTATION IN HUMAN-COMPUTER INTERACTION

4.5 COMPLEX MISSIONS AND TECHNOLOGIES CREATE INFORMATION FRICTION

CHAPTER 5: SPECIAL OPERATIONS IN AL-ANBAR

5.1 PARTICIPANT-OBSERVATION IN IRAQ

5.2 EXTERNAL INSTABILITY OF THE IRREGULAR BATTLEFIELD

5.3 INTERNAL CONSENSUS AND CONTENTION

5.4 HAPHAZARD ADAPTATION

5.5 EXPECT HIGH INFORMATION FRICTION

CHAPTER 6: INTERFERENCE PATTERNS

6.1 THE ADMINISTRATION OF VIOLENCE

6.2 COGNITIVE PROSTHETICS IN THE TASK FORCE

6.3 SOCIOTECHNICAL APPLICATIONS

6.4 THE INTERFERENCE VARIETY OF INFORMATION FRICTION

CHAPTER 7: TARGET FIXATION

7.1 DISTRIBUTED COGNITION IN TARGETING

7.2 PERCEIVING TARGETS

7.3 INTEGRATING MULTISOURCE DATA

7.4 ARTICULATING RAIDS

7.5 THE INSULATION VARIETY OF INFORMATION FRICTION

CHAPTER 8: THE BATTLE OF BRITAIN

8.1 CASE SELECTION AND METHODOLOGY

8.2 DISTRIBUTED COGNITION IN AIR DEFENSE

8.3 EXTERNAL STABILITY OF AIR DEFENSE

8.4 INTERNAL CONSENSUS ABOUT AIR DEFENSE

8.5 EXPEDIENT ADAPTATION OF THE SYSTEM

8.6 AN INTEGRATED ENTERPRISE

CHAPTER 9: CONCLUSION

9.1 CASE COMPARISON

9.2 OPEN THEORETICAL QUESTIONS

9.3 POLICY IMPLICATIONS

9.4 INTEGRATION IS THE REVERSE SALIENT