3001071 Confrontation Analysis

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE AUG 1999 2. REPORT TYPE

3. DATES COVERED 00-08-1999 to 00-08-1999

4. TITLE AND SUBTITLE Confrontation Analysis. How to Win Operations Other Than War

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Office of the Assistant Secretary of Defense (OASD),Command &Control Research Program (CCRP),Washington,DC,20301

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution unlimited

13. SUPPLEMENTARY NOTES The original document contains color images.

14. ABSTRACT

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT

18. NUMBEROF PAGES

315

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

About the CCRP

The C4ISR Cooperative Research Program (CCRP) hasthe mission of improving DoD’s understanding of thenational security implications of the Information Age.Focusing upon improving both the state of the art andthe state of the practice of command and control, theCCRP helps DoD take full advantage of the opportunitiesafforded by emerging technologies. The CCRP pursuesa broad program of research and analysis in informationsuperiority, information operations, command andcontrol theory, and associated operational concepts thatenable us to leverage shared awareness to improve theeffectiveness and efficiency of assigned missions. Animportant aspect of the CCRP program is its ability toserve as a bridge between the operational, technical,analytical, and educational communities. The CCRPprovides leadership for the command and controlresearch community by:

n articulating critical research issues;n working to strengthen command and control

research infrastructure;n sponsoring a series of workshops and symposia;n serving as a clearing house for command and

control related research funding; andn disseminating outreach initiatives that include the

CCRP Publication Series.

This is a continuation in the series of publicationsproduced by the Center for Advanced Concepts andTechnology (ACT), which was created as a �skunkworks� with funding provided by the CCRP under theauspices of the Assistant Secretary of Defense (C3I).This program has demonstrated the importance ofhaving a research program focused on the nationalsecurity implications of the Information Age. It developsthe theoretical foundations to provide DoD withinformation superiority and highlights the importanceof active outreach and dissemination initiativesdesigned to acquaint senior military personnel andcivilians with these emerging issues. The CCRPPublication Series is a key element of this effort.

Check our website for the latest CCRP activities and publications.

www.dodccrp.org

DoD C4ISR Cooperative Research Program

Senior Civilian Official and Chief Information Officer, OASD (C3I)Mr. Arthur L. Money

Director, Performance AssessmentDr. Margaret E. Myers

Director, Research and Executive Agent for CCRPDr. David S. Alberts

Opinions, conclusions, and recommendations expressed or impliedwithin are solely those of the authors. They do not necessarilyrepresent the views of the Department of Defense, or any other U.S.Government agency. Cleared for public release; distribution unlimited.

Portions of this publication may be quoted or reprinted withoutfurther permission, with credit to the DoD C4ISR CooperativeResearch Program, Washington, D.C. Courtesy copies of reviewswould be appreciated.

Library of Congress Cataloging-in-Publication Data

Howard, Nigel. Confrontation analysis : how to win operations other than war / Nigel Howard. p. cm. -– (CCRP publication series) Includes bibliographical references. ISBN 1-893723-00-3 pbk 1. United States—Armed Forces—Operations other than war. I. Title. II. Series. UA23.H568 1999 355.4—dc21 99-22842 August 1999 CIP

ConfrontationAnalysis:

Operations Other Than WarHow to Win

Nigel Howard

i

Table ofContents

List of Tables and Figures ...................................... iii

Acknowledgments ................................................. vii

Preface ................................................................... ix

Chapter 1 ............................................................... 1

Chapter 2 ............................................................. 21

Chapter 3 ............................................................. 61

Chapter 4 ............................................................. 95

Chapter 5 ........................................................... 133

Chapter 6 ........................................................... 155

Chapter 7 ........................................................... 189

Chapter 8 ........................................................... 215

Chapter 9 ........................................................... 253

Chapter 10 ......................................................... 275

Bibliography ........................................................ B-1

iii

List of Tablesand Figures

Table 1. Confrontation between Allied Forcesand Rebels ............................................................. 9

Figure 1. Traditional relationship betweenpoliticians and military .......................................... 14

Table 2. UN commander confronts ethnic militia .. 23

Table 3. Local commander confronts roadblock .. 25

Figure 2. The six phases of conflict resolution ..... 30

Table 4. The Drunk, His Wife, and His Friend ...... 74

Table 5. Status of Northern Ireland peace talkswhen Tony Blair took over .................................. 119

Table 6. Status of Northern Ireland talks afterBlair shifted position ........................................... 124

Table 7. Northern Ireland options in1993-1994, showing in detail the position takenby the British and Irish governments in theDowning Street Declaration................................ 141

Table 8. Moryan government no longer wants acease-fire ........................................................... 167

Table 9. The general has thought up a newcard, “Blame the government” ............................ 175

iv

Table 10. After the Moryan president hasaccepted the UNFORMOR position ................... 180

Table 11. Bosnian government refuses todiscuss a cease-fire............................................ 221

Table 12. Dilemmas facing characters intable 11 .............................................................. 226

Table 13. What made UNPROFOR angry?The situation preceding table 11 ........................ 228

Table 14. The dilemmas facing characters intable 13 .............................................................. 229

Table 15. Ultimatum to the Bosnian Serbs ......... 233

Table 16. Grand strategic pressure on theBosnian Serbs .................................................... 237

Table 17. Ultimatum to the Bosnian Serbs withcontext cards added from grand strategicmodel ................................................................. 241

Table 18. Dilemmas in table 16 .......................... 243

Table 19. Dilemmas in tables 15 and 17 ............ 244

Table 20. Real pressure brought to bear at laston the Bosnian Serbs ......................................... 247

Table 21. Dilemmas in table 20 .......................... 248

Table 22. Resolution (deceptive) of theoperational drama, with grand-strategic context 250

Table 23. Ultimatum to the Bosnian Serbswith threatening grand strategic context ............. 264

Table 24. Dilemmas facing characters intable 22 .............................................................. 269

v

Figure 3. Computer screen showing briefingon character role for Bosnian government ......... 283

Table 25. Confrontation between representativesand suppliers ...................................................... 297

vii

Acknowledgments

This book is based on research commissioned bythe UK Defence Evaluation and Research Agency

(DERA) at the beginning of 1997 into howConfrontation Analysis might be applied to PeaceSupport Operations. The idea of turning the researchinto a book came from Dr. David S. Alberts, Directorof Research and Executive Agent for the C4ISRCooperative Research Program. For this I am mostgrateful to him.

Andrew Tait (now with SAIC, then with DERA) wasresponsible for suggesting the original researchproject. His suggestion was taken up by GeorgeBrander and Paul Willis of DERA, while Peter Murray-Jones also took a keen interest. Many persons withknowledge or experience of peace support managedto find time to assist with the research, includingMajGen Pennefeather (Commandant-General of theRoyal Marines), Brig Alastair Duncan, Col RobertStewart (Retd), LtCol Philip Wilkinson, Brig B.C.Lambe, Prof K.C. Bowen, Michael Codner and EdwardFoster (Royal United Services Institute for DefenceStudies), Prof James Gow, Indjana Harper and VesnaDomani-Hardy. Capt David Fifield helped to makemilitary contacts, and Brig R. Lambe (Retd), SteveLea, George Rose and Graham Mathieson wereamong many at DERA who helped by responding tothe ideas. The basic ideas have themselves beenworked out in collaboration with Peter Bennett (nowat the UK Department of Health), Prof Jim Bryant and

viii

Prof Morris Bradley, as well as through interactionswith numerous academic and business colleaguesover many years.

Finally, I must express my gratitude to Dr. Richard E.Hayes and Richard Layton, Evidence Based Research,Inc., and their staff, Lynne Jennrich and Margita Rushingfor their work on drafts of the book, and Meg Rittler forher work on the figures and cover design.

© British Crown copyright 1998/DERA. Published withthe permission of the Defence Evaluation andResearch Agency on behalf of the Controller of HMSO.

ix

Preface

This book presents a simple idea. A PeaceOperations campaign (or Operation Other Than

War) should be seen as a linked sequence ofconfrontations—in contrast to a traditional, warfightingcampaign, which is a linked sequence of battles. Theobjective in each confrontation is to bring about certain“compliant” behavior on the part of other parties, untilin the end the campaign objective is reached. This isa state of sufficient compliance to enable the militaryto leave the theater.

If this simple idea is accepted, we can show how thenew technique of Confrontation Analysis (derived fromGame Theory via a development called Drama Theory)can be applied. Thus we can show how to win anOperation Other Than War.

Since this book was written, further research carriedout in the Bosnia theater has clearly revealed thatSFOR commanders—from platoon commanders tothe overall theater commander—are doing it already.They are winning confrontations, or campaigns, madeup of linked sequences of confrontations on a day-to-day basis.

They are doing it, however, without a clear, uniformsystem of concepts specifically designed for aconfrontational campaign. Using practical good sense,they are instead taking doctrinal concepts developedprimarily for warfighting, and adapting them for use inconfrontations. For example, they are using concepts

x

of artillery targeting to plan how to “target”noncompliant parties (i.e., a local Mayor and policechief who are refusing to provide security for returningrefugees from a different ethnic group).

Such common-sensical adaptation of standardwarfighting systems and concepts is admirable. Andit works.

We believe, however, that a system that treatsconfrontations as confrontations, distinguishing themfrom battles both conceptually and in terms ofplanning procedures, will enable strikingimprovements to be made so that our forces becomestill more effective. In particular, it will make it possibleto use the powerful techniques of ConfrontationAnalysis described in this book.

These are ideas that still have to win acceptance. Thisbook aims to lay them out for your consideration.

1

Chapter 1

The Need toReconceptualize

Operations OtherThan War

“War is a mere continuation of politics by othermeans.”

—Clausewitz

From 1945 until 1989, U.S. and Allied defenseforces took as their first priority preparations for

a superpower conflict that never actually occurred.Thus billions of dollars were spent in pursuit of atheoretical construct.

Since 1989, the political assumptions on which thisconstruct was based have shifted. High-intensitysuperpower conflict, the first and most importantcontingency we had to prepare for, vanished from theimmediate agenda because only one superpowerremained. The result was that instead of a conflictconceived and planned for using theory alone, defenseforces now must prioritize lesser threats of which theyhave real experience.

Although this is real experience, as distinct from atheoretical deduction, it is hard to make sense of intraditional military terms. To the theorist using

2 Confrontation Analysis

traditional models, it is messy and disappointing. Thishas put the defense community in a difficult position;it has to ask for and use public money in pursuit of avision, “persuasive for peace, decisive in war,preeminent in any form of conflict” (Joint Chiefs of Staff,1997), whose conceptual and doctrinal underpinningsit needs to clarify.

What is this messy, real-world experience?

Traditional Clausewitzian missions do come up, as inthe case of the Gulf War (Operation Desert Storm).But the traditional character of even this missionseems, in retrospect, less than obvious. Traditionalmilitary objectives were, in the event, attained easily,with minimal loss of life to the Allies and frightful lossesto Iraq. Yet the political problem was poorly resolved.

Other missions are not at all Clausewitzian. Peaceoperations in general (see Alberts and Hayes [1995],Maxwell [1997], and Wentz [1997]) have to do withpolitical stability and humanitarian assistance, ratherthan physically compelling an enemy to submit to ourwill. Objectives in these missions change and developas the mission goes on. Cooperation is requiredbetween military forces and non-governmentalorganizations that share responsibility for missionobjectives and must take them over when the militaryleaves. Economic, political, and psychologicalpressures are important. Military commanders mustnegotiate with conflicting parties rather than fight them.

Despite these non-Clausewitzian characteristics, onepoint made by Clausewitz (1968, 1st edition 1832) ismore relevant than ever. War and war preparations

3Chapter 1

are political acts, whose nature varies with their politicalbackground and aim.

Thus, to understand war in the era beginning after theCold War, and to begin to provide better conceptualunderpinning for defense budgets and plans, we willtry to understand the new political structures thatcharacterize this era.

The New World Order

What are defense forces now defending?

To be frank, what happened after 1989 was theeconomic and military triumph of a single, unified worldsystem that U.S. President George Bush called theNew World Order (Bush, 1990). This system makessurprisingly specific and detailed demands:

…that all must be capitalist, democratic,tolerant, non-racist, and non-sectarian; allowequal opportunities and freedom of speech andthe press; protect human rights; not letgovernments greatly incommode world tradeor capital movements; and be peaceful exceptwhen enforcing these demands.

We admit, of course, that no sooner had Bush utteredthe words “New World Order” than the concept wasdenounced as an absurd chimera by everycommentator. What is significant is that all recognizedat once what he meant and were able, unanimouslyand without coordination, to agree as to what it wasthey considered absurd. Their outraged denials,repeated at intervals ever since, merely gave greatersubstance and wider dissemination to the concept,


helping to make it the powerful coordinating frameworkit has become.

The coordinating power of this universal understandingis immense. When the United States and its allies seekto impose peace and order anywhere in the world, allknow more or less what kind of order they aredemanding. Moreover, there is a definite tendency forpopulations everywhere to demand it for themselves,despite local efforts to persuade them not to.

Russia or China may still provide centers of resistanceto the New World Order. Their ability to do so isdoubtful, and decreases as they become moredependent upon and involved in the system. As theyand others join it, the system will of course evolve. Itis essentially dynamic; however, fundamental changecan be expected to be continuous and evolutionary inthe manner of capitalist development, rather thanviolent and disruptive.

The system now called the New World Order can beseen, in retrospect, to have been growing within the shellof the old, bipolar world since 1945. The old Clausewitzianassumptions that war takes place between more-or-lessequal nations is carried out by military professionalsleading citizen soldiers, and is largely military in characteras distinct from political, held true exceptionally, if at all,from 1945 until 1989. This is because, under the guiseof the Free World combating Communism, Bush’s NewWorld Order was developing.

Now that the system has emerged into the open, howdoes it increasingly enforce its disciplines?

Minor, particularly economic transgressions arepunished by economic sanctions. Transgressions that

5Chapter 1

are not only violent but also sufficiently gross, well-publicized, or economically damaging, such as in theGulf, Bosnia, Ireland, or Oklahoma, evoke forcefulresponses. Defense forces are used then.

Among forceful interventions, the Gulf War generallywas counted a success, Somalia a failure, and Bosniaa mixture of failure followed by the hope of eventualsuccess after the United States came in and got tough.Northern Ireland has taken a turn for the better. Thethreat of internal terrorism (e.g., the Oklahoma bomb,the Tokyo subway gas attack) resembles an ongoingwar against an ever-varying enemy, with new methodsof attack continually met by new methods of defense.The Arab-Israeli problem seems incapable ofresolution as long as the United States, because ofinternal political divisions, remains unable to intervenewith sufficient force.

How Can the Strong Defeat the Weak?

Even over the Gulf War, counted a success, questionsare asked. At the end, why did Allied forces refrainfrom unseating Saddam Hussein? Rogers (1997)suggests that it was because the Allies fearedSaddam’s biological weapons, which he had delegatedto the control of his commanders in the event thatoverall central command was lost. Although we donot believe this, it shows the kind of power that mightpotentially be wielded by a player that has beencomprehensively defeated.

Somalia and Bosnia expose in other ways theinadequacy of conventional military responses to newkinds of threat. Internal terrorist threats show this moreclearly. Consider not only the Oklahoma and World


Trade Center bombs, the IRA’s economic targeting, andthe Tokyo subway attack, but the highly successfuldestruction of the Bank of Sri Lanka by Tamil Tigers. Ineach case, modern technology and terrorist or guerrillatactics have enabled the weak to take on the strong.

The problem of defense in the modern world is theparadoxical one of finding ways for the strong to defeatthe weak. Obviously, this should not be an impossibletask; however, doing it with maximum effectivenessand minimal loss of life and treasure does require anew approach.

Confrontations, Not Battles

What kind of success can the weak aim for againstthe strong?

The weak can never hope to defeat the New WorldOrder by fighting battles, the predominant mode ofwarfare that defense forces prepared for for 40 years.At most, the weak can hope for single, isolated victoriesagainst local outposts. Some may see these assignificant because they believe the system is sounpopular that their attacks will be copied and so fragileit will then collapse. Others aim, more realistically, toget away with something the system forbids: to commitgenocide against local ethnic neighbors (Rwanda,Bosnia); to annex or transform a nation without ademocratic mandate to do so (Iraq, the IRA beforeEaster 1998); or to grab weapons to loot and rapeneighbors (urban riots, Albania).

In opposing such rebels, the system’s guardiansmust make or reject demands. They must say either:“You must give up…” (genocide, aggression, etc.)

7Chapter 1

or “We refuse to…” (publish your manifesto, releaseyour comrades from prison, etc.). In either case,defense forces must sustain and win not a battlebut what we may call a confrontation. This is asituation in which victory consists of compelling,persuading, or inducing others to submit to our willwithout, if possible, using violence, althoughviolence, as well as other threats and inducements,must be a credible part of our armory.

Forty years of preparation for the world’s greatest battlehave not equipped defense forces for winningconfrontations. Our aim in this book is to begin toremedy this comparative lack of preparation.

We propose to investigate confrontations in generalusing the technique of confrontation analysis, bywhich the commander of a peace operation may planand execute a strategy for fighting and winning it.We would like to claim, in fact, that by using thistechnique it is possible to take a more logical, ordered,and defensible approach to this problem than to theanalogous problem of winning battles.

This is not because emotion, friction, irrationality, andthe Clausewitzian “fog of war” are less present inconfrontations than they are in battles. On the contrary,they are found in all types of conflict. What we claim isthat we have a method for analyzing these and otherfactors systematically and scientifically.


A General Model

Confrontation analysis does not apply a general modelindiscriminately to every confrontation. On the contrary,we model each confrontation separately, picking outand taking advantage of its special features.

However, to show what a confrontation is, table 1 setsout a simple, general model of one taking placebetween the Allies and certain unidentified Rebelsagainst the New World Order. Here, the Rebel positionis that they should not be required to give in, but Alliedforces should concede their demands, the precisenature of which, in this particular case, we do notspecify. The Allied position is that the Rebels shouldgive in without any concessions being made.

These positions are displayed in table 1 using themetaphor of a card-table. This is simple. It worksas follows:

• Each player (participating party) holds certaincards, representing its yes/no policy options. Intable 1 players are listed on the left, with theircards listed below their names (e.g., theRebels hold the cards “Give in” and “Retaliate”;Allied forces hold the cards “Concede” and“Crush Rebels”).

• In general, a player can play any combination ofits cards (although some combinations may beinfeasible). In this manner a player chooses andimplements a policy.

After each player chooses a combination of its cardsto play, a column of cards laid out on the card-tablerepresents a projection of the situation as it would be

9Chapter 1

Table 1. Confrontation between Allied Forces and Rebels.

determined by those policy choices. The choice by aplayer to play a particular card is shown in table 1 bya heavy, framed cell representing a card; the choicenot to play it is shown by a white cell. Column R, forexample, represents the future in which the Alliesconcede the Rebels’ demands and do not crush them.The Rebels, accordingly, do not give in (their demandshaving been met) and do not retaliate either (action tocrush them not having been taken).


How Positions and Fallback Positions AreShown in a Card-Table

This, of course, is precisely the Rebels’ position (i.e.,the solution the Rebels demand). The Rebels’ positionis shown by an appropriately labeled column.

The Allies’ position (i.e., the Allied solution, that theRebels give in and not retaliate, while the Allies neitherconcede nor crush them) is shown in column A.

Column t shows the threatened future: What theparties implicitly or explicitly threaten to do if theirpositions are not accepted. The Allies say, “If you don’tgive in, we’ll crush you.” The Rebels say, “If you try,we’ll retaliate.”

This concept addresses an item of utmost importancein a confrontation: what, in the last resort, a partyconveys to all and sundry that it will do if its demandson the situation are not met. We call it the party’sfallback position.

Note that the players’ ways of conveying their fallbackposition, which is a particular kind of conditionalintention, may be direct or indirect. A player maysometimes convey its fallback position by denying it.Imagine, for example, what it would mean if theRussian president declared he has no intention ofinvading a certain country, but that it shouldnevertheless discontinue its anti-Russian policies.

The implicit, although explicitly denied threat in sucha statement would not necessarily be credible. It would,however, be clearly made. The important point is thatat a climactic “moment of truth” in a confrontation, each

11Chapter 1

party’s fallback position (whether or not it is credible)is clearly conveyed to the others.

In a card-table model, parties’ fallback positions arerepresented by certain selections from their owncards. Putting these selections together, we obtain afuture, called the threatened future. This is column tin table 1. Here the Allied forces are implicitlythreatening to crush the Rebels (often merely bringingarmed forces into a theater is to implicitly threaten touse them); the Rebels are threatening retaliation inthat case.

What This Model Represents

Many examples roughly fit our generalized model. In a(highly simplified) model of the Northern Irelandconfrontation toward the end of the IRA’s period ofarmed struggle, the Rebels could be the IRA. The Alliedforces would be or the governments of Britain andIreland. “Giving in” might then mean accepting theAnglo-Irish offer of negotiation in return fordecommissioning arms. “Retaliation” might mean acontinuing terrorist campaign. “Conceding” might meanhanding over power to Sinn Fein and the IRA. “Crushingthe Rebels” might mean continuing to fight terrorism.

This is a high-level example involving governmentsand national movements as players. But the samepattern appears at other levels. The first British troopsin Bosnia, for example, took on the task of clearingroadblocks set up by different factions. We could fitour model to the kind of situations they faced. TheAllied forces of our model would be the local force ledby a British colonel. Rebels would be the local ethnicmilitia manning a particular roadblock. “Giving in” would


mean dismantling the roadblock. “Conceding” wouldmean accepting its continuance. “Crushing the rebels”might have meant using tanks to eliminate theroadblock. “Retaliation” might have meant opening fireon British troops.

Do We Know How to DoConfrontations?

How should a commander deal with a confrontation?Does he know how to win it? What does “winning” mean?

Although individual commanders have beensuccessful in individual cases, there is no general,trained military competence in this area. Yet ourargument is that winning confrontations is the clue toconducting and winning peace operations. Just as atraditional military campaign may be conceptualizedas a linked sequence of battles, so a peace operationmay, we suggest, be seen as a linked sequence ofconfrontations. Continuing the comparison: winning atraditional campaign consists of achieving overallmission objectives through a sequence of battles, eventhough objectives may not be fully met in eachparticular battle, and some battles may be avoidedrather than fought. Winning a peace operation maybe defined similarly. We merely have to replace theword “battle” with the word “confrontation.” The aim ina peace operation is to achieve overall objectivesthrough conducting, and as far as possible winning, asequence of confrontations.

Yet compare the differences in preparedness betweenfighting a traditional campaign and a peace operation.Current military doctrine and training give excellent

13Chapter 1

guidance as to how to conduct individual battles andplan a victorious campaign through a whole sequenceof such battles. Complete, detailed instructions arenot, of course, given. Training and doctrine provideguidelines, adaptable by the commander to meetchanging circumstances. In this way a commander istaught how to plan a strategy and how to implement itby devolving responsibility for its various elementshorizontally and vertically throughout his command.By comparison, with the guidelines available for manyother human activities (e.g., setting up and running agovernment department) these guidelines must bejudged to be highly scientific and effective.

Where do we find comparable guidelines on how tofight and win individual confrontations and wholesequences of them? Where is the training not only inhow to devise and follow a strategy, but in how todevolve it to responsible units?

Figure 1 suggests an explanation for this comparativelack of trained understanding. It depicts the traditionalrelationship, implicitly assumed by Clausewitz andother theorists, between politicians and the military.

Traditionally it was assumed to be the job of politiciansto resolve conflicts peaceably, if they could and wishedto do so (see top part of figure 1). When conflictresolution failed or it was decided to use force instead,the politicians directed the military accordingly,instructing it to forcibly achieve political objectives (seelower part of figure 1). The job for the military was toconduct armed conflict and report back to politicianson their progress. This and only this was what themilitary was trained to do; actual, armed conflictdefined military professional competence.


The development of nuclear weapons following WorldWar II caused one kind of change in this model: thechain of command was modified to ensure thatpoliticians rather than generals made the decision, ifany, to escalate to nuclear war. That decision was feltto be, as Clemenceau reportedly said of war in general,too serious a business to be left to generals. Thus,politicians began to intervene in decisions previouslyleft to the military. Such intervention became morefrequent as improved communications made it morefeasible. It was extended to more types of conflict asthe feeling grew that any use of force by a superpoweror ally of a superpower might start a process ofescalation; theoretical mechanisms accounting forsuch escalation were spelt out by Kahn (1965). In thisway, nuclear weapons gave rise to a doctrinal conceptof limited war.

Figure 1. Traditional relationship between politicians and military.

15Chapter 1

At the same time, another kind of change was takingplace. Members of nuclear alliances no longer couldbe challenged by forces strong enough to do so,because of fear of nuclear escalation. The onlychallenge that could still be mounted against them wasa challenge by the weak, such as the challenge of lowintensity conflict faced by Napoleon’s armies in Spain.

As we have said, meeting a challenge mounted bythe weak is essentially a matter of conducting asequence of confrontations. Objectives cannot beachieved by force alone: the weak are not strongenough and the strong face an enemy that avoidsdecisive encounters by vanishing into the environment.Although such operations are conducted by themilitary, they can succeed only by reaching a politicalsolution, and this solution cannot be reached by high-level decision makers alone because it must begrounded in the hopes and fears of the foot-soldiersfighting on behalf of the weak.

Consequently, the move to this kind of conflict had anopposite result to the move to limited war; it meantthat the military was required to take over many of theconflict resolution activities traditionally assigned topoliticians. Low-intensity conflict, it turned out, was toomessy, detailed, local, and political to be left entirelyto politicians to resolve.

Again, low-intensity conflict usually poses no directthreat to the integrity of the powerful nations involved,but only an indirect threat through the undermining ofinternational order. Because of this, the objectives oflow-intensity conflict are often far from clear. Oftenobjectives are ill-defined resultants of compromise atthe political level between different groups or different


nations in a coalition. As a result, the political directionshanded to the military are vague and ambiguous.

While limited war and low-intensity conflicts comprisedmost of the reality faced by defense forces in the post-war era, it was imperative for them, while the ColdWar lasted, to train and study for the high-intensity,total superpower conflict that never actually occurred.This was their most serious task, on which nationalsurvival and the future of the world ultimatelydepended. The fact that this war never happened may,paradoxically, be a result of the fact that it was sosoundly prepared for. This cannot be accounted afailure. It was a success.

The price of this success was a comparative lack ofdoctrine and training for limited war and low-intensityconflict. This is the lack we hope to begin, at least,to fill.

How To Do It, In Outline

Confrontation analysis, as said, proceeds by buildingspecific models incorporating the specific details ofeach confrontation. It can be used to build models ateach level, strategic, operational, and tactical. Thesemodels can be linked to enable a commander’sstrategy to be devolved into strategies for eachsubordinate level and linked to the strategies of playersrelated horizontally to his command, such as coalitionpartners or non-governmental organizations.

Models so constructed will be strategic, looking forwardto winning a whole operation through a sequence ofconfrontations. They will be readily capable ofchanging to meet changing contingencies, and such

17Chapter 1

changes will be readily propagated through the systemof linked models.

By presenting such a system in a simple, clear way,accessible to all professionals involved, we would hopeeventually to provide doctrinal guidance and trainingfor the at-present, vaguely defined task of peacesupport that is at least as good as that provided forfighting battles.

We would propose that eventually a theatercommander in a peace support operation would haveanalysts on his staff trained in confrontation analysis.At each stage, starting with his first notification of themission he is tasked with, these specialists wouldmodel the problem at each command level.

The commander’s first need is, in general, tounderstand the problem facing his political masters,because understanding the intent of his commanderis necessary for him to understand his own mission.To help him, analysts would model the peace missionas part of the world-political problem and the specificpeace-support problem, as he sees it.

When a commander receives a specific mission,perhaps in vague, nonspecific terms that result fromcompromises between political actors, analysts wouldwork with him to model the specific confrontations heis directly responsible for handling. Such analysis,incorporating the commander’s own assumptions,should yield the two following results:

• Establishment of a sequence of steps to applyrequisite pressure on other parties to bring theminto compliance with the commander’s position(i.e., mission objectives). We shall see how to


derive this sequence based on dilemmas (change-inducing points of stress) we and others face, andhow these are handled to achieve our objectives.

• Invention of an immersive role-playing exercisethrough which the commander and other officersinvolved can digest and criticize the analysis,rehearse their interactions with other parties, andbecome familiar with other-party points of view.

Analysts would then work with lower-levelcommanders and representatives of othercomponents to build devolved and horizontally linkedmodels. Both the analysis and the role-playing basedon it would be updated as the confrontation evolves.The role-playing would be a kind of war-gaming, butwith a firm analytic basis.

This, in outline, is the system we hope eventually willfill the present gap in doctrine and training forOperations Other Than War (OOTW). Chapter 6 willshow how it might work. It contains the Frontline Play,an attempt to dramatize the situation of an imaginarycommander of a peace support operation in a fictitiouscountry who decides to analyze his problem usingconfrontation analysis.

Meanwhile, chapters 2 and 3 give a detailed accountof what confrontation analysis means.

Summary of Chapter 1

Following the end of the Cold War, defense prioritiesgenerally have shifted to a new kind of mission, broadlydescribed as defending the New World Order againstRebels who are militarily weaker. To understand this

19Chapter 1

kind of mission, the concept of winning a campaignby fighting a linked sequence of battles needs to bereplaced with that of winning an OOTW by conductinga linked sequence of confrontations.

A simple, card-table model of a general confrontationshows the position and fallback position of each party.A commander’s mission is to get all parties’ willingcompliance with his position. In doing this, he needsto apply pressure on other parties in ways traditionallythought of as being a politician’s job. A professionalapproach to this requirement is needed. It can besupplied by confrontation analysis.

21

Chapter 2

Handling a MultilevelConfrontation:

The Six-Phase Model

Winning an Operation Other Than War (OOTW),we have said, is a matter of conducting and

winning a linked sequence of confrontations. Thequestion is: How does a commander form andimplement a strategy for this?

This is the key question. To begin to prepare ananswer, in this chapter we will look at the process aconfrontation goes through, leading it to being resolvedor breaking out into conflict. The commander’s aim is,of course, to have it resolved on his terms.

His strategy for achieving this is not primarily a matterof physical activities, as it is with warfighting. It is amatter of communication. At the same time, hispossession of a credible capacity for warfighting isusually central to it.

As with warfighting, his strategy needs to beimplemented on many levels. He should be able togive directives that link together the manyconfrontations occurring on different levels andimplement them through a cohesive, multilevelstrategy. To make this point more clearly, we beginwith a concrete, albeit simplified, example.


Example: Removing Roadblocks

Table 2 shows a card-table model of the situation facedby a UN commander who has been tasked to make acertain ethnic militia force stick to an agreement it hassigned. This model differs from the prototype in chapter1 (see table 1) in that both parties are taking the sameposition. This does not, however, mean they have nofurther problems. The numbers alongside players’names in this model represent their preferencerankings (the order in which they prefer the threefutures shown). The most preferred future for thatplayer is given the number 1, the next most preferred,number 2, and so on.

We are making the following assumptions:

Assumption 1— The agreement states that themilitia should cease ethnic cleansing (i.e.,attacking villages and relocating members ofanother ethnic community) and allow freemovement of people. The commander’s cardsinclude the following options: (a) give or withholdsupport and aid to the militia, which constitutesthe effective local government; (b) if the militiadisallows free movement, the commander mayforce it by removing roadblocks and escortingtravelers; (c) if the militia flagrantly violates theagreement, the commander may bombard themilitia’s headquarters, as he was doing until themilitia signed the agreement.

Because the militia has now signed the agreement, itshares the commander’s position that the agreementbe kept (column P); however, the commandersuspects that the militia does not actually intend to

23Chapter 2

Table 2. UN commander confronts ethnic militia.

allow free movement, which would encouragedisplaced villagers to return to their homes, explainingthe question mark placed on this card. Hence, theconfrontation is not really resolved. [Note: The word“intend” needs clarifying here. In what sense does acomplex, multifarious organization such as an ethnicmilitia intend anything? Answer: A player’s objectivesand beliefs are considered to be the result of internalconfrontations between internal factions (subplayersbelonging to and identifying with the player). In our


case, the militia leadership (one subplayer) may wantto allow free movement, but the leadership’s followers,who are local villagers operating locally (and constituteanother subplayer), do not. The commander, weassume, suspects that the followers’ view may prevail.In this sense, the commander suspects the militiaintends to defect from the agreement.]

Assumption 2— The current threatened future(column t) is not the threatened future whoseactual implementation (bombarding of the militiaheadquarters) led to the signing of theagreement. It is the future the commander isnow threatening to enter on (by implementing hispart of it) if the militia does not allow freemovement. Simultaneously the militia isthreatening to implement its part if thecommander does not give aid and support.

The term “default future,” attached to column d, is ageneral term for the future now being implementedthrough the parties’ current policies, and which willtherefore continue unless and until those policies arechanged. The militia has given up ethnic cleansing,but it has not yet allowed free movement. Thecommander has not started giving aid and support;nor has he started forcibly removing roadblocks.

Note that the cards “bombard militia HQ” and “resumeethnic cleansing” are included in the model althoughthey are not played in any character’s position orfallback position, nor in the default future. They arethere because they are active possibilities in players’minds. A card-table model should include all cardsthe players consider relevant. It must include all cards

25Chapter 2

Table 3. Local commander confronts roadblock.

needed to define players’ positions and the threatenedfuture; but it may include more, as in this case.

In summary, the UN commander has bombarded theethnic militia until it agreed to give up ethnic cleansingand allow free movement. The commander thenpromised to give the ethnic militia aid and support.He has not done so yet, and the militia has not yetallowed free movement; thus, the agreement has notbeen fully implemented.

Linkages Between Different Levels

Next, to illustrate linkages between two levels ofcommand, table 3 models the situation of a localbattalion commander faced with a roadblock. The localmilitia’s position (M) is that the roadblock stays. In light


of his commander’s intent (as shown in table 2), thelocal commander’s position is that it be removed orhe will forcibly remove it (t). If the militia then fires onhis troops from surrounding hilltops, the commanderwill call in air strikes (tI).

Note that the players are considering two possiblethreatened futures. This is because the ethnic militiahas not made its fallback position clear (i.e., whetherit will fire on UN troops if roadblocks are forciblyremoved). This ambiguity indicates that theconfrontation is still in the process of building up to aclimax. At a proper climax, all positions are made clear.

The local commander’s card-table is linked to hiscommander’s card-table; it represents theimplementation at the local level of the commander’soverall position and fallback position.

This statement is clearly true in a general way;however, it is important that the exact manner of thelinkage be decided by the local commander based onhis particular circumstances. It is not determined inany mechanical way by the commander’s position.What is true is that the commander’s position shouldimply that the local commander must make sure theroadblock is removed and give a general idea of thesanctions he can use: first, forcible removal, thenadequate punishment of any retaliation on the part ofthe militia.

The linkage between these two levels is mutuallyreinforcing. If the commander took up his position andfallback position only at the level of his discussionswith the militia leadership, without making sure hislower-level commanders implemented corresponding

27Chapter 2

positions at their respective level, his position wouldbe far less credible. Similarly the credibility of the localcommander’s position is enhanced by its beingsupported by the position of his commander.

Conflict Resolution Is Not a Matter of TakingPhysical Action

Is not conflict resolution just a matter of the commanderissuing orders to all local commanders to forciblyremove roadblocks and call in reinforcements ifnecessary to counter opposition?

The answer is no. Dealing with a confrontation is notjust a matter of taking such-and-such actions,contingent or otherwise, upon the actions of others,although such actions may be a part of it. This isshown, in our case, by the fact that forcible removal ofthe roadblock is not part of the local commander’sposition in table 3. His position is that the local militiashould remove it.

This is far from a trifling difference. Handling aconfrontation is a matter of resolving a conflict throughdialogue. The aim is that actions eventually takenshould be willingly agreed to by all parties, ideally seenby each party as fulfilling its own objectives.

This would be ideal in the case we are examining. Ifthe local UN commander gets his way in this manner,then the message going up from the grassroots to themilitia leadership will be, “Agree to remove roadblocks!We’re in favor.” Forcible removal of roadblocks, onthe other hand, would run the risk of increasing localopposition to their removal, or cause grassrootsdemands that the militia leadership resist this. It might


thus work against the UN operational commander’smission objectives.

We are emphasizing, through this example, thedifference between physical action, including the useof force, and resolving a conflict. The latter is not donedirectly through force, but through dialogue conveyingthe credible threat of force, leading, if successful, togenuine agreement.

The difference must be stressed because it is oftenobscured by the fact that dialogue itself may requirephysical acts, even large-scale ones, to demonstratecredibility. For example, warlike preparations, evenup to beginning to use force, may be necessary tosend the message that force will be used unlesscertain conditions are agreed to. But this is stilldialogue, even though lives may be lost. It is dialoguebecause the aim is to send a message rather thandirectly to achieve objectives. This can be seen fromthe simple fact (Alberts and Hayes, 1995) that theelement of surprise, always desirable when force isused directly to achieve objectives, must be absent.When forceful preparations, or force itself, are usedto send a message, it is essential for the other sideto know about them.

It appears then that force, although often an integralpart of conflict resolution as carried out by armies,should be avoided if possible. To avoid using it, itsuse must be made credible.

What exactly is going on here?

29Chapter 2

The Six Phases ofConflict Resolution

We need to explicate the place of force (or, moregenerally, actually realized conflict) in the process ofconflict resolution. Figure 2 shows the general conflictresolution process divided into six phases. We willdiscuss these one by one.

Phase 1: Scene-setting

In this phase the problem to be resolved is set beforethe parties by the context, by a higher authority, or bya preceding confrontation. For a commander, thismight be when he is tasked with a mission, or whennew circumstances arise during a mission.

For the confrontation in table 3 we can be fairly specific.Scene-setting consists of the roadblock set up by thelocal militia and the receipt of information about it bythe local commander. These are the events that bringthem together in a confrontation.

The Informationally Closed Environment

As figure 2 indicates, scene-setting must logically setup a so-called “informationally closed environment”within which the confrontation may be resolved. Thisconcept is important. Issues between parties can beresolved only on the basis of information available tothem at the time, and this information must remainfairly stable during the course of negotiations.Resolution is impossible if new, relevant informationcontinually upsets nascent understandings andcommitments. Following are two contrasting examples

30C

onfrontation Analysis

Figure 2. The six phases of conflict resolution.

31Chapter 2

that illustrate how demanding the requirement forinformational closure can be:

• Recently a state came to a crisis because ofpersistent feuding between the private armies oftwo politicians. As a last resort, the army chief ofstaff took over and locked the politicians in ahotel room with phones cut off until they couldreach agreement. In doing so, he enforcedinformational closure.

• During Operation Market Garden (the Allied dropon Arnhem in World War II), a reconnaissancepilot observed German heavy tanks in an areawhere British paratroopers were scheduled todrop. This indicated the presence of an SSpanzer division; however, there was resistanceto this genuine new information. Senior Alliedstaff had spent much time and effort agreeing tobattle plans on the basis of the best availableinformation. This naturally created a mind-setaccording to which conflicting new informationshould, if possible, be seen as incorrect. Thealternative, after all, might require lengthyrenegotiations of positions (in drama theoryterms) and redrawing of plans throughout largeorganizations, which would have cost a greatdeal in terms of time, resources, and trustbetween players. The report was ignored.

The second example shows how informational closure,if taken too far, can have pathological effects. Itnonetheless shows how necessary it is for conflictresolution, in this case, resolution of internalconfrontations between Allied staffs. The point is thatnegotiations require stability, at some level, in the facts


accepted as common ground between negotiatingparties. The operational need is to resolveconfrontations in a way that allows for flexibleresponses to new information, while preserving mutualconfidence and understanding. Hopefully confrontationanalysis will help us to meet this need.

Phase 2: Buildup

In this phase, dialogue takes place between theconfronting parties to bring them into full confrontation.As figure 2 indicates, in this manner parties take upfinal positions within a common reference frame. Thisneeds explaining.

The Concept of Common Reference Frame

Although communication between confrontingparties does not require agreement between them,it does require a degree of mutual understanding.They must know the meaning of terms used by eachother and know that each other knows that theyknow. The following analogies illustrate this: It is nouse for me to threaten you with a gun if you think itis a popsicle. Nor can I bribe you with a popsicle ifyou take it for a gun.

The game-theoretic term common knowledge is usefulhere. It stands for what each party knows and knowsthat others know and knows that others know thatothers know…and so forth. Using this term, we cansay that confrontation requires “common knowledgeof the assumptions underlying communicationsbetween parties.”

33Chapter 2

Without such common knowledge, they cannotadequately communicate threats and promises to eachother; this requires that each must not only understandthe other, but know that the other understands them,and so forth.

This essential set of communications assumptionsshared between parties is their common referenceframe. It is this set of assumptions that tables 2 and3, showing confrontations with a militia force, aremodeling. We consider a common reference frameto be specified by designating a card-table, togetherwith an indication of how each player would rankthe various possible futures in order of preference.These preference rankings can be conveyed byassigning players priorities for the various futures,as in our models.

The point here is that communication requirescommon knowledge of what the other is talking about(i.e., who the players are and what their cards are)and of each player’s assumed preferences foroutcomes (e.g., without common knowledge ofpreferences, what is meant as a promise may betaken as a threat, or vice versa).

Following are key points concerning commonreference frames:

• Simplicity is required for a common referenceframe. Only if a common reference frame is verysimple can each party be sure that it is commonknowledge (i.e., can really think, “I know that youknow that I know…what I mean.”). An example iswhen U.S. negotiator Richard Holbrookeconfronted Bosnian Serb commander General


Mladic and Serbian President SlobodanMilosevic in Dobranovci, near Belgrade, inSeptember 1995. Their common reference framewas very simple. It contained two players, NATOand Serbia. The cards under discussion wereNATO’s continued bombing of the Bosnian Serbarmy and the Serbian withdrawal of guns fromaround Sarajevo. “This,” said Holbrooke, “unlessthat.” All details that might distract from thissimple message receded into the background.(See Silber and Little, 1996).

This could be considered an over-simplificationof a complex reality. The Bosnian governmentand Bosnian Serbs were independent actors.Milosevic recently had obtained a kind ofmandate to negotiate for the Serbs as a whole,but on hearing Holbrooke’s demands, he said,“Come and tell Mladic yourself.” Holbrookeagreed; but Mladic was adamant. Holbrooke left,threatening NATO action. Milosevic warnedMladic, telling him that unless he agreed, NATOwould destroy his army. Mladic finally agreed.

A more complex model could be built; however,any model can be made more complex byincluding more detail. It also can be madesimpler by generalizing and omitting details.Simple models are better representations of thestripped-down common reference frames ofplayers at the climax of a confrontation thanones that include distracting details. Such detailsas those mentioned above would merelyillustrate our earlier point that a player’sdecisions are determined by confrontations

35Chapter 2

between subplayers within that player (such asthe face-off between Milosevic and Mladic).

• A common reference frame may or may notrepresent parties’ actual views. This point toois important. If one party succeeds in deceivinganother (e.g., by concealing the fact that itpossesses certain weapons) then the commonreference frame does not include the card, “Usethose weapons,” even though, if the truth wereknown, both parties would include that card intheir considerations. Similarly, a party maydeceive another by pretending to have a weaponit does not have, or pretending (falsely) to havethe will to use a particular weapon. Goingfurther, suppose that the second party is notdeceived by some such pretense, but pretendsto be. Then both parties will in fact share thesame frame, while communicating in terms of acommon reference frame believed in by neitherof them. If these parties resolve their conflict,they do so within a common reference frameknown by at least one of them to be inaccurate.

• The common reference frame, beingcommon, does not reflect the parties’differing values. Each party generally considersitself to be in the right and thinks that others’values, and hence their way of describing thesituation, are wrong. Thus they often usedifferent terminology for elements of thecommon reference frame. What one calls“freedom fighters” another may call “terrorists”and a third “drug runners.” A common referenceframe, in our sense of the term, still exists,


provided that all parties know what each othermeans, and hence know what each other isreferring to using different terminology. Thusthey are referring to the same players and cards(which for purposes of analysis we may name aswe like, using neutral terms that best suit theparty we are communicating with) and to thesame preferences, even though different partieswith different ideologies may give differentreasons for and evaluations of these.

Final Positions

As figure 2 indicates, something more is required forthe parties to build up (during the Buildup phase) to afull confrontation. The parties must take up positionsand fallback positions within the common referenceframe, as shown in tables 2 and 3, and these too mustbe common knowledge.

During the Buildup phase they may experiment withvarious positions to see how the others react. TheBuildup phase comes to an end when the partiesadopt positions they consider final within what theyconsider a final common reference frame. In the end,if the parties are to resolve matters, they must takea stand.

They then face what we call a “moment of truth,”defined technically as a frame together with positionsand fallback positions for each party.

At this point, one of two things must be true. Either allparties take the same position and, according to theircommon reference frame, can trust each other toimplement it, or this is not the case.

37Chapter 2

Suppose it is the case. Then the parties have solvedtheir problem as defined by their frame and theirobjectives within it. A more detailed model, or a modelrepresenting different issues, might show they havedifferences. Within this frame, they have none. Asillustrated in figure 2, they therefore proceed next tothe Resolution phase.

This does not necessarily mean the parties are beinghonest with each other. A solution according to theirproclaimed positions and common reference framemay not be a true solution. One may be deceivinganother. Each may be deceiving all the rest. If thereis deception, it necessarily consists of each deceiveracting as if it has reached honest, trusting agreement.This is because successful deception requires thedeceiving party to behave, not in accordance withthe facts it believes to be true, but in accordance withthe pseudofacts it wishes to be accepted. Thebehavior of such parties (to the extent that they aresuccessful) therefore will be indistinguishable fromthe behavior of parties that are not deceiving eachother, but genuinely believe in the version of realitythey are projecting.

For example, the end period of the Nazi-Soviet Pact,initially signed in August 1939, may be a case of mutualdeception. Although evidence of Soviet intentions islacking (see Davies, 1996, pp. 1000-1013), it seemsfrom their force dispositions that by June 1941 eachwas preparing a surprise attack on the other. The Nazishappened to strike first. Meanwhile, as far as theirdiscussions went, the pact stood. This case actuallytakes the deception one twist further than we havedescribed, in that not only did they deceive each other,


they also disbelieved each other; however, each mayhave mistakenly believed that it was believed.

Phase 3: Climax

We have seen that one possibility at the end of theBuildup phase is that the parties find they have solvedtheir problem, in that the dialogue between themrecords full, trusting agreement, even though,because of deception, the agreement may not turnout as expected.

What about the second case? Here parties eitherdisagree on the terms of a resolution or openly distrusteach other’s intention to carry it out. They then moveto the Climax phase as seen in figure 2. Here theyface what is correctly called a moment of truth;something has to give. If it does not, they move intothe Conflict phase, which consists of each undertakingto carry out its fallback position.

Fear of this outcome, which is in general not likedby one or more parties, puts pressure on them tochange; therefore, something must give, and whatmust give is either their fixed positions within thegiven frame or the fixed frame itself. In other words,to avoid the Conflict phase, they must change oneor more of the following:

• Objectives— Their own or others’ objectives(encapsulated in their positive positions)

• Threats —Their own or others’ implicit threats ofwhat they will do if they do not get their objectives(encapsulated in their fallback positions)

39Chapter 2

• Boundaries —Their own and others’ beliefs aboutthe boundaries of their negotiations (modeled bythe relevant set of players and cards)

• Missions— Their own or others’ general aimspursued in the confrontation (resulting in thepreferences they attach to various combinationsof cards).

For example, look at the encounter described betweenU.S. negotiator Holbrooke and the Serb leaders. Atthe climactic moment of truth, the Serbs changedposition. They agreed to withdraw their weapons. Didthey also change their mission? Were they stilldetermined to cleanse Sarajevo of Muslims? Mostprobably they were. The grudging nature of Mladic’schange of position indicates that his concept of whatthe Serb mission should be did not change; however,Milosevic succeeded in overruling Mladic, andMilosevic may by then have had a different conceptof the Serbs’ mission. In that case, we would call thechange a mission change as well as a change ofposition, although the mission change would be seenas precarious, depending on the balance of powerwithin the Serb coalition.

Emotion and Reason

How then is change brought about? The answer is bymeans of emotion (positive and negative) and rationaldebate (or arguments in the common interest).

Holbrooke must have expressed the anger of theUnited States and the international community at theSerb bombardment of Sarajevo. (Note that ranting andarm waving are not the only ways of expressing anger.


An expressionless face in answer to another’s emotionmay suffice.) However expressed, such negativeemotion would have helped to make it credible thathe would bomb the Serb army into oblivion. Mladic,too, must have expressed anger, signaling hisdetermination not to give in. At the same time, as askilled negotiator, Holbrooke may have expressedpositive emotion toward the Serbs contingent onchanging their mission. The negative emotion of fearfinally would have propelled the Serbian change.

Such emotions necessarily will have played anessential role; however, emotion on its own is not aseffective as when it is supported by rational argumentsin the common interest, based on logic, and if possible,including the production of evidence.

Holbrooke would have pointed out that neither sidewished to see the Serbs destroyed. (Note that it wasnecessary to make this clear to the Serbians because,had it been part of NATO’s mission to destroy them,the threat to do so would not have placed pressure onthem to withdraw, because NATO’s mission wouldhave been to destroy them whatever they did.) Forhis part, Mladic may have tried to appeal to a supposedSerbian–United States common interest in opposingMuslim expansionism.

The function of rational arguments in the commoninterest, using logic and evidence, is to build credibility.Mere emotion can appear to be not serious becauseits effect may be considered transient; however, if Ican convince you that my position and fallbackposition, in addition to being invested with strongpositive and negative emotion, also are based on logicand evidence and are in the common interest, they

41Chapter 2

become very credible. This is because offering reasonsmakes it harder for me to give in and easier for you.Often, neither reason nor emotion, on their own, arecredible motivators; together, they are.

Rational common-interest arguments also tend, ifsuccessful, to weld the confronting parties togetherinto what is seen as one party, pursuing a unified setof interests. This is the converse of the fact, alreadynoted, that the preferences and attitudes of a playerare determined by confrontations between thesubplayers composing that player.

Subplayers, by continually solving their subconfrontationswithin a player, in effect continually construct the playerto whom they belong as a purposeful entity able to takedecisions and, through its representatives, expressemotions. Similarly, when any confrontation is properlyresolved, good feelings between the parties andsentiments of loyalty toward the agreement reached arecreated. If organizational requirements also are met,these feelings tend to create a single superplayercomprised of the original players.

Some further points need to be made about emotionand reason:

• Warning and Notification —If I do not need tochange my preferences to make my threat orpromise credible (because they are things I wouldwant to do anyway), my threat is more like awarning and my promise more like a notification.To convey this, I may show lack of emotion.

• Deception —Conversely, to deceive you intothinking that my threat is merely a warning, mypromise a mere notification, I may show lack of


emotion. For example, Holbrooke may have triedto show the seriousness of the U.S. commitmentby deliberately showing no emotion.

• Irrationality —If I have no good reasons forpreferring to reward or punish you, I may makemy promise or threat credible by strong positiveor negative emotion accompanied by evidenceof irrationality so that you think, with delight orhorror, “He is mad enough to do it!” This is morelikely to be effective, the more immediate is thepossibility of implementing the threat or promisebecause we know that emotions often aretemporary. For example, your anger is morelikely to be effective if your guns are pointing atsomeone than if it will take 3 weeks to transportthem from the United States.

Phase 4: Conflict

If a change in positions or common reference frametakes place in the Climax phase, it requires a return tothe Buildup phase, as shown in figure 2. This isnecessary to communicate the new configuration ofpositions within the new reference frame and toestablish that they are common knowledge.

It is possible change still may not take place. Emotionsand the search for reason and evidence may not beenough to overcome inescapable evidence (e.g., Hitlerhas invaded Poland), unforesakeable values (e.g.,Europe should be democratic rather than Nazi-ruled),and unshiftable positions (e.g., Poland shall not beoccupied by Germany; if it is, Britain will declare war).If change is impossible, by assumption the charactersmust resort to their fallback positions; that is, move to

43Chapter 2

the Conflict phase (e.g., World War II). Here the partiesmust prepare for what we have called the threatenedfuture, the appropriate term for the conflict as imaginedor foreseen by the parties.

At this point, the parties split up (their bid to achievetheir position having failed), and each decides if itnow intends to implement the fallback position it hasimplicitly or explicitly committed to. This phase is aserious decision point. Internal subplayers belongingto each player now may confront each other, somedemanding to carry out their commitment, othersunwilling to do so. Estimates on what other partieswill do are important to a party making this decision.Will they carry out their commitments to thethreatened future? What alternative courses of actionwill they take?

These internal confrontations also can be modeledby the card-table method, using models linked to theoriginal overall table.

In Athens in May 1993, United States and Britishenvoys Cyrus Vance and David Owen, backed byBosnian President Slobodan Milosevic, persuadedBosnian Serb leaders to sign the Vance–Owen planfor dividing Bosnia while keeping it as one nation.According to Vance’s right-hand man: “Vancesuddenly tells them straight out that this is the last-chance café, the U.S. Air Force is all prepared to turnBosnia and Serbia into a wasteland.” This was thethreat Milosevic had been hammering home to them.So the Bosnian Serbs signed, subject to the approvalof its parliament, which rejected the plan. For a timeafter, U.S. officials tried to persuade Europeans toagree to implementation of their threat. They failed.


This United States–European subconfrontation endedwith the United States shifting position and agreeingnot to anger its NATO allies by “act[ing] alone in takingactions in the former Yugoslavia.” There was aconcomitant change in United States values (i.e., thevalues it was pursuing in former Yugoslavia), initiatedby President Bill Clinton and his wife Hillary, who reada book that persuaded them the Balkans were doomedto violence. We interpret this as follows: In the Climaxphase of the West-versus-Serbs confrontation, theSerbs decided to call the West’s bluff. The playersthen moved to the Conflict phase, in which the West’sdecision (through the resolution of confrontationbetween the United States and Europe) was to not doas they had threatened to do (Silber and Little, 1996,chapter 21).

Thus the Conflict phase may end in an Implementationphase in which the threatened future is carried out orone in which it is flunked. There is a third alternative.There may be a return to the Climax phase if one ormore parties decides, as a result of internalconfrontations, that after all they are prepared tochange, and therefore request to reopen negotiations.They may then cycle back from the Climax phase to arenewed Buildup phase.

This happened in the example of the Holbrooke–Milosevic–Mladic confrontation of September 1995.When Holbrooke left the meeting, negotiations movedinto the Conflict phase. Milosevic and Mladic then hada subconfrontation as subplayers within the playerSerbs. This was resolved by Mladic caving in. TheSerbs then called Holbrooke back.

45Chapter 2

The Use of Force

In this example, the Western threat was to use force. Isa threatened future necessarily one of armed conflict?

At the political level, a threatened conflict may notinvolve force at all. Economic sanctions, for example,may be the only threat. The mere involvement of themilitary tends to carry the implicit threat of the use ofarms, even though not necessarily their direct orimmediate use.

In classic United Nations peacekeeping operations, ifformer combatants reject their position in favor of acontinuing cease-fire, the fallback position of theUnited Nations is simply to withdraw UN troops. Thereis no theater-level threat of United Nations use of force,although this is threatened at lower levels if there areattacks on UN troops, who are supposed to defendthemselves if attacked. Although the UN does not itselfthreaten to use force at theater level, UN withdrawalmay lead to resumption of armed conflict between theparties, as happened in May 1967, when thewithdrawal of UN troops was followed in June by theSix-Day War (Alberts and Hayes, 1995, p. 18).

Phase 5: Resolution

If parties are to finally resolve their differences, ratherthan fall into Conflict, according to figure 2, they mustcycle between phases of renewed Climaxes andrenewed Buildups until they converge on a commonposition within a frame that ensures them of eachother’s trustworthiness. Each time they cycle throughthe Climax phase, they risk moving into the Conflict


phase. If they successfully resolve their differences atthe Climax level, they move to the Resolution phase.

After the parties sense they have reached anunderstanding sufficient to resolve their problem, theymust communicate to give mutual reassurance thateach understands what is required. This takes placein the Resolution phase. Here details are discussedand understandings translated into more or lessunambiguous commitments.

It is always possible that this consideration of detailsand firming up of an overall understanding may revealflaws sufficient to differentiate the parties’ overallpositions or cause serious mistrust. That willnecessitate a return to the Buildup phase followed bya renewed Climax phase, with the renewed danger ofconflict. To avoid this, the detailed discussions needto be carried out in an atmosphere of mutual goodwill(positive emotion) to encourage parties to see eachothers’ viewpoint, find win-win solutions, and trust eachother; however, this general atmosphere cannot meanthat detailed differences are not explored. That is thereason for this phase. Exploring differences gives riseto negative as well as positive feelings. The challengein this phase is to avoid going back to the Buildup andthe Climax phases by keeping the overall emotionaltone positive while allowing negative feelings to growup and die down, but never take over.

Even after Germany’s unconditional surrender in WorldWar II, it was necessary for the Allies to sit down andnegotiate detailed issues of governance with Germanofficials. After receiving Japan’s surrender, Britishofficials in some former colonies had to negotiate thecontinued temporary employment of Japanese troops

47Chapter 2

as a police force. The overall solution, unconditionalsurrender, did not determine all these details. Nor didit determine what the implicit fallback positions ofGerman or Japanese officials were in thesenegotiations. It must have been non-cooperation invarious degrees. What if this non-cooperation hadbeen multiplied and increased by general failure tonegotiate satisfactory solutions? This is what musthave happened in the negotiations between Germaninvading forces and the at-first-welcoming inhabitantsof the Ukraine and other areas of the Soviet Union.We suppose the result was first a return to the Buildupphase where the position taken was “Treat us better.”In the immediately ensuing Climax phase, theUkrainians came to a realization that this positionwould be rejected. Finally, the Ukrainians reached aConflict phase where they decided to continue the warwith armed resistance.

Note that discussions at the Resolution phase neednot solve all the problems raised. By recursion, thatwould mean all details had to be solved beforeoutline agreement could be reached, which is absurd(e.g., agreeing the U.S. Constitution did not requiresettling all political differences between parties). Formany problems, discussions will result in mereagreement on parameters for future interactions. TheResolution phase looks at details solely to confirmthe agreement reached within the simplistic, overall,common reference frame arrived at following the lastClimax phase.


Modeling Considerations

Each difference or group of differences emerging inthe detailed negotiations of the Resolution phase canbe modeled by a submodel, where differences arerepresented by differing positions. These submodelsthen can be linked.

At one extreme, each separate issue would bemodeled separately. At the other extreme, alldifferences would be brought together in a large card-table. (Note that there are no practical limits as to howmany cards you can put in a card-table model.) Thechosen modeling procedure should reflect a decision(which itself can be guided by building different models)as to whether to divide issues into separate groups tosolve separately, or to solve them together. Solvingissues together (i.e., modeling them in one card-table)allows consideration of trade-offs between issues.Modeling them in separate but linked card-tables doesnot allow consideration of trade-offs (i.e., it is difficultto model an agreement that consists of a quid proquo between the issues represented in one card-tableand those represented in another). Separate but linkedconsideration does allow the effect of issues on oneanother, inasmuch as reaching a certain agreementin one card-table may have an effect, through causallinkages, on players’ preferences and the cards theyare able to play in another, linked card-table.

Phase 6: Implementation

We have seen several possible kinds ofImplementation of what is decided in a conflictresolution process. What is implemented may includeany of the following:

49Chapter 2

• Resolution of the confrontation— What isimplemented is the common position of all partiesin their common reference frame, with detailshaving been confirmed in the Resolution phase(e.g., the overall agreement between parties inNorthern Ireland reached during Easter 1998)

• False resolution— A common, agreed positionexists but is not implemented because parties tryto deceive others (e.g., the end of the Nazi–Soviet pact in 1941)

• Conflict— Implementation of the threatenedfuture, where all parties decide in the Conflictphase to implement their fallback positions (e.g.,Britain’s 1939 declaration of war on Germany)

• Flunked conflict— Resolution has not beenachieved, and the parties declare the threatenedfuture should be enacted; however, it is notenacted because one or more of the partiesdecides not to implement its fallback position(e.g., the 1993 aftermath of Serbian rejection ofthe Vance–Owen agreement).

These are four ways a confrontation may end. A fifthway is by interruption (i.e., by a breach of theinformational closure that stabilizes the assumptionsunderlying the confrontation). If the parties receivenew, relevant information from exogenous sources thatchanges their assumptions and expectations ascreated in the Scene-setting phase, then all bets areoff. If the parties to the present confrontation continueto interact, it will be through conducting or resolving anew confrontation, one that starts with a new scenebeing set (e.g., a new confrontation started when the


Japanese attack on Pearl Harbor changed theassumptions underlying the confrontation between theUnited States and British parties as to whether theUnited States should enter the war. The confrontationwas speedily resolved in favor of United States entryinto the war).

Implementation and the Unexpected

Interruption, by definition, creates an unexpectedfuture. The same generally is true of the other fourways a confrontation may end. These other fourendings attempt to implement a known future, but thisattempt generally fails because it contains surprises.We cannot foresee the future.

This is more than a matter of uncertainty (a known setof possibilities with known probabilities attached tothem) or risk (a known set of possibilities with unknownprobabilities). Uncertainty and risk apply to the futureas projected by parties beforehand. The future that isactually realized is usually one the parties did not thinkof as a possibility.

The lack of foreseeable future is well known to warriors.As Clausewitz emphasizes, parties’ projections of whatwill happen if a threatened future is one of armedviolence usually fall wide of the mark. The point forconfrontation analysis is that these projections, notthe reality, are what persuade players to choose onepath or another.

51Chapter 2

Unexpected Contingencies and the Need forPositive Feelings

Agreements between parties are necessarilycontingent on the assumptions about the future theparties made at the time, even if they know theseassumptions will turn out wrong. They know theiragreement will have to be reinterpreted by each ofthem under circumstances not foreseen in theagreement. This means that each agreement involvingfuture cooperation between parties needs to containa general card for each one to interpret this agreementwith the other’s interests in mind. To make the playingof this card credible, a general cementing of feelingsof long-term goodwill or love is needed betweenparties. Such feelings are well known to warriors. Aplatoon going into action needs each member tocredibly play the card, “I’ll risk my life to save yours,whatever happens”; hence the love between soldiers.Similarly, cooperation between different nationalcontingents in a combined command requires thebuildup of positive feelings.

Reversible andIrreversible Decisions

Every event is irreversible, in one sense. If it happened,it happened. Yet the logic of the conflict resolutionprocess as shown in figure 2 denies this. To resolveor try to resolve our differences requires a time andspace where the only things that happen are like theputting down and picking up of cards. They are seenas reversible. We need to be able to say, “They saidthat, but we can change their mind. We can reversewhat was said.” This time and space, created in the


Scene-setting phase by informational closure, is aplace where everything is reversible becauseeverything is seen primarily as a message, symbolicof what we intend to do, rather than something donefor its own sake. The logic follows this thought: Nothingis irreversible until the end (i.e., until one of the fiveendings).

This means that until the Implementation phase, oruntil the conflict resolution process is interrupted, allthat can happen is communication of intentions,beliefs, values, and reasons.

Holbrooke walked out on Milosevic and Mladic at theirmeeting in September 1995, thereby taking theconfrontation to the Conflict phase, but he was ableto resume talks with them when they relented, thusmoving back again to the Climax and from there to anew Buildup.

Because nothing irreversible can happen during thefirst five phases, any cards whose playing is by natureirreversible cannot be played until the Implementationphase. This rule is merely a matter of definition. Itmeans that a significantly irreversible decision wouldbe taken to represent the beginning of a newconfrontation, different from the old one inasmuch asthat decision would now be irreversibly fixed. By thesame token, if a card is by nature reversible, then animportant, even necessary way of signaling anintention to implement it may be to start doing so.

The NATO air strikes that accompanied the sameHolbrooke–Milosevic–Mladic meeting had, of course,irreversible effects in damaging Serbian equipmentand taking Serbian lives; but their real function was to

53Chapter 2

signal NATO’s conditional intention to continueattacking the Serbs. This is shown by the need to letthe Serb leadership know the strikes were going on,and why, even though by the principle of surprise(Alberts and Hayes, 1995, pp. 29-30), the attackswould have been more effective in destroying theSerb’s war-fighting capacity if they had been unawareof the strikes or the intention behind them.

The air strikes’ function as a signal depended on theirbeing reversible in the large (i.e., on the possibility ofstopping them if the Serbs relented). Of course, ifcontinued long enough, the air strikes might havecaused so much damage as to create an irreversiblychanged situation, a fact that set a kind of deadlinefor the confrontation to be resolved. Yet the air strikeswere, by nature, sufficiently reversible that the signal,“We’ll bomb you,” had to be, at a certain point,reinforced by starting to do it or it would not havebeen credible. The point at which this need forcredibility kicked in was partly set by the West’sprevious undermining of its own credibility.

By contrast, the Cold War nuclear threat, “We’ll takeout Moscow,” could not be and was not expected todemonstrate its credibility by being carried out. It wastoo irreversible.

You Can Destroy a Parking Lot Only Once

There is only one Moscow; hence by carrying out athreat to nuke Moscow, you lose a card from your hand.You have one less means of exercising pressure.

As an example, there is a possibly apocryphal storythat tells of a Bosnian Serb leader who was opposed


to a cease-fire receiving a phone call from NATO.“Look out the window. See your parking lot? With yourRange Rover in it? That noise is a British jet. It’ll bewith you in two minutes. Know what’s in its sights?Your parking lot. Now perhaps you’d like to changeyour mind… The General? He is standing beside me.”The Bosnian Serb changed his mind. The point of thestory is that if the parking lot had actually beendestroyed, such pressure could no longer be placedon the Bosnian Serb leader. Destroy his house, hisfamily, and everything he values, and he has nothingleft to lose. Consequently it may be impossible to putany pressure on him to concede.

This contrast between destroying other-party assetsto win a battle and to win a confrontation could not bestarker. Each asset I destroy increases my relativephysical strength but loses me a bargaining card;therefore, it is counter-productive to destroy other partyassets when the following situations exist:

• I have sufficient preponderance of physicalforce not to need the physical strengthening Iget from destroying his assets. In general, theUnited States and its allies have this kind ofpreponderance in the post-Cold War world,where we are typically “the strong fighting theweak”; nevertheless, guaranteeing the physicalsecurity of our forces may require selectivedestruction of assets.

• I don’t need to start destroying his assets tomake it credible that I’m prepared to. Thisrefers to the message-sending function of assetdestruction. The problem I may face is the needto use up some of my bargaining assets to make

55Chapter 2

the rest credible. This problem may be solved byminimizing actual destruction while maximizingits psychological effect and, at the same time,sending the right messages to accompany it.

War-Fighting That Is NotMessage-Sending

We should stress that the discussion in this chapterhas concerned a military campaign whose purpose isto conduct a confrontation and, if possible, resolve iton our terms. In this context, the primary purpose ofmuch of the military action that takes place is to senda message; however, this describes only one kind ofcampaign, the kind usually called an OOTW, althoughthe threat of war normally forms part of it, at least inthe case of peace operations. Peace is kept by crediblythreatening war. This is why such missions areundertaken by the military.

In chapter 1, we argued that such missions areincreasingly common in the post-Cold War world;however, military missions of the traditional,Clausewitzian kind also occur. Here the mission issimply and directly to destroy the military assets ofthe other side, thereby lessening or eliminating theirability to do us harm. Message-sending has little todo with it.

The purpose of Clausewitzian war-fighting is, ingeneral, to set the scene for a subsequent peace inwhich the victor can dictate terms, having achieved amonopoly of military force. As Clausewitz says, whilestressing that war is always a means to a political end:“Whatever may take place subsequently, we must


always look upon the object as attained, and thebusiness of war as ended, by a peace.” Clausewitzianwar is undertaken, not primarily to send a message,but to change the facts on the ground so thatsubsequent conflict-resolution processes may have amore favorable outcome.

We can fit direct, purely military action into our model.It can be seen as belonging to the Scene-setting phaseof figure 2. War of this kind is launched by politicianswho foresee that if they and others enter negotiationsequipped with the cards they presently hold, the resultwill be unfavorable to them; therefore, they act to takeaway certain cards held by others, cards consisting ofeffective military response. Surprise and othergenerally accepted principles of war apply fully.

Preemptive action of a traditional military kind alsomay be taken to interrupt a process of conflictresolution to be able to start another on more favorableterms, after others have been deprived of certaincards. This may happen in two following ways:

• The aggressor may intend from the beginning totake the preemptive action. It may simply use theprocess of confrontation and conflict resolutionto lull others into lowering their defenses.

• Alternatively, the aggressor’s participation inconflict resolution may be genuine up to acertain point. It then receives new, exogenousinformation, the result of which is to raise itsdemands above what it can reasonably expectothers to accept, given the present distribution ofcards among the players. So it decides toforcibly take certain cards from others’ hands.

57Chapter 2

Hitler’s decision to interrupt discussions with theSoviet Union and invade Russia may be an exampleof the latter kind, looked at from a short-termviewpoint. In a longer-term perspective, it wasprobably a case of the first kind. He always intendedto attack the Soviets eventually.

Another example might occur if Iraq were discoveredto possess a nuclear bomb, and the Allies knew itslocation. To prevent him threatening the West with it,the decision might be made to destroy the weapon bya commando or air raid. If successful, this wouldremove the card, “Use nuclear weapon,” fromSaddam’s hand.

These examples are Clausewitzian in that war is beingwaged for its physical effect in removing another’sability to harm us. Clausewitzian war may, of course,also be waged by a defender of the peace in reactionto an act of aggression.

We are not seeking to say anything original abouttraditional warfighting, merely to contrast it with OOTWand to point out the essential difference. Whereastraditional war-fighting merely precedes and sets thescene for the message-exchanging activity of conflictresolution, an OOTW consists in large part of thismessage-exchanging activity itself. It consists ofconducting a sequence of confrontations andattempting to resolve them in such a way as to resolvethe issues in accordance with our objectives. That iswhy the primary function of most actions taken inOOTW is to send a message.



An example of a UN force dealing with roadblocksshowed how confrontations at different commandlevels are linked together. Handling them in acoordinated way reinforces the effectiveness ofmessages sent, and we point out that handlingconfrontations is essentially a matter of message-sending, not physical operations.

Conflict resolution is a natural human process divisibleinto six stages: Scene-setting (when the problem isposed within a given context); Buildup (when playerstake positions and fallback positions within a commonreference frame); Climax (when they must eitherredefine their positions and their frame, or else fallinto conflict); Resolution (by having changed they findthey agree and can trust each other to implement theiragreement); Conflict (when having failed to agree, theyare faced with having to carry out their fallbackpositions); and Implementation (when they carry outor decide not to carry out either the agreement theyhave reached or the conflict they have fallen into).

A common reference frame embodies the minimal setof assumptions players need to share to communicate.It needs to be simple, it may be deceptive (becauseplayers can deceive each other), and it is interpreteddifferently by players with differing values.

At the Climax, players use emotion and reason(positive and negative) to change themselves and oneanother. Reason and emotion are each relativelyineffective without the other. Arguments based on theplayers’ common interests are most effective;consequently, players conducting confrontations tend

59Chapter 2

to build up the preferences of a superplayer to whichthey all belong.

Conflict may be flunked (i.e., not implemented becauseof unwillingness to carry out threats that have beenmade), just as agreements may be betrayed. Becausethe future cannot be foreseen, conflict outcomes inany case rarely turn out as projected, and agreementsneed goodwill between participants because they willneed to be maintained in circumstances not foreseenwhen they were launched.

During conflict resolution, nothing irreversible can bedone, because all that is supposed to happen (untilthe Implementation phase, provided the process is notinterrupted) is dialogue. Credibility may requirebeginning to carry out threats, but that is still regardedas a way of sending a message. There is a problemwhen we send a threatening message by destroyingassets. This defeats the purpose of the messagebecause the threat does not exist when the assetshave been destroyed; therefore, it may be best todestroy low-value or replaceable assets provided thishas sufficient psychological impact.

61

Chapter 3

How Parties PressureEach Other:

Cooperation and Trust

Our continuing aim is to show how a commandercan form a strategy to win a confrontation in the

sense of getting his position genuinely accepted.

Of course, the most excellent strategy conceivablecannot by itself guarantee victory, any more than itcan in battle-fighting. Superior strength and moraleare also important, just as they are in battle-fighting,although in a different way. There is a furthercomplication in that mission objectives themselvesmay change in the course of a confrontation, and withthem the very definition of success; nevertheless, it isimportant, as in warfighting, to have the best possiblestrategy for obtaining given objectives. How is thisdeveloped?

In the confrontation shown in table 2, how should thecommander work out a strategy for getting the militiato comply with the agreement to cease ethniccleansing and allow free movement? How should hedevolve his strategy to other levels of his command,such as the level of the commander in table 3?

To answer these questions, one part of the resolutionprocess is of prime importance. It is the Climax phase,


as shown in figure 2, where changes take place incharacters’ attitudes. In this chapter we analyze thesechanges and show how they are brought about.

Understanding them is the key to success. It must be,because the commander requires these changes togo in the direction of his objectives.

As we discuss below, the commander’s job as a peace-operations professional is to build on the fact that thereis a tendency for changes to occur unless and until afull resolution is reached (i.e., until all parties haveconverged on a single solution they can trust eachother to implement). He must direct and orchestratethese tendencies to ensure that resolution is achievedas close as possible to his position.

Games vs. Drama

Therefore we take up the question, “What drives thechanges in attitudes, beliefs, and objectives that takeplace in the Climax phase shown in figure 2?

To lay the foundations for an answer, we will look atthe main assumptions underlying von Neumann’stheory of games and the more general approach withinthe social sciences that calls itself “rational choicetheory” and reaches its fullest development in gametheory. (For game theory, see von Neumann andMorgenstern, 1959. For a clear account of its presentreformulation, following decades of development, seeOsborne and Rubenstein, 1994.)

Our actual question is debarred by game theory. Thesubject defines itself in such a way that the question,“What drives these changes?” cannot be asked. It

63Chapter 3

cannot because rational choice theory has a mission.It tries to take the idea as far as it can that all choice isrational; and the term rational is used to mean“pursuing fixed, given preferences within a fixed, givenframe.” To assume rationality in this sense is toassume that the kind of change depicted in figure 2 isimpossible. Characters’ beliefs about the frame andtheir preferences for possibilities within it (althoughnot their positions) are assumed to be fixed. Of course,exogenous changes in these beliefs and preferences(interruptions) brought about by new informationcoming from outside, are allowed, just as we allowthem. But change brought about merely by characterspressuring each other is assumed not to occur.

Why do we discuss this rationalistic view taken bygame theory if it excludes the possibility of answeringour question? One reason is that the reader may havecome across it. It is important throughout the socialsciences, it dominates our understanding ofeconomics, and it is influential in political science,negotiation theory, and military theory.

That is not our main reason for discussing it. The mainreason is that in pursuing its limited definition ofrationality (“the pursuit of fixed preferences within afixed frame”) game theory uncovers numerousdilemmas, the best-known being “prisoner’s dilemma”(see in particular the early discussions by Schelling,1960; Rapoport, 1964, 1966; Howard, 1971). Thesedilemmas show in various ways that the pursuit of fixedpreferences within a fixed frame is not what it seems;to achieve outcomes high on one’s list of preferences,it may actually be advantageous to forego pursuingone’s preferences because characters generally need


to make threats and promises they would prefer notto carry out.

• The West wanted the Serbs to stop invadingBosnia. To achieve this, the West had to threatento do something it preferred not to do, attack theSerbs. Fixed, known pursuit of the latterpreference (not to attack) prevented satisfactionof the former (Serb cessation of violence).

• The Serbs did not want to be bombed. They alsowanted to take over much of Bosnia. When theWest made its ultimatum credible, Serbianpursuit of its preference for taking over Bosniaprevented it from satisfying its first, greaterpreference, cessation of bombing.

Game-theoretic dilemmas are created by a tension orcontradiction between the values a character wantsto pursue and the credibility it (the character) needsto have to pursue those preferences. The dilemmasso created are dilemmas not just for the theorist, butfor the players themselves. They are in paradoxicalsituations: the prisoners in prisoner’s dilemma actuallysee that if each chooses so as to guarantee the worstpossible outcome for itself (given the other’s choice),both do better for themselves than if each guaranteesthe best possible outcome for itself (again, given theother’s choice). (Howard, 1971, p. 45).

This illustrates that game-theoretic dilemmas bringplayers face to face with psychologically intolerableparadoxes and gives us a clue to explaining thechanges that take place in the Climax phase of aconfrontation. We make the following hypothesis:players make changes in their common reference

65Chapter 3

frame, their own and others’ preferences, and thepurportedly final positions they have taken up in anattempt to eliminate paradoxical dilemmas. Wehypothesize, therefore, that emotions are aroused ina predictable way by game-theoretic dilemmas. Aplayer is pointed in the direction of dilemma-eliminatingchanges. Reason and evidence are invoked to justify,if possible, changes that emotion has pointed to.

For example, Western public opinion eventuallybecame sufficiently emotionally aroused against theSerbs to eliminate the dilemma the West faced, thatwe preferred not to attack the Serbs even though weneeded to make credible our threat to do so. Our anti-Serb feelings, aroused by this dilemma, wererationalized by demonizing the Serbs (i.e., bypropagating the idea that they were evil and supportingit by selectively highlighting the atrocities theycommitted while ignoring those committed againstthem). After we achieved such rationalization, ourpreferences changed. We now preferred to attack theSerbs rather than let them continue attacking Bosnia.

The proposition that emotion evokes the use of reasonand evidence to try to bring about dilemma-eliminatingchange is the basis of the discipline, drama theorythat underpins confrontation analysis. (See Howard,Bennett, Bryant, and Bradley, 1992; Howard, 1994,1994a, 1996, 1998; Bennett and Howard, 1996;Bryant, 1997. See also Nigel Howard Systems, 1992–1997).

It is, perhaps, a disconcerting proposition if onebelieves that reason and evidence should be used totry to reach objective, unbiased conclusions rather thanto prove a point given beforehand. This belief is not


generally well founded. We find the very epitome ofobjective reasoning in scientific method, the closestmethod for arriving at objective truth. Even so, studentsof scientific method stress that scientists too must havea hypothesis to prove or disprove before selectingevidence or using reason (see Popper, 1959; Kuhn,1962). The important thing about the collocation ofreason and evidence is not that it is impartial ordisinterested—in the sense of having nothing to prove.That is not the case. Its importance is that, if effectivelypresented, it compels belief in what it aims to prove.

That is precisely its function in a drama-theoreticconfrontation. Emotion may make me heard; I needreason and evidence to make me believed. Thereason I need to be believed is that I need to makemy position credible.

Why the Drama Metaphor?

The word “drama” replaces the word “game” in thetheory that underlies our approach. Practical usersneed not bother with fundamental theory;nevertheless, they may ask why. What does the term“drama” point to that “game” obscures? Both gamesand drama are role-playing, leisure activities that areused to illuminate life itself. The game metaphor wasused by the military for training purposes long beforegame theory was conceived. Do we need a change ofmetaphor to understand the basic nature of OOTW?

To see that we do, compare behavior in game-playingwith behavior in drama. A game lays out a fixed set ofchoices (sequential or simultaneous), specifies whatoutcomes (stochastic or deterministic) to expect fromeach combination of choices, and requires players to

67Chapter 3

have specific preferences over outcomes. It isforbidden either to vary your preferences (e.g., preferanother to win, or to invent other choices, such asmove another’s pieces). The result is that as long asactors play the game, they have nothing to do butpredict what others will choose and choose the most-preferred outcome for themselves, given theirpredictions. This is precisely what game theoristsmean by rationality.

By contrast, in a drama we see characters beingemotional and irrational and participating in rationaldebate. In doing so, they pass through a crucible thatchanges them. Both their value systems and their viewsof reality change. The very rules of the game change.

The chief interest of drama is in seeing how thishappens. The main thing we learn from drama is not,as with games, how to be rational in the sense ofpursuing fixed preferences against others with fixedpreferences, all the time keeping to fixed assumptionsabout what is possible. Instead, we learn how players(now called “characters”) in interaction with each otheruse reason (in another sense of the term) and emotionto change their common assumptions and to work onthe value roots of their own and others’ preferencesto change them also.

The way a drama ends differs from the ending of agame. A game ends with the victory of one side. Losersare left in an artificial state of frustrated discontent. Itis artificial in the sense that only the arbitrary rules ofthe game prevent the losers from doing anything aboutit. If they choose to operate outside these rules, theymay resort to throwing the pieces on the floor, wreckingthe grounds, or attacking their opponents.


By contrast, a drama ends when none of the charactershas anything left to hope for or to fear. Thischaracterizes a state of stability, with no furthertendency for characters’ expectations to change,although new external information (interruptions) maydisrupt this state. In a real-world drama, thedénouement never turns out as expected, a factpassed over in fiction by not showing what happensafterwards as a result of the agreements andunderstandings reached.

The end of a drama is nevertheless characterized bycomplete stability of expectations, which may bebrought about in two different ways represented bytragic and happy endings. A tragic ending is stablebecause characters’ fears have been realized and theirhopes destroyed; a happy ending is stable becausecharacters’ hopes are realized and fears banished. Ineither case, they no longer have hopes or fears.

From a military commander’s viewpoint, what mattersis that the ending of a drama represents stability, whilethe ending of a game does not. Losers of a game lookforward to another round, if they do not demand areplay or resort to violence. Characters at the end of adrama are content, seeing no alternative to acceptanceof what they have.

Such contentment, enforced or willing, must be thecommander’s object in an OOTW. The parties needto be brought into a state of contentment with a futurethat consists of fulfillment of the commander’s mission.They may, of course, be in various states of discontentand conflict in regard to other matters, particularlydetails. The aim of a peace-support operation isgenerally to bring about a broad degree of peace and

69Chapter 3

order, not to settle every issue between inhabitants ofa region; nevertheless, in regard to the broad societalissues that he has to settle, the commander’s objectmust be to make the parties content with his position.

The means for obtaining that kind of stability throughcontentment are examined in drama theory.

The Need for Game Theory

We must stress that despite its willingness to violategame theory’s basic assumptions, drama theory needsgame theory. In some places it needs game theorydirectly. When for any reason characters cannotinteract with each other to negotiate a resolution (i.e.,when preplay communication is disallowed) dramatheory reduces to game theory.

This, in particular, is the state of affairs when partiesenter the Conflict phase as shown in figure 2. At thispoint they have failed to reach agreement and mustdecide whether to implement the fallback positionsthey are committed to implementing or, if not, whatelse to do. They must make this decision having cutoff communications with each other–unless, that is,they decide to go back to the Climax phase by askingto reopen negotiations on the ground that they have achange to announce. Failing this, the calculations theymust make include hypothesizing if the others will stickto their fallback positions. If so, what should I do? Ifnot, what else will they do? What will they think I amgoing to do? All these responses are game-theoretic.

This is true only of interactions between the charactersas such. Internal confrontations between thesubcharacters making up a character generally


continue to be drama-theoretic. Even so, it is clearlyone way in which game theory is part of drama theory.

Warfighting itself is essentially game-theoretic. In thetraditional Clausewitzian model, the military takes overwhen politicians fail to resolve a confrontation in whichConflict (i.e., the Conflict phase as shown in figure 2)means war. In such case, the development of a militarystrategy to destroy the enemy’s war-fighting capacityis essentially an application of game theory, regardlessof whether quantitative game theory (the form in whichit is usually presented) is considered useful. On theother hand, internal relations between the differentunits of a joint or combined force are drama-theoretic.

Dependence on Game-Theoretic Dilemmas

A fundamental reason drama theory needs gametheory is that it uses the dilemmas produced by gametheory to analyze, predict, and understand thepressures on characters in a Climax phase to changetheir positions, preferences, and common referenceframe. The following logical derivation of drama theoryshows how.

• Initially we suppose that characters see theircommon reference frame as fixed, and hencesee themselves in a game. This is because theyhave assumed this frame for purposes ofcommunication and for the time being, cease toquestion it.

• As the characters see themselves in a game,they try to behave rationally in the game-theoretic sense. This brings them up againstgame-theoretic dilemmas.

71Chapter 3

• The dilemmas then cause the characters to feelpositive and negative emotions. The emotionsfelt depend on the dilemmas the characters faceand the ways they deal with them.

• Emotions may cause them to behave irrationally(i.e., not optimally), change their preferences, orsearch for new cards to play or new charactersto introduce. By these means the dilemmas maybe eliminated.

• As an alternative, the characters may solve theirdilemmas by changing their positions.

• A general alternative to these methods ofdilemma-elimination is deception. Instead ofbecoming irrational or changing preferences,cards, or positions, characters may pretend to doso. Note that deceitful persuasion changes thecommon reference frame just as much as non-deceitful persuasion.

• What makes deceit attractive for one charactercreates disbelief in another. Such disbeliefmust be overcome to effect the desired changein the common reference frame. To overcomeit, characters construct logical arguments,show evidence, and appeal to generallyaccepted standards.

• Characters’ arguments cannot be value-free; tomake sense, arguments generally must assumethe pursuit of some common interests orobjectives. If objectives are not shared,arguments based on them will seem insincere orunappealing. Characters are thus led to constructrational arguments in the common interest.


• Successful arguments of this kind have the effectof building up the preferences and attitudes of asupercharacter formed by an alliance of thecharacters. Individual characters’ preferences donot become the same as the supercharacter’s,but become such that solving subcharacterconfrontations becomes a mechanism causingthe supercharacter to function as a character.

• The supercharacter generally will be a characterin a larger drama, in which it too tries to behaverationally and so confronts dilemmas.

Starting from the assumption that characters try to berational within a fixed frame, drama theory shows howthey are led to behave irrationally and change theframe, so creating the possibility of rational behaviorat a higher level. This is the deep sense in which dramatheory depends on game theory.

The Cooperation Dilemma

Having seen in general how the encounter with game-theoretic dilemmas leads players to change, we lookin detail at how this happens.

There are six dilemmas that can place pressure oncharacters: cooperation, trust, deterrence, inducement,threat, and positioning. We will examine each anddiscuss the kinds of change necessary to overcome it.Each dilemma is illustrated with examples, including areference to the roadblocks-removal problem in tables2 and 3.

The cooperation dilemma faces a character that istempted to defect from its own position. (This arises

73Chapter 3

because my position generally will contain things I donot want, but have included as concessions to others.I may be tempted not to carry them out).

A cooperation dilemma faces a character at a cardtable when, by changing its own selection of cards,the character can move from its own position to a futurethat it likes just as well or better. Table 2, column P,illustrates. Here, the ethnic militia has accepted thecommander’s position, thus also adopting that positionfor itself. Actually, the militia prefers the future (notshown on the table), otherwise the same as P, in whichit does not play the card, “Allow free movement.” Themilitia can move to this future (if the commander carriesout his part of P) merely by changing its own selectionof cards; therefore, the militia’s own preference givesit a cooperation dilemma.

The cooperation dilemma can be thought of as the“drunk’s dilemma.” Imagine an alcoholic whose wifethreatens to leave him if he will not stop drinking. Hedoes not want her to leave, so he swears he will stop.He knows, and knows that she knows, that if she stayshe will not be able to keep his promise, but will contineto drink.

General Statement of the Cooperation Dilemma

Generally stated, the cooperation dilemma faces acharacter that belongs to a group (subset) of characters,all of whom can, by changing just their own cardselections, move from the character’s position toanother future they all like just as well or better.

For example, assume the drunk depends on a visitingfriend for his supplies of booze, as shown in table 4.

74C

onfrontation Analysis

Table 4. The Drunk, His Wife, and His Friend.

75Chapter 3

The friend’s position, F, is that the husband should beallowed to drink (the friend likes someone to drink with)and the wife should let him have the money to do so(rather than forcing him to rely on the friend). Thehusband (the drunk) still swears to his wife that he willquit; thus the drunk and the wife share the position P.The drunk accepts P because he fears the threatenedfuture t; however, he knows and she knows that hewill be tempted to get his friend to assist him indefecting from position P: both the friend and the drunkprefer column d over position P. They can get to dfrom position P just by changing their own joint cardselection, assuming the wife does not change hers.Note that the wife may not be able to detect a movefrom P to d. Because d is the default future (i.e., thefuture they are presently in, and will continue to be inif present policies continue to be pursued), the wifesuspects that despite his promises, he will secretlycarry on drinking, supplied by his friend.

Eliminating the Cooperation Dilemma

A character that faces a cooperation dilemma must, ifit wants itself and others to genuinely accept itsposition, make them believe an incredible promise.How can the character do this?

The character may abandon its position. For example,the ethnic militia may openly say it does not agree tofree movement, or the drunk may tell his wife he willnot give up drinking. In each case (although not in allcases), this lands the character in other dilemmas;however, it eliminates the cooperation dilemma.

If a character does not abandon its position, it musttry to make its promise credible.


What emotion then moves the character? In general,positive emotion, arising from the urge to reassure,will cause the character to feel and show goodwill,friendship, or love (the appropriate term depends onthe types of characters and relationships involved). Ifthe character seems sulky or reluctant or aggressive,its promises will not easily be believed.

Positive rationalizations go with positive emotion.The husband, in an attempt to be convincing, mayexplain how and why he has decided to change hislife. The militia may give reasons why it has changedits mind and decided to agree with free movement.Unless reasons are given, the character’s conversionwill be unconvincing.

Emotion and reason are to some extent substitutes.Strong emotion without adequate reasons mayconvince others that the character can be trusted,temporarily at least. They may think, “He feels thisstrongly enough to go against his own interests ordesires” (irrationally, in the game-theoretic sense ofrationality). On the other hand, strong reasons mayconvince others of the character’s trustworthiness,even if the reasons are accompanied by the wrongemotional signals. Emotion and reason are mosteffective when they accompany and support eachother. Then the character’s emotion shows that itspreferences are in the process of changing (so that inthe game-theoretic sense, the character must betemporarily irrational, while moving betweenpreferences); the character’s reasons give cause tobelieve the change will hold.

Finally, a character may take irreversible actions toeliminate the cooperation dilemma by removing its own

77Chapter 3

temptation to defect. For example, the drunk may tellhis accommodating friend to leave the house. Thecommander of the ethnic militia may make publicbroadcasts threatening discipline against those whoset up roadblocks, thereby irreversibly tying hisreputation as a leader to their removal and makinghimself not prefer not to take action against them.

Deceit, Disbelief, and Rational Arguments in theCommon Interest

Is a character that projects emotions and gives reasonsin the manner suggested above deceitful? Possibly,but not necessarily.

Irreversible actions actually may change preferences,as in the examples given. In addition, attitudes, values,preferences, and beliefs can change for emotional-rational reasons. We propose that it is just these forcesof emotion and rationalization that cause genuinechange. In convincing others, you may convinceyourself. Indeed, if genuine change were impossible,deceit would be impossible. It generally is not possibleto make others believe the impossible.

However, it is important that many of the abovemethods of eliminating the cooperation dilemma mayindeed be carried out with deceit (i.e., not carried outin actuality, only in pretense). Deception means thatthe dilemma is eliminated from the characters’common reference frame (their common view of thesituation, needed for communication). That is all thatdilemma elimination requires. Any deceit, if successful,will not be discovered, at least until the Implementationphase, as shown in figure 2. Whether it is discovered


then, later, or ever depends on characters’ informationat later stages.

Now any situation in which deceit is profitable is onewhere disbelief is appropriate. Any of the dilemma-elimination methods described above may becounter-productive and induce disbelief: “They wantme to believe this; maybe I shouldn’t.” Hamlet sayson such an occasion: “Methinks the lady doth protesttoo much.”

We derive a theorem: “No one should ever believeanyone, because if you tell me something I can deducethat you want me to believe it, which gives me a reasonnot to, since presumably you would want me to believeit whether it were true or not (since in either case youwould succeed in making it part of our commonreference frame).”

To rescue us from this conceptual abyss, recall that itis the characteristic of reason and evidence to compelbelief. They have this characteristic because theymake deceit difficult or impossible. Of course, “difficult”is not “impossible.” Reason and evidence may requireexhaustive scrutiny to compel belief. However, it istheir essential characteristic that they can do sobecause they can uncover deceit.

This, then, is why reason and evidence are needed toback up the more primitive method of emotion inattaining credibility. They are needed to overcomedisbelief. We may speculate that humans developedrational argument in the course of evolution simplybecause they developed the capacity to deceive. Wemay reason as follows: After humans began to findan evolutionary niche in their enhanced ability to

79Chapter 3

cooperate, deception became an advantageousstrategy for an individual or subgroup by enabling themto share in the payoff from cooperation without payingtheir dues; in the presence of deception, disbeliefbecame advantageous; and given disbelief, the abilityto compel belief by rational argument becameadvantageous in turn.

Rational argument requires an ultimate framework ofthe common interest within which to operate, becauseit is always possible to go behind any argument orevidence and ask, “Why? How? Where from?” For thisbasic reason, arguments must in the end be foundedon common values to be convincing. Commoninterests will, however, always exist when there is needto make a promise credible because such a needimplies a common preference for having the promisebelieved and kept (as in column P) rather than notbelieved (as in column t). This common interest,generalized, can be made a foundation for commonvalues and hence for acceptable arguments.

The ethnic militia, if it truly wants to convince thecommander of its sincerity, may draw upon a rangeof common-interest arguments and a fund of possiblecommon values in favor of freedom of movement andagainst setting up roadblocks. Similarly the drunk maydraw on the values and interests he and his wife havein common in their marriage, all of which will besacrificed if he continues to drink. In each case, thecommon-interest arguments made may need to bebacked up by evidence of the value change theyargue for.

In summary, we can derive from characters’ needs tomake their promises credible part of a process by which


they build up or maintain a conception of their commoninterest and so become an effective single supercharacterin larger superconfrontations. From other dilemmas, wewill derive other parts of this process.

Friction

Whether it is easy or hard to change attitudes, beliefs,values, and preferences depends on the strength ofwhat must be changed. It is easy for me to prefer thatyou precede me through a door, although note that Ifeel an access of goodwill toward you even as I makethis slight preference change. It would be hard for meto prefer to die to save you.

The many factors that make it hard to change a commonreference frame are called friction. They involve manythings. In general, friction arises when values are hardto change and evidence hard to escape.

In this general sense, however, values may be morechangeable than the word suggests, and involve no morethan trivial preferences or marginally different ways ofapplying deeply held value systems. For example, anofficer who values discipline may make many smalldecisions in the course of applying it; these representadjustments to what he means by this overarching value,and thus they are mini-value changes.

Why Do People Keep Their Promises?

Why is it we often can rely on people to keep a promiseto do things that they would prefer not to do if theyhad not promised? Often it is because the friction theyfind in changing their preferences between futures isslight enough to be overcome by the emotional urge

81Chapter 3

to come to an agreement and prefer to do as promised.They feel an emotional, preference-changing,dilemma-eliminating urge, however slight. To see this,recall how you feel when promising to do somethingyou would prefer not to. You tend to feel and projectgoodwill, provided you mean to keep the promise. Thatprojection of emotion tells others you can be trusted.If it is not there, people would find it disturbing (provideda preference change is indeed required; provided youare promising something you would have preferrednot to do, rather than just notifying people of somethingyou would want to do anyway). The same projectionof positive emotion occurs when people change theirbeliefs to reach an agreement (dilemma eliminationby belief-change).

In a further twist, strong, principled believers in keepingtheir promises may show their trustworthiness byprojecting no emotion, signaling that having made apromise, their principles infallibly kick in to make themwant to keep it. Such people lay themselves open tobeing misunderstood by strangers.

When stakes are high, such as in matters of peaceand war, friction is too great for such socialmechanisms. Merely promising something will notchange preferences toward performing it.

The militia in table 2 wants to convince the commanderof its sincerity, to keep him from forcibly freeing up theroad network, but there is too much friction involved(the militia wants to keep roadblocks in place as asource of revenue, to keep out returning refugees, andto control movements of groups that might threatenthe militia) for the militia to change its preferences andgenuinely prefer free movement. The militia cannot


help being insincere, despite wishing to convince thecommander of its sincerity.

Eliminating the Cooperation Dilemma byChanging Position

So far in our discussion of ways of eliminating thecooperation dilemma, we have assumed that a playerwill maintain the aim of having its position accepted.

A player’s proclamation that it aims to achieve itsposition may be deceptive. We have allowed for this.The drunk or the ethnic militia secretly may intend todefect, after having obtained acceptance of its position,but a character practicing deception still wishes to haveits position accepted: it needs this to be able to defectfrom it. In light of his wife’s ultimatum, “Stop or I’ll leaveyou,” the drunk cannot carry on drinking until he hasmade his wife believe he genuinely intends to stop. Inlight of the commander’s ultimatum, “Stop or we’ll useforce,” the militia commander cannot carry onobstructing roads without first convincing thecommander he intends to stop.

A dishonest negotiator is still a negotiator that aims tohave its position accepted.

By contrast, another way to escape the cooperationdilemma is to actually change one’s position (i.e., toswitch to a position that does not suffer from thatparticular dilemma in that particular way, although itmay suffer from other dilemmas and even from anothercooperation dilemma).

What emotions accompany the abandonment of aposition? If a player has been determined to maintainits position, sorrow and despair are feelings that help

83Chapter 3

it to withdraw emotionally from the objectives it haspursued, reassess the situation, and pursue lessambitious aims. These emotions generally accompanya committed player’s change of position.

In table 4, the husband would feel sorrow and despairif his inability to stop drinking made him decide togive up his wife (i.e., if he decided that the only wayto solve his cooperation dilemma at P was to shiftposition by accepting the threatened future t,consisting of his wife leaving him while he continuesto drink with his friend).

Alternatively, a player may welcome a change ofposition. This is likely to be so with a player facing acooperation dilemma: “Because I got myself into thisdilemma by accepting a position that is inferior to atleast one other possible position (the one I am temptedto move to), I can probably think of other positions Iwould prefer to go to, if I could get other players toaccept them. For example, there may be a position Ihave relinquished unwillingly, under threat ofsanctions, but would like to go back to it if I canpersuade others to agree.”

A player who for such reasons would welcome a changeof position may seize upon the dilemmas that beset itspresent position as an opportunity to persuade othersthat it might be a good idea to move away from it. Wecan suppose, for example, that the ethnic militia in table2 and the drunk in table 4 each have accepted positionP solely to escape t. This being so, they may use theirown cooperation dilemma at P as an argument forabandoning or modifying this position. The drunk maysay, “I know I agreed to stop drinking, but you knowhow hard it will be for me. Would you really be able to


trust me? Maybe we should think of some compromisearrangement I would actually be able to live with.” Themilitia commander may say, “Agreed. Free movement.But it will be tough to get my people to agree. I don’tknow if I can do it. Is there a compromise position itwould be easier for me to sustain?”

The emotions that accompany attempts to changeposition in the latter case (when the aim is to move toa preferred position, as distinct from one that isaccepted only because it is realistic) are, like despairand sorrow, emotions of withdrawal. The emotionaltone here is skeptical, detached, and deflationary,rather than despairing, because the character needsto deflate emotions and puncture arguments thatsuggest commitment to the present position bypointing to other ways of solving its dilemmas.

If the wife or UN commander beams forth positiveemotion and constructive arguments as to how thedrunk may stop drinking or the militia commandermay free up the roads, the drunk or militiacommander, if they want to argue for a move toanother position, will try to deflate the emotion anddebunk the arguments by being skeptical, realistic,and emotionally uninvolved with the problem. In thisway, they hope to initiate acceptance of another, moreacceptable position.

The Trust Dilemma

I face a trust dilemma when you are tempted to defectfrom my position.

In card-table terms, a trust dilemma confronts mewhen a group of one or more characters (not including

85Chapter 3

me) can, by changing just their own card selections,move from my position to another future they all likejust as well.

Note that when we all share a common position and Ihave a trust dilemma, it must constitute a cooperationdilemma for someone else.

In table 2, the commander has a trust dilemmabecause he cannot trust the militia to allow freemovement. In table 4, the wife’s trust dilemma is thatshe cannot trust her husband not to continue drinkingalcohol supplied by his friend.

Is this the same dilemma stated twice? It is not,because my inability to trust you is my dilemma; yourinability to be trustworthy is yours. There is a differencebetween not being able to trust myself (thus beingpotentially untrustworthy to others) and being unableto trust others. Consider, for example, the case whenthe ethnic militia in table 2 wants to escape itscooperation dilemma by changing position, while theUN commander wants to get rid of his trust dilemmaby persuading the ethnic militia to change its attitudesor incentives. His trust dilemma is the militia’scooperation dilemma. Their perspectives are different.

When there is no common position, my trust dilemmadoes not need to be someone else’s cooperationdilemma. In table 3, the local commander has a trustdilemma because he is asking the local militia to agreeto something he cannot trust them to implement.

For example, if the militia agrees to remove theroadblock (column C), it can move from C to its ownposition, M, simply by taking back the “Removeroadblock” card (the militia prefers M to C). The militia


can move unilaterally from the local commander’sposition to a future it prefers. This gives the localcommander a trust dilemma. The lesson he can drawfrom it is this: If the militia agrees, he will need tomake sure it does remove the roadblock, and doesnot put it back when he goes away.

The same dilemma faced the United States whenPresident Richard Nixon and Secretary of State HenryKissinger attempted to negotiate with North Vietnamto end U.S. involvement there. Put simply, the U.S.position was: “We’ll leave Vietnam if you agree not toinvade the South.” North Vietnam replied: “How canwe agree to such a thing? Would you believe us if wedid?” The U.S. response was: “Agree, damn you, orwe’ll bomb you until you do.” And the United Statesdid just that; it dropped more bombs on Hanoi thanhad been dropped in the whole of World War II. TheNorth Vietnamese finally said: “Okay, we agree.” TheU.S. response was: “We don’t believe you.” How couldwe? Still, the United States left. North Vietnam theninvaded the South.

Clearly this account simplifies the negotiations;nevertheless, it captures the essence of the UnitedStates’ trust dilemma. In a moment of truth, negotiatorsthemselves simplify the issues to make sure they havea common reference frame.

Eliminating the Trust Dilemma

As with the cooperation dilemma, I can eliminate mytrust dilemma by giving up my position. For example,the local commander can cease to require theroadblock to be removed. The wife can agree to letthe husband drink.

87Chapter 3

The emotions that go with giving up a position one iscommitted to are generally negative toward oneself(e.g., despair, sorrow). Accompanying rationalizationsconsist of finding reasons why the new position is notso bad, after all, and may even be better than the oneabandoned. As with the cooperation dilemma, it is alsopossible that the position being moved to is alreadypreferred, in which case the trust dilemma may bewelcomed as an argument in favor of changing position.

To maintain my position against a trust dilemma, Imust find ways to eliminate another’s temptation todefect from it. The required emotional tone is thenpositive and constructive, showing goodwill,cooperativeness, and empathy with the other. This istrue even if I impose an automatic sanction toeliminate the other’s temptation.

Suppose the local commander lets the local militia knowthat, after the roadblock has been taken down, areconnaissance satellite will spot it if the roadblock goesup again, and rocket fire automatically will be calleddown. Despite the fact that violence is being threatened,a positive, friendly tone will convey this information mosteffectively, because its object is to elicit cooperationwith the commander’s position. The commander’s toneshould imply that violence will be unnecessary becausethe militia will cooperate. Violence is there to stabilizethe agreement and provide a reason for the parties totrust each other.

Rationalizations accompanying the projection of suchpositive emotion also should be positive, aimed atconstructing a common position on the basis ofcommon interests. For example, the local commander,even while giving notice of his automatic sanction,


should adduce all kinds of common-interest argumentsin favor of freedom of movement, such as rebuildingthe economy and bringing the country forward to jointhe community of advanced nations. The aim is to bringabout willing compliance with his position. He shouldassume this will be forthcoming and welcome it in aspirit of goodwill.

The wife might let her husband know that she canalways spot his drinking and will leave without furtherwarning if he continues one more time. This isautomatic retaliation. At the same time, she shouldbuild on their love for each other and their commoninterest in the marriage succeeding provided he givesup drinking.

Deceit, Disbelief, and Positive Argument

The local commander’s threatened satellitesurveillance and automatic calling of fire may not existor be as efficient as he suggests; the wife may not beable to spot her husband’s drinking as infallibly as shesays; thus, these communications may be more orless deceitful. Because this is so, the other side maybe more or less disbelieving. The fact that one sidehas reason to produce belief, even if unjustified, makesdisbelief rational for the other side.

To combat disbelief, concrete evidence may beproduced that automatic retaliation is a reality.Alongside or instead of this, positive arguments in favorof the position advocated are a way to deflect disbelief.A bald sanction invites disbelief and hence resistance,leading to thoughts of how to get around it. Positivereasoning on the basis of common interests directsattention away from such reactions and toward seeing

89Chapter 3

the threat of automatic retaliation as an excuse forgiving in to what, judged by such reasoning, seemspreferred on its own merits.

Thus rational common-interest arguments, positiveemotion, and evidence of automatic retaliation are, in away we have seen before, both possible substitutes foreach other and strongest when used together. If mysystem for automatic retaliation is objectivelyunconvincing, I may compensate for this by goodcommon-interest arguments, presented sympathetically.At one extreme, I may be able to do without retaliationaltogether and change another’s mind merely bycommon-interest arguments in favor of my position; viceversa, I may compensate for weak arguments presentedbadly by strong evidence of mechanisms for retaliation.

Modifications of Position

Your arguments will be most effective if you takeaccount of others’ attitudes and beliefs and payattention to their objections, incorporating them intoyour position if possible.

In this way, their objections can be used to modifyyour position itself. Assuming that your aim is tomaintain your position, not to change it, you maynevertheless find that one way to safeguard theessential features of a position may be to modify it tomake it more acceptable to others. This works asfollows. Building upon common interests naturally canlead to the addition of cards that make our positionmore favorable to the other, eliminating the dilemmaby making our position better for them than thetemptation to defect. Thus the wife or local commandermight promise the husband or local militia something


they have been wanting as a reward for going on thewagon or removing the roadblock.

Such inducements will be more credible if stated interms of the common interest rather than in theinterests of those you are inducing, because common-interest arguments give them reason to suppose youwill want to carry out the promise you are making.They also will be more credible if first suggested bythemselves, because this makes the inducementsseem less likely to be a trick or trap.

History cannot be rewritten; nevertheless, it isinteresting to speculate how the United States mightpossibly have eliminated its trust dilemma with NorthVietnam. Such speculation might include creating newcards to put in characters’ hands: automatic retaliation(U.S. forces stationed nearby and conductingreconnaissance), positive arguments (suggestion ofa U.S.–Vietnamese alliance against Russia andChina), or modification of position (a federal Vietnamwith South Vietnamese rulers maintained in powerthrough elections).

Friction

Suppose, as the UN commander, you want to eliminatea trust dilemma without substantially changing yourposition. You proceed to assess evidence andarguments produced by the other side (the ethnicmilitia) to prove it does not intend to defect. As you doso, it will be inherently harder to eliminate your dilemma(a) the more mistrustful you are and (b) the greateryour estimation of the cooperation dilemma the otherwould face if it were to accept your position (or doesface, if it has already accepted it).

91Chapter 3

These are distinct sources of friction. The first,mistrustfulness, should be avoided. A highlysuccessful businessman said, “I trust everyone untilthey give me reason not to.” He was explaining howto make money out of a tendency to be over-mistrustful(arising from the fact that in modern society wefrequently deal with strangers).

Naïve, mistaken acceptance of others’ assurances isbest avoided, not by mistrustfulness, but by realisticallyassessing the reasons for their cooperation dilemmaand helping to counter them in a positive spirit ofcooperation, even if, as noted, you are asking them tocooperate with damaging acts of automatic retaliationsuch as jets bombing roadblocks.

The bad effects of mistrustfulness are that, instead ofencouraging others to genuinely adhere to yourposition, it invites them to treat it as a position that willhave to be abandoned. By projecting a skeptical,detached attitude and treating your trust dilemma(which is their putative cooperation dilemma) assomething hard to overcome, you give the impressionthat you consider your own position untenable.

Willing Changes of Position

I actually may prefer to move, with others, to anotherposition. If this other position does not suffer from thetrust dilemma I face at my present position, I maywelcome my trust dilemma because it gives me anargument for switching.

Suppose the wife actually wants to leave her husband,the drunk. Her position, that she will stay if he will stopdrinking, is one she would be happy to relinquish. She


will then welcome her trust dilemma as giving her areason to leave him. Whatever he says or does shewill not want to trust him. Her mood will be skepticaland uninvolved so as to deflate any emotion andargument he may use to try to persuade her.

Suppose the local commander (belonging to acoalition nation that lacks high standards of militarytraining) has been bribed to let the roadblock stay.His position, that it be moved, is then one he wantsto give up. He accepts with detachment his trustdilemma, the fact that he cannot trust the local militianot to put back the roadblock as soon as his back isturned. He deduces from it that there is no point tryingto get the roadblock moved.


To win a confrontation by getting others to accept hisposition, a commander needs to understand howchanges come about at the Climax phase. Game-theoretic dilemmas are the key. They face players thattry to pursue fixed objectives within a fixed frame,causing them to feel emotion and to seek reasons tochange their preferences and the frame. The metaphorof dramatic interaction seems better than game-playing for describing such processes oftransformation; hence confrontation analysis is basedon drama theory, an extension of game theory.Through trying to influence each other by argumentsin the common interest, dramatic characters tend tobuild up the attitudes and interests of a supercharacter.

Six game-theoretic dilemmas pressure characters ata moment of truth: the dilemmas of cooperation, trust,

93Chapter 3

deterrence, inducement, threat, and positioning. Thecooperation dilemma faces a character that belongsto a group that can do better by defecting from it (thecharacter’s) own position. It is the dilemma of a drunktrying to persuade his wife he means to stop drinkingafter she has threatened to leave him if he does not.Eliminating this dilemma without greatly changingone’s position means making an incredible promisecredible. It requires positive emotions andrationalizations as to why you do not now prefer todefect. Such protestations may be deceitful, invitingdisbelief, which finally can be overcome only by soundreasoning and evidence. Friction to be overcomeinvolves hard-to-change values and beliefs. Somefriction is relatively slight, like that involved in keepingeasily-kept promises. Promises involving high stakesneed to be guaranteed with convincing reasons,conversions, or sanctions.

One way of overcoming a dilemma is to change yourposition to one that does not suffer from this dilemma(though it may suffer from others). If you are committedto your position, emotions of sorrow and despair mayhelp you to rationalize the switch to another position.You may already prefer the other position, in whichcase you may use the dilemmas of your presentposition as arguments in favor of a move.

The trust dilemma faces a character when a groupnot including the character is tempted to defect fromits (the character’s) position. If it does not wish to giveup its position, it must eliminate the trust dilemma byeliminating others’ temptation to defect, either byarranging sanctions against it or by reasonedpersuasion and modification of its position to others’


benefit. The required emotional tone is positive, evenwhen arranging punishing sanctions. Deceit is atemptation and disbelief problem. Mistrustfulness iscounter-productive because it is negative in tone.

95

Chapter 4

More Pressure:Dilemmas of Deterrence,Inducement, Threat, and

Positioning

The Deterrence Dilemma

I face a deterrence dilemma in relation to a characterthat is under no pressure to accept my position. It

feels no pressure because the threatened future doesnot deter it from rejecting my position; rather, itencourages it to.

In card-table terms, I have a deterrence dilemma inrelation to a character opposed to my position thatprefers the threatened future to it (my position).

Unlike the cooperation and trust dilemmas, thisdilemma arises only when parties take differentpositions. There are conflicting positions in the localcommander’s confrontation (see table 3), but not inthe problem faced by his superior (see table 2). Intable 4, the drunk and his wife share a position, but itconflicts with that of the friend.

The high-level, general confrontation in the formerYugoslavia between the Serbs and the West placedthe West in a deterrence dilemma for several years.


The West wanted the Serbs to desist from ethniccleansing. If the Serbs refused, the West threatenedto impose economic sanctions, and actually carriedout this threat. The economic sanctions were nodeterrent because the Serbs as a whole (taking intoconsideration the dynamics of the interaction betweenthe Bosnian Serbs and the Serbian government)preferred the threatened future, “Ethnic cleansing witheconomic sanctions,” to “No cleansing, no sanctions,”which was the West’s position.

Consider the local commander’s confrontation in table3. The local militia has not yet taken a definite fallbackposition; therefore, what is represented is a point inthe Buildup phase of figure 2 at which both parties,the local commander and local militia, have taken uppositive positions. The commander wants theroadblock removed; the militia wants it to stay. Also,the local commander has effectively stated thefollowing fallback position: “If you don’t remove theroadblock, I’ll forcibly remove it. If you then fire on mytroops, I’ll call in air support.” The local militia has notyet stated a fallback position. Whether a deterrencedilemma now follows depends on the fallback positionthe local militia chooses. A refusal to remove theroadblock (without a threat to fire on Allied troops) putsthe militia in a deterrence dilemma because the localcommander prefers the threatened future t (forcibleremoval of roadblocks) to M, the militia’s position;therefore, this would not pressure the commander toaccept M.

97Chapter 4

Eliminating the Deterrence Dilemma

The deterrence dilemma, like other dilemmas, maybe eliminated by giving up one’s position. Forexample, the West might have agreed to let the Serbshave their way, leading to most of Bosnia beingdivided up between Serbs and Croats, with large-scale genocide of Muslims. Such a withdrawal policywas advocated by many in the West. The local militiacould solve its deterrence dilemma, if it had one, byagreeing to remove the roadblock.

Suppose, however, that a character is determined toretain much of its position. An alternative way ofeliminating the deterrence dilemma is to escalate to ahigher level of retaliation to make the threatened futureworse for the other party. This generally involvesthinking up, or bringing onto the agenda, new cardswith which to punish them. This process of thinkingup and making credible new punishment cards isdriven by negative emotions such as anger andindignation. It is rationalized by demonizing the partyto be deterred; that is, seeing them as wicked andevil, so that it becomes right and necessary to considerextreme reprisals against them.

In the overall confrontation between the West and theSerbs, Western public opinion demonized the Serbs(while ignoring, or paying less than proportionateattention to atrocities committed against them byCroats and Muslims) until it was possible for seriousmilitary intervention against them to become a crediblepart of the threatened future, eliminating ourdeterrence dilemma.


The West speedily demonized Saddam Hussein (forexample, relative to President Assad of Syria andothers) when he invaded Kuwait, and it becamenecessary to get rid of a deterrence dilemma bythreatening military force against him.

In table 3, the militia can get rid of its deterrencedilemma by adopting the threatened future tI, ratherthan t; that is, by threatening to fire on Allied troops.Though this invites retaliation from the air (this beingpart of the commander’s fallback position), it curestheir deterrence dilemma because the danger of alliedcasualties means the commander prefers M (themilitia’s position) to tI. In making this threat, they arelikely to be driven by feelings of resentment and anger.

Escalation and De-escalation—Conciliation as aResponse to the Deterrence Dilemma

Thinking up new, more damaging threats is anegative, escalatory way of resolving the deterrencedilemma. It may be necessary, but it is risky. It runsthe risk of triggering a process of mutualdemonization and mutual escalation, in which eachparty responds to threat escalation by furtherescalation, driven at each stage by increasingfeelings of hatred and anger rationalized by viewingthe other as more and more evil.

This risk is greatest when characters have roughlyequal ability to escalate their threats. Clausewitz(1968; 1st edition 1832) generally assumes warbetween equally matched states. Accordingly, hesees mutual escalation as an inherent tendency ofwar. By definition, he thinks war implies “a sort ofreciprocal action, which must logically lead to an

99Chapter 4

extreme” in which “even the most civilized nationsmay burn with passionate hatred of each other” (p.103). Clausewitz, however, does not see why suchpassions are aroused; that is, he does not see theirdrama-theoretic function.

As we saw in chapter 1, war today is not generallyClausewitzian. One side (the United States and itsallies) normally has superior capacity to escalate itsthreats. This imbalance may diminish the risk of mutualescalation, often enabling the allied side to cure itsdeterrence dilemmas by one-sided escalation,provided it is done in the right manner.

Consider a positive, conciliatory way of resolving thedilemma. This is to add new cards that improve(sweeten) our position in the estimation of thethreatened party. In this way the threatened futurebecomes worse for them than our position, notbecause we have made it worse, but because wehave made our position better. The emotion that goeswith this method is the positive one of reaching outand sympathizing with the other ’s needs.Concomitant rationalizations are designed toaccomplish the following:

• Prove to ourselves that others have genuineneeds that deserve to be met

• Establish our position, as now modified andimproved, as a win-win outcome that they oughtto accept.

Following are some examples:

• The local militia, rather than escalate bythreatening to fire on the local commander’s


troops, might try to persuade him to accept theroadblock as a means of keeping peace in thearea. “We’ll always let UN troops through,” themilitia might say. “We can use the roadblock asa way of helping you accomplish your mission.”

• In February 1994, U.S. envoys reportedly offeredPresident Tudjman of Croatia a deal: stopfighting the Muslims of Croatia and we will letyou take back the Krajina region from the Serbs(Silber and Little, 1996). The ensuing de factoalliance between Croats and Muslims led to asuccession of setbacks for the Serbs thatpersuaded the Bosnian Serbs to accept theDayton agreement. Thus, the United Statesshifted its position in favor of the Croats topersuade them to stop fighting the Muslims, andjoin them in fighting the Serbs.

• The 1978 Camp David agreements betweenIsrael and Egypt brokered by U.S. PresidentJimmy Carter were made possible by the UnitedStates sweetening its position for both sides bypromising each of them large sums in aid.

By adding new cards that sweeten our position, wedo run the risk of creating a cooperation dilemma forourselves. Others may suspect we will not keep ourpromise. If this occurs, we must use previouslydescribed positive-emotion methods for overcominga cooperation dilemma by giving reasons why weshould be trusted. Our cooperation dilemma is theirputative trust dilemma.

Note that conciliation and trust-creation both involvepositive emotion. Following are the differences:

101Chapter 4

• Conciliation consists of modifying our position tomake it more attractive to others, therebyeliminating a deterrence dilemma

• Trust creation eliminates a cooperationdilemma by giving reasons why we shouldcarry out our promise (i.e., why we shouldprefer our own position to a possibletemptation to defect from it).

Conciliation needs to emphasize how our positionbenefits them, trust creation how it benefits us.Common-interest arguments have the virtue ofemphasizing both points at once.

Conciliation (i.e., enhancing our position from theviewpoint of the other side) also has been discussedas a method of getting rid of the trust dilemma. Thedifference is that removal of the trust dilemmaaddresses a situation in which both parties share thesame position, as with the UN theater commander innegotiations with the ethnic militia. It assumes eitherthat this is the situation or that the trust dilemma is themain objection to making it so (as with U.S.negotiations with North Vietnam). Removal of thedeterrence dilemma, by contrast, tries to rectify asituation in which the other side is refusing to acceptour position because they prefer the threatened future.It is a mistake to confuse these situations.

U.S. policy in negotiating with North Vietnam, if ouranalysis is correct, made just this mistake. It dealt witha trust dilemma (i.e., “How can we trust them not toinvade the South after we leave?”) using methodsappropriate to a deterrence dilemma (bomb them untilthey agree not to invade the south). These methods


brought the North Vietnamese to the conference table,but left our trust dilemma unresolved.

Conciliation Combined with Escalation

Conciliation as a method for overcoming thedeterrence dilemma has a disadvantage; it may meanwe must make distasteful concessions. Conciliationbecame particularly discredited by British PrimeMinister Chamberlain’s attempts to conciliate Hitler inthe Munich crisis of 1938; conciliation was famouslyused then as an alternative to escalation, becauseBritain and France were unwilling to threaten war.

Conciliation may be effectively combined withescalation in a manner that makes large concessionsunnecessary. Pure escalation has a negativeemotional tone and rationalization; it tends torationalize a preference for anything that is againstthe others’ interests. Consequently others will fear thattheir interests will not be safeguarded in any settlementthey might discuss with us. (Recall that the exactdetails of any settlement are discussed in theResolution phase following overall agreement.)Consequently, rather than discuss a settlement at thisstage, they will look for possibilities of counter-retaliation to lay the basis for a more balanced solution.In this way, pure escalation encourages counter-escalation, if possible.

Combining escalation with conciliation in a “tough cop,tender cop” routine, gives the impression that whiledetermined to punish refusal to settle, we are willingto be sympathetic to others’ interests if they do settle.This impression may count for more than any concreteconcessions you offer, allowing actual concessions

103Chapter 4

to be relatively slight. The point is that others neednegative feelings and rationalizations directed againstus to take the path of escalation. Our concern for theirinterests undermines such negative feelings andmakes it harder for them to justify escalation. In termsof the internal confrontations that determine theirpolicies, conciliation gives the doves within themarguments against the escalation-favoring hawks.

Friction

The deterrence-dilemma friction consists of anyinherent difficulty in making others believe that thethreatened future is worse for them than your position.This friction may be great or small.

For example, in 1995 at Wright-Patterson Air Forcebase in Dayton, OH, Holbrooke took Milosevic andBulatovic (the Montenegrin President) into theNintendo room, a map center equipped withcomputers that allowed the user to overfly the Bosnianterrain. At this point the positions of the Serbs andthe United States were close, differing only in regardto whether a certain area should be under Serbcontrol. Showing him the area in dispute, Holbrookesaid, “Am I seeing right? There’s nothing there. Justmountains. No houses, no villages.” Bulatovic said,“That’s right, but this is Bosnia.” Holbrookeresponded, “Look at what you’re fighting for. There isnothing there.” He was producing reason andevidence in favor of the U.S. position by showing theBosnians how close it was to their own, and hencehow much they should prefer it to the threatenedfuture. (See Silber and Little, 1996, p. 373.)


When at the Athens meeting in 1993 Cyrus Vancetold the Bosnian Serbs “the U.S. Air Force is allprepared to turn Bosnia and Serbia into awasteland,” he was giving reasons why they shouldconsider the threatened future to be worse for themthan the U.S. position.

Overcoming this type of friction is a matter of workingon others’ beliefs and values to affect their comparisonbetween our position and what they can hope for ifthey reject it. Making our position or fallback positioncredible, given the extra cards we may have added tothem, is a separate matter of solving either acooperation dilemma or one of the three remainingdilemmas examined below. It is not a matter of thedeterrence dilemma as such.

Subcharacters Aiming to Change Your Position

We have said you can eliminate a deterrence dilemmain a conciliatory manner by changing your position inthe sense of giving up something; however, such achange of position may be just what certainsubcharacters belonging to your organization prefer.They may use your deterrence dilemma as anargument in favor of the position change that is theirobjective.

• For opinion holders in the West who preferrednot to oppose the ambitions of Hitler, the SovietUnion, China, North Vietnam, the Serbs, andSaddam Hussein, the West’s deterrencedilemma (“How can we deter them?”) was not aproblem but an opportunity. They used theanswer, “We can’t,” as a reason to stopopposing.

105Chapter 4

• Similarly, for opinion holders in the 1980s in theSoviet Union who wanted the Soviets to movetoward the West’s position, the Sovietdeterrence dilemma (“How can we match theReagan arms buildup to negotiate fromstrength?”) was not a problem. They too wouldhave argued, “We can’t. That’s why we shouldgive in.”

When internal confrontations take place to determinethe policies of large characters such as the West orthe Soviet Union, a subcharacter’s true objectives maybe revealed by its attitude toward dilemmas. One thatreally wants to achieve the organization’s statedposition will show negative emotion in face of adeterrence dilemma. One whose true aim is to changethat position will be more detached and objective.

Detachment and objectivity go with a tendency toregard the frame as fixed, encouraging positionalchange because if the frame is fixed, then the onlyway to escape from a dilemma is to change position.

The Inducement Dilemma

The inducement dilemma is the other side of thedeterrence dilemma: In eliminating my deterrencedilemma, I give you an inducement dilemma.

In card-table terms, a character has an inducementdilemma if another’s position (though different fromits own) is as good for it as the threatened future.

It follows that by successfully solving a deterrencedilemma I give someone else an inducementdilemma. However, I may give myself one as well.


When the West, in its overall confrontation with theSerbs, finally started threatening armed intervention,it replaced its deterrence dilemma with an inducementdilemma. It preferred the Serbs’ position (“We continueethnic cleansing”) to implementing its new fallbackposition (i.e., attacking the Serbs). This becameobvious to the Serbs as the West, in onesubconfrontation after another, failed to follow up onits threats. In terms of the five possible endings in figure2, each confrontation ended with the implementationof a threatened future flunked by the West.

After Munich, as Hitler’s ambitions proved to beunappeased, Britain and France started to rearm,thereby making war part of their threatened fallbackposition. In this way they may have eliminated theirdeterrence dilemma (Hitler might have preferred toback off rather than fight) but replaced it with aninducement dilemma (Hitler’s disbelief in theirwillingness to fight).

These examples illustrate that while my deterrenceand inducement dilemmas may be similar in theireffects (in that both may allow my opponents to havetheir way undeterred by me), they are different in theircauses and need to be tackled differently. Mydeterrence dilemma is a matter of my opponents’preferences. I must ensure that they prefer to acceptmy position rather than provoke me into taking up myfallback position. Having ensured this, I may still facean inducement dilemma, which is a matter of my ownpreferences. The question here is, do I prefer to givein to them rather than implement the threatened future?If so, I have an inducement dilemma to overcome tomake my deterrence fully credible.

107Chapter 4

Avoiding Escalation by Accepting theInducement Dilemma

By overcoming my inducement dilemma, I give theother side a deterrence dilemma. This forces them tochoose between giving in or overcoming that dilemma.The latter choice gives me another inducementdilemma to overcome, and so on. For both sides to tryto overcome both dilemmas means embarking on acycle of escalation.

Therefore the question arises: Is it possible tonegotiate in a situation when both sides have aninducement dilemma and neither has a deterrencedilemma? The answer is yes.

Under the nuclear deterrence regime of the Cold War,both sides were in this position. Each must havepreferred, at each confrontation, to give in to the otherrather than start a nuclear war. Because this situationwas clearly symmetrical (balance of terror), neitherside simply gave in; they negotiated.

Hitler and Stalin negotiated the Nazi–Soviet pact in asimilar way. Each preferred a range of positions to aRusso–German war because for each of them thosepositions meant a license to gobble up their neighbors,their military forces being the only ones in Europe thatpresented a serious threat. (Each knew, of course,that after this gobbling-up process had reached itslimits, all bets would be off.)

Often in such negotiations the threatened future istactfully left unmentioned by the parties. Theynegotiate positively, advancing their own positions andattacking others’ on the grounds of each position’scontribution to the common interest. Because all


positions being discussed are preferred to thethreatened future by all parties, continually comparingthem with that threatened future is unnecessary.

Balanced negotiations of this kind, in which each side,rather than suffer a breakdown, would prefer to accepta range of positions, are the norm in civilized relations(e.g., most trade negotiations). Such negotiationsituations are not stable. They contain dilemmas,causing them to move; however, they are preventedfrom escalating and kept within stable bounds bydirecting the destabilizing effect of the inducementdilemma toward impelling characters to negotiate asingle position. This is a way of eliminating thedeterrence dilemma that is a non-escalatory alternativeto preferring the threatened future to others’ positions.

After they have agreed to a single position, thecharacters generally will face cooperation and trustdilemmas. Foreseeing such dilemmas will be one ofthe factors leading them to accept this or that position.Ideally the main factor leading them to agree will bepassionately argued rational arguments in the commoninterest, even when the common interests involvedare as crude as in the case of the Nazi–Soviet pact.

Other Ways of Eliminating the InducementDilemma

Should we then see balanced negotiations (i.e., initialacceptance of the inducement dilemma followed byrational, common-interest debate) leading to itselimination through convergence to a single position,as the norm for defense forces. Should we aim for it?

109Chapter 4

We probably should when considering relationshipsbetween internal actors (the subcharacters within analliance whose confrontations determine its policies).We are then concerned with cooperativeconfrontations, as discussed in chapter 10. Weprobably should not when our defense forces facerebels against the New World Order. In such caseswe typically find an extremist approach on the rebelside (i.e., a willingness to escalate threats to a pointwhere they can dictate the outcome), and anasymmetry of potential power with the Allies, led bythe United States, having superior escalatory capacity.

Opponents motivated by a philosophy requiring themto refuse any kind of compromise may reject theidea of negotiating on the basis of common interests.They may work themselves up into preferring abreakdown (i.e., the threatened future), no matterhow dangerous, over any position acceptable to us.Our forces then may need to eliminate ourdeterrence dilemma by putting on the table cardsthat both punish and conciliate the other sufficientlyto make them prefer at least one acceptable positionto a breakdown. This means combining adequatedeterrence with a conciliatory posture. It meanscarefully lining up our guns to point at their head,then saying, “Right, let’s talk about your problems.”It also means eliminating our inducement dilemmaby preferring the breakdown (firing our guns) toacceptance of their extreme position.

Eliminating our inducement dilemma gives them adeterrence dilemma, forcing them to either modify theirextreme position or escalate by making the breakdownstill worse for us. We are trying to ensure the former


response. To forestall the latter, we must eliminateour inducement dilemma so thoroughly as to preferthe breakdown to their extreme position, no matterhow much they escalate. In general we have thecapacity to do this because of our preponderance ofmilitary power, although it is true that new threats ofchemical or biological attack on cities may enableterrorist groups to make the West back down.

U.S. and Allied preparations for the Gulf War, bothmilitary and psychological–political, exemplifiedthorough elimination of our inducement dilemma; itcovered all possibilities of Iraqi retaliation and preparedus for the worst. The problem seems to be that wefailed to cure our deterrence dilemma (i.e., SaddamHussein still preferred war to leaving Kuwait). We mayhave assumed too much common interest betweenhim and the Iraqi people, who were the ones thatsuffered from the war.

What, then, is the recipe for eliminating an inducementdilemma when the other side will not budge from anunacceptable position? Negative emotion andrationalizations are needed; however, they do not needto be directed entirely at the other side as they mustbe to eliminate the deterrence dilemma, where weneed to think of cards that will hurt them. It is now ourown preferences and underlying values that concernus. We need to think sufficiently poorly of the extremistposition being rejected and reconcile ourselvessufficiently to the threatened future, to prefer the latterto the former. It can be a case of “hate the sin (theposition being rejected), not the sinner (the characterstaking that position).” A feeling of martyrdom (“Bytaking this position they are forcing us to accept the

111Chapter 4

threatened future”) may be the greatest negativefeeling invoked against the other side.

The major argument leading a wide coalition of nationsto support Operation Desert Storm was theunacceptability of letting Saddam Hussein get away withthe annexation of Kuwait. Many reasons were given forthis, taking account of different parties’ interests.Hostility to Saddam Hussein was not important, exceptfor the Allies who had to gird themselves to put theintervention card on the table. They were the ones whohad needed to make an effort (unsuccessful though itturned out) to overcome the deterrence dilemma as wellas the inducement dilemma.

Modifications or reappreciations of the threatened futurethat make it more acceptable to us, but not to the otherside, also help eliminate our inducement dilemma.

For example, at the outbreak of World War I, as tosome degree with any hostilities, there was patrioticpride and excitement at the prospect of using violence.This precisely fitted the bill, making the threatenedfuture more attractive to us while encouraging creativeideas as to how to make it worse for them.

It might seem there also could be modifications of theother side’s position, as distinct from reappraisals ofit, to make it less attractive to us. There are suspectedcases of this. For example, the suspicion that PresidentFranklin Roosevelt permitted Pearl Harbor to happenor that in February 1994 Muslims planted the bombthat blew up their own people in Sarajevo. Here, covertdirect intervention is suspected of having made aposition worse in someone’s eyes, thus overcomingan actual or potential inducement dilemma. Why must


it be covert? It cannot be done openly because it isinconsistent to encourage something that wedenounce (i.e., evil aspects of the other’s position).Inconsistency is against reason, and reason isnecessary to compel belief at a moment of truth. Weadmire inconsistency and confessed self-manipulationin our leisure moments, not at the moment of crisis ina confrontation.

Friction and Subcharacter Conflicts—theInappropriateness of Cost-Benefit Analysis

The inducement dilemma may be hard to overcome,even for a character that can easily overcome thedeterrence dilemma. It is relatively easy for the UnitedStates, leading the West, to threaten action againsttyrants and murderers; it is harder to carry it out,particularly when U.S. citizens’ lives may be lost. Thiscan lead to a commander being given absurdmandates such as, “Threaten to do it; but don’t do itwithout permission.”

What is needed is the political will to take action ifnecessary, supported by an understanding that suchwillingness to carry out threats will mean that they willless often have to be carried out (provided thedeterrence dilemma is first overcome, so that thethreats are adequate). There may always besubcharacters within our character that want to shiftposition toward acceptance of an opposing positionthat is giving us an inducement dilemma. They willpoint to disadvantages of the threatened future as ifour preference for it should be determined by cost-benefit analysis. But this tool is inappropriate whencomparing the threatened future with a position we

113Chapter 4

are rejecting, because we then have a paradoxicalneed to prefer to carry out our threats in order not tohave to carry them out. Concepts such as honor,patriotism, solidarity, ethics, and adherence to principlemeet this need. These emotional concepts have norole in economics, but they do have a vital role inresolving confrontations in which military action is acard that needs to be made credible.

The Threat Dilemma

Often the threat dilemma coincides with theinducement dilemma, but it is conceptually andpractically different. It occurs when I cannot be trusted;I cannot even trust myself to implement my part of thethreatened future if we move into the conflict phase oftable 2.

In card-table terms, I face a threat dilemma when Ican move, just by changing my selection of cards, fromthe threatened future to another future I like just aswell. Therefore, at the moment of truth, others willsuspect that I am bluffing.

We saw in chapter 2 how the West flunked theimplementation of Vance’s threat “to turn Bosnia andSerbia into a wasteland” made in May 1993 to theBosnian Serbs in Athens. The West faced a threatdilemma: it preferred doing nothing to launching abombing campaign. It is unclear if the Serbs suspectedthem of bluffing, but the failure must have damagedWestern credibility for the future. (See Silber and Little,1996, p. 282.)

The dilemma often coincides with the inducementdilemma because the temptation I feel to refrain from


implementing the threatened future may be atemptation to accept your position. However, I mayhave other temptations to defect to from the threatenedfuture. In any case, the two dilemmas are different.

The threat dilemma is like the cooperation and trustdilemmas in that it arises from parties looking forwardto what is likely to happen in the Implementationphase. It is a problem for me because others suspectI might not prefer to carry out my threat if and whenthe time comes to do so; therefore, they discount mythreat as incredible.

The inducement dilemma, like the deterrence dilemma,is grounded in the tug-of-war that takes place at aclimactic moment of truth. I am under pressure thenand there to give in to your position, because to do sois as good for me or better than the threatened future;therefore, for purposes of the inducement-dilemmaargument, the threatened future is assumed to becredible, not incredible. By contrast, my threatdilemma is all about questioning the credibility of thethreatened future. You are disinclined to believe I willcarry out my threat. If I walk out threatening to do so,you think I will either come back, having changed mymind, or do something else instead.

Getting Rid of the Dilemma

Despite this difference between the two dilemmas,many of the emotions and rationalizations we use toovercome an inducement dilemma work as well for athreat dilemma, for the simple reason that both canbe overcome by raising our valuation of the threatenedfuture. In this way we overcome the inducementdilemma by raising its valuation relative to another’s

115Chapter 4

position; we overcome the threat dilemma by raisingit relative to any temptation we have to defect.

Helpful emotions are defiance, anger, indignation, andthe martyred feeling of being forced into the threatenedfuture by others’ intransigence. Temptations may bedownplayed, with reasoning as to why we would notwant to do that anyway. The most general, all-purposerationalizations center on elevating our preference forthe threatened future. There are two kinds: first,evocation of principles such as honor, self-respect,integrity, and the need to keep commitments; second,evaluation of our fallback position as beinginstrumentally the best way, or at least a good way, todeal with the situation created by others resorting totheir fallback positions.

The British continually have used both kinds ofargument to justify their continuing fight againstterrorism in Northern Ireland, despite the public’sfrequently-stated preference for “leaving them to fightit out.” The same two kinds of argument were used inthe United States to justify the continuing involvementin Vietnam and, more recently, the decision to fightSaddam Hussein.

Subcharacter Conflicts Over the Threat Dilemma

It is essential to note that friction to be overcome ineliminating the threat dilemma is felt in theImplementation phase itself. Its effect at the momentof truth or in the preparatory Conflict phase isderivative. It derives from anticipation of the problemsthat will or would arise during implementation.


In the Implementation phase we are faced withneeding to carry out our threats, as the United Statesdid in launching Operation Desert Storm, or failing todo so, as when the United States could not convincethe Europeans they should follow through on CyrusVance’s threat to “turn Bosnia and Serbia into awasteland.” Hard questions of ethics and self-interestarise at this point that were glossed over when makingthe threat. Many who previously supported our positionnow change their minds and join those who werealways against it in arguing that we should not carryout our threat. As in the case of the inducementdilemma, such opposing subcharacters may use cost-benefit arguments. Threats having failed, it may seemthat the realistic thing to do is to abandon our position.

The point is that these difficulties are foreseen,accurately or inaccurately, and possibly withexaggerations in either direction, by those who, at themoment of truth, are faced with assessing thecredibility of our fallback position. The more difficultiesthey foresee for us, and the more reasonable theythink we will be in succumbing to them, the lesscredible our threat, and the more likely it is that we willbe faced with the hard choice of put up or shut up.

The Argentineans were amazed when the British fleetset out to recapture the Falklands, having confidentlyforeseen that Britain’s threat to do so was futile.

The Positioning Dilemma

The positioning dilemma occurs when I preferanother’s position to my own. In card-table terms, Ilike the other’s position column more than mine.

117Chapter 4

This is likely to occur when, under pressure ofdilemmas, I have abandoned a position I occupiedwith others and accepted one that I find less preferred,leaving others with whom I shared the first position tocontinue to occupy it.

A player might abandon a preferred position in favorof one it likes less because of arguments based onrealism. At its old position, the player faced dilemmas,which, unlike the players that still occupy that position,it came to consider insurmountable; however, now itfaces a dilemma in arguing with remaining proponentsof its old position because it prefers what it is arguingagainst to what it is arguing for. The player standsaccused of insincerity and dishonorable conduct infailing to stand and fight for what it once believed in.

The accusation is one of failure to live up to principlesthe player previously defended. A common exampleis that of two friends. In their youth they supportedleft-wing causes. In middle-age, one continues tosupport left-wing causes while the other argues thatthe positions they once shared are unrealistic.

Feelings of guilt at having failed to live up to once-defended principles accompany this dilemma. Theseguilt feelings arise out of the dilemma. They arestructural. We are wrong to think of guilt feelings asbeing caused by the abandonment of higher feelingssuch as altruism in favor of lower ones such asselfishness. The same guilt is felt in response to apositioning dilemma in which lower principles areabandoned in favor of higher ones. To see this,consider two middle-aged friends who thieved and tookdrugs in their youth. One has reformed, primarily onrealistic grounds (deterred by society’s punishments


for such behavior), but also on the grounds that thebehavior is wrong. His friend will accuse him ofdeserting lower principles for higher ones, and if hismain reason for changing position is realism, he willfeel guilty in face of this accusation.

The positioning dilemma is essentially a guilt dilemma.It is overcome by a change of principles (or areassignment of weights between different principles)that eliminates the dilemma by making the new positionpreferred to the old one. The emotion accompanyingthis change is the zeal and enthusiasm of a convert,rationalizing negative attitudes toward the old positionand positive ones toward the new. FormerCommunists become zealous defenders of free-market capitalism. Former drug addicts campaignvehemently against drugs.

Often such converts are respected for their supposedinside knowledge of the position they are attacking.Actually, their knowledge may be biased, having beenthrough strong emotional pressures.

Modeling Northern Ireland Negotiations

To see how the positioning dilemma may affect peaceoperations, we will model the negotiations in NorthernIreland preceding the talks that led to the Easter 1998agreement. Here we will see a number of dilemmasat work, in addition to the positioning dilemma.

Table 5 shows the moment of truth between Britain,the Irish Republican Army (IRA), and the Unionistsbefore the reconvening of peace talks in 1997. Itfocuses on the point when a new government underPrime Minister Tony Blair took over from the

119Chapter 4

Table 5. Status of Northern Irelandpeace talks when Tony Blair took over.

Conservatives, under which an IRA cease-fire brokedown in early 1996. This cease-fire was meant toprecede all-party peace talks. It broke down over theBritish government’s insistence on backing theUnionist demand that Sinn Fein (the IRA’s politicalwing) not be admitted to talks unless the IRA firsthanded in some of its weapons.


When Blair took power, the parties’ positions wereas shown in table 5. In column I, the IRA was offeringto cease fire (but not disarm) if it could then beadmitted to peace talks. In column U,B, the Unionistposition, shared by the British, was that Sinn Fein beadmitted to talks only if they disarmed. If this wasrejected, the British and Unionists were threateningto go ahead with talks without the IRA (in fact, talkswere ongoing, but marking time). Thus the threatenedfuture consisted of the IRA neither ceasing fire nordisarming, the British not admitting them to talks, andthe Unionists not quitting talks. The default future wasthe same.

In this table, as before, the numbers beside characters’names represent their preference rankings for thevarious futures, derived from the values they are tryingto pursue in this confrontation. Number 1 is attachedto the most preferred future, 2 to the next mostpreferred, and so on. From these preference rankings,some of the dilemmas can be read off. Others areindicated by question marks (i.e., a card has a questionmark on it if there is some doubt if it would actually beplayed or not played).

For example, the distrust felt by each side toward theother is indicated by question marks beside theUnionists’ “Not quit talks” choice in column I and theIRA’s “Cease fire” choice in column U,B. This isbecause the IRA suspected that if talks took place onfair terms, as in column I, the Unionists would sooneror later quit such talks, while the Unionists suspectedthat the IRA would sooner or later break the ceasefire even if, as in column U,B, they previously had givenup some of their arms, which the Unionists in any case

121Chapter 4

suspected they would not. (Even if they gave up mostof them, the Unionists suspected, the IRA could alwaysrearm.) This means that both the IRA and Unionistshave trust dilemmas. Each also has a potentialcooperation dilemma because each can envisage talksgoing in directions so unacceptable that they wouldprefer to return to violence or quit the talks.

These dilemmas gave the characters arguments toreject each others’ positions. At the same time, eachwas inclined to regard its own position as second-bestand to argue that, in light of the above dilemmas, theBritish should move with them to another position, onein which the other side was excluded from negotiations.For the Unionists, this preferred position was thedefault future d, the same as the threatened future t.For the IRA, it was the same as their position I, butwith the Unionists provoked into playing their “Quittalks” card.

Britain’s Dilemmas

What caused movement was the combination ofdilemmas faced by the British. They had not only aninducement dilemma (they preferred I to t), but also adeterrence dilemma (the IRA, because they would findit impossible to retain internal discipline if theydisarmed ahead of talks, preferred t to U,B).

It was the British deterrence dilemma that wasdecisive. Unless you can get rid of this dilemma, youare not in the game. Meanwhile, the IRA had neitheran inducement nor a deterrence dilemma. It was theirotherwise unsolvable deterrence dilemma thatprecipitated the Blair government’s change of position.The Blair government accepted the IRA’s position (i.e.,


that it should be admitted to talks without disarming).In table 5, the government shifted position fromcolumns U,B to I.

The same deterrence dilemma faced the Unionists. Ithad the effect on them too of making them want tochange position, but in a different direction. TheUnionists used their deterrence dilemma, like theirother dilemmas, to argue that they and the Britishshould take the status quo (the default future d, thesame as the threatened future t) as their position.

Why did the same imbalance not precipitate a changeof position on the part of Blair’s Conservativepredecessors? The simple explanation, often given,is that they depended on Unionist politicians to keeptheir majority in Parliament, and hence stay in power.This might not have been enough to keep them alignedwith the Unionists if a positioning dilemma did not lieahead if they changed position. Consider the situationof these Conservative politicians. They preferred, forgood reasons, the position U,B to the IRA position I.This meant they need not think about their majority inParliament. Instead they could think of all kinds ofprincipled reasons, such as not giving in to terrorists,for staying with the Unionist position. Blair’sgovernment, by contrast, went against its ownprinciples in shifting position.

The Blair government went against its principles butembraced political realism. The solution U,B in table 5was simply unrealistic. The Blair government might havetried to make it realistic by escalating other threats tothe IRA, such as potential loss of U.S. and Irish support,followed by a heavier crackdown on terrorism. TheUnionists suggested these extra threats, but the

123Chapter 4

Unionists wanted the threatened future, and all suchthreats were, in fact, things the Unionists wanted tohappen for their own sake.

New Moment of Truth—Britain’s PositioningDilemma

When the Blair government shifted position, a newmoment of truth emerged, as shown in table 6. Britainnow shared the IRA’s position. The IRA respondedby calling a new cease fire, without disarming, givinga clear signal of support for what was now a jointBritish–IRA position. (Notice how this cease-fire, beingreversible, acted merely as a signal of intent, not asan element of the implementation. Evidently itremained possible that the implementation, when itcame, would not contain the cease-fire card.)

The threatened future was now that the talks wouldgo ahead without the Unionists. The Unionists alwayssaid that if Sinn Fein was admitted to talks withouthaving disarmed, they would quit. Blair now called theUnionists’ bluff, declaring that the train, which underthe Conservatives had threatened to leave withoutSinn Fein, would now leave, if necessary, without theUnionists; however, in light of the Unionists’ declaredfallback position, the default future was now one inwhich the Unionists would refuse to join the train (i.e.,it was the same as the new threatened future).

This new threatened future naturally changed thedilemmas parties faced. It was no longer the Unioniststhat preferred the threatened future to their ownposition; it was the IRA. The Unionists now faced aninducement dilemma, not wanting to accept theiropponents’ position I,B. Yet the Unionists preferred


this position to t, because the latter would have leftimportant decisions to be made in consultationbetween the British government and their swornenemies. If the train was going to leave, they had tobe on it.

In addition, they now faced a deterrence dilemma.Their position, U, which they now preferred to t,remained unrealistic. It was still the case (more so, infact, than before) that t brought no pressure on theIRA to concede. The IRA preferred t to U even morethan before. This deterrence dilemma gave the British

Table 6. Status of Northern Ireland talks after Blair shifted position.

125Chapter 4

an argument for refusing to accept U, even though indoing so the British had a positioning dilemma. (Thegovernment actually preferred U to its new positionI,B.)

To overcome their now-pressing deterrence dilemma,Unionist militants argued in favor of escalation (i.e.,campaigns of disobedience and disruption). Unionistleaders, however, saw little future in becominglawbreakers and allowing the IRA to claim the mantleof respectability.

Other dilemmas as before faced the IRA and theUnionists, if their positions were to be accepted. TheIRA and the British now had no deterrence dilemma;they were able to put pressure on the Unionists togive in. Nor did the IRA have an inducement dilemma.

The British did have an inducement dilemma, andworse, a positioning dilemma. In talking to theUnionists, the British had to argue against a positionthey preferred in favor of one they disliked. Theirargument could not, therefore, be based on anysupposed virtues of I,B as compared to U. They hadto admit that U was better. Their point was merelythat U was unrealistic.

The Unionists, meanwhile, could not rely on Britain’sinducement dilemma. They might hope that the newthreatened future t in table 6 would prove unbearableto the British, who would eventually be tempted tobreak off talks with the IRA and resume them with theUnionists. This hope was outweighed by the fear thatthe Blair government would face no such dilemma.The anguished Unionists finally accepted the positionI,B and all-party talks began.


How to Get Rid of the Positioning Dilemma

In this example, Britain eliminated its positioningdilemma by getting the Unionists, whose position theypreferred to their own, to accept their position. Thegovernment did so using the argument that theUnionist position was unrealistic.

Undoubtedly, they must also have argued that theirposition, which they shared with the IRA, was notas bad as they had formerly painted it. As long asthe Unionists continued to reject realisticarguments, British negotiators were under pressureto eliminate their positioning dilemma, not just byappeals to realism, but also by adjustments to theirbeliefs and values.

The emotion that drives such elimination isembarrassment at being unable to give meritoriousarguments for something that you are prepared toaccept for reasons other than its merits. This producesnegative emotion directed at the character you arearguing with, together with rationalizations building upthe merits of the advocated position as against theone rejected. In extreme form, this emerges in theemotions of a convert, who generally feels morestrongly in favor of its new position and against its oldone than those who never took the old position.

Subcharacter Conflicts Over the PositioningDilemma

The friction involved in overcoming the positioningdilemma is simply the difficulty you find in overturningthe beliefs and values by which you once justified yourold position against your new one. Scientists accepting

127Chapter 4

a new theory, for example, must go methodicallythrough each of the old arguments to see why pointsthey previously rejected are now acceptable. Similarly,converts to a new morality have to critically review oldlife situations in which they made wrong choices.

This is what you must do to justify your new position.Meanwhile, for subcharacters who think we shouldhave kept to the old position, our positioning dilemmais welcomed as an argument against the new position.It may similarly be welcomed by those who favor athird position, as they can use the dilemma to pointout that we are being illogical.

When negotiations break down and characters enterthe Conflict phase of figure 2, each character generallysplits into subgroups, within which confrontations takeplace between those who want to proceed with thefallback position and those who want to restartnegotiations or otherwise flunk the Conflict phase. Insuch a confrontation, those whose position is“Implement our fallback position” may be accused ofinconsistency on the grounds that until now they haveadvocated another solution, the one they were urgingat the Climax phase.

For example, following the breakdown of the Vance–Owen initiative in 1993, Secretary of State WarrenChristopher urged the European allies of the UnitedStates to support lift and strike, (i.e., lift the armsembargo and supply the Muslims with arms whileconducting air strikes against the Bosnian Serbs). Liftand strike was the threat Vance and Owen had usedunsuccessfully to try to get the Serbs to agree toVance–Owen. The European response was, “We stillsupport Vance–Owen. Why don’t you?”


Wanting to continue to press for the Vance–Owenposition, the Europeans accused Vance ofinconsistency in having abandoned it. Owen himselfargued against the air strikes he previously had usedas a threat, saying, “You will not solve the problem at10,000 feet.” The Europeans used what had nowbecome the Americans’ positioning dilemma, theiradvocacy of lift and strike even though they preferredVance–Owen, as an argument against the U.S.position. The same argument was used by others who,judging that Vance–Owen could not be revived,advocated withdrawal or other positions.

This example illustrates another point. What was thethreatened future during the Buildup and Climaxphases becomes a position to be taken up in theConflict phase. The future that you were using as athreat, all the time hoping you would not have to carryit out, becomes a position that you may have to arguefor against opposition from other subcharacters.


I face a deterrence dilemma in relation to a characteropposed to my position, who prefers the threatenedfuture to my position. The problem I face is that myfallback position does not pressure this character toaccept my position. I can eliminate my dilemma bygiving up my position. Alternatively, I can escalate toa higher level of retaliation by thinking up new cardsto punish the opposing character. Anger andindignation, leading to demonization of the opposingcharacter, accompanies this creation of new cards.

129Chapter 4

Conciliation is another way of eliminating the dilemma.To pursue conciliation, I retain the importantcharacteristics of my position while sweetening it untilthe opposing character prefers it to anything it can obtainfrom the threatened future. A tough-cop, tender-coproutine is a way of combining escalation with conciliation.

Certain subcharacters within our character mayprefer us to abandon our position and accept that ofanother player. They will tend to use a deterrencedilemma in relation to that other player as anargument for abandoning our position, on thegrounds that it is unrealistic.

I have an inducement dilemma if another’s position(different from my own) is as good for me as thethreatened future. This puts me under potentialpressure to accept that position, rather than go to thethreatened future. I can overcome this dilemma byfinding reasons to prefer the threatened future to theirposition, but in so doing I give them a deterrencedilemma. I escalate the conflict between us.

To avoid escalation, each side must accept itsinducement dilemma. It must try to eliminate it byworking out a common position with the other side.This means using passionately presented rationalarguments in the common interest.

Against an extremist opponent, the best strategy maybe, first, to eliminate our deterrence dilemma by mixingconciliation with a threat sufficiently strong that theycannot hope to escalate their way out of it; second, toeliminate our inducement dilemma by preferring tocarry out this threat rather than accept any extremistposition; third, to negotiate an acceptable common


position that will not now be extremist. Under suchcircumstances, cost-benefit analysis will be aninappropriate tool for evaluating the threatened future.Concepts such as honor, commitment, and adherenceto principle are appropriate.

I face a threat dilemma if I would be tempted to defectfrom the threatened future, if it were implemented.Unlike the inducement dilemma (with which it may,but need not, coincide) the deterrence dilemmaaffects characters at the Climax phase because theylook forward to how it would affect them if they movedto the Conflict phase and considered whether toimplement a future (the threatened future) fromwhich they are tempted to defect.

I face a positioning dilemma when I prefer another’sposition to my own. This is likely to occur when I haveabandoned a position I occupied with others andmoved to one I find less preferred, but more realistic.My dilemma will cause me to feel guilty when arguingfor the new position against the old, unless I canchange my principles sufficiently to prefer the new one.Having made this change, I will, as a consequence,tend to uphold the principles supporting my newposition even more strongly than those who havealways held it.

An alternative way for me to get rid of my positioningdilemma is to get those who still hold my old positionto move with me to my new one. This is how the Britishgovernment eliminated the positioning dilemma it facedwhen arguing with the Unionists in favor of admittingSinn Fein to talks although the IRA had not yetdecommissioned arms.

131Chapter 4

A positioning dilemma may face subcharacters of acharacter that has to put up or shut up (i.e., decidewhether to implement its fallback position afternegotiations break down). Subcharacters whopreviously argued for internal implementation of apeaceful solution now may find themselves arguingfor implementation of a punishing policy that was usedin the higher-level drama as a threat to pressure othersinto accepting the peaceful solution. Othersubcharacters may accuse them of abandoning theirprinciples in switching from a peaceful policy of theinternal game to a punishing one.

133

Chapter 5

When All DilemmasAre Eliminated

What happens when all six dilemmas areeliminated?

We have seen that while any dilemmas remain,characters are under rational and emotionalpressure to get rid of them; therefore, the situationis not fully stable (i.e., there is discontentment andpressure to change). This must continue until nodilemmas are left.

Is the situation then fully stable, with fully contentedcharacters? This question is important for thecommander of a peace operation that has at least apartial goal to bring about stability. The answer is yes,with several qualifications.

The Final State

It is true that the final state, where there are nodilemmas that give rise to pressures for change, isstable; all characters agree on a single position, andno character or group has any temptation to defectfrom this position. This theorem of the final state isproven in the mathematical appendix. It can be madeintuitively clear as follows:


• If all parties agree on a single position, then theabsence of any cooperation dilemma ensuresthat no character or group has any temptation todefect from it.

• If there is not a single position, then the only wayfor characters to eliminate their inducementdilemmas without creating deterrence dilemmasfor each other is to converge to a single position.

• If there is a single position and no cooperationdilemmas, then no other dilemma can existexcept the threat dilemma.

Note that the threat dilemma can still exist; however,this is unimportant because if all agree on a positionthat they can trust each other to carry out, then anylack of credibility in what they threaten to do, if theydo not trust each other, ceases to matter.

The final state is, then, a completely satisfactory solution,satisfying both emotion and rationality. Correct?

That is correct; however, we have said there arequalifications to bear in mind.

Tragic and Happy Endings

The first qualification concerns what is meant bysatisfactory. A drama may end tragically as well ashappily. (It also may end tragically in some respects,happily in others.) Both kinds of ending are satisfactoryto the audience of a drama (everything is settled);however, a tragic ending is unlikely to meansatisfactory completion of a commander’s mission.

135Chapter 5

We are really using the word satisfactory in a scientificand aesthetic sense. Whether the ending is happyor tragic, there is stability in the form of contentment(i.e., no character has anything left to hope for or tofear). In either case, this comes about partly throughmodification of initial hopes and fears during thecourse of the drama. At a tragic ending, broadlyspeaking, hopes have been destroyed and fearsrealized. At a happy ending, hopes have beenrealized and fears banished. Clearly, a commanderwants stability at a happy ending, although perhapsonly after hopes have been appropriately modified,rather than at a tragic one.

For an example of the tragic type of stable ending,consider the following generalized model of howcountries escalate their differences until they gladlygo to war. Let each implicitly or explicitly threaten tofight unless its political and diplomatic position isaccepted. Let each then eliminate its inducementdilemma by rationalizing a preference for war ascompared to the other’s position. Then, to eliminateany possibility of the other responding to this by furtherescalation, let it take its rationalization so far that waritself becomes its position, any compromise with theother being considered worse than war. War has thenbecome a totally satisfactory resolution; that is, it isnow a shared position from which no group of playersis tempted to defect. Pursue this reasoning further andwe find players escalating the type of war they considerbetter than any of the opponent’s positions until theyare committed to Clausewitzian total war with itsabsolutist demand for unconditional surrender.


The process in which friends and neighbors fromdifferent ethnic groups start to murder each other, asin Bosnia and Rwanda, may be explained as anotherkind of tragic ending. Each group knows, from taleshanded down, that the group its neighbors belong tohas in the past massacred them. Each group hearsofficially denied rumors that this group is arming togive itself the capability of doing it again; therefore,each group takes the precaution of arming itself.Finding now that its fears are confirmed, each groupfears that the other now prefers to massacre it, andhence rationalizes a change in its attitudes so that itprefers to massacre the other first. The future, “Eachtries to massacre the other before it can be massacreditself,” is now necessarily interpreted by any unit withina group as meaning “We succeed in massacring themfirst.” Unless and until this unit is massacred, this isthe logical way for it to interpret this future. Next, therationalizations each group has used to justify itspreference for massacring the other, based onhanded-down stereotypes, means that it now prefersthis to any compromise or settlement. Hence, “Eachtries to massacre the other” has moved from beingalmost the worst outcome for either to becoming atragic “totally satisfactory resolution.”

In general, escalation to a totally tragic ending happensin the following way. Each party has an inducementdilemma, which it solves by rationalizing a preferencefor the threatened future as compared to the other’sposition. Each party then has a deterrence dilemma,which it solves by demonizing the other and thinkingup credible cards that make the threatened futureworse for the other. Each then has an inducementdilemma, which it solves…, and so on. This cycle of

137Chapter 5

rationalizations continues until “Each doing the worstit can to the other” becomes a shared position elevatedby each above any possible compromise.

To break this cycle, each side needs to start solvingits inducement dilemma not by preferring conflict tocompromise, but by suggesting modifications of itsown or the other’s position to create a single positionboth can share. This is what happens whennegotiations proper start (e.g., during the 3-weekconference in Dayton, OH, that led to the Bosniaaccords). Parties’ thoughts are then directed towardagreeing on a single position, although they may notsucceed. Escalation, on the other hand, tends to gothrough its successive stages when the parties areseparated, lobbing pronouncements and symbolicactions at each other from a distance whileaddressing either their own constituents or interestedthird parties. It is when parties are not formallynegotiating that there is need for escalation to becontrolled and thought given to the construction of asingle positive position. Analyzing how to do this maybe the main contribution of confrontation analysis.

Dealing with the Details

A second qualification to bear in mind in interpreting“totally satisfactory resolution” is that the degree ofresolution of a conflict is relative to the model we areusing. Zooming in on a model to see more detailuncovers potential disagreements that disappear whenwe zoom out.

This is literally true of boundary disputes: the more wezoom in on an agreed boundary, the more potentialdisagreements are revealed about where exactly it


should run. It is also true of agreements on matterssuch as cessation of violence, economic arrangements,prisoner exchanges, and so on.

With this in mind, reconsider figure 2, where characterscycle between the Buildup and Climax phases untileither they fall into conflict or reach a single, totallytrustworthy position. If the latter, they look at the details(in the Resolution phase) to see if their agreement isreally stable (i.e., do they mean the same thing andcan they trust each other). If the answer is yes, theyproceed to the Implementation phase.

This is correct in theory; but in applications, is it realisticto expect a single, totally trustworthy position?Imperfect agreements reached with much distrustremaining seem more likely. Our answer is thatimperfect, mistrustful elements of the agreement arehidden by zooming out. This is realistic. Parties, toreach agreement, deliberately use ambiguous or over-general formulations; therefore, we model them withcard-tables where much detail is covered over by afew, general cards until all that is left is a single, totallytrustworthy position. That reflects the methods of real-life negotiators.

Nevertheless, you must look into the detail you havecovered up. That is, what parties do in the Resolutionphase. They do it to check that the overallunderstanding they have reached is sufficiently soundand reliable, not to settle all details, which would be anever-ending task. Card-table models can be used tosee the details of an agreement, as well as its overallshape. This is accomplished in the following steps:

139Chapter 5

• Start with a simplified model, representingparties’ general positions, to deduce conclusionsabout dilemmas and dilemma-elimination.

• Subject your conclusions to exploration andcriticism by adding cards and characters to themodel, particularly any that seem likely tooverthrow the conclusions. Assess their effect onhow parties, at the moment of truth, may sum upissues in the complicated model in terms ofanother simple model, perhaps different from thefirst. A principle tool for such assessment will beanalysis of subconfrontations betweensubcharacters. From your new simple model,deduce new conclusions about dilemmas anddilemma-elimination.

• Criticize your new conclusions again by addingdetails to the simple model. Continue until youhave a satisfactory model.

Technically, the technique of card-table modeling canincorporate any number of cards and characters. Thereason for using a simple model is not practicality; it isrealism. Simple models are realistic because thecharacters themselves, at a moment of truth, mustand will use a simple common reference frame to feelsure they understand each other. To model the framein terms of which they reach agreement by, a complexmodel would be unrealistic. Complex models, ascomplex as possible, nevertheless should be built.They are built to model not the moment of truth, butthe process by which characters look into theadequacy of their simple models.


Example of a Detailed Model (Northern Ireland,1993)—Context Cards

To illustrate, table 7 shows a complex model used torepresent and explore the position supposedly takenby the British and Irish governments in the DowningStreet Declaration of 1993.

The model not only represents the cards that, underour interpretation of the two governments’ position,should be played by the various parties innegotiations, it also shows a number of cards that(under the position) it was assumed would be playedby external characters. These form part of theassumed environment of the negotiations, and areplaced below the others in a sectioned-off part of thetable headed “Context.”

In this Context section, the symbol ~ is put againstcards about which either assumption (“will be played”or “will not be played”) could be made. Above thecontext line, the same symbol is put against cardswhere the Downing Street declaration took no position.

Note that filling in and interpreting a table like thisrequires detailed knowledge. This is useful; but it alsounderlines the limitations of a complex model. It isuseful in assessing the impact of the details thatunderlie general statements and raising questions thatmight be important. We can take a comprehensivelook at a mass of detail before summing it up in asimple model.

It underlines the model’s limitations by showing why acomplex card-table will be an unrealistic model of acommon reference frame. Characters cannot assumethat such details are common knowledge between

141C

hapter 5

Table 7. Northern Ireland options in 1993-1994, showing in detail the positiontaken by the British and Irish governments in the Downing Street Declaration.


them, hence they cannot put them into a model ofwhat they are sure each other knows. They know thateach other’s knowledge of details varies, with eachappreciating the details of its own situation much betterthan the other’s.

Implementing an Agreement—the Role ofSubcharacters Within a Character

How then are details settled that have been glossedover by a simple, overall agreement? Each characterconsists, in principle, of an organization, and each ofthe chiefs that have accepted the agreement mustorder it to be implemented within their organization.

This is another reason why the overall agreementcannot include much detail. High-level officials cannotdecide all the details of a complex operation; they mustdevolve them. Hence the orders given to implementan agreement, and so the agreement itself, must besimple and general.

Do lower-level decision-makers then simply fill in thedetails in accordance with their local knowledge? Notexactly. What happens is that the resolution of aconflict generally gives rise to a set of subsidiaryconflicts, as it is implemented through confrontationsbetween lower-level characters within eachorganization. Some of these confrontations are withinorganizations, some between them. When we modelthe overall agreement, we must realistically try toassess how such sequences of lower-levelconfrontations will work out in practice, just as thecharacters will try to assess the same thing at amoment of truth.

143Chapter 5

For example, in 1993 Serbian President SlobodanMilosevic, a player on the Serbian side who at thattime favored acceptance of the Vance–Owen plan,recommended it to the Bosnian Serbs on the groundsthat it could not actually be implemented. It would leaveSerbian gains more or less untouched. However, thisview was not generally accepted on the Serbian side,so it did not effectively become the Serbian view. Theincident shows both how characters attempt torealistically assess what different agreements wouldmean in practice and how a character’s views ariseout of confrontations between subcharacters, hereMilosevic and the Bosnian Serbs. (Silber and Little,1996, p. 279).

Often an agreement looks forward to the conflicts thatwill arise in its implementation by setting up asubsidiary organization to help by handing outrewards and penalties, as was done following the1995 Dayton accords.

There may be problems when a character innegotiations has insufficient power to order theimplementation of what it has agreed. In modelingterms this is not a problem. It is simply a matter ofcorrectly describing the cards given to characters. If acharacter cannot bring about a cease-fire, it shouldnot have a “Cease fire” card. Its card should be calledsomething else, perhaps “Recommend cease fire.”Names should indicate characters’ actual powers. Thisagain reflects how characters themselves will modelthe situation at a moment of truth.


Unforeseen Contingencies

A third qualification to bear in mind when describing aresolution as totally stable is that it only reflectscharacters’ projections of a possible future, to whichthe actual future is unlikely to conform.

Partly this is because of the previous point. Plans mustbe realized through sequences of confrontationsbetween lower-level subcharacters, and the results ofthose confrontations, whether resolutions, conflicts,or defections, cannot be foreseen in detail. Their totaleffect may make implementation of an agreementbetter or worse than foreseen.

We do not know the future, so the plans we agree onare not generally fulfilled, even when we honestly tryto fulfill them and all lower-level conflicts are resolved.Unforeseen contingencies arise, requiring theagreement to be interpreted to apply in circumstancesit never envisaged, so that it may need renegotiatingat various levels. This provides yet another reason foran overall agreement to be simple and general, as itmust provide a framework within which suchadaptations can be made. A clear yet generalagreement allows flexible responses to changingcircumstances while assuring subcharacters that theagreement remains in force, so that if they implementit they can continue to expect cooperation from eachother and support from higher authorities. Agreementsthat go too far into detail run the risk of being seen tooobviously not to apply when circumstances change.

For example, the history of the United StatesConstitution, or other constitutions, shows how asimple, general agreement is clung to, being

145Chapter 5

reinterpreted when necessary, because the frameworkit provides for subcharacters’ activities is too importantand valuable to lose.

We have said before that an agreement that mustsurvive under circumstances it did not foresee needsto be entered into in a positive spirit of goodwill, toreassure each subcharacter that others will strive toreinterpret and renegotiate it in ways that respecttheir interests.

When Conflict ResolutionBreaks Down

The conflict resolution process by nature may alwaysbreak down. It can fall into the Conflict phase of figure 2,the phase where parties prepare to implement theirfallback positions, having failed to bring about sufficientchange to enable them to move from the Climax phaseback to the Buildup phase.

In peace support operations, conflict normally meansresort to or continuation of armed violence. Suchviolence may not involve our own force. Others maydo the fighting, such as when UN peacekeeperswithdraw from a deteriorating situation (cf. UNEF’s1967 withdrawal from the Middle East). Armedviolence normally results. In other kinds ofconfrontation, such as between different componentsor coalition partners on our own side, conflict normallymeans a standoff with failure to cooperate, causing adiminution in operational effectiveness (e.g., failure toshare intelligence or meet minimum interoperabilityrequirements leading to loss of life through friendly-


fire incidents, equipment malfunction, or inability togive timely support).

In disaster relief or humanitarian operations,breakdown of conflict resolution may mean similarstandoffs, reducing operational effectiveness throughfailure to cooperate.

The Details of Implementing a Threatened Future

When parties in the Conflict phase, having failed tomake their threats sufficiently credible or awesome todeter others, are faced with having to carry them out,the emotions aroused may be strong enough toconvert the threatened future itself into a formalresolution of the conflict, as in the case of the tragicending discussed above. Such a tragic ending is acomplete, dilemma-less, totally satisfactory resolution;however, we do not say formally that it goes throughthe Resolution phase of figure 2. We say it goesthrough the Conflict phase.

We say this to maintain an essential differencebetween the Resolution and Conflict phases of figure2. Characters in the Resolution phase sit downtogether to examine the details of their sharedposition. In the Conflict phase, each character meetsseparately to discuss within itself (i.e., each holdingdiscussions between its own subcharacters) theimplementation of its fallback position. The charactersat the level of the confrontation as a whole do notcommunicate. (Their attempts to communicate wouldbe defined as attempts to go back to the Climax phaseand restart negotiations.)

147Chapter 5

Can the characters in a three-or-more personconfrontation get together and communicate insubgroups?

This too can be outlawed as a matter of definition.Characters must have a common view, as part of theircommon reference frame, regarding which subgroupsof characters would get together if the threatenedfuture goes into the Conflict phase. That common viewshould determine the definitions of characters andsubcharacters (i.e., we should define communicatinggroups of characters as subcharacters of a singlecharacter).

All this is a matter of formal definitions and modelingprocedures, but it reflects important real-worlddistinctions.

For example, after the breakdown of the Vance–Oweninitiative, the United States and European countriesconferred as to the next step they should take, finallydeciding not to implement their fallback position (liftand strike); likewise, the Bosnian Serbs conferred withthe Serbian government. The subconfrontationbetween them entered its own Conflict phase asMilosevic angrily cut off military supplies to the BosnianSerbs, although afterwards he resumed supplies. Thismeans that in analyzing the climax of the 1993confrontation over Vance–Owen, we should regardthe Serbs as one character, comprising the BosnianSerbs and Serbian government as subcharacters, andthe West as another character comprising the U.S.and European governments. This reflects the way theparties perceived which parties they belonged to(Silber and Little, 1996, p. 287).


The Conflict Phase Is Game-Theoretic

From this we get a useful theoretical insight. Thedecision problem facing characters in the Conflictphase, after they have absorbed the preference andbelief-changing impact of emotions andrationalizations generated at the Climax phase, andprovided they do not decide to go back to the Climaxphase, is essentially game-theoretic. All the methodsof present-day game theory (which has, in the main,turned its back on the tendency to explore and payattention to dilemmas) are applicable in principle atthis stage.

There are some qualifications to this. The internalconfrontations between subcharacters are not at allgame-theoretic. Moreover, characters may lookforward to communications in later confrontations withparties they are now cut off from by the breakdown ofnegotiations, and their preferences in their currentsituation may be influenced by the positions theyforesee in such later confrontations.

For now they must make a decision in the mannerassumed by game theorists (i.e., withoutcommunication, and hence by choosing the bestcourse of action within a fixed, given framework ofbeliefs and preferences). They must do this whetheror not the threatened future has become a tragicbut nevertheless final resolution caused by emotionsgenerated in the climax (i.e., whether or not theystill face dilemmas in their relations with othercharacters). If they do, the situation requires thatthey ignore them and simply pursue the dictates ofinstrumental rationality.

149Chapter 5

The traditional Clausewitzian military mission is bestunderstood in this sense. Rapoport (1968, pp. 69-77)points this out, while nevertheless emphasizing thelimitations of traditional quantitative game theory as adecision-support tool. Clausewitz assumes that militaryspecialists are given a mission to carry out againstenemy forces if political negotiations have brokendown and changed national attitudes have broughtabout passionate hatred between nations. Themission’s objectives are political in that its aim is touse force to create a situation where resumed politicalnegotiations will be more favorable to the nation. Itsconduct is non-political. During this military interlude,negotiations with the enemy concerning the issuesbeing fought over are out of the question, althoughtactical-level negotiations may be carried on over suchissues as cease-fires to remove bodies, localsurrenders, and declarations of open cities.Meanwhile, relations between the different players onour side, including the government, variouscomponents of the military, and our allies, are not atall game-theoretic. Confrontations between theirdiffering views are supposed to lead to willingcooperation in meeting the war’s objectives.

The Fog of War

Because the Conflict phase is game-theoretic, playersin this phase encounter the fog of war in Clausewitz’ssense. This is because Clausewitz’s fog of war hastwo causes not present in negotiations when we aretrying to guess at other parties’ true attitudes andbeliefs to pressure them into accepting our position.


The first cause of the fog of war, absent duringnegotiations, is that physical actions and operationsare being undertaken for their physical effects;therefore, they are undertaken secretly. If the reasonfor undertaking a physical action (e.g., abombardment) is to send a message, you will not tryto make it secret. If you do, the message may not bereceived. When the aim of the action lies in itsdestructive physical effects, keeping it secret has theadvantage of preempting countermeasures. Enemysecrecy causes fog.

In addition, physical operations tend to meetunforeseen contingencies, causing them to unfold inways we did not expect. This creates more fog as units’ideas of what is happening to each other diverge.

The second main cause of fog in the Conflict phase isthat, because we are not negotiating, we have nomessages to interpret. Someone sending you amessage is trying to make you understand something,and revealing themselves. There is a problem of whatto believe or not believe. That problem may be calledthe fog of confrontation. It is different in kind from thefog of war, where the problem is one of having nomessages to interpret.

Flunking and Defecting

We have discussed the Resolution phase and theConflict phase of figure 2 as if they are followed byimplementation, respectively, of the current commonposition and threatened future.

151Chapter 5

They may be. But we have said these are not the onlytwo possibilities. A third is interruption, caused by newexogenous information overturning the characters’common reference frame (e.g., Pearl Harbor upsettingthe frame of Anglo–United States negotiations). Apartfrom this, two other possibilities are the betrayedresolution and the flunked conflict. The first consistsof one or more characters defecting from the commonposition (e.g., the Nazis’ betrayal of the Nazi–Sovietpact). The second consists of one or more defectingfrom the threatened future (e.g., the West’s failure topunish Serbian rejection of the Vance–Owen plan).

There are, therefore, five ways a confrontation mayend, usually setting the scene for new ones. Of these,the decision to flunk is, we have said, game-theoretic(i.e., instrumentally rational). The preceding Climaxphase may have radically altered our attitude. Also,we may look forward to future confrontations in whichour credibility may be affected by the actions we takenow. However, given these considerations, we flunkbecause we consider flunking to be the best thing forus in light of our predictions of what others will do.

Consider the classic case of a terrorist who hasthreatened to blow up the plane he is on if his demandsare not met. They are not. It is put-up-or-shut-up time.He now must ask himself, given the emotionalconfrontation he has been through, and thinking ofthe future he must look forward to if he gives in,whether he really prefers death. Given his operativevalues and beliefs at the time, the decision he makeswill be instrumentally rational.

Is the decision to defect from an agreed position alsogame-theoretic? In principle, yes. Having gone through


the Resolution phase where the parties confirm theiragreement with each other, each must then separatelydecide whether to keep its part of the agreement.

Logically, this must be so. There must be some pointat which parties decide whether to keep to theiragreement, otherwise there would never be anydoubt about this, as there clearly is. Yet the game-theoretic nature of this decision is not so clear as inthe parallel case of the Conflict phase. For this thereare several reasons.

In the Resolution phase, parties generally have usedpositive emotion and common-interest argumentationto make themselves prefer to keep to the agreementrather than pursue any temptation to defect. In thisthey may have been successful, particularly if thediscussions have involved subcharacters within eachmain character. The result is that when they maketheir separate, game-theoretic decisions as to whetherto defect, they often find they now prefer not to.

Secondly, during the Resolution phase parties mayhave agreed to an implementation plan institutingincentives for lower-level decision makers to stick tothe agreement; they will have done so in order toeliminate temptations to defect. (For example, theDayton accords invited into the area the NATOImplementation Force.) By affecting the incentivesof subcharacters within each main character, thisagain makes the main characters prefer, on thewhole, to stick to the agreement.

Thirdly, parties who have reached an agreement oftenlook forward to a relationship where they will benefitfrom cooperating with each other in future

153Chapter 5

confrontations; therefore, it is important for them to buildup and maintain positive credibility (i.e., the belief thatthey will not defect from agreements). This gives thema reason not to defect from the present agreement.

To see the importance of the last two factors, comparetwo different scenarios in which a British colonel getslocal militia forces to agree to remove its weaponsfrom a certain area. In Scenario 1 the Britishcommander will, immediately after the agreement,withdraw his battalion to be replaced by forces ofanother nationality (e.g., Russian), with which Britishforces have poor liaison, and which the local militiathink are likely to be more sympathetic to their side.In Scenario 2, the British battalion stays to overseethe keeping of the agreement. We may suppose thatunder Scenario 1, the British colonel and local militiacommander, knowing what will happen next, gothrough a Resolution phase in which they expendmuch effort and goodwill convincing each other theiragreement will be kept; however, no matter what theysay, it is more likely to be kept under Scenario 2.


When all dilemmas have been eliminated, it isnecessarily the case that all characters agree on asingle position and can be trusted to implement it(see Appendix). In this sense, a totally satisfactoryresolution has been reached, subject to a number ofprovisos. First, “satisfactory” is meant in a scientificor aesthetic sense, even though in this sense a tragicending with hopes destroyed is as satisfactory as ahappy one in which hopes are fulfilled. In real life, wegenerally prefer happy endings. These are brought


about by managing the resolution process so thatcharacters attempt to overcome inducementdilemmas by negotiating a single position, rather thanby escalation. Secondly, resolution is model-dependent. A more detailed model, such as isexamined by the characters in the Resolution phase,may reveal disagreements that require renegotiationof what has been agreed. The card-table modelingtechnique may be used to examine the details of anagreement. Thirdly, most details are settled, not in theagreement itself, but in the course of confrontationsbetween subcharacters during implementation,confrontations that may or may not be resolvedsatisfactorily. Fourthly, we cannot know the future, soimplementation may turn out unsatisfactorily becauseof contingencies not foreseen in the agreement.Finally, a perfectly satisfactory resolution may bereached in terms of the common reference frame andcharacters’ communications, with one or morecharacters nevertheless intending or later deciding todefect from it, after having deceived the others intothinking they were trustworthy.

The same considerations apply, in a different way, ifthe confrontation ends not in agreement, but in conflict(which for defense forces usually means armedconflict). In the Conflict phase, characters separatelylook into details and confer between theirsubcharacters as to how to implement their fallbackpositions. Detailed consideration may cause them tochange their mind, and either return to the Climaxphase or simply not carry out their threats. When theydo attempt to implement the threatened future, itusually turns out unexpectedly, leading them into freshconfrontations.

155

Chapter 6

The Front-Line Play:A Dramatization of a

Confrontation Analysis

Most of this chapter consists of a play. It was writtenby the author under a contract with the United

Kingdom Defence Evaluation and Research Agency(DERA). It was performed at a seminar onconfrontation analysis held at DERA (PortsdownWest) on February 5, 1998. Its object was to conveyas vividly as possible how a commander tasked witha peace mission, and knowing nothing aboutconfrontation analysis, might use the method toformulate and begin to implement a strategy forwinning an Operation Other Than War (OOTW). It isreprinted here with the permission of DERA to givethe reader an example of the formulation andimplementation of a confrontation strategy.

A fictionalized peace mission is used, although withobvious resemblance to real ones, to focus attentionon the process of analysis and strategy formulationrather than on the details of a particular case. Thecharacters are British because the underlying researchwas done with British defense forces.


Would It Be Done This Way?

In the play General Deloitte, commanding a combinedjoint task force, analyzes his situation by calling in aconfrontation analyst on his staff. The general knowsnothing about the method; it is something he has justheard about. He analyses the immediate problem athis level, then asks the analyst to further analyse thefollowing items:

• Grand strategic problem that gave rise to hisbeing tasked with this mission

• Other problems at his own level, bothsimultaneous with the immediate one and to beexpected as the campaign develops (i.e., thelinked sequence of confrontations)

• Internal problems of coordination between differentforces in his coalition and different components

• Problems at lower levels, consisting of tacticalconfrontations that should be resolved in waysthat support and are supported by hisoperational strategy.

What can be said about the organizationalarrangements under which the general does this?Although better than doing no confrontation analysis,they are imperfect. The general and his chief of staffshould have learned the method in normal trainingprocedures. They would not then depend on the analystfor guidance at every step, as they do in the play.

As he does in the play, the analyst would depend onthem for the strategic decisions that would underpinhis more detailed, lower-level models. For these

157Chapter 6

models he would need more staff than himself, andthe models would need to be developed throughinteractions with responsible commanders at all levels.

In the next chapter we explore the principles ofconstructing a confrontation strategy. While readingthe play, remember that here we are showing thefundamentals of the method by taking the case of acommander who has encountered it for the first timeand is learning it at the same time he is using it. Thisis not meant to be ideal.

The Play

[The scene is a partitioned area of the banqueting hallof a tourist hotel near Morubwe, the capital of the NorthAfrican state of Morya. The hotel has beenrequisitioned for HQ UNFORMOR (United NationsForce in Morya). The particular area we are in is theoffice of the commander (UNFORMOR), Major-General Eric Deloitte, CBE. It contains the general’sdesk, a large hotel dining table with chairs set roundit, and a flipchart.]

[Commander UNFORMOR is British because Britainhas been designated the framework nation for the UN’sintervention in Morya. Under him is a British brigadiercommanding the Anglo-Egyptian land forces, a U.S.Air Force brigadier-general commanding the air forces,and a colonel of the Armée Francaise commandingthe aviation force, a battalion of helicopters availableto support the land forces.]

[General Deloitte, a determined-looking, middle-agedman in combat dress, is standing looking discontentlyout of the window. Seated at the table is his chief of


staff, Brigadier Ray Jones. The chairs are drawn backfrom the table and papers are scattered on it. Evidentlya meeting has just finished. On the flipchart arescrawled the names of some of the main protagonistsin the Moryan drama]:

MORYAN GOVERNMENT

ISLAMIC REVOLUTIONARY ARMY (ISRA)

GOVERNMENT OF PELUGYA (long commonborder, supplies ISRA)

FRANCE (tends to support Moryan govt)

EGYPT (strongly against ISRA)

ARAB COUNTRIES (some support for ISRA)

[Brigadier Jones is busy at the table sorting throughpapers and making notes.]

DELOITTE: Is everything ready for the pressconference?

JONES [still writing notes]: Yes sir. At the airport.Fourteen hundred. We should leave 1330.

DELOITTE [irritably]: What’s the right way to handlethis? It’s no good, you know, Ray. We’re not there.

JONES [raising his eyebrows and turning to look atthe General’s back.]: Really, sir? This morning’smeeting was the most thorough review of the situationwe’ve had. I’m just writing up the notes for you. Eachcomponent reviewed the situation from their viewpoint.Intelligence gave a good overall assessment. We knowwith fair accuracy what’s going on, how the reliefconvoys are getting through, etc.… As to the press,

159Chapter 6

I’ve seen you handle them. I’m sure you can refrainfrom giving away anything that’ll further inflame thesituation.

DELOITTE [disgustedly]: That, precisely, is not thepoint. This press conference should be my first movein a planned campaign to reach my objectives. Itshouldn’t be a matter of giving nothing away, of givingno hostages to fortune. But I’ve nothing resembling aplan. [Turning to face Jones] That’s the trouble, Ray.Soldiers know how to fight battles. I can plan for abattle. But this, this is not anything resembling a battle.The objective here is to get your way without the useof force. We are not trained to do that. That’s apolitician’s job.

JONES [wisely]: Peacekeeping isn’t a soldier’s job,but only a soldier can do it.

DELOITTE: Who said that? Never mind, it’s quiteright… [Goes back to gazing morosely out of thewindow. Then starts again, waving his hand at theflipchart.] You know, Ray, I’ve a hunch the Moryangovernment, not ISRA, will be the one that refuses tocooperate in peace talks. Apparently we still have notgot the President to agree to a time and place for anemergency meeting with ISRA. That’s why they calledme out of the meeting half an hour ago. I told them tokeep trying.

JONES [surprised]: The government make difficulties?Why should they? Our intelligence is that they’re underthe most pressure to end the fighting. They have mostto lose from it continuing.

DELOITTE: That was the position. My hunch is thatthings have changed. I suspect they think all this world


publicity against ISRA has tipped the balance in theirfavor. They think it’ll force us to intervene on their side.So, the worse things get for them the better. That, Isuspect, is their thinking.

[Jones shakes his head disgustedly.]

DELOITTE: Meanwhile, ISRA will continue theircampaign, regardless what the world thinks. And I’msupposed to do something about it. Without usingforce. But what? [Musingly] You know, I’m trying toremember something I heard recently. About a newthing called confrontation analysis. That’s what I need.After all, that’s what PJHQ has thrown me into. Aconfrontation. That’s the term for it.

JONES [hesitantly]: Sir…

DELOITTE: Yes?

JONES: We’ve someone on the HQ staff who’s trainedin confrontation analysis. He is a civilian analyst inOA called Bright, Mark Bright. DERA sent him along.We haven’t used him in that capacity. OA have hadtheir hands full programming databases for us. But Icould get him in to tell you about it if you like.

[The Commander looks at Jones thoughtfully, as ifhis mind were on other things. Then he shakes hishead as if to reproach himself for indecision, andglances at his watch.]

DELOITTE: Send him in.

JONES: Yes sir.

[Jones gets smartly to his feet and leaves the room.The general returns to his morose vigil at the window.

161Chapter 6

Almost at once Brigadier Jones returns leading MarkBright, an intense and slightly supercilious young manin civvies. The general turns round. Bright, withoutactually saluting, stiffens and inclines his head as if toindicate military readiness.]

JONES: This is Dr. Bright, sir.

DELOITTE [regarding him curiously and somewhathumorously]: Sit down, doctor. [Bright does so.] Tellme about this new discipline DERA has trained you in.

BRIGHT [promptly]: Yes sir. Confrontation analysisis an OA tool. You analyze situations where each partytakes a position, meaning a suggested solution, whatit suggests everyone should do, itself included.Generally, parties’ positions don’t coincide. They’reproposing different solutions. So there’s a conflict ofwills. As well as their positive positions, each party ina confrontation has a fallback position, what it saysit’ll do if its position is not accepted.

DELOITTE: That defines a confrontation?

BRIGHT: Yes sir. In a way confrontations correspondto battles. To win a war, you win a linked sequence ofbattles. To win a peace support operation, you couldsay, you fight and win a linked sequence ofconfrontations.

DELOITTE [in a lively tone]: Well, suppose I’m in aconfrontation. How do I analyze it?

BRIGHT: Well, once you’ve decided on everyone’spositions, you look for the so-called dilemmas facingeach party. Essentially, dilemmas are credibilityproblems. For example, I have a threat dilemma if youthink I would not carry out my threat. In other words, if


you think I’m bluffing. I have a trust dilemma if I couldn’ttrust you to carry out my position, even if you agreedto it. There are four other dilemmas.

DELOITTE: They sound realistic enough.

BRIGHT: They’re extremely realistic, in myexperience, sir.

DELOITTE: Can you always find these dilemmas? Inany confrontation you look at?

BRIGHT: They’re guaranteed to exist, sir, unless anduntil all parties have agreed on a single solution theycan trust each other to carry out. In other words, ifthere’s no dilemma, there’s no problem. If there’s aproblem, the parties are necessarily facing dilemmas.So by eliminating all dilemmas, you solve the problem.

DELOITTE [bridling]: You do, do you? On whoseterms? Whose solution do you end up supporting?

BRIGHT: We have to make sure it’s our solution thatwins, don’t we, sir?

DELOITTE [loudly]: And how the hell do we do that?[He leans forward with his hands wide apart on thetable looking down at Bright, who sits across from him.]

BRIGHT [quietly, looking down at the table]: It’s amatter of making a plan and following it. Winning aconfrontation is like winning a battle, only different.The details are different, but the principle’s the same.You exploit the other side’s weaknesses and our ownstrengths. First off, you analyze the dilemmaseveryone faces. Dilemmas are points where we canexert pressure on them, or, if we’re not careful, themon us. Next, you make a plan to eliminate dilemmas

163Chapter 6

in a sequence of operations that will bring the othersinto full compliance with your objectives. [He looks upto find the general staring fixedly at him. He pausesuncertainly, then continues more quietly than before.]It’s a logical process, sir. Supposedly, it’s as logical,or more so, than winning a battle. Of course, you’vestill got the fog of war to contend with. I mean, youhave to make assumptions because you can’t be sureof the facts, and your analysis is dependent on thefacts you put in. But you’ve got to act. So you makethe best assumptions you can, whilst being ready tochange your plan, without vacillation or confusion, ifyour assumptions prove to be wrong… [His voicefalters as the general continues to glower at him.]

DELOITTE: You’re quoting Clausewitz.

BRIGHT: Well…

DELOITTE: That’s straight Clausewitz.

BRIGHT: Yes sir.

DELOITTE: What does winning mean when you’re ina confrontation?

BRIGHT [without hesitation]: Bringing others into full,willing compliance with your objectives.

DELOITTE [turns his back and walks to look out ofthe window. After a pause]: Okay, suppose I want toanalyze the confrontation I’m in now. How do I do it?What do I need? [Looking at his watch.] How muchcan I do in one hour?

[Bright looks appealingly at Jones, who returns hislook impassively. The general continues to gaze outof the window, his back to them.]


BRIGHT: Er, well, sir, I could take you through a broad,high-level analysis. I’d sort of ask you a series ofquestions and show what follows from the answers.You’d be responsible for all assumptions, andobviously decisions taken on the basis of the analysiswould be yours. I’d be responsible for the process ofanalysis, showing you how to do it…step by step…

JONES [imperturbably]: Equipment needed?

BRIGHT: Just a flipchart. In an hour, we might…mapout a broad policy, though probably it’d raise a lot ofquestions for intelligence to answer. We’d have to domore detailed analysis and checking later on. Thenwe could make detailed plans to be implemented atvarious levels and by various components. [Lookinganxiously at Jones while still talking to the general,whose back is still turned to them] I’d have to start,sir, by asking what decisions have to be made oractions taken in an hour’s time.

[A long pause.]

DELOITTE [turning round briskly and coming to sit atthe table]: Of course. [Waving his hand at the flipchart]:There’s your equipment. Ray, perhaps you’d bettersit in on this. I may need your views on some ofthe…er…assumptions. [To Bright]: Let’s go.

BRIGHT [standing up awkwardly and going to theflipchart]: Yes sir. Er…

DELOITTE [sitting easily, his hands wide apart on thetable]: To answer your first question. I have to give apress conference at 1400 hours. I could just regard itas a nuisance, and give away as little as possible.That would be our usual style, I suppose. [He looks

165Chapter 6

mischievously at Brigadier Jones, who does notrespond.] Instead, I want to handle it as the first stepin a PLAN…to achieve my OBJECTIVES. So, at 1400I’ll be facing reporters and cameramen from all theworld’s media. These are the people who’ve whippedup public hysteria in Britain, America, and Europe overthe Friday massacre attributed to ISRA. Dead babieson television. All that kind of thing. Horrified the wholeworld. The Cabinet met to discuss it yesterday. TheDefence Secretary contacted the CJO, who called mein. I got back from PJHQ early this morning.

BRIGHT: And you came back with a new mission, sir?

DELOITTE [sardonically]: My mission, if you can callit that, is to get the politicians out of trouble. Somehowor other, I have to stop ISRA massacring people, or atleast, stop it getting into the news. Once again, we’vebeen landed with what ought to be the politicians’ job.But CJO overrode my objections. Said we’ve got todo something.

BRIGHT [hesitantly]: So, sir, your objective is…? Howwould you state your strategic objective?

DELOITTE [after pausing to think]: To get a cease-fire between ISRA and the government, with anagreement between them to start discussing asettlement. That’s my objective. To be achievedwithout using force.

BRIGHT: If possible…

[The general does not answer.]

BRIGHT: I mean, it’s possible the only way to getyour way without using force is to be ready to useforce, if necessary. Wouldn’t you agree?


[The general gives a short, bitter laugh. Brigadier Jonessmiles wanly.]

DELOITTE: You’re right, in general. But you’restraying into a conceptual minefield. Doctrine, so faras we’ve got one, distinguishes peacekeeping, whereyou don’t use or threaten force, from peaceenforcement, where you do. We’re supposed to behere in a peacekeeping role. On the other hand, apeacekeeping situation is one where the parties areprepared to be peaceful. In Morya, they’re not. Sowe’re in a peace enforcement situation with apeacekeeping mandate. What do we do about that?

BRIGHT [rising to the challenge, picking up a pen andfolding up the flipchart to disclose a blank sheet ofpaper]: I can’t, of course, tell you what to do, sir. [Jonesraises his eyebrows at this naïve remark.] All I can dois help you analyze the situation to see what’s possible,that is, how far it’s possible to get peace withoutthreatening force. And if so, how. Now…

First Narrated Interlude

[As he raises his pen, the narrator steps forward andlifts his hand. The three characters freeze in position.]

NARRATOR: General Deloitte, assisted on requestby Brigadier Jones, now answers Mark Bright’squestions as to the positions currently being taken bythe parties. This enables Mark to draw up this card-table. [He folds up the blank sheet on the flipchart todisclose a sheet, shown here as table 8.]

NARRATOR: As Mark explains to the general, hehas modeled the confrontation by giving eachprotagonist certain cards to play or not play. The

167Chapter 6

various parties’ positions are then shown by listing, inseparate columns, the cards each party proposesshould be played.

So the UNFORMOR position, column U, is that bothsides should cease fire and join peace talks.UNFORMOR then will not call air strikes against ISRA(although this is what the Moryan government wants)nor recommend UN withdrawal (which is what ISRAwants). Note the name given to the card, “Recommendwithdrawal.” The general can’t actually decide on a

Table 8. Moryan government no longer wants a cease-fire.


withdrawal, he can only recommend it to his superiors.Hence the name.

ISRA’s position, column I, is that the UN shouldwithdraw, and therefore that the general shouldrecommend this to his superiors. He shouldn’t callair strikes against ISRA positions. ISRA and theMoryan government should be left to fight it out. ISRAexpects to win the resulting civil war and install anIslamic regime.

The Moryan government’s position, column M,apparently has just changed. It seems unwilling toattend the meeting the general has called, whichindicates that it’s now taking the position that thereshould be no cease-fire or peace talks. Instead,UNFORMOR should intervene on its side by callingair strikes against ISRA. Until now, while continuallyasking for air strikes, the government hasn’t objectedto the idea of a cease-fire followed by talks.

These are the parties’ positive positions. But aconfrontation consists not just of positive positions. Itincludes parties’ fallback positions as well, theunilateral actions they say they’ll take if others don’tconvincingly agree to their positions. If all parties carryout their fallback positions, we get something calledthe threatened future. This is shown in column t. It sohappens that in this case it’s the same as the defaultfuture (column d), the future to be expected if everyonecontinues their present policies without change. Thisis a future under which ISRA and the Moryangovernment will continue to fight while UNFORMORlimits itself to humanitarian aid. So UNFORMOR doesnot call air strikes nor recommend withdrawal, while

169Chapter 6

ISRA and the Moryan government neither cease firenor join talks.

This is the future that has led, through one particularincident, to the present crisis. The general situation isthat government forces are maintaining control of thecenter of Morubwe and other big towns, while ISRAcontrol most of the country. ISRA are being suppliedmilitarily by the neighboring government of Pelugya,which has a radical, fundamentalist Islamicgovernment, while bombarding government-controlledareas, causing deaths which tend to get reported bythe Western media. The particular incident which hasnow caused an international outcry and led to thegeneral being told to do something is the bombardmentof a crowded market, which caused sixty deaths.

And the background to all this?

Three years ago ISRA supposedly won an election. Butthe secular government of Morya, headed by PresidentSaldin, hoping for Western support, annulled theelection results. ISRA thereupon began a bloody revoltand Saldin asked for Western help. The UN approvedthe sending of UNFORMOR with a mandate not to takesides, but to protect humanitarian relief missions whileencouraging the parties to agree on fresh, UN-supervised elections. ISRA have rejected this plan,arguing that they have already won elections. PresidentSaldin has gone along with the plan until now in orderkeep Western support, but he really wants the West tointervene militarily and help him defeat ISRA.Meanwhile, the UN has banned military flights andimposed sanctions against Pelugya for supplying ISRA,while UNFORMOR has brought in aid to the populationsof the cities and sent convoys to aid rural areas, entering


into negotiations with ISRA to allow them passage. Thusit has built up cooperative relations with both sides inpursuit of its humanitarian objectives.

[The narrator waves his hand, causing the charactersto come back to life. He walks off the stage and leavesthem to it. The general stands up and leans forwardacross the table, gazing intently at the flipchart.]

Resumption of the Play

BRIGHT: So it seems this [indicating the flipchart] isthe moment of truth in the drama between you, ISRA,and the Moryan government. Obviously, there areother dramas. There’s the higher-level, GrandStrategic drama between governments—the U.S.,Britain, France, the Moryan government, Pelugya,various Arab governments. That’s the drama that ledto the formation of UNFORMOR, that is, it’s the dramathat defines your superior’s intent at two levels above.

DELOITTE [sardonically]: If there is such a thing.

BRIGHT: I mean, sir, that the requirement for acommander to understand his superior’s intent, twolevels up, is not very clear in this kind of UN peaceoperation. In this kind of case superior’s intent is theresultant of political interactions between governments.So understanding that grand strategic drama isessential. So is understanding the lower-level, tacticaldramas that are also going on between the forces youcommand and other forces, as well as NGO’s, aidorganizations, communities, and so on. You have todirect those lower-level dramas to implement a whole,cohesive plan for achieving your objectives.

171Chapter 6

JONES [glancing at his watch]: We don’t have time forall this, Dr. Bright. The general meets the press at 1400.

BRIGHT: I understand that, sir. Analyzing those otherdramas will have to be done later, in follow-up work.[Gesturing at the flipchart] This is the one we’llconcentrate on now, but we have to remember thoseothers in the background because we’ll be makingassumptions about them.

DELOITTE [as if talking to himself]: How do I get themto agree on column U?

BRIGHT: Ah. Yes, sir. First we have to analyze thedilemmas everyone faces.

DELOITTE [with a quick gesture]: Show me those.

BRIGHT: I’ll have to ask you some questions aboutplayers’ preferences. First…Isn’t it true that both ISRAand the Moryan government prefer this, the threatenedfuture (the same as the present, default future) to ourposition, column U? Isn’t that correct? [As Bright refersto each future, he points to the corresponding columnon the flipchart.]

DELOITTE: Hmm…I don’t know about that. ISRAwants us out…Saldin wants us to intervene againstISRA…

BRIGHT: Right. The present future is not the positionof either of them. But in the case of the Moryangovernment, surely it hopes, by milking the presentsituation, to get the UN to tell us to intervene againstISRA. Notice, here we’re talking about its assumptionsconcerning the grand strategic drama.


DELOITTE: In that sense I believe you’re right.Correct, Ray?

[Brigadier Jones nods.]

BRIGHT: So, taking into account those assumptionsabout the grand strategic drama, the government doesprefer to continue with the current future, the same asthe threatened future, rather than move with ISRA toour position. [They both nod.] Okay. Now,ISRA…Surely ISRA believe the current conflict will endin their victory.

JONES: That’s why they want us out, Dr. Bright.

BRIGHT: Yes, in other words they’d most of all preferthis, their position [indicating column I]. But betweencontinuation of the present conflict with us here butremaining neutral, and a cease-fire with peace talks(which is our position), they’d prefer the present conflict.

DELOITTE: Correct.

BRIGHT [triumphantly]: Then we have it. We face adeterrence dilemma. Fatal. We must solve it, or ourposition is untenable. Unrealistic.

DELOITTE [sardonically]: Presumably you’re tryingto say that because they prefer being where they are,they’re under no pressure to accept our position.

BRIGHT: Exactly. Or rather, they prefer what we’rethreatening them with, which in this case happens tobe the same as the present, default future, to ourposition. So, as you say, we’re placing no pressureon them to accept our position. U is, at present, simplyout of the question. That’s our deterrence dilemma:we’ve got no deterrence.

173Chapter 6

DELOITTE: I take your point. I congratulate you, Dr.Bright, in having gone to the heart of my dilemma. [Hewalks up and down, hands behind his back, frowning.]The position seems to be rather hopeless.

[Jones looks at the general with concern.]

BRIGHT: Er…excuse me. [Jones looks at him sharply.]May I go through the ways of solving a deterrencedilemma? [The general stops walking and looksinquiringly at him.] One: you can, of course, changeyour position. You must do that if all else fails. Two:you can persuade one of the other parties that thethreatened future, or anything they might do in reactionto it, actually makes them worse off than our position.

DELOITTE: But we’ve just said they prefer it.

BRIGHT: But why do they? Take the Moryangovernment. They only prefer the threatened futurebecause they’re hoping for certain grand strategicreactions, press hysteria, leading to a demand forintervention against ISRA. Can’t we persuade themthat will never happen?

JONES: We can’t get into playing politics.

BRIGHT [to the general]: Sir, could we brainstormthis a bit? Without analyzing the grand strategic drama,which we haven’t time for, perhaps there’s somethingyou could do, make recommendations to your politicalmasters, to get a grand strategic message sent toPresident Saldin to discourage him from expectingintervention. Or is there some way you could use, orthreaten to use, the press conference?

DELOITTE [staring at him]: Of course there bloodyis! Ray, remember what Intelligence was telling us.


There’s evidence the Moryan government may havebombed their own people, that this whole thing wasset up by them. That fits in with the way they seem tohave suddenly switched their line against peace talks.

JONES: But sir, it was decided not to usethat…evidence.

DELOITTE [belligerently]: Decided? So what? [ToBright] Let’s think about this. If at the press conferenceI were to present evidence that the government itselfis to blame for the bombing…[Thinking again] But no.Press reaction might be unfortunate.

BRIGHT: Sir, like most threats, that might be moreeffective as a threat than it would if carried out. If Imay, let me give you a card…

Second Narrated Interlude

[The narrator steps forward and lifts his hand, freezingthe actors in position. He raises the sheet on theflipchart, revealing table 9.]

NARRATOR [pointing to rows and columns in table 9to illustrate what he says]: Mark now adds the row,“Blame the government,” to represent a card held byUNFORMOR. The general then decides to declarehis intention to play this card if there’s no movementin others’ positions, specifically, if the Moryan presidentrefuses to talk. In other words, he makes it part of anew fallback position for UNFORMOR, part of a newthreatened future, shown in column t. Brigadier Jones,incidentally, points out that ISRA have in fact blamedthe Moryan government for the bombing, whichindicates that for UNFORMOR to blame them will bepart of ISRA’s position, too (column I).

175Chapter 6

[Turning to the audience]: Now, what’s going on here,looked at from a technical viewpoint? The general hasreacted to his deterrence dilemma, pointed out to himby Mark, by reframing the situation, adding a card tohis hand, and inserting that card into the threatenedfuture. This eliminates the dilemma by making thethreatened future worse for the Moryan governmentthan his (the general’s) position. Notice the emotionwith which he carries out this reframing. He becomesangry at the Moryan government. But he doesn’t getangry on perceiving his dilemma (then he was merely

Table 9. The general has thought up anew card, “Blame the government.”


dejected, walking up and down, fearing he might haveto abandon his position). Nor did he get angry, muchearlier, at the Moryan reluctance to talk. He got angrywhen starting to think of a way to make the threatenedfuture worse for them. Anger is the emotion thataccompanies thinking up an adequate threat.

Having thought up the card, the general’s angerdisappears while he considers how to use it. Does heface any dilemmas in making it credible that he willuse this card? Watch…

[He waves his hand again, bringing the actors back tolife.]

Second Resumption

BRIGHT [pointing to the new card]: “Blame thegovernment.” If you make that part of the threatenedfuture…

DELOITTE [thoughtfully]: …then from PresidentSaldin’s point of view, that threatened future would bepretty bad. With us blaming them for the bombing, howcould they expect the West to intervene on their side?

BRIGHT: Exactly. Adding this card makes thethreatened future worse for them than your position,and so it eliminates your dilemma. Instead, it givesthe Moryan government a dilemma, an inducementdilemma, consisting in the fact that they now preferour position to the threatened future.

DELOITTE: You call that an inducement dilemma?

BRIGHT: Yes, sir. It puts pressure on them to acceptour position. It’s good for us, bad for them. We, on the

177Chapter 6

other hand, don’t prefer their position to the threatenedfuture. We’ve no inducement dilemma…Let’s see,now, what about a threat dilemma…

DELOITTE: Hopefully I will not have to actually playthe new card…The threatened future is one thatneedn’t actually be implemented…

BRIGHT: Correct, sir! It’s a threat that lies in thebackground, hopefully making them change, becauseif there’s no change, it represents what’s expected tohappen…But, if I may, can we look at other dilemmas?Is the new card, “Blame the government,” one you’dwant to play, sir? For its own sake, I mean, apart fromputting pressure on the president?

JONES: No!

DELOITTE: Er… no. I’d obviously prefer not to takesides, upset world opinion, act against the very sidethat’s favored by our own and alliedgovernments…Definitely not preferred.

BRIGHT: Then, sir, we do face a threat dilemma,consisting in the need to make credible a threat we’drather not carry out. How to overcome it? Well,assuming we don’t want to change position, we canuse emotion. Getting angry could make them thinkwe’ll carry out the threat, even while preferring notto…Better, can we rationalize a change of preferences,that is, give reasons why we would actually prefer toor be forced to carry out the threat? Can we think ofany such reasons?

JONES: May I make a suggestion, sir?

DELOITTE: Yes?


JONES [delicately]: You might indicate that it wouldbe your duty, if they refuse to talk, to reveal theevidence that puts blame on the government. Thereis, perhaps, a perception on their part of the Britishcharacter…

BRIGHT: Right! I see what you mean! A perceptionthat the British can be unreasonably obstinate inpursuit of inexplicable principles! [Jones nods.] Thatway, sir, you could simultaneously admit all thereasons why you’d regret revealing this evidence tothe press, while making it clear that you’d feel forcedto do so. On principle…

DELOITTE [decisively]: Ray, can we get through toSaldin, on the phone?

JONES: Yes sir. The communications room…

DELOITTE: I’m going to do this right now, before thepress conference. I can see just how to put this to him.

[He leaves the room. Bright becomes plunged inthought in front of the flipchart. Jones stands up, walksin front of the table, folds his arms, and gazesthoughtfully at Bright’s back.]

JONES: Dr. Bright, may I ask you something?

BRIGHT [turning round]: Of course.

JONES: Is there an…ethical aspect to use of theseprocedures? It seems somewhat…Machiavellian. Nota straightforward way to deal with people.

BRIGHT: I see. You’re right, in a sense, when we’reusing the procedures in this way, for decision supportof one party. Then it’s a matter of helping that party to

179Chapter 6

get its own way. We do it by helping it to behavenaturally, that is, transmit emotions and argumentsappropriate to its situation. That’s all right, as far as itgoes. But it also means making its behaviorconsciously appropriate and functional, while leavingall the other parties to stumble around and mess uptheir chances. So the situation is asymmetrical. You’reshowing one side how to win.

JONES [nodding thoughtfully]: Which is what we want,of course…

BRIGHT: I assume so. I’d like to assume an Alliedpeace-keeping mission is generally in the right, andought to get its way. In another kind of situation, wecan use the same kind of procedures for mediationsupport instead. That means working simultaneouslywith all the parties, getting them to solve dilemmas insuch a way as to work out a cooperative solutiontogether…

[He tails off as the general re-enters the room. Jonessmartens up and turns his attention to his commander.]

DELOITTE: Right. New situation. Assume thepresident’s changed his position.

Third Interlude

[The narrator steps forward, raises his hand to freezethem in position, and lifts another sheet on the flipchartto reveal table 10.]

NARRATOR: General Deloitte’s blunt speaking toPresident Saldin has shifted the Moryan governmentinto acceptance of UNFORMOR’s position. Thegeneral now wants to press home this victory. So Mark


Table 10. After the Moryan president hasaccepted the UNFORMOR position.

sets up a card-table in which UNFORMOR and thegovernment occupy the same position.

He then asks, what is now UNFORMOR’s fallback? Not“Blame the government.” That’s not necessary any more,now the government has accepted UNFORMOR’sposition. Does then the fallback consist of doing nothing?If so, the general will face another deterrence dilemma,since ISRA (the only player now openly rejecting his

181Chapter 6

position) will prefer such a threatened future to thegeneral’s position.

They reluctantly decide that threatening to call airstrikes is all they have to bring pressure on ISRA. Thisis an action the general would be most unwilling totake. It would jeopardize the relief convoys that dependon ISRA’s cooperation to get through. Also, it mightcause ISRA to retaliate against UN personnel…butas soon as this possibility is suggested, Mark pointsout that it is not in the model. It has to be modeled asan important ISRA card. He puts it into the card-tableand makes it part of the threatened future, ISRA havingmore than once indicated that this is what it would doif UNFORMOR were to order air strikes.

The discussion now continues…

[The narrator waves his hand, bringing the charactersback to life, and leaves the stage.]

Third Resumption

BRIGHT: Okay, so ISRA will threaten us withretaliation. Fair enough. But is their threat credible?Haven’t they got a threat dilemma? Does retaliatingagainst UN personnel make the threatened futurebetter for ISRA, in and for itself, disregarding anypressure it puts on us?

DELOITTE [dubiously]: Hard to say, in and for itself.That is, forgetting about possible internationalrepercussions.

BRIGHT: Just a moment, sir, we shouldn’t forget aboutthem. They’ll be part of the grand strategic game,which, remember, we have not analyzed, but have to


consider when analyzing this one. Internationalrepercussions…Let’s see…you mean that if theyretaliate against UN personnel, it’s likely the UN andthe powers behind it would demand interventionagainst ISRA?

DELOITTE: Precisely. Don’t you agree, Ray?

JONES: I do, sir.

BRIGHT: Okay, then if we take grand strategicrepercussions into account, ISRA should have a threatdilemma in threatening retaliation. If they realize it. Imean, if they realize the likelihood of repercussions,they should prefer not to retaliate against UNpersonnel…If they don’t realize it, perhaps we shouldpoint it out to them! Which suggests somethingelse…We might at the same time get rid of our ownthreat dilemma, I mean the fact that they’ll think usreluctant to carry out air strikes.

DELOITTE: There you’ve lost me, I’m afraid. Go overthat slowly.

BRIGHT: Yes. I mean this, sir. We’ve established, Ithink, that heavy, sustained air strikes against ISRApositions would solve our deterrence dilemma. Thatis, rather than suffer that, they’d prefer to accept ourposition, which is a cease-fire and peace talks.

JONES: …Although we’ve said their acceptancemightn’t be very trustworthy…

BRIGHT: Right, agreed. I’m not forgetting that. Thatis to say that accepting our position might give ISRA(and the government, too, come to that) a cooperationdilemma, in that they’d be accepting a position it paysthem not to honestly implement. Likewise it’d give us

183Chapter 6

a trust dilemma, we’d have got them to agree tosomething we can’t trust them to do. But those aredilemmas to be tackled once we get overt acceptanceof our position. Getting overt acceptance is the firststep. That’s what we’re discussing now.

DELOITTE: One step at a time.

BRIGHT: Exactly, sir. Now, to get overt acceptancewe must make it the best alternative for them that theycan see. That means, first, making the alternative, thethreatened future, worse for them, that is, solving ourdeterrence dilemma. Air strikes do that. Correct?

DELOITTE: I believe so.

BRIGHT: Second, it means making air strikes credible.That means getting rid of our threat dilemma. Makingthem believe we prefer to or will be forced to carry outair strikes, if they don’t accept our position.

DELOITTE [concentrating on this]: …Even though Idon’t, in fact, prefer to carry them out, because, aswe’ve said, it’ll mean an end to relief convoys plus alikelihood of retaliation.

BRIGHT: Correct, sir, but we have to get rid of thatimpression in the minds of ISRA. How do we do that?Well, we could emphasize to them the certainty ofinternational repercussions if they retaliate. If theybecome convinced of that, their retaliation maybecome unlikely, eliminating that particular reason forour dislike of air strikes.

DELOITTE: Okay. But how do we eliminate our otherreason for disliking air strikes? The fact that it willdisrupt relief convoys?


BRIGHT: Er, maybe I’m confused here. I had in mind,again, emphasizing international repercussions as thereason why…

DELOITTE [exuberantly, completing his sentence forhim]: …as the reason why I’ll be forced to call airstrikes if they refuse a cease-fire, regardless of whetherI like air strikes or not! Jolly good! This works, Ray. I’dbeen planning, in the usual kind of way, to dampendown the tendency toward escalation. Now I’m seeingthat tendency as my strength. I mean to use it like afollowing wind. Unless we get a cease-fire and talkspronto, then the pro-intervention movement in the Westwill force my hand.

JONES [dubiously]: There certainly is such amovement, sir, pressing for intervention against ISRA.

DELOITTE [walking up and down with hands behindhis back]: Indeed there is. What this has made mesee [waving his hand at the flipchart] is that to avoidintervention, I shouldn’t seem to oppose it. Thosestrong pro-intervention forces are what give my threat,the threat of air strikes, credibility. And without suchcredibility, ISRA will not agree to a cease-fire, thingswill continue to get worse, until eventually we’re forcedto intervene on a larger scale and with less chanceof success.

BRIGHT [hurriedly]: Of course, what you say at thepress conference and through other channels may stillfail to make air strikes credible, sir. ISRA may still thinkthey can call our bluff, given our past history of threatsnot followed up. In which case, credibility may requirestarting to carry out air strikes, in other words, shiftingthe default future to one that contains air strikes.

185Chapter 6

DELOITTE [after walking up and down silently for awhile]: Right. So I may have to start air strikes, whilemaking it clear that if they agree to a cease-fire andtalks, air strikes will cease. We should, therefore, startwithdrawing our people; it’ll take a while anyhow tolocate targets properly. [He stops, looks at his watchand addresses Mark.] Dr. Bright, this has been useful.I’m off to the press conference now with a clear overallplan to pursue. I want you to continue this analysis.Brigadier Jones will make sure you get the help youneed, including access to information—the, er,assumptions you need to put into your model. Whatwill you look at next?

BRIGHT: I’d like to explicitly model both the grandstrategic as well as the military strategic games, sir. Ithink we need to be clear as to how they work, howthey impact on your own, operational level. [Thegeneral nods.] Then I’ll need to model some of theconfrontations going on at tactical level. We shouldplan to send a single, unified message at all levels.That way we can bring it about that ISRA and theMoryan government are being pushed in the samedirection by pressures from their own grass-roots aswe are pushing them at the higher level. That’ll beimportant for making air strikes credible. It’ll be stillmore important when it comes to making sure theykeep to our position, that they actually implement it,as distinct from merely agreeing to it to keep us offtheir backs.

DELOITTE [who has been listening intently]: Right.Carry on. [Turning to Jones] Ray…

JONES [opening the door for him]: Yes sir. Your carmust be waiting. I’ll see Dr. Bright gets what he needs.


[They leave together. Mark Bright collapses on a chairand grins at the audience.]

BRIGHT: Whew! Well, that seemed to go all right.What did you think?


The chapter dramatizes the situation of a commanderwho, knowing nothing of confrontation analysis,formulates and begins to implement a confrontationstrategy with the help of an analyst who happens tobe on his staff. This is not how it would ideally bedone, but serves to illustrate the concept of aconfrontation strategy before a discussion of itsprinciples in the next chapter.

General Deloitte has been tasked to get two warringparties to cease fire and start peace talks. He is not touse force. He has to address a press conference inan hour, and is anxious to formulate a plan. The analystcalled in to help him draws up a card-table on aflipchart, points out that he has a deterrence dilemmaand suggests directions in which he might look to tryto eliminate it. This prompts the general to think of away forward by adding a new card to his threat againstone of the parties. After discussion of how to removehis resultant inducement and threat dilemmas, heimplements this plan through a phone call, whichbrings the party concerned into apparent compliancewith his mission objectives (i.e., they now say theyaccept his position). Of course, this party may have acooperation dilemma, so that the general may still havea trust dilemma; this, however, can be dealt with later.

187Chapter 6

A deterrence dilemma remains in relation to the otherparty, which, they decide, can only be eliminated bythreatening to use force. A threat of retaliation againstUN personnel can, they decide, itself be deterred. Thegeneral’s own threat and inducement dilemmas canbe removed in the eyes of the other party. The generalnow has a clear plan upon which to proceed in dealingwith the press conference. He leaves after asking theanalyst to build models of the grand strategic, militarystrategic, and tactical levels.

189

Chapter 7

Formulating AConfrontation Strategy

The last chapter showed a confrontation strategyin use, although not as it should be used, not

implemented in an organized way by personnel trainedin how to do it. We saw it adopted in an ad hoc mannerby a commander relying on a single expert whohappened to be on his staff.

Nevertheless, we now have an example to fix on.Next is a discussion of the principles that shouldunderlie the formulation and implementation of aconfrontation strategy.

What Is a ConfrontationStrategy?

We have said that Operations Other Than War(OOTW), and peace operations in particular, can beviewed as sequences of linked confrontations;however, the confrontations involved are not all onone level.

To see what a confrontation strategy is we need tosee how it is implemented at various levels. We beginby discussing its implementation at grand strategic,military strategic, operational, and tactical levels ofcommand. There are many other levels within each


of these four; however, discussion of these broadlydefined levels will present the general idea.

Then follows a discussion of what confrontationstrategy conducted on these four levels consists of.To illustrate, we use two examples: the roadblock-removal in tables 2 and 3 found in chapter 2 andGeneral Deloitte’s strategy in chapter 6.

How Different Levels of Command Implement aConfrontation Strategy

First, the grand strategic or political level is that atwhich, in the play, Western public opinion was seento be putting pressure on Western governments tostop ISRA from massacring people. A confrontationstrategy at this level is primarily the responsibility ofpoliticians. They are dealing with the public, with eachother, and with foreign governments. As in allconfrontations, they are trying to orchestrate theemotions and arguments provoked by variousdilemmas to get everyone to play their tune.

Some of the cards politicians can play are military;therefore, the military’s role is to advise them on thepracticability and likely consequences of playing oneor another military card. The military also needs tounderstand the grand strategic confrontations thatpoliticians are involved in to understand the intent oftheir superior and their superior’s superior.

At the military strategic level, senior military staffmaintain and deploy a nation’s total defense forces inpursuit of national political objectives. Having receiveddirections from the grand strategic level, this level mustissue strategic directives to operational-level

191Chapter 7

commanders. In our terms, these directives empowera commander, in the case of an OOTW, to implementa confrontation strategy against certain other playersin pursuit of certain objectives. They also give himcertain types of cards he can play; therefore, it isnecessary for staff at the grand strategic level to havea broad understanding of the confrontation strategyrequired of their OOTW commanders to pursue. Inaddition, they are involved in various, hopefullycooperative, confrontations at their own level (e.g.,among themselves or with the military staff of othercountries with whom they are required to act in coalition).

The next level, the operational level, is the level atwhich General Deloitte is operating in the play. Havingreceived his strategic directive from the British nationaldefense staff, his task is to play his cards in such away as to achieve his mission objectives. To do so,he must confront various local parties and try to getthem to comply with what he wants. In the play he isplanning to do this and, at one point, actually leavesthe room to do it. In this, he is implementing aconfrontation strategy.

At the operational level General Deloitte, like the staffat grand strategic level, will conduct various otherconfrontations, hopefully cooperative, such ashorizontal ones with other units in a joint or combinedforce and internal ones with his own staff. In the playwe see him in a sequence of such confrontations withhis chief of staff and Mark Bright. (These personalconfrontations are not analyzed, just acted out.)

Finally, the general’s confrontation strategy needs tobe analyzed and broken down into strategies that canbe carried out by units reporting to him at the tactical


level. The significant thing about the tactical level is,in general, that this is where things, physical things,actually get done.

What things? In the case of battle-fighting, tactical-level units carry out and support destructive militaryoperations. In the case of an OOTW, destructivemilitary operations are not necessarily required.When they are, their primary function often is to senda message rather than to cause physical destruction.Non-destructive operations, such as assistingrefugees and helping with disaster relief, may berequired. Even then, the ultimate objective is to getothers to carry on such operations, perhaps afterreceiving initial or interim help, rather than for themilitary to do so indefinitely.

Generally it is true in an OOTW that the objective isnot so much to do physical things as to get others,those on the spot who will remain after we have gone,to do or refrain from doing things. (Operations suchas disaster relief may seem to be an exception. Herethere are certainly physical things to be accomplished.Yet even here the need to obtain cooperation withother agencies and local bodies is often paramount.)

In a sense the above is true of all military operations.Clausewitz himself says that after having “render [-edthe enemy] incapable of further resistance” we “compel[him] to fulfill our will” (1968, 1st edition 1832; p. 101).In Clausewitz’s traditional view, the military task, assuch, was to “render him incapable of furtherresistance.” After that was done, politicians took overto dictate to him our will. In an OOTW, by contrast,the military’s task is to induce or compel others to fulfill

193Chapter 7

our will, if possible without first rendering themincapable of further resistance.

This is the objective and essential principle of aconfrontation strategy. In relation to what is done atthe tactical level, the point is that while we areconducting a confrontation, we are not essentiallydoing anything except communicating. Here“essentially” means that we may be doing other things,but all such other things will have communication asat least one of their functions, and it is this functionthat our confrontation strategy is concerned with.

In devolving a confrontation strategy to the doing, ortactical level, we again must devolve, not a number ofphysical tasks as such, but a number of cards to beused in confronting others and getting them to complywith our objectives. The essential difference,nevertheless, between the operational and tacticallevels is that tactical level confrontations generally areconducted at the local, grassroots level.

In devolving an operational level strategy, the essentialprinciple is as follows: To plan tactical-levelconfrontations to change others parties’ foot-soldiersin such a way that the message they send up to theirtheater-level commanders reinforces the message weare sending to those commanders at the operationallevel. This is the essence of how to devolve aconfrontation strategy to the tactical level.


Linked Confrontations in a ConfrontationalCampaign

We see that an OOTW campaign is not a singleconfrontation, but a number of confrontations atdifferent levels linked to each other. Following is a listof some of those confrontations:

• An operational commander’s mission is definedfor him at the military strategic level afterpoliticians in a confrontation at the grand strategiclevel have decided to play certain military cards. Itis the political confrontation at the grand strategiclevel that defines his superior’s intent and hissuperior’s superior’s intent.

• The operational commander is then involved invarious (hopefully cooperative) horizontalconfrontations with other units in a combined orjoint force; he must resolve these to properlydeploy his military assets.

• His main confrontation is with other parties thatmust be brought into compliance with hismission objectives.

• Finally, the operational commander must devolvehis main confrontation strategy to the tacticalunits he commands. These then face numerouslocal confrontations at the tactical level.

We are now close to what a confrontation strategy is.But first we need to say how a theater commandermight, in a simple, step-by-step process, form andimplement a confrontation strategy for his mainconfrontation, that in which he must get other partiesto comply with his mission objectives. Links with other

195Chapter 7

confrontations will be discussed where relevant to thisdescription of a confrontation strategy in action. Afterdescribing a confrontation strategy in action, we cansay in general what a confrontation strategy is.Following is a step-by-step description.

A Confrontation Strategy in Action, Step by Step

In step 1 , a commander builds a model of hisconfrontation and identifies within it the dilemmasfaced by all parties. These include his own and others’weak and strong points (i.e., the points at which heand others are open to being changed).

In the play, Mark Bright helps General Deloitte andhis chief of staff build a model, which is essentiallybuilt and owned by the general, with all substantiveassumptions belonging to the general, while Markprovides expert advice and guidance on what aconfrontation model is. Mark might argue for orquestion a certain assumption on the basis of hisknowledge of the theoretical question being answeredby it. He does this quite often in the play. After he issure the general knows the meaning of a question, healways accepts the general’s answer. After arriving ata first model, Mark helps the general identify that hefaces a deterrence dilemma. He does not, as perhapshe should, go on to check that he has identified alldilemmas; instead, he immediately asks the generalto consider what to do about this one. He is takingshort cuts based on his knowledge of model building.After all, the general asked him to do what he can inan hour.

In the model shown in table 2, the commander mightidentify himself as facing a trust dilemma and the ethnic


militia as facing a cooperation dilemma. On the otherhand, the commander does not have a threat dilemma.Rather than allow the militia to keep roadblocks inplace, he would prefer to remove them forcibly. Hisfirst preference would be that the militia should removethe roadblocks themselves.

In step 2 , after the commander has identified thedilemmas in his model, he chooses a subset ofdilemmas he will attempt to eliminate andsimultaneously establishes the means by which hewill attempt such elimination. He must use hisjudgment and knowledge of the situation to estimatehow much friction would have to be overcome inbringing about various dilemma-eliminating changes.

General Deloitte decides to eliminate his deterrencedilemma by introducing a new card, “Blame thegovernment,” and making it part of his fallbackposition. Realizing that this will give him a threatdilemma, he next decides to make the Moryanpresident believe he will feel bound to play this cardif the president does not back a cease fire. He spendssome time discussing the feasibility of these changes(i.e., how much friction he must overcome inintroducing this card and making the president believehe will play it). Notice that the two dilemmas he iseliminating are connected: the second is broughtabout by the elimination of the first. Notice too thatthere is a strong connection with the grand strategicdrama. Bright does not analyze it, he does not havetime, but the connection is as follows. The Moryanpresident is trying to appeal over the general’s headto international opinion and Western governmentsto get them to send directives to the general that willchange his preferences (i.e., make him prefer the

197Chapter 7

Moryan position (M) in table 8 to the threatened futuret. The general will have to prefer it, the presidentreckons, if he is directed to implement it.) In this way,the president is trying to overcome a deterrencedilemma, consisting in the fact that the general isunder no pressure to accept the president’s position.The president is trying to overcome this dilemma bymaking moves in the grand strategic confrontationthat will, he hopes, affect the present drama throughlinkages between the two confrontations.

In table 2, the commander finds he has a problem.The militia has signed the agreement but is continuingto maintain roadblocks. When questioned, the militialeadership denies that there are roadblocks or blamesthem on local groups. Meanwhile the militia pointsout that the commander himself is violating theagreement by withholding aid and support. Shall thecommander give up on the agreement and move tothe threatened future t (i.e., forcibly remove allroadblocks)? (This was IFOR policy after the Daytonagreement.) Or shall he try to change the militia’spreference for unofficially keeping roadblocks inplace? He decides to try the latter. He devolves hisstrategy to the tactical level, directing localcommanders such as the one in table 3 to persuadelocal militia to remove roadblocks. He tells the militialeadership that this is what he is doing. He is, he says,helping them to make their policy effective with localgroups so that aid and support can start to flow.Simultaneously, he starts a campaign on localtelevision telling local militia groups what is the officialpolicy of their leadership. He reasons that if localcommanders, thus supported, succeed in their localconfrontations, then the leadership of the ethnic militia


will change its preferences and abandon its unofficialpolicy of maintaining roadblocks. (Note that this is alink that he sees between the two levels.) He will thenlet them have the promised aid and support.

Step 3 is to implement the planned dilemma-eliminating measures. Note that by definition no onecan know, when entering a moment of truth, how theywill be changed by it. This applies less to thecommander, who is deliberately instigating a momentof truth to cause specific dilemma-eliminating changes,than to those he is interacting with, who are, incomparison with him, unprepared for the pressuresthey are about to face; nevertheless it applies in somedegree to both sides.

In the play, General Deloitte gets on the phone toPresident Saldin and lets him know he had better cometo the cease-fire talks and back a cease fire, or he willbe publicly blamed for being responsible for themassacre. This, he must see, will undermine his effortsto enlist Western support. President Saldin now facesa moment of truth.

The UN commander in table 2 sends directives to hislocal commanders, describing his confrontationstrategy and how they can contribute to it, giving themappropriate cards to play but leaving the details oftheir individual strategies to their initiative ascommanders. If they have questions or suggestions,such as additional cards they might want to have intheir hand, they discuss them with him. At the sametime he informs the militia leadership of what he isdoing, presenting it as a way of helping them to exertcontrol over local groups and holding out to them theprospect of aid and support. Finally, he launches a

199Chapter 7

television campaign while his local commandersconfront local groups manning roadblocks. In this way,the militia leadership is brought to a moment of truth.

Step 4 , following the moment of truth, at whichchanges were meant to take place, is a return to step1 to build a model representing the new, changedsituation. If there is now a satisfactory solution, withall parties taking the same position and trusting eachother to carry it out, then the commander has achievedhis mission objectives (perhaps in a modified form). Ifnot, dilemmas remain, which the commander identifiesat this point. He then moves on to step 2, and so on.

In the play we saw how the commander, havingbrought the Moryan government into compliance withhis objectives, models the new situation and worksout how to do the same to ISRA.

Suppose, in table 2, that the UN commander’sstrategy succeeds in getting the militia to complywith the removal of roadblocks. There are now nomore dilemmas. The problem is resolved within thismodel. Other, more detailed models, or modelsfocusing on other issues, would no doubt revealdifficulties. In addition, implementation of theagreement may meet unforeseen problems. Forexample, the commander may find that the aid andsupport he promised to the militia, and which he hasheld up by refusing to certify that the first stage ofpacification has been reached, is now held up forother reasons. To get around this problem, he mustresolve a horizontal confrontation with the highrepresentative responsible for aid coordination.


Note: A confrontation between two parties is verticalif they are related as superior and subordinate in thesame organization. This means that one partygenerally has a card consisting of formally issuingorders or directives to the other, although often thisdoes not settle the matter; if the other does not agree,it may be able to ignore or re-interpret directives givento it, or appeal against them. The distinction betweenvertical and horizontal relations becomes fuzzy when,as is often the case in peace operations, a commanderis responsible both to his national headquarters andto an international organization. For the sake of ourexample, we are assuming that the high representativehas no authority over the UN commander (as in thecase of the Dayton agreement, where the UN highrepresentative was responsible only for coordinatingcivilian aspects of the agreement). Therefore,negotiations between the UN commander and the highrepresentative in our example are horizontal. Neitherhas authority over the other.

The Confrontation Strategy Itself

The above step-by-step description of theimplementation of a confrontation strategy is not yet adescription of the strategy itself, but it enables us togive the following definition. A confrontation strategyconsists of a plan for cycling through the above foursteps until a resolution of the confrontation is reached;therefore, it consists of a sequence of dilemma-revealing models of characters’ positions such that:

• A plan for dilemma-elimination is associated witheach model except the last.

201Chapter 7

• The next model in the sequence is created byimplementing the plan on each model exceptthe last.

• The last model has no dilemmas.

This is our suggested characterization of aconfrontation strategy. To understand it correctly,several points must be kept in mind.

First, each plan for dilemma-elimination associatedwith a model in the sequence is a plan for bringingabout a moment of truth, at which changes may takeplace. We have said that on entering a moment oftruth no one can tell exactly how it may change them.This, as well as the general fact that the future tendsto hold surprises, makes it an essential characteristicof a confrontation strategy that it should be looked atanew, and reformulated if necessary, at each stage,after each moment of truth.

Even though it is constantly reformulated, it isnevertheless desirable for a whole strategy to exist,so that assets can be mobilized and relevantinformation collected to fulfill it.

Second, it is important to realize that a plan fordilemma-elimination associated with a particular modelin the sequence may be linked to other, relatedconfrontations. We have looked at some examples ofsuch linkages between the particular confrontation acommander is involved in and other confrontations.

For example, the confrontation between GeneralDeloitte and President Saldin was linked to the grandstrategic confrontation between the West, the Moryangovernment, ISRA, and other countries (e.g., the Arab


nations, France, the bordering state of Pelugya). Partof this linkage consisted in the fact that since GeneralDeloitte’s job was to fulfill his mission, his preferenceswere largely linked to the policy decisions of the West.If Western politicians wanted him to side militarily withthe Moryan government, that likely would become hispreference. The result was that to eliminate a dilemmain his confrontation with the general, President Saldincould try to alter the latter’s preferences by means ofmoves in the grand strategic confrontation he wasconducting with the West.

The confrontation in table 2 between the UNcommander and the ethnic militia leadership wassimilarly linked to the confrontations between localcommanders and local militia groups; that is,decisions by a superior could decide the preferencesof a subordinate. However, interestingly, the oppositewas also true in the case of the ethnic militia. Thecommander’s plan for eliminating a trust dilemmadepended on lower-level militia attitudes influencingthe preferences of their leadership.

After apparently resolving his confrontation with theethnic militia, we saw the UN commander findinghimself unable to implement the agreement reachedbecause of problems with a horizontally relatedpartner, the high representative responsible forcoordinating aid. Technically this meant thecommander had to resolve a horizontal confrontationto eliminate a cooperation dilemma.

203Chapter 7

Applying the Principles of Warto Confrontations

It is clear that a confrontation strategy is different froma strategy for warfighting, yet some of the sameprinciples apply; therefore, it may be useful to askwhich generally known principles of war continue toapply to confrontations. This may help to bring outwhat is different about OOTW. We are not attemptingany more than this; we do not aim to establish a list ofprinciples of OOTW.

Seven Principles of War

Alberts and Hayes (1995, pp. 28-37) ask how far theprinciples of war apply to peace operations. Theysurvey seven principles of war and conclude thatsimplicity and unity of purpose (rather than ofcommand) remain important. Unity of command, theypoint out, is often unachievable when there is a needto coordinate strategy between different nationalcontingents; it has to be replaced by the attempt toachieve unity of purpose. Achieving this also helps tofocus on the objective. Consequently they suggest thatunity of purpose can replace this principle also.

Thus, Alberts and Hayes endorse simplicity as aprinciple for peace operations and subsume the twoprinciples, unity of command and focus on theobjective, under the one principle, unity of purpose.They suggest adding consensus planning, adaptivecontrol, and transparency of operations as principlesto be followed in peace operations. Four otherprinciples that apply in warfighting, taking the offensive,concentration of superior force, taking the enemy by


surprise, and maintaining security, are, they consider,less applicable and may even do harm.

This is certainly right if we look at the physicaloperations involved, as we do when consideringwarfighting. We have said, for example, that aphysical operation such as a bombardment or atroop movement whose primary purpose is to senda message cannot sensibly be kept secret. Thusthe principles of surprise and security, applied inways that would make sense if the primary purposewere destruction of the enemy, can do harm ifapplied in OOTW.

The Contrast Between Warfightingand Confrontation

The contrast involved here is important; it perhapsneeds re-emphasizing. If we are bombarding theenemy to force negotiation, our bombardment isactually a message to the effect: “Look what willhappen to you if you don’t come to the bargainingtable.” Its physical effects in actually destroying theenemy’s assets (as distinct from communicating ourdetermination to do so) are a disadvantage to us aswell as to the enemy, because the more assets wehave already destroyed, the less remain to bedestroyed, and the less significant is our threat tocontinue destroying them.

Counter-insurgency forces, for example, by destroyingthe homes and families of guerrilla-fighters, oftencreate large numbers of homeless, bereaved fighterswith nothing left to lose and much to revenge.

205Chapter 7

Consequently, to send the message, “look what willhappen to you,” the ideal is to combine a maximum ofshock and awe with a minimum of actual destruction.Keeping secret the destruction you have wrought iscounter-productive. By contrast, if the intention is notto send a message but to actually destroy the enemy’sfighting potential, then the more secure he feels andthe less he knows about what is happening to him thebetter, because his ignorance and contentment reducethe effectiveness of his countermeasures andmaximize the damage done to him.

On the other hand it may be a sensible strategy, aswe have seen, to interrupt negotiations to weakenanother’s bargaining position by physically takingcertain cards out of the other party’s hand. Tocomplicate matters, such physical operations maybe combined with message-sending. For example,operations such as interdicting terrorists’communications or destroying their arms dumps mayreduce their ability to harm us, and thus either takecards from their hand or reduce the effectiveness ofcertain cards they might play. Such operationsobviously require secrecy and surprise at the tacticallevel, but if carried out while we are conducting aconfrontation with terrorist leaders, they may servetwo functions: to send a message (“This is what youcan expect if you don’t concede”) as well as to diminishthe effectiveness of terrorist action. The message mayeven be, “You can expect your effectiveness to keepdiminishing as long as you hold out.”

An operation that has dual functions, message-sending combined with physical effects, may need abalance of secrecy and openness. It remains true thata message as such needs to be clearly received and


correctly interpreted if it is to do its job. To destroyterrorists’ effectiveness will not work well as a messageif the terrorists think they have suffered less loss ofeffectiveness than they actually have, or mistakenlybelieve that we mean our strikes against them to beour last. Yet these mistaken beliefs actually wouldbenefit us if our actions were not meant to send amessage, but were carried out purely to diminish theterrorists’ physical effectiveness.

The contrast between something done for its actualeffect and the same or a similar thing done to send amessage is striking. We have said that a confrontationstrategy (the multilevel, coordinated object describedin the last section) is not essentially a matter ofconducting physical operations, but of sendingmessages. It is not surprising that we get somewhatdifferent results when we look at how the Alberts-Hayes principles of war apply to message-sendingrather than when they are applied to physical activitiesas such.

Applying the Principles to a Confrontation Strategy

Consider now the seven principles in turn.

Simplicity is of utmost importance, essentially for thereasons given by Alberts and Hayes, but we can addsome more. We have seen that an operationalcommander’s confrontation strategy needs to beunderstood and implemented in a coordinated wayby tactical echelons below him and other task-forceunits horizontally related to him. He also needs to beable to explain it clearly to military strategic and grandstrategic levels above him to enlist their support inpressuring parties from above or expanding the cards

207Chapter 7

available to him, when necessary. This is theconfrontation-strategy need for simplicity. Simplicityis also necessary because the common referenceframe through which the commander communicateshis position to other parties must be simple; therefore,his confrontation strategy can be simply expressed interms of it. The commander in table 2, for example,might explain his strategy by presenting and explainingtables 2 and 3.

Unity of purpose, including, as Alberts and Hayessuggest, focus on the objective, also magnifies theeffectiveness of a confrontation strategy. Suppose acommander is pressuring another party to take acertain action, say to cease fire, withdraw armaments,or permit the passage of refugees, and to this end isusing certain implied threats and promises. We haveseen that he is more likely to succeed if the otherparty’s grassroots organization, superiors, and alliesare receiving versions of the same message, and soare inclined to tell it to make the concession. Such aunanimous chorus is achieved only by getting unity ofpurpose behind a message dictated by an agreedconfrontation strategy. We have also seen thatobtaining agreement between different coalitionpartners on a common confrontation strategy is itselfa task for a (separate, hopefully cooperative)confrontation strategy.

Taking the offensive is shown by Alberts and Hayesto be an inappropriate principle for peace operations,if we think in terms of a physical offensive; however,as applied to a confrontation strategy, we mightreinterpret it to mean taking the initiative in forcing amoment of truth so that a certain set of dilemmaeliminations, planned by us in advance, can occur.


This is to take and hold the initiative in terms of theargument. It is analogous to the battle-fighting conceptof forcing the enemy to respond to us, rather than viceversa. To avoid confusion, it might be better to avoidmetaphor and call it taking the initiative rather thantaking the offensive.

Is concentration of superior force another principle wecan usefully reinterpret in application to a confrontationstrategy? “Superior force” might be taken to mean anarray of superior arguments for use at a preplannedmoment of truth, provided we understand that thearguments we have in mind need not be veryintellectual, but may include points such as “We’ll bombyou to bits if you don’t.” Here “argument” may be thewrong word. “Force” may also be misleading.“Sanctions” is perhaps the right term, because it mayrefer to positive enticements such as offers of aid aswell as negative ones. This suggests that“concentration of superior sanctions” might be the rightname for this principle.

Concentration of sanctions may be unnecessarybecause of what appears to be an essential differencebetween confrontations and battle-fighting. In a battle,assets used in one place cannot at the same time beused in another; therefore, they must be concentratedbefore an assault. By contrast, negative assets usedin a confrontation, that is, the cards you can play toout-escalate another, are not necessarily used. Thehope is that you will merely have to threaten their use.The same negative assets can therefore be used indifferent places at once. Following are two examples:

• All the commanders confronting local militia inour roadblocks example can simultaneously

209Chapter 7

threaten to call down air strikes. They will notneed that many planes.

• The British Empire, while it lasted, was helddown with a small number of troops. Eachpotential rebel feared that those troops would beused against him.

Positive assets (sanctions that are promises ratherthan threats) are different. They will be used up in asuccessful operation because using them up is, bydefinition, the objective. Their quantity must besufficient to cover the different points where theyare used.

In sum, it seems that assemblage of superior sanctionsmay be the principle that should replace concentrationof superior force.

Surprise is another principle inapplicable to OOTWwhen we consider only physical operations. Given theneed to build trust and verify agreed physicalconditions (such as separation of forces ordisarmament), peace forces should not createuncertainty by behaving in unexpected ways (Albertsand Hayes, 1995, p. 30). Indeed, this follows from ourassumption that actions in a confrontation are seenprimarily as communications. Our argument is asfollows. One of the characteristics of a surprise actionis that it is uninterpreted. You wonder what it means.That defeats the purpose of an action that is supposedto send a message. The whole point of that is for youto know what it means. Prior notice tells other partieswhat an action means.

The principle of surprise, however, may be applicableto the content (i.e., meaning) of the messages you


send in a confrontation, as distinct from the form (theaction itself). Producing a surprising new argument,offer, or threat, as General Deloitte in the play doesin threatening to blame the president for themassacre, makes it hard for the other to produce aconvincing defense, because they have not beengiven time. Lawyers in television courtroom dramaswin this way. In the other direction, a surprisingconcession or compliment, even if insignificant, maycause favorable changes in others’ perceptions ofyour motives just because it is surprising. A surprisedraws attention to itself. The principle of surprise,understood in this way, can be used alongside takingthe initiative and assembling superior sanctions. Theyare mutually reinforcing methods of obtaining otherparties’ willing compliance.

Finally, Alberts and Hayes cast doubt on the principleof security, pointing out how over-emphasis onphysical security, e.g., ensuring minimal risk to troops,may reduce the effectiveness of peace operations.This follows from asking, “What message does thisaction send?” Strict security precautions say, “I don’ttrust you.” This is not compatible with “Trust me.” Onthe other hand, it is quite compatible with the message,“I’m ready to smash you,” which may, in othercircumstances, be the intended message. In relationto physical security, all that confrontation analysis cansay in general is what it says in relation to otherphysical activities: “Pay attention to the message youwant to convey.”

On the other hand, surprise requires security, and wehave said that surprise may be advantageous in itselfand help us to take the initiative and assemble superiorsanctions; however, this applies to the content of

211Chapter 7

messages, not to physical actions. To ensure surprisein our confrontation strategy, and also to preservesimplicity and avoid misunderstandings, it is advisableto preserve security in the process of confrontation-strategy formation; however, in implementing aconfrontation strategy, disclosure of our strategy isusually the best tactic because it reinforces themessage being sent and helps to avoidmisinterpretation. Similar reasons are given by Albertsand Hayes for their principle of transparency ofoperations. They are the reasons why, in our example,the UN commander decides to inform the ethnic militialeadership of his devolved strategy of confronting localmilitia roadblocks.

This seems to leave us, in this area, with a ratherawkward hybrid slogan: Security of planning plustransparency of operations.

Summing up, the principles of war apply rather badlyto OOTW considered as physical activities; however,many of them seem to apply better to the formationand implementation of a confrontation strategy, whichis primarily a matter of sending messages.


A confrontation strategy for a particular confrontationconsists of a plan for sequentially eliminatingdilemmas, each in a specific way, so that the end resultis full compliance on the part of all parties with ourmission objectives. A strategy for a peace campaignas a whole is a plan for winning in this manner a linkedset of confrontations.


A strategy needs to be adaptive to changingcircumstances and new information, such as the factthat a particular confrontation is not resolved in themanner expected. At the same time it needs to besufficiently robust to give continuing guidancethroughout the inevitable fluctuations of expectationsthat occur in the course of a campaign.

For these reasons, a strategy must have a clearlystated objective; be simple; and be carried out withunity of purpose. As has been pointed out, otherprinciples of war, such as surprise, concentration offorce, taking the offensive, and security do not applyto peace operations in the same way as they do towarfighting; however, applying superior force andtaking the initiative are applicable to confrontationstrategy at the level of persuasion, as distinct fromphysical operations.

A strategy needs to be understood and implementedat various vertically linked levels of command and byvarious horizontally linked components and coalitionforces. We must understand the manner in whichdifferent confrontations are linked to formulate aconfrontation strategy. A confrontation is linked toothers that its conditions and objectives affect or areaffected by. It is linked to higher-level confrontationsthat set the scene for it and to lower-levelconfrontations that its Implementation phase sets thescene for; also, a confrontation is linked to the moredetailed confrontations that take place in its Resolutionand Conflict phases.

Confrontations are linked together strategically if thepositions and the sequence of dilemma-eliminationsadopted in one confrontation are linked to those

213Chapter 7

adopted in others. In the removing-roadblocksexample, a confrontation strategy worked out by thetheater commander could be devolved to the level ofthe local commander by linking their confrontations.Similarly, the commander’s confrontation with non-compliant parties at theater level should be linked tothose of horizontally related characters such as othercomponent and coalition forces. Achievingconfrontation strategies that are coordinated betweenhorizontally and vertically linked partners may requireconsensus planning and negotiation of strategies.Above all, it requires that the commander’s strategybe simple and his objectives clear and agreed.

215

Chapter 8

Analysis of theBombardment of

Sarajevo, February 1994

On February 5, 1994, a bomb exploded inSarajevo’s market square. It killed 69 people and

left 200 wounded. Blame naturally fell on the BosnianSerbs, who at the time were bombarding the city withmortars placed on the surrounding hills. They at oncedenied responsibility and blamed the Muslimdefenders of the city. Nevertheless, there was aninternational outcry against the Serbs. As a result, theBritish commander of the United Nations ProtectionForce (UNPROFOR) in Bosnia, LTG Sir Michael Rose,found himself tasked with getting the Serbs to ceasefire and withdraw their weapons in order to fend offNATO air strikes on their positions.

In this chapter we examine this crisis in some detail,going through the most complex confrontationalanalysis of a real-world Operation Other Than War(OOTW) that has been done so far. As we do so, wewill discuss some of the problems of real-worldconfrontation analysis.


Sources for the Analysis

Research for the analysis of Sarajevo, February 1994,was carried out by the author in early 1997 under acontract issued by the UK Defense Evaluation andResearch Agency (DERA) for a report on the use ofdrama theory in operational analysis (DERA, 1997). Itis reported here by permission of DERA.

The research included interviews with a number ofBritish officers who had served in Bosnia; however,they were not interviewed for their involvement in theparticular confrontation analyzed, the crisis over theBosnian Serb bombardment of Sarajevo in February1994. All factual assumptions made in the analysisare drawn from published sources. If they areinaccurate, the fault lies either with those sources orwith our misinterpretation of them, and we can onlyhope that those in the know, in particular General Roseand his staff, will make allowances accordingly.

Even if we had been able to interview actualparticipants, they would not have been the best sourcefor assumptions to go into our analysis. Memories arebiased, rationalized reconstructions of what was feltand thought at the time. We tend to think that whatdid not happen, could not have happened. We do notthink so at the time. The best source of assumptionsfor confrontation analysis is the judgments of thoseinvolved while the confrontation is going on. This ishow assumptions are made in the fictitious casedescribed in the Frontline play in chapter 6.

Incidentally, the reader will notice that the play,although fictitious, has a plot that is based on the crisisin Sarajevo in February 1994. The play was written by

217Chapter 8

taking that real-life scenario, simplifying it, equippingit with imaginary characters, transferring it to a non-existent North African location, and imagining that itwas dealt with by a commander who decides to useconfrontation analysis.

Conclusion of the DERA Report

The conclusion to keep in mind and examine criticallyas you read this chapter and later ones is as follows:If the UNPROFOR commander in Sarajevo had usedconfrontation analysis in his efforts to get the twowarring parties to cease fire, he might have achievedhis mission objectives more quickly and satisfactorily.

In the DERA report, the conclusion is spelled out asfollows:

• The first stage, bringing the Bosnian governmentinto apparent compliance, might have beenachieved in a way that was planned andintended beforehand, without depending onemotional, spontaneous, unplanned reactions ofthe commander. Emotions might have beenused, but in a rationally planned way.

• The second stage, bringing the Bosnian Serbsinto apparent compliance, might have beenachieved sooner by the UNPROFORcommander on the operational level instead of atthe last moment, after many concessions hadbeen made, as a result of developments on thegrand strategic level.

• Even if the result had differed little from theactual one, it could have been achieved by thecommander in a more purposeful way because


from the beginning he would have been able toseize and maintain the initiative.

Putting Pressure on theBosnian Government

When the bomb exploded in a square crowded withshoppers, the international outcry brought to a heada growing demand for intervention to stop the Serbbombardment of Sarajevo. The UNPROFORcommander actually was engaged at the time innegotiations to this end, tackling the problem thatwhile the Serbs were predominant in heavyweapons, they were weaker in infantry than the(mainly Muslim) Bosnian government forcesdefending the city, who were ready to take theoffensive if Serb actions were halted.

At this time, the UNPROFOR commander conceivedhis mission as primarily humanitarian (delivery ofaid, prisoner exchanges). After the bomb fell, he wentto Belgrade, met with the overall UNPROFORcommand and the UN special attaché, and cameback with a mission to get the Serbs and the Bosniangovernment to agree to a cease-fire and withdrawalof heavy weapons.

Simultaneously, at the grand strategic level, the U.S.and French governments began pressing for a NATOultimatum to the Bosnian Serbs: “Cease fire andwithdraw heavy weapons or we launch strategic airstrikes against you.” Although it supported his effortsto get Serb agreement, this was of serious concern tothe UNPROFOR commander, who feared that NATOair strikes would make it impossible to continue his

219Chapter 8

humanitarian mission and would endanger the livesof UN and non-governmental organization personnel.His reasoning followed this path: “If my mission ishumanitarian, let me do it. If you want to change mymission to fighting the Serbs, I must withdraw UNpersonnel and abandon the humanitarian mission.”

The commander proceeded urgently with his newmission in the hope of forestalling air strikes. He hadalready laid the groundwork. In previous discussionshe had obtained the agreement of commanders ofSerbian and government forces to four points: acease-fire; withdrawal of Serbian heavy weapons toat least 20 kilometers from Sarajevo if not placed underUN control; the interposition of UN troops; and dailymeetings to oversee implementation of theseagreements. When the crisis broke, however, BosnianPresident Alija Izetbegovic decided he would nowmake no agreement with the Serbs. Rather, his policywould be to take advantage of the outcry to persuadethe international community to take action on theBosnian government’s behalf, specifically by lifting thearms embargo against it and taking military action suchas air strikes against the Serbs.

The result was that the commander discovered thatthe president, who was at that moment beinginterviewed by CNN and declaring the need for airstrikes against the Serbs, had told his commandersnot to attend the UNPROFOR-arranged meeting.

Angrily the commander let it be known that if theBosnian government did not attend the meeting, hewould publicly blame Muslims for the market-square


incident and for blocking a cease fire. Table 11 is ourattempt at a card-table model of the commander’smoment of truth at this juncture.

In table 11 the symbol ~ indicates that a charactertakes no position as to whether a particular card shouldbe played or not; that is, its position is that the cardmay or may not be played. The Bosnian Serbs, forexample, took the position (column BS ) thatUNPROFOR should not blame them for the market-square bomb nor call air strikes against them, whilethey themselves should not cease fire or withdrawweapons; in return they would not retaliate againstUN personnel. They took no position as to whetherUNPROFOR should blame the Muslims or the Bosniangovernment should cease fire; obviously they wouldhave preferred both these cards to be played, but couldhardly demand or expect them, given their ownproposed actions. The numbers indicate preferencerankings as before, with the most preferred future giventhe number 1, the next most preferred the number 2,and so on. In assigning preference rankings to columnscontaining entries with ~, the most probable decisionis assumed. For example, it is assumed in this tablethat if the Bosnian Serb position, BS, were accepted,then the players’ expectation in regard to the cardsleft unspecified (“Blame Muslims” and the Bosniangovernment’s “Cease fire” card) would be that neithercard would be played: UNPROFOR would not blamethe Muslims and the Bosnian government would notcease fire. This is why the column BS is assigned thesame preference ranking for each player as columnd, even though it is compatible with either t or d.

221Chapter 8

Table 11. Bosnian government refuses to discuss a cease-fire.


The following assumptions are made in the model:

• The UNPROFOR position (column U) was thatboth sides should cease fire while U(UNPROFOR) remained impartial, not blamingor attacking anyone. Note that it was importantfor U that BG (the Bosnian government) as wellas BS (the Bosnian Serbs) should cease fire. BGhad more infantry, although fewer heavyweapons, than BS. Also, it often initiatedincidents, even though it might lose militarily, toget world opinion on its side.

• The commander could assume that the BSposition was as shown (column BS), eventhough BS had agreed to U’s four points. This isbecause BS had agreed unwillingly, arguing thatas BG forces outnumbered theirs, they neededto use artillery to respond to BG-initiatedincidents. When BG rejected the four points, itcould be assumed that BS would reject themalso in favor of its previous position.

• The BG position was that they, the victims,should not be asked to cease fire. U should taketheir side, blaming BS and calling air strikesagainst them. BS could do as they wished, (i.e.,BG took no position as to which BS cards shouldbe played).

• The assumption behind the fourth column isthat, by angrily interrupting the president’sCNN interview and directly threatening him, thecommander made the threatened future one inwhich he would publicly blame the Muslimsand the bombardment would continue. This, of

223Chapter 8

course, was very much what the Serbswanted. The threatened future put no pressureon BS; however, it did pressure BG to acceptU’s position.

• The fifth column states that, if present policiesand actions continued unchanged, there wouldbe no cease fire and U would take no action.

Other Parties in the Background

There were at least two other important parties in thebackground of the drama we are analyzing: theSerbian government of Slobodan Milosevic and theCroatian government of Franjo Tudjman. At this pointboth were in transition from one policy to another. TheSerbian government was attempting to getinternational sanctions against it lifted by distancingitself from the Bosnian Serbs, who had advanced sofar with its backing. The Croats were engaged in U.S.-sponsored negotiations to restore their alliance withthe Bosnian government against the Bosnian Serbs,having in the most recent phase of the conflict beentaking land and expelling Muslims in unspoken alliancewith the Serbs. Alliances were reforming against theBosnian Serbs. This, we assume, increased theirfeelings of persecution and discrimination while makingthem more anxious to take Sarajevo while they could.At the same time, it meant that the Bosnian andCroatian governments played no active role in theconfrontation we are analyzing.

Note that we are following and will continue to followan important rule of realistic modeling, as stated inthe play in chapter 6 by Mark Bright while advising thegeneral. Mr. Bright points out that factors outside the


model that influence things in it must be taken intoaccount in making assumptions. They should not beignored just because they are not explicitly named inthe model.

Analysis of Table 11

What dilemmas face the characters in this model?Take the UN commander first.

U has no deterrence dilemma against BG. BG preferscolumn U to t (t would undermine internationalsympathy for BG); therefore U’s threat, if credible,puts pressure on BG to accept column U. The resultis that the UN commander’s threat concernedBosnian President Izetbegovic; he broke off hisinterview with CNN.

U does have a deterrence dilemma against BS, whogreatly prefers t (with the West blaming the Muslimswhile their bombardment continues) to U. The resultis that the Serbs were under no pressure, and wouldhave been delighted if Izetbegovic had refused toattend the meeting.

U has a threat dilemma; it prefers not to blame theMuslims for the market-square bomb. That wouldmean giving up impartiality and possibly going againstthe grand strategic policy of NATO and the UN. Theresult is that the UN commander’s anger, withoutwhich the Bosnian president might have dismissedhis threat. This anger showed either or both of thefollowing:

• That U’s preferences were changing in favor ofcarrying out the threat (preference change)

225Chapter 8

• That U would carry it out regardless of whetheror not it preferred to do so (irrationality).

U has an inducement dilemma in that t is arguablyworse for U than BS, a fact that puts pressure on U toaccept BS. The reaction was anger and defiance onthe part of the UN commander as the BG position,with accompanying need to threaten t, puts pressureon him to abandon his objective and accept BS.

U has no positioning or cooperation dilemmas. Itprefers its own position to the positions of the othercharacters and to any future it could reach from itsown position.

U has a dilemma of trust in that both BS and BGwould prefer unilaterally to defect from U’s positionby breaking the cease-fire. It is even preferable forboth if both defect: preferable for BS because it wantsto keep up the pressure on Sarajevo, preferable forBG because it wants to keep up pressure on theinternational community to intervene on its behalf.There is no immediate reaction to this dilemma. Butnote what happens next.

BG gives in to U’s pressure and accepts U’s position.BG and BS army commanders then attend a meetingwith U. In this meeting all agree with U’s position;however, even at the time U does not give muchcredence to this agreement. U does not decide to giveup. Instead, U decides to monitor whether theagreement is observed while continuing, with positiveemotion, to stress its desirability to both sides in anattempt to persuade them to keep it. All the signs arethat U feels it has a trust dilemma.


Table 12. Dilemmas facing characters in table 11.

Other Parties’ Dilemmas

We have looked at U’s dilemmas. What about othercharacters’ dilemmas?

All the dilemmas facing the parties are set out in table12. It is striking that the Serbs face no dilemmas. Thisis partly because their own position is what they wantto see happen and is such that they do not need totrust anyone. They have no cooperation, trust, orpositioning dilemmas. Also, the tug-of-war between Uand BG has made the threatened future one they like,although the other two do not. They therefore haveno inducement, threat, or deterrence dilemmas. Theyare sitting pretty.

The Bosnian government has three dilemmas:

• Deterrence: The threatened future, t, pressuresno one to accept BG’s position. Both U and BSprefer t to BG. The reaction was abandonmentof position BG.

• Inducement: BG prefers U to t. The reaction wasacceptance of position U.

227Chapter 8

• Trust: Even if U were to agree to BG’s position,BG might doubt whether they would carry it out,given the UN commander’s strongly expressedpreference for remaining impartial and not takingsides. The reaction was possibly greaterwillingness to abandon position BG.

How UNPROFOR Got Angry—The Model ThatPreceded Table 11

It seems probable that the card, “Blame Muslims,” intable 11 was thought of by the UNPROFORcommander as something he might actually do onlywhen he realized the position the Bosnian governmentwas taking. To see this, we draw up the card-table intable 13 showing the moment of truth before he thoughtup this card.

Both Bosnian sides here prefer t to U’s position (BGdoes so because t places pressure on the internationalcommunity to intervene on BG’s side); therefore, Unow faces a strong deterrence dilemma (i.e., it ispressuring no one to accept its position). U’s reactionis anger and demonization of BG, motivating U to thinkup the card, “Blame Muslims,” and make it credible aspart of the threatened future t shown in table 11.

Note that if we look at the analysis formally, it mightseem that U might equally well have demonized theSerbs. However, the Serbs had at least agreed to talk.The Muslim-dominated government was refusing to doso. Also, the Serbs already had been so demonized bythe media and Western public opinion that U would needlittle rationalization, and hence little demonization, tojustify turning against them.


Table 13. What made UNPROFORangry? The situation preceding table 11.

This, then, is an example of emotion and rationalizationleading to a change in the set of cards assigned tocharacters in a frame.

In table 14 we set out all the dilemmas in table 13.

• U has a deterrence dilemma (as said) and a trustdilemma (for the same reasons as in table 11).

• BS has no dilemmas.

229Chapter 8

Table 14. The dilemmas facing characters in table 13.

• BG has a deterrence dilemma (no one ispressured to accept its position) and a trustdilemma (U would not want to implement BG’sposition, if it were agreed).

BG’s dilemmas obviously gave it serious problems in thistable; however, it had a strategy to overcome them,consisting of a strong appeal to the internationalcommunity, which had it in its power to change the missionassigned to U and thereby change U’s preferences. If itcould get U on its side, this would be enough to achieveits position, as its position did not specify any particularactions on the part of the Serbs (i.e., any Serb actionswould be compatible with its position).


After the BosnianGovernment Was Brought intoApparent Compliance

A commander’s mission objective, expressed interms of confrontation analysis, is to bring othersinto compliance with his position. Now at this pointin our analysis, it might seem that BG, at least, hasbeen brought into compliance.

A commander must, however, distinguish betweenapparent compliance (overt acceptance of his position)and actual compliance (in addition, the intention toimplement that position). He must be aware that acharacter facing a cooperation dilemma may solve itby deceit (i.e., accepting a position while intending toact in non-compliance with it).

In this case the estimate is that BG has been broughtinto apparent compliance only, because if a cease-fire were implemented, they would prefer to break it.Although the UNPROFOR commander has achievedthe bringing of BG into apparent compliance, BG nowhas a cooperation problem and the commander a trustproblem in that BG will want to defect from position U.

Problems in Determining a Character’s Position

Has BS also been brought into apparent compliance?A superficial reading might suggest so. We have saidthat all parties, including BS, have at this point formallyaccepted U.

This would be a mistake, illustrating the fact thatdetermining a party’s position by the rule, “It is whatit says it is,” may be harder in practice than it seems

231Chapter 8

in theory. A real-life character has manysubcharacters saying different, contradictory things.Even the same subcharacter may contradict itself.What do we do then?

Judgment is necessary, but a general rule ininterpreting others’ messages is, “Pay attention to theiremotions.” One thing we know is that a characterwhose emotional signals and rationalizations arepredominantly negative toward others, angry, defiant,and so forth, must be taking a position opposed totheirs. We know this because such emotions are notappropriate to solving cooperation or trust dilemmas,which are the only ones left between parties that sharea common position.

If a negatively emoting party formally states that itshares your position, it is contradicting itself by itsnegative rationalizations. If these are as open andpublic as its formal statement of agreement, if theycome across more strongly than the formal statement,and if they state a clear alternative position, they shouldbe read as stating the party’s actual declared position.

In this light, consider the present case. At this point,the signals being sent by BS and BG are quite different.BG is showing signs (i.e., fear of the threatened futureand a conciliatory attitude) of the impact of thepressures needed to make them comply. BS isshowing opposite signs. The NATO ultimatum to theSerbs, “Cease fire and withdraw or we bomb you,”was issued on the day that agreement apparently wasreached. Instead of reacting with fear, depression, orconciliation, BS reacted with defiance, threatening toshoot down 40 NATO planes in the first wave of attack.More seriously, they threatened retaliation against UN


personnel. This, together with their continuing strongarguments for column BS, meant that this was theirposition. They were not even in apparent compliancewith U.

One interpretation, possibly that of the commander,is that the NATO ultimatum itself, by infuriating BS,had made them non-compliant. It is equally possiblethat BG’s compliance made the BS non-complianceobvious. While BG was defying U, BS may havethought it had nothing to lose by seeming conciliatory.

In either case, Serbian defiance as now revealedmeans that the correct model of the moment of truthat this point is that in our next model, table 15. HereBS has de facto not agreed to U’s position, eventhough formally it has done so. Only BG has agreed.In addition, the threatened future is that U will blamethe Serbs and call air strikes against them. BS thenwill retaliate against UN personnel.

Modeling the Grand Strategic Drama

Note that the threatened future in table 15 comes fromthe NATO ultimatum rather than from the words of theUNPROFOR commander, who was strongly opposedto air strikes because he feared retaliation.

The NATO ultimatum emerged from a grand strategicdrama going on in parallel to the operational dramarepresented in table 15. We attempt to model thishigher-level drama in table 16, before discussing howit is linked to table 15. Again, our assumptions aboutthe grand strategic model use published sources only,and readers who may know them to be wrong areasked to make allowances.

233Chapter 8

First, what cards are in play and who holds them?

The United States controls the card, “Bomb Serbsinto compliance.” Such bombing would be done byNATO, but this card is assigned to the United Statesbecause, first, the United States has the necessarymilitary assets and, second, NATO as a whole, apartfrom Britain, is prepared to support the United Statesin this.

Table 15. Ultimatum to the Bosnian Serbs.


Russia has the important card, “Back Serbs againstNATO.” Russia’s pro-Slav sympathies and desire tohave a friendly Slavic nation controlling formerYugoslavia incline it to play this card, if necessary, tooffset what it sees as Western anti-Serb bias.

The Bosnian Serbs have the same cards in thisconfrontation as at the operational level (see table 15).

Finally, Britain has the card, “Publicly opposebombing,” a card it is inclined to play to keep areputation for impartiality and lessen the danger to itsvulnerable troops engaged in a humanitarian mission.

What positions are parties taking?

The United States and Britain (B) take somewhatdiffering positions, but because their positions areclose, we have collapsed them into one. We have donethis by leaving it open whether, under the joint position,B would publicly oppose bombing. B proposes to doso (to show impartiality and protect its troops), andthe United States is willing to let it do so. The UnitedStates, Britain, and the rest of NATO agree that theBosnian Serbs should cease fire and withdrawweapons, obviating the need for the United States tobomb the Serbs or for the Bosnian Serbs to retaliateagainst UN personnel.

Where Britain differs from the United States and therest of NATO is in being opposed to the U.S. fallbackposition, “Bomb Serbs into compliance.” Britain’s ownfallback position is to continue publicly to opposebombing. Britain takes this fallback position out ofpreference (to show impartiality and protect its troops)and to deter the United States from taking its fallbackposition by making that position less comfortable for

235Chapter 8

the United States (and so helping to give the UnitedStates inducement and threat dilemmas that mightlead it to change its fallback position). Britain is nottaking this fallback position as a way of inducing theBosnian Serbs to accept the joint position U,B.

Russia (R) at this stage has not taken a clear positionon any issue except that it is against NATO bombingof BS (it has agreed to close air support, but excludedpunitive air strikes). It has not taken the position ofbacking the Serbs against NATO, but is implicitlythreatening to do so if NATO bombing is implemented.

Meanwhile, the United States, Britain, and the rest ofNATO take no position on whether Russia shouldback the Serbs against NATO. That, they consider,is up to Russia. But if Russia should carry out itsimplicit threat to back the Serbs while NATO isbombing them, there would be serious concern abouta deterioration of relations between Russia and theUnited States, even fear of a new East-Westconfrontation arising over the Balkans.

The threatened future, t, is that the United Statescommences to bomb the Serbs into compliance (theaction would be taken by NATO, but our assumptionis that the decision would be taken by the UnitedStates); Russia responds by backing the Serbs againstNATO, with all the geopolitical risks that entails; theBosnian Serbs respond by retaliating against UNpersonnel; and the British publicly oppose bombing.Whether or not it is credible, this is the future impliedby the parties’ current fallback positions.


General Positions

In table 16 we have made free use of the symbol ~ torepresent general positions (i.e., to show that a partyhas not taken a clear position as to whether certaincards should be played). The symbol ~ means that acard may or may not be played. Hence, whenincorporated into a character’s position, it means thatthe character takes no position as to the playing ofthat card.

We must use this symbol here in order to be realistic.Your position is something you are willing to go to thethreatened future for. If you are not willing to go that far,then at the moment of truth it becomes clear thateffectively you are not taking a position on that card(i.e., your position is compatible both with it being playedand with it not being played). This is the case with theU.S. attitude to Britain publicly opposing bombing; theUnited States does not like it, but it is known it will acceptit. Similarly with the Bosnian Serb attitude to Russiabacking it against NATO; the Bosnian Serbs would likeit, but are not demanding it (at least not in this model,although if we modeled Serb-Russian negotiations, wemight find the Bosnian Serbs demanding it). Russia hasnot taken a firm position on any issue except NATObombing of the Serbs.

Use of the symbol ~ means that the different positionsof different parties may be compatible (e.g., in table16 Russia’s position is compatible with both of the othertwo, although they are incompatible with each other).

237Chapter 8

Table 16. Grand strategic pressure on the Bosnian Serbs.


Linkages Between Models—Use of Context Cards

We now have models of two confrontations, at theoperational and grand strategic levels, that areclearly linked.

First, we can see that the reason the Bosnian Serbshave the same cards in each model is that they areplaying simultaneously in both confrontations. At thegrand strategic level they are sending messages tothe United States, Russia, and Britain, while at theoperational level they are negotiating with theUNPROFOR commander’s staff and through it,indirectly with the Bosnian government.

Second, it is clear that the U.S. fallback position in thegrand strategic game (“Bomb the Serbs intocompliance”) partly dictates the UNPROFORcommander’s fallback position at the operational level(i.e., it means his fallback must contain the card, “Callair strikes against Serbs”); however, it does not dictatewhich other cards he must play. Because he is theman on the spot, responsible for interacting with theBosnian Serbs and having immediate responsibilityfor deciding if they have complied, he must decide atwhat exact point this fallback position should beimplemented. Realistically, we should regard him asholding and negotiating with the card, “Call air strikes,”while knowing (and knowing that the Serbs know, etc.)both that he has been directed to threaten its use andthat his decision to play it or not may be overridden bya U.S. decision in the grand strategic game. Thus heplays with this card knowing that at any time he maybe tapped on the shoulder and told what to do with it,but fearing that if he cannot get agreement, he mustuse it.

239Chapter 8

The Bosnian Serbs’ tactic makes use of this. Theyplan to negotiate with him while simultaneouslysending messages (both through him and through theRussians) in the grand strategic game. Their aim is tomake the threatened future in that game so bad forthe United States that it will back off from its fallbackposition, thereby making UNPROFOR back off fromits fallback position in the operational game. TheBosnian Serbs plan to do this partly by threatening toretaliate against UN personnel, partly by getting theRussians to threaten to back them against NATO, and,to a lesser extent, partly by the threat of Britishopposition to bombing.

The UNPROFOR commander’s tactic, on the otherhand, is driven by the fact that while opposed tobombing, both personally and in line with his nationalgovernment’s policy, to avert it he must crediblythreaten to do it. While credibly threatening, he can atthe same time use his discretionary powers to try towater down the ultimatum (e.g., let the Serbs keeptheir weapons in place provided he can monitor them),thereby making conciliatory adjustments to his positionto make it more acceptable to the Serbs. Hedesperately tries both tactics to get an agreed positionbefore the NATO ultimatum expires; however, hisbargaining position is strengthened by his ability touse the “nice-guy-mean-guy” argument that, eventhough he does not like bombing, he must order it ifthe Serbs do not comply.

Thus the models are linked first, because theBosnian Serbs are playing in both games at once,using the same cards, and secondly, theUNPROFOR commander’s choices in one game arepartly dictated by those made in the other, although


he does exercise some initiative as a subordinate;that is, he has cards he can play or not play and canchoose his negotiating strategy.

What does the linkage do to characters’ preferences?For example, to know if the United States and Britainhave a deterrence dilemma, we need to decide if theBosnian Serbs prefer t to U,B in table 16. Is this thesame as whether they prefer t to U, BG in table 15?

It is the same question, and will receive the sameanswer, provided we follow our constant advice, whichis to consider relevant factors outside the model whenmaking assumptions within a model.

In answering the question relative to table 16, we needto recall that in table 15 UNPROFOR is threatening toplay the card, “Blame the Serbs,” as part of t while atthe same time seeking to make compliance with U,BG less onerous for the Serbs.

In answering the same question in the context of table15, we need to recall that in table 16 the Russians areindicating they will play the card, “Back Serbs againstNATO,” as part of t, and the British are indicating theywill publicly oppose bombing.

This consideration of external factors can be doneformally by adding context cards to a model. Thismeans adding cards below a line labeled “Context,”as in the detailed model of table 7. We can add cardsshowing the assumptions made about any externalfactors. To show how this works, table 17 is the sameas table 15, but with added contextual assumptionsabout what would be happening in table 16. Theseassumptions are those stated above.

241Chapter 8

Table 17. Ultimatum to the Bosnian Serbs withcontext cards added from grand strategic model.


Dilemmas in the Grand Strategic andOperational Dramas

We have looked at players’ positions and tactics inthe grand strategic and operational confrontations asNATO presents its ultimatum to the Bosnian Serbs.We have not spelled out the dilemmas they face.

The dilemmas in the grand strategic confrontation areshown in table 18.

The United States and Britain face an inducementdilemma and the United States faces a threat dilemmain that t could lead to an East–West confrontation, aswell as retaliation against UN personnel; therefore, itseems that the United States and Britain would preferthe Bosnian Serb position to t (this is their inducementdilemma) and the United States would prefer not tobomb the Serbs (its threat dilemma). Britain’s reactionsinclude fear of an East–West confrontation (and ofendangering British troops on the ground); this fearwas conveyed to the Bosnian Serbs by David Owen,who was moved by it to conduct diplomacy against airstrikes (Silber and Little, 1996, pp. 311-313; Owen,1996, chap. 7). The U.S. reaction includedrationalization of preference for t over column BS byissuing an ultimatum and following the ultimatum,arguing that NATO credibility is now at stake.

The United States and Britain face a deterrencedilemma in relation to the Bosnian Serbs, who see anEast–West confrontation in which Russia takes theirside (this is the BS view of t) as likely, in their view, toend with the West making compromises in their favor,and hence as being preferable to position U,B. U.S.and British reactions include attempts to get Russia

243Chapter 8

to threaten the Bosnian Serbs with lack of support.U.S. President Clinton phoned Russian PresidentYeltsin and British Prime Minister Major visited Yeltsinin Moscow to request Russia to pressure the BosnianSerbs to accept the position U,B.

The United States and Britain face a trust dilemmabecause they cannot trust the Bosnian Serbs tocontinue to implement the agreement U,B, even if theyshould seem to comply (i.e., even if they begin toimplement it in the short term).

Russia faces a threat dilemma in that Russia, like theUnited States and Britain, also fears a renewed East–West confrontation and would prefer not to play thecard, “Back Serbs against NATO,” as part of t. Thereaction included conflicting conciliatory and angrysignals from Moscow, ending, after a visit by BritishPrime Minister to the Russian president, withacceptance by Russia of a joint Anglo-Russian positioncompatible with the U,B position in table 15 and a

Table 18. Dilemmas in table 16.


threat to the Bosnian Serbs that if they rejected this,Russia would not back them against NATO.

The next step in the resolution process is that Russia,in an attempt to eliminate its threat dilemma, shifts itsposition and brings about a new moment of truth;however, before looking at this, we will look at thedilemmas found in table 15, where the UNPROFORcommander negotiates at the operational level withthe Bosnian Serbs. Table 17 is the same as table 15except that context cards have been added to showinfluences from the grand strategic level.

The dilemmas found at this level are shown in table 19.

Table 19. Dilemmas in tables 15 and 17.

U (UNPROFOR) and BG (the Bosnian government,now sharing U’s position) face deterrence and trustdilemmas. Column t puts BS under no pressure toaccept column U,BG (this is the deterrence dilemmafacing U and BG.) In any case, because BS alreadyhave formally accepted it, any further agreement ontheir part to do so could hardly be trusted (the trustdilemma facing U and BG). Reactions include U

245Chapter 8

attempts to improve U,BG from the BS viewpoint byreinterpreting the agreement to allow BS guns to stayin their positions subject to monitoring. For BG, onthe other hand, these dilemmas are welcome asarguments in favor of abandoning the position U,BGand resorting to a position it much prefers (i.e., thethreatened future t); therefore, BG urges strictcompliance with the agreement U,BG, hoping for non-compliance and the implementation of t.

U also has inducement and threat dilemmas. TheUNPROFOR commander himself openly states thathe prefers not to call air strikes against the Serbs, andhis dislike of this is supported by the grand strategicfear of East-West conflict. The reaction includes amixture of anger and despair on the part of U, whichnevertheless keeps trying to get an agreement.

BG also faces a cooperation dilemma (it prefers notto keep to the agreement) and a positioning dilemma(it prefers BS to U,BG) because under BS it wouldhope for eventual Western intervention. The reactionis that these dilemmas are quite welcome for BG. Ithas been forced into unwilling acceptance of columnU,BG. If this column is implemented, BG no doubtintends to defect from it in due course, breaking thecease-fire, perhaps gaining some ground, butprovoking further BS attacks on the basis of which itcan make further appeals to international opinion.

How Real Pressure Was Brought to Bear onthe Serbs

Note that in these models the Bosnian Serbs face nodilemmas. Their position BS is preferred by them andstable for others, so that they have no cooperation, trust,


or positioning dilemmas. Also, t is attractive to thembecause it would involve Russian support, so that theyhave no inducement or threat dilemmas. Also, they haveno deterrence dilemmas because t is feared by Russiaand the Western players, while BG likes BS.

All this changes after the British Prime Minister’s visitto Moscow and Russia’s subsequent agreement toadopt a joint position with Britain and put pressure onthe Serbs to accept it. In table 20, the British-Russianjoint position is shown in column B,R. Here Russiaproposes to give its backing to the Serbs in return forSerb agreement to cease fire and withdraw weapons,while Britain publicly opposes bombing. Note that thisis compatible with the U.S. position. This is not to saythat the United States likes all its elements; it dislikesRussian backing for the Serbs and British oppositionto bombing. Its compatibility with the U.S. positionmerely means that the United States is prepared toaccept it, rather than go to its fallback position.

U, B, and R are now united against BS in demandingthat the Serbs cease fire and withdraw; however, thesignificant point is that Russia’s fallback position hasshifted. If BS rejects B,R, then Russia no longer willback the Serbs. This becomes clear to the BosnianSerbs when Russian Ambassador Churkin visits theSerbs and suggests position B,R. On the same visit,Russia assures the Bosnian Serbs of its willingnessto back the Serbs against NATO if the Serbs will agreeto position B,R by undertaking to send Russian troopsto supervise the withdrawal of Bosnian Serbs heavyweapons. This changes everything for BS. It is nowimplicit that, if they refuse this offer, Russia will notback them against NATO. The threatened future t isnow one in which the world, including Russia, is against

247Chapter 8

them. Also, t now holds fewer fears for the UnitedStates and Britain (no more fear of a new East–Westconfrontation). Britain’s fear of Bosnian Serb retaliationagainst UN personnel remains, but affords little comfortbecause U.S. troops are not threatened, and it is theUnited States that controls the card, “Bomb Serbs.”

Table 20. Real pressure brought tobear at last on the Bosnian Serbs.


The dilemmas arising in table 20 are shown in table21. The United States and Russia have only thedilemma that they cannot trust the Bosnian Serbs tokeep a cease-fire for long, after they have agreed toone. Britain has, in addition, the inducement dilemmathat it prefers the Bosnian Serbs’ position to t (becauseof the threat t poses to British troops). This is a problemfor Britain, but does not help the Bosnian Serbsbecause it is not Britain’s fallback position, but that ofthe United States, which makes t unpleasant for theBosnian Serbs. The Bosnian Serbs have a deterrenceand an inducement dilemma of a kind that they cansee no way to solve.

The Bosnian Serb’s response is to seize upon the offerby B,R as a position it even prefers to its own becauseit includes Russian backing. BS is joyful; arrivingRussian troops are greeted by cheering crowds.

Table 21. Dilemmas in table 20.

249Chapter 8

There is now agreement by all parties on the positionB,R (the intersection of the two compatible positionsU and B,R). This is reflected in the operational drama,as shown in table 22, where all parties now take thesame position. The remaining dilemmas are those ofcooperation and trust, indicated by the question markson the cease fire cards played by the Bosnian Serbsand the Bosnian government. Note that the Bosniangovernment here is accepting the position column thatis worst for it among the three columns shown. Thisarose, as we have seen, because the Bosniangovernment was threatened with a still worse column(not shown here) where UNPROFOR would blame itfor the market square massacre; however, the factthat it prefers either of the other two columns shownto the one it is accepting emphasizes its cooperationdilemma (and UNPROFOR’s trust dilemma). It cannotbe trusted to stick to the agreement. Reactions to thesedilemmas include the following: deceit on the part ofthe Bosnian Serbs and the Bosnian government,hiding their intention to break the cease-fire; disbeliefon the part of U, leading to the UNPROFORcommander monitoring the cease-fire and trying to findsanctions that will make defection unattractive andbenefits that will make the agreement attractive to theBosnian Serbs and the Bosnian government.


Table 22. Resolution (deceptive) of theoperational drama, with grand-strategic context.

251Chapter 8

What This Analysis Has Achieved

In this chapter we analyzed the dilemmas faced by theUNPROFOR commander and certain characters heinteracted with in 1994. We tried to show that theirreactions to these dilemmas can explain the emotionsfelt and initiatives undertaken by the characters.

We explained how and why the UNPROFORcommander, faced with Bosnian governmentintransigence, reacted angrily, thereby making crediblea newly minted threat to place public blame on theMuslims. This brought the Bosnian government intofearful, apparent compliance with UNPROFOR’sposition. There remained the problem that the Serbs,although they had declared themselves willing tocomply, now showed unequivocally by their actions,public statements, and defiant emotional attitude thattheir position now was one of non-compliance. Theywere eventually brought into apparent compliance byUNPROFOR’s threat of bombardment; however, thisthreat was made credible only by grand strategic action(the NATO ultimatum) and became adequate onlywhen Russia, again acting on the grand strategic level,made it clear that the Serbs would not have its backingagainst NATO. After it was made credible andadequate, the threat achieved apparent Serbcompliance, although the Serbs remained defiant. TheSerbs took consolation from the fact that Russia,thought to be their ally, had become involved. Theeventual attitudes of both Serbs and Muslims indicatedthat their compliance was more apparent than real.

Our detailed explanations of these events are posthoc; moreover, explanation by itself is of little value.


The important question is whether this analysis, if doneat the time, would have helped the UNPROFORcommander achieve his mission objectives. Thisquestion is addressed in the next chapter.


To illustrate how the approach throws light on a real-world confrontation, this chapter reports an analysiscarried out for DERA of the episode in February 1994when the UNPROFOR commander was tasked withstopping the Bosnian-Serb bombardment of Sarajevo.The analysis was drawn from published sources only.Interviews were conducted with officers who had servedat various levels in the Bosnian mission, but not at thisstage or during this particular crisis. An explanation ofthe behavior and reactions of the various parties is givenusing drama-theoretic principles.

253

Chapter 9

Would This AnalysisHave Helped?

Analyzing a confrontation, as in the previouschapter’s analysis of the Sarajevo crisis of

February 1994, does not answer the question, “Howmight confrontation analysis have helped thecommander’s operation-level decisions, if the methodhad been available to him at the time?”

To answer this, we will need to assume that ouranalysis is broadly correct, despite the fact that it isbased on published accounts written after the event,rather than on participants’ on-the-spot judgments. Notonly are published accounts a poor guide to how thingsseemed at the time, they also have the disadvantageof concentrating on events at the operational and grandstrategic level, rather than the tactical level. The effectis that we will not be able to say much about how thecommander might have devolved his operationalstrategy to lower levels.

Despite these difficulties, we will project the analysisbackward, beginning with the situation after the marketsquare bomb, and imagine what might have happenedif the commander had been able to formulate andfollow a confrontation strategy guided by our analysis.


Rerunning the History of a Crisis

On February 5, 1994, let the commander begin bybuilding the model in table 13. At this point, he hasrealized that the Bosnian president will not allow hisofficials to attend cease-fire talks because he is tryingto use the situation to bring about internationalintervention on his side.

Determining End States

In the course of building this model, the commanderhas defined his end state. This is column U, the futurein which BS (Bosnian Serbs) ceases fire and withdrawsweapons, BG (Bosnian government) also ceases fire,U (UNPROFOR) remains impartial (not blaming Serbsnor calling air strikes), and there is no retaliationagainst UN personnel.

This is the same as U’s position, although a character’send state is not always its position. Your position is yourovertly declared end state, what you propose to the othercharacters as a solution. We have seen that a characterwith a cooperation dilemma has a temptation (perhapsworking with others) to move from its position, ifaccepted, to another future that it prefers. It may solvethis dilemma by deceit, so that its secret intention andactual end state is to carry out its temptation, not itsposition. Again, a character may have reluctantlyaccepted a position while planning for it to be rejectedin favor of another. In this case, the latter rather thanthe former will be its actual end state.

In table 13, it appears that each party’s end state isidentifiable with its position; thus, the commander, inbuilding this model, has determined not only his own

255Chapter 9

end state, but the other parties’ too. He has identifiedtheir goals.

End states found in a model of a common referenceframe, which is what a card-table is, are necessarilysimple. Simplicity is desirable. It is also realistic. In aconfrontation, parties’ end states, representing whatthey want others to accept, are necessarily simple.At the same time, the commander’s staff will look atthese end states in more detail to explore theirimplications and consequences. They do this byadding extra cards, both internal (i.e., above thecontext line) and external (below it), as in table 7.Simplicity, clarity, and certainty are lost in this process,but important points may be uncovered and difficultquestions raised. Going into detail in this way is alsonecessary to devolve a strategy to the tactical level.

To find end states for a mission in a card-table model,you obviously must model that mission, and themission must be a confrontation (i.e., an OperationOther than War [OOTW]). In our example, thecommander has modeled the mission he has justreceived, not his previous and continuing humanitarianmission, and has determined his end state for this new,obviously confrontational mission. Lower-level,devolved models will enable determination of endstates for lower-level missions.

Conversely, to determine his superior’s end state, andthe end state of his superior’s superior, the commandermodels the grand strategic drama by building themodel in table 16. In doing so, he identifies hissuperior’s intent by modeling how Britain (his country),the United States (supported in this by the rest ofNATO), and Russia see the application of pressure to


the Bosnian Serbs at the grand strategic level. Observethat his superior’s intent appears, not as a clearlydefined mission, but as a position in a political dramabeing played out between national governments. Theend state of this conglomerate superior is column U,Bin table 16.

The commander’s end state is, in general, a moredetailed implementation of his superiors’ end state,just as the end states in strategies devolved from hisstrategy will be more detailed implementations of hisown end state. In moving directly to achieve his ownend state he is helping his superiors to achieve theirsand using his subordinates to implement matters ofdetail within his.

In the case we are considering, the commander findsa problem with his superiors’ model of the situation.They are implicitly assuming that the Bosniangovernment shares the position of the U.S. and Britishgovernments, or at least accepts what thesegovernments think is good for it. BG (the Bosniangovernment) is not even a player in the grand strategicmodel. This reflects the way the governments see thismoment of truth. They see themselves as negotiatingwith each other over how to bring the Bosnian Serbsinto line. BG does not enter into it.

How should the commander deal with the fact that,contrary to his superiors’ assumptions, BG is planningto take advantage of their blindness by refusing tosettle, knowing that the Serbs will be blamed for theresultant perpetuation of the bombardment?

In light of his superiors’ intent, the answer is clear. Hemust use his own initiative to bring BG into line.

257Chapter 9

Centers of Gravity and Sequences ofDecisive Points

Use of some familiar military terminology may helpdescribe what the commander does. We can say heattains his end state by finding a sequence of decisivepoints by which to destroy the centers of gravity ofnon-compliant parties.

The significance of looking for other parties’ centersof gravity is this. Chapter 7 contained a description ofthe commander as cycling through a number ofdilemma eliminations to achieve his end state;however, formal analysis of dilemmas is never enough.It is essential to look at the real-world context in whichdilemmas arise to see how to eliminate them in thedesired manner (i.e., without pressuring others intofurther escalation or into adoption of yet more contrarypositions). This can be described as a matter ofdefining and attacking the center of gravity of theirnon-compliance.

As an example, the center of gravity of the Bosniangovernment’s non-compliance could be defined as itsproject of working on Western public opinion to bringthe West to its side. This center of gravity was effectivelyattacked by the UNPROFOR commander’s threat toblame the Muslims for the market-square bomb.

This illustrates the point that to define a non-compliantparty’s center of gravity, it is helpful to first decide whatits policy is. Is its position non-compliant? If so, it musthave a policy by which it hopes to get acceptance ofthat position, and that policy must be overthrown, asthe UNPROFOR commander overthrew the Bosniangovernment policy of prolonging the Serb bombardment


to arouse international opinion against the Serbs. Onthe other hand, it may be that its overt position iscompliant, and its non-compliance lies in its presumedintention to defect from it. In that case, it is the policy ofdefection that must be overthrown. Its center of gravitymay be the belief that defection will be possible or thebelief that it will be profitable.

In seeking to define centers of gravity, it usually isnecessary to analyze (formally or informally) theinternal drama going on within other players. Thepresent policy of non-compliance has been arrived atas the resolution of a confrontation between variousinternal subcharacters. Our aim must be to launch anattack on certain agreed underpinnings of that policy,thereby throwing it into crisis and bringing about a shiftto a policy of compliance. The commander’s task is toeliminate dilemmas at his own level, working wherenecessary through subordinates, coordinating withothers on his own level, and requesting assistance asnecessary from higher levels.

Each round of dilemma elimination at his own levelmay be described as a decisive point for thecommander. His confrontation strategy as a whole maybe described as a plan to move through a sequenceof decisive points, success at all of which will eliminatenon-compliant centers of gravity and bring all partiesinto sincere compliance, so achieving his end state.

Planning and Following a Confrontation Strategy

To show how he proceeds, we will describe severalcycles of dilemma elimination. These will illustrate bothhow he might have planned a confrontation strategyand how he might have carried it out.

259Chapter 9

Step 1: We have said that he starts by building table13. The dilemmas found here are set out in table 14.BS has no dilemmas. Both U and BG have deterrencedilemmas and trust dilemmas. Neither BS nor BG isunder pressure to accept the commander’s position.

Step 2: The commander must solve his deterrencedilemma against at least one of the other parties (i.e.,bring pressure on at least one). Now he knows thathis superiors are thinking how to bring BS (the BosnianSerbs) into line. They are not even thinking about BG;therefore, he decides that he must deal with BG.

Now BG’s deterrence dilemma means that its positionis weak at the operational level because it has nomeans of inducing the commander or BS to accept itspositive position (column BG). Its strategy is to remedythis by operating on the grand strategic level to changethe commander’s preferences from above. He canforestall this if he can overcome his deterrencedilemma toward BG using a credible, useful threat thatwill induce it to comply. (He may not be able to bring itinto true compliance because it may still intend todefect from his position, but apparent compliance is afirst step.)

The commander brainstorms with his staff forpossible cards. The commander considers the BGcenter of gravity is its desire to swing internationalopinion on its side; therefore, an appropriate card topressure BG with is, “Blame Muslims for marketsquare bomb,” a card suggested by the evidencethat they were responsible. He provisionally addsthis card to his hand.


Having done so, he builds table 11, which includesthis new card in his fallback position. He finds, onanalyzing this table, that he has solved his deterrencedilemma but given himself inducement and threatdilemmas (in that he prefers not to use the new card).He decides to overcome them either by showing anger(as was actually done in 1994) or in some other wayfound by brainstorming and role-playing the problem(e.g., he might represent Muslim blameworthiness asa logical deduction from BG’s unwillingness to adoptthe position U).

Solving his own dilemmas will, he decides, heightenBG’s inducement dilemma to the point where it cansolve it only by accepting his position.

Step 3: At this step he successfully carries out, or, inplanning mode, rehearses, these dilemma-eliminations. Rehearsal methods described in chapter10 are a good way of testing this phase of aconfrontation strategy.

Step 4 consists of a return to Step 1. Because theplanned dilemma-elimination has succeeded (or, inplanning mode, has been assumed to havesucceeded), in shifting BG’s position to overt, ifunreliable, acceptance of his position, the commanderbuilds the model in table 17.

This model has context added from the grand strategicdrama because this now seems to be of overridingimportance. To establish this context, the commanderalso builds table 16, thereby modeling the problemfaced by his superior.

In the new model at his own operational level, he facesup to the problem of bringing BS into apparent

261Chapter 9

compliance. He identifies the dilemmas in the modelas being those in table 19. He finds that he facesdilemmas of deterrence, inducement, and threat, aswell as trust. BS has no dilemmas. BG has dilemmas,but because BG is still clearly aiming for a positionother than the one it is overtly taking, it welcomes them,regarding them as arguments against its position andin favor of the position it would rather be taking.

Step 2 : What can the commander do? In neither theoperational nor the grand-strategic level table doesBS now face any dilemmas. No pressure is beingbrought on BS.

This is disconcerting. The commander decides to usebrainstorming and role-playing to search for the BScenter of gravity. He finds that BS does not fear t ineither model because BS sees R (Russia) backing it ifNATO attacks. It seems, then, that BS’s belief inRussian backing is the BS center of gravity. From hissuperior’s model in table 16, the commander finds thatR has an inducement dilemma and a threat dilemmain backing BS in the event of NATO air attacks.

That is useful information. The commander decidesto use it to attack the BS center of gravity (afterchecking with his superiors, which he needs to dobecause the tactic of using these Russian dilemmashas grand strategic implications). His plan is to pointout to BS that R will certainly not back it when thechips are down. Getting this message acrosseffectively should eliminate his (UNPROFOR’s)deterrence dilemma and give BS an inducementdilemma by making BS prefer his position to t.


Simultaneously, he plans to eliminate his owninducement and threat dilemmas and give BS adeterrence dilemma by pointing out that NATO iscommitted by its own ultimatum, and cannot now backdown. NATO, unlike the UN, cannot afford to losecredibility. NATO’s commitment is such that even thethreat of retaliation against UN personnel is notsufficient to make column BS preferred to t; such athreat merely angers Western public opinion andincreases the U.S. preference for t over the BosnianSerbs’ position, and, he will point out, U.S. preference,not British, is what makes t credible. He plans to pointout that Russia knows all this, and therefore will itselfback down; this is why it has not even taken a firmposition. He will point out the inevitable conclusionthat Bosnian Serbs’ intransigence will lead to NATOair action plus withdrawal of Russian support. He willtry to get this message reinforced by all sources incontact with the Bosnian Serbs, thus achieving unityof effort. For good measure, to take away any crumbof comfort the Bosnian Serbs may find in thethreatened future t, he will add the assurance that whileBritain does at present publicly oppose bombing, itwill not do so if NATO takes action.

The commander builds these plans from analysis ofdilemmas in conjunction with appreciation of theattitudes of the Bosnian Serbs. He has one problem,though. To overcome his inducement dilemma, hemust find some way of dealing with his own reluctanceto use air strikes.

He does not like air strikes. He fears their damagingeffect on his ongoing humanitarian mission and prefersnot to risk the lives of UN personnel. Yet his analysis

263Chapter 9

makes it clear that to get rid of his inducement dilemmaand give the Bosnian Serbs a decisive deterrencedilemma he must find some way of preferring t to theBosnian Serbs’ position. He finds an answer to thisdilemma in the principle that he must try to implementhis superior’s intent. His superior, in this case, is acongeries of governments swayed by media-hypedpublic opinion and arriving at decisions through aprocess of political confrontation. Right or wrong, thatis his superior. It is not, in this case, manifestly wrong.It is clear that the firm line being taken by the UnitedStates and supported by the rest of NATO (apart fromhis own government) means that his superior isswinging in the direction of maximum pressure to makethe Bosnian Serbs comply; therefore, clearly hiseffective preference, mandated by his superior’s intent,is for t rather than BS, much as he dislikes t.

BS simply must give in.

Step 3: BS does give in (or its role-players do, if weare still in planning mode). Another decisive point ispassed. This occurs because U convinces BS thatthe true model it faces if it does not shift its position isthat in table 23. Here, t involves neither R backing theSerbs nor B publicly opposing bombing, but continuingair strikes preferred by U to BS.


Table 23. Ultimatum to the Bosnian Serbswith threatening grand strategic context.

265Chapter 9

The Rerun Compared to the Reality

Observe that this rerun of the past achieves thesame end result as the reality, but achieves it in adifferent way.

In reality, U got BG to overtly accept U’s position bythreatening to blame the Muslims for the market squarebomb. This achieved the first round of dilemmaelimination described in our rerun; however, ratherthan being done in a planned and thoughtful way, withan adequate appreciation of risks and ramifications,the elimination was done impulsively and emotionally.The commander got angry and made his threatcredible in that way.

The commander’s instincts were right, and he wassuccessful; moreover, when there is little time to act,action must be fast and decisive. As a whole, U(including the commander, his staff, and UNPROFORcoevals) would have benefited from a clearer, morecommunicable understanding of what he was doingto win a confrontational battle. A major benefit mighthave been to win in such a way as to enlist more willingsupport from BG for U’s position (e.g., by making itseem to BG that U’s motivation was to prevent BGfrom ruining its case at the court of Western publicopinion). BG’s actual support, after being coerced byangry threats, was unwilling and reluctant.

The second round of dilemma elimination was notachieved at all; not, that is, at U’s operational level. Unegotiated with BS throughout the 10 days of theNATO ultimatum in terms of the model in table 17 anddid not succeed in eliminating any dilemmas from it.


This was the model U had created by successfullybringing BG into line. In it, BS faced no dilemmas,and hence U could bring no pressure to bear on BS.Aware of this, U tried to fend off air-strikes byconcessions to BS in terms of the locations to whichheavy weapons must be withdrawn and the degree oftheir supervision, hoping thereby to entice BS intowithdrawal. In this way U tried to solve its deterrencedilemma and give BS an inducement dilemma bysweetening U’s offer. Apparently U failed in this, asBS simply failed to withdraw its weapons (Silber andLittle, 1996, pp. 316-317).

U was saved, according to our analysis, bysimultaneous action on the grand strategic level.Effectively, the politicians came to U’s rescue. AfterBritish Prime Minister Major’s visit to Moscow, RussianAmbassador Churkin saw the Bosnian Serbs at theirheadquarters in Pale, told them that Russia backedthe NATO ultimatum, and offered them Russianbacking if they accepted its conditions. He made thislast offer by saying he would send Russian troops tomonitor the withdrawal of BS weapons. Note that hedid not explicitly tell BS that if they rejected NATO’sconditions, then Russia would leave them in the lurch.It is undiplomatic to spell out your threats. To get willingcooperation, a positive note is necessary. In failing totell them that Russia would, of course, back themunder all circumstances, the ambassador was beingsufficiently explicit and at the same time positive, asthe Serbs made clear by joyously accepting hisconditional offer of support.

Such was the reality. In our rerun, U takes advantageof precisely the same configuration of grand strategicdilemmas to solve the problem at operational level,

267Chapter 9

although note that in doing so, it first checks its strategywith its grand strategic superiors. Checking its strategyin this way also enables it to achieve usefulcoordination with the grand strategic level, givingreassurance that the messages BS receives on thatlevel will confirm what U is telling it, the essential pointof which is that it cannot rely on Russian backingagainst NATO air strikes. Note that U’s argument thatBS cannot rely on R depends on the very Russianinducement and threat dilemma (that R feared a t thatmight lead to a new confrontation between East andWest) that led Ambassador Churkin to visit Pale andmake the same point.

In our rerun, U uses these powerful arguments to winthe confrontation at operational level. BS gives in. Wecannot, of course, say with certainty that U could haveachieved this. BS might have persisted in its grandstrategic efforts until R communicated directly with it.What we can say is that if U attempted dilemmaelimination along the lines sketched in our rerun, itwould have been pursuing a confrontation strategycoordinated between grand strategic and operationallevels, and thereby would have been pursuing itssuperior’s intent. It is certain, too, that such a strategy,coordinated between different levels, would have beenfar more effective than the ineffective and ultimatelydamaging efforts to sweeten its offer to BS that werepursued in reality.

In military terms, U’s strategy would have seized andmaintained the initiative by attacking BS’s center ofgravity rather than, as in reality, trying unsuccessfullyto conciliate BS by progressively abandoning U’sown position.


The Next Step in Our Rerun of Reality—Stabilizing the Agreement

Step 4 (A New Step 1): Assuming we have coercedBS to overtly comply with its position, U returns to thedrawing board. Its new model of the situation is thatseen in table 22. The dilemmas it finds in this are setout in table 24. BS and BG both have temptations todefect, preferring to break the cease-fire not onlyindividually but collectively. They would prefer, in otherwords, to resume hostilities. BS would prefer thisbecause they wish to gain ground while they can, beingaware of the growing strength of a Croat–Muslimcoalition being formed against them. BG would preferit, now or in the near future, partly because they hopeto be able to drive the Serbs back and partly because,if they fail, the renewed bombardment of Sarajevomight swing world opinion in favor of intervention ontheir behalf.

The agreement is holding for the time being becauseit has just been signed and UN troops have beeninterposed between the two sides. In thesecircumstances, neither side wants to be seen to bethe first to break it; nevertheless, it requires stabilizingbefore one or the other moves against it.

Step 2: How is the agreement to be stabilized? Theproblem is to change the perception by BS and BG ofthe advantage to them of an end to the cease-fire.Unfortunately published sources give too littleinformation for us to say how this problem can be (orwas) tackled. Threatened media exposure of violationsmay deter BG but not BS, whose image in the mediacan hardly get worse. Credible threats of reprisals maybe needed to deter BS. Recognizing that the dilemmas

269Chapter 9

Table 24. Dilemmas facing characters in table 22.

facing BS and BG are cooperation dilemmas, for whicha positive tone is appropriate, U makes sure that allthreats are part of a harmonious, positive stabilizationpackage constructed with full consultation andincluding carrots, however symbolic, as well as sticks.

This package must be implemented at the tacticalas well as operational level. Consequently thecommander’s staff, after putting all the detail it canthink of into an expansion of the overall model, drawsup guidelines for local commanders. Within theseguidelines, local commanders can analyze theirconfrontations and draw up local packages toeliminate dilemmas in them. As in the localcommander’s model in table 3, BS and U units oftenwill take differing positions in these local models(e.g., in regard to whether certain arrangementsconform to the agreement). Negative emotions thenwill be demanded, within the overall positive tone ofthe agreement.

There are potential problems of coalition relations inregard to Russian troops guarding some BS weaponcantonments. These will have to be managed in a


simultaneous, coordinated manner at operational andtactical level.

Step 3: Suppose their dilemmas of cooperation andours of trust are solved, either in reality or role-playingmode. The agreement is then stable, for now;however, it is likely that external events, interruptions,may disturb it by changing the conditions under whichit holds.

Predictions Made by Drama Theory andConfrontation Analysis

Rerunning reality as we have done highlights the kindof predictions drama theory makes. It is important tounderstand that they are not deterministic. Neither arethey probabilistic.

A character may react in various ways to the dilemmasit faces; however, if the assumptions fed in are correct,the analysis identifies the dilemmas each charactermust react to and allows us to predict that it will reactin such a way as to try and eliminate those dilemmas.What reactions are available to it will depend on theconcrete details of the particular situation (i.e., on thefriction it encounters in trying to change its moment oftruth in one way or another). Ways of changing themoment of truth may include changing position,changing preferences, thinking up new cards to playor taking action (e.g., military strikes) to deprive othersof cards.

In this way, the theory uses models in an unusual way.We cannot, as in most other approaches, build amodel, manipulate it formally to derive predictions, thencome back to reality to apply those predictions. From

271Chapter 9

our model-building and formal analysis, we derive notpredictions, but questions about reality in these forms:What means are available for X to try and solve such-and-such dilemmas? It is our answers to thesequestions that provide us with the predictions we feelable to make.

Use of a confrontation model, rather than turning usaway from reality, turns us back to it with searchingquestions to be answered.

Comments on This Rerun

This particular rerun of reality has suffered from beingbased on assumptions taken from third-party,retrospective, published sources, rather than the viewsof actual participants at the time. This affects morethan the reliability of the analysis, although we haveasked readers to discount inaccuracies. Drama theoryis based on the game as perceived by the players.Their views, not those of others, determine their actionsand reactions; therefore, theoretically it needs to bebased on their views at the time.

Relying on published sources also means that we haveincomplete data, particularly as regards confrontationsat the tactical level. We could not say much abouthow BS and BG might have been brought into actual,not just apparent, compliance with a cease-fire.

Accepting all this, what difference might it have madeif a confrontation strategy had been formed and used?It appears that the same result would have beenachieved in bringing the BS and BG into apparentcompliance with a cease-fire, with certain differences.


The bringing of BG into apparent compliance mighthave been done in a planned way rather than throughspontaneous reactions. Further, it might have beenplanned to make BG more willing to comply, ratherthan feeling coerced.

The bringing of BS into apparent compliance might havebeen achieved sooner on the operational level insteadof being achieved, after unavailing concessions hadbeen made, on the grand strategic level. In achievingthis result on the operational level, the commanderwould have been implementing his superior’s intent.Some of the tactics used, in particular, the forecastingof Russian reactions, might have had to be checkedwith his superiors. Others would not. Coordinationbetween the operational and grand strategic levelswould have been enhanced.

Achieving the result on the operational level, in themanner outlined, would have meant some differencesin the result on the grand strategic level. The actualresult involved Russia. If the commander had achievedit in the manner we discuss, Russian involvementmight not have been necessary. Perhaps it wasdesirable. Perhaps Russian involvement was a grandstrategic objective. This emphasizes the necessity fora commander to check with his superior before usingtactics with higher-level implications. His checkingmight have led to an improved plan for obtainingRussian involvement, if this was an objective.

However much or little the putative result would havediffered from the actual one, it would have beenachieved (if a confrontation strategy had been used,at the initiative of the commander) instead of by last-minute action at a higher level after the commander

273Chapter 9

had made significant concessions from his original endstate. It would have enabled the commander, fromthe beginning, to seize and maintain the initiative.

Insufficient data in public sources make it hard to saymuch about turning apparent compliance by theBosnian parties into actual compliance. It may or maynot have been achievable using available assets andwithin given guidelines. If it was not, attempts toformulate a confrontation strategy would have revealedthis, enabling inadequacies to be reported upward andrequests for extra support clearly justified.


The analysis reported in the previous chapter is usedto show how in a real-world case the commander inan OOTW might have formulated and implemented aconfrontation strategy. Two rounds of dilemmaeliminations that might have been carried out as partof such a strategy are suggested.

Comparing this rerun with what actually happened,the conclusion is that the first round of dilemmaeliminations might have been achieved much as inreality, but in a more planned way and perhaps withgreater effectiveness in inducing willing compliance.The second suggested round of dilemma eliminationswas not achieved at all by the commander on theoperational level; it required grand strategic action. Ifit had been carried out by the commander, he wouldhave been operating a strategy coordinated betweenoperational and grand strategic levels, seizing andmaintaining the initiative, avoiding costly concessions,and carrying out his superior’s intent.


While we hope this attempted rerun may be useful, itshould be pointed out that it makes use of hindsight,whereas confrontation analysis is most effective whenused to understand and plan for an actual, ongoingoperation, not one that already has taken place. Thisis so because various futures that seemed only toopossible at the time seem by the exercise of hindsightto have small probability just because they did nothappen. This diminishes the effectiveness and impactof confrontation analysis. Use of hindsight also makesthe rerun questionable, if not useless, as a test of thetheory or technique.

275

Chapter 10

Immersive Briefing andMediation Support

The main topic of this book has been how to buildand use a confrontation strategy to win an

Operation Other Than War (OOTW).

In dealing with this, we have several times mentionedtwo related topics we now discuss in more detail.

The first is role-playing. We will discuss a method ofwriting briefings for role-players called “immersivebriefing.” This is a drama-theoretic method of lookingat a set of linked confrontations from the viewpoint ofeach party, thinking how they must see the situationand how they must see the way each other must seeit, and so forth, and basing on this a set of briefingsfor role-players to take the parts of real-life characters.

The second topic is mediation support. Generally wehave assumed that confrontation analysis is used forunilateral decision support, helping one party deal withothers. A commander would use confrontation analysisin this mode to bring external players into compliancewith his end state. But, a commander also must dealwith confrontations that are essentially cooperative,in that the other players are on his side and hissuperior’s intent is that they should all coordinate theiractivities to achieve given objectives. Mediationbetween the parties is then more appropriate than


support to just one player. For this we will discuss amodified form of confrontation analysis.

Immersive Briefing

In step 2 of a confrontation strategy, you decide on abatch of dilemmas to eliminate and think how to do it.In step 3 you carry out your elimination plan, or if youare planning rather than implementing a strategy, youthink through what the results are likely to be. Fromthere you go on to the next cycle, modeling the newsituation you have brought about or imagined.

This was the process we described in chapter 7 and gavean example of in chapter 9. However, when planning orchecking your strategy, you can do better than justthinking through what its results are likely to be.

There is an inherent difficulty involved. On going intoa moment of truth you open up your beliefs and valuesto being changed; hence, it is theoretically impossibleto be sure how you will come out of it. While you cannotbe sure, an excellent way of stretching yourimagination and letting you feel the forces of changebeforehand is role-playing.

Role-playing is widely used for this purpose. Its resultsgenerally are seen by good actors as abysmal. Thereason is simple: however good you are, you cannotdo much with a bad script. Having a good script is not,however, a matter of having words to learn. It is amatter of knowing the life situation of the characteryou are acting, meaning not its personality orcharacter, but what it is trying to achieve, and whyand how, and what it thinks others are trying to achieve,and why and how. Knowing these things about your

277Chapter 10

character, you can start to act, not by putting onanother person’s mannerisms, but by putting on itslife, playing the game it has to play.

This requirement of good acting is precisely therequirement we have for exploring what may happenat a moment of truth. We need to be able to experiencethe moment of truth as closely as we can, and so weneed to recreate the characters we will be playingagainst. If we can be helped to throw ourselvescompletely into the problems and viewpoints that othercharacters inhabit, we can recreate them as we wantthem, try out things with them, and have the addedbonus of a deeper insight into their viewpoint. In thisway we can check assumptions made in the courseof planning a strategy (e.g., whether a non-compliantcharacter, placed in a certain situation, has anyalternative other than to give in).

Immersive briefing is a drama-theoretic way ofproducing briefings for role-players. Unlike most role-playing methods, it tends to produce good acting. Itseffectiveness derives from its analytical basis. It putsparticipants into the positions of other characters, sogetting them to understand their subjective feelingsand perspectives and enabling them to come up withcreative ideas (e.g., this character might react in such-and-such a way). Participants also come to understandtheir own confrontation strategy on a deep, intuitivelevel, enabling them to criticize it and suggestimprovements.


What Comprises a Briefing

An immersive briefing is constructed using specialsoftware. With its aid you create a plot situation wherea number of role-players can be briefed to interact,without scripts, as characters in a set of interconnectedconfrontations. There are no rules and no points toearn. The method is based on confrontation analysisof a situation, but briefings can be given to and usedby role-players with no knowledge of drama theory orconfrontation analysis.

Separate briefings are given to each role-player, whomay be an individual or a team. A character in theconfrontation that is being role-played is generally anorganization (e.g., a country or a an ethnic group).Individual role-players are told to act as responsiblerepresentatives of their organization. The briefings theyare given resemble those given to a commander takingover tasks from a predecessor. They should be bothcomplete and concise. A role-player taking on acharacter is briefed on the following items:

• Character background, including its organization,the internal subcharacters and subconfrontationsin its organization, and its relevant history

• Values and motivation, recognizing that acharacter will have many kinds of value systems,selfish and unselfish, long- and short-term, andthat these often conflict

• Current projects (current goals)

• Current relationships with other characters

• Other characters’ backgrounds, values, andprojects and the relationships between them

279Chapter 10

• Confrontations it is involved in with othercharacters.

The logic of this briefing is that a character’sbackground and relationships explain its values, itsvalues explain its projects, and its projects, becausethey are interdependent with the projects of others,explain the confrontations it is involved in.

In relation to each confrontation, a character is toldthe following facts:

• Who is involved (characters)

• What each character can do about it (cards)

• What position each is taking (i.e., what solutionto the conflict each recommends)

• What fallback position each character says it hasin case the others will not accept its position (orit believes will not genuinely accept it)

• What future threatens if these fallback positionsare implemented

• What future is forecast if currently implementedpolicies continue. This may or may not be thesame as the threatened future

• What strategy it and the other characterscurrently are pursuing as each tries to get theothers to accept its position.

A character may be role-played by a team of up toabout five people who digest and discuss theircharacter’s briefing among themselves. During internaldiscussions, these teams are segregated in separaterooms, in separate corners of a large room, or even in


separate locations, as when the drama is played overa computer network.

After receiving and digesting their briefings, characters(as played in role by role-players) are allowed tocommunicate with each other to implement theirstrategies (i.e., so that each character can try to getothers sincerely to accept its position forimplementation). For this each character tries to makecredible the threats (often implicit) and promisesspecified in its strategy. Communication may be througharranged meetings, fixed by intermediaries or by e-mailover a network. Various communication media may beused. Communication by public announcement (e.g.,giving a press conference), knowing that anothercharacter will read about it, can be simulated, such asby sending out a general e-mail purporting to be a newsreport of a press conference. Internal discussionscontinue between meetings.

Records of communications between characters helpin later analysis. They can be kept automatically ifcommunication is over a network.

What an Immersive Briefing Represents

An immersive briefing really represents a particularcharacter’s memory, or at least, the part of it that isrelevant to a particular set of confrontations. If acharacter is an organization, its briefing represents itsorganizational memory. That is why a briefing may belikened to the briefing given a new commander whenhe takes over responsibilities from a predecessor.

Three things follow from the fact that a briefingrepresents a character’s memory:

281Chapter 10

• A character can consult its briefing at any time.

• As a situation develops into a new one, abriefing becomes out-of-date. Like a diary thatstops at a certain date, it records things asthey were, or seemed to be, at a point now inthe past. Replacement of the briefing by amore up-to-date one is possible. Alternatively,characters can be left to make their own notesof new developments.

• Briefings are subjective and differ from oneanother. Character A may be briefed differentlyfrom B about past events, about its own oranother’s position in a particular conflict, and soforth. Certainly characters’ briefings will differ inthe values reflected in them, because eachbriefing will reflect the values of the characterbeing briefed. Consequently briefings aresecond- and third-order subjective, as well asfirst-order; that is, characters are briefed as to(their beliefs) about others’ beliefs and values,others’ beliefs about their own and others’beliefs, and so forth (see Bennett, 1977).

Using Software

The package for authoring an immersive briefingproduces a suite of software briefings, one for eachcharacter, reflecting the characters’ differentviewpoints and information. Current software providesbriefing information both in text form and card-tableform. Text or card-table is called up in a computerscreen window by clicking on menu items or picturesof the characters and the confrontations betweenthem. Further clicking on different parts of a card-table


calls up textual information about characters, positions,fallbacks, and strategies in that confrontation. Figure3 shows how a briefing written for the character,“Bosnian Govt,” looks before any pictures or menuitems have been clicked. This briefing was one of anexperimental suite of briefings written for DERA.

Pointing and clicking through an immersive softwarebriefing is, even without role-playing, an effective wayof absorbing a confrontation analysis. It might also beused, therefore, to present and distribute analyses.Users could brief themselves by looking through thebriefings given to different characters and comparingtheir viewpoints.

Friction and Change in Immersive Role-Play

Immersive briefings generally are designed to bringcharacters to a moment of truth. Recall what this means.A moment of truth is a point at the climax of aconfrontation when characters achieve the following:

• They finally understand each other (or think theydo), including each others’ positions andfallback positions.

• They are, as a result, brought up against theneed to change their fixed views of themselves,their situation, and each other if they are to avoidfalling into conflict.

At such a point characters tend to change theirdefinition of the confrontation they are in by redefiningthe characters in it, the cards they can play, and theirpreferences between futures, as well as their positions;however, their ability to do so is limited by the weightof existing values, evidence, mission definition, and

283C

hapter 10

Figure 3. Computer screen showing briefingon character role for Bosnian government.


so forth (i.e., by friction). Emotion enables them tobreak out of existing assumptions and routinereactions, and so to overcome friction.

Role-players accordingly are told that the frames andpositions their briefings give them are initial framesand positions only. They have inherited them from theirpredecessors but they can change them. In thisrespect, immersive role-playing differs from game-playing or ordinary role-playing. The only constraintsare imposed by the following circumstances:

• The past, consisting of characters’ backgroundsand the events and projects that brought them totheir current impasse

• Characters’ expectations of each other, and thedifficulty they may find in changing these. It willbe of little use to me to undergo a conversion sothat I prefer to carry out a promise or threat ifyou continue to believe I will not. Equally, achange of position or discovery of a new optionnot listed in the card-table will need to becommunicated convincingly

• Role-players’ sense of responsibility for properlyrepresenting their character. An immersive role-play can, in theory, be spoiled by anirresponsible player who acts out of role. At theother extreme, role-players may need to bereminded that they are free to change, as someare over-responsible. Such reminders can begiven in debriefing sessions arranged to followeach round of negotiations.

Over-responsibility for the reality of your character isa particular problem in role-playing past real-life

285Chapter 10

events, as players tend to think they should do as theirreal-life counterparts did; therefore, it is best tofictionalize past events, at least to the extent ofchanging characters’ names.

The degree of friction a character meets shoulddepend on its briefing, which tells it how importantcertain values are to it and which facts it sees asincontrovertible, and on the pressures brought to bearon it in interactions with others. Emotions engenderedin a role-player may enable it to overcome evenconsiderable friction to reframe the situation. Role-players should experience the fundamental paradoxof drama theory (i.e., the more I feel hemmed in byfixed constraints, the more emotion I feel, hence themore I am inspired to think of changes, so that themore unchangeable a situation feels, the morechangeable it becomes). The realism of thisexperience is enhanced if players first run a role-playunder instructions that they should stick strictly to theirbriefings. Having seen and discussed the result of this,they may then replay the scenario with instructionsgiving them more liberty to innovate.

Immersive Role-Playing as a Means of Prediction

Suppose an immersive role-play is based on accurateanalysis of a real-world situation. Then we study theforces at work in the real world by studying role-playerbehavior. This can yield predictions of a sort, as whenwe find that a character seems to have no alternativeother than to comply, or vice versa, when we find thatthere are no pressures on it to do so.

We cannot reliably get such predictions simply byrunning a role-play and observing the results. Any


prediction of this sort should be subjected to criticismand discussion aimed at overthrowing it. The role-play itself should be used to draw attention toquestionable assumptions.

The point is that a drama-theoretic prediction, ifbased on the presumption that there is only one wayto overcome a dilemma, such as by complying withour position, itself generates pressures for itsoverthrow because it motivates the characterconcerned to find some other way. The only purepredictions generated, even if our model andbriefings are accurate, are to the effect thatcharacters will face and attempt to eliminate certaindilemmas. Further predictions are obtained byasking: “What kinds of dilemma elimination are likelyto succeed?” Answering this question depends onestimating the friction players will meet in differentdirections, hence any useful predictions depend oncommon-sense estimates of friction.

How Immersive Role-Playing Helps

How, then, could immersive methods help acommander develop a confrontation strategy? Wesuggest in the following ways:

• Foreseeing likely reactions— Role-players’behavior can be used to foresee possible real-world reactions. Firm predictions of this kindshould be subjected to criticism of assumptionsbefore being accepted.

• Encouraging criticism of key assumptions—Dilemmas can inspire role-players not only to

287Chapter 10

change their briefings but to criticize theassumptions made in them.

• Asking key questions— Assumptions madeabout beliefs and preferences in the briefingsgiven to role-players are key to the developmentof a confrontation. Questioning them raises keyquestions for a commander to answer.

• Investigating assumptions— Assumptionsexposed to criticism can be investigated boththeoretically (i.e., asking “What dilemmas arestrengthened or weakened if this assumption isvaried?”) and in real life (i.e., intelligence can bedirected to answering specific questions aboutother parties’ beliefs or preferences).

• Understanding the implications of aconfrontation analysis— A complex analysiscan be hard to digest. Role-playing givesintuitive understanding.

• Understanding intuitively the gradient of amoment of truth —The term “gradient” is usedby analogy with a physical system, where it is anobject showing the directions a system is tendingto move. The drama-theoretic gradient of amoment of truth is an object that sums up itstendency to change as characters attempt toeliminate dilemmas. Intuitive understanding ofthe gradient of a confrontation is what acommander needs to orchestrate the entirecompass of beliefs, attitudes, and emotions to tiltthe gradient toward his position.


• Rehearsing real-world interactions— Weperform better in a situation after havingdeveloped skills and intuition through rehearsal.

• Enhancing unity of effort across differentcommands and levels— Joint role-playing bythose responsible for operations at differentcommands and levels can generate commonunderstanding of the confrontation, enhancingunity of effort.

• Appreciating the viewpoints of otherparticipants— An immersive briefing does nottake a single viewpoint, nor even a neutral one.It takes each viewpoint and its viewpoint oneach other’s viewpoint and…so forth; thus itprovides a conceptual platform from which tosurvey all the viewpoints.

Mediation Support

Unilateral Decision Support—General Considerations

Most of our discussion of how to win an OOTW hasassumed that the requirement is to support thedecisions of one character, ourselves, in confrontationwith others. In other words, we have been assumingthat our confrontation analysis is in unilateral mode.

What does this imply? First, that all information weuse is collected and kept by ourselves, except whenwe selectively decide to release it to others. Forexample, we may construct, for role-playing purposes,a version of each character’s viewpoint and itsviewpoint on others’ viewpoints, and so forth; however,all these viewpoints are based on our own information,

289Chapter 10

we do not go directly to other characters and ask fortheir viewpoints.

Second, unilateralism implies that our strategy issimilarly confidential. We tell others only those partsof it we wish them to know. They know the messageswe send them, or should know them. Messages sentby physical action, such as bombing or delivery of aid,should be accompanied by clear explanations lest theybe misinterpreted and send a message that was notintended. Such clear explanations do not amount to afull disclosure of our confrontation strategy.

Unilateral mode is the default mode for players thatneither belong to a common organization nor have anacknowledged common purpose. Each must thenhave a strategy for pursuing its ends and must sendmessages to others as part of its strategy. Its endsmay be selfish or benevolent, atomistic or communal,but they are its own ends.

However, we saw in chapter 3 that the very process ofconflict resolution between separate parties tends togenerate, through the emotions and arguments theydevelop in trying to influence each other, a commonpurpose and a common organization to support thatpurpose. We sometimes express this by the phrase,“every drama is a character.” This is how nations developand private revenge is replaced by public justice.

The development of law is only one example (althoughan extreme and important one) of a general processby which organizations develop formal and informalprocedures for conflict resolution. The essentialcharacteristic these procedures share is that they arepublic and open, dependent on every step taken in


them being common knowledge among the players,whereas the formation and implementation of aconfrontation strategy is essentially private. Eventelling other characters my confrontation strategy doesnot make it common knowledge because saying thatit is my strategy may be merely part of my strategy. Imay be lying.

By contrast, many structures and procedures of anorganization must be public or they defeat theirpurpose. It makes no sense to have a secret capitalcity or a secret constitution. Many of these publicorganizational procedures have an important conflict-resolution function. This is obviously true of arbitrationand appeal procedures. Command procedures resolveconflicts by specifying which officer has to make adecision when there are conflicting views.

Informal conflict-resolution procedures are more tricky.Suppose an office makes decisions by consensus,but a new recruit does not know that and expects themanager to decide on his own. This example illustratesboth the need for procedures to be commonknowledge if they are to be effective and the fact thatinformal procedures are free to change and developwith the players that operate them.

It is important to realize that each character, inchoosing how to go through the public steps of anopen conflict resolution process, is still pursuing itsown, private, intuitive confrontation strategy. (Here by“intuitive,” we mean “naturally developed, not derivedfrom knowledge of drama theory or confrontation-analysis methods.”) Realizing this, we see that theunilateral mode of implementing a private confrontationstrategy is not replaced by open institutions for conflict

291Chapter 10

resolution. It continues to be pursued within them. Thedifference they make is that each private strategist’schoice of cards to think up and play is constrained bythe rules of the institution it is playing within, becauseto break these rules would sometimes incurpunishment, at other times would be self-punishing indiminishing the impact of the common-interestarguments the strategist is using to make others acceptits position.

Thus, unilateral mode continues universally, even withinconflict-resolution procedures; however, it is often seenas a Machiavellian pursuit of selfish interests to thedetriment of the common purpose. Why?

We all have a common interest in strengthening theconflict-resolving institutions in our society andinfluencing them to develop in ways that advancecommon ends. Clear, intelligent development ofprivate confrontation strategies is thought to endangerthese institutions by showing how to take advantageof them.

A generally excellent book on how to negotiate, Fisherand Ury’s Getting to Yes (Fisher and Ury, 1983), headsits first chapter, “Don’t Bargain Over Positions.” Thiswould seem to contradict the whole of drama theory.On examination, it does not. To begin with, the authorsrecommend that each negotiator should solve itsdeterrence dilemma, if any (i.e., have a fallbackposition that puts pressure on the other). If the situationis not to escalate, it follows that they must solve theirinducement dilemmas by converging to a commonposition. The authors implicitly assume this. Given it,the essence of their advice is, “Don’t talk aboutpositions, use rational arguments in the common


interest.” Within the given structure, that is excellentdrama-theoretic advice. Not talking about positionsdoes not mean they do not exist. Rational argumentsin the common interest depend on the fact that foreach negotiator, even the other’s position is preferableto the threatened future, avoiding which is a priorityfor both. The arguments depend on finding thecommon interests responsible; hence, the argumentsimplicitly but clearly refer to positions.

In general, the animus against private strategies leadsto much hypocrisy. It is as if unilateral mode is seenas part of our sinful nature. It cannot be avoided ordenied, but must be discouraged by being starved ofintelligent development.

We suggest that this attitude is partly correct.Unintelligent, cynical expositions of how to form andimplement private strategies do undermine societaland organizational conflict-resolving institutions;however, drama theory shows how these veryinstitutions develop out of the pursuit of privatestrategies, as private strategists find a need to arguefor the common interest. Drama-theoreticconfrontation strategies should strengthen, notundermine, the institutions.

Meanwhile, the animus against private strategiesexists. It means that unilateral mode often will be seenas appropriate only when parties are not part of acommon organization or are not pursuing anacknowledged common purpose. Even then, it willseem appropriate only if we approve of the purposesour side is pursuing. Fortunately, these conditions aremet if we limit ourselves, as we have largely done, toconfrontations between our defense forces and

293Chapter 10

potential rebels against the New World Order. This isthe post-Cold War world described in chapter 1.

Mediation Mode

We now ask, “Can we use drama-theoretic insightsinto the private strategies pursued within open conflict-resolving procedures to improve the conflict-resolvingprocedures themselves?”

The answer is that we can. Confrontation mediation(i.e., doing confrontation analysis in mediation mode)involves doing a confrontation analysis that is seenby and is common knowledge for all characters. Theaim is not to help one side to win, but to help theprocess of conflict resolution itself. We want to moveit in a positive direction, toward a cooperative, happyending, rather than negatively, toward a tragic one.

The following question arises: Knowing that partiesare pursuing private strategies, how can we trustthe information they submit to an open, common-knowledge confrontation analysis? We will answerthis and other questions after outlining theprocedure, giving a simple example to make thediscussion concrete.

How Confrontation Mediation Is Done

Step One— Analysts go to each side and interviewthose involved. Views are solicited and carefully noted,without criticism. This is important, not only to build agood model but so that participants realize their viewshave been incorporated (i.e., the model is not biasedagainst them).


Step Two— The confrontation is analyzed using datafrom Step 1. A model is built. There are now two cases,as follows:

• The model shows full and harmoniousagreement. It is presented to the parties. If theyconfirm the agreement, the process ends;otherwise, it goes to Step 3.

• The model contains dilemmas. It is presented tothe parties, perhaps through immersive role-playing with fictional names substituted for actualones. In discussions, dilemmas are high-lightedand dilemma elimination encouraged throughpositive discussion and arguments in thecommon interest.

Step Three— Parties confer among themselves to seeif they can confirm with their internal subcharactersany changes in attitude or beliefs generated at Step2. The process then goes back to Step 1.

For example, suppose a memorandum ofunderstanding must be negotiated to provide forcooperation between two forces, ourselves andanother. Our team at the negotiations (call themrepresentatives) want to make commitments that thoseresponsible for fulfilling the commitments (call themsuppliers) see as extreme, but representatives seeas necessary. Each side, representatives andsuppliers, has a reasonable case, but sees the otheras unreasonable. Each side is interviewed.

Step One— In interviews, representatives giveexcellent reasons for the need for the commitment,suppliers for the uncertainty of being able to fulfill it.Neither side will budge.

295Chapter 10

Step Two— The simple model in table 25 (first threecolumns) is built and presented to the parties. After role-playing they converge to the compromise in column C.Representatives agree not to make the commitmentformally because suppliers are unsure they can fulfill it,while suppliers agree to make every effort to fulfill it,although they cannot guarantee success.

Step Three— Each side confirms its changed positioninternally.

Step One, round two— Interviews reveal that each sidemeans to keep the agreement, but distrusts the other.

Step Two, round two— The new model is presented.Each side’s trust dilemma is discussed and eliminated.

Step Three, round three— Newly acquired trust inthe other side is confirmed internally, reported onreturn to Step One, round three, and reconfirmedby each party to the other at Step Two, round three.End of process.

Harnessing Private Strategies to theMediation Process

Let us now answer the question, “How can model-builders in this process trust the information they aregiven in interviews, knowing that each party will betrying to influence the process in its favor?”

By choosing the right party to trust for each piece ofinformation, a mediation model can be madedeception-proof in the following sense: each party’sattempts to influence the process in its favor will helpthe process rather than hinder it.


To show how this works, we will use as an examplethe process by which mediation resolved the simpleconfrontation between representatives and suppliersin table 25. In the first model, constructed in the firstStep 2, each side has an inducement dilemma(preferring the other’s position to t) and a threatdilemma (preferring not to carry out its threat, if theother does).

Now observe that it is in neither side’s interest to admitthat it faces these dilemmas; therefore, if either of themrefuses to admit to a dilemma, we can point out thatwhat matters to it is not, actually, whether it has thedilemma, but whether the other believes that it has.Seeing this, it can be encouraged to produce rationalcommon-interest arguments and give evidence toconvince the other it does not have the dilemma. Itsuse of common-interest arguments in particular canbe encouraged by pointing out their greatereffectiveness in influencing the other’s beliefs, whichis what it needs to do.

If, on the other hand, either party cannot believe thatthe other is unconvinced, the purportedly unconvincedparty is shown that what matters to it, again, is notthat it is unconvinced, but that the other should believethat it is; therefore, it can be encouraged to use rationalcommon-interest arguments and produce evidence asto why it ought to remain unconvinced.

This general principle, of laying the burden ofproducing conviction on the one who needs to do so,and suggesting how it should be produced, will, iffollowed consistently, bring about convergence tocommon beliefs through the use of reason andevidence, passionately argued.

297Chapter 10

Table 25. Confrontation between representatives and suppliers.

Moreover, in this process of convergence, characterswill have made appeals to and explorations of theircommon interests. Their motivation will have beento convince each other, but the effect will be toconstruct a view of themselves as having interests incommon that will prompt them to think of win-winsolutions and compromises.

Continuing with our example, suppose thatrepresentatives and suppliers have found acompromise solution and in Step 2, round two, foundthat they cannot trust each other. Each is told that itsproblem is not whether it is trustworthy, but that theother should believe that it is. As before, it needs toproduce rational common-interest arguments andevidence to convince the other, and may in the processconvince itself, if it had in fact intended to defect, thatdefection is against its interests.

In this way, the process again brings aboutconvergence to common beliefs with accompanyingreframings of the situation that actually may change


perceived facts, as well as beliefs. The generalprinciple to be observed is that the unbelievability, forgood structural reasons, of a character’s evidence canbe used in the mediation process to persuade thatcharacter to provide better arguments and evidence,based on the parties’ common interests, simplybecause it is always other parties, not itself or themediator, that it needs to convince.

When Should It Be Used?

Theoretically confrontation mediation is usablewhenever parties are willing to participate; however, itmust be done by external experts, who require funding.The party that funds these experts naturally lays itselfopen to the suspicion that it is pursuing a privatestrategy. Such suspicions are certainly aroused byattempted mediation in many OOTW confrontations.

To overcome this problem in a fundamental manner,confrontation mediation, like established conflict-resolution procedures such as legal systems, needsto be funded by an organization to which all partiesbelong and owe loyalty.

This is because the above suspicion is justified. Thefunding organization must be pursuing a privatestrategy; we all are. The question is, with what aim?The only answer that will satisfy the parties is thegeneral, long-term, common interest of the partiesthemselves. To ensure this aim requires an institutionsuch as the law, that is owned by an organization (here,the nation) to which all parties belong, yet is keptseparate from its policy institutions (because policiesmay be what the confrontation is about).

299Chapter 10

Arrangements like these can be made within anorganization such as a firm or government agency. Amediation unit can be set up with guarantees ofindependence. It might even be done within a coalitionof defense forces assembled for a specific OOTW.Our argument is that unless it is done, confrontationmediation will be used at most sporadically.

Informal Use—Being Your Own Mediator

When formal mediation is impracticable, the ideas ofmediation may be of use to those implementing theirown, private confrontation strategies.

There is a sense in which I need to be my ownmediator. If I wish to avoid tragic outcomes, as bydefinition I should, then I want the process as a wholeto go in a positive direction, in addition to wanting it togo in my own particular direction. I can help this tohappen by observing myself and other parties inmediation mode and giving mediation-type advice tomyself and them as needed. Familiarity with theprinciples of mediation support is useful to all involvedin confrontations.

Another consideration is that confrontations vary alonga continuum of expected cooperativeness accordingto the degree to which cooperation is expected ordemanded of them. This is not a matter of howcharacters feel. Often parties in the most hopefullycooperative relationships have the strongest negativefeelings, just because they cannot cooperate. Viceversa, parties trying to kill each other may feel verylittle toward each other, or may even feel friendly.Emotions, as we have seen, have particular functions


at a moment of truth within a confrontation; they donot make confrontations cooperative or not.

That is brought about by organizational relationships.Parties are expected to cooperate when they are infunctional relationships within the same or relatedorganizations. Hostile armies confront each other non-cooperatively in our sense, whereas armies in acoalition, who may frustrate and infuriate each otheras much or more, do so even though theirconfrontations are hopefully cooperative.

It is this variable factor of expected cooperativenessthat seems to determine how strongly privateconfrontation strategies are deplored, regarded asdisruptive, and denied funding. Organizations deploreMachiavellian behavior within their ranks. We mighthope to use mediation support instead of unilateralconfrontation support to deal with such problems, exceptthat entrenched positions and angry attitudes oftenfound in hopefully cooperative confrontations may makeit hard to get participation. In such cases, being myown mediator may be the best solution available.


This chapter deals with two topics not coveredelsewhere in the book. Immersive briefing is a drama-theoretic tool for briefing role-players to enable themto take the parts of characters in a confrontation thathas been analyzed. Each role-player is told about itscharacter’s background, values, projects, relationshipswith other characters, and confrontations with others.In relation to each of its confrontations, it is told whothe characters involved are, what their positions and

301Chapter 10

fallback positions are, and which tactics they are using.Each character in a drama gets its own briefing,different from the others. A character may be role-played by a team or an individual.

Briefings are composed using a software tool anddistributed as software packages, through whichinformation is accessed by clicking and pointing. Afterbeing briefed, role-players interact with each other totry to resolve their problem. Interactions can be viavarious media, from e-mail to personal meetings.

Immersive role-playing can help a commander todevelop a confrontation strategy by enabling him atstep 2 to test out the likely effects of his strategy onother players. The results of such testing need to bethoroughly criticized and discussed.

Mediation support is contrasted with unilateral decisionsupport as an alternative mode of using confrontationanalysis. Instead of being used to develop a privatestrategy for one player, confrontation analysis in thismode is used to model a problem for the equal benefitof all players. Information is obtained from all anddistributed to all. The modeling process is used to helpthe conflict resolution process move toward a happyending rather than a tragic one.

It might seem problematic to rely on the informationsupplied by each participant in confrontationalmediation because each will want to give informationthat will bias the procedure toward its own ends;however, this is not a problem because this privatemotivation of each participant can be used to drivethe process forward. This is done by pointing out to aparticipant whose information is doubted that it is not


the truth of its information so much as its acceptanceby others that matters to the participant. In this way,the participant is encouraged to use rational argumentsin the common interest to convince others.

Mediation mode may be considered most appropriateto problems where parties belong to a commonorganization or have acknowledged common ends,whereas unilateral mode may seem right only whenthese conditions are not met; however, unilateral modealso encourages happy endings, if used intelligently.

B-1

Bibliography

Alberts, D.S., & R. Hayes. Command Arrangementsfor Peace Operations. Washington, DC:National Defense University Press. 1995.

Bennett, P.G. & Howard, N. “Rationality, Emotion andPreference Change: Drama-Theoretic Modelsof Choice.” European Journal of OperationalResearch. 92: 1996. 603–614.

Bennett, P.G. “Toward a Theory of Hypergames.”OMEGA 5(6): 1977. 749–751.

Bryant, J. “All the World’s a Stage: Using DramaTheory to Resolve Confrontations.” OR Insight.10(4): 1997. 14–21.

Bush, G. Speech by U.S. President in September 1990following Iraq’s invasion of Kuwait. 1990.

von Clausewitz, C. On War. Edited with introductionby A. Rapoport. London: Penguin Books. 1968;1st edition 1832.

Davies, N. Europe: A History. London: Pimlico. 1996.

DERA. Drama-Theoretic Tools for OperationalAnalysis. Report under contract number CDA/P/564. 1997.

Fisher, R. & W. Ury. Getting to Yes. London:Hutchinson. 1983.

B-2 Confrontation Analysis

Howard, N. Paradoxes of Rationality:Theory ofMetagames and Political Behavior. Cambridge,MA: M.I.T. Press. 1971.

Howard, N. “Drama Theory and Its Relation to GameTheory. Part 1:Dramatic Resolution vs. RationalSolution.” Group Decision and Negotiation. 3:1994.187–206.

Howard, N. “Drama Theory and Its Relation to GameTheory. Part 2: Formal Model of the ResolutionProcess.” Group Decision and Negotiation. 3:1994a. 207–235.

Howard, N. “Negotiation as Drama: How ‘Games’Become Dramatic.” International Negotiation.1: 1996. 125–152.

Howard, N. “What Is Drama Theory?” In ModellingInternational Conflict: Proceedings of aConference by the Institute of Mathematics andIts Application. Southend-on-Sea: IMA. 1997.

Howard, N. “Drama Theory: The FundamentalTheorems” in Modelling International Conflict:Proceedings of a Conference by the Instituteof Mathematics and its Application. Southend-on-Sea: IMA. 1997a.

Howard, N. “N-Person ‘Soft’ Games.” Journal ofOperational Research Society. 49: 1998. 144–150.

Howard, N., P. Bennett, J. Bryant, and M. Bradley.“Manifesto for a Theory of Drama and IrrationalChoice.” Journal of Operational ResearchSociety. 44(1): 1992. 99–103.

Joint Chiefs of Staff. Joint Vision 2010. 1997.

B-3Bibliography

Kahn, H. On Escalation: Metaphors and Scenarios. NewYork, Washington: Frederick A. Praeger. 1965.

Kuhn, T. S. The Structure of Scientific Revolutions.Chicago: University of Chicago Press. 1962.

Maxwell, D. “Adaptive Command and Control: FutureMission Statements in Peace Operations.”Proceedings of Third International Symposiumon Command and Control Research andTechnology, National Defense University,Washington, DC. 1997.

Cooperation—or Conflict (Newsletter). Birmingham,UK: Nigel Howard Systems. 1992–1997.

Osborne, M., and A. Rubinstein. A Course in GameTheory. Cambridge: MIT Press. 1994.

Owen, D. Balkan Odyssey. London: Indigo (CassellGroup). 1996.

Popper, K. The Logic of Scientific Discovery. London:Hutchinson. 1959, original German edition 1934.

Rapoport, A. Fights. Games and Debates. New York:Harper. 1960.

Rapoport, A. Strategy and Conscience. New York:Harper. 1964.

Rapaport, A. Preface to Clausewitz On War. London:Penguin Books. 1968.

Rogers, P. “War in the 21st Century: Likely Trends andTheir Possible Control.” In ModellingInternational Conflict: Proceedings of aConference by Institute of Mathematics and ItsApplications. Southend-on-Sea: Institute ofMathematics and its Applications. 1997.

B-4 Confrontation Analysis

Schelling, T.C. The Strategy of Conflict. Cambridge,MA: Harvard University Press. 1960.

Silber, L., and A. Little. The Death of Yugoslavia.London: Penguin Books (BBC Books). 1996.

Tait, A. “Drama Without Tears.” Tutorial Papers, YoungOR 10. Birmingham, UK: Operational ResearchSociety. 1998.

von Neumann, J., and O. Morgenstern. Theory of Gamesand Economic Behavior. Princeton, NJ: Universityof Princeton Press. 1959; 1st edition 1944.

Wentz, L.K. “C3I for Peace Operations: Lessons fromBosnia.” Proceedings of Third InternationalSymposium on Command and ControlResearch and Technology. Washington DC:National Defense University Press. 1997.

About the Author

Nigel Howard is Managing Director and Chairmanof ISCO Ltd and Visiting Professor at Sheffield

Business School, Sheffield Hallam University, UK. Hisearly work in game theory focused on the practical“dilemmas” created when one tries to take a purely“rational” approach to human interactions. His workhas remained practical, while remaining focused onthis same problem. His methods were applied, in thelate sixties, to the first SALT agreement, nuclearproliferation, and Vietnam; they have since been usedon many business and governmental policy problems.In recent years he has been the chief developer of anew approach—“drama theory.” He has taught at theLSE and the universities of Pennsylvania, Waterloo,Ottawa, and Aston. From 1987 to 1997 he publisheda bimonthly research letter, Cooperation or Conflict,now an Internet site (www.nhoward.demon.co.uk\drama.htm). From 1997 onwards he has applied atechnique derived from drama theory—ConfrontationAnalysis—to defense problems, in particular to PeaceOperations. Currently he is engaged in a project for aCommand and Control System for Confrontations. Inthe non-defense sector, he is involved in work on bothmergers and acquisitions, and also in designing a“virtual hospital” for medical personnel training.

3001071 Confrontation Analysis

Documents

Transcript of 3001071 Confrontation Analysis