How The Weak Win Wars

How the Weak Win Wars
How do the weak win wars? The likelihood of victory and defeat in
asymmetric conflicts depends on the interaction of the strategies weak
and strong actors use. Using statistical and in-depth historical analyses
of conflicts spanning two hundred years, Ivan Arreguı́n-Toft shows
that, independent of regime type and weapons technology, the inter-
action of similar strategic approaches favors strong actors, while oppo-
site strategic approaches favor the weak. This new approach to
understanding asymmetric conflicts allows us to makes sense of how
the United States was able to win its war in Afghanistan (2002) in a few
months, while the Soviet Union lost after a decade of brutal war
(1979—1989). Arreguı́n-Toft’s strategic interaction theory has implica-
tions not only for international relations theory, but for policymakers
grappling with interstate and civil wars, as well as terrorism.
I VA N A R R E G U Í N - T O F T
is Fellow in the International Security Program,
at the John F. Kennedy School of Government, Harvard University.
He has authored numerous conference papers and his articles have
appeared in International Security and the Cambridge Review of
International Affairs. He is a veteran of the US Army where he served in
Augsburg, Germany as a military intelligence analyst from 1985 to 1987.
CAMBRIDGE STUDIES IN INTERNATIONAL RELATIONS: 99

A Theory of Asymmetric Conflict
Editorial Board
Steve Smith (Managing editor)
Thomas Biersteker Phil Cerny Michael Cox
A.J.R. Groom Richard Higgott Kimberley Hutchings
Caroline Kennedy-Pipe Steve Lamy Michael Mastanduno
Louis Pauly Ngaire Woods
Cambridge Studies in International Relations is a joint initiative of

Cambridge University Press and the British International Studies
Association (BISA). The series will include a wide range of material,
from undergraduate textbooks and surveys to research-based mono-
graphs and collaborative volumes. The aim of the series is to publish
the best new scholarship in International Studies from Europe, North
America and the rest of the world.
CAMBRIDGE STUDIES IN INTERNATIONAL RELATIONS
100 Michael C. Williams

The Realist Tradition and the Limits of International Relations
99 Ivan Arreguı́n-Toft
A theory of asymmetric conflict
98 Michael Barnett and Raymond Duvall (eds.)
Power in Global Governance
97 Yale H. Ferguson and Richard W. Mansbach
Remapping Global Politics
History’s revenge and future shock
96 Christian Reus-Smit (ed.)
The Politics of International Law
95 Barry Buzan
From International to World Society?
English School theory and the social structure of globalisation
94 K. J. Holsti
Taming the Sovereigns
Institutional change in international politics
93 Bruce Cronin
Institutions for the Common Good
International protection regimes in international society
92 Paul Keal
European Conquest and the Rights of Indigenous Peoples
The moral backwardness of international society
91 Barry Buzan and Ole Wœver
Regions and Powers
The structure of international security
90 A. Claire Cutler
Private Power and Global Authority
Transnational merchant law in the global political economy
Series list continued after index

How the Weak Win Wars:
A Theory of Asymmetric Conflict
Ivan Arreguı́n-Toft
CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo
CAMBRIDGE UNIVERSITY PRESS
The Edinburgh Building, Cambridge CB2 2RU, UK
Published in the United States of America by Cambridge University Press,

New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521548694
# Ivan Arreguı́n-Toft 2005
This book is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.
First published 2005
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Arreguı́n-Toft, Ivan.
How the weak win wars: a theory of asymmetric conflict / Ivan Arreguı́n-Toft.
p. cm. — (Cambridge studies in international relations; 99)
Includes bibliographical references and index.
ISBN 0 521-83976-9 — ISBN 0-521-54869-1 (pbk.)
1. Asymmetric warfare — Case studies. 2. Military history, Modern — Case
studies. I. Title. II. Series.
U163.A776 2005
355.4’2 — dc 22 2004058131
ISBN-13 978-0-521-83976-1 hardback

ISBN-10 0-521-83976-9 hardback
ISBN-13 978-0-521-54869-4 paperback
ISBN-10 0-521-54869-1 paperback
Cambridge University Press has no responsibility for

the persistence or accuracy of URLs for external or
third-party internet websites referred to in this book,
and does not guarantee that any content on such
websites is, or will remain, accurate or appropriate.
To Monica
Epigraph
Do not press a desperate enemy
Sun Tzu
Contents
List of figures page x

Preface xi
Acknowledgments xiii
List of abbreviations xv
1 Introduction 1
2 Explaining asymmetric conflict outcomes 23
3 Russia in the Caucasus: the Murid War, 1830—1859 48
4 Britain in Orange Free State and Transvaal: the
South African War, 1899—1902 72
5 Italy in Ethiopia: the Italo-Ethiopian War, 1935—1940 109
6 The United States in Vietnam: the Vietnam War,
1965—1973 144
7 The USSR in Afghanistan: the Afghan Civil War,
1979—1989 169
8 Conclusion 200
Appendix 228
References 235
Index 243
ix
Figures
1 Percentage of asymmetric conflict victories by

type of actor, 1800—2003 page 3
2 Percentage of conflict victories by type of actor
over time 4
3 Expected effects of strategic interaction on
conflict outcomes 39
4 Strategic interaction and asymmetric conflict outcomes,
1800—2003 45
5 Strategic interaction and conflict outcomes when weak
actor received no external support 46
x
Preface
This is a book about power, and how common understandings about

power can lead to disaster.
The term ‘‘asymmetric conflict’’ is meant to bracket the broad topic of
inquiry in the fewest words and syllables — yet it suffers from a whiff of
academic conceit and ivory tower detachment.
The real topic at hand is naked brutality.
In war the primary recipients of this brutality should be soldiers.
They are trained to supply it, within limits; and they expect to be
injured or killed by other soldiers in the course of their duties. But
nowadays war’s brutality is less and less often restricted to soldiers
(some would say it is a myth that it ever was). It is perhaps an unin-
tended consequence of the attempt to use the Geneva Conventions (and
subsequent instruments of international humanitarian law) to protect
infants, the injured, the sick, the mentally ill, the crippled, small chil-
dren, women who do not bear arms, and the elderly, that it is precisely
these human beings, and not soldiers, who have increasingly become
targets of knives, rifle butts, flame, and flying metal. They are targets
because desperate men find it useful to shelter behind and among them,
while their enemies lack either the will or the ability to strike them
without also striking say, the nine-year-old girl huddled nearby.
In asymmetric conflicts — those in which one side is possessed of
overwhelming power with respect to its adversary — this is especially
true. It is true because the weak are desperate. It is true also because the
strong cannot abide the offense of resistance: if power demands obedi-
ence then resistance to overwhelming power supplies proof of evil or
madness; and neither the evil nor the mad need be treated as fellow
human beings.
xi
Preface
The real brutality of war is missing from most social science analyses
of war. It is missing because we are ignorant: most of us have never
directly experienced the horror whose analysis has become our life’s
work. It is missing because it is necessary: to get close to the reality of our
subject would be intolerable, unbearable. And some cruelties cannot be
described. There are simply no words in any language capable of
bearing the weight of their experience. Finally, the brutality of war is
missing from most social science analyses because it is useful: it allows
us to detect patterns and make generalizations that may someday
persuade others to alter how conflicts are resolved — to end those
ongoing and to prevent them from escalating to violence alltogether.
It is in this spirit I offer this analysis, flaws and all.
xii
Acknowledgments
I have read many of these acknowledgments sections over the years.

They almost always strike me as alternately maudlin and boring. Mine
will be no different.
I was trained at the University of Chicago in the last decade of the
twentieth century. It was a challenging process. One might call it benign
neglect; intended — designed even — to enable me to recognize, frame,
and answer good questions with little help beyond my own resources.
To two of my mentors then, John Mearsheimer and Stephen Walt, I owe
thanks for my intellectual self-sufficiency. Like an Army Ranger I can
now be dropped solo into an academic wilderness and I will always
find my way, survive, and fulfill my mission.
But the ability to work alone has not altered my inclination to
work with others, nor in any way dulled my love of teaching. I owe
thanks to more colleagues at the University of Chicago than I can
list here. But for good conversations, penetrating criticism, and unstint-
ing support I especially want to thank Ann Davies, Sharon Morris,
Jordan Seng, David Michel, Kim Germain, Brett Klopp, Andy Kydd,
Susan Liebell, John McCormick, Jen Mitzen, David Edelstein, and Paul
Kapur.
Thanks to Daniel Verdier, Charles Lipson, and especially Charles
Glaser, for insightful criticism on early versions of this work, for their
help in training me by example, and for their support.
I also owe thanks to three men who in my view count as towering
intellectuals: Peter Digeser, Bernard Manin, and Lloyd Rudolph.
Without their patient efforts and generosity I’d never have learned to
be circumspect about what it is I think I know.
For friendship and support vital to this intellectual enterprise I owe a
special debt of gratitude to Tom Reisz, Ty (and Lynn) Aponte, Queta
xiii
Acknowledgments
(and Ron) Bauer, and Marvin Zonis. Each in his or her own way helped
to make me a better man.
Here at Harvard’s Kennedy School of Government I have been
blessed also with outstanding criticism and support. I’d especially
like to thank Steve Miller, Sean Lynn-Jones, John Garofano, Arman
Grigorian, and Rose McDermott.
I also owe thanks to Andrew Mack, without whose originality, keen
criticism, advice, and support this book would never have been
possible.
Thanks also to the Institute for the Study of World Politics and Smith
Richardson Foundation for generous support.
Finally, I owe a debt to my families. My father Alfredo I thank for
reminding me that I do what I do because it’s who I am, rather than
because I have expectations of social science as a career. My mother
JoAnne I thank for her absolute love of ideas, as well as, of course, her
faith in me come what may. My stepmom Susie, and my sisters, Kristine
and Lesley, I thank for the inspiring passion with which they live their
lives. My uncle Paul I thank for his patience, and unflagging support
over these many years. My in-laws I thank for putting up with me; and
for forgiving me (now and then) for not having chosen to become a
lawyer.
Above all else and always I thank the three dearest to me in this
world, without whom even a life as rich as mine has been would not
have been much worth living: my wife Monica, my son Samuel, and my
daughter Ingrid Anne.
xiv
Abbreviations
ARVN Army of Vietnam

COIN Counterinsurgency
DMZ demilitarized zone
DRA Democratic Republic of Afghanistan
DRV Democratic Republic of Vietnam
FLN Front de Libération Nationale
GVN Government of Vietnam
GWS guerrilla warfare strategy
KLA Kosovo Liberation Army
NATO North Atlantic Treaty Organization
NVA North Vietnamese Army
SAM surface-to-air missile
VC Viet Cong
xv
1 Introduction
And there went out a champion out of the camp of the Philistines,
named Goliath . . .
And all the men of Israel, when they saw the man, fled from him,
and were sore afraid . . .
And David said to Saul, Let no man’s heart fail because of him; thy
servant will go and fight with this Philistine . . .
And Saul armed David with his armour, and he put an helmet of
brass upon his head; also he armed him with a coat of mail. And David
girded his sword upon his armour . . . And David said unto Saul, I
cannot go with these; for I have not proved them. And David put them
off him. And he took his staff in his hand, and chose him five smooth
stones out of the brook, and put them in a shepherd’s bag . . . and his
sling was in his hand: and he drew near to the Philistine.
. . . And the Philistine said to David, Come to me, and I will give thy
flesh unto the fowls of the air, and to the beasts of the field.
And David put his hand in his bag, and took thence a stone, and
slang it, and smote the Philistine in his forehead . . . and he fell upon
his face to the earth. So David prevailed over the Philistine with a sling
and with a stone . . . I. Samuel 17
Why do the strong lose to the weak?

Because we expect strong actors1 to defeat weak actors in con-
tests ranging from wars and fist fights to business competitions
1
‘‘Actors’’ in this context mean states or coalitions of states, although the same dynamics
would apply to governments fighting against rebels or rival national or ethnic groups in
civil wars. ‘‘Conflicts’’ in this analysis mean wars (1000 battle deaths per year), although
again, similar dynamics may apply in conflicts which are not wars, such as terrorism,
trade wars, and labor disputes. Because this analysis focuses on explaining asymmetric
conflicts I exclude those few wars in which the ratio of forces changed dramatically
(toward symmetry) between the start and end of a conflict.
1
and sports contests, the fact that the strong sometimes lose is
puzzling.2
Relative power and realist international relations

theory: the strong do as they will . . .
As far back as Thucydides’ description of the wars between Athens and
Sparta, the link between power and conflict outcomes has been the root
principle of realist international relations theory.3 More power means
winning wars, and less power means losing them. And defeat in war
means death or slavery. This is not the same thing as saying that either
international relations scholars, or political and military elites, imagine
that raw material power is the only thing that explains who wins or
loses a battle, campaign, or war. Many things — ranging from resolve,
technology, strategy, luck, leadership, and even heroism or cowardice —
can lead to unexpected outcomes. But power is useful. It is useful both
because in the real world enough of it can overwhelm deficiencies in the
other categories, and because in the theoretical world it is quantifiable,
measurable, and comparable in a way that luck or leadership, for
example, are not.
If it is true that power matters most, then in very asymmetric conflicts —
conflicts between actors with wide disparities in power — the strong
2
Power is one of the trickier concepts in international relations theory. Here I follow a
long tradition by introducing a quantifiable proxy for power which is an admittedly
imperfect one. By ‘‘strong,’’ for instance, I mean an actor whose relative material power
exceeds that of its adversary or adversaries by large ratio (see below). ‘‘Weak’’ and
‘‘strong’’ then only have meaning in a particular dyadic context: Italy in 1935 is weak
compared to the Soviet Union, but strong compared to Ethiopia. By ‘‘material power,’’ I
mean the product of a given state’s population and armed forces. Other quantifiable
proxies for state power have been proposed and used over the years; including iron and
steel production, gross national product (GNP), and so on. However, no single measure
appears to be sufficient on its own; and GNP, perhaps on balance the most useful, suffers
because these data were not kept prior to the 1920s. For a review and analysis of the
literature on empirical and quantifiable measures of relative power, see Nutter, (1994:
29—49). On the empirical measurement of power in asymmetric conflicts in particular, see
Paul (1994: 22).
3
My use of the term ‘‘realist IR theory’’ throughout this essay refers to a simple version of
realist theory that has two key tenets: (1) there is no authority above states capable of
regulating state interactions, and (2) all states have some capacity to harm other states. As
a result, states seek to increase their relative power by various means, including buying or
manufacturing armaments, and forming alliances. Power in this view is expected to have
a number of positive consequences for states that acquire it: it can deter other states from
attack, cow them into concessions, or defeat them in wars. For a cogent summary of realist
IR theory, see Mearsheimer (2001: chapters 1—3). On the limits of power in relation to
objectives, see Waltz (1979: 188—192).
2
Introduction
100 strong actors

80 71.5
weak actors
60
40 28.5
20
0
1800–2003, n = 200
Figure 1. Percentage of asymmetric conflict victories by type of actor,

1800—2003
should always win.4 Indeed, a review of all asymmetric wars fought

since 1800 supports this claim, as seen in Figure 1.
Again, ‘‘strong’’ and ‘‘weak’’ only have meaning in particular conflict
dyads; though as noted above these may include individual actors or
coalitions of actors. Moreover, in this analysis I’ve sharpened the puzzle
by making the strong much stronger than the weak. A literally asym-
metric conflict, for example, only requires a slight disparity — say, 1.1:1.
But in this analysis the aim is to test competing explanations against the
assertion that relative power matters most. For this reason the disparity
in power is raised to 10:1, and then adjusted to account for the fact that
strong actors — great powers and superpowers in particular — often have
other security commitments that constrain the application of all their
resources to a single conflict.5
That said, since 1816 strong actors have won more than twice as many
asymmetric conflicts as weak actors. On the other hand, since in this
analysis strong actors outpower weak actors by a large margin, it
remains puzzling that strong actors have lost such fights as often as
they have. What explains these unexpected outcomes?
A number of answers seem plausible besides bad luck. Perhaps the
strong actors lost because they were squeamish in some way. Perhaps
4
As Mearsheimer puts it, ‘‘There are definite limits to the utility of measuring force levels.
After all, even a cursory study of military history would show that it is impossible to
explain the outcome of many important military campaigns by simply comparing the
number of forces on each side. Nevertheless, it is clear that if one side has an over-
whelming advantage in forces, the glaring asymmetry is very likely to portend a decisive
victory’’ (Mearsheimer, 1983: 168). See also Mack (1975: 107).
5
‘‘Power’’ in this analysis is represented as the halved product of a strong actor’s armed
forces and population at the start of a conflict versus the simple product of the weak
actor’s armed forces and population. Data for this survey come primarily from Small and
Singer (1983), and from the 1992 revision of that data set. Additional data are from
Laqueur (1976); and from Ellis (1995). For a comprehensive list of the cases used in the
statistical analysis, see Appendix.
3
strong actor
100 88.2 79.5 weak actor
80 65.1
60 48.8 51.2
34.9
40 20.5
20 11.8
0
1800–49, n = 34 1850–99, n = 78 1900–49, n = 43 1950–99, n = 43
Figure 2. Percentage of conflict victories by type of actor over time
authoritarian strong actors win asymmetric conflicts but democratic

strong actors lose them. Perhaps they were irresolute, or poorly led.
Perhaps weak actors had come into possession of sophisticated military
technology of some sort,6 and this tilted the balance enough that strong
actors lost interest in victory when the costs of conflict and occupation
suddenly exceeded the expected benefits. Thinking more about it, and
since these explanations seem to resonate more or less with different
historical periods, it would be useful to know whether the distribution
of outcomes is consistent over time.
It isn’t. If the total record of asymmetric conflicts since 1816 is divided
into discrete time periods, a striking trend emerges: strong actors have
been losing asymmetric conflicts more and more over time (see Figure 2).
From 1800 until 1849, strong actors won 88.2 percent of all asym-
metric conflicts. That proportion dropped slightly to 79.5 percent in the
next fifty-year period. But starting in 1900, the number of asymmetric
conflicts won by strong actors began to fall off significantly, down to
65.1 percent through 1949. By the last fifty-year period — 1950 to 1999—
strong actors won only 48.8 percent of all asymmetric conflicts.7
Here then are two puzzles represented graphically. On the one hand,
realist international relations theory leads us to expect that in a two-
actor conflict, the larger the ratio of forces favors one actor the more
quickly and decisively that actor will win; and this is supported in
6
In this analysis, technology is presented as a power multiplier or divider, not an incre-
ment of power itself. Power is captured — crudely but sufficiently — by the multiplication of
population and armed forces. This leads to some distortions — e.g., nuclear weapons and
maritime vs. continental power distinctions do matter — but the impact of these distortions
is marginal on the overall analysis.
7
The colonial wars that distinguish this period are arguably a special case. But, if so, they
must be special in a way that overcomes the expected effects of relative power (i.e. they
still challenge realist IR theory’s primacy of power pillar). The overall trend is the same
whether the data are divided into fifty-, ten-, or five-year periods. The four-fold division
represented here is valuable analytically because it represents more data per period, and
because it presents the trend more clearly. Cases from the period 2000—2010 (Afghanistan
2002, and Iraq 2003) were not included because the period has not yet lasted five years.
4
Introduction
Figure 1. On the other hand, strong states have lost nearly 30 percent of all
conflicts in which they out-powered their adversaries by a factor of at
least 5:1. In addition, as shown in Figure 2, strong states have been losing
such fights more and more often over time. What explains this trend?
A good way to begin is to think about what was different in the early
nineteenth century that may have favored strong actors so dramatically
as compared to strong actors at the end of the twentieth century. Again,
a number of plausible explanations come to mind. Perhaps early strong
actors won because of their technological advantages: artillery, firearms,
and blue-water navies must have been tremendous force multipliers.
Perhaps the strong actor defeats concentrated in the last period were
due to the rise of national self-determination as a norm of interstate
politics? Nationalist resistance to European rule might have made con-
quest and occupation too costly. We might also observe that after World
War I and especially World War II, the number of authoritarian strong
actors declined. And, after 1991, the Soviet Union collapsed and ceased
to be an authoritarian actor in interstate politics. If authoritarian strong
actors fight asymmetric conflicts better than democratic strong actors,
perhaps the nature of the actor explains the trend.
This speculation points to an important utility of the trend observation
independent of the fact that until recently it remained unidentified: an
ideal theory of asymmetric conflict outcomes should be able to account
for both the fact of strong actor failure and the trend toward increasing
strong actor failure over time. Explanations of the trend which don’t
plausibly explain strong actor failure will be less persuasive than those
which can. Also, accounts of strong actor failure inconsistent with the
trend will be less persuasive than those which can explain both.
In sum, the problem with our international relations-theory-informed
expectations is that they appear to be only partly right, and only in
aggregate. When the aggregate data are divided into discrete time
periods, the expected correlation between power and victory becomes
significantly less useful as a guide to policy.
Competing explanations of asymmetric conflict

outcomes
As observed by Andrew J. R. Mack more than a quarter century ago
(Mack, 1975: 176), few international relations scholars have advanced
explanations focused specifically on the subject of asymmetric
5
conflict,8 and with the exception of my own work and Mack’s, none
has advanced a general explanation of asymmetric conflict outcomes
(Arreguı́n-Toft, 2001).
In this book I argue that although relative power matters, the inter-
action of the strategies actors use matters more than how much power
they have at the start of their conflict. This strategic interaction thesis
may seem intuitively plausible to readers, but before its explanatory
power can be evaluated competing explanations must be introduced
and explored. As noted above, a good way to think about these explan-
ations is in terms of how well they explain both the outcome and the
trend puzzles.
This section offers four competing explanations of strong actor failure
in asymmetric conflicts and the trend toward increasing strong actor
failure. Some, such as the arms diffusion explanation or democratic
social squeamishness argument, are strongest when explaining the
trend, but offer equivocal explanations of why strong actors lose.
Others, such as Mack’s interest asymmetry argument, are good at
explaining strong actor failure, but cannot account for the trend. A
sound general theory of asymmetric conflict outcomes should be able
to explain the fact of strong actor failure in a way consistent with the
observed trend, and vice versa. What follows is an introduction to these
competing explanations.9 In Chapter 2, the logic of each explanation
8
Chief among those Mack cites are Katzenbach (1962: 11—21); and Taber (1965). More
recently, T. V. Paul devoted a book to the question of why the weak start wars against the
strong. Paul’s threshold of analysis for asymmetry was a power ratio of 1:2, where power
was measured in traditional — that is to say, material — terms. On Paul’s definition of
asymmetry, see Paul (1994: 20).
9
These three hardly exhaust the possibilities, but they offer the strongest general explan-
ations of both unexpected outcomes and the trend. Three additional explanations
worthy of note are (1) social structure; (2) the rise of nationalism; and (3) Cold War
bipolarity. The social structure argument links key aspects of a given society’s social
structure to the military effectiveness of the forces it is capable of fielding. In this view,
the reason states such as the United States and USSR lost their respective fights against the
North Vietnamese and mujahideen is because their societies were not as efficient at
producing effective militaries as were the North Vietnamese and mujahideen. On the
importance of social structure, see Rosen (1995: 5—31). The rise of nationalism argument is
that the post-World War II period in particular was an era in which nations came to see
self-determination as a necessity which could not be put off because the structural
changes forced on them by colonial powers had begun to irreparably destroy their social
fabric. The logic of this argument is that nationalists are more stubborn and more willing
to risk death in pursuit of autonomy than people motivated by other concerns. See, e.g.,
Wolf (1973); and Anderson (1991: 9—10, 36). The Cold War bipolarity argument is that the
USSR and United States intervened more often in order to counter the perceived interests
of the other superpower. Each justified these interventions by means of domino logic,
6
Introduction
will be reduced to testable propositions which will later be compared in

the historical case studies — some formally and some informally —
against those derived from my own explanation.
The nature of the actor

One way to explain how a strong actor loses a war against a weak actor
is to argue that authoritarian actors fight wars better than democratic
actors.10 If true, then isolating the key differences between how each
type of actor fights could explain why strong actors sometimes win
asymmetric wars quickly, while other times they lose, or ‘‘win’’ at a cost
far out of proportion to their initial estimates. If authoritarian regimes
fight better than democratic regimes, and if fewer strong actors over
time have been authoritarian, then this explanation could explain both
the fact of strong actor failure (democratic strong actors lose asym-
metric conflicts) and the trend (fewer tough authoritarian actors mean
fewer strong actor victories).
Consider that authoritarian regimes are defined by several key char-
acteristics.11 First, authority for making domestic and foreign policy is
restricted to a single person or a small group of people. Second, author-
itarian regimes maintain strict control over the public’s access to infor-
mation about the consequences of domestic and foreign policies.
Finally, public attempts to criticize or change domestic or foreign policy
are often punishable by severe sanctions, such as death, torture, or
indefinite imprisonment.
In theory, these characteristics have four important implications for
warfighting effectiveness. First, authoritarian regimes should be able to
mobilize resources more effectively than democratic regimes because
they control the public’s perception of a war’s legitimacy, how it is
being fought by both sides, and the outcomes of specific battles. Second,
on the battlefield authoritarian regimes can coerce their soldiers to fight
which inflated even minor skirmishes into conflicts of vital importance in the struggle
between the West and the Socialist world. In other words, bipolarity implies the reduction
of all extra-core fights to mere proxy wars between the United States and USSR. This is
another way of saying that due to constant US and Soviet support and interference there
effectively were no asymmetric conflicts during the Cold War.
10
For a comprehensive introduction to the question of regime type and military effec-
tiveness, see Desch (2002: 5—47). For a pessimistic view of the military effectiveness of
democratic regime types at war, see, e.g., Waltz (1967); LeMay and Smith (1968: vii—viii);
Blaufarb and Tanham (1989); and especially Iklé (1991: xiv).
11
Mack discounts the argument that regime type matters in asymmetric conflicts. See
Mack (1975: 188—189).
7
by threatening to execute them for ‘‘cowardice,’’ where cowardice is

defined as any hesitation to engage the enemy regardless of objective
circumstances. Third, authoritarian soldiers may not be constrained by
the laws of war regarding noncombatants such as prisoners of war and
civilians in combat zones. Instead of having to provide captured enemy
soldiers with food and medical services, for example, an authoritarian
regime’s soldiers may be ordered to murder them, thus conserving
resources. Finally, lack of responsibility to a cost-bearing public may
allow authoritarian regimes to sustain higher combat casualties in pur-
suit of military objectives than can democratic regimes.
If an axiomatically just cause, just conduct, fearless soldiers, disre-
gard for the laws of war, and casualty insensitivity make authoritarian
regimes more effective in battle than democratic regimes, then the trend
toward increasing strong actor failure would have to be explained by
the fact of decreasing participation in asymmetric conflicts by author-
itarian strong actors. This seems intuitively persuasive: the progressive
dissolution of authoritarian states after 1918 (Austria-Hungary, the
Russian Empire, the Ottoman Empire), 1945 (the Third Reich, fascist
Italy, and imperial Japan) and 1991 (the Soviet Union) implies fewer
authoritarian strong actors involved in asymmetric conflicts.
Problems with the argument

But there are a number of problems with both the logic of the nature-of-
actor argument and the evidence used to support that logic. First, even
assuming that there are real differences in the capabilities of author-
itarian and democratic actors in war, why assume that regime types
stay constant in war? Most democratic actors become more conserva-
tive in war, restricting many of the civil liberties that define them as
democracies (one thinks most famously of the fate of Japanese
Americans in the United States after December 1941). Authoritarian
regimes may also become more liberal under stress: Stalin’s rehabilita-
tion of the Orthodox church during World War II is a classic example.
And the threat need not be existential to have this effect: the United
States passed the USA Patriot act — which allowed the state greater
freedom to pry into the private affairs of its citizens — in the wake of the
terrorist attacks of September 11, 2001.
Second, an authoritarian regime’s control over information about the
justness of a war only benefits a regime that wishes to fight a war its
own public would otherwise think unjust. The argument therefore
generally assumes that publics in authoritarian regimes would consider
8
Introduction
their government’s wars unjust, when there’s no way to know, a priori,

whether this is likely. Moreover, even assuming total resource mobili-
zation, the tight control of information necessary to achieve that effi-
ciency may impose severe limitations on the total amount of resources
available to mobilize in the first place: the command economies typical
of authoritarian regimes are notorious for being inefficient. This is the
unifying theme of a number of recent comparisons of authoritarian and
democratic regimes at war. According to this very sound logic, demo-
cratic states will, all other things being equal, be relatively more effect-
ive at fighting and winning wars than authoritarian states.12 The use of
extreme coercive measures to motivate soldiers is also logically suspect.
Regimes resorting to these methods may achieve a tactical advantage in
rare circumstances, but such a system can only motivate soldiers to take
actions they know to be excessively costly or futile. To soldiers as yet
uncommitted, this would signal that their leaders care nothing for their
lives, and thus ultimately decrease their combat effectiveness. Such
soldiers may desert at any opportunity, and cease to function without
the threat of harsh reprisals. They may even resort to murdering their
commanders. The systematic harm of noncombatants — a strategy
referred to here as barbarism — also has drawbacks. For one thing, it is
technically quite difficult to murder prisoners or noncombatants with-
out one’s opponents finding out about it. Once they do, they fight
harder, avoiding surrender at all costs (they also often start reprisals
in kind). Another problem is that sustained resort to barbarism ruins
soldiers for conventional missions:13 Here Richard Rhodes records the
complaint by Eastern Territories Commander Johannes Blaskowitz of
‘‘excesses’’ by Heinrich Himmler’s Einsatzgruppen in Poland in a memo
to the German High Command:
It is wholly misguided to slaughter a few ten thousand Jews and Poles
as is happening at the moment; for this will neither destroy the idea of
12
See Lake (1992); Reiter and Stam (1998: 259); and Reiter and Stam (2002). On the
contrary argument that regime type doesn’t matter much, see Desch (2002).
13
See, e.g., Stanislas Andreski, who argues that Latin America has had few interstate
wars because its states have soldiers who specialize in domestic oppression (torture,
murder, rape, and so on) (Andreski, 1968). The logic is the same: soldiers used for
barbarism will become ineffective as regular combat soldiers, putting states who face
threats from other states’ regular soldiers at increased risk. This logic is supported by a
study of military disintegration by Bruce Watson. Watson argues that when, for example,
a platoon of the US Army 20th Infantry Division under the command of Lieutenant
William Calley murdered civilians in the village of My Lai in 1968, they were not acting
as soldiers, but as a mob. See Watson (1997: 151).
9
a Polish state in the eyes of the mass of the population, nor do away
with the Jews . . .
The worst damage affecting Germans which has developed as a
result of the present conditions, however, is the tremendous brutal-
ization and moral depravity which is spreading rapidly among pre-
cious German manpower like an epidemic. (Rhodes, 2002: 9—10)
This is why authoritarian regimes who resort to this strategy almost

invariably develop ‘‘special forces’’ or ‘‘para’’ militaries to do the work.
Even within the Einsatzgruppen, Himmler worried that certain prac-
tices (e.g., the unauthorized murder of Jews by SS units tasked with
other functions) would damage discipline and hence, mission effective-
ness (Rhodes, 2002: 187). Finally, the ratio of acceptable combat casual-
ties to the value of military objectives should logically depend on the
relative populations of the combatants, not their regime types: a small
authoritarian regime could not win by selling its soldiers lives cheaply
for objectives that in any other regime type would matter little.
On the evidence side, the nature-of-actor argument is difficult to support
in its general form, because authoritarian regimes don’t win wars more
often than democratic regimes. This would appear to limit the power of
this explanation to account for unexpected asymmetric conflict outcomes.
On balance then, although the nature-of-actor argument may offer
important insights into key causes of strong actor failure, by itself it is
unlikely to stand as a sound general explanation of why strong actors
lose asymmetric conflicts.
Arms diffusion
A second explanation for strong actor failure begins by observing the
trend that accelerated after World War II. During the war, Allied and
Axis powers struggled to defeat each other in the developing world.
Throughout Asia and Africa, in particular, both sides shipped, stored,
and distributed relatively sophisticated small arms and ammunition,
including semi-automatic rifles, portable mortars, hand grenades, and
machine guns. After the war, these arms remained in the developing
world, along with a considerable number of indigenous soldiers expert
in their effective use.14 When colonial powers returned to their former
14
Examples include the Hukbalahap in the Philippines and the Malayan communists in
Malaya. For an account of the Hukbalahap insurgency in the Philippines from the
insurgent perspective, see Pomeroy (1964). For an account of the Malayan Emergency
and British responses to it, see Thompson (1966); Stubbs (1989); and Ramakrishna (2002).
10
Introduction
colonies after the war, they were often violently opposed by these
better-trained and armed soldiers. The net effect was to increase the
costs of conquest and occupation as compared to the expected benefits;
and strong actors lost wars against weak actors because they failed to
anticipate these higher costs.
Thus, the logic of the arms diffusion argument is that weapons
technology equals power, and as a result of this fact, ‘‘weak’’ actors
were not as weak as anticipated. As Eliot Cohen puts it:
The enormous increase in the quality and quantity of arms in the
hands of Third World nations, coupled with increased organizational
competence in the handling of such weapons, renders many local
conventional balances far more even than before. (Cohen, 1984: 162)
The empirical evidence appears to support this logic, because the

sharp rise in strong actor failures in asymmetric conflicts correlates
with the increase in small arms availability to developing countries
following World War II.

There are three problems with the logic of the arms diffusion explan-
ation. First, the argument assumes that the costs of conquest or occupa-
tion rose relative to the benefits, or that the benefits remained constant.
Second, the acquisition of a weapons technology in no way assures an
increase in combat effectiveness.15 Third, even assuming an increase in
combat effectiveness attributable solely to arms diffusion, any absolute
increase in effectiveness may be far outweighed by the relative technol-
ogy capabilities of strong actors.
The Cold War that began soon after the conclusion of World War II in
Europe altered states’ perceptions of the value or benefits of occupation
in the developing world. Especially following the invasion of South
Korea by the North in 1950, a new understanding of the linkage of
values came to dominate state calculations of power and interest. In the
context of a new US strategy of communist containment, domino logic
could transform an intrinsically low-value country — such as the then
15
For a recent and comprehensive treatment of the relationship between arms, arms
transfers, and conflict outcomes, see Craft (1999: 92—93), and especially p. 121. Craft
briskly summarizes hypotheses on the proposed relationships between arms and out-
comes, and undertakes a sophisticated statistical analysis of them. Craft concludes that
‘‘Arms transfers that take place during a war do not predict war outcomes (winners,
duration, or casualties) to commonly accepted statistical significance’’ (Craft, 1999: 121).
11
South Korea — into a vital US security interest. The logic was the same
for the Soviet Union. Thus, even if the absolute costs of conquest,
occupation, or stabilization of distant countries in the developing
world were rising, so were the perceived benefits.
There is no doubt that weapons technology can increase a military’s
combat effectiveness. But logically, it could also decrease it. Many key
military-technical innovations — ranging from the machine gun to the
tank — decreased combat effectiveness until the proper mix of strategy
and tactics for their effective use was established.16 This is why, in his
remarks above, Cohen is careful to specify ‘‘that organizational compet-
ence in the handling of such weapons’’ is a key component of any
expectation of increased military effectiveness. Moreover, there is no
reason to assume that weapons systems that are highly integrated
into the doctrine, training, tactics, and strategy of an industrial power
can be assimilated into militaries working within different historical,
geographical, or social contexts.17 One risk in particular is that the
acquisition of unassimilated military technology may tempt a pre-
mature or inappropriate shift in strategy, with disastrous results.18
Finally, in ‘‘Command of the Commons,’’ Barry Posen shows that
even the most efficient integration of doctrine, technology, and tactics
may produce counterproductive specialization: the United States — and
the United States is not alone among advanced industrial powers in this
regard — is particularly good at leveraging firepower and maneuver
16
This may explain the striking empirical finding that increased availability of weapons
to an actor increases its likelihood of defeat in war (Craft, 1999: 73).
17
Mearsheimer, for example, notes that his theory only applies to large-scale armored
warfare, and hence only to states whose geography favors such warfare (there is no
attempt to assess the distribution of such territory as a percentage of a total) (see
Mearsheimer, 1983: 15). Thus, the same technology (and associated doctrine) which
allows states to achieve decisive results in one geographical setting, may prove useless
in another (on this point, see also Posen, 2003). Cultural factors may also constrain the
adoption of non-indigenous technologies. Rosen notes that after World War II, Japanese
businessmen attempted to duplicate the success of US industry by adopting a host of US
business practices, only to abandon them due to cultural incompatibilities (see Rosen,
1995: 16). Finally, Chris Parker argues that not all weak states are equally successful at
assimilating nonindigenous military technology. Insofar as assimilation and military
effectiveness are the same thing (or at least covary), then anything which hampers
assimilation should hinder effectiveness (Parker, 1999).
18
Usually, actors seek the technology to facilitate their adopted strategy. When the
reverse occurs, disaster can result (see, e.g. Johnson, 1973: 50; and Karnow, 1983: 182,
610). Perhaps this century’s most famous philosopher of guerrilla warfare, Mao Tse-tung,
warned against both premature escalation to conventional confrontation, and its opposite,
something he came to call ‘‘guerrillaism’’ (staying on the strategic defensive too long) (see
Hamilton, 1998: 28; and Paret and Shy, 1962: 35).
12
Introduction
advantages in open terrain, but less effective in closed, and, in particular,

urban terrain (Posen, 2003).
But even assuming that arms diffusion increased the absolute cap-
abilities of armed forces in the developing world, it is hard to argue a
relative increase over the advances maintained by industrial powers.
Weak actors, for example, may have gained reliable compact automatic
weapons, but strong actors gained combat helicopters, night vision, and
better communications capability.
In sum, arms diffusion offers an at best equivocal explanation of the
trend toward increasing strong actor failures after World War II. It is
particularly weak when relied upon to explain unexpected outcomes.
By itself, then, arms diffusion will probably not constitute a sound
general explanation of asymmetric conflict outcomes.
Interest asymmetry
A third explanation for strong actor failure in asymmetric conflicts is
Andrew J. R. Mack’s. Mack’s explanation for how weak states win wars
has three key elements: (1) relative power explains relative interests; (2)
relative interests explain relative political vulnerability; and (3) relative
vulnerability explains why strong actors lose. According to the logic of
this argument, strong actors have a lower interest in a fight’s outcome
because their survival is not at stake, whereas weak actors have a high
interest in a fight’s outcome because their survival is at stake (Snyder
and Diesing, 1977: 190; Mack, 1975: 181). Mack introduces the import-
ant concept of political vulnerability to describe the likelihood that an
actor’s people (in democratic regimes) or competing elites (in author-
itarian regimes) will force an actor’s leaders to halt the war short of
achieving its initial objectives (Mack, 1975: 180—182). A strong actor’s
low interests imply high political vulnerability, and a weak actor’s high
interests imply low political vulnerability. Mack argues that this polit-
ical vulnerability explains why the strong lose to the weak (Mack, 1975:
194—195): delays and reverses on the battlefield will eventually encour-
age war weary publics or greedy elites to force the strong actor’s leaders
to abandon the fight. Mack’s argument therefore reduces to the claim
that relative power explains strong actor defeat in asymmetric wars:
power asymmetry determines interest asymmetry (high power equals
low interest), which varies inversely with political vulnerability (low
interest equals high vulnerability), which varies inversely with out-
comes (high vulnerability equals low probability of victory). Interest
13
asymmetry is the key causal mechanism and Mack’s argument is in this

sense an interest asymmetry thesis.
Mack applies this logic to the case of US intervention in Vietnam,
where it appears to provide a strong explanation of that war’s unex-
pected outcome. According to Mack, the United States lost the war
because it had less to lose than North Vietnam. Over time, the United
States failed to coerce North Vietnam and was eventually forced by an
angry and frustrated American public to withdraw short of achieving
its main political objective: a viable, independent, noncommunist South
Vietnam.

Mack’s interest asymmetry thesis has at least three weaknesses. First,
relative power is a poor predictor of relative interest or resolve in peace
or war. In peacetime, a strong state may act as if its survival is at stake
when it is not. A state whose leaders and citizens imagine it to be
‘‘leader of the free world,’’ for example, might rationally calculate that
although the defeat of an ally in a distant civil war would be materially
insignificant, its own survival as free world leader depended on a
favorable outcome. These calculations are often intensified by domino
logic, in which a series of individually insignificant interests are linked
so that their cumulative loss constitutes a material threat to survival.
Both identity survival and domino rationales, for example, affected the
US decision to intervene in the civil war in Vietnam.19 In war, once
strong actors get involved in a fight — even one acknowledged to have
been initially peripheral to their interests — their resolve to win may
increase dramatically. This was as true of Soviet calculations in
Afghanistan as of US calculations in Vietnam.20
Second, the operation of political vulnerability, which Mack uses to
explain weak actor success, presupposes a span of time. Mack’s argu-
ment assumes rather than explains a weak actor’s capacity to avoid
defeat and impose costs on its stronger adversary. This leaves us
wondering why some asymmetric conflicts are over quickly yet others
drag on.21
19
See, for example, Herring (1986: 70); and Karnow (1983: 169—170, 377—378, 399, 423).
20
On Soviet calculations in Afghanistan, see Magnus and Naby (1998: 68, 122). For apt
examples from the US intervention in Vietnam, see Herring (1986: 222).
21
Mack recognizes this weakness and suggests that guerrilla warfare strategy explains
the longer duration of asymmetric conflicts (Mack, 1975: 195). But this argument suffers
14
Introduction
Third, if the interest asymmetry thesis is right there should be little or

no variation over time in the distribution of asymmetric conflict out-
comes when relative power is held constant. But, as shown in Figure 2,
in which relative power is held constant, weak actors are increasingly
winning asymmetric conflicts.
In sum, Mack’s interest asymmetry thesis is not so much wrong as
incomplete. It is weakest when explaining actor interests as a function
of their relative power and strongest when explaining strong actor
failure as a consequence of political vulnerability.
Democratic social squeamishness

A variant of the nature-of-actor argument worthy of brief discussion on
its own is that in one particular kind of asymmetric conflict, small wars,
regime type matters so much as to constitute a strong general explana-
tion of both the trend and a strong actor’s defeat. In How Democracies
Lose Small Wars, Gil Merom argues that essentially, democratic strong
actors can’t win small wars because the state is constrained by society to
avoid the sacrifice — both in the form of sustained and inflicted casual-
ties — necessary to win:
My argument is that democracies fail in small wars because they find it
extremely difficult to escalate the level of violence and brutality to that
which can secure victory. They are restricted by their domestic struc-
ture, and in particular by the creed of some of their most articulate
citizens and the opportunities their institutional makeup presents
such citizens. Other states are not prone to lose small wars, and
when they do fail in such wars it is mostly for realist reasons.
(Merom, 2003: 15)
Merom adds that ‘‘Essentially, what prevents modern democracies

from winning small wars is disagreement between state and society
over expedient and moral issues that concern human life and dignity,’’
and ‘‘Achieving a certain balance between . . . the readiness to bear the
cost of a war and the readiness to exact a painful toll from others — is a
precondition for succeeding in war’’ (Merom, 2003: 19).
The argument reduces to the claim that, in small wars, insensitivity to
friendly casualties and a willingness to maximize violence against an
opponent are necessary for victory; and that modern democracies are
from two related problems: (1) weak actors do not always defend with guerrilla warfare
(this limits the generality of Mack’s theory); and (2) some defenders using guerrilla
warfare strategy are defeated quickly (this limits his theory’s explanatory power).
15
structurally constrained in both categories. The particular relationship

between society and the state makes democratic strong actors too
squeamish to win small wars.

As with the interest asymmetry thesis this one is both carefully rea-
soned and well supported by historical case evidence. But, overall, it
suffers from most of the same problems as the nature-of-actor argu-
ment. There are three additional problems.
First, like Mack, Merom restricts his explanatory universe. Mack’s
analysis restricts itself to asymmetric conflicts between actors in which
the weak actor invariably applies a guerrilla warfare strategy. Merom
goes one further: asymmetric conflicts in which the strong actor is
invariably a ‘‘democracy’’ and the weak actor invariably uses a guerrilla
warfare strategy. Merom also follows Mack in claiming that, for the
democratic strong actor, small wars are axiomatically ‘‘not existential’’
(Merom, 2003: 21). This may be objectively true, especially in hindsight,
but one can hardly think of an example of a democratic state launching
a small war it didn’t claim (and its leaders and citizens did not in some
real measure believe) was of vital importance to its survival; albeit via
domino logic.
This in itself is not all that damning for either thesis, because both
kinds of asymmetric conflict continue to remain important foreign
policy problems for both strong and weak actors. But for the democratic
social squeamishness argument, a second problem is that we have
important examples of democratic strong actors initiating brutal stra-
tegies and still losing small wars (e.g., the French in Indochina, 1946—54;
and the United States in Vietnam, 1965—75); as well as authoritarian
regimes doing the same thing and losing (e.g., the Nazis in Yugoslavia,
1941—43; and the Soviet Union in Afghanistan, 1979—89).
The authoritarian examples call our attention to a third problem with
Merom’s argument. It implies that the mechanism of success in small
wars reduces to a willingness to deliberately and systematically target
enemy noncombatants.22 Again, the record of success in the use of this
strategy is mixed at best. Merom goes wrong by assuming that because
22
It also places too much causal weight for outcomes on the shoulders of strong actors.
What if the claim by the weak that they are willing to resist to the death is more than
simply propaganda? To the extent that such resolve — most famously represented by the
North Vietnamese during their fights with France and the United States in Indochina — is
16
Introduction
the strong actor holds additional brutality in reserve, its conduct on the
battlefield may still be described as restrained, and when the strong
actor later abandons the field that ‘‘restraint’’ becomes the causal
focus.23 There is something to this argument: my own research suggests
that in the post-World War II period barbarism was a sound strategy for
winning a small war yet losing the subsequent peace. But Merom seems
to rule out the possibility that democratic strong states may hold brutal-
ity in reserve — unused — not because their societies constrain them but
because they sense, rightly, that the utility of further brutality may be
either marginal or even negative.24 Moreover, democratic states facing
insurgencies have another option besides escalating brutality (barbar-
ism) — an option which historically has been both successful and rare:
conciliation. Britain experienced success with this option in the Malayan
Emergency of 1948, and the United States successfully supported Ramon
Magsaysay’s reform efforts to overcome the Hukbalahap in the Philippines
in 1952.
A final problem with the democratic social squeamishness argument
is that it is not well tested. Merom’s analysis skillfully brings together
political and game theory threads to explain what Mack, for example,
assumes: democratic strong actors are politically vulnerable to setbacks
on the battlefield. But Merom’s analysis ignores the success and failure
rates of authoritarian states — states less constrained by (1) dependence
of the state on society for military forces; (2) a general preference for
measures short of war by society; or (3) by the ability of society, through
real, this places strong actors in the position of either having to resort to genocide, or
withdrawing. It makes more sense in such cases to look at least as carefully at factors on
the ‘‘weaker’’ side as on the stronger.
23
Merom is hardly alone here. A recent example focusing on the case of US intervention
in Vietnam is C. Dale Watson’s The Myth of Inevitable US Defeat in Vietnam, in which
Watson argues that had the United States been less restrained it could have won (Walton,
2002: 5). The problem with this argument, and a weakness in Merom’s, is that it deliber-
ately ignores actor restraints that had nothing to do with societal pressure (e.g., ‘‘let’s do all
we can without starting another world war’’).
24
This argument goes right back to the first of the modern-era small wars — indeed, the
war from which the term itself, guerrilla, originated. Napoleon Bonaparte’s forces in Spain
were harassed and attacked at all turns by Spanish guerrillas from 1807 to 1814, and they
responded to these attacks with extremely brutal reprisals, including torture, mass mur-
der, rape, and wholesale property destruction. These only appeared to stimulate even
further resistance, creating a spiral of reprisal that did not end until the French left Spain in
1814. So clear was the connection between barbarism and increased social resistance that
before pursuing the French into France, Britain’s Wellington laid down the strictest
restrictions on the conduct of his own troops in French towns (see Fremont-Barnes,
2003: 53—58, 67, 90).
17
charismatic representatives, to alter the state’s foreign policy or military

strategy.
These problems aside, Merom’s account of the political vulnerability
of democratic small actors in small wars is both excellent and timely. It
expands on Mack’s description of why and how democratic strong
actors are politically vulnerable, and offers a useful explanation of
why democratic states choose the strategies they do, and why they
stick to failing strategies even after they’ve received a clear indication
that their strategy for victory is in fact failing.
Strategic interaction
My own explanation for weak actor success in asymmetric conflicts is
more general. I argue that although relative power, the nature of the
actor, arms diffusion, and interest asymmetry all matter, the best pre-
dictor of asymmetric conflict outcomes is strategic interaction. According
to this thesis, the interaction of the strategies actors use during a conflict
predicts the outcome of that conflict better than competing explan-
ations. If we think of strategies as complex but discrete plans of action
which include assumptions about the values of objectives, as well as
tactical and leadership principles and rules of engagement, different
interactions should yield systematically different outcomes independ-
ent of the relative power of the actors involved.
In Chapter 2, I argue that for purposes of theory building the universe
of real actor strategies — blitzkrieg, attrition, defense in depth, guerrilla
warfare, terrorism, and so on — can be reduced to two ideal-type strategic
approaches: direct and indirect. My central thesis is that when actors
employ similar strategic approaches (direct—direct or indirect—indirect)
relative power explains the outcome: strong actors will win quickly
and decisively. When actors employ opposite strategic approaches
(direct—indirect or indirect—direct), weak actors are much more likely to
win, even when everything we think we know about power says they shouldn’t.
My explanation of strategic choice and the trend focuses on Kenneth
Waltz’s concept of state socialization: the idea that actors imitate the
successful policies and strategies of other actors, while avoiding failed
policies or strategies (Waltz, 1979: 127). I argue that socialization of this
kind works in regions, and that after World War II the developed and
developing world imitated diametrically different patterns of success.
When these patterns came to interact systematically, as they did after
World War II, weak actors won more often.
18
Introduction
Why study asymmetric conflict outcomes?

Resolving the puzzle of how the weak win wars is important for several
reasons. First, in terms of US foreign policy in the coming decades,
many of the problems of asymmetric conflict — such as catastrophic
terrorism, and the necessity, legitimacy, and costs of military interven-
tion in ethnic conflicts and civil wars — will remain vital concerns;
particularly because for the foreseeable future the United States will
be a strong actor by default relative to any of its potential adversaries.
Recent wars in Iraq, Kosovo, Afghanistan, and again, Iraq, are illus-
trative in this regard. All were asymmetric conflicts in which the United
States, often leading a coalition of states, attacked a much weaker state or
non-state adversary. In the first Gulf War (1990), the US-led coalition
launched a conventional assault against an Iraqi military bent on defend-
ing Kuwait (and later Iraq) with a conventional defense. Iraq was quickly
overwhelmed and, having achieved its stated political objective of forc-
ing Iraq from Kuwait, coalition forces declared victory and left.
In Kosovo in 1999, the United States, under the auspices of NATO,
attacked Serb-supported forces in Kosovo from the air, hoping to force a
halt to well-documented depredations against ethnic Albanians there.
This clumsy conventional attack strategy was not directly opposed by
Serb fighter interceptors or surface-to-air missile batteries; and it did not
prevent Serb-supported forces achieving their objective of depopulating
Kosovo of ethnic Albanians and destroying the Kosovo Liberation Army
(KLA) as a military force. The three most remarkable features of this war
were (1) that NATO was able to justify intervention even though none of
its member states was directly threatened by Serb depredations in
Kosovo — in essence, NATO violated international law in order to uphold
a humanitarian principle; (2) the Albanian refugees came back; and (3)
many observers counted operation Allied Force as a resounding success
for the independent effectiveness of air power.
In Afghanistan in 2002, US forces attacked the Taliban with a brilliant
strategy of providing close-air, logistical, and artillery support to the
Taliban’s last surviving domestic adversaries, the Northern Alliance.
Essentially, the war pitted the Northern Alliance in a conventional
attack against a Taliban stuck with a conventional defense. Most analysts
predicted a repeat of the Soviet experience of 1979, including failure. But
having spent nearly a decade abusing much of Afghanistan’s population
and bribing potential adversaries with drug money, the Taliban was in
no position to launch a popularly based guerrilla defense such as had
19
faced the Soviets in the 1980s. And the US decision not to commit major
combat forces prevented the Taliban from rallying religiously conserva-
tive Afghanis to the cause of ejecting non-Muslims from Afghanistan.
The Taliban lost the war and decamped to Northwestern Pakistan, where
today they strike at vulnerable US and government forces and wait and
train to take control of Afghanistan once US-supported forces weaken
there.
Finally, the second US-led assault on Iraq in 2003 in key ways resem-
bles the Boer War of 1899—1902. Like that war, the world’s pre-eminent
military power engaged another army in a conventional campaign that,
after a few setbacks, quickly ended in an overwhelming victory. But,
also like the Boer War, the losers refused to surrender, and instead
switched from a conventional to a guerrilla warfare defense strategy.
The result has been increasing difficulty for US and Allied forces. These
forces were never designed, trained, or equipped for extended occupa-
tion duties. They continue to suffer daily casualties from unseen ene-
mies and may, as a result, be increasingly tempted to inflict reprisals on
noncombatants. This book explains both the conditions under which
the US can win such fights and why it will certainly lose its fight to build
democracies in Afghanistan and Iraq so long as it pursues its current
strategy (excessively militarized and with the wrong mix of armed
forces).
Second, although asymmetric conflicts are the most common type of
conflict they are among the least studied by international relations
scholars.25 If there are conflict dynamics unique to asymmetric con-
flicts, or if an analytical focus on asymmetric conflicts enables us to
attain valuable insights into more symmetrical conflicts, then a general
explanation of asymmetric conflict outcomes is not only desirable, but
necessary, both to reduce the likelihood of unnecessary conflicts and to
increase the likelihood of US success when relying on the force of arms
to advance its political objectives.
25
See Paul (1994: 4); and Mack (1975: 176). My own survey of all wars from 1816 to the
present reveals the following conflict type distribution: asymmetric conflicts = 52 (14
percent), probable asymmetric conflicts = 141 (37 percent), symmetric conflicts = 28 (7
percent), and missing data = 156 (41 percent). The total number of recorded wars since
1800 was 377. Probable asymmetric conflicts include Singer and Small’s ‘‘extra-systemic
wars’’ (colonial wars numbered 300—454). Missing data include most of Singer & Small’s
‘‘civil wars’’ (numbered 600—982). If probable asymmetric conflicts are added to actual
asymmetric conflicts, they comprise 51 percent of the distribution of all types of war since
1800. Even without adding probable asymmetric conflicts, however, actual asymmetric
conflicts are twice as common as symmetrical conflicts.
20
Introduction
Third, the whole question of how the weak win is in itself fascinat-
ing. The contests between David and Goliath, Hannibal and Rome at
Cannae, Henry V and the French aristocracy at Agincourt, Germany’s
blitzkrieg into the Soviet Union in June of 1941, and even the World
Heavyweight championship fight between Muhammad Ali and
George Foreman in Zaire, are remembered chiefly because their out-
comes were so unexpected. More recently, we’ve witnessed the unex-
pected defeats of US elite forces in Somalia, Russian Federation forces
in Chechnia, Israeli forces in Lebanon, and the Taliban in Afghanistan.
Explaining how and why these unexpected outcomes happen is itself
a worthwhile endeavor. It is an effort that can result in important
additions to our understanding of power in international relations
theory, as well as a guide to US policymakers struggling to respond
to global terrorism.
Finally, a theory of asymmetric conflict outcomes could help us
understand how and why weak states respond to strong and strength-
ening states in the international system (Elman, 1995; Walt, 2002; Nye,
2004). It’s an important question, because if the United States — the
world’s current pre-eminent power — does not act wisely and with
restraint, it could well provoke countervailing alliances that eventually
overwhelm it or undo its previous economic, political, and military
advantages.
For all these reasons developing a theory of asymmetric conflict out-
comes is vital.
Plan of the book

The remainder of the book is organized as follows.Chapter 2 introduces
the strategic interaction thesis more fully. It defines key terms, reduces
competing explanations to testable propositions, and reports the results
of a statistical analysis of all asymmetric wars fought between 1816 and
1998. This statistical analysis supports the strategic interaction thesis.
Chapters 3—7 contain the book’s historical case studies of asymmetric
conflict. These five case studies were chosen for several reasons. First,
each case study is drawn from one of the four major historical periods of
analysis: 1816—49, 1850—99, 1900—49, and 1950—99. This helps us track
outcomes across large historical trends, and lends a better intuitive fit
between the raw statistical data and the facts of each conflict. Second,
the cases include variation on key causal variables, such as regime type,
arms diffusion, and strategic interaction. Finally, each of the cases
21
contains multiple strategic interactions. Thus, although only five his-

torical cases are examined, together they provide a total of sixteen tests
of the strategic interaction thesis.
Chapter 3 examines the fight between the Russian Empire and the
Murids in the Caucasus from 1830 to 1859. The Russians won this
twenty-nine-year conflict, which involved three distinct strategic inter-
actions. Chapter 4 examines the war between the British Empire and the
allied Boer republics of Orange Free State and Transvaal from 1899 to
1902. The British won this three-year war, which was the most costly of
their entire colonial history. Chapter 5 examines the conflict between
fascist Italy and Ethiopia from 1935 to 1940. The Italians used mustard
gas in order to win this war against the vastly outgunned and out-
numbered Ethiopians. Chapter 6 examines the fight between the Soviet
Union and Afghan mujahideen from 1979 to 1989. The Soviets lost this
brutal decade-long war which either killed or made refugees of fully
half the Afghan civilian population; and they laid the foundation for the
accession of one of the most conservative religious movements in
history: the Taliban. Chapter 7 analyzes one of the most complex and
important conflicts of the post-World War II era: US intervention in the
Vietnamese Civil War. Chapter 8 contains a summary of the book’s
arguments, its conclusions, and its relevance to international relations
theory and foreign policy (in particular, US foreign policy).
22
2 Explaining asymmetric conflict
outcomes
I’m a speed demon, I’m a brain fighter, I’m scientific, I’m artistic, I plan
my strategy. He’s a bull, I’m a matador . . .
Muhammad Ali, Zaire (1974) in Gast (1996)
[T]aking practical account of the area we wished to deliver . . . I began

idly to calculate how many square miles: sixty: eighty: one hundred:
perhaps one hundred and forty thousand square miles. And how
would the Turk defend all that? No doubt by a trench line across the
bottom, if we came like an army with banners; but suppose we were
(as we might be) an influence, an idea, a thing intangible, invulnerable,
without front or back, drifting about like a gas? Armies were like
plants, immobile, firm-rooted, nourished through long stems to the
head. We might be a vapour, blowing where we listed. Our kingdoms
lay in each man’s mind; and as we wanted nothing material to live on,
so we might offer nothing material to the killing. It seemed a regular
soldier might be helpless without a target, owning only what he sat on,
and subjugating only what, by order, he could poke his rifle at.
T. E. Lawrence (1926)
Actors in a conflict of interests each come to that conflict with three

things: (1) an estimate of the resources immediately available to fight
with, relative to a potential adversary’s; (2) a plan for the use of those
resources in pursuit of a specified objective (strategy); and (3) an estimate
of resources potentially available once the battle has been joined (again,
relative to a potential adversary’s). As observed in Chapter 1, the
conventional wisdom regarding conflict outcomes is generally derived
from comparisons of each actor’s standing armed forces, economic
capacity, and population at the onset of hostilities. In general, the
actor with the greatest amount of these resources is expected to win,
and to win in proportion to its power advantage.
23
This conventional wisdom is a problem, however. First, relying on it

for explanation leaves many conflict outcomes unexplained. Second, as
an underlying assumption of policy it can have disastrous conse-
quences. In this chapter, I show that in addition to knowledge of each
actor’s available resources, explaining outcomes demands an estimate
of the consequences of the interaction of each actor’s strategy. I argue
that all actor strategies can be sorted into one of two main approaches —
direct and indirect — and that asymmetric conflict outcomes depend on
which of two ideal-type interaction patterns obtains. If strong and weak
actors use a strategy representing the same strategic approach — direct
against direct, or indirect against indirect — strong actors should win as
the conventional wisdom suggests. If, however, strong and weak actors
employ strategies representing opposite strategic approaches — direct
against indirect or indirect against direct — weak actors are much more
likely to win than the conventional wisdom allows for. This is the
strategic interaction thesis.
Explaining asymmetric conflict outcomes

The strategic interaction thesis takes Andrew Mack’s pre-theory of
asymmetric conflict as its starting point. In Mack’s model, relative
power determines relative interests, and results in a conflict in which
the weak actor is systematically more motivated to fight than the strong
actor. Relative interests, in turn, determine relative political vulnerabil-
ities: because weak actors are necessarily more motivated to fight and
win than strong actors, they will be unlikely to quit short of achieving
their political or military goals. Finally, Mack argues that in a long war
the high political vulnerability of the strong actor will eventually force it
to quit, regardless of whether it has an authoritarian or democratic
regime type. This explains why the weak win.
The strategic interaction thesis is a general theory of asymmetric
conflict outcomes. It improves on Mack’s pre-theory in three areas.
First, relative interests determine relative political vulnerabilities, but
in the strategic interaction thesis relative interests are not explained by
relative power. I argue that actor interests are too complex to be usefully
reduced to Mack’s simple formula. Not only have many states, for
example, maintained an interest in the outcome of a conquest or an
occupation far greater than that implied by their relative power, but
state interests often shift once a conflict is under way. Second, an actor’s
regime type matters. It affects the costs and risks of adopting certain
24
strategies (which then alter the likelihood of victory or defeat). Third,

and most importantly, the strategic interaction thesis explains why
some asymmetric conflicts end quickly while others drag on.1
Interests and vulnerabilities

Interests may be thought of both as a proxy for the probability of national
solidarity — that is, broad agreement within a given society and between
its elites and its people — and as a description of it. Although prior to the
Japanese attack on Pearl Harbor, many Americans opposed US interven-
tion in the war in Europe, since 1945 World War II has come to be
regarded as a war fought for the highest of all actor interests: survival
(see e.g., Mack, 1975: 183). Survival — or less dramatically, self-defense — is
almost universally recognized as a legitimate reason to go to war.
When interests in the use of force are low, this is another way of
saying that there is little consensus within a given polity on the need for
the risk and sacrifice of war. Any war effort that is undertaken is
vulnerable to domestic political opposition should things go wrong.
An actor with very high interests will need to be defeated on the battle-
field, while an actor with very low interests need not be defeated
militarily. This is what is meant by political vulnerability.
The corollary of political vulnerability is what might be called poli-
tical license, or just license. When interests are high (survival), a polity
places few limits on the quantity of resources extracted and mobilized
for war, or the way these resources are used. Civil rights can be
abridged or suspended, and the laws of war may be ignored (Mack,
1975: 186, 187). The US experience in World War II again serves as an
example: Japanese Americans were imprisoned during the war in
direct violation of the US Constitution; and Japanese civilians were
deliberately targeted by US firebombing campaigns once the Japanese
main islands came within range of US strategic bombers.2
In sum, actor interests and political vulnerability vary inversely. The
higher an actor’s interests in the issues at stake, the less vulnerable it
will be to being forced to quit a fight before a military decision. The
lower an actor’s interests, the more vulnerable it will be.
1
Mack attributes protracted wars to the use of guerrilla warfare strategy by weak actors.
But he also recognizes that guerrilla warfare strategy is very old, and thus cannot by itself
explain why strong actors didn’t lose as often in the past as after World War II.
2
This was true even though, objectively, by the time US bombers came into range of the
Japanese main islands, no credible case could be made for US survival being threatened.
25
Political vulnerability and conflict duration

Even assuming that one actor is much more interested in the outcome
than the other, the actor with lower interests can still win. The ideal
conflict outcome for an actor with low interests is a fait accompli: a
victory so swift and so final that before a strong actor’s polity can
even begin to debate its utility the conflict is already resolved.
It is highly doubtful, for example, that the United States had a high
interest in the rout of Cuban-supported insurgents on the tiny Caribbean
island of Grenada in 1983. US survival could only have been at stake in
the highly dubious domino logic sense, or in the sense of a national
identity as ‘‘defender of the free world.’’ But the fight in Grenada was
over so quickly there was no time for the US public to consider the need
to intervene, much less to oppose it, before US victory had become an
accomplished fact.
If a fait accompli is the ideal outcome for an actor with low interests in
a conflict prior to a fight, then the worst outcome would be a protracted
war. In a war that drags on there is time to consider the need to resort to
arms, the value of the objectives sought by force, the morality of how
force is applied, and above all the opportunity costs. Committees and
town meetings convene, the press explores all aspects of the conflict,
and political and military elites are called upon to justify the expense of
what was once called ‘‘the appeal to heaven.’’
This euphemism calls to mind a more subtle reason that delay
hurts strong actors — and democratic strong actors in particular. In
ancient times many cultures held that the outcome of a fight repre-
sented the expressed will of a divine being. This meant that when
two parties were aggrieved they could settle their differences by
personal combat; it being presumed that victory in combat indicated
the will of an all-good, all-powerful, and all-knowing supreme being
(Lea, 1968: 86—161), and was therefore legitimate. It may well be that
this tradition of trial by combat, as it came to be known, has never
entirely disappeared from the collective European and North
American consciousness. If true, then a war’s duration — especially
an asymmetric war in which the strong actor’s victory is expected
quickly — becomes a measure of its legitimacy: if the war is a just
one, why is it taking so long? Put differently, the trial-by-combat
legacy implies that the chief source of political vulnerability for
strong actors has less to do with any assessment of ex ante actor
‘‘interests’’ and more to do with the strong association of outcomes
26
with legitimacy, especially when there is a gross mismatch in relative

material power.
In asymmetric conflicts, as in the Grenada example above, there is a
strong expectation that any war will be over quickly. But some asym-
metric wars drag on well beyond what was anticipated before the fight.
If power implies victory, and a lot of power implies quick and decisive
victory, what then explains why power causes quick and decisive
victory sometimes but not others?
My answer is strategy. The interaction of some strategies makes the
way clear for power to determine the outcome, whereas the interaction
of other strategies causes wars to be protracted, thus dramatically
abridging power’s expected effect on the outcome.
Political vulnerability and regime type

In his pre-theory of asymmetric conflict, Mack argues that because both
authoritarian and democratic regimes face resource trade-offs in wars,
political vulnerability constrains strong actors of both types. As Mack
puts it:
Despite these obvious points, my main contention — that limited wars
by their very nature will generate domestic constraints if the war
continues — is not disproved. In terms of the argument put forward
here, ‘‘politics’’ under any political system involves conflict over the
allocation of resources. In closed or centrist polities, these conflicts will
by and large be confined to the ruling elite—but not necessarily so.
(Mack, 1975: 189)
It is a weak argument, and Mack’s discussion previous to this comment

supports the claim that on balance an actor’s regime type affects actor
conduct, and hence, conflict outcomes. As noted in Chapter 1, there are
good reasons to assume — resource trade-offs notwithstanding — that
regime type affects conflict outcomes. But how exactly? I argue that
authoritarian regimes have an advantage in a particular kind of war: a
war in which the regime’s opponents try an indirect defense strategy,
such as terrorism, guerrilla warfare, or nonviolent resistance. In such a
war, the costs and risks of employing the harshest measures are lower
for authoritarian regimes, while the benefits — which don’t vary with
regime type — remain potentially high.3 Authoritarian regimes enjoy
3
I say ‘‘potentially’’ because although logically, indirect defense strategies can be more easily
overcome by depredations against civilians and other noncombatants than conventional
27
two advantages over their democratic counterparts. First, authoritarian

regimes have much more control over what information reaches their
domestic audiences: wars can be made to seem both necessary and
fairly fought. That same control over information can also obscure the
true nature of the violence to international audiences. Moreover, when
evidence of barbarism does reach an authoritarian regime’s public (or
interstate audience) it can be justified as limited reprisal for atrocities
committed by an adversary. Second, even when domestic audiences
become aware of barbarism, they are in no position to alter state policy
or strategy. As Chapter 5 makes clear, Italy’s invasion of Ethiopia was
aided by both advantages.
In other kinds of war and other strategic interactions, however, there
is little evidence to support the claim that authoritarian regimes fight
better or win more often than democratic regimes. On the contrary,
with the exception of long wars (Bennett and Stam, 1998), democratic
states appear to win wars more often than authoritarian states (Reiter
and Stam, 1998). Thus, as in this analysis, the proper question isn’t ‘‘do
authoritarian regimes fight more effectively than democratic regimes,’’
but rather, ‘‘under what circumstances do authoritarian regimes fight
more effectively than democratic ones?’’
Vulnerability and strong actor defeat

Strong actors will more often lose asymmetric conflicts that are pro-
tracted. Mack’s description of the process by which a strong actor
eventually quits the fight serves well here:
In the metropolis, a war with no visible payoff against an opponent
who poses no direct threat will come under increasing criticism as
battle casualties rise and economic costs escalate . . . Tax increases may
be necessary to cover the costs of the war, a draft system may have to
be introduced, and inflation will be an almost certain by-product. Such
costs are seen as part of the ‘‘necessary price’’ when the security of the
nation is directly threatened. When this is not the case, the basis for
consensus disappears. In a limited war, it is not at all clear to those
groups whose interests are adversely affected why such sacrifices are
necessary. (Mack, 1975: 185—186)
defense strategies, the empirical support for barbarism’s increased effectiveness is equivocal.
For a fuller exposition of the counterproductive consequences of the use of barbarism in
asymmetric conflicts, see Arreguı́n-Toft, 2003.
28
Merom’s thesis takes this argument further still. The key causal
relationship is between political will and costs. As Mack and others
have pointed out, asymmetric conflicts are those in which, in general,
the stronger actor will be both more able to absorb unexpected costs
and at the same time, paradoxically, more sensitive to unexpectedly
increased costs. This is because elites and, in democratic strong actors,
publics, have an ex ante expectation of quick and decisive (i.e., low-cost)
victory. A protracted war is already, therefore, an unexpectedly costly
war. Again, the key puzzle is why don’t all asymmetric wars end
quickly?
The strategic interaction thesis

The essence of the strategic interaction thesis is that there are two patterns
of strategic interaction — same-approach and opposite-approach — and
these determine the likelihood that the strong actor will win or lose. This
section explores the logic of this argument.
Strategy
Strategy, as defined here, is an actor’s plan for using armed forces to achieve
military or political goals.4 Strategies incorporate actors’ understandings
(rarely explicit) about the relative values of these objectives.5 Strategy in
this sense should be distinguished from two closely related terms: grand
strategy and tactics. Grand strategy refers to the totality of an actor’s
resources directed toward military, political, economic, or other object-
ives. Tactics refer to the art of fighting battles and of using the various
arms of the military — for example, infantry, armor, and artillery — on
4
The meaning of strategy is both complicated and constantly evolving. Mearsheimer
uses perhaps the simplest definition — ‘‘the plan of attack’’ (see Mearsheimer, 1983: 28—29).
For a discussion of strategy and its evolution, see B. H. Liddell Hart (1967: 333—346);
and J. P. Charnay (1994: 768—774).
5
This lack of explicitness is an important component of strategy because guessing wrong
about how an adversary values its objectives can lead to unexpected outcomes. US
strategy in Vietnam, for example, assumed that after sustaining a certain level of casual-
ties, North Vietnam would no longer be willing to support the insurgency in the South.
The search for this breaking point, and uncertainty over whether it would have any
political utility, bitterly divided the Johnson administration. There may in fact have
been a breaking point in Vietnam, but as US Secretary of Defense Robert McNamara
concluded in 1967, reaching that point would result in virtual genocide (Rosen, 1972:
167—168; Mueller, 1980: 497—519; Brown, 1980: 525—529; and Karnow, 1983: 454, 596).
29
terrain and in positions that are favorable to them.6 Grand strategy,

strategy, and tactics all describe different points on a continuum of a
given actor’s means toward a single end: compelling another to do its will.
The following typology of ideal-type strategies is a useful starting
point for analysis:
Offense strategies (strong actor):
conventional attack
barbarism
Defense strategies (weak actor):

conventional defense
guerrilla warfare strategy (GWS)
Both actors have other strategic options besides those listed here. On
the indirect side, for example, strong actors might choose a ‘‘hearts
and minds’’ or conciliation strategy over barbarism. Weak actors
might choose nonviolent resistance or terrorism instead of a GWS.
Moreover, barbarism could be used defensively and GWS offensively
(though in both cases they would count as representatives of an indirect
strategic approach). In this analysis I assume (1) strong actors initiated
the asymmetric conflict in question;7 and (2), these ideal-type strategies
are war-winning rather than war-termination strategies.8
Conventional attack
Conventional attack means the use of armed forces to capture or
destroy an adversary’s armed forces, thereby gaining control of that
opponent’s values (population, territory, cities, or vital industrial and
communications centers).9 The goal is to win the war in a decisive
6
This definition is a paraphrase of one from the Littré Dictionary as quoted by Charnay
(1994: 770).
7
Strong actors are often but not always the initiators in asymmetric conflicts. Paul counts
twenty weak-actor-initiated conflicts from 394 BC to 1993 (Paul, 1993: 3—4), of which
eleven are included here.
8
In eight asymmetric conflicts (4.1 percent) the outcome was effected by a war termina-
tion or conciliation strategy. Conciliation strategies include the use of bribes, offers of
amnesty, power sharing, or political reforms, and do not require armed forces to imple-
ment. Examples of conflicts ended by a conciliation strategy include the Murid War
(1830—59), the Third Seminole War (1855—58), the Malayan Emergency (1948—57), the
British—Cypriot Conflict (1954—59), and the Philippine—Moro Conflict (1972—80). In the
case of the Malayan Emergency, e.g., see Ramakrishna (2002).
9
The distinction between forces and values derives from the nuclear war strategy
literature; as in ‘‘counterforce’’ (attacking the enemy’s forces) vs. ‘‘countervalue’’ (attack-
ing the enemy’s cities and population) targeting.
30
engagement or a series of such engagements by destroying the adver-

sary’s physical capacity to resist.
In the most common pattern of a conventional attack strategy an
attacker’s forces advance to capture a defender’s values or strategic
assets — say a capital city, industrial or communications center, or
bridge or fort — and the defender moves to thwart that effort. A battle
or series of battles follows, sometimes marked by lulls lasting entire
seasons, until one side admits defeat.
Barbarism
Barbarism is the deliberate or systematic harm of noncombatants (e.g.,
rape, murder, and torture) in pursuit of a military or political object-
ive.10 Unlike other strategies, barbarism has been used to target both an
adversary’s will and its capacity to fight. In a strategic bombing cam-
paign, for example, when will is the target the strong actor seeks to
coerce its weaker opponent into changing its behavior by inflicting pain
(destroying its values).11 In a counterinsurgency (COIN) campaign,
when will is the target the strong actor may attempt to deter would-be
insurgents by, for example, a policy of reprisals against noncombatants.12
But strong actors in a counterinsurgency can also target a weak
actor’s physical capacity to sustain resistance by, for example, imple-
menting a concentration camp policy.13 Historically, the most common
forms of barbarism are the murder of noncombatants (e.g., prisoners of
10
See, for example, Walzer (2000: 151). Chemical and biological weapons have been
traditionally included in this category because they are inherently indiscriminate.
Deliberate destruction of a defender’s natural environment (through deforestation, drain-
ing of swamps, etc.) is also a violation of the laws of war for the same reason (see Perry and
Miles, 1999: 132—135).
11
The classic work here is Schelling (1966: chs. 1 and 4; see also Pape, 1990: 103—146; and
Pape, 1996). In theory it is possible to use strategic air power to target an adversary’s
capacity to fight by using air forces to destroy or interdict supplies, demolish key com-
munications points (railroad junctions, bridges, and airfields), or arms factories. If it were
possible to do so without killing noncombatants this would count as a direct attack
strategy. But in practice — even taking into account advances in precision guided muni-
tions — strategic air power is a too blunt weapon, and noncombatants are killed out
of proportion to the military necessity of destroying the targets. NATO’s strategic air
campaign in Kosovo in 1999 is a case in point (see, e.g., Independent International
Commission on Kosovo, 2000: 92—94; and Daalder and O’Hanlon, 2000).
12
Such reprisals typically include executing randomly selected civilians in retaliation for
the killing of an occupying soldier (see, for example, Asprey, 1994: 108; and Arreguı́n-Toft,
2003).
13
Insurgent intelligence and support networks depend on the participation of sympa-
thetic noncombatants, and concentration camps disrupt these networks (see, for example,
Hamilton, 1998: 59; and Krebs, 1992: 41—42).
31
war or civilians during combat operations); concentration camps;14 and

since 1939, strategic bombing against targets of little or no military
value.15
Conventional defense
Conventional defense is the use of armed forces to thwart an adver-
sary’s attempt to capture or destroy values, such as territory, popula-
tion, and strategic resources. Like conventional attack strategies, these
target an opponent’s armed forces. The aim is to damage an adversary’s
physical capacity to attack by destroying its advancing or proximate
armed forces. Examples include most limited aims strategies,16 static
defense, forward defense, defense in depth, and mobile defense.17
Guerrilla warfare
Guerrilla warfare strategy (GWS) is the organization of a portion of a
society for the purpose of imposing costs on an adversary using armed
forces trained to avoid direct confrontations.18 These costs include the
loss of soldiers, supplies, infrastructure, peace of mind and, most
14
The British used concentration camps as a counterinsurgency (COIN) strategy during
the South African War. Although not intended by the British, as many as 28,000 Boer
women and children died in these camps — more than the combined total of combatant
casualties from both sides. On the use of concentration camps as a COIN strategy, see Ellis
(1976: 111). On their use and consequences in the South African War, see De Wet (1902:
192—193); Pakenham (1979: ch. 29 and 607—608); and Krebs (1992).
15
The Allied bombing of Dresden is a common example (see Schaffer 1985: 97—99). On the
subject of strategic bombing as coercion against Nazi Germany more generally, see Pape
(1996: 260—262). In terms of Pape’s argument, strategic bombing that targets noncomba-
tants would count as barbarism. When air power is used to target enemy forces, it would
count as a conventional attack strategy. Attacks on infrastructure and industry are more
problematic: noncombatants are not deliberate targets, but those who use this strategy
know beforehand that noncombatants will be systematically killed in such attacks.
16
An example of a defensive limited aims strategy would be Japan’s attack on the US
Pacific Fleet at Pearl Harbor in December of 1941. An example of an offensive use of
limited aims strategy would be Egypt’s attack on Israel in October of 1973.
17
For full summary descriptions, see Mearsheimer (1983: 48—50).
18
In Mao Tse-tung’s revolutionary guerrilla warfare strategy, a GWS eventually evolves
into a full conventional confrontation. In this analysis such an evolution would count as a
shift in strategy from an indirect to a direct strategic approach. A related strategy is
terrorism, which often has political objectives similar to GWS. The logic of most terrorism
mirrors that of coercive strategic bombing. A largely urban phenomenon, terrorism
generally seeks either to inflict pain on noncombatants so they will pressure their govern-
ment to accede to the terrorists’ political demands, or to delegitimize a government as a
means of replacing it. This implies that the strategy will be most effective when citizens
have a say in government policies. On terrorism as an insurgency strategy, see Merari
(1993). On suicide bombing as strategy, see Pape (2003).
32
important, time.19 Although GWS primarily targets opposing armed

forces and their support resources, its goal is to destroy not the capacity
but will of the attacker.20
GWS requires two essential elements: (1) sanctuary (physical, e.g.,
swamps, mountains, thick forest, or jungle — or political, e.g., poorly
regulated border areas or border areas controlled by sympathetic
states), and (2) a supportive population (to supply fighters with intelli-
gence, supplies, and replacements). The method of GWS is well sum-
marized by perhaps its most famous practitioner, Mao Tse-tung:
In guerrilla warfare, select the tactic of seeming to come from the east
and attacking from the west; avoid the solid, attack the hollow; attack;
withdraw; deliver a lighting blow, seek a lighting decision. When
guerrillas engage a stronger enemy, they withdraw when he advances;
harass him when he stops; strike him when he is weary; pursue him
when he withdraws. In guerrilla strategy, the enemy’s rear, flanks, and
other vulnerable spots are his vital points, and there he must be
harassed, attacked, dispersed, exhausted, and annihilated.
(Tse-tung, 1961: 46)
GWS is not a strategy for obtaining a quick defeat of opposing forces

(Hamilton, 1998: 27). Moreover, because guerrillas cannot hold or
defend particular areas (save isolated base areas), they do not provide
security for their families while on operations or when demobilized to
await new missions. GWS is therefore a strategy that requires placing
key values (e.g., farms, family, religious or cultural sites, and towns)
directly into the hands of the adversary. Logically then, important costs
of adopting a GWS depend on the purpose and restraint of the adver-
sary.21 When invading or occupying forces do not exercise restraint in
the use of force, or when their political objective is the destruction rather
than coercion of a weak actor’s people, GWS can become a prohibitively
expensive defensive strategy.22
19
On this point especially, see Samuel Huntington’s remarks in Hoffmann et al. (1981: 7;
see also Cohen, 1984: 157).
20
For general introductions to GWS, see Laqueur (1976); Ellis (1995); Asprey (1994); and
Joes (1996). On Chinese and Cuban variations of GWS, see Tse-tung (1961); Katzenbach
and Hanrahan (1955: 321—340); Guevara (1961); and Debray (1968).
21
In March 1900 the British captured the first Boer capital, Bloemfontein. Surviving Boer
commanders gathered to decide whether to surrender or keep fighting. They were closely
divided, but tipping the balance in favor of continued — guerrilla — war was faith in British
civility. The Boer found their faith unjustified (see De Wet, 1902: 192—193).
22
The question of whether an indirect defense — in particular, nonviolent resistance — can
be effective against a ruthless strong actor is taken up by Gene Sharp and others (see
McCarthy and Sharp, 1997; Sharp, 2003). The case of Kosovo from 1998 to 1999 suggests
33
Every strategy has an ideal counterstrategy. Actors able to predict
an adversary’s strategy in advance can therefore dramatically improve
their chances of victory — or at a minimum inflict unexpectedly high
costs on an adversary — by choosing and implementing that ideal
counterstrategy. Mao, for example, argued that ‘‘defeat is the invariable
outcome where native forces fight with inferior weapons against
modernized forces on the latter’s terms.’’23 Mao’s maxim suggests that
when the weak fight the strong, the interaction of some strategies will
favor the weak while others will favor the strong.
Building on Mao’s insight, I argue that a universe of potential
strategies and counterstrategies can be aggregated into two distinct
ideal-type strategic approaches: direct and indirect.24 Direct strategic
approaches — e.g., conventional attack and defense — target an adver-
sary’s armed forces with the aim of destroying or capturing that adver-
sary’s physical capacity to fight, thus making will irrelevant. They feature
soldier-on-soldier contests along with codified rules as to their conduct
and a shared conception of what counts as victory and defeat. Indirect
strategic approaches — e.g., barbarism and GWS — most often aim to
destroy an adversary’s will to resist, thus making physical capacity
irrelevant.25 Barbarism targets an adversary’s will by murdering, tor-
turing, or incarcerating noncombatants. GWS attacks an adversary’s
will by targeting enemy soldiers, though noncombatants may be targets
as well. This constant-if-incremental loss of soldiers, supplies, and
equipment, with little chance of a quick resolution, is aimed at the
balance of political forces in the stronger actor’s homeland. Same-
approach interactions (indirect—indirect or direct—direct) imply defeat
for weak actors because there is nothing to mediate or deflect a strong
actor’s power advantage. Barring a battlefield miracle, these interac-
tions should therefore be resolved in proportion to the force applied. By
that nonviolent resistance against an adversary bent on genocide will only prove effective
when it provokes an armed external intervention (see, e.g., Independent International
Commission on Kosovo, 2000).
23
Quoted in Mack (1975: 176, emphasis in original).
24
This reduction of strategies to two mutually exclusive approaches is well established in
the strategic studies literature. See, for example, Corvisier and Childs (1994: 378); and
Liddell Hart (1967: 197, 361—364); see also, Galtung (1976).
25
For a similar definition see Pape (1990: 106—107). If coercive power is the product of will
and physical capacity, then either approach can win: reducing an opponent’s capacity to
zero makes its willingness to fight irrelevant; and reducing its willingness to fight to zero
makes its capacity irrelevant.
34
contrast, opposite-approach interactions (direct—indirect or indirect—

direct) imply victory for weak actors because the weak refuse to engage
where the strong actor has a power advantage (i.e., ‘‘on the latter’s
terms’’).26 They therefore tend to be protracted, and time favors
the weak.27
When strategic interaction causes an unexpected delay between the
commitment of armed forces and the attainment of military or political
objectives, strong actors tend to lose asymmetric conflicts for two rea-
sons. First, although all combatants tend to have inflated expectations
of victory (Blainey, 1988: 53), strong actors in asymmetric conflicts are
particularly susceptible to this problem (see e.g., Mack, 1975: 181—182;
Herring, 1986: 144—145). If power implies victory, then an overwhelm-
ing power advantage implies an overwhelming — and rapid — victory.
As war against a Lilliputian opponent drags on, overestimates of success
force political and military elites to either escalate the use of force to
meet expectations (thus overtly increasing the costs of a conflict which
were expected to be slight), lie, or look increasingly incompetent. Either
way, for democratic strong actors, domestic pressure to end the conflict
is likely to increase. The longer the war drags on the greater the chances
the strong actor will simply abandon the war regardless of the military
state of affairs on the ground.
Second, strong actors anxious to avoid increasing costs — such as
declaring war, mobilizing reserves, raising taxes, or sustaining add-
itional casualties — may be tempted to use barbarism because they
believe it to be a cost-saving strategy for victory.28 But in the post-
World War II era, barbarism is a difficult strategy to prosecute effect-
ively. Barbarism that is effective militarily demands ‘‘special’’ forces,
trained and equipped to accomplish their demanding ‘‘mission.’’ It
demands, above all, thoroughness in its application. Even the Nazis
found both demands beyond their capabilities in World War II.
26
In GWS, an attacker’s armed forces are physically avoided or only engaged on favor-
able terms. In a blockade or strategic bombing campaign against a direct defense in a
limited war, the strong actor’s destructive power is deflected because such attacks invari-
ably place the noncombatant population between attackers and political elites.
27
On the importance of conflict duration as a cost of conflict, see Mearsheimer (1983: 24);
and Katzenbach and Hanrahan (1955: 324—326).
28
Alexander Downes persuasively argues that both democratic and authoritarian
states are likely to employ barbarism in these circumstances because they believe it is a
cost-saving strategy (see Downes, 2003). My own research shows that barbarism rarely
is effective militarily (it tends to be counterproductive) and is almost always
counterproductive politically (see Arreguı́n-Toft, 2003).
35
The result was a resistance backlash that dramatically increased the costs
of Nazi occupation in the conquered territories they administered.29
But even when militarily effective barbarism is risky: for democratic
strong actors it carries the possibility of domestic political discovery
(and opposition30) and for actors of either regime type who are not
nuclear powers, it carries the risk of external intervention.
Strategic interaction: explaining the trend

This is not a book about how and why actors choose the strategies they
do. Nor do I offer a general explanation of the reasons some actors shift
strategies during a war while others don’t.31 Rather, I take actor strat-
egies as given and then map the consequences of strategic interaction
on outcomes.
That said, my explanation for the trend toward increasing strong
actor failure is suggested both by the timing of the biggest shift
in outcomes favoring weak actors (1950—99), and by the logic of
Kenneth Waltz’s argument that actors in a competitive international
system ‘‘socialize’’ to similar policies and strategies. As Waltz argues,
‘‘The fate of each state depends on its responses to what other states
do. The possibility that conflict will be conducted by force leads to
competition in the arts and the instruments of force. Competition
produces a tendency toward the sameness of the competitors’’ (Waltz,
1979: 127).
29
The Nazis did better at developing forces capable of prosecuting barbarism than
they did at achieving a thoroughness in barbarism’s application. The few survivors
and unindoctrinated witnesses who escaped, e.g., the cruel murdering pits of the
Einsatzgruppen, spread word of these atrocities, thereby increasing the costs of civil
administration and military operations throughout occupied territories. Nazi barbarism
raises a problem for the thesis at hand, however, in that, unlike most barbarism, which is
intended to be coercive (i.e., a means to an end), the bulk of Nazi barbarism (viz., that
aimed at the mentally ill, handicapped, Jews, homosexuals, and Roma) was not intended
to coerce but to destroy utterly the groups targeted: it was less a strategy and more an end
in itself. Many of these murders — e.g., against Jews in Poland and further east — were
initially justified as COIN measures when they were demonstrably nothing of the kind.
However, a smaller but real proportion of Nazi barbarism — e.g., in occupied Yugoslavia —
was in fact intended to serve a COIN utility in occupied territories. On the creation of the
Einsatzgruppen and the challenges of training people to mass murder of noncombatants,
see Bartov (1992); and especially Rhodes (2002). On the Nazi use of barbarism as COIN
strategy in Yugoslavia, see, e.g., Hehn (1979).
30
Mack correctly emphasizes that barbarism is judged in proportion to the relative power
of the actors: weak actors will be forgiven abuses for which strong actors will be hanged
(see Mack, 1975: 186—187).
31
In 77.5 percent of asymmetric conflicts neither actor switches strategies. Winners and
losers stay with the same strategy they started the war with.
36
By ‘‘socialization,’’ Waltz means that states imitate each others’ suc-

cesses and avoid what they count as failures. Over time, this means
actors will have a tendency to converge on the most successful integra-
tion of arms, doctrine, training, strategy, and tactics; where ‘‘success’’ is
measured by what won the last big war. I argue that socialization works
regionally, and that after World War II different patterns of success
emerged in Europe and Asia. In Europe, success in war was measured
by a state’s capacity to produce and deploy large mechanized and
combined arms forces designed to destroy an adversary’s armed forces
on relatively open terrain; capturing its values without the need for
costly battles of annihilation: a blitzkrieg. This model was imitated by
the United States,32 its European allies, and the Soviet Union.33 In Asia,
success was measured by the ability to sustain a protracted conflict
against a technologically superior foe on relatively closed terrain:
a GWS. Mao’s long fight for and eventual conquest of China was a
model consciously imitated by Algerian rebels, the Vietminh, the
Hukbalahap, Cuban insurgents, Malayan communists, and, to a large
extent, Afghanistan‘s mujahideen.34 The blitzkrieg model is a direct
strategic approach; the guerrilla warfare model, an indirect one.
When the two interacted systematically — as they did with greater
frequency following World War II — the strong actor lost more often.35
The fact that in nearly 78 percent of asymmetric conflicts losing actors
don’t switch strategies suggests that actors on the verge of armed
conflict or defeat are not entirely free to choose an ideal strategy. This
is true for two reasons. First, forces, equipment, and training — all
closely integrated — are not fungible. Each strategic approach will be
facilitated by certain kinds of forces and technology and undermined
by others. Moreover, the development and prosecution of an actor’s
32
See, e.g., Cohen (1984: 179) on this point.
33
It is true that the major European colonial powers — France and Britain — each possessed
large and specialized colonial militaries before World War II. In both World War I and
World War II the ‘‘colonial’’ nature of these armies were sometimes cited as reasons why
French and British land forces did so poorly in each war, especially in the early phases.
But, following the war, during which Britain in particular attempted to socialize to the
blitzkrieg standard, the ability of its troops to conduct counterinsurgency operations,
while still greater than that of the United States and Soviet Union, was diminished
compared to what it had been prior to World War II.
34
On Mao’s revolutionary guerrilla warfare as a template, see Katzenbach and Hanrahan
(1955: 322); and Hamilton (1998: 18).
35
Nationalism proved necessary but not sufficient to account for the success of the GWS
pattern against states possessed of blitzkrieg forces. Both GWS without nationalism and
nationalism without a GWS will lose to a strong actor’s direct approach assault.
37
ideal strategy or counterstrategy may be blocked by entrenched organ-

izational interests or traditions.36 Second, actors prioritize threats: if the
United States and Soviet Union, for example, identified each other as
the primary threat, and both calculated that the most likely area of
direct confrontation would be the heart of Europe, then adopting inte-
grated equipment, doctrines, and strategies favorable to winning that
sort of war made sense. But the nature of threats can shift faster than an
actor — even one as technologically advanced as the United States — can
shift strategies. The inertia of decades of preparation to win one kind of
war can become a crippling liability when faced with an entirely dif-
ferent kind of war in a different terrain and climate. This is why strong
actors — mostly European — lost so many conflicts in Asia in the period
following World War II. This is why the United States is currently doing
so poorly in its ‘‘war’’ against terrorism (Arreguı́n-Toft, 2002).
Main hypotheses: strategic interaction and conflict

outcomes
This section explores the logic of four distinct strategic approach inter-
actions, and explains how hypotheses derived from each can be use-
fully reduced to a single hypothesis. The expected relationship of
strategic approach interactions to outcomes in asymmetric conflicts is
summarized in Figure 3.
Direct offense vs. direct defense

In this interaction both actors make similar assumptions about the
priority of values over which they fight. Both can therefore be expected
to agree about the implications of a catastrophic loss in battle, the rules
of war, or the capture of a capital city. Because nothing mediates
between relative material power and outcomes, strong actors should
win quickly and decisively.
Hypothesis 1: When strong actors attack using a direct strategic approach,
and weak actors defend using a direct strategic approach, all other things being
equal, strong actors should win quickly and decisively.
Direct offense vs. indirect defense

Unlike direct strategies, which involve the use of forces trained and
equipped to fight as organized units against other, similarly trained and
36
See, e.g., Cohen (1984: 165—168); Krepinevich (1986); Asprey (1994: 36); and Marquis
(1997).
38
weak actor
strategic
approach
direct indirect
strong
actor strong weak
direct
strategic actor actor
approach
weak strong
indirect
actor actor
Figure 3. Expected effects of strategic interaction on conflict outcomes

(expected winners in cells)
equipped forces, indirect defense strategies (such as terrorism, GWS,

or nonviolent resistance) typically rely on forces difficult to distin-
guish from noncombatants when not in actual combat. As a result, an
attacker’s forces tend to kill or injure noncombatants during operations,
which tends to stimulate weak actor resistance.37 Most important,
because indirect defense strategies sacrifice values for time (see
Katzenbach and Hanrahan, 1955: 325—326), they necessarily take longer
to resolve so long as weak actors continue to have access to sanctuary
and social support.38 In asymmetric conflicts, delay favors the weak.
Hypothesis 2: When strong actors attack with a direct strategic approach
and weak actors defend using an indirect approach, all other things being equal,
weak actors should win.
Indirect offense vs. direct defense

Because the overwhelming force available to the strong actor implies
success against a weak adversary attempting a conventional defense,
an indirect offensive strategy in this context targets a defender’s will to
resist. Prior to the advent of strategic air power and long-range artillery
37
Logically it could undermine resistance but empirically, it doesn’t (see Arreguı́n-Toft,
2003).
38
Although a GWS requires sanctuary and social support, mere access to them in no way
mandates its adoption. French forces, for example, had access to both as they faced defeat
in the Franco-Prussian War in 1870. France actively considered adopting a GWS after the
disastrous defeat at Sedan. Yet threatened with the loss of Paris, France surrendered
instead (see Howard, 1961: 249—250).
39
(such as the V-1 and V-2 rockets of World War II), blockades and sieges
were the primary means of coercing adversaries in this way. Today,
strategic bombing campaigns39 and economic sanctions40 are the most
common forms of indirect offense against direct defense when the
attacker is the stronger actor.41
The logic of this interaction could go either way. On the one hand,
and as imagined by such early air power theorists as Douhet and
Mitchell, the threat of attacking an opponent’s population centers and
industry — skipping over entrenched standing armies rather than
attacking them directly — could by itself coerce that opponent into
changing its behavior.42 It could also turn a defender’s people against
its own government, making it rational for them to force their leaders to
capitulate and spare the citizens further injury. On the other hand, the
injury and death of noncombatants — in particular, children — could
increase resistance among citizens who might formerly have been
opposed to the war or neutral. It might bring them closer to their
government rather than alienating them. A defender’s government
might also disperse key resources more widely, protecting them from
future destruction from the air. And, as Pape notes, modern nation
states are adept at redistributing the burdens of blockades and eco-
nomic sanctions away from defense assets (Pape, 1997: 93, 109). Over-
all, I expect strong actors to lose these interactions because they are (1)
time-consuming; (2) tend toward barbarism;43 and (3), in the post-World
39
Robert Pape has shown that strategic bombing or ‘‘punishment’’ strategies rarely work
(and they cannot work against indirect defense strategies, such as GWS) (see Pape, 1996:
ch. 6; see also Clodfelter, 1989). If Pape is right and tactical air power is effective as a means
to coerce an adversary, then tactical air support that accepts collateral damage should
become more common; and ‘‘human shield’’ defense of, say, armored or transport col-
umns, will become an increasingly common countermeasure.
40
See , e.g., Pape (1997: 90—136). Pape undertakes an updated look at an old debate about
the effectiveness of economic sanctions to advance noneconomic political objectives and
concludes that such sanctions are likely to be effective only in rare circumstances.
41
When the attacker is the weaker actor, terrorism and insurgency are the most common
forms of an indirect offense against a direct defense. I assume strong actors are the
attackers because (a) I am building and testing a theory of asymmetric conflict outcomes,
and (b) the strong attacking the weak is by far the most common pattern.
42
See Douhet (1921); and Mitchell (1925). On the effectiveness of strategic air power in the
first Gulf War, see Press (2001). There remains a healthy debate about whether the 1999 air
campaign against Slobodan Milosevic’s government counts as successful case of coercion
by air power. For an argument against, see Daalder and O’Hanlon (2000). For an argument
for air power’s effectiveness, see Stigler (2003: 124—157).
43
Strategic bombing campaigns usually start out with the intent to spare noncombatants —
often to the point of putting pilots and air crews at increased risk (e.g., by flying lower or
40
War II period of high nationalism, barbarism tends to be militarily and

politically counterproductive (Arreguı́n-Toft, 2003).
Hypothesis 3: When a strong actor attacks with an indirect strategic
approach against a weak actor defending with a direct approach, all other
things being equal, the strong actor should lose.
Indirect offense vs. indirect defense

Indirect defense strategies presuppose a certain level of restraint on
behalf of attackers.44 When strong actors employ a strategy that
ignores such restraint, weak actors are unlikely to win, both because
there would be no one left to win for, and GWS depends directly on
a network of social support for intelligence, logistical support, and
replacements.45 Barbarism works as a COIN strategy because by
attacking either or both of the essential elements of a GWS — sanctuary
and social support — it destroys an adversary’s capacity to fight.46
slower) — but in most cases these campaigns escalate until either noncombatant casualties
are simply accepted (as in US bombing of North Vietnam during the ROLLING THUNDER
campaign), or they become intended targets (as in the case of the firebombings of Dresden
and Tokyo and the atomic bombings of Hiroshima and Nagasaki). A strict interpretation of
the laws of war may make a strategic air campaign that accepts ‘‘collateral damage’’ (viz.,
death or injury to noncombatants) a war crime. The issue is tricky: the laws of war permit
collateral damage so long as that damage is proportional to the military necessity of a
target’s destruction. But where strategic air power has no military utility, and this has been
established by bomb damage assessments, then collateral damage subsequent to those
assessments would constitute a war crime. It would also count as barbarism, even though
‘‘collateral damage’’ is by definition unintended. This is true because although the specific
deaths and injuries caused by the attacks were not intended, they were generally accepted
and therefore systematic.
44
Harry Turtledove explores the counterfactual of what would have happened had
Gandhi attempted his nonviolent resistance strategy (an indirect defense strategy akin
to GWS and terrorism) against Hitler’s Third Reich instead of against the British. In
Turtledove’s fictional account, Gandhi’s resistance is crushed and Gandhi and Nehru
are executed (see Turtledove, 2001: 231—262).
45
Mao Tse-Tung once analogized the relationship between insurgents and citizens in a
people’s war by likening the fighters to fish and the people to the sea. Effective counter-
insurgency would therefore require either altering the terrain (making the sea transpar-
ent) or killing, expelling, or imprisoning the people (drying up the sea). Both would count
as barbarism (see Mao Tse-Tung, 1968: 284). This interaction is best exemplified by the
Serb strategy in Kosovo from 1998 to 1999, prior to the intervention of NATO in Operation
Allied Force. In combination with Serb paramilitary forces, regular armed forces of the
Yugoslav military engineered the flight of ethnic Albanians so effectively that they
destroyed any chance that the KLA would become an effective insurgency within Kosovo.
46
As noted above, the problem with barbarism isn’t that it can’t win wars, but that it can
only win in special circumstances (comprehensively applied by special forces and in a
COIN context) and it almost always results in a lost peace. This is an historical trend which
doesn’t make itself obvious until the return of the colonial powers to ‘‘their’’ colonies
following World War II. Murder, torture, and brutality could deter resistance and cow
41
Hypothesis 4: When strong actors employ barbarism to attack weak actors

defending with a GWS, all other things being equal, strong actors should win.
Each of the four interaction outcome hypotheses describes an
interaction of either same-approach or opposite-approach strategic
interactions. It follows that all four may be tested as a single hypothesis.
Hypothesis 5: Strong actors are more likely to win same-approach interac-
tions and lose opposite-approach interactions.
Alternative hypotheses
As noted in Chapter 1, there are three alternative explanations of both
strong actor failure in asymmetric conflicts, and of the trend toward
increasing failure over time. Each generates testable propositions.
For example, the logic of the arms diffusion argument clearly implies
that better arms for the weak actor make strong actor failure more
likely. This in turn implies that arms themselves are an important
component of ‘‘power,’’ and by extension, the conflicts in question
were less asymmetric than they seemed (the weak were less weak
when better armed). If true, the high diffusion of military technology
to the developing world after World War II might account for the trend
in strong actor failures as well.
Hypothesis 6: The better armed a weak actor is, the more likely it is that a
strong actor will lose an asymmetric conflict.
The logic of the nature-of-actor argument also yields testable
propositions:
Hypothesis 7(a): Authoritarian strong actors win asymmetric wars more
often than do democratic strong actors.
Hypothesis 7(b): Authoritarian strong actors win asymmetric wars in which
the weak actor uses an indirect strategy more often than do democratic strong
actors.
The interest asymmetry thesis provides an alternative explanation of
asymmetric conflict outcomes in two senses. First, it holds that relative
power explains relative political vulnerabilities (power and vulnerabil-
ity vary directly), and second, it holds that regime type does not matter.
It therefore yields two testable propositions:
Hypothesis 8: Relative material power explains relative interests in the
outcome of an asymmetric conflict.
social support prior to World War I. But following World War II the same methods tended
to increase resistance, thus increasing the costs of conquest and especially occupation.
Strong actors willing or able to follow barbarism through to its extremes (e.g., the French
in Algeria under General Massù) could still win wars, but no longer could these methods
win the peace.
42
Hypothesis 9: Authoritarian and democratic strong actors share roughly

equal political vulnerability in a prolonged asymmetric conflict.
Evaluations of hypotheses 5, 6, and 8 will be included in the conclud-
ing section of each historical case study. Hypotheses 7a, 7b, and 9 can
only be evaluated across cases. Evaluations of these hypotheses will
therefore appear in the overall conclusion to the book.
A quantitative test of the strategic interaction thesis

Even if the strategic interaction thesis is plausible and logically sound,
before it can be pronounced a theory it needs testing. Here I test the
strategic interaction thesis in a quantitative analysis of 202 asymmetric
conflicts fought between 1816 and 2003. The aim is to determine
whether there is a statistically significant correlation between strategic
interaction and asymmetric conflict outcomes, but I also test alternative
arguments, such as the argument that the weak won wars after World
War II because they were aided by the superpowers.
Coding and cases

The basic method of coding cases was to examine the history of each
war in the Correlates of War data set. A conflict was coded asymmetric if
the halved product of one actor’s armed forces and population
exceeded the simple product of its adversary’s armed forces and popu-
lation by 5:1 or more. If the strong actor used armed forces to attempt to
destroy a weak actor’s forces or capture (but not destroy) values, it was
coded a conventional attack. If the weak actor used armed forces
against the strong actor’s forces in an attempt to thwart these attacks,
it was coded a conventional defense. A coding of barbarism was
reserved for strong actors that systematically targeted noncombatants,
employed indiscriminate weapons, or accepted collateral damage in
a strategic bombing campaign after bomb damage assessments cast
doubt on the military necessity of the campaign as a whole. A weak
actor was coded as using a GWS if it sought to impose costs on the
strong actor with armed force while avoiding pitched battles. Each
conflict dyad was coded with one of four strategic interactions (direct—
direct, direct—indirect, indirect—indirect, or indirect—direct),47 before
47
Most asymmetric wars contain a single strategic interaction from start to finish, but a
few, such as the South African War or the US intervention in Vietnam, contain multiple-
sequential or multiple-simultaneous interactions, respectively. In multiple-sequential
43
being reduced to one of two interaction types (same—approach or

opposite—approach).48
The key variable of analysis is strategic interaction (STRATINT)
as compared to conflict outcome (OUTCOME). If strategic interaction
causes change in conflict outcome, then a shift in the value of
strategic interaction across the case universe should be matched by
a corresponding shift in outcome. The STRATINT variable was coded
‘‘0’’ if the strategic interaction was same-approach (direct—direct or
indirect—indirect), and ‘‘1’’ if it was opposite-approach (direct—indirect
or indirect—direct). The OUTCOME variable was coded ‘‘0’’ if the strong
actor lost, and ‘‘1’’ if it won.49
Strategic interaction and conflict outcomes

Cross tabulations established that strategic interaction and asymmetric
conflict outcomes are associated, and that the relationship is statistically
significant (see Figure 4).50
The results thus support hypothesis 5.51 Strong actors won 76.8
percent of all same-approach interactions, and weak actors won 63.6
interaction conflicts, strategies change, but in temporal sequence: one side’s strategic shift
is quickly followed by another’s. Multiple-simultaneous interaction conflicts are those in
which a single actor or an actor and its allies pursue different strategies against the same
adversary within a single theater of operations. The empirical distribution of conflict
types is as follows: single, 134 (77.5 percent), multiple-sequential, 29 (16.8 percent), and
multiple-simultaneous, 10 (5.8 percent). In this analysis, conflict outcomes are explained
by strategic interaction outcomes.
48
In the main analysis, the relatively few wars with multiple interactions were reduced to
single interactions. In multiple-sequential conflicts the final interaction was used to
represent the overall conflict: the South African War was coded same-approach because
it ended with an indirect—indirect interaction. In multiple-simultaneous conflicts, inter-
actions were averaged: US intervention in Vietnam was coded opposite-approach because
although some interactions were same-approach, on balance, the contest was decided by a
direct-indirect interaction. The chief consequence of these reductions is a tougher test for
the strategic interaction thesis, because collapsing interactions increases the impact of
relative material power on outcomes. Because strong actors have a greater material
capacity to adapt to failure than do weak actors, collapsing interactions hides strong-
actor rather than weak-actor failures. I also tested the thesis using strategic interactions
(rather than conflict outcomes) as the dependent variable, thus un-collapsing the data
analyzed here. See fn. 51.
49
Stalemates and ongoing conflicts were coded losses for the strong actor.
50
Pearson chi-squared (1), 14.56, p < 0.001.
51
An analysis of the relationship between strategic interaction and interaction outcomes
(as opposed to war outcomes) produces an even more striking finding: weak actors win
23.1 percent of same-approach and 78.4 percent of opposite-approach interactions. This
relationship is statistically significant (Pearson chi-squared (1), 40.95, p < 0.001). An
analysis of the impact of external noncombat support for weak actors did not refute the
strategic interaction thesis; even when weak actors received no support, they were still
44
Percent of victories
100 strong actor
76.8 weak actor
80 63.6
60
36.4
40 23.2
20
0
same approach, n = 151 opposite approach, n = 22
Figure 4. Strategic interaction and asymmetric conflict outcomes,

1800—2003
percent of all opposite-approach interactions. Weak actors were there-

fore nearly three times more likely to win when fighting strong actors in
an opposite-approach strategic interaction.
Testing an alternative explanation: external

support for weak actors
How many of these weak actor victories resulted from external sup-
port?52 Logically, strong actors should rarely attract external support
when combating weak actors, whereas weak actors may often attract
some support.53 Weak actors could therefore be winning asymmetric
conflicts because they are supplied from outside, rather than because
they use a favorable counterstrategy. My data set allows comparison of
key causal variables while controlling for external support for the weak
actor.54 The results are summarized in Figure 5:
Figures 4 and 5 are very similar. The fact that weak actors are slightly
worse off implies that external support matters, but the key finding
is that even when they receive no external support, weak actors are
still three times more likely to win opposite-approach interactions
than they are same-approach interactions. Essentially, the effects of
three times more likely to win opposite-approach interactions than same-approach inter-
actions (Pearson chi-squared (1), 11.38, p < 0.001).
52
Support in this context means arms, logistical support, and perhaps military advisers,
but not combat troops.
53
The United States received support — including actual combat units — from Turkey
during the Korean War and Australia during the Vietnam War. In these contexts, however,
support was less about materially affecting the balance of forces available to the strong
actor, and more about ratifying and legitimizing US policy.
54
There were not enough cases of positive external support to make a statistically
significant finding. As a result, I report only the relationship of strategic interaction and
outcomes when the weak actor received no external support.
45
100
Percent of victories
81.9 strong actor
80 weak actor
56.3
60 43.8
40
18.1
20
0
same-approach, n = 138 opposite-approach, n = 16
Figure 5. Strategic interaction and conflict outcomes when weak actor

received no external support
strategic interaction overwhelm the effects of external support for weak

actors.
Strategic interaction and conflict duration

A key causal mechanism of the strategic interaction thesis is time:
same-approach interactions should be over quickly, whereas opposite-
approach interactions should be protracted (and weak actors tend
to win protracted wars). An analysis of the average duration of
same-approach and opposite-approach interactions supports this
claim: the mean duration of opposite-approach interactions was 2.27
years; of same-approach interactions, 1.57 years (1.66 years was the
overall mean).
Strategic interaction and long-term trends

In terms of the trend, both opposite-approach interactions and weak
actor successes have increased over time: from 1800 to 1849, 5.9 percent
of interactions in thirty-four asymmetric conflicts were opposite-
approach. From 1850 to 1899, 10.1 percent of interactions in sixty-nine
asymmetric conflicts were opposite. From 1900 to 1949, 16.1 percent of
thirty-one asymmetric conflicts were opposite, and from 1950 to 1999,
21.1 percent of thirty-eight asymmetric conflicts were opposite.
In sum, the data analysis supports three key hypotheses relating
strategic interaction to asymmetric conflict outcomes. First, strong
actors are more likely to lose opposite-approach strategic interactions.
Second, opposite-approach interaction conflicts last longer than same-
approach interactions. Third, the frequency of opposite-approach
46
interactions has increased in rough proportion to strong actor failure

over time.55
This analysis is limited because some data are missing: many civil and
colonial wars recorded neither the quantity of forces committed nor the
strategies actors used. Although these defects are balanced by statistical
controls, even a perfect data set would support only a correlation
between variables, not causation. Thus, although the data analysis
might have refuted the strategic interaction thesis, only in combination
with a careful comparison of historical cases can the thesis be confirmed.
Conclusion
Actors in a conflict of interests each come to that conflict with resources,
a plan for the use of those resources, and hopes for help from allies or
soon-to-be-deployed weapons in development. There is no question
that all other things being equal an abundance of resources is good to
have in a war. But if the strategic interaction thesis is right, the inter-
action of each actor’s plan for the use of those resources — whether
grand or meager — turns out to be even more important.
How the remaining chapters are organized

The remainder of this book is taken up with a systematic analysis of
five historical case studies. Each case study has the same structure:
historical background (including geography and climate), actor interests
prior to the fight, how the war was fought, its outcome, and its after-
math. The focus of each chapter is provided by each competing
argument’s expectations: how did each actor’s power and regime
type affect its interests and strategies? How did comparative military
technology work? Was one side politically vulnerable, and if so, how, if
at all, did that vulnerability contribute to the war’s outcome?
55
The proportion of strong actor defeats as compared to the increased percentage of
opposite-approach interactions is greater in the final fifty-year period than in the other
periods. This suggests that other factors — such as an established norm of anti-colonialism,
a rise in nationalism, the spread of free trade regimes, and interference by superpowers
during the Cold War — explain some of the trend. However, as the data make clear, the
strategic interaction thesis remains the most important causal variable.
47
3 Russia in the Caucasus: the Murid
War, 1830–1859
The Caucasus may be likened to a mighty fortress, strong by nature,

artificially protected by military works, and defended by a numerous
garrison. Only thoughtless man would attempt to escalade such a
stronghold. A wise commander would see the necessity of having
recourse to military art, and would lay his parallels, advance by sap
and mine, and so master the place. Veliaminov (Baddeley 1908)
When you do battle, even if you are winning, if you continue for a long
time it will dull your forces and blunt your edge; if you besiege a
citadel, your strength will be exhausted. If you keep your armies out in
the field for a long time, your supplies will be insufficient . . .
Therefore I have heard of military operations that were clumsy but
swift, but I have never seen one that was skillful and lasted a long time.
It is never beneficial to a nation to have a military operation continue
for a long time. Sun Tzu (1988)
Russia’s attempt to annex the Caucasus began in earnest nearly two
hundred years ago. Since many contemporary studies of interstate
politics begin with caveats about the sensitivity of their findings to
particular periods of history,1 it seems worthwhile to say why I think
the careful study of so distant a conflict is still useful.
Appropriateness
Clearly there are many things about the place and time which appear
unique to both. The character of warfare two hundred years ago was
1
For example, in Theory of International Politics, Kenneth Waltz restricts the explanatory
power of structural realism to the post-Westphalian time period; and in Conventional
Deterrence, John Mearsheimer restricts the explanatory power of his argument to the
post-World War II time period. See Waltz (1979), and Mearsheimer (1983).
48
Russia in the Caucasus
different from that of the post-World War II era. In the period in

question — stretching from the time of the French Revolution to the
American Civil War — two differences in particular stand out.
First, the profound sense of honor, duty, and glory associated with
war was a major source of motivation for both officers and men. These
things strongly motivated a willingness to take risks and make sacri-
fices; and, most particularly, bravery in battle. Bravery in battle was in
turn widely believed to affect combat outcomes, and so it did.
Second, the staggering costs of the Murid War in terms of time (thirty
years) and lives (well over half-a-million men) call into question our
preconceived notions about the meaning of victory in war; or less
prosaically, of the relative costs of achieving a given political objective
by military means. Russia’s conquest of the Caucasus gave it few new
strategic advantages save stability on its southern frontier. The region
was costly to control, difficult to traverse and, until the industrial age
(which dawned quite a bit later in Russia than elsewhere), its wealth in
oil lay fallow.
These differences aside, there remain enough similarities during the
time of the Murid War to make it relevant to our larger purpose. There
was still an interstate system, with regime types well described as
democratic or authoritarian. Weapons technologies were different
from today, but strategies were similar, and the interaction between
weapons technology and strategy were analogous as well. Most import-
antly, strategic interaction mattered as much in the nineteenth century
as at Hannibal’s victory at Cannae in 216 BC, or Norman Schwarzkopf’s
in the Kuwaiti theater of operations in 1991.
The Caucasian theater

The idea that as a comprehensive territorial space stretching from the
Caspian to the Black Sea, the Caucasus form an unrivaled natural
fortress, is repeated by every commentator on that struggle, and on
both sides. As Napoleon did before them, and as Hitler was to do after
them, the Russians lamented both the difficulties of climate and of the
densely forested and sheer mountainous terrain:
And when the Russians employed the argument so much favoured by
later dictators, that they were producing order from chaos and
improving the lamentable state of the roads . . . Shamil had his answer
ready: ‘‘You say my roads are bad and my country impassable. It is
well: that is the reason why the powerful White Tzar and all his armies
49
who march on me ceaselessly can still do nothing against me. I do not

compare myself to great sovereigns. I am Shamil — an ordinary
Avar.’’ (Blanch, 1960: 129)
In effect, the mountains of Daghestan proper served as a sort of inner

keep — a virtually unassailable final barrier to conquest — while the
hills and dense forests of Chechnia served as a kind of outer wall or
moat — considerable barriers to conquest in their own right (Baddeley,
1908: xxxv).
The lack of roads and the density of Chechnia’s beech forests severely
restricted the utility of Cossack light cavalry, while at the same time
exposing any line of march to an endless series of ambushes. In order to
attack and defeat mountain strongholds, the Russians had to bring up
sappers and heavy artillery. Yet powder, shot, and field pieces required
considerable logistical resources, and these in turn required heavy
security. The Russians were slow to adapt, yet adapt they did, though
with mixed effect:
In front of the advance guard, behind the rearguard, and on both
flanks of the column for its full length went the sharpshooters, with
their reserves and mountain guns. On the level or on open places these
flanking lines or chains kept at a good musket shot from the column,
but on entering a forest they marched as the ground permitted, striv-
ing as much as possible to keep the enemy’s fire at a distance, for it was
too deadly when directed at a compact body of troops. The soldiers
called this ‘‘carrying the column in a box.’’ (Baddeley, 1908: 269)2
Give up the artillery and supplies, and you increased mobility, but
upon reaching the mountain strongholds your troops would find it
impossible either to advance or to hold their ground. In the end,
Russia adapted to the difficulties of Caucasian terrain in two ways:
first, by cutting trees,3 and second, by building roads and forts.
2
Note that this ‘‘column in a box’’ strategy is the precise analog of the World War I/World
War II naval convoy strategy, with the Chechens taking the place of ‘‘wolf packs,’’ and the
beech trees taking the place of the opaque North Atlantic. The disadvantage of both
systems was also similar: more and more combat units had to be tied down to escorting
logistical assets; and, worse still, there was simply no glory in the monotonous drudgery
of escort duty.
3
Commentators have argued, in many ways convincingly, that Chechnya and Daghestan
fell to the axe, not the sword. The strategy of terrain alteration is a logical, if drastic,
counterinsurgency strategy. Yermolov’s tree-cutting policies, amounting to systematic
deforestation, anticipate the later US strategy of defoliating large areas of South and
North Vietnam with Agent Orange; as well as Saddam Hussein’s drainage of the southern
swamps of Iraq in order to contain and destroy the Shiite guerrillas whose insurgency was
based there.
50
Yet for all their defensive advantages, the mountains and forests did
not constitute an unambiguous asset. On the contrary, as Baddeley
argues, the difficulties of terrain and climate exacerbated the disunifying
effects of ethnicity, language, and the adats (Baddeley, 1908: xxii).
According to Baddeley, the Murids would win so long as they were
able to maintain a politically united front, and alliance defections would
ultimately doom them. If true, the terrain of Chechnia and Daghestan
presented Shamil with a military advantage but a political disadvantage.
Background
Russia’s conquest of the Caucasus took place in three broad phases. The
first occurred during the reign of Peter the Great, following immedi-
ately upon the heels of his successful war with Sweden in 1721. Taking
advantage of an Afghan invasion of Persia, Peter himself led the expe-
dition south, and his success resulted in the transfer of sovereignty of
various administrative districts from Persia to Russia. This phase fea-
tured very little military conflict, yet such conflict as did take place
foreshadowed what was to come: both in the forests of Chechnia and
the mountains of Daghestan, the Russians experienced ‘‘difficulties’’
(Gammer, 1994: 2). Russia’s success in this first phase was due to the
fact that its conflict was mainly with Persia, not with either the
Chechens or the Avars (the main ethnic tribe of Daghestan):
The whole thing was accomplished in a few months, and in the joy of
victory the Russians failed to notice how tremendously pleased the
Persians were that they could hand over to the Russians this restless,
unprofitable, and, in the last analysis, entirely independent land of
brigands. (Essad-Bey, 1931: 213)
The second phase took place during the reign of Catherine the Great.
In 1763 she established the fortress of Muzlik in the heart of Kabardia;
and after fourteen years of hard struggle, Kabardia was conquered and
occupied by a new Cossack regiment, the Mozdok Cossacks. Other wars
of expansion were fought during Catherine’s reign — Georgia was briefly
occupied, then abandoned — but Russian activity in the Caucasus
remained half-hearted until just before her death, when the barbarous
sacking of Tiflis provoked her into declaring war on Persia. The Russians
defeated the Persians everywhere, but Catherine died during the war
and her son Paul withdrew Russia’s forces from the Caucasus, intending
to relinquish all Russian possessions there. He was unsuccessful,
51
however, and after his death in 1800 his son Alexander agreed to accept a
dying Georgian monarch’s offer ceding Georgian sovereignty to Russia.
It is only in 1800, therefore, that historians mark the beginning of
Russia’s sixty-year-long attempt to conquer the whole of the Caucasus.
The annexation of Georgia, like the later annexation of Manchuria by the
Japanese in 1933, provoked a series of escalating military commitments
which contained an inexorable expansion logic: in order to hold on to what
we have, we have to expand to acquire more resources; but the resources
we expend to expand make our new possessions tenuous; and so we must
expand again just to hold on to what we have (Snyder, 1991: 3—4, 8).
The period from 1800 to 1815 marked a difficult time for Russia in the
Caucasus, but especially in the west. Russia’s de jure annexation of
Georgia proved difficult to achieve de facto; and the Empire faced war
with the Ottomans, Persia, as well as an invasion by Napoleon’s France
during this time. Not until the defeat of Napoleon was Russia once
again able to turn its sustained attention to its southern flank.
The third and final phase of the conquest of the Caucasus thus begins
roughly in 1816, and lasts until the capture of Shamil in 1859. This war
eventually came to absorb the complete attention of the Russian
Empire; along with the lives of many of its most talented officers, one
of its most beloved poets, and hundreds of thousands of its serf soldiers.
It is this third phase of the attempted conquest that constitutes the
central focus of this chapter.
Prologue to war: Yermolov, Veliaminov, and barbarism

When Mikhail Yermolov acceded to command of the Army of the South in
1816, his name was already a legend among Russians. Not only was this
gargantuan and leonine Russian general revered for his dash and bravery
in numerous Napoleonic engagements (he had been awarded the Cross of
St. George at the age of 16, and was promoted to colonel after the battle of
Austerlitz in 1805), but he is forever remembered as the most beloved of
commanders by the soldiers he commanded (Baddeley, 1908: 95).
Larger than life, Yermolov devoted himself unstintingly to his tsar’s
will and to the glory and security of Russia. Baddeley sums up his goal
as follows:
Yermóloff’s central idea was that the whole of the Caucasus must, and
should, become an integral part of the Russian Empire; that the exist-
ence of independent or semi-independent States or communities of
any description, whether Christian, Mussulman, or Pagan, in the
52
mountains or on the plains, was incompatible with the dignity and

honour of his master, the safety and welfare of his subjects.
(Baddeley, 1908: 99—100)
What was his strategy for achieving this end? Yermolov was hardly the
first or last to come upon what he considered an ideal counterinsurgency
strategy: barbarism. The idea was to divide the region to be conquered into
smaller regions, each of which would be attacked and razed piecemeal:
Baulked, the Russian generals decided on a ruthless system of

reprisals. The native population must be broken, crops and villages
razed, wells choked, orchards cut down, vineyards trampled. They
must be brought to heel. No quarter must be given — not that any was
ever asked by the fighters — and the campaign must be brought to a
close without further delay. (Blanch, 1960: 23)
Yermolov’s strategy was simple: to get at the inner keep (the mountain
strongholds of Daghestan) it was first necessary to subdue the
Chechens. To that end he began by building a fort called Grozny —
literally, ‘‘menace.’’ Yermolov intended to use Grozny as a base of
operations for a succession of raids into Chechnia proper. These raids,
as noted above, had the aim of conquering Chechnia by destroying it.
The Chechens would either submit to Yermolov and become subjects of
the tsar, or they would suffer and die: every man, woman, and child.
For a time it seemed as though Yermolov would succeed. He felled
trees, built roads, and completed a line of fortresses stretching from the
Caspian to the Black Sea. He razed villages, blasted mountain fortresses
to rubble, and forced tribe after tribe to submit to Russia. So great
seemed his momentum that, by the end of June in 1820, he declared to
the tsar ‘‘the subjugation of Daghestan, begun last year, is now com-
plete; and this country, proud, warlike, and hitherto unconquered, has
fallen at the sacred feet of your Imperial Majesty’’ (Yermolov as cited by
Baddeley, 1908: 138). Though few could have guessed it at the time his
assessment was to prove premature. Success followed on success, and
not until 1824 did Yermolov begin to pay heed to rumors of religious
fanaticism. Having ‘‘conquered’’ Daghestan, his command witnessed a
spontaneous religious uprising in Chechnia.
In December, 1825 Alexander I died unexpectedly, and Yermolov,
thinking the tsar’s brother Constantine would accede to the throne,
swore his and his command’s allegiance to Constantine. Unbeknownst
to Yermolov, however, Constantine declined to accept the crown, and it
passed instead to the tsar’s son Nicholas. Yermolov’s premature
53
commitment to Constantine woke Nicholas’s suspicions, however.

Although allowed to continue in the command for another year, the
accession of Nicholas I effectively marked the end of Yermolov’s career.
Yermolov was eventually dismissed in disgrace; forced even to beg for
his own escort home from the Caucasus. His legacy to the contest for the
Caucasus is a mixed one. On the one hand, he and Veliaminov invented
the Caucasian Corps system, and they built a chain of fortresses which
were to serve as crucial bases for the later conquest of Chechnia and
Daghestan. On the other hand, his barbaric strategy directly paved the
way for the rise of Muridism and Shamil, and for the historically unpre-
cedented unification of the Caucasian peoples into a solid fighting
front — a front implacably opposed to Russian sovereignty.
Russia’s interests
During the Murid War the sprawling Russian Empire was ruled from
St. Petersburg by two successive tsars: Nicholas I from 1826 to 1855, and
Alexander II from 1855 to 1859. Although individual tsars tended to
represent opposite political and social philosophies — father seeking
reform, son repression, grandson reform, and so on — all the tsars in
question shared an absolute secular and religious authority during
their respective reigns.
Russia’s great advantage lay in her own system of government, that
autocratic power which, coupled with the existence of serfdom,
enabled her to fill the ranks of her armies at will, and, yet more
important, to secure her conquests by a vast system of land settlement
on the Cossack principle of military tenure. (Baddeley, 1908: 236)
This is an argument worthy of elaboration, because in many ways

Russia constituted a medieval regime. Only a tiny minority of its
citizens were literate, and the serf system amounted to slavery. In effect,
Russia at that time had no real ‘‘public’’ which could exercise an inde-
pendent influence on its foreign or military policies. There was no
centralized collection or widespread dissemination of casualty or battle
reports. Finally, to the extent that Russia’s social and political elites
were engaged in the Caucasian campaigns, it is fair to say they were
mostly engaged romantically.4
4
No more visible or durable evidence of this romanticism is needed than Mikhail
Lermontov’s poems and stories of the Caucasus (see Lermontov, 1966).
54
Russia’s regime type was the very definition of authoritarian. And

Russia’s interests in the Caucasus were simple: conquest, annexation,
conversion, and pacification. These interests had not changed signifi-
cantly since the reign of Peter the Great, and they did not change during
the course of the war.
Murid interests
In 1816 Russia’s adversaries in the Caucasus could best be described as
independent tribes and Khanates, each of which resisted Russian
advances independently. Although most Christian regions eventually
were bribed or browbeaten into swearing loyalty to Russia, Russia’s
Muslim regions and Khanates were generally dealt with by military
means (Gammer, 1994: 6). Regardless of faith, language, or loyalty,
however, these areas were generally governed by hereditary ruling
families. In formal terms, they were authoritarian regimes; though it
should be added that until the advent of Muridism, no single tribe’s
ruling family enjoyed a degree of authority comparable to that of the
Russian tsars. Instead, rulers exercised an authority restricted by
ancient laws called adats, which among other things specified appro-
priate rewards and punishments and codes of conduct for the sexes.
The adats also institutionalized such practices as kanly or blood feud,
which again, until the advent of Muridism, made unified resistance to
Russian advances all but impossible.
After 1830, however, the non-Christian regions and Khanates of
Chechnia and Daghestan came under the sway of an absolute ruler,
who like the tsar claimed both secular and religious authority. Khazi
Muhammad was acclaimed first imam of Daghestan in 1830, and his
creed has been called Muridism (see below). Khazi Muhammad was
killed in battle at Gimri in October of 1832 (he was survived, in that battle,
‘‘as if by a miracle,’’ by a grievously wounded Shamil). In November of
1832, Hamzad Beg was proclaimed second imam of Daghestan. Like his
predecessor, he spread the Murid creed, and his military actions were in
the main confined to consolidating support among wavering Caucasian
tribes. Yet Hamzad Beg, whose treachery was widely blamed for Khazi
Muhammad’s fall at Gimri, was himself cut down by treachery; and in
September of 1834, Shamil was acclaimed third imam of Daghestan. It was
a title he was to hold for twenty-five years. It is no exaggeration to say that
in terms of absolute authority Shamil’s leadership very quickly equaled
that of the tsar who was to constitute his main opponent, Nicholas I.
55
Prior to the advent of Muridism, the Caucasian tribes fought tacti-

cally brilliant but strategically uncoordinated engagements with the
Russians. Their interests in such attacks were gain or personal fame.
After the advent of Muridism, Shamil organized the Murids into a
sophisticated order of battle based on a system of tens. Ten ‘‘greater’’
naı̈bs (famous leaders and devoted followers, the equivalents of gen-
erals) commanded 100 ‘‘regular’’ naı̈bs (unfailingly loyal, the equiva-
lents of colonels), who in turn commanded 1,000 lesser naı̈bs or
murshids (skilled tacticians, the equivalents of captains). Enlisted
men and officers alike were motivated by the same burning passion:
to free their lands of the infidel, where the term itself combined a
secular (ethnic and political) and a religious (non-Muslim) meaning.
Instead of gain and fame being the goal of engagements, they became
by-products of service to God. The terms of that service were defined by
the imam, the mouthpiece of God:
Muridism was a heady brew of mystic and absolute power, though even
after it had become, to the Russians, synonymous with resistance, it was
variously interpreted: there were the Murids of the Tarikat who never
took up arms, as opposed to the Murids of the Ghazavat who fought a
Holy War fanatically. To these last there was no other interpretation of
the Prophet’s teachings. If to live in peace meant submitting to the
Infidel rule, there could be no peace. While the Tarikat abhorred vio-
lence and, in the face of force, counseled a withdrawal to some inner
spiritual sanctuary, this was not a doctrine which came easily to the fiery
Caucasian tribes. Most of them felt that, in this issue, the Tarikat must be
modified, or adapted, to meet the more bellicose tenets of the Koran,
which promised short shrift to an Infidel foe. (Blanch, 1960: 58—59)
This analysis is remarkable in its reminder that, as a religious phil-

osophy, Muridism in and of itself did not demand armed resistance. It
is therefore not the arrival of Muridism in Daghestan, nor its revival in
the 1820s, which explain the waxing resolve of the Caucasian tribes to
resist imperial Russia. It is the winning out of a particular interpretation
of Muridism by successive imams which made Muridism the potent
force it became — a force which made political unity possible, bribery
difficult, and self-sacrifice desirable.
Shamil’s accession therefore marks both a shift in the quality of the
forces opposed to Russia, and a shift in their interests: from fame, glory,
and loot, to the ejection of infidel Russians from the territory of the
Caucasus and the establishment of a theocracy to govern the lands from
the Black Sea to the Caspian Sea.
56
The Murid War

The Russian Empire’s population in 1816 is given by Singer and Small
as approximately 51 million, of which 800,000 were under arms
(Singer & Small, 1983). Gammer gives the total number devoted to the
Caucasus at 30,000 in the 1830s, escalating to 200,000 in the 1850s
(Gammer, 1994: 24).5 There are no satisfactory population and armed
forces data for the Caucasian tribes,6 though Gammer gives them at
200,000 and 40,000—50,000 respectively (Gammer, 1994: 22). There is
therefore no controversy regarding the relative material power of the
two actors: Russia was clearly the strong actor, and the Murids were the
weak actor.
The Murid War featured three different strategic interactions. The
first interaction, from 1827 to 1845 coincides with the accession of
Paskyevitch and Grabbé to command of the Southern Front. Here the
overall Russian strategy was barbarism punctuated by direct attacks
on Murid strongholds. The Murids pursued a GWS throughout. The
second interaction, from 1845 to 1855 coincides with the accession of
Voronzov and Freitag to command. The beginning of this period fea-
tured a Russian direct attack strategy against a Murid GWS, though it
ended with a return to barbarism. Finally, the third interaction, from
1855 to 1859 coincides with the accession of Bariatinsky and
Yevdomikov to command. By this time, the main Russian strategy
had switched to direct attacks and conciliation while, again, the
Murids maintained a GWS.
Interaction one: barbarism/conventional attack vs.

GWS
Paskyevitch’s tenure began uncertainly. He inherited the beginnings of
a nationalist-religious uprising which had for the first time shown that
the Russians could be defeated. The start of his command thus
5
Gammer doesn’t specify a time period here, but he presumably means during the 1830s.
He also notes that more than 80 percent of these troops were regular army units (see
below). Note that the lack of any change in relative power, combined with the large shift in
the number of troops the Russians were willing to the fight, already suggests that Mack’s
proposed link between relative power and relative interests is weak at best.
6
And by ‘‘Caucasian tribes’’ I mean Chechnia and Daghestan only. As will become clear
below, these two areas constitute the geographic area during the phase of the asymmetric
conflict in which we are chiefly interested.
57
coincides with both the apparent conquest of Daghestan, and the rise of
Muridism (Baddeley, 1908: 237, 239).
The years 1827—30 passed quietly enough in terms of conflict,
although, all during this time, the first Murid imams were discreetly
gathering strength and adherents throughout the Caucasus. By 1831,
however, Khazi Muhammad, first imam of Daghestan, had rallied a
considerable body of dedicated troops and begun raids against tribes
allied to Russia. His Murids experienced some successes, but overall
their first attempts proved a failure. Muhammad was a learned and
holy man: he simply lacked the charisma and strategic skills needed to
exploit victories and maintain allies.
What emerges from accounts of battles in this early period is the theme
of the importance of reputation in attracting alliance support. The
Caucasus of the 1830s was a bandwagoning world:7 success meant allies,
defeat meant isolation.8 Khazi Muhammad’s final defeat came in an heroic
defense-to-the-death of Gimri aôul, in which he lost his life and, though
‘‘mortally wounded,’’ Shamil, his student and chief lieutenant, escaped.
The Russians responded to this religious threat in Chechnia and
Daghestan by renewed barbarism:
Greater Tchetchnia, in turn, was devastated with fire and sword, and a
hatred sown and watered with blood, the traces of which are still
visible after seventy years. (Baddeley, 1908: 275)
The ultimate effect of such policies was to guarantee the survival of

Muridism of the Ghazavat. Russia had not yet gained the capacity to
prosecute barbarism on a scale that could damage a Murid insurgency.
Shamil recovered from his wounds in a short time and Russia’s puni-
tive raids guaranteed a ready audience for his message. From 1833
to 1837 the Murids grew stronger and stronger, uniting more and
more tribes and forever demolishing the image of Russians as militarily
invincible. Even Russian military victories only strengthened Shamil’s
grip (Baddeley, 1908: 304—305). Again and again, Russia’s COIN
strategy — kill everyone and destroy everything — backfired.
7
Bandwagoning means that the more allies you have, the more you will attract; and the
reverse is true also, the defection of any ally implies the defection of every ally. On
bandwagoning and its interstate implications, see Walt, 1987: 19—21.
8
This is the essence of the Guevara/Castro foco theory: in opposition to Mao’s maxim that
social organization must precede military adventures, Castro and Guevara held that in
Latin America (Cuba specifically), military successes were a necessary precondition of
social support. See e.g., Blaufarb and Tanham, 1989: 12.
58
In 1839, Shamil again escaped almost certain capture after the

Russians surrounded and defeated his forces — after frightful losses
on both sides — at Akhulgo. By 1840, however, he was once more on
the offensive; and from 1840 to 1842, Shamil’s Murids made themselves
feared by Russia’s regiments and allies. Shamil was careful during this
period to avoid pitched battles; choosing instead to attack only weak or
isolated forces, and withdrawing before help could arrive. From 1843 to
1844 it was therefore Shamil, and not the Russians, who enjoyed the
military initiative (Baddeley, 1908: 364).
Paskyevitch, whose strategic plans favored the defensive, tended to
play into Shamil’s hands. Yet the Murids were not the only ones learn-
ing important lessons. On the Russian side, a talented group of Russian
junior officers were learning both how to fight, and how not to fight, the
Murids. In 1845 Paskyevitch was replaced by Count Voronzov.
Interaction two: conventional attack vs. GWS

Arriving fresh from Nicholas’s war room with new reinforcements and
an impossibly ambitious battle plan, Voronzov’s inquiries quickly con-
vinced him of the folly of the tsar’s strategy: an all-out offensive aimed
at capturing or killing Shamil himself. Yet he was too faithful a servant
to challenge his master’s will. It was thus almost with a sense of pre-
destined doom that Voronzov set out to capture Shamil in the very heart
of Daghestan:
If God is not pleased to bless us with success we shall nevertheless
have done our duty, we shall not be to blame, and we can then turn,
somewhat later, to the methodical system which will bear fruit, though
of course not so quickly as a victory over Shamil himself.
(Voronzov as quoted by Baddeley, 1908: 387)
Shamil had by this time acquired full mastery of the strategic and
tactical requirements of defense against such adventures. His strategy
was again a GWS: avoiding direct confrontation with the enemy on his
outward march, abandoning even fortifications if necessary to avoid
contact, and then attacking the retreating armies when supplies and
morale were low, and when exhaustion and frostbite had weakened
them:
His opportunity would come later on when Nature, his great ally, had
done her work, and the invaders, worn with toil, weak from privation,
uninspirited by successes in the field, would have to face the
59
homeward march over the barren mountains of Daghestan or through

the forests of Itchkéria. Then indeed he would let loose on them his
mobile hordes, break down the roads in front of them, seize every
opportunity of cutting off front or rear guard, of throwing the centre
with its weary baggage train and lengthy line of wounded into confu-
sion, and give the men no rest by day or by night.9
(Baddeley, 1908: 388)
In the event, Voronzov was only saved from complete annihilation by

the bravery and foresight of General Freitag who, experienced in
Caucasian warfare, had predicted almost exactly the fate of the exped-
ition, and prepared in advance to relieve his superior upon the first hint
of danger. This is what happened.
On June 15, Voronzov, opened his campaign to catch Shamil with
21,000 men, and 42 pieces of artillery (Gammer, 1994: 153). Voronzov
chased the Murids to Andi Gates, where his spies had assured him
Shamil would make a stand. He found the site abandoned and in ruins,
its population already evacuated by Shamil. After waiting three weeks
at the ruined site, Voronzov then had to decide whether to push ahead
in the hope of catching Shamil at another fortress (Dargo), or to return
empty handed to Tiflis (Tblisi). Voronzov decided to advance, but the
difficulties of terrain, combined with the increasing drain on supplies,
soon forced a halt. He compounded his error by sending a body of
troops — half the combat strength of each unit in his order of battle —
back along the march route to gain additional provisions and evacuate
his wounded. This so-called ‘‘Biscuit Expedition’’ proved a disaster. All
along the route, long stretched-out columns of Russian soldiers became
isolated. Following Suvorov’s maxim that ‘‘the head doesn’t wait for the
tail,’’ impetuous Russian vanguards rushed ahead and were cut off
from supply units carrying wounded. The Murids then attacked and
annihilated the isolated columns piecemeal. As the wounded multi-
plied, the columns slowed still further.
9
Shamil’s strategy was the same as that used by the Russians against Napoleon during
his invasion of Russia. Then it was the Russians who avoided confrontation, destroyed
crops in the path of the invader, and waited until ‘‘General Winter’’ had weakened and
dispersed French formations. Yet here, scarcely thirty years later, was Vorontsov playing
the part of Bonaparte, marching toward almost certain destruction at the head of an
overconfident, overburdened army. Moreover, Vorontsov had not arrived from St.
Petersburg unaccompanied: on the contrary, his reputation and fame provoked dozens
of Russia’s noble sons to flock to his standard in search of glory in the Caucasus. To the
hostile and snowy wastes of Daghestan, these dandies brought elaborate tea services,
personal slaves, plush tents and carpets, meticulously tailored uniforms, brandies, phea-
sants, and fine cigars.
60
The expedition returned, but it cost Voronzov dearly. Only after reach-
ing Shamil’s capital, Dargo, and finding it, like Andi Gates, abandoned
and in ruins, did the magnitude of Voronzov’s position become clear to
him: isolated deep within the mountains, surrounded on all sides by
nimble and well-supplied foes, he realized that even to regain the foot-
hills would take a miracle. Quickly he gave the order to withdraw, and
the entire formation began the retreat. Almost unbelievably, Voronzov’s
withdrawal was plagued by the same mindless adherence to maxims
which nearly destroyed the Biscuit Expedition. Once again, the columns
became separated. Once again, the nimble Murids cut the isolated for-
mations to ribbons. As casualties mounted and food and ammunition
dwindled to nothing, Voronzov sent five couriers to seek help from
General Freitag. Amazingly, all five couriers reached Freitag, who set
off without delay to relieve Voronzov. Gammer gives Voronzov’s losses
at 984 killed (including three generals), 2,753 wounded, 179 missing, 3
guns, a great sum of money in coins, and all the expedition’s baggage
(Gammer, 1994: 156). Freitag’s foresight saved Voronzov’s expedition;
and Voronzov, true to his earlier pronouncement, then set about organ-
izing a more methodical destruction of Shamil and the subsequent con-
quest of the Caucasus. Over the next decade, this involved nothing more
complicated or less deadly than the deforestation of Chechnia.
The year 1845 was to mark the high point of Murid power. In 1846,
Shamil attempted a conventional military offensive against Russian
positions in Kabardia. His hope was that Kabardia’s people would
rise up spontaneously in support of Muridism, and his bold stroke
would completely sever Russian communications while leaving the
Murids in control of a continuous geographic space between the
Caspian and Black Seas. In the event, Shamil failed, mainly due to the
bravery and doggedness of Freitag. Freitag, fully alerted to the strategic
consequences of failure, immediately set off in pursuit of Shamil’s
invading Murid cavalry with a woefully inadequate hodge-podge of
garrison troops, Cossacks, and regular army formations. Shamil, isol-
ated from his vaunted intelligence resources, did not realize that the
force pursuing him was so weak. Neither, it turned out, did the
Kabardians, who waited to see the outcome of the contest before declar-
ing their support. After unexpected resistance at a Kabardian fort,
Shamil faced the prospect of being trapped between a stubborn fort,
and a ‘‘relief force’’ headed by the much respected Freitag. He with-
drew, and later set about fulfilling his promise of prosecuting the steady
and methodical conquest of the Caucasus.
61
Interaction three: conciliation vs. GWS

Into this final period, from 1854 to 1859, marches the last of the famous
names among Russia’s commanders in the Caucasus: Bariatinsky.
Voronzov had set about undermining Shamil’s power base in two
ways. First, he accelerated the deforestation of Chechnia. Second, he
fostered the concentration and development of Russia’s artillery.
Cleared forests and wide roads made way for the heavy guns, which
were then positioned, with frightful losses, so as to batter the Murid
strongholds into dust one by one. Moreover, it finally occurred to the
Russians that the source of Shamil’s foodstuffs, horses, ammunition,
and powder was in fact Chechnia. Voronzov therefore tasked
Bariatinsky with systematically interdicting the Murids’ logistical sup-
port, which further weakened Shamil.
Without sufficient supplies, Murid offensives became more and more
difficult to mount. This victory drought led in turn to a weakening of
alliance support, and in this period once staunch allies began to defect
to Russia: at first a trickle, then later a flood.10
Bariatinsky, who eventually replaced the aging Voronzov, added the
final blow by initiating a heretofore unprecedented policy: clemency.
This was a policy the Murids did not expect, and it was one to which
they proved powerless to respond.
Three things brought about Shamil’s defeat. The cumulative force of

Russian arms . . . the leadership of Prince Bariatinsky, who was
appointed supreme commander in place of the old ailing Voronzov;
above all, Shamil’s cause was lost by internal dissensions . . . Terror
had always been [Shamil’s] strongest weapon: it was ever the guiding
force in Asiatic warfare. Now Bariatinsky’s clemency was proving the
stronger weapon. The tribes feared Russian guns — but not Russian
vengeance. They had discovered their enemy to be terrible in battle,
but magnanimous in conquest. So they ceded over in their thousands,
leaving Shamil and his remaining Murids isolated, a diminishing
force, withdrawing each day farther into the high mountains.
(Blanch, 1960: 390—391)
10
This implies that a sine qua non of effective GWS is a strong degree of social organization
which makes bribery and coercion ineffective. The defection of the tribes suggests that
Muridism was no substitute for nationalism as we have come to know it in the post-World
War I context. For a general discussion of the relationship of nationalism to armed
resistance, see Wolf, 1973.
62
Bariatinsky’s clemency strategy was unconventional from the

Russian point of view, and it proved effective. Although Shamil lost
fewer troops in every engagement than the Russians, his ability to
replace these losses constantly diminished. Russia could always draw
on resources far beyond Shamil’s reach; whereas Shamil’s resources,
human and material, were constantly and increasingly within Russia’s
grasp; a fact which, once fully appreciated by Russia, was turned to
such advantage that it ultimately led to Shamil’s defeat.
It is very likely, in other words, that had the Chechens not already
been weakened by decades of punitive raids, the Russians would never
have been able to cut the trees. With the forests intact, it would have
been as difficult as ever to bring heavy artillery to bear on the moun-
tains. Finally, Bariatinsky’s clemency may have been taken for weak-
ness, as Yermolov had previously warned.
Outcome: Russia wins

The final Murid fortress, Gounib, fell on August 25, 1859. Shamil
surrendered, and Blanch argues that he did so this time because only
in this way could he guarantee the safety of his family (Blanch, 1960:
408; Gammer, 1994: 286).
Thus did the Russians finally conquer the Caucasus after nearly sixty
years of costly struggle. During the course of the conflict Russian
military strategy varied depending on the philosophy of the general
in command of operations there, as well as the particular tsar ruling in
St. Petersburg. After the Napoleonic Wars, the Southern Front’s first
commander, Mikhail Yermolov, introduced barbarism: a strategy that
demanded the use of methods proscribed in European warfare, such as
the killing of non-combatants, torture, rape, and the despoiling of wells,
crops, and livestock. These things of course happen in every war, but
Yermolov made them policy. These were adopted by Paskyevitch, and
resorted to later by Voronzov as well.
Yet, in every case, barbarism’s effectiveness proved mixed. For every
short-term military advantage gained, the Russians suffered a corres-
ponding political cost. Ultimately, barbarism made it impossible to coerce
the Caucasian tribes; and if they couldn’t be coerced, they could only be
destroyed; and that only at a tremendous cost in lives and treasure.
The three interactions coincide with the fortunes of each side, and
explain as much or more than any other factor, the outcome of the war.
Yermolov’s barbarism worked militarily. Its impact was such that he was
63
able to declare the subjugation of the tribes in 1825. Russia employed an

indirect strategy and the tribes had responded with an indirect strategy.
Yet Russia had not yet developed the logistical reach necessary to make
the barbarism decisive. As a result, the barbarism Yermolov and
Paskyevitch unleashed against the Caucasian tribes — isolated and deva-
stated — left smoldering embers. The Murids — Khazi Muhammad,
Hamzad Beg, and Shamil, successively fanned those embers into an
inferno which nearly succeeded in ejecting the infidel Russians from
the Caucasian theater. The greatest successes of the Murids coincided
with the period of greatest strategic opposition: Russia employed a direct
strategy, and the Murids countered with an indirect strategy. Yet the
Russians did not give up. They switched strategies yet again to the
indirect strategy of conciliation, to which the Murids had no other
answer than their own indirect strategy, GWS.
Bariatinsky’s conciliation strategy deserves special mention for
two reasons. First, unlike all previous strategies, its aim was more to
end the war than to win it. Bariatinsky was the first commander to treat
his enemies as warriors and men instead of bandits and savages. He
sought to persuade human beings, rather than destroy animals. He
treated prisoners in accord with European standards, abolished torture
and rape, and provided food and clothing to the women and children of
his defeated adversaries. Second, such a policy would have been
impossible under Nicholas I. Like his grandfather, Alexander II was a
reformer and a firm believer in the power of kindness toward a
defeated enemy.
Analysis: competing explanations of the Murid

War’s outcome
The Russian Empire won the Murid War, and its victory is in most ways
a perfect representative of asymmetric conflict outcomes during the
first fifty years of the 200-year period of examination that comprises this
book’s analytical interest. In keeping with the spirit of the times, it was
widely viewed as a conflict which pitted a civilized nation against
‘‘unruly savages,’’ and, as such it was one in which few limitations in
the quantity or quality (barbarism) were observed. Before comparing
the relative explanatory merits of nature of actor, arms diffusion, and
interest asymmetry explanations of the outcome, two additional points
should be established.
64
First, relative power appears to have overwhelmed the effects of all

competing arguments, including even strategic interaction. That is, the
overwhelming advantage of the Russian Empire — and, most import-
antly, its unflagging resolve to employ those resources for an indefinite
period — provides the most satisfying and intuitive explanation of the
war’s outcome. Second, this extreme cost insensitivity was only made
possible by the fact of an absolute ruler and an illiterate population. The
distribution of both conditions across time and space is limited, which
makes this case a tougher test of competing explanations of asymmetric
conflict outcomes.
That said, there are a great many lessons to be learned from how the
war was fought independent of its outcome. Strategic interaction
explains why victory cost the Russian Empire so much more — in
time, blood, and treasure — than any other great power could have
been willing to pay under similar circumstances. Regime type proved
vital for both strong and weak actors, arms diffusion played a small and
unexpected role, and leadership, terrain, and climate were also import-
ant parts of the explanation of this particular war and its outcome.
Interest asymmetry is not much use in explaining this war.
Assessing the relative explanatory merits of competing explanations
of asymmetric conflict requires answering four basic questions. First,
what were the interests of the actors prior to the conflict and how are
these best explained? Second, what was each actor’s regime type, and
how did this affect strategy and political vulnerability? Third, what role
did military arms diffusion play in the conflict’s outcome? Finally, what
were the strategies each side employed, what was the rationale given
for their selection or change, and how did they affect the eventual
outcome of the war?
Actor interests
Russia’s interests were not explained by its material power relative to
that of the Murids. Russia’s interests appear to have been settled in the
geopolitical context of the late eighteenth century, and to have simply
continued momentum after that. In the context of Russia as an empire,
its interests must best be described as expansion. Less obviously, the
fact that Russia’s tsars were both secular and religious leaders meant
that opposition to the will of the Tsar had the aspect of opposition to
God — especially when the opposition came from non-Christians. The
very existence of resistance therefore became a casus belli.
65
The Murids fought for their survival, but they also maintained the
positive goal of ejecting Russia from the Caucasus. The negative goal
could be explained by their relative material power, but not the positive
goal.
Regime types, strategy, and vulnerability

As noted above, Russia’s regime type was virtually unique in its day: an
essentially feudal empire ruled by an autocrat. How did this affect
Russia’s strategy and political vulnerability?
The effects on strategy were mixed. On the one hand, debates among
Russia’s political elites revolved around two questions: (1) when would
the tribes be subdued?, and (2) how many men would it take? The
question of strategy was largely left up to Southern Front commanders:
Yermolov, Paskyevitch, Voronzov, and Bariatinsky. In any case, authori-
tarian Russia had no need to consult Russian public opinion before
committing resources to the fight. Nor would it be held accountable for
failure, delay, or barbarism. On the other hand, the personal prefer-
ences of the tsars did affect strategy in two cases. First, Nicholas sent
Count Voronzov to the Caucasus with both massive reinforcements and
a specific plan and timetable for their use. Second, after Nicholas’s
death, his son Alexander II made it clear that barbarism would not be
a strategic option. He fully supported Bariatinsky’s use of clemency,
bribes, and concessions in order to demobilize Shamil’s remaining
allies.
In terms of political vulnerability, Russia’s regime type effectively
precluded this. According to the interest asymmetry argument, we
should expect to see even authoritarian regimes subject to trade-off
constraints when involved in prolonged conflicts. The practical result
is a similar degree of political vulnerability as that expected in demo-
cratic regimes. But do we see it here? In fact, Russia was involved in two
potentially serious interstate conflicts during the Murid War, and
neither caused the expected vulnerability.
Its first fight was with Persia in 1826, just prior to the advent of
Muridism. The Persians invaded from the south, and threatened to
capture Georgia. Fought on conventional terms, the ponderous
Persians were time and again defeated by much smaller Russian for-
mations. The second potential source of Russian vulnerability was the
Crimean War (1853—56). This war affected the conduct of the Murid War
in two ways. First, and most obviously, the troops mobilized to relieve
66
Sevastopol were transferred from the Crimea to the Caucasus immedi-

ately upon the war’s end.11 This meant a quick and significant influx
of fresh troops for Russia. Second, and more importantly, the outcome
of the Crimean War removed all hope of relief or support from either
Britain or the Ottomans. Gammer and Baddeley both argue that
this proved a crushing blow to Shamil’s morale, convincing him for
the first time that his war with Russia was hopeless, and eventually
infecting his forces with resignation (Gammer, 1994: 291; Baddeley,
1908: 459).
In short, we don’t see the trade-off’s mechanism in this case. There is
in fact no sense of political vulnerability either in Russia’s willingness to
fight in the Caucasus or its soldiers’ conduct there, at all.12
Merom’s democratic social squeamishness argument receives mixed
support. On the one hand, no comparable democratic regime (and few
authoritarian ones) could have sustained such costs in order to defeat
Shamil. On the other hand, the proximate cause of Shamil’s downfall
was not brutality — which had been tried before but had never proven
decisive — but kindness.
Arms diffusion
Military technology played a relatively minor role in the Murid War.
In the first interactions of the war, Russian military technology —
including firearms, artillery, and even clothing — were much more
poorly adapted to warfare in Chechnia or in Daghestan than that of
their adversaries. In this sense, their technology was less advanced. Yet,
as time passed, this situation improved to the point where Russian
military technology equaled that of their adversaries, and in artillery
far surpassed it.
As all the commentators agree, artillery proved to be the decisive
technology, but its effectiveness depended on very specific circum-
stances, and on strategy and doctrine (Baddeley, 1908: xxxii). For
example, the Russians generally launched campaigns from forts situ-
ated in low hills or plains near river systems. Even had the Murids
gained artillery capability (as they in fact did briefly), its utility against
the Russians in these forts would have required (1) considerable escort
11
And note that Russia’s defeat did not weaken the tsar’s authority or imply a change of
regime.
12
Again, see Baddeley, 1908: 236.
67
to counter the Cossack cavalry formations which would be operating in

their ideal environment, and (2) the capacity to advance against forts by
sap and mine — capabilities neither the mountain Murids nor the forest
Chechens possessed. When the Murids did attempt to employ captured
Russian artillery against the Russians, they met with disaster. First,
their lack of skill with the technology caused them to miss their targets,
but, more importantly, Shamyl’s insistence on hanging on to the guns
once the engagement turned disastrous cost the Murids heavy
casualties.
For their part, the Russians were only able to employ artillery effec-
tively after they had deforested Chechnia, and come to accept the high
casualty rates necessary to advance and position the guns to destroy
Murid fortifications (Blanch, 1960: 90).
If relative power doesn’t explain relative interests, and regime type
explains political vulnerability (or, in this case, its absence), what best
explains the outcome of the Murid War?
I argue that it is strategic interaction, but with an important caveat.13
Given the disparity in material resources between the two actors, and
given the constancy of their interests over time, it is remarkable that the
Murid War lasted twenty-nine years. As has been made clear in this
case, there are three reasons it took so long to subdue the Murids. First,
the terrain and climate made offensive operations very difficult, even
assuming the best leadership, training, and equipment. Second, there is
the problem of Russian arrogance and incompetence, both of which
caused the Russians to underestimate their adversaries. Third, and
most importantly, the Murids’ GWS systematically sacrificed fortresses
for time. When Yermolov prosecuted a barbarism strategy against the
Caucasian tribes, Russia won. When, after Paskyevitch, Voronzov
switched to a direct attack strategy, the Murids won. And when, finally,
Bariatinsky switched to a conciliation strategy, the Russians won. The
final strategic interaction, ‘‘same approach,’’ ended the war.
13
The caveat is that for authoritarian regimes with large populations, the implied lack of
cost sensitivity tends to overwhelm all other factors. Fortunately, the distribution of such
regimes is low: China and Iran may be the only contemporary analogues of nineteenth-
century Russia.
68
Conclusion
The Murid War was an asymmetric conflict between the Russian
Empire, and a coalition of Caucasian tribes under the banner of
Muridism.
Russian interests were higher than that expected by the interest
asymmetry argument. Russia wanted to conquer and annex the
Caucasus, and it was willing to spend hundreds of thousands of lives,
millions of rubles, and over half a century to accomplish this objective.
Russia’s dubious possession of Georgia proved to be one of two reasons
Russia refused to abandon its crusade in the Caucasus, even after
Catherine’s decision to leave Russia’s possessions there fallow, and
Paul’s wish to abandon them altogether.14 As an ally, Georgia’s con-
tribution to the war effort was neither inconsiderable nor decisive;
though it gave the Russians a huge strategic advantage, lying as it
did between another Muslim power, the Ottoman Empire, and
Russia’s Muslim foes. Access to Georgia’s ports on the Black Sea also
made it possible to ship supplies and reinforcements from Odessa or
Sevastopol.
Its only real strategic interest, therefore, lay in its desire to protect
Georgia. It is impossible to argue, however, that the possession of
Georgia itself materially or strategically benefited the Russian Empire
in any reasonable proportion to the resources Russia was to expend to
secure Georgia by conquering the Caucasus.
Murid interests were in establishing their political and religious
independence from Russia. To achieve this aim the Murids were willing
to fight to the end — and they did.
In sum, hypothesis 8 — relative material power explains relative
interests in the outcome of an asymmetric conflict — is not supported
in the Murid War. Relative power proved a poor predictor of relative
interests in this case — at least on the Russian side. Did relative interests
explain political vulnerability?
Russia was simply not politically vulnerable. According to the inter-
est asymmetry argument, its preponderance should have made it
14
The other reason was simple ethnic chauvinism: how dare these ‘‘ignorant savages’’
resist the will of the great white tsar? The language of both the commentators and original
sources makes this view of the Caucasians clear: they are constantly referred to as savages,
and their behavior is described as cunning where Europeans are described as brave or
brilliant. The resistance of these ‘‘uncultured children’’ to the tsar was therefore con-
structed as an affront to civilization itself.
69
politically vulnerable. The lack of a serious threat from the Caucasus

should have made even an authoritarian Russia sensitive to resource
trade-offs, and politically vulnerable in this sense. Yet, during the
Murid War, Russia faced two potentially serious interstate conflicts,
and neither caused the expected vulnerability. Strong actor regime
type therefore matters more than is allowed by the interest asymmetry
argument.
Is the nature-of-actor argument therefore a better explanation of the
outcome of the Murid War than strategic interaction? No. Regime type
remained constant on both sides, but the fortunes of each side varied
markedly during the war’s three interactions. Regime type had little
effect on the strategies chosen or prosecuted by either side. On the
Russian side, what may be said is that Nicholas often lent a direct
hand in strategic planning, and often with detrimental effects. It should
also be added that merit was hardly the basis for assignment to com-
mand in the Southern Front: on the contrary, Russian officers with
politically unpalatable ideas, or running from debts, or suffering from
broken hearts, often ended up in command in the Caucasus. On the
Murid side, what may be said is that Shamil’s authority for a time
transcended all ethnic and territorial cleavages to make possible a
sophisticated and effective GWS. Unlike the Russian army, advance-
ment and command in the Murid army was based exclusively on merit.
But, as can be readily appreciated, the differences between the two sides
are not artifacts of their regime types, which were the same.
Arms diffusion played almost no role in the Murid War, but its
limited impact was the opposite of that predicted by the logic of the
arms diffusion argument: the possession of modern artillery pieces by
the Murids led them to abandon their comparative advantages in
mobility and seriously weakened them. Hypothesis 6 — the better
armed a weak actor is, the more likely it is that a strong actor will lose
an asymmetric conflict — is therefore not supported in the Murid War.
In terms of strategic interaction, the war played itself out in three
interactions. During the first interaction, the strategic interaction was
same-approach: the Russians pursued a barbarism strategy and the
Murids a GWS. The Russians won militarily, but the political conse-
quences of this first interaction led to the rise of militant Muridism.
During the second interaction, the strategic interaction was opposite-
approach: the Russians pursued a conventional attack strategy and the
Murids continued their GWS. The Murids won. But, during the final
interaction, the strategic interaction switched to same-approach again:
70
the Russians pursued a conciliation (indirect) strategy and the Murids,

again, GWS. The Russians won, and Shamil went into exile. With
one important caveat, then, hypothesis 5 — strong actors are more
likely to win same-approach interactions and lose opposite-approach
interactions — receives considerable support. The caveat is that the out-
come of the war was determined more by the unwavering application of
unlimited resources than by strategic interaction. On the other hand,
the costs to the Russian Empire in time, lives, and treasure is best
explained by strategic interaction.
In sum, if we wish to explain why Russia won the Murid War, we
cannot rely on relative power alone. Relative power clearly mattered.
But what mattered most was strategic interaction.
71
4 Britain in Orange Free State and
Transvaal: the South African War,
1899–1902
They were conquerors, and for that you want only brute force —
nothing to boast of, when you have it, since your strength is just an
accident arising from the weakness of others. They grabbed what they
could get for the sake of what was to be got. It was just robbery with
violence, aggravated murder on a great scale, and men going at it blind —
as is very proper for those who tackle a darkness. The conquest of the
earth, which mostly means the taking it away from whose who have a
different complexion or slightly flatter noses than ourselves, is not a
pretty thing when you look into it too much. What redeems it is the idea
only. An idea at the back of it; not a sentimental pretense but an idea; and
an unselfish belief in the idea — something you can set up, and bow down
before, and offer a sacrifice to . . . Conrad, Heart of Darkness (1985)
The war between the British Empire and the governments of

Transvaal and Orange Free State which began on October 11, 1899,
was once called the Anglo-Boer War, or Boer War for short, but histor-
ians now call it ‘‘The South African War ’’ in recognition of the import-
ant role played in it by black Africans.
The South African theater

European interest in South Africa has its origins in the development of
trade contacts with India and the Far East. It was in this context that the
first major European economic and security interest — Cape Town — was
established by the Dutch East India company in 1652. The Dutch sub-
sequently lost and regained the town, and the colony, until the close of
the Napoleonic Wars (1815), at which time Britain finally and firmly
established its rule. Britain sought to establish a colony which could
safeguard its security interests, while at the same time ameliorating
72
Britain in Orange Free State and Transvaal
domestic unemployment and unrest. In 1820, 4000 British settlers

arrived at Cape Town:
These nineteenth-century British immigrants, however, did not assimi-

late into the Afrikaner population as earlier French Huguenot and
German arrivals had done. They came from an urbanized, industrial-
izing society in Britain and in South Africa they tended to settle in the
towns, the largest of which therefore became predominantly English-
speaking. (Smith, 1996: 15—16)
This urban—rural split exacerbated ethnic and linguistic differences,

and eventually laid the roots for two convictions. On the British side,
immigrants became convinced of their ethnic superiority to local
Afrikaners (Smith, 1996: 16). At the same time Afrikaners came to be
suspicious that these Uitlanders (outlanders) were not interested in
sharing power or property in South Africa, but only in capturing it all
for themselves. This was especially true after Britain’s parliament out-
lawed slavery in the Colony in 1834. Since slavery was seen then to be
vital to the rural economy, many Boer were deeply embittered by this
blow to their way of life. The abolition of slavery thus prompted a mass
emigration into the South African interior — an emigration now known
as The Great Trek (1835—37). Thousands of Boer families took to wagons
with all their property, and at great personal risk (many were attacked
and killed by the then powerful African tribes such as the Xhosa and
Zulu), moved inland, eventually to establish the Boer republics of
Natal, Orange Free State, and further north and east, Transvaal.
These were difficult times for the Afrikaners, who ran into hostile and
numerous African tribes along the way. From the African perspective,
the times were darker still. White settlers poured into their ancestral
lands, more often than not killing or enslaving them as they advanced.
Atrocities took place on both sides as native Africans were steadily
displaced and dispossessed. In 1838, the Boer general Pretorious
defeated the Zulu king Dingaan at the Battle of Blood River, and
Afrikaners began to concentrate in Natal. By 1843, however, the
British had annexed Natal and, five years later, Transorangia (later to
become Orange Free State). After a number of sometimes bloody dis-
agreements between the Boer trekkers and the Crown, however, Britain
eventually agreed to allow limited autonomy for the two inland Boer
republics, Transvaal (1852) and Orange Free State (1854).
There followed nearly three decades of relative peace. In 1870, dia-
monds were discovered in Kimberley, a town between Orange Free State
73
and Cape Colony (later quickly and illegitimately annexed by Britain in

1871). The ‘‘diamond rush’’ of 1870—71 brought a considerable influx of
Uitlanders — which included the then 18-year-old Cecil Rhodes — into the
previously pastoral and rural communities of Orange Free State. The
miners and prospectors drank and gambled, and were an early shock to
the conservative, pious, and hard-working Boer farmers who still con-
stituted a majority within the republics.
These tensions were only aggravated after Sir Theophilus Shepstone
announced by proclamation the annexation of Transvaal as a British
Crown Colony in 1877. This move was deeply resented by the Boers,
who argued that the move was a direct violation of the Sand River
Convention of 1852, in which Britain had previously guaranteed their
sovereignty.
Disagreements and bitterness aside, Britain’s de jure sovereignty in
South Africa implied British de facto responsibility for the security of
Europeans within its declared jurisdictions:
Thus British power succeeded in achieving for the Transvaal state

what the Afrikaners had been too weak to achieve for themselves.
British annexation may have removed their ‘‘sacred independence’’
but by the end of 1879 it had also secured the Transvaal against its most
powerful African neighbours. This crucial development freed the
Afrikaners to make a bid to regain their independence during the
following year. (Smith, 1996: 28)
In 1879 the British destroyed the Zulus as a military force. The Zulus
had been the last major African tribe to resist white imperialism in
South Africa. With this last threat to Boer survival eliminated, the
stage was set for a major confrontation between the Boer and the
British. Although Shepstone’s annexation proclamation had promised
Transvaalers ‘‘a separate government, with its own laws and legisla-
ture,’’ after 1879 ‘‘the territory was ruled directly, as a Crown Colony, in
a tactless, authoritarian way . . . ’’ (Smith, 1996: 28).
As resistance in Transvaal mounted, and threatened to spill into open
rebellion, Transvaal’s President Kruger waited for the results of a gen-
eral election in Britain. The election brought Gladstone and the Liberals
to power, and Kruger waited for policy to change in accordance with
Gladstone’s speeches on British policy in South Africa:
In the Transvaal, declared Gladstone, ‘‘we have chosen most unwisely,

I am tempted to say insanely, to place ourselves in the strange pre-
dicament of the free subjects of a monarchy going to coerce the free
74
subjects of a republic to compel them to accept a citizenship which

they refuse’’. (Smith, 1996: 29)
Yet, once in power, Gladstone balked: how could he trust the prospect
of a South African confederation to the uncultured hands of the likes of
Kruger and his cronies? In the event, Gladstone refused to grant
Transvaal its own sovereignty and soon the whole region was in open
revolt. The three-month war culminated in a humiliating British defeat
at the Battle of Majuba Hill (February 27, 1881):
The battle, in which only a few hundred troops were engaged and
where British losses, including the death of Colley himself, accounted
for about a third of the total for the whole war, soon acquired great
symbolic importance. Like the failure to relieve General Gordon at
Khartoum, it came to be regarded in Britain as a blot upon the national
honour which Conservatives pointed to as an example of Liberal
mismanagement. ‘‘Remember Majuba’’ became the rallying cry with
which many British soldiers were to go into action in 1899.
(Smith, 1996: 31—32)
Thus, by 1884, three years after the Battle and two years before the
discovery of gold, Transvaal became the South African Republic offi-
cially in the London Conventions. All the powers of a sovereign and
independent state save the right to determine foreign policy (and
expand beyond current borders) thus reverted to the Republic.
Then came the discovery of gold at Witwatersrand in 1886. The gold
rush which followed had three important effects. First, it led to a
massive new influx of Uitlanders. It transformed the demography
and economies of the Boer republics. Second, the discovery of gold
created for Kruger the problem of maintaining Boer control of the
state,1 while at the same time providing him the opportunity to use
the revenues to buttress his state’s physical security and independence
by purchasing arms and maintaining a state monopoly on dynamite
manufacture (Smith, 1996: 54).
1
This problem, which was to become for the British a primary casus belli, revolved around
the question of the franchise. The Transvaal, for example, restricted the franchise to
Afrikaner burghers by means of legal requirements for residency. This meant the bur-
geoning population (a majority in Transvaal) of Uitlanders were not allowed the vote, and
their protests over disenfranchisement, sent even to Queen Victoria herself, allowed
Britain to use the franchise issue to disparage the Boers as corrupt and authoritarian in
the local and British press.
75
Kruger also granted a railroad concession (another monopoly) to a

Dutch company. Both decisions were to prove crucial during the first
interaction of the later war. Third, the gold and diamonds created
extremely wealthy (and hence powerful) extra-state interests in the
region. The conservative — and by British standards corrupt — govern-
ment of ‘‘peasants’’ led by Kruger was a constant thorn in the side of the
region’s great capitalists, such as Cecil Rhodes and Alfred Beit. These
wealthy men were hardly shy of using their influence to undermine
Kruger, and their meddling ultimately culminated in the fiasco of the
Jameson Raid.
On December 29, 1895, Dr. Leander Starr Jameson and 500 Chartered
Company2 police rode across the border into Transvaal in an attempt to
spark a rebellion of Uitlanders in Johannesburg, providing a pretext for
British intervention in Transvaal. No rebellion occurred, and Jameson’s
raiders were defeated and forced to a humiliating surrender at the
battle of Doornkop.
The main effects of the Jameson Raid were three. First, it provided a
powerful impetus to the development of an Afrikaner nationalism
which had been sparked at Majuba Hill. Second, it confirmed in the
minds of Kruger and Steyn (his Orange Free State counterpart) that
Britain was behind attempts to take back Transvaal’s sovereignty.
Third, it caused Kruger to prepare seriously for war, most especially
by building forts and investing in the latest and best military
equipment.
The eve of war: British interests

British interests in the fate of South Africa were overwhelmingly repre-
sented by two individuals: first, and foremost, Alfred Milner; second,
Joseph Chamberlain, colonial secretary. Milner, as British high commis-
sioner, was the Empire’s ‘‘man on the spot’’; while Chamberlain, as
colonial secretary, was responsible for reporting to the queen,
Cabinet, and Parliament on all of the Empire’s colonial affairs. Milner
was convinced of the absolute necessity of a show of military force in
order to coerce the Boers into ceding sovereignty (essentially, acceding
to annexation); and then later, into abiding by whatever concessions the
Boer granted while under such duress (Pakenham, 1979: 77).
2
The chartered company was owned and run by the multimillionaire and then prime
minister of Cape Colony, Cecil Rhodes.
76
Convinced that annexation by force was the only way to assure

British control over the two republics,3 Milner fretted over the prospect
of peace:
What could be more alarming than the prospect of Kruger’s conces-
sions leading to a settlement that he and every British South African
were certain would prove to be a sham? ‘‘Very great feeling of depres-
sion,’’ he wrote on 23 July, ‘‘as I can see no good is coming of the long
struggle against S.A.R. misgovernment. British public opinion is going
to be befooled and that is the long and short of it.’’
(Pakenham, 1979: 82)
This passage highlights the importance key political elites placed on

the power of British public opinion.4 And in this case public opinion
was to be ‘‘managed’’ by means of the press, which was believed to
exercise a direct influence on British public opinion:
But Milner’s other cronies from his Pall Mall days were now in a
position to show their loyalty, and show it they did . . .
All these newspapers taught British public opinion in the last
few weeks of the dangers of delay, and of being fobbed off without
a settlement on the franchise. It was The Morning Post and The
Times, both staunchly pro-government, that banged the jingo drum
loudest. (Pakenham, 1979: 86)
Milner exercised a great deal of influence over Chamberlain. He was

not only charismatic, but connected. Milner pushed and bullied and
threatened and cajoled in order to get his 10,000 troops sent to Natal.
Now, under Milner’s breathtakingly comprehensive pressure,
Chamberlain was to put a particular spin on Kruger’s latest proposals
which would foreclose any chance of a negotiated settlement:
3
Milner’s concept of control included certain good intentions. He believed Kruger and
his cronies were anachronisms well gotten rid of, and that Milner’s own administration
would raise the level of civilization of all the people living in the republics, whether Boer
or British.
4
Here is Pakenham’s paraphrase of Milner regarding the threat of a British public
undermining ‘‘the tough necessities’’ of Britain’s South Africa policies: ‘‘Above all, the
‘unctuous rectitude’ (quoting Rhodes’s famous sneer) of the British public must not be
allowed to ruin the settlement. No votes for the coloured people in the Transvaal at all
costs. There was only one set of laws in the Transvaal that the Uitlanders considered really
‘excellent’: The laws ‘to keep the niggers in their place’’’ (Pakenham, 1979: 123—124). This
passage is important for two reasons. First it counts as evidence of what Mack would call
‘‘political vulnerability.’’ Second, the passage highlights the hypocrisy of Britain’s position
regarding South Africa’s blacks. In his arguments to the Cabinet (Pakenham, 1979: 113,
520), Chamberlain went so far as to argue that South Africa’s blacks would be better off as
British subjects (it was a base lie).
77
The final proof of this came from [Kruger’s] latest offer. To couple the
offer of a five-year franchise with two conditions he knew to be unaccept-
able — on suzerainty and non-interference on behalf of British subjects —
this conclusively proved that he did not ‘‘want peace’’. The only way to
make him disgorge the franchise was to put a pistol to his head.
The Uitlanders [non-Dutch colonists] were treated like ‘‘an inferior
race, little better than the Kaffirs or Indians whose oppression has
formed the subject of many complaints’’. (Pakenham, 1979: 91)
But was the franchise really the casus belli the British made it out to
be? No. Even as a justification for war to the British public, it had
serious drawbacks:
[The franchise] may have served its turn ‘‘to get things forrarder’’ in
South Africa and as a ‘‘splendid battle cry’’ to rally the support of
British public opinion, but Chamberlain knew that if it came to a casus
belli, grounds for this would have to be found elsewhere. British public
opinion would not support a resort to war over the difference between
a seven or a five years’ franchise. (Smith, 1996: 314)
In other words, British policymakers were constrained in their argu-

ments and actions by a sense of what the British public would or would
not accept. But Jan Smuts and other Boer moderates were not privy to
such information, and as a result they crafted a peace proposal
designed to assess Britain’s true intentions. Britain’s rejection of the
proposal confirmed to Smuts and Kruger that the franchise was not the
real issue, the real issue was instead ‘‘British supremacy and Transvaal
subservience’’ (Smith, 1996: 354).
There is another important feature of Chamberlain’s arguments to
the Cabinet: precedent-setting effects:5
But the issue now went further than the grievances of Uitlanders, or
natives. What was now at stake was no less than ‘‘the position of Great
Britain in South Africa — and with it the estimate formed of our power
and influence in our colonies and throughout the world’’. Such were
Chamberlain’s formal arguments to the Cabinet. (Pakenham, 1979: 91)
In effect, Chamberlain argued that if Britain didn’t threaten the use of

force, ‘‘others’’ would think Britain weak and seek to take advantage of
5
In The Geography of Ethnic Violence, Monica Duffy Toft elaborates the difficulties multi-
ethnic states face in negotiating settlements short of violence with independence-minded
ethnic groups when doing so will set a precedent for other groups (Toft, 2003).
78
that weakness. Keeping in mind that Victorian Britain possessed a

sprawling colonial empire, this was a serious consideration. But
Chamberlain had a final argument based on his low opinion of Boer
culture and resolve. He argued that sending troops would make their
use unnecessary because Kruger was bluffing, and faced with force he
would back down (Pakenham, 1979: 91—92).
But did the British want war? Did their interests extend so far as to
risk an all-out fight with the tiny Boer republics? It’s a difficult question
to answer, but the question of Britain’s war-willingness is important if
we accept that a willingness to expend blood and treasure in a war is a
good indicator of vital interests. The less willing to fight a war, the less
vital the interests at stake.6
Smith argues that in fact the British did not want war, and his
indicator of British intentions is Britain’s lack of military preparations
for war (Smith, 1996: 337—338). But there is one obvious problem with
this interpretation: there may have been other reasons why Britain did
not seriously prepare for a war in South Africa. Chief among these
reasons is lack of regard for the military capabilities of one’s adversary.7
Smith himself notes that the British were used to ‘‘small wars’’ which
required very few troops to prosecute (Smith, 1996: 3). He also makes it
clear that the British did not rate the Boer military threat very highly
(Smith, 1996: 339). Pakenham takes the argument a step further:
Only a few days after his arrival Lansdowne cabled to ask his views
about the first crucial questions. [Penn Symons’s] reply: a mere two
thousand extra troops would make Natal safe right up to its northern
apex (hemmed in though it was by the two republics).
(Pakenham, 1979: 74)
6
There were economic interests — gold and diamonds — in South Africa which commen-
tators have long argued were the real reason Britain was willing to go to war. The most
notable such argument was leveled by the socialist J. A. Hobson, then correspondent for
the Manchester Guardian. In his influential book, The South African War: Its Causes and
Effects, Hobson argued that British policy was hostage to capitalist interests, especially
those of the ‘‘goldbugs,’’ such as Cecil Rhodes and Alfred Beit (Hobson, 1900). Subsequent
scholarship has refuted this thesis soundly (see Smith, 1996: 393—412). But even had it not
been refuted, the argument would have reduced to the claim that South Africa counted as
a vital British interest, one over which it was willing to fight and die.
7
Another interpretation is that the British were unprepared for any military action of a
serious nature. For a thorough discussion of Britain’s overall lack of preparedness for a
major war, see Hamer, (1970: chap. 6). Hamer shows that British lack of preparedness was
not specific to the fight in South Africa, but would have been evident in anything other
than a ‘‘small war.’’
79
The real question was, what would the Boer do? In September, the
two Boer republics could field a total of 54,000 men, whereas Britain
had a total of 10,000 men to guard Natal and Cape Colony. If the Boer
attacked, would they seek to directly engage these garrison forces in
strength, or would they adopt a raiding strategy, pushing only a few
thousand commandos into Natal and the Cape? British military intelli-
gence, under the command of Major-General Sir John Ardagh, pre-
dicted a raiding strategy:
How did Ardagh and his department come to this conclusion, that
would seem so astounding in the light of events? Military Notes gives
the answer: the Boers were not regarded as a serious military adver-
sary. As fighting men, they were expected to be inferior to the Boers
who had beaten Colley’s small force at Majuba. Boer generals, used to
fighting Kaffirs, knew nothing of handling large bodies of men . . .
(Pakenham, 1979: 74—75)
There is, of course, no way to definitively establish Britain’s will-

ingness to go to war. However, as these passages make clear, ‘‘prepara-
tions for war’’ in this context does not make a good indicator of British
intentions. Moreover, the weight of documentary evidence available
suggests that whatever Britain wanted, Milner had his heart set on war
as the only means of accomplishing British political objectives in South
Africa. In short, the British did want war, but not the war they got.
The eve of war: Boer interests

We have established that Britain wanted to annex the republics, take
over their political and economic management, and thereby facilitate
the establishment of a unified political space — a South African
Federation — under British control. What then were the Boer interests?8
Essentially, the two Boer republics — Orange Free State under the
leadership of Steyn, and Transvaal under the leadership of Kruger —
wanted political independence from Britain. In negotiations leading up
8
To be fair and accurate there were no ‘‘Boer’’ interests: ‘‘Despite the direct and painful
testimony of Boer generals and participants on commando, the rifts and divisions on the
Boer side tended to be glossed over. An Afrikaner nation did not exist in 1899 and
Afrikaners in different parts of South Africa had been moulded in different contexts
and states by very different experiences during the course of the nineteenth century.
There were more ‘‘Boers’’ in the Cape Colony than in the two Boer republics.’’ (Smith,
1996: 7). However, we may nevertheless discern a common denominator across republics:
freedom from British rule.
80
to war, the Boer proved themselves remarkably pragmatic and willing

to compromise on many of the key issues — for example, the franchise —
raised by Chamberlain and Milner. Yet by the eve of war two features of
this conflict of interests made war inevitable. First, Britain had tipped
its hand and made it clear that it would accept nothing less than
annexation, while the Boer made it equally clear that although they
were willing to redress a number of British grievances, they were not
willing to compromise on the key issue of sovereignty. Second, neither
side trusted the other’s declarations and overtures. To make matters
worse, the British dramatically underestimated both the Boer military
capacity, and their resolve to use that capacity. The end result was that
both sides exchanged ultimatums, and when the conditions of neither
were met, the Boer mobilized for war.9
The South African War

Smith provides the most cogent summary of the South African War in
terms of its historical importance, and its relevance to our central
question:
The South African War of 1899—1902 was the most extensive, costly
and humiliating war fought by Britain between the defeat of Napoleon
in 1815 and the outbreak of the First World War in 1914. It involved
over four times as many troops as the Crimean War and cost more than
three times as much in money. It was a colonial war, by far the largest
and most taxing such war fought by Britain during the century of her
imperial pre-eminence, and the greatest of the wars which accompa-
nied the European scramble for Africa. The British government
resorted to war in 1899 to end a situation of Boer—British rivalry over
the Transvaal, where political power was concentrated in Boer hands;
to advance the cause of South African unification; and to establish
British power and supremacy in southern Africa on a firmer basis. Two
Boer republics . . . fought for their independence in an anti-colonial
war against full incorporation into the British Empire. The combined
Afrikaner population, of what were then two of the world’s smallest
states, amounted to less than 250,000. By March 1900, some 200,000
British and Empire troops were involved in a war in South Africa with
Boer forces fielding less than 45,000. (Smith, 1996: 1)
9
Although the Boer initiated the war, this is only of technical interest. Britain is respon-
sible for the war itself: on this there is no disagreement.
81
There is no controversy concerning the relative power of the two

sides in this fight. The UK was the strong actor, while the two Boer
republics — even including the possible support of citizens from Cape
Colony and Natal — were out-powered well in excess of 5:1.
Interaction one: conventional attack vs.

October–December, 1899
In the initial interaction of the war the Boer locally outnumbered the
British garrisons — in Cape Colony and Natal — by 3:1. Here is
Pakenham’s characterization of Boer strategic thinking as the likelihood
of war became more certain:
[Smuts] then launched into a feverish plan for a military offensive. Its
keynote was a blitzkrieg against Natal before any reinforcements
could arrive. The numerical advantage would then lie in their favour
by nearly three to one — that is, forty thousand Boers against fifteen
thousand British troops. By throwing all their troops against Natal,
they would capture Durban before the first ships brought British
reinforcements. In this way they could capture artillery and stores
‘‘in enormous quantities’’. They would also encourage the Cape
Afrikaners in the interior to ‘‘form themselves into a third great repub-
lic’’. The international repercussions, Smuts continues, would be dra-
matic. It would cause ‘‘an immediate shaking of the British Empire in a
very important part of it’’. Britain’s enemies — France, Russia, and
Germany — would hasten to exploit Britain’s collapse.10
(Pakenham, 1979: 102—103)
The first half of Smuts’s vision seems hardly controversial in light of

events. What damned this otherwise fine strategic plan was time and
the hesitation of Orange Free State’s President Steyn. All through
September Steyn hesitated as the British troop-ships steamed ever
nearer Durban. By the time both republics mobilized it was too late.
By 9 October the majority of British troop-ships had already landed at
Durban.
10
Although the strategy itself is sound, it is difficult to read this sort of domino reasoning
without, first, recalling similar reasoning on the part of the Japanese prior to their attack
on Pearl Harbor in 1941; and, second, realizing how common such thinking was in that
day. Milner and the British were certainly thinking along similar lines.
82
And as to Boer tactics, equipment, and training, it must be added that

although capable of conventional offensives against regular forces in
fortified positions, their organization and training hardly made them
ideal prosecutors of Smuts’s strategy:
Their elected civilian leaders were made commandants — appointed,

that is, to lead the five hundred to two thousand burghers of each
commando in battle. In this commando system, it was no one’s job to
train the burghers . . . the men were left to fight as they had always
fought — with the tactics of the mounted frontiersman. If the enemy
were superior in numbers, they would provoke the enemy’s attack,
dismount, take cover and shoot, remount and ride away. In European
military manuals it was a formula known as ‘‘strategic offensive,
tactical defensive’’. The Boers had never seen the manuals.
(Pakenham, 1979: 105)
The war started with attacks by Boer commandos (essentially,

mounted infantry with light artillery support) against the British in
three areas: Dundee to the south, Mafeking to the north, and
Kimberley to the northwest.
The British were dangerously exposed north of the Tugela river in
Natal. Britain’s general on the spot, Sir George White, wondered
whether he should withdraw from Dundee to positions south of the
Tugela, and he was answered in the negative by both local military
authority (the Dundee garrison’s commander, Sir Penn Symons) — who
did not consider the Boers much of a military adversary — and by
political authority (Natal’s governor, Sir Walter Hely-Hutchinson) —
because it would encourage the Zulus to revolt (Pakenham, 1979: 109).11
The first armed clash of the war took place on October 20 at Talana
Hill just outside Dundee in Natal. The Boer were under the command of
Lucas Meyer, and the British were under the command of Penn
Symons, who was just sitting down to breakfast as the first Boer shells
plunged into Dundee. Symons was neither a fool nor a military genius.
He organized the defense of Dundee ‘‘by the book.’’ His tactics in
engaging the Boer were: (1) artillery preparation; (2) infantry attack
(ending with a charge with fixed bayonets); and finally, (3) a cavalry
charge to ‘‘cut off the enemy’s retreat.’’ And although a number of his
subordinates worried about the target that closely packed troops might
11
Milner later made similar arguments and issued similarly dire warnings regarding the
revolutionary potential of the Cape Afrikaners in the event of military defeat or a lack of
sufficient military presence in the Cape.
83
present to sustained and rapid rifle fire, Symons also believed in keep-
ing his infantry in close rather than open order (so as to maximize
punch). In this Symons was to prove wrong, but he should not be
blamed as careless or incompetent on that account because, as
Pakenham notes, ‘‘In the whole of Europe there was no body of soldiers
that had ever seen the concentrated fire of the magazine rifle, with the
muzzle end facing them’’ (Pakenham, 1979: 132).
In effect, the curtain had just risen on a match between a highly
mobile Boer military organization outfitted with dramatically augmen-
ted firepower,12 and a ponderous, hidebound, and overconfident yet
superior (in terms of numbers) British imperial army.
In the battle itself, in which the British won the artillery duel (eigh-
teen guns against three), Symons and many of his officers and infantry
died forcing the Boer commandos off Talana Hill, only to discover that
most of Meyer’s commando had escaped on horseback. Unbeknownst
to Symons, his cavalry, which was supposed to cut off Meyer’s retreat,
had run into trouble and surrendered hours earlier. The British thus
won the hill, but did they win the battle?
The fight for Talana Hill proved typical of most encounters between
the British and Boers up until ‘‘Black Week’’ in December. Although the
Boer often had better guns, they were never locally sufficient to alter the
outcome of the ‘‘second act’’ of the three-act pattern initiated by Symons
and his peer commanders in the early interactions of the war. British
infantry tended to rush straight against an entrenched Boer infantry in
close order, sustaining heavy casualties against the concentrated fire-
power of the new Mausers. To the British officers leading their men into
battle, the advent of smokeless powder proved to be disconcerting to
say the least. To the British, the dug-in Boer were invisible until the
advancing infantry were right on top of them. This made it difficult for
officers on the ground to disperse, concentrate, advance, and retreat
their infantry effectively. The result was high casualties. After fixing
bayonets for the final charge, the British usually found their objective
abandoned: a few dead Boer and in the distance, a majority of the
commando trotting away on horseback. Where the British in fact had
cavalry, as in the case of Boer General Kock’s premature advance
against White at Ladysmith, decisive — if brutal — outcomes were
12
Kruger’s foresight had armed the Boers with the Mauser ’98 — an accurate, reliable,
magazine-fed, breech-loading rifle firing smokeless rounds.
84
possible.13 However, the British never had a sufficient cavalry presence

in South Africa to counter Boer hit-and-run tactics.14
In the next few weeks the two forces met in battle after battle, which
the British had the worst of. In November, White was trapped at
Ladysmith and all pretense of a short war ended. White’s internment
at Ladysmith was perhaps the greatest British strategic blunder of the
war. White had in fact already been ordered to fall back south of the
Tugela River in the event of encountering superior forces. His orders, in
other words, were to avoid becoming besieged at all costs. How then
did it happen, and why?
Essentially, White believed — as all his peers at that time — that the
Boer were not a worthy military opponent; and that fighting ability
aside, their morale would be so fragile that one decisive British victory
would send them packing. Yet he was outnumbered by Joubert, the
famous Boer general, by two to one. What made him willing to risk
entrapment by offering battle against his specific orders? Pakenham
provides an answer worth quoting at length:
To some extent, [White] was merely expressing the conventional

British general’s ignorance of the realities of large-scale war. For half
a century Britain had fought small wars against the disunited and ill-
armed tribesmen of India and Africa. Often these wars had begun with
13
A British journalist on the scene provides the following description of the cavalry
charge against Kock’s retreating commando at Elandslaagte: ‘‘The charge of two hundred
horsemen galloping across a plain is designed to be an irresistible force. It does not stop
simply because the enemy would like to surrender . . . The charging line of horsemen
caught them broadside, like the steel prow of a destroyer smashing into the side of a
wooden boat. People heard the crunch of the impact — and saw the flash of the officers’
revolvers, and heard the screams of the Boers trying to give themselves up . . . Back came
the cavalry for a second charge. (‘Most excellent pig-sticking . . . for about ten minutes, the
bag being about sixty,’ said one of the officers later.) Again the shouts and screams . . . But
a story had got round that the Boers had abused a flag of truce and, anyway, the order
was: no prisoners’’ (Pakenham, 1979: 143—144).
14
Three problems arose in relation to fielding an effective cavalry in South Africa. First,
the British found it impossible to get enough horses in theater. They stripped horses from
other colonies as far away as India and Australia. But the transport took a heavy toll, and
those horses which survived to reach the theater were often too weak to be used as cavalry
mounts. Second, the climate and terrain of South Africa was hard on horses, who not only
needed extra time to acclimatize, but could not afterward be used in sustained galloping
(which killed them). By the time the British understood this, they had effectively
destroyed their existing cavalry as a fighting force. Finally, the later British policy of
farm-burning (a COIN strategy) had the effect of restricting the mobility of cavalry
formations to the railways, because the policy not only destroyed forage for Boer horses,
it destroyed forage for any horses. In the end, the British gave up trying to employ cavalry,
and hit upon the wiser strategy of imitating the Boer by using their horses as mounts for
trained infantry.
85
shattering reverses; small bodies of men, surrounded by savages who

gave no quarter, had fought to the last cartridge. In due course, the
main British army would come on the scene and inflict a crushing and
permanent defeat on the enemy. There was little strategic, or even
tactical, maneuvering, by European standards. To transport and sup-
ply his men in desert or jungle — that was the chief problem of the
British general. The actual fighting was usually simple by comparison.
So it had been for Roberts in Afghanistan, Wolseley in Egypt and
Kitchener at Omdurman. The campaign might take months, but the
decisive battle could be fought in a matter of hours. War was a one-day
event, as practised on Salisbury Plain. Hence it was as natural for
White to try to deal a knock-down blow to Joubert, as for Symons to
try to deal one against Lucas Meyer. (Pakenham, 1979: 155)
White’s forces were beaten in the field by the Boer on November 2,

and White retreated to Ladysmith, where he and his remaining troops
remained trapped until February 28, 1900. By the end of the first week
of November 1899, the British had been effectively bottled up in three
sieges: Kekewich was stuck with Cecil Rhodes in Kimberley, Baden-
Powell was under siege at Mafeking, and White was trapped with his
‘‘field force’’ at Ladysmith. The three sieges significantly complicated
the military campaign against the Boers, because White’s entrapment
(and for political reasons, Cecil Rhodes’s in Kimberley) now forced the
British to divide their forces in order to relieve the sieges, instead of
concentrating them for an irresistible rush toward the Boer capital
cities.
Black Week: December 11–18, 1899

The task of relieving these sieges first fell to the South African theater’s
new commander-in-chief, Redvers Buller. Lord Methuen, advancing
to relieve Kimberley, was repulsed by well-entrenched Boer at
Magersfontein on December 11. On December 15, while rushing to
relieve White at Ladysmith, Buller was himself repulsed by entrenched
Boer under the skillful command of Louis Botha — one of the youngest
and most capable of the Boer generals. Three days after this reverse,
Buller was relieved of his command and replaced by Lord Roberts, with
Lord Kitchener as Roberts’s chief of staff.
This last week of reverses came to be known as ‘‘Black Week’’ in the
British press. In fact, although the military situation was actually favor-
ing the British at the end of December — they had, after all, taken the
offensive, and the Boers, although inflicting higher-than-usual (for the
86
British) casualties, were in retreat — public outrage over the week’s

reverses reached crisis proportions. This outrage had in fact become a
constraint on Britain’s conduct of the war:
To expand the British army in South Africa was not merely a question
of recruiting the troops, hiring the ships, and sending them steaming
off to South Africa. It was complicated by two political pressure
waves, freak storms of public opinion, now making the windows rattle
in Whitehall . . . The second was ominous for the whole world: a
shock-wave of Anglophobia vibrating across the Continent, precipi-
tated by the war, and prolonged by Britain’s failure to win it.
(Pakenham, 1979: 257)
Again, perceptively I think, Pakenham continues trying to isolate

what lay at the root of British public concern over Black Week. He
argues that it was ‘‘disappointment in victory so long delayed’’:
The people of Britain had had war on the cheap for half a century.
Small wars against savages: the big-game rifle against the spear and
the raw-hide shield. Small casualties — for the British. To lose more
than a hundred British soldiers killed in battle was a disaster suffered
only twice since the Mutiny. Now, in 1899, they had sent out the
biggest overseas expedition in British history to subdue one of the
world’s smallest nations. It would have been odd if the public had not
shared the government’s confidence in a walk-over. The resulting
casualties were thought shattering: seven hundred killed in action or
dead of wounds, three thousand wounded since October . . . This was
at the root of the public’s humiliation. Then the spasm of bitterness
passed. (Pakenham, 1979: 258)
This passage highlights two key issues: the context of British casualty
sensitivity, and the degree to which pride and shame motivated British
actions. And what of British war aims? Now that the costs were mounting,
what of the anticipated gains? Paraphrasing Asquith, Pakenham adds:
But he warned people that it would be ‘‘grotesque’’ to get these

reverses out of proportion. He compared the present ‘‘humiliations
and mortifications’’ with periods of real national crisis during the
Napoleonic War or the Indian Mutiny. How would Marlborough,
Wellington or Havelock have survived this ordeal by telegraph —
every blow they struck and every blow they received made subject
to hourly scrutiny by the public? The struggle now went much deeper
than a mere question of ‘‘asserting and maintaining our position in
South Africa. It is our title to be known as a world power which is now
upon trial.’’ (Pakenham, 1979: 258)
87
If the war had not begun as a vital security interest for the British,
Asquith’s remarks make it clear that it had now become one.15 This
passage also points to what may be the first media complaint regarding
the clash between a public’s access to information about the costs and
conduct of a war, and the degrees of freedom granted local military
authorities in winning it. As will become clear in the next interaction of
the war, it would not be the last such complaint.
With the accession of Roberts and Kitchener to command — and more
importantly the arrival of massive British reinforcements — the final
stage of the first interaction of the war was now set. There were blun-
ders on both sides, but by the end of June 1900, Roberts’s ‘‘steamroller’’
had captured Bloemfontein (March 13), Johannesburg (May 31), and
finally, Pretoria (June 5). In the euphoria over his capture of
Bloemfontein, Roberts issued his first proclamation (March 15):
amnesty for Boer regulars (they could keep their property if they turned
in their rifles and swore an oath of neutrality) but not for Boer leaders.16
The three sieges had also been raised; Kimberley (February 15),
Ladysmith (February 28), and Mafeking (May 17).
The war was over — or rather, it should have been. But Roberts’s
triumphant march to Pretoria marked the close of only the first third of
the war, which was due to stretch on, in much more brutal form, for
another two years.
Interaction two: conventional attack vs. GWS

Bloemfontein to Pretoria: March–June, 1900
The second interaction of the war had its genesis, on the Boer side, in
failure. Christiaan De Wet — among the most talented of the Boer
generals, and in fact the single most talented guerrilla leader, — thought
through the pros and cons of a Boer shift to a GWS after General Cronje’s
surrender at Paardeburg:
15
This opinion was shared by the Gladstonian (Liberal) wing’s most outspoken member,
Henry Labouchère, who ‘‘now declared his belief in the doctrine of ‘my country right or
wrong’. The danger of Britain’s being humiliated in front of the other Great Powers, he
said, outweighed the moral disadvantages’’ (Pakenham, 1979: 267).
16
Roberts believed his capture of Bloemfontein would ‘‘knock all the fight’’ out of the
Orange Free State burghers, and ‘‘left to themselves, [they would] accept the amnesty, take
the oath of allegiance, and disperse to their homes’’ (Pakenham, 1979: 400).
88
But [De Wet] had seen enough to realize that, quite apart from the
blunders of Boer generals like Cronje, the overwhelming numerical
superiority of the British now demanded new strategy from the Boers.
Indeed, the commando system was best suited not to large-scale, set-
piece battles, but to smaller-scale, guerrilla strikes. A smaller group
could make better use of their best asset, mobility; and their worst
defect, indiscipline, would prove less of a handicap.
(Pakenham, 1979: 348)
The British had made mistakes too. First, underestimating their

adversary, and second, failing to adapt their forces and tactics to exploit
vulnerabilities in the Boer method of fighting. But the British, who had,
after all, ‘‘won,’’ showed little interest in learning from their mistakes.
After his capture of the Boer capitals, Roberts returned to a hero’s
welcome in London. Kitchener replaced Roberts as commander-in-chief
in November of 1900, and was not interested in the mistakes of other
British generals:
Kitchener displayed no interest in learning from the mistakes of Buller

and Methuen. He probably attributed their own failures, as Roberts
did, to their lack of self-confidence. Of the revolution in tactics — of
the new, invisible war of the rifle-plus-trench — he showed himself
supremely unaware. (Pakenham, 1979: 351)
Other British generals, however, had worried about the tactical and
strategic difficulties presented by the enemy, terrain, and climate.
As early as October of 1899, Buller had planned to build and equip
a large force of irregular colonial troops trained to fight like the Boer.
It was Buller who had foreseen the problem of lack of mobility, and it
was Buller who anticipated the Boer’s switch to GWS (and its
rationale):
In this type of pioneering, colonial society, there was no highly organ-

ized machinery of administration, and the central government carried
little influence or authority. ‘‘Time has not yet glorified the seat of
Government with a halo of sentiment,’’ wrote Buller. ‘‘To every man
his own home is the capital. Hence there is no commanding centre by
the occupation of which the whole country or even a whole district can
be brought into subjection; no vital spot at which a single blow can be
struck that will paralyse every member of the body. There are living
organisms which can be divided into a multitude of fragments without
destroying the individual life of each fragment.
(Pakenham, 1979: 398)
89
This was part of Buller’s warning to Roberts: the Boer were a body
with no head. But as we have already seen, Roberts had little faith in
Buller’s opinion:
Roberts . . . stuck to the conventional idea of surrender. Capture the
capital and you have cut off the head of the enemy. Their spirit must
die too. And he had long intended to use political means — the kind
that was often used in Indian frontier wars — to smooth his march to
Pretoria. On 15 March he offered an amnesty for every Free State
burgher except the leaders. (Pakenham, 1979: 399)
The British were therefore slow to realize their strategy was not going
to win the war. They continued to hold to the idea of capturing values
and offering limited amnesty as a way to end Boer resistance.17 The
Boer, by contrast, began planning for a new strategy right after the
capture of Bloemfontein.
That strategy was GWS. Its chief architect was De Wet, who put
forward his arguments for a strategic shift at a war council held just
four days after Roberts’s capture of Bloemfontein. The strategy had
three key elements. First, it needed a new kind of force — one made up
exclusively of men who were fiercely dedicated to the Boer cause. After
the fall of Bloemfontein, De Wet sent his exhausted commando home on
ten days furlough, fully aware that many of them would not come
back. But this was by design: those who did return could be counted
on. Second, it needed a force which could operate independently of the
ox-driven wagon trains which typically accompanied Boer commandos
in the field. Third, De Wet urged a switch from attacking concentrations
of British troops, to attacking British communications, which were
extremely vulnerable. Above all, De Wet argued that even a few small
victories could rally the morale of the burghers far out of proportion to
their military impact (Pakenham, 1979: 408—409).
The new strategy thus promised new hope — at least of avoiding
defeat. Something else — something vital — was not discussed as part
of the war council’s deliberations: the presumption of British moral
conduct in war. One reason the Boer had developed as a mobile military
organization — including mobile transport — was precisely due to the
character of warfare against native African tribes, who did not often
respect the notion of noncombatant immunity. Boer trekkers thus kept
17
This would not last long, however. Within a month, Roberts would detach fully half his
forces — 20,000 men — to both guard his communications and begin the process of looting
and burning farms as a COIN strategy (see below).
90
their women and children on wheels, ready to flee their farms on short
notice. But since the destruction of the Xhosa and Zulu as military
threats, the Boer had become more settled and less mobile. This
meant values exposed to the British. But as De Wet makes clear in his
later account of the war, it never occurred to the war council to question
the safety of Boer women and children under British occupation:
Many a smart, well-bred daughter rode on horseback and urged the

cattle on, in order to keep out of the hands of the pursuers as long as at
all possible, and not to be carried away to the concentration camps,
which the British called Refugee Camps (Camps of Refuge). How
incorrect, indeed! Could any one ever have thought before the war
that the twentieth century could show such barbarities? No. Any one
knows that in war, cruelties more horrible than murder can take place,
but that such direct and indirect murder should have been committed
against defenceless women and children is a thing which I should
have staked my head could never have happened in a war waged by
the civilized English nation. (De Wet, 1902: 192—193)
Of course, not all those attending the war council had such faith in
British moral restraint. Louis Botha, commanding the last effective
conventional Boer army in the field, objected to the change in strategy
on humanitarian grounds:
But, however practicable, was a guerrilla war a ‘‘civilized war’’? This

was the question that obviously troubled the consciences of the
Transvaal leaders, especially Botha, and explains why they clung so
long to the strategy of regular warfare on the eastern front. Botha must
have read enough military history — Smuts certainly had — to know
what a guerrilla war inevitably entails for civilians. Sherman’s march
through Georgia, the Prussian treatment of the French franc-tireurs;
they cast ugly shadows on the veld, these international precedents.
And a guerrilla war in South Africa, however gentlemanly the main
combatants professed to be, threatened to have elements of savagery
absent from warfare in more civilized states.18 (Pakenham, 1979: 500)
18
This is a critical debate. Key members of both sides appeared to accept the idea that a
portion — ranging from negligible to complete — of the responsibility for the depredations
which often form the heart of COIN operations is often placed on the defenders. So it was
during the concentration camps controversy (see below). When cornered in Parliament on
the issue of Britain’s farm-burning and camp policy, for example, Lord Broderick
answered ‘‘for the twentieth time, that the policy of sweeping the country had been forced
on them by the guerrillas. Some of the women had been assisting the enemy; others had
been abandoned by them; none of them could be simply left out on the veld to starve’’
91
On May 3, Roberts resumed his irresistible advance to Pretoria by

way of Johannesburg. He detached half his force — 20,000 men — to
guard his line of communications and to begin COIN operations
designed to deny support to De Wet by disarming Boer farmers and
burning ‘‘selected’’ farms whose men were away on commando.
Interaction three: barbarism vs. GWS

This marked the beginning of a COIN strategy which would later
harden into barbarism. As Roberts approached Johannesburg, he nego-
tiated an armistice with the Boer defenders. They agreed to surrender
the town to the British on one condition: that they be granted 24 hours to
extract their army (including, it turned out, ammunition, supplies, and
gold bullion). In retrospect it seems a blunder on Roberts’s part.
Part of the failure, ironically, must be blamed on the very strategy
Roberts had set into motion in dividing his forces: barbarism. There is
almost always a gap between the initiation of such a strategy and its
expected military impact. This creates an early danger that barbarism —
in this case, burning farms and forcing women and children into con-
centration camps — will stimulate the resolve of the resistance while
only having a negligible impact on guerrilla operations. This is what
happened here; many burghers had accepted Britain’s amnesty offers
and returned to their farms, only to find them burned, their cattle dead,
and their women and children missing. They were outraged, and many
returned to fight with De Wet, Jan Smuts, and Koos De la Rey.
(Pakenham, 1979: 540). The Boers themselves appeared to accept this argument in part,
‘‘here was a daunting moral problem. Was it fair to the volk (women and children, as well
as the menfolk) to involve them in such a savage kind of war? For the women and children
it would be like a return to the dark pages of voortrekker history, when their grandparents
had struggled against the Kaffirs: women and children pressed into service, each farm a
commissary and an arsenal; their homes looted and burnt; then forced to choose between
going as refugees to the cities, or following the laagers into battle’’ (Pakenham, 1979: 500).
The real issue, which would come to dominate international conventions on the laws of
war in the next century, was whether, as Pakenham argues, such consequences were
inevitable. In Europe’s past, dominated as it was by sovereigns who derived their legiti-
macy by ‘‘divine right,’’ or who occupied the position of defender of the faith, rebellion
and heresy were closely related, and neither were protected by the laws of war governing
the treatment of prisoners, the granting of quarter, or conduct toward noncombatants in
rebel areas. On the contrary, sovereigns, and later states, were permitted to use any force
or depredation necessary to crush rebellion and stamp out heresy (Parker, 1994: 43). This
slowly changed over the years, so that by the time of the famous Hague Convention of
1899, the matter of rebellion received considerable attention. The issue was not resolved
there, however, nor was it resolved in the 1907 convention that followed (Roberts, 1994:
121—122).
92
For his part, as Roberts advanced, he became frustrated with the

delayed victory. Where were the Boer armies to resist his advance?
Why wouldn’t they stand and fight? All along his advance from
Johannesburg to Pretoria, De Wet and the other guerrilla generals
harassed his communications, captured his troops and supplies, and
generally humiliated his strung-out troops. As a result, Roberts’s
new COIN strategy was about to take a more systematic and more
brutal turn.
Roberts reached Pretoria and marched through its abandoned, dusty
streets, on June 5, 1900. During the next five months, as he continued his
advance through the last Boer towns to the east, he waited in vain for
the approach of Kruger, Steyn, or any of the Boer generals — De la Rey,
De Wet, Smuts, or Botha — for terms of surrender. In November he sailed
for England, while Kitchener took over as the new commander-in-chief.
Under Kitchener’s command, the farm burning continued and intensi-
fied. It may have made sound military sense, but it was not easy to
prosecute — at least not at first:
The worst moment is when you first come to the house. The people
thought we had called for refreshments, and one of the women went to
get milk. Then we had to tell them that we had to burn the place down.
I simply didn’t know which way to look . . .
I gave the inmates, three women and some children, ten minutes to
clear their clothes and things out of the house, and my men then
fetched bundles of straw and we proceeded to burn it down. The old
grandmother was very angry . . . most of them, however, were too
miserable to curse. The women cried and the children stood by hold-
ing on to them looking with large frightened eyes at the burning house.
They won’t forget that sight, I’ll bet a sovereign, not even when they
grow up. We rode away and left them, a forlorn little group, standing
among their household goods — beds, furniture, and gimcracks strewn
about the veldt; the crackling of fire in their ears, and smoke and
flames streaming overhead. (Pakenham, 1979: 466—467)
The practice of burning farms — ostensibly only those of ‘‘rebels’’ but

in practice nearly every farm in reach — came to involve much more
than just burning farm houses. Horses were taken, and cattle were either
taken or slaughtered. By the time a British patrol had done its duty, in
other words, they had effectively deprived noncombatants of food and
shelter. The British decided to solve this problem by putting the
women, children, and old men into concentration camps, for their
own safety.
93
In theory at least this would have been a brilliant COIN strategy. On

the one hand, the British could argue that, after all, they couldn’t just
leave these women and children to fend for themselves against the
elements, including black Africans. On the other hand, the real purpose
of the practice — to deprive the guerrillas of intelligence and logistical
support — could be kept secret (Krebs, 1992: 41—42).
The problem with the theory was that neither Roberts nor Kitchener
ever concerned themselves with logistical matters, and for this reason
British forces had suffered a far higher percentage of post-battle casual-
ties than usual. To put it bluntly, Roberts had left the management of his
field hospitals in thoroughly incompetent hands. That, plus his own
‘‘reform’’ of the system of regimental supply, and the direct actions of
the guerrillas themselves, had caused thousands of casualties — not only
from battle, but from typhoid. Now these same defects were to strike
down and ruin the benevolent justification for the internment of Boer
noncombatants. Women and children began to die in the camps, and
they began to die in alarming numbers.19
The Boer war may count as the first major war in which the literate
soldiers of a democratic government participated. Many of the stea-
mers and trains coming from and going to London contained mail —
letters to and from the homes of regular soldiers in the field. The British
public too, as we have already observed, was both literate and able to
influence British foreign policy. The British public found out about the
conditions in the camps, and the alarming death rate, through the
efforts of a well-meaning humanitarian relief worker named Emily
Hobhouse.
Hobhouse began her tour of the camps in January of 1901, and what
she saw horrified her: poor sanitation, starvation rations, insufficient
clean water, and insufferable crowding. By April she had seen enough.
She raced back to England to report, and her report proved to be a
political bombshell:
19
In May of 1901 there were 93,940 whites and 24,457 blacks in these ‘‘camps of refuge.’’
The death tolls had been rising steadily, and alarmingly, for months. In May, 550 died; in
June, 782; in July, 1675. By October the death rates were 34.4 percent per year for whites of
all ages — 62.9 percent for children in the Orange River Colony, and 58.5 percent for
children in Transvaal. ‘‘At individual camps like Mafeking, the October figures repre-
sented an annual death-rate of 173 percent’’ (Pakenham, 1979: 548). Pakenham calculates
that ‘‘at least twenty thousand whites and twelve thousand coloured people had died in
the concentration camps, the majority from epidemics of measles and typhoid that could
have been avoided’’ (Pakenham, 1979: 549).
94
instantly, the sentiment of the country was aroused and had it been
allowed its true expression, not only would the camps then and there
have been adequately reformed, but very possibly the war would also
have dwindled in popularity and been ended.
Government officials feared just such an occurrence.
(Krebs, 1992: 52)
Lloyd George and the Liberals made a great deal of the conditions of
the camps, and it is in these speeches in Parliament we find supporting
evidence for the claim that even as early as 1900 the British recognized
a norm of noncombatant immunity, the violation of which was con-
sidered sufficient grounds for a change of government:
Why pursue this disgraceful policy, he asked; why make war against
women and children? It was the men that were their enemies. ‘‘By
every rule of civilized war we were bound to treat the women and
children as non-combatants.’’ The novel method of warfare adopted
was all the more disgraceful because it would prolong, not shorten, the
war. ‘‘We want to make loyal British subjects of these people. Is this the
way to do it? Brave men will forget injuries to themselves much more
readily than they will insults, indignities, and wrongs to their women
and children.’’ He concluded . . . ‘‘When children are being treated in
this way and dying, we are simply ranging the deepest passions of the
human heart against British rule in Africa . . . It will always be remem-
bered that this is the way British rule started there, and this is the
method by which it was brought about’’. (Pakenham, 1979: 539—540)
In the end, the government did not change. There were three reasons
for this. First, the Liberal Party was itself divided on the issue of
Britain’s war with the Boer. Second, the British public was divided as
well: some felt that the Boers were to blame for their suffering and for
the conditions in the camps.20 Others were rightly horrified at the
thought of British participation in a system which had such fatal con-
sequences for women and children. Third, the problems of the camps
were not mysteries, nor were they expensive to correct. While the
political crisis over the camps raged through August, effective reforms
20
Hobhouse’s tour was later repeated by special commissions from both the government
and the opposition. The government’s commission was headed by Millicent Fawcett, who
blamed many of the conditions in the camps on the ‘‘filthy and superstitious habits’’ of the
Boer women themselves: ‘‘Boer women were ‘ill-nature[d]’, unnatural. British women, it
was understood, would not make war. The Boers were seen as primitive, unchanged since
their arrival in South Africa from Holland two hundred years earlier. This put them on a
lower scale of civilisation than the British, different in what would have been seen as a
racial way while they were also a different class — a nation of peasants’’ (Krebs, 1992: 44).
95
followed in September. By October the death-rates had stabilized and

then began to plummet sharply.
In the meantime, the war itself proceeded apace. Kitchener inherited
an army of 120,000 regulars and 20,000 irregulars with which to oppose
a Boer fighting force of at most 20,000. He divided his troops into more
mobile columns, which he then sent out after the Boer commandos
much as one would beat the brush for game. The strategy — essentially
a search and destroy mission on a massive scale — was inefficient and
destructive. It was inefficient because the Boer were still so much more
mobile than the British, and the constant wearing away at them had
reduced the average size of effective units to as low as one or two
hundred. It was destructive not only of Boer farms and cattle, but of
British horses and morale.21 Kitchener’s next innovation was the use of
barbed wire and blockhouses to create a massive net covering the
countryside:
On the periphery, the barriers served as offensive, not defensive,

weapons; not as cordons to keep out the enemy, but as cages in
which to trap them, a guerrilla-catching net stretched across South
Africa. By May 1902, there would be over eight thousand blockhouses,
covering 3,700 miles, guarded by at least fifty thousand white troops
and sixteen thousand African scouts. (Pakenham, 1979: 569)
The blockhouse and barbed wire net was expensive, but combined
with all the other measures it finally broke the back of Boer resistance.
The farm burning and concentration camps made it impossible for the
Boer commandos to replace mounts or replenish their food supplies,
and it made intelligence spotty. This resulted in close calls, where only
luck and over-riding horses saved many commandos from capture. In
addition, the British had begun arming native Africans and employing
them as scouts and spies. The effects of this innovation were dramatic:
suddenly the ponderous British columns began to find their ‘‘game.’’
Finally, Kitchener and his staff hit upon a novel idea: why not stop the
practice of shipping Boer women and children off to concentration
camps?
21
Here Pakenham quotes Allenby on the hardships of COIN operations: ‘‘The struggle, as
described in Allenby’s letters, took shape and dissolved like a fog. There were no lines or
fronts, no battles — mere skirmishes with an invisible enemy, whose only aim, apparently,
was to run faster than their pursuers . . . The physical strain of the three-month trek, and
the moral stain of making war on women and children, left him exhausted and ill’’
(Pakenham, 1979: 527).
96
Viewed as a gesture to the Liberals . . . it was a shrewd political move.

It also made excellent military sense, as it greatly handicapped the
guerrillas, now that the drives were in full swing. Indeed, this was
perhaps the most effective of all anti-guerrilla weapons, as would soon
emerge. It was effective precisely because, contrary to the Liberals’
convictions, it was less humane than bringing them into the camps,
though this was of no great concern to Kitchener.
(Pakenham, 1979: 581)
Another problem had already plagued the Boer war effort: what to do
with prisoners. Unlike Boer prisoners, who remained prisoners
throughout the remainder of the war, British prisoners were often
returned to their own lines within hours of being captured:
It is much to be regretted that we were unable to keep them, for had we
been in a position to do so, the world would have been astonished at
their number. But unfortunately we were now unable to retain any of
our prisoners. We had no St. Helena, Ceylon or Bermuda, whither we
could send them. Thus, whilst every prisoner which the English cap-
tured meant one less man for us, the thousands of prisoners we took
from the English were no loss to them at all, for in most cases it was
only a few hours before they could fight again. (De Wet, 1902: 227)
It is a tribute to the dedication and skill of commanders such as De

Wet, Smuts, and De la Rey that the Boer war effort lasted as long as it
did against such overwhelming odds. Surrender terms were signed in
Pretoria on 31 May 1902.
Outcome: Britain wins

Although all concerned agreed that Kitchener would eventually cap-
ture or destroy all the remaining commandos, the decision to surrender
was far from easy or settled. At the war council that finally decided the
issue there were six arguments for why the Boer should seek terms from
the British:
no food for women and children, and no means to continue the war;
the concentration camps (this was for propaganda purposes; Botha
had actually admitted ‘‘one is only too thankful nowadays to know
that our wives are under English protection’’; the ‘‘unbearable’’ condi-
tions caused by Kaffirs . . . Kitchener’s proclamation of 7 August,
which threatened the confiscation of burghers’ land; the impossibility
of keeping British prisoners; in short, no hope of success . . . The
delegates voted for Kitchener’s peace by an overwhelming majority:
97
fifty four to six. It was the bitter end, but the alliance stood firm.
(Pakenham, 1979: 603—604)
Britain had already officially annexed Orange Free State — renamed

Orange River Colony — back in May of 1900; and it annexed Transvaal in
October of that year. Now the fighting ended, and Milner could at last
preside over the formation of a federated South African Republic. Here
Pakenham sums up the outcome of the war:
In money and lives, no British war since 1815 had been so prodigal.
That ‘‘tea-time’’ war, Milner’s little ‘‘Armageddon’’, which was
expected to be over by Christmas 1899, had cost the British tax payer
more than £200 million. The cost in blood was equally high. The War
Office reckoned that 400,346 horses, mules and donkeys were
‘‘expended’’ in the war. There were over a hundred thousand casual-
ties of all kinds among the 365,693 imperial and 82,742 colonial sol-
diers who had fought in the war. Twenty-two thousand of them found
a grave in South Africa: 5,744 were killed by enemy action (or accident)
and shoveled into the veld, often where they fell; 16,168 died of
wounds or were killed by the action of disease (or the inaction of
army doctors) . . .
On the Boer side, the cost of the war, measured in suffering, was
perhaps absolutely as high; relatively, much higher. It was estimated
that there were over 7,000 deaths among the 87,365 Boers — including
2,120 foreign volunteers and 13,300 Afrikaners from the Cape and
Natal who served in the commandos of the two republics. No one
knows how many Boers — men, women and children — died in the
concentration camps. Official estimates vary between 18,000 and
28,000. The survivors returned to homesteads devastated almost
beyond recognition. Several million cattle, horses and sheep, that
had comprised their chief capital, had been killed or looted.
(Pakenham, 1979: 607—608)
Analysis: competing explanations of the South

African War’s outcome
Assessing the relative explanatory merits of competing explanations of
asymmetric conflict requires answering four basic questions. First,
what were the interests of the actors prior to the conflict and how are
these best explained? Second, what was each actor’s regime type, and
how did this affect strategy and political vulnerability? Third, what role
did military arms diffusion play in the conflict’s outcome? Finally, what
were the strategies each side employed, what was the rationale given
98
for their selection or change, and how did they affect the eventual
outcome of the war?
Actor interests
Britain’s interests were not explained by its material power relative to
that of the two Boer republics that opposed it. Britain acted as if the
survival of its empire was at stake. It marshaled three arguments in
support of this calculation. First, although the Suez Canal was opened
in 1869, because it could be easily closed (sabotaged or mined) in time of
crisis, Britain calculated that political control over Cape Town was a
vital security interest. Second, because a majority in Cape Colony were
Afrikaners, Britain calculated that it could not compromise or back
down regarding its efforts to replace Kruger and Steyn with more
amenable leaders. If it failed in Orange Free State and Transvaal, it
would only be a matter of time before the Afrikaners in Natal and
Cape Colony, encouraged by British weakness, facilitated the annexa-
tion of Natal and Cape Colony by the two Boer republics. Britain would
then lose control of its vital port at Cape Town. Finally, British policy-
makers calculated that failure to put down resistance from such tiny
upstart republics would make Britain appear weak to its Great Power
rivals, encouraging them to exploit British weakness, and eventually
unraveling the tapestry of its empire.
The Boer republics did fight and act as though their survival was at
stake — and so it was. The Boer leadership understood early and clearly
that Britain’s aim was annexation: the end of their political independ-
ence and sovereignty.

At the time of the South African War, Britain was a democracy. It had a
large, literate population with a broad franchise. Its elites worried about
public opinion and had to take public opinion into account when
formulating public policy (Smith, 1996: 399—400). What then may be
said about the relationship between Britain’s regime type and the
strategies its leaders chose in South Africa?
Chamberlain’s and Milner’s first difficulty was how to convince the
British public of the need to threaten the use of force in South Africa. All
during the summer months of 1899, these two men attempted and
failed to find a casus belli adequate to mobilize public support. Their
99
best effort resulted in the franchise issue — this gained them sympathy,
but was insufficient to justify risking British lives over. In the event, as
we have seen, they were saved from having to justify the use of force by
Kruger’s ultimatum and subsequent mobilization. Self-defense was a
casus belli every Briton could rally behind. Had Britain had an author-
itarian regime, it could have dispensed with this concern over justifying
its actions to an informed and empowered public.
How many troops to send became the second strategic decision
which would be constrained by public opinion. Britain could not send
too many troops because this would be expensive, and they would
moreover imply an interest out of proportion to the stated aim of merely
securing Natal and Cape Colony from attack. Besides, what was the
need? Part of the propaganda campaign leading up to the ultimatums
had caricatured the Boer as backward, venal brutes: one good crack on
the snout should be sufficient to send them packing — by Christmas at
the latest. The public expected a walk-over, and they expected it on the
cheap. An authoritarian Britain could have sent as many or as few
troops as it wished, without the constraint of public expectations.
A final strategic decision subject to regime-type constraints was the
decision to begin burning farms and concentrating noncombatants into
camps. This was a bottom-up decision, started in the field by Roberts as
a sort of reprisal for the Boers’ refusal to ‘‘fight fair’’ and give up after
the capture of Bloemfontein, Johannesburg, and Pretoria. Yet its polit-
ical impact was such that it very nearly forced Britain from the war. Had
Britain been authoritarian, Emily Hobhouse would have been impri-
soned or possibly executed as a Boer spy. Her entire family would have
been locked up, and her report suppressed.
As to regime type and political vulnerability, nothing emerges more
strongly in this case than the sense that British political elites lived in
constant fear of, as Cecil Rhodes was apt to put it, that ‘‘unctuous British
rectitude.’’ Regime type mattered because in democratic Britain there
was a gap between what was necessary and what was permissible. In
an authoritarian Britain, anything necessary would have been permis-
sible. In democratic Britain this was not the case, and leaders paid more
than mere lip service to this constraint. As we have already seen, British
interests do not explain its vulnerability in this case. For one thing, its
interests were higher than those predicted by the interest asymmetry
argument. For another, even though the British public of that day may
have agreed that ‘‘upstarts must be put in their place,’’ British public
opinion made it difficult for Britain to initiate hostilities in South Africa,
100
or to do so with the necessary number of troops. In sum, democratic

Britain was vulnerable to certain specific actions which violated the
public’s sense of right and wrong in a way that an authoritarian Britain
would not have been. Yet the necessary condition for the operation of
this vulnerability was time. Had the war in South Africa indeed been
over by Christmas of 1899, then the public’s unctuous rectitude would
have had no room to operate. As it was, victory was delayed. This delay
was due first, to underestimating the enemy, and second, to the inter-
action of the strategies adopted by the British and the Boers in March of
1900: a conventional attack vs. a GWS.
Arms diffusion
In the South African War, the arms diffusion argument appears to gain
strong support. After the Jameson Raid, the two Boer republics
embarked on a major arms acquisition and modernization program.
They gained rifles comparable or superior to those of the British, and
their artillery in particular was actually better than that fielded by the
British. Moreover, the Boer had not neglected to train their artillerymen
to employ their new field pieces effectively (Nasson, 1999: 59). There
were, however, two problems regarding artillery that the Boer simply
could not overcome. First, they depended upon contracts to re-supply
ammunition for the guns which were subsequently bought out by Britain
(Nasson, 1999: 59). The Boer succeeded to some extent in manufacturing
their own ammunition, but it was never enough to match demand.
Second, modern as they were (outranging the British guns in most
cases), they were still too few. Once the full British expeditionary force
landed and began to move inland, the relatively few guns the Boer had
could do nothing to affect the strategic balance of forces. Boer mobility
was hampered by the guns and their heavy logistical trains, so that the
preferred method of fighting — establishing a well-defended position in
difficult terrain, then jumping to horse when overwhelmed — would have
been impossible had the Boer continued to rely on their artillery.
The new and highly efficient Mauser rifles Kruger and Steyn had
purchased for their burghers were a different story. These proved
highly effective throughout the war, but particularly in its opening
stages, when British troops were ordered to rush Boer in entrenched
positions. The British had never before run into the business end of
a breech-loading, magazine-fed rifle firing smokeless cartridges
(Pakenham, 1979: 184, 238). It was some time, and not a few bloody
101
engagements, before the British learned tactics that could minimize

their soldiers’ exposure to concentrated rifle fire from entrenched posi-
tions. It is fair to say that they did not learn quickly. But their sheer
weight of numbers made it less imperative to learn as time went on:
they were taking heavier casualties than anticipated, but until the Boer
switched strategies from conventional defense to GWS, they were
winning.
The South African War took place in three interactions. In the first
interaction, an overwhelmingly superior force of British imperial
troops — blunders notwithstanding — decisively defeated their adver-
saries and captured their capital cities. The strategy employed by the
British was conventional attack: they sought to bring the war to an end
by destroying the Boer armies in the field (or forcing them to surren-
der), and by capturing state capitals — Bloemfontein for Orange Free
State, and Pretoria for Transvaal. In this first interaction, the two Boer
republics countered British strategy with a conventional defense strat-
egy: they sought to block the British advances. Both sides made mis-
takes, but overall the British had the worst of it. Yet by dint of
overwhelming numerical superiority, it became clear that the Boer
could not defeat the British with a conventional defense strategy.
Indeed, the brighter generals among the Boer recognized their earlier
strategy as counterproductive, insofar as the fighting style, training,
armament, and discipline (or rather, indiscipline) of the Boer burghers
made them poor ‘‘regular’’ soldiers.
Thus, after the capture of Bloemfontein, the Boer switched to a GWS.
This shift in strategies marked the opening of the second — and far more
brutal — interaction of the war. For the first five months after the capture
of Bloemfontein, the British continued a conventional attack strategy.
When it became clear that the combination of amnesty offers and capital
city occupation would not cause the Boers to surrender, the British
switched to a barbarism strategy. Why should we characterize British
strategy in this way?
There is some controversy about one aspect of British strategy: the
concentration camps. Clearly, the British intended the combination of
farm-burning and internment of Boer noncombatants as a COIN strat-
egy. In this sense alone, the policy falls under the heading of depreda-
tions against noncombatants. But as always we must qualify such
102
judgments. It is true that the practice of confining Boer women and

children led to more deaths — due to malnutrition and plague — than all
the battle casualties on both sides combined. But it is equally clear that
this particular effect was never intended by the British. Yet after the
conditions in the camps were improved, the British then stopped the
policy of forced internment in order to deliberately cause hardship to
Boer noncombatants and hobble the guerrillas. On balance, local com-
manders violated the laws of war regarding noncombatant immunity
in a systematic way and as a COIN strategy. Moreover, we have suffi-
cient testimony from both sides in parliamentary debates to confirm
that, overall, the British accepted the principle of noncombatant immu-
nity as a law of war.22
On the Boer side, there were atrocities and violations of the laws of
war as well. The three most common complaints from the British were
(1) abusing the white flag of truce;23 (2) the use of prohibited weapons;24
and (3) the wearing of khaki.25 Yet at no time did these approach the
22
This was particularly important in this instance because, as is well known, a large
portion of the justification of imperialism was the gift of ‘‘civilization’’ brought to those
‘‘less civilized.’’ Britain’s deliberate violation of the laws of civilized conduct undermined
its moral claim to be civilizing the allegedly uncivilized Boer burghers.
23
Pakenham records a number of claims to this effect, but does not trouble to support or
refute such claims. Here is one account, however, which appears to support the British
claim in at least one case: ‘‘The Boers reappeared with a white flag. They pointed out that
the donga was full of wounded, who would unavoidably be shot, if Bullock insisted on
fighting. They chivalrously offered to let the wounded be removed, before going on with
the fight. Meanwhile, less chivalrously, about a hundred Boers had crept round the side of
the donga, and emerged holding their rifles pointed at the heads of the Devons . . . Bullock
and the rest of the Devons, and the unwounded gunners, were bundled off as prisoners’’
(Pakenham, 1979: 247).
24 The Hague Convention of 1899 had prohibited the use of ‘‘dum-dum’’ bullets — at least
against European troops (the British had used them liberally in Sudan against the
Dervishes). Sanctioned military rounds featured bullets made of lead with a copper jacket
(to help maintain the shape of the bullet in flight and to reduce fouling of the rifle barrel). If
the points of such bullets were cut or cross-hatched, however, they would become hot in
flight and dramatically expand upon contact with a solid body. The likelihood of death
when struck with such a round was thus much higher than when struck with a full-
jacketed round.
25 In the later months of the war, Boer commandos were in such a state of poor supply that
the soldiers often rode in rags or half-naked. They therefore thought little of liberating the
fine uniforms of their captured foes. There is no evidence that Boer commandos sought to
use such uniforms to deceive their adversaries (a violation of the laws of war), but
Kitchener nevertheless ordered that any Boer caught in khaki be summarily shot. It
should be added that by 1900 many British officers and men had begun to adapt their
uniforms to the conditions of the veld — especially the headgear (pit helmets were widely
abandoned in favor of the floppy Boer veld hats). It is thus probably the case that British
soldiers were as often mistaken for Boer, as Boer for British.
103
scale of British violations. Moreover, such violations as did occur were

never deliberate policy on the Boer side.
In sum, the strategic interaction in the first and final interactions of the
war was ‘‘same-approach.’’ In the first interaction, from October 1899 to
March 1900, the British pursued a direct attack strategy which the Boer
countered with a direct defense strategy. Facing imminent defeat, the
Boer switched to a GWS in the second interaction of the war, from April
1900 to May of 1902. The British responded by switching to a barbarism
strategy in the final interaction of the war, from June of 1900. The out-
come was that the British won both the first and final interactions of the
war, and the Boer capitulated at Vereeniging on May 31, 1902.
Conclusion
The South African War was an asymmetric conflict between Great
Britain — then the world’s pre-eminent great power — and two of the
world’s tiniest states — Orange Free State and Transvaal.
Britain’s interests were many, but its willingness to go to war in South
Africa was due to its overriding interest: undisputed political control of
the four South African republics and, by extension, security for a vital
route of communications between Britain and its colonial possessions
in India. Contrary to the interest asymmetry thesis, which holds that the
stronger actor in an asymmetric conflict will have a low interest in a
conflict, Britain calculated that control of Orange Free State and
Transvaal was vital for three reasons. First, control of Cape Town was
key to maintaining sea communications with India. Second, and related
to the first reason, allowing the Boer republics to ‘‘dictate’’ to the British
might encourage the majority population of Cape Colony — Afrikaners —
to seize control of the Cape. Third and finally, allowing the Boer repub-
lics to ‘‘dictate’’ to the British would make the British appear weak-
willed in relation to its European great power rivals, thus encouraging
interference or intervention from those rivals. Employing a domino-like
logic throughout, the British therefore calculated that war was neces-
sary in South Africa.
Boer interests were in maintaining their independence from Britain.
The presidents of the two republics guessed that Britain’s real interests
were not — as they constantly maintained — the franchise rights of
Uitlanders and more civil conduct toward ‘‘Kaffirs.’’ They guessed
that Britain wanted to end their independence, and they were proved
right in the end.
104

interests in the outcome of an asymmetric conflict — is not supported
in the South African War. Relative power was a poor predictor of
relative interests in this case: at least on the British side. Did relative
power and interests explain political vulnerability?
Again, on the Boer side, it is difficult to answer this question. The lack
of a solid transnational Afrikaner national identity severely under-
mined the effectiveness of the later Boer shift to a GWS. Without such
an identity, or where such an identity was weak, the Boer GWS was
vulnerable to British amnesty offers:
General Christian De Wet later described these proclamations as
‘‘deadly lyddite bombs which . . . shattered Afrikanerdom.’’ Some
13,900 of the Boer commandos voluntarily surrendered their arms by
July 1900, this figure representing 26 percent of those liable for military
service in the two republics, or about 40 percent of the number origin-
ally mobilized on the outbreak of war. (Smith, 1996: 7)
We should have expected a ‘‘fight for their lives’’ to have made the
Boer defenders much more resolute than this. Yet, on the other hand,
when faced with a choice between their political independence and
their national survival (a choice which must in some sense depend on a
surprisingly benevolent conception of post-war British rule), the Boers
chose national survival:
Six months earlier, the tide had set finally against them — their com-
mandos immobilized and starving, beaten at last by blockhouses and
food burning; their womenfolk threatened by Kaffirs. And, to set
against all these sacrifices, there was no prospect of foreign interven-
tion, or any corresponding gain. Negotiate now, said Botha and De la
Rey and Smuts, while we still have control of our destiny, and can
keep the volk together as a nation. Fight on, and the volk will die (or
suffer a fate worse than death). The threat was not only to the lives of
individuals, but to the continued existence of the nation.
(Pakenham, 1979: 602)
British political vulnerability may be explained in part by its over-
whelming material superiority over the Boer republics. Certainly it
explains a large portion of Britain’s underestimation of the costs and
risks of a war with the Boer. But is relative material capability the best
explanation? No. Two other factors play into British political vulner-
ability and explain it better: regime type and time to objective.
As leaders of a democratic state, British policymakers could not make
policy without taking public reaction and judgments into account. In
105
terms of the political theory of the laws of war, there were two loci of
vulnerability for Britain: jus ad bellum — the justness of Britain’s resort to
force in South Africa; and jus in bello — the justness of its conduct of a war
in South Africa. In authoritarian regimes these vulnerabilities are man-
aged by means of control over access to information about both loci,
and by the fact that information notwithstanding, there is no mechan-
ism for connecting popular will with foreign policy. As we have already
seen, Kruger’s ultimatum spared British elites from difficulties over jus
ad bellum: the British public recognized the legitimacy of self-defense as
a justification for the resort to arms. But what of the moral conduct of
the war itself? Here is where Britain’s regime type made a big differ-
ence. The concentration camp controversy created the possibility of
Britain’s withdrawal from the war short of achieving its political object-
ives. We have already observed the reasons why this possibility was
never realized, but the point is that in an authoritarian Britain with-
drawal short of achieving the political objective due to jus in bello
concerns would not have been a possibility.
Regime type therefore matters more than allowed by the interest
asymmetry argument. However, the most important determinant of
vulnerability is time to objective. This case makes clear the importance
of time to objective in war: it is a positive cost on almost equal footing
with blood and treasure. Had the war been over by Christmas — as widely
anticipated by British elites and the British public — these concerns over
the legitimacy of a resort to force and the moral conduct of war would
have had much less space in which to operate. Indeed, in this general
category of conflict, we expect wars to be quick and decisive almost by
definition, since the gap in the ratio of resources available to each actor
exceeds five to one. If vulnerability is activated or caused by a delay in
time to objective, it then makes sense to ask what caused the delay?
It wasn’t arms diffusion. Although the Boer had better arms than the
British initially, their possession of a military technology advantage
correlates with their rapid defeat, not victory. Thus, although hypoth-
esis 6 — the better armed a weak actor is, the more likely it is that a strong
actor will lose an asymmetric conflict — receives some support here, on
balance the technological advantage enjoyed by the Boer did not affect
the outcome of the war. It marginally increased the costs to the British in
terms of killed and wounded soldiers, but only a shift in strategy saved
the Boer from defeat in March of 1900.
I argue that it was this shift in strategy to a GWS by the Boer in March
of 1900 that explains the delay which nearly forced Britain from the war.
106
The British won the first interaction of the war, and, after the fall of
Bloemfontein, the Boer leadership gathered to discuss surrender or a
strategy for continued resistance. After heated debate over the appro-
priateness of a GWS, and the risks under which it might place Boer
noncombatants, the Boer leaders opted to continue their resistance by
employing a GWS.26
It was five months before the British adjusted to this change in
strategy. During this time, Roberts advanced from city to city, capturing
each in turn and posting amnesty proclamations. When the Boer didn’t
give up, and in fact struck back at British communications and isolated
units, the British intensified their earlier policy of collective reprisals
against Boer noncombatants. They switched to a barbarism strategy,
and although this created its own risks in terms of British public
opinion, it also proved militarily effective against the Boer — especially
by 1901 (Pakenham, 1979: 500—501).
In evaluating the utility and effectiveness of barbarism as a COIN
strategy, however, we should keep in mind the evidence that the
strategy for shortening the war contained the powerful probability of
making a later peace costly or unstable:
Dundonald agreed that collective punishments were neither fair nor

politic. Attacks on the railway were not the work of locals; ‘‘and when
once these farms were burnt the country round became a desert and
their owners inveterate haters of the British.’’
(Pakenham, 1979: 479—480)
It must remain one of the great counterfactuals of the conflict, but it

seems likely that had the British not switched strategies, the war either
would have continued for a much longer time, or the continued delay
would have itself become a political issue which would have resulted in
at least a change of government, and very likely, a settlement short of
Britain’s ex ante political objectives (in other words, a defeat for Britain).
26
The decision also involved a recalculation of Boer interests: was it really the capital
cities that represented Boer freedom, or was it something else? Here Pakenham recounts
Boer deliberations following the capture of Bloemfontein: ‘‘But the man who had inspired
Botha and Smuts was Steyn. The Free State President had seen his own capital subjected to
the same humiliations two months before, and had realized that it was not a city, but the
illimitable veld, that was the true symbol of the volk. When Kruger’s despairing telegram
reached him on 1 June in his hide-out near Lindley (by an oversight, Roberts had left the
telegraph lines to the northeastern Free State intact), Steyn’s reply was characteristically
blunt. We shall never surrender . . . Steyn’s reply was the most important telegram of the
war’’ (Pakenham, 1979: 457—458).
107
It is also clear that, as in the Murid War, the barbarism strategy lost in
post-war political effectiveness anything it might have gained in war-
time military effectiveness:
And the backward nation of farmers from whose misrule the British
were going to save South Africa ended by being granted self-government
in the Orange River Colony and the Transvaal. Just eight years later,
rallying around an Afrikaner nationalism at least partly fueled by the
memory of the thousands dead in the concentration camps, whites in the
Cape Colony and Natal joined those of the former republics to become
citizens of the Union of South Africa in the beginning of a new era of
white government in South Africa. (Krebs, 1992: 54)
In sum, strategic interaction is the strongest explanation of the South

African War’s outcome. Once it became clear that the capture of the
Boer capitals would not force the Boer to surrender, the British had
three strategic choices. First, continue with the present strategy, per-
haps sending more troops, equipment, and horses to the theater.
Second, offer the Boers terms: something which could save face yet
end the war. Third, switch to a barbarism strategy: take the gloves off
and go after Boer noncombatants, either by holding them hostage or by
killing them outright in reprisal for continued resistance. The Boer
faced an even starker choice after the fall of Bloemfontein: either give
up or switch to a dangerous GWS. They chose a GWS, and the war
lasted another two years. During that period, which opened a window
of political vulnerability, British policymakers had to face the possibi-
lity that outrage over the delay itself, but especially over the concentra-
tion camp deaths, might actually force them from the war, or at least
force them to make political concessions which would have destroyed
the justification for going to war in the first place.
108
5 Italy in Ethiopia: the Italo-Ethiopian
War, 1935–1940
We are valiant formations

Of indomitable warriors
Carrying the flags
Into Eastern Africa.
The symbol of glory,
Of peace and civilization
Will be made to shine once more
By the Italic Victory . . .
Slavery and barbarity
Are doomed to vanish
The very moment
The Roman eagles appear . . .
From an Italian marching song in the Abyssinian campaign
We flung ourselves on the machine-gun nests to clear them out, we

hurled ourselves against the enemy artillery. We held firm against
their bombs and their containers of mustard gas. We put their tanks
out of action with our bare hands. We cannot reproach ourselves with
any taint of cowardice. But against the invisible rain of deadly gas
that splashed down on our hands, our faces, we could do nothing. Yet
I say again that we have no cause to be ashamed: we could not ‘‘kill’’
this rain. Haile Selassie
The war between fascist Italy and Ethiopia began on 2 October 1935.
Italy was the strong actor by a wide margin.1 It was a sharp conven-
tional engagement between a well-armed and well-supplied yet poorly
led Italian army, and the poorly armed and poorly supplied Ethiopian
1
Using Singer and Small’s data (1992), the halved ratio of relative material power comes
to 24.23 to 1 in favor of Italy.
109
army. By May of 1936, the emperor of Ethiopia, Haile Selassie, had fled,
and the Italians marched triumphantly into Addis Ababa, the capital.
Yet the war did not then end. It continued as a guerrilla war from 1936
to 1940, when Britain joined the fight. The Italians surrendered to the
British in North Africa on May 5, 1941.
The Ethiopian theater

Ethiopia lies in the centermost portion of the Horn of Africa. In 1935, it
was surrounded by the colonial possessions of four European states. To
the southwest and northwest lay Kenya and Anglo-Egyptian Sudan. To
the south and east Italian and British Somaliland respectively. To the
east and northeast, French Somaliland (Djibouti) and Eritrea (then a
colony of Italy).
Ethiopia was thus landlocked. Its main route to the sea was a single
railway line running from Addis Ababa, the capital, to the French-
controlled port of Djibouti on the Red Sea. It is a country of two
distinctive topographies and climates. From Addis Ababa south to the
Somalilands stretches a hilly and mountainous country which slowly
descends to the Ogaden desert: hot, dry, and flat. From Addis Ababa
north to Sudan and Eritrea, stretches some of the most rugged and
impassable mountains in all Africa.
The country had but one paved road, which had been built for the
emperor’s coronation in 1930. The remainder of the country was tra-
versed by means of mule tracks and undeveloped roads, often impas-
sable during the spring and summer rains, which regularly drenched
the region from May until September of each year.
Ethiopia’s people spoke Amharic, an ancient Coptic Christian dia-
lect, and most Ethiopians were Christians. The state itself consisted of a
federation of semi-independent regions each ruled by a noble family,
and all under the nominal control of an emperor in Addis Ababa. It was
a feudal system. Del Boca provides a description of the capital,
Ethiopia’s best-developed city in 1935:
There was only one generator, the generator that supplied electric light
and power for the ‘‘Little Gebbi,’’ the emperor’s palace, only one pas-
sable hotel, the Imperial, and only one hard-surface road, the one
which had been asphalted for Haile Selassie’s coronation. There
were no drains, no system of waste disposal; all the refuse was simply
dumped into the streets and the sky was always black with vultures
and crows. (Del Boca, 1969: 31)
110
Italy in Ethiopia
Such descriptions, though accurate, would later be used by Italy to

justify its conquest and colonization of Ethiopia.
Italian interest in Ethiopia dates from the waning years of the nine-
teenth century. Having established a viable colony in Eritrea, the
Italians decided on a military campaign further inland. On January
26, 1887, a force of 500 Italians were annihilated by a force of 20,000
Ethiopians led by Emperor Yohannes IV at Dogali.2 This forced the
Italians to postpone their conquest efforts for several years.
In fact, after the death of Yohannes and the accession of Menelik II,
relations between the two states thawed, and these friendly relations
were marked by the signing of a treaty and the gift from Italy’s King
Umberto to Menelik of 28,000 rifles and 28 cannon. In the Treaty of
Uccialli, Menelik recognized Italy’s right to Eritrea and agreed to allow
Italy to conduct some Ethiopian foreign affairs. But after the Italians
chose to interpret the treaty as recognition that Ethiopia had become a
‘‘protectorate’’ of Italy, relations quickly soured. In order to soothe
Menelik and convince him of the advantages of protectorate status, in
1893 the Italians made him a gift of 2,000,000 cartridges for the rifles
they had previously given him. No sooner had Menelik received the
cartridges than he denounced the treaty and began using the threat of
Italian invasion to rally the Ethiopian chiefs around him:
The Italians then very obligingly fostered his efforts to develop
Ethiopian nationalism and solidify his own power. Using his treaty
denunciation as a pretext, they organized, under the leadership of
General Oreste Baratieri, the Governor of Eritrea, an expeditionary
force of 17,700 men, a considerably larger force than other colonialist
countries were accustomed to using, and marched south into Ethiopia.
(Coffey, 1974: ix)
But the Italians chose an invasion route which appeared poised to

cross through Axum, one of Ethiopia’s most holy sites and the spiritual
center of the Coptic Church. This allowed Menelik to rally more chiefs
to his cause, and he assembled an army of more than 100,000 warriors
armed with, among other things, the rifles and ammunition the Italians
had already given him.
But fielding an army of such size proved difficult for Menelik. His
hungry soldiers threatened to desert him when the Italians, already
established in an impregnable position at Sauria, decided to end the
2
Such apparently prohibitive odds were the norm on colonial frontiers.
111
campaign quickly by launching a surprise attack on the assembled

Ethiopians.
Baratieri divided his forces into three columns who were all to reach
key hills overlooking the Ethiopian positions and then, at the proper
signal, descend in a crushing pincer movement to annihilate the
Ethiopians. Unfortunately for the Italians, Baratieri failed to adequately
scout the approaches and his maps were inaccurate. The columns
became separated and the waiting Ethiopians wiped out a flank column
before it could reach its assigned positions. The next morning (1 March
1896), as the remaining two columns went into action near Adowa they
were unaware that their left flank was entirely unprotected. After being
halted by lead elements of the Ethiopian army, the Italians were envel-
oped and though they fought bravely, 12,000 died before the sun set,
ending the battle (the 5000 survivors owed their lives to this Ethiopian
tradition of ending every battle at sunset).
This crushing defeat at the Battle of Adowa marked the only success-
ful repulse by an African state of a European attempt at colonization.3 It
solidified Ethiopian nationalism and proved an intensely humiliating
loss for the Italians — so humiliating that it would still count as a
primary casus belli for fascist Italy some thirty-nine years later.4
In the meantime, Italy licked its wounds, nursed its resentments, and
waited. In 1906, it signed a secret treaty with France and Britain, in
which the latter two states:
acknowledged Italy’s priority in Ethiopia, with the stipulation that
Italy would never hinder the operation of the French-owned railway
3
It also runs counter to the predictions of the strategic interaction thesis: a strong actor
and weak actor both used direct strategic approaches and the weak actor won.
4
Italy’s fascist dictator, Benito Mussolini, was 13 years old at the time of the Battle of
Adowa, and he personally experienced Italy’s humiliation over the defeat (Coffey, 1974:
22). Del Boca characterizes the effects of this humiliation this way: After Adowa, Alfredo
Oriani had written:
We have signed a peace but there will be no peace. We will never give up
Africa — the war will be resumed. The defeat seared their minds and the words
of the boastful patriotic ditties the soldiers had sung, words such as these stung
them like whips:
Baldissare, hey, beware
Of these black folk with woolly hair!
Menelik, you’re dead — here comes
A shower of lead, not sugar plums.
By and large Italians could not bring themselves to accept the fact that the
‘‘barbarous Shoan hordes’’ had succeeded in annihilating a European army.
(Del Boca, 1969: 9—12)
112
Italy in Ethiopia
from Addis Ababa to Djibuti, or the flow of Blue Nile water from
Ethiopia’s Lake Tana into the White Nile, which fed England’s depend-
encies, Egypt and Sudan. (Coffey, 1974: 11)
The eve of war: Italian interests

Benito Mussolini came to power in Italy in 1922, when Italy’s King
Vittorio Emmanuele called on him to form a government. During the
two years it took him to consolidate his power by intimidating or
assassinating his opponents, Ethiopia petitioned for membership in
the League of Nations:
When the Ethiopian government, at Tafari’s instigation, applied for
League of Nations membership in 1923, Mussolini supported the
application because he didn’t think he could defeat it, even though
England opposed it. The British argued that Ethiopia should not be
admitted into a society of civilized nations because Ethiopians still
practiced slavery . . . Ras Tafari had already taken steps, not always
effective, toward abolishing slavery. His application prevailed and
Ethiopia became a League of Nations member. (Coffey, 1974: 13)
Essentially, Britain had argued that Ethiopia was not civilized

enough to merit membership.5 After years of bitter internecine quar-
rels, Ras Tafari, a powerful Shoan noble, was later crowned emperor of
Ethiopia under the throne name of Haile Selassie.
The emperor was an educated, intelligent, and charismatic man who
spoke French fluently and planned to reform and develop Ethiopia
along the lines of a European state.6 He was handicapped by his
country’s poverty, and by the difficulty of establishing and maintaining
central authority over the independent princes who ruled Ethiopia’s
distant regions. Above all, he was distracted by fascist Italy’s increas-
ingly obvious acquisitive intentions.
Selassie had attempted to reach an accommodation with Italy in the
Italo-Ethiopian Treaty of Amity, Conciliation and Arbitration of August 2,
5
This was a theme which was to underlay all discussions and negotiations among the
three European powers. It is difficult to estimate how much this influenced the bargain-
ing, but the memoirs of the participants make it clear it proved to be a major, if at times
unspoken, factor (see below).
6
He maintained a number of foreign advisors close to the throne, including an American,
Everett Colson, a Swede, General Eric Virgin, and an Englishman, Sir Sidney Barton (His
Britannic Majesty’s minister). His coronation on November 2, 1930 was directly modeled
on the coronation of King Edward VII in London.
113
1928, but Mussolini tended to use the more liberal provisions of the
treaty to prepare for his invasion of Ethiopia. Del Boca recounts how
Mussolini’s efforts to build a road from Assab to Dessie was stymied by
the emperor (although technically allowed under the treaty), and how
Mussolini’s propaganda machine characterized the emperor’s objec-
tion as tantamount to the rejection of civilization:
The emperor’s reluctance to furnish the country which had a common

frontier with the most rapid route for an invasion was fully justified,
of course, and Mussolini knew it. But from this time on we were the
archangelic messengers of light and civilization, the Ethiopians the
lowest, most deeply damned race on earth, the personification of all
that was evil and abominable. Utterly puerile as this counterblast was,
the majority of Italians honestly believed this rubbish was true;
consequently, when the hour struck, they gladly acclaimed the act of
aggression. (Del Boca, 1969: 14)
From 1930 on then, Italy began preparing for the invasion, conquest,
and occupation of Ethiopia. All it needed was a decent pretext for
invasion.
The incident which proved to be the ‘‘Sarajevo’’ of the Italo-Ethiopian
war took place at the distant outpost of Wal Wal in the Ogaden desert on
December 5, 1934. Mussolini had already sent a large expeditionary
force to Eritrea consisting of 200,000 troops, 7000 officer, 6000 machine
guns, 700 cannon of every caliber, 150 tanks, and 150 pursuit and
bomber planes (Del Boca, 1969: 21). But Selassie, clearly understanding
Mussolini’s intentions, had prudently moved all his armed forces a safe
distance away from Eritrea, precisely so as to avoid any chance of a
border ‘‘incident’’ which Italy could inflate into a pretext for invasion.
He also set about seeking help under the articles of the Covenant of the
League of Nations.
But in the Ogaden the Italians had stationed troops at Wal Wal, 60
miles within Ethiopian territory. When a survey commission escorted
by Ethiopian troops journeyed to the wells there, they were halted by
Somali Dubats (African soldiers under Italian command). When the
Ethiopians protested that they would not withdraw because Wal Wal
was Ethiopian territory, the Italians replied that Wal Wal was Italian
territory — part of Italian Somaliland. The Ethiopians at first far out-
numbered the Italians, and they refused to withdraw because to do so
might be construed as acknowledging Italian claims on the area.
Tensions increased, as did Italian reinforcements, until on 5 December
114
Italy in Ethiopia
1934, shots were fired and in the mêlée which followed, the Ethiopians
were routed.7
What was most striking about the incident was the way in which Italy
managed it, and the way in which the world powers reacted. Italy
instantly claimed it had been the victim of unprovoked aggression on
its own territory; it demanded, in the most insulting terms possible, an
apology and reparations from the Ethiopians. The Ethiopians protested
that they had been attacked by Italians 60 miles within their own
borders, and made every effort to downplay the incident, while at the
same time refusing to acknowledge guilt or apologize. At the League of
Nations, most representatives simply assumed the Italian territorial
claims were correct, and Ethiopia was strongly encouraged to accede
to Italy’s demands.
Despite a few rumors that Wal Wal might actually be in Ethiopia, most
newspapers accepted the Italian assertion that it was in Somaliland,
especially after Italy sent a note to the League on December 16 renew-
ing the assertion and decrying Ethiopia’s charges.
Eventually, a correspondent in the press room of the League of
Nations . . . on or about December 20, did something no one else had
thought to do. He looked up at a map on the wall in front of him. It was
a map of Africa, issued by the Italian Geographical Institute at
Bergamo, which showed that Wal Wal was indeed at least sixty miles
inside Ethiopia according to the terms of the 1897 treaty which settled
the 1896 war between the two countries.
The Italian delegation, apprised of this evidence, reacted immedi-
ately. Baron Pompeo Aloisi, Rome’s . . . representative in Geneva,
demanded that the map be removed from the press room because it
was obsolete. It took no account of certain modifications of the 1897
treaty which were made in 1908.
No sooner had the map been removed than the correspondents, now
full of enterprise, went all the way to the League library in the north
wing of the main building, where they found an Italian government
map of Ethiopia, issued by the Colonial Office in 1925. This map also
showed that Wal Wal was at least sixty miles inside Ethiopia.
(Coffey, 1974: 19)
In short, regardless of who fired the first shot, Italy was in the wrong;
having invaded Ethiopia and established an armed outpost within
Ethiopia’s territory. But none of this made any real difference. Italy
7
It was impossible to answer the question of who fired the first shot (Coffey, 1974: 8; Del
Boca, 1969: 19), and the battle itself was decided by Italian armored cars, against which the
Ethiopians had no effective countermeasures.
115
continued to bluster, fabricate, and bully; in the end Mussolini used the
incident as a way to justify sending more troops and supplies to the
theater. The Ethiopians patiently applied for League assistance — even
arbitration — but none was forthcoming. The two dominant great
powers in the League — Britain and France — each negotiated separate
deals with Mussolini in which the end result was to give Italy a free
hand in Ethiopia.
What then were Italy’s true interests in Ethiopia? The Wal Wal inci-
dent is useful in this regard, because as Mussolini’s long-awaited pre-
text for the conquest of Ethiopia, it also prompted the Italians to
advance arguments and justifications for that conquest. These come
down to three things: reputation, responsibility, and resources.
National humiliation over the defeat at Adowa was deeply felt by
most Italians, Mussolini personally,8 and the military especially. A big
reason for the attack and conquest of Ethiopia then, would be to erase
that humiliation.9 Beyond this simple redress, however, there were two
other reputation issues at stake: the reputation of Italy as colonial power
and the reputation of fascism as a political and ruling ideology within
Italy. Italians had also been humiliated by their lack of promised spoils
from their ‘‘victory’’ over Germany and Austria-Hungary in World War I
(Coffey, 1974: 25).
The responsibility interest centered on Italy’s conception of itself — a
conception which gained sympathy from Britain and France at the
time10 — as a civilizing European power:
8
Coffey adds that in a secret memorandum Mussolini argued for a fait accompli strategy
in Ethiopia. Mussolini calculated that so long as British and French interests in the area
were respected, he could get away with the conquest: ‘‘[Mussolini’s] dreams of colonial
glory and his bitter boyhood memories of Italy’s defeat at Adowa forced him to hope so.
‘This problem has existed since 1885,’ he concluded in his secret memorandum’’ (Coffey,
1974: 22). Griffin points out that it is typical of fascism to refer to aggressive colonial myths
and aspirations as ‘‘problems,’’ which must then be ‘‘solved’’ by military or totalitarian
measures (Griffin, 1995: 74).
9
Baron Aloisi, Italy’s representative at the League of Nations, makes this link clear:
‘‘There was the earlier defeat at Adowa which would have to be ‘washed out in blood’’’
(Coffey, 1974: 73).
10
Here is how Britain’s foreign minister, Sir Samuel Hoare, put it to the House of
Commons: ‘‘We have always understood Italy’s desire for overseas expansion,’’ he told
the House. ‘‘In 1925 [it was actually 1924] we ceded Jubaland [a bleak piece of East Kenya
bordering Italian Somaliland] to Italy. Let no one in Italy suggest that we are unsympa-
thetic to Italian aspirations. We admit the need for Italian expansion’’ (Coffey, 1974: 105).
This ‘‘need’’ for colonial possessions was therefore a norm of European power politics,
and hence a legitimate aspiration for European powers outside of Europe. There was
another logic at play: precedent setting: ‘‘An Abyssinian victory might result in serious
116
Italy in Ethiopia
[Mussolini] also proclaimed that Italy could ‘‘civilize Africa, and her
position in the Mediterranean gave her this right and imposed this
duty on her.’’ Then as a warning to other colonial powers he said Italy
did not ‘‘want earlier arrivals to block her spiritual, political, and
economic expansion.’’ (Coffey, 1974: 21)
The negative argument had the form tu quoque: after all, Italy was only
asking to do in Ethiopia what all the other European powers had done
for centuries with impunity (Del Boca, 1969: 26; Coffey, 1974: 59—60, 75).
Finally, the weakest of the three was the resources interest. Mussolini
calculated that developing and colonizing Ethiopia would be expen-
sive, and might take as long as fifty years to return dividends.
Nevertheless, such arguments were common among Italy’s efforts to
justify its aggression: ‘‘Italy had no other choice but to expand.’’ Part of
Italy’s resources problem had in fact been manufactured by the success
of Mussolini’s earlier efforts to increase the Italian birthrate.
A final argument has to do with Mussolini himself and his craving for
personal power. Coffey argues that part of Mussolini’s ambitions
abroad were pathological: since he had already acquired as much
power from the Italian people as they had to offer, his satisfaction
came to depend on gaining power over others abroad (Coffey, 1974:
20). Mussolini recognized all these interests in conquest, but his trump
card interest was one we have encountered before: self-defense
(Mussolini, 1995: 74—75).11
The point is that Italy’s preponderance vis-à-vis Ethiopia did not
reduce its interests in conquest or colonization.12 It was not greed
disaffections within British possessions. The Abyssinian victory over the Italians at
Adowa in 1896, he said, was one of the governing factors which led to the Anglo-
Egyptian campaign in 1898, because British authorities then felt strongly that the blow
to white prestige in that area required a victorious campaign of whites against blacks’’
(Coffey, 1974: 112). Of course, the real difficulty with this argument was that Ethiopia and
Italy were not yet at war, and preventing war would have been relatively simple (close or
threaten to close the Suez Canal). Moreover, Ethiopia was entitled to assistance under the
provisions of the League Covenant. In other words, to prevent the proposed domino
effects of a white defeat in black Africa, all that was necessary was to prevent the war.
11
When Ethiopia’s skillful diplomacy forced Italy to tip its hand, Mussolini, now
deprived of the opportunity to realize his fait accompli strategy, resorted to a sunk costs
variant: the war must go on because Italy has already spent millions of lire on the
adventure (see Coffey, 1974: 75, 142).
12
Mussolini maintained that it was not bad generalship which lost the day at Adowa, but
insufficient resources: ‘‘For the lack of a few thousand men, we lost the day at Adowa. We
shall never make that mistake. I am willing to commit a sin of excess, but never a sin of
deficiency’’ (Del Boca, 1969: 21). Unlike the constraints faced by democratic Britain in the
South African War, fascist Italy could send overwhelming forces (and illicit weapons) to
Ethiopia without fear of public recrimination.
117
alone which motivated Italy to seek out a colony in Ethiopia. Nor was
the desire to conquer Ethiopia an entirely fascist affair: colonies had
long been a dream of Liberal Italy. Italians came to see their conquest of
Ethiopia as both necessary and just, and this opinion was distributed
across classes and political affiliations in Italy.13 Most depressing of all,
Italy’s aims evoked more sympathy than condemnation from other
states. The fact of Ethiopia’s membership in the League of Nations
achieved little more than frequent embarrassment for Britain; and
Ethiopia’s careful and patient efforts to secure for itself the help to
which it was entitled under the provisions of the League Covenant
ultimately came to nothing.
The eve of war: Ethiopian interests

As a state, Ethiopia was considerably less unified than Italy. Its nominal
ruler was Haile Selassie, an emperor with titular control over hundreds
of allied princes, most of whom jealously guarded their independence
but nevertheless remained loyal to the emperor.
Ethiopia’s interests were survival, but it is clear from all sources that
up until Italian and Eritrean soldiers actually crossed the border into
Ethiopia in 1935, and even for some months afterward, the emperor
counted on outside intervention from his de jure allies: Britain, France,
and other influential members of the League of Nations. In other
words, many of the early battles — among the bloodiest and hardest
fought engagements — were not motivated by a sense of inevitable
doom. What then explains Ethiopian interests in the first three months
of the war?
The answer is honor: personal, familial, tribal and, possibly, national.
All Ethiopian boys were raised on the tales of the defeat of the vaunted
Italian army at Adowa at the hands of Emperor Menelik. Ethiopia’s
barefoot and lightly armed warriors sacrificed their lives bravely and in
their thousands as a consequence of this deeply ingrained sense of
honor.
13
Del Boca argues that ‘‘The vast majority of Italians, particularly the younger generation,
hailed the colonial enterprise with sincere enthusiasm. They were fighting for a place in
the sun that other great powers had enjoyed for years or for centuries, and to a country as
poor and overpopulated as Italy, the conquest of Ethiopia meant jobs and a patch of land
for millions of unfortunates’’ (Del Boca, 1969: 26).
118
Italy in Ethiopia
The Italo-Ethiopian War

The Italo-Ethiopian war began on October 2, 1935, when Haile Selassie
reluctantly ordered the mobilization of his 30,000 northern troops to
their defensive positions opposite an expected invasion of three Italian
armies. The next morning, 250,000 Italian troops under the command of
General De Bono crossed into Ethiopia from Eritrea.14 The war was
fought in five strategic interactions.15
Interaction one: conventional attack vs. conventional defense

The fight for Ethiopia began as a contest between the superior (in
numbers and technology) forces of Italy and the poorly organized and
poorly supplied forces of Ethiopia.
Ethiopia’s emperor had forced his eager generals to respect a wide
buffer zone between Eritrea and Ethiopia, and so the initial weeks of the
war were anticlimactic. The Italians, with advanced radio communica-
tions, motorized transport, and total air superiority, advanced by cau-
tious steps from Eritrea toward their first objective: Adowa. They were
not seriously opposed.
Military analysts at the time, in fact, questioned Italy’s choice of
invasion routes. Why advance through some of the least developed
and most vehicle-impassable terrain in Africa, when Italy could simply
have advanced through the Ogaden from Italian Somaliland right up to
Addis Ababa? The answer is that Italian strategy was also guided by the
need for revenge: to erase the humiliation of Adowa it was necessary to
capture it — preferably after a hard fight and a decisive victory over the
Ethiopians. But such was not to be.
In fact, these earliest weeks of the war saw virtually no fighting. The
war took on the following pattern. Italian troops, led by the Eritreans,
would advance a few miles at a time under cover of air support. Such
skirmishes as did take place usually pitted these excellent Eritrean
troops against the equally excellent Ethiopians. After the superior fire-
power of the Eritreans decided the issue, the Ethiopians withdrew, and the
14
These troops included ‘‘two highly efficient [Eritrean] divisions — the 1st and 2nd
Eritreans’’ (Mockler, 1984: 54). The Eritrean divisions were made up of black Eritrean
soldiers and NCOs under the command of white Italian officers.
15
After Italy declared war on France in June of 1940, the Ethiopian conflict became a fight
between Britain and Italy. In this sixth and final interaction, the contest was decided by
conventional forces using conventional attack and defense strategies. In May of 1941
Italy’s forces in Ethiopia collapsed and surrendered. Because the conflict ceased to be
asymmetric as defined in this analysis, this final interaction is not discussed here.
119
Italians advanced cautiously again. The final advance into Adowa was
in fact unopposed by the Ethiopians.16 Anxious Italians, and especially
Mussolini, seemed disappointed by the lack of a decisive engagement.
Meanwhile, the emperor set about rallying, organizing, and moving
his troops to meet the invader. Ethiopians had no air support and no
motor transport. They had very few radio sets, and when they did
communicate using these sets, they were forced to do so in the clear
(which meant that, combined with total air superiority, the Italians
enjoyed strong communications intelligence). Most of the emperor’s
soldiers went barefoot, and many did not even have rifles, but were
armed instead with spears and swords. Many of those who did have
rifles, had old ones, or had only a few rounds of ammunition for them.
The logistical difficulties of mobilizing a coordinated conventional
defense against the Italians were therefore staggering.
In addition to mobilization difficulties, the emperor was being begged
by his most loyal generals to oust a number of traitors before they
revealed the emperor’s defensive plans, or defected at a decisive moment
in a crucial battle. But when faced with evidence that one of his generals,
Gugsa, was in the pay of the Italians, the emperor refused to take action
against him, claiming that many of his generals took money from the
Italians but they were all loyal to Ethiopia. Gugsa was a traitor, however,
and he had promised the Italians he would bring 10,000 men with him to
the Italian side. In November, he made his move, but both he and the
Italians were embarrassed to discover that by the time he reached Italian
lines his much-vaunted 10,000 men had dwindled to a mere 1200. Gugsa
himself proved to be a thorn in the Italians’ side, constantly pestering De
Bono to advance to Makalle (the seat of Gugsa’s family’s power).
But De Bono was cautious. He responded to Mussolini’s increasingly
frequent demands for decisive action by requesting delay after delay
before advancing:
Mussolini’s bewilderment and impatience at such telegrams from de

Bono was hardly surprising. De Bono had spent almost a year preparing
16
Coffey highlights what would soon become a morale problem for the Italians: ‘‘The
reason the Third Eritrean Brigade had not been ‘employed in the occupation of the valley,’
after spearheading the drive to capture it, was that they were black, and the honor of the
‘reconquest’ of the area, as soon as it had been rendered safe by the blacks, was reserved
for the white Italians. The monument whose unveiling he witnessed was inscribed: to the
dead of Adowa, avenged at last’’ (Coffey, 1974: 184). The prowess and battle honors of
the black units soon began to cause serious morale problems for the Italians, especially the
Black Shirt Divisions (see Coffey, 1974: 262).
120
Italy in Ethiopia
for the invasion of Ethiopia. Then, after an unobstructed advance of

twenty-five miles into the country in three or four days, during which
time his airplanes hadn’t been able to find any enemy armies, he had
decided it would be too dangerous to continue, even to a place like
Makalle, the Ethiopian ‘‘defender’’ of which was now in his employ.
(Coffey, 1974: 194)
Delays in Ethiopian mobilization, combined with excessive caution

on the Italian side, best explain the relative lack of action in the theater
of operations until December. In November, Mussolini finally grew
weary of exhorting a virtually unopposed De Bono to advance and
engage the enemy, and replaced him with Badoglio. Unfortunately,
Badoglio, who had complained behind De Bono’s back that De Bono
was not aggressive enough, soon began sending his own telegrams
requesting delays and more supplies. Unlike De Bono, however,
Badoglio was about the meet the enemy in force.
The Battle of Tembien

The emperor was no fool. After assessing the terrain, and the strengths
and weaknesses of his own soldiers, he determined that the only hope
of success lay in a GWS. On the first day of mobilization, October 2, he
had advised his generals and soldiers this way:
‘‘Soldiers, I give you this advice, so that we gain the victory over the
enemy. Be cunning, be savage, face the enemy one by one, two by two,
five by five in the fields and mountains.
Do not take white clothes, do not congregate as you’ve done now.
Hide, strike suddenly, fight the nomad war, snipe and kill singly.
Today the war has begun, therefore scatter and advance to victory.’’
In these very words lay the key to a possible victory. the Emperor
knew if he could persuade his Ethiopian warriors and chiefs to con-
centrate on guerrilla warfare against the Italians, they might have at
least a slim chance to win. But it was not easy to tell an Ethiopian
warrior he should sneak up behind his enemy and attack him from the
cover of a rock or bush. To these proud men, a war was series of battles
or one climactic battle. Each battle should properly be fought in a
single day, from dawn to dusk, in an open field, with the winner of
the last battle being acknowledged by both sides to be the winner of
the war. (Coffey, 1974: 162)
In this passage lay both the possible salvation of Ethiopia, and the
explanation of its doom. Ethiopian warriors marched and fought in the
open and wore the white shamma (a kind of toga) instead of camouflage.
121
Selassie was soon forced to acknowledge, therefore, that his ideal strat-
egy could not be pursued for cultural reasons (Coffey, 1974: 311).
By mid December both sides were bracing for their first major con-
frontation. After directly ordering Badoglio to advance to Makalle, the
left wing of the Italian army had advanced and captured the town in
November. Enduring constant bombardment from the air, the massed
armies of Ras Seyum, Ras Kassa, and Ras Mulugeta attacked the weak-
est point in the Italian line at Tembien. Suffering frightful losses, they
crossed the Takkaze ford and captured a number of key bridges and
passes. Within days the advance threatened to cut the Italian army in
two, forcing it to retreat back to Eritrea in ignominy.
For the Italians, the situation was as critical as it was unexpected. The
poorly coordinated attacks of Ras Kassa and Ras Seyum had proven the
bravery and ferocity of the Ethiopian soldiers. They swarmed and
destroyed Italian tanks and armored cars, often armed only with spears
and swords. Their astonishing local mobility enabled them to fight
hand-to-hand, a form of fighting at which they excelled, and against
which Italian air support was useless. By January the situation looked
grim:
The trackless terrain was ideal for the Ethiopians and suicidal for the
Italians. Only airplanes could get at the advancing Ethiopians in such
country. But since they were now learning to spread out and take
cover, the airplanes would have to be armed with something more
effective than bombs and bullets.
Because of the foresight of Mussolini and De Bono, Badoglio found
that he did have something more effective, and though his country
had signed an international agreement never to use it, Mussolini, just a
few days previously, on the sixteenth, had reiterated an earlier author-
ization to use it. Badoglio could congratulate himself now for his
foresight in banishing the news correspondents to Eritrea. As long as
they didn’t actually see what he was about to do, they would probably
believe him later when he denied having done it. (Coffey, 1974: 263)
There is some controversy over the Italian decision to use gas, and
over its existence in theater.17 The best evidence, however, supports the
17
Del Boca argues that De Bono had refused to avail himself of this option, and that it was
Badoglio who took the initiative in its use at the prompting of one of Mussolini’s ‘‘Roman
rages’’ (Del Boca, 1969: 78). The resort to the use of gas in the south seems especially
egregious, because Graziani’s 60,000 well-supported troops were opposed by 7000 starv-
ing Ethiopians under the command of Ras Desta. Yet Graziani still felt ‘‘compelled by
necessity’’ to resort to bombing enemy positions with mustard gas canisters: ‘‘A few days
after the arrival of the Duce’s telegram, on Badoglio’s orders, the Takkaze fords were
122
Italy in Ethiopia
claim that Badoglio (and Graziani, commander of the 60,000 Italian

troops advancing from Italian Somaliland) resorted to the use of mus-
tard gas (yperite) in order to compensate for his own incompetence. The
use of mustard gas was a clear violation of the laws of war. The Italians
themselves had signed the Geneva Convention of June 17, 1926 pro-
hibiting the use of gas as a weapon of war except in self-defense against
a gas attack (Del Boca, 1969: 80). The Italians had earlier justified their
shipping of mustard gas (680 tons) to the East African theater by
arguing that they needed it for self-defense in the event of an
Ethiopian chemical weapons attack (Coffey, 1974: 104). This was
implausible, to say the least. Against the defenseless Ethiopians the
air-dropped gas crushed the Ethiopian threat to Badoglio’s communi-
cations. The Italians could now breathe easy again.
Interaction two: conventional attack/barbarism vs.

The use of gas altered the entire character of the war. The Ethiopians
had been prepared — psychologically and, to some extent, tactically — to
deal with bombing, artillery, machine guns, and even tanks, but they
could not fight the gas. Here Ras Imru describes what it was like to
experience a gas attack:
I myself narrowly escaped death. On the morning of December 23,
shortly after we had crossed the Takkaze, we saw several enemy
planes appear. We were not unduly alarmed as by this time we were
used to being bombed. On this particular morning, however, the
enemy dropped strange containers that burst open almost as soon as
they hit the ground or the water, releasing pools of colorless liquid.
I hardly had time to ask myself what could be happening before a
hundred or so of my men who had been splashed by the mysterious
fluid began to scream in agony as blisters broke out on their bare feet,
their hands, their faces. Some who rushed to the river and took great
gulps of water to cool their fevered lips, fell contorted on the banks and
writhed in agonies that lasted for hours before they died. Among the
victims were a few peasants who had come to water their cattle and a
number of people who lived in nearby villages. My chiefs surrounded
me, asking wildly what they should do, but I was completely stunned.
I didn’t know what to tell them, I didn’t know how to fight this terrible
rain that burned and killed. (Del Boca, 1969: 78—79)
drenched with mustard gas, while Graziani — in his own words — ‘terrorized the inhab-
itants’ of the villages scattered round Jijiga by dropping on them container after container
of poison gases’’ (Del Boca, 1969: 78).
123
It was not just soldiers and retreating soldiers who were gassed in
these aerial assaults, but lakes and streams and pastures as well.
Ethiopian soldiers relied for food on livestock they drove with them
to the battle site. The mustard gas killed not only women and children
but also these cattle, by the thousands. Medical facilities were only the
most rudimentary, and such as there were had become subject to
frequent air attacks by Italian bombers. Since many of these hospitals
were staffed by volunteers from ‘‘civilized’’ countries, such as Sweden
and Britain, Mussolini had ordered these strafings and bombings to
scare away witnesses to his use of gas. This created the rather ridiculous
situation in which the Italians were committing one war crime (attack-
ing clearly marked Red Cross hospitals) in order to conceal another.
The Italian response to the increasing evidence of both practices
followed a typical pattern. It began with outright denial: that the
claim that photographic evidence had been fabricated,18 or that report-
ers were in fact communist agents. As evidence mounted, Italy argued
that it had been forced to do some of these things, but on a small scale;
and that they were justified as reprisals against Ethiopian atrocities (Del
Boca, 1969: 81).19
Whether Italy was believed or not, the end result was that a much-
hoped-for great power intervention on Ethiopia’s behalf never hap-
pened. Instead, the war dragged on, with the desperate Ethiopians
fighting a losing battle against a rain of death from the air.
18
Del Boca relates a chilling story which supports the argument that many in the Italian
military didn’t know about Italy’s resort to gas. It seems a general in Italy’s military
intelligence service and his aide had been snooping into the mail of a London news
correspondent when they came upon some suspicious photographs: ‘‘they found several
photographs of Ethiopians whose bodies were covered with sores. These photographs
struck them as extremely suspect and a few minutes later they were on my desk. I looked
at them and took them straight to Professor Castellano, then the leading authority on
tropical diseases. He examined them and confirmed what I had suspected — there could be
no doubt, he said, that the sores on the bodies of these Ethiopians had been caused by
mustard gas. We stared at each other in deep embarrassment. After an awkward silence,
Castellano added, ‘Still, leprosy produces almost identical sores,’ and he handed me a few
photographs of lepers so that I could compare them with the others. As I could not spot the
difference, I suddenly made up my mind: I would substitute the pictures of the lepers for
the original photographs and let the package go on its way’’ (Del Boca, 1969: 80). This later
enabled the fascist propaganda machine to declare the photographs, which later appeared
in print, as ‘‘fabrications’’ intended to falsely malign the honor of Italian arms.
19
Tales of Ethiopian atrocities had become a major fascination for average Italians.
Alleged abuses included mutilating Italian corpses (in particular, cutting off their geni-
tals), and especially the use of dum-dum bullets. The Italians also claimed that Ethiopian
soldiers were using Red Cross hospitals as shelters from air attack.
124
Italy in Ethiopia
Although the use of mustard gas killed or injured noncombatants as

well as combatants, its use in this interaction of the war was directed
against Ethiopia’s soldiers. In other words, it was used in order to
coerce Ethiopia by destroying its capacity to fight, not to punish
Ethiopia by destroying its will to fight. In this sense, Italian strategy
must best be described as conventional attack supplemented by
barbarism.20
After four more battles, and the liberal application of mustard gas on
retreating Ethiopian armies, the only remaining organized opposition
between Eritrea and Addis Ababa was the emperor’s Imperial Guard.
After consultations with his advisors and surviving generals in which
he attempted and again failed to convince them to adopt a GWS,
Selassie decided to march to meet the Italians at Mai Chew with the
best-trained and best-supplied forces Ethiopia could field. His plan was
to feint a frontal assault against the Italian main position and then
swing around with his main force to attack a weak point in the Italian
left. The plan might have worked, but it depended on surprise and a
degree of timing difficult to achieve under even the best of circum-
stances. In the event, the Italians intercepted a radio transmission con-
taining all the details of the emperor’s planned attack, and they altered
their defensive arrangements to counter it.
The morning of the battle, March 31, 1936, the Ethiopians moved into
position against the waiting Italians. Even with foreknowledge of the
Ethiopian plans, the Italians were nearly overwhelmed in a few places.
However, the tide of battle soon turned, and the emperor was forced to
retire.21
As at all of the other battles, the Italian air force liberally doused the
retreating Ethiopians with wave after wave of mustard gas mixed with
high explosive. The retreat became a rout, and the emperor fled to
20
This is supported by Sbacchi’s argument that the fascists considered mustard gas to be
‘‘a means of legitimate warfare’’ (Sbacchi, 1985: 73).
21
Del Boca argues that the Italian need for a decisive engagement was so great that many
felt actual disappointment over the low Italian (read: white) casualty figures in the battle:
‘‘While no words can do justice to the heroic stand made by the Alpini and the gallantry of
the Italian officers who set a shining example to their men,’’ the ‘‘battle for the empire’’ was
in fact won by the colored troops who shed their lifeblood so generously for the common
cause. This afforded little comfort to Mussolini who hoped to certify that, through the
state, his people had acquired a strength as solid and durable as cement. Indeed, for the
next few months, he was filled with bitter disappointment. Denis Mack Smith interpreted
his feelings correctly: ‘‘The Duce grieved over the fact that the fallen Italians had not even
numbered 2,000 and that the war had been won at too low a price to reinvigorate the
national character to the extent Fascism required’’ (Del Boca, 1969: 172).
125
Addis Ababa, closely pursued by the Italians. On May 2 he quietly

slipped out of the capital with his wife and family on a train bound for
Djibuti. The war was over and Ethiopia had been crushed.
Interaction three: barbarism vs. GWS

But the war was not over. It should have been, and might have been, but
for the continued use of mustard gas in the hunt for the emperor’s
remaining loyal generals. As noted above, the aerial application of
mustard gas (especially from sprayers attached to the wings) was a
far from discriminate weapon. Its use during the war itself had already
stirred much resentment and caused terrible tragedies in Ethiopia. Now
the Italians had won the war, but they continued to use yperite as a
counterinsurgency weapon,22 and many of Ethiopia’s water resources
were rendered deadly as a result. Livestock and mules continued to die,
as did ordinary men, women, and children entering streams to wash
themselves or their clothing.
On July 28, 1936, after almost two months of relative peace, a group of
resistance leaders launched a four-pronged assault on the capital. It was
repulsed, but after the rains ended in November, Graziani, the newly
appointed viceroy of Africa Orientale Italiana (or Italian East Africa:
AOI), ordered the capture or destruction of the rebel leaders. Italian
columns fanned out into the countryside. By December, all the remain-
ing resistance leaders had been captured: all were shot on Mussolini’s
orders save Ras Imru, whom Graziani, uncharacteristically, spared.
But Graziani’s brutal methods soon backfired. Resentment built to
the point that an attempt was made on his life. On February 19, 1937,
Graziani was seriously wounded by a hand-grenade fragment during
the assassination attempt. In fury and anger, the Italians ordered the
most brutal of reprisals: for three days the capital city became a hunting
ground, as soldiers, Black Shirts, and even common citizens vented
their fury and frustration by killing, raping, and burning at will (Del
Boca, 1969: 221—222).
Unable to find his would-be assassins, Graziani took the opportunity
to murder the entire Ethiopian intelligentsia, every member of the
Young Ethiopian Party, and all the officers and cadets of the Holeta
22
‘‘Not only was gas used throughout the war, but afterward as well to break down the
resistance of the Ethiopian patriots, as we know from a telegram dated September 11,
1936, sent by Graziani . . . ‘Today our air force will carry out reprisals, dropping various
asphyxiating gases between those who have submitted and those who have not . . . ’’ (Del
Boca, 1969: 82).
126
Italy in Ethiopia
[military] Academy (Del Boca, 1969: 222). Graziani’s methods were

heartily approved in Rome, but
the wish was nevertheless expressed that the executions be carried out
in the utmost secrecy with no witnesses present, an instruction that,
needless to say, was never followed. The apparent rebuke rankled
Graziani, who wired back in order to justify himself, ‘‘I cannot deny
that some Ethiopians have shouted as they faced the firing squad,
‘‘Long live Ethiopia!’’ I beg to state, however, that executions ordered
in consequence of the attempt on my life are invariably carried out in
isolated spots where no one, I repeat, no one can witness them.
(Del Boca, 1969: 222—223)
But of course they were witnessed. Worse still, on May 20, 1937,
Graziani ordered 297 monks of Debra Libanos, Ethiopia’s most famous
monastery, to be shot. The enormity of this atrocity had the opposite of
its intended effect: instead of cowing the population, it incensed them.23
Italian outposts came under fire, and convoys were ambushed and
destroyed. Whole regions, in particular Gojjam and Beghemder, rose
in revolt.
In addition, Graziani appeared to have become mentally unbalanced
by the attempt on his life and his failure to find the assassins. He
reportedly locked himself into his bedroom at night, and had an entire
battalion assigned to his personal security. By November, his eccentric
behavior and his counterproductive policies eventually prompted
Mussolini to replace him with the duke of Aosta — a man reputed to
be humane as well as intelligent.24
Interaction four: conventional attack vs. GWS

The duke inherited an impossible situation. His intention had been to
attempt to reverse the revolts by a combination of direct military action
against the most powerful opposition, combined with an amnesty and a
halt to summary executions. But the Ethiopians had ceased altogether
23
Mockler explains the consequences of the accumulation of brutalities: ‘‘Killings and
atrocities, however, no longer acted as a deterrent, for the [Ethiopians] had been driven to
desperation. The internal feuds and quarrels which had made so many of them accept, if
not welcome, an alternative to their Shoan rulers, were submerged in the face of a
common enemy. Their leaders had been treacherously killed. Their clergy had been
treacherously massacred. Like any race faced with extermination they instinctively
rebelled’’ (Mockler, 1984: 184).
24
It is not clear why Mussolini chose the duke of Aosta to replace Graziani (Mockler,
1984: 186). All agree that Graziani needed to be replaced, and that the duke would have
been the best candidate for the position, had not Graziani already ‘‘poisoned the well’’
with his brutal pacification policies.
127
to trust the Italians. To get that trust back was possible, but only after
years of hard work and consistent policy. Mussolini was hardly likely to
be patient in this regard, and the duke’s main military advisor, General
Cavallero, was soon complaining behind the duke’s back to Rome that
the duke’s lenience was counterproductive.
In March through April of 1938, the Italians launched a series of
major military campaigns intended to crush the rebellions in Gojjam
and Beghemder. The invasion of Gojjam was undertaken by 60,000
troops, and appeared to achieve all its aims. But no sooner had
Gojjam been declared pacified than new threats appeared.
Cavallero decided to try a different strategy. Instead of deploying
regular forces in overwhelming numbers as at Gojjam, he created three
flying columns or Gruppo Bande which he dedicated to the capture of
Abebe Aregai, one of the most exasperating of the resistance leaders,
who had been trying for months to relocate his forces in the much more
forbidding terrain of Menz.25 The fighting, which lasted from the end
of the rains in October to December of 1938, proved inconclusive.26
Aregai succeeded in reaching Menz and joining Dejaz Auraris, and
both survived repeated Italian efforts to destroy them. The Italians
withdrew.
If Graziani’s pacification policies sparked a unified rebellion against
Italian rule, the duke of Aosta’s more lenient policies and more con-
ventional assaults had not proven effective at solving the military
problem. At the close of 1938, Ethiopia was still in full revolt. Its
resistance leaders were still at large, and the sixty-five Italian battalions
in Ethiopia were forced to live inside their forts. In March of 1939 the
duke was recalled to Rome for ‘‘talks.’’ Meanwhile, Hitler invaded
Czechoslovakia, and Europe began to brace itself for war.
Interaction five: conciliation vs. GWS

In Ethiopia 1939 was a different sort of year. There was no repeat of the
previous year’s intensive military campaigns against intransigent rebel
leaders for two reasons. First, the costs of such operations in terms of
25
Menz was the heartland of the Shoan kingdom. It rose 10,000 feet above sea level, and
occupied 850 square miles of plateaux. Its governor was a respected Shoan noble named
Dejaz Auraris, who was another of the resistance leaders the Italians desperately sought to
capture or kill.
26
Part of the problem was that although Cavallero’s bande were more mobile and
independent, they lost proportionately in firepower. Much of the fighting which took
place from October to December, therefore, was virtually hand-to-hand. Although casual-
ties were high on both sides, it proved impossible to achieve decisive results in this way.
128
Italy in Ethiopia
money had become staggering. The previous year’s operations alone

had cost three times as much as the entire bill for the administration and
development of the colony, and the results had been inconclusive.
The second reason was a new focus on preparations to defend
Ethiopia in the event of a war breaking out between the European
powers. Mussolini did not intend to draw on support from Ethiopia
in the defense of Italy, but he made it clear that in the event of war
Ethiopia was on its own. For this reason, the Italians spent much of the
year and proportionately more of their resources organizing an army
for the defense of the AOI from an attack by British or French forces. As
a result, Italian efforts against rebellion in Ethiopia shifted emphasis
from coercion to bribery, amnesty, and negotiation. Ironically, now that
they had given up trying to win a COIN war against the Ethiopian
resistance, this combination of policies proved to be more effective than
anything the Italians had so far attempted.
As the year came to a close, Europe was at war, but although Poland
had fallen, the situation in Europe had stabilized. Italy remained neu-
tral, and as a result, tensions between Britain, France, and Italy in East
Africa eased.
World War II in East Africa

By 1940, the only significant resistance leader still holding out was
Abebe Aregai. The Italians spent the first part of the year attempting
to gain his trust by bribes, promises of amnesty, and promises of
independence after submission to Italian rule. These negotiations
ended in March, however, after the Italians were warned that the
planned submission of Aregai was in fact an ambush.
In April, Germany invaded Norway and in May, France. In June,
Italy declared war on a defeated France, prompting Britain’s Prime
Minister Winston Churchill to call Mussolini ‘‘the jackal of Europe.’’
Italy’s declaration of war instantly transformed the strategic map of
North and East Africa. In June of 1940, the AOI had 250,000 troops
supported by 200 aircraft, as compared to Britain’s 10,000 troops and 20
aircraft (Mockler, 1984: 208). A major concern for the British was being
sandwiched between Italy’s considerable armies in Libya and the AOI
in East Africa. But the Italians were not thinking about offense in the
Spring of 1940:
‘‘I saw the Duke of Aosta this morning,’’ [Italian Foreign Minister
Ciano] wrote in his Diary for April 6. ‘‘He tells me that for him it is
129
not only impossible to take the offensive but will also be extremely
difficult to hold his actual positions because the Anglo-French are now
well-equipped and ready for battle and the population, among whom
rebellion is still smouldering, will rise as soon as they have the sensa-
tion that we are in trouble.’’ (Mockler, 1984: 208)
The French in Djibouti were prepared to join with their British allies
in the defense of East Africa. But they were soon knocked out of the war,
and Djibouti’s commander was forced to become a passive bystander in
the Anglo-Italian fight.
That fight did not last long. Although the British had hoped to spark a
rebellion in AOI by the return of Haile Selassie 27 (supported by British-
led guerrillas along the lines of T. E. Lawrence’s adventures in Arabia),
the fight itself was soon reduced to a series of conventional strikes and
counterstrikes, with the British initially getting the worst of each
encounter. In August of 1940, the Italians even succeeded in routing
British forces in British Somaliland.
But Italian success in Somaliland emboldened them to try their luck
against Egypt from Libya. These forces, under the inept leadership of a
rehabilitated Graziani, advanced several miles into Egypt, but stopped
at Sidi Barrani. In December, the British Army of the Nile counter-
attacked, and the Italians were shattered: ‘‘At one stroke the threat to
Cairo and the hope of linking the two halves of Italy’s African Empire
were removed; and at one stroke Italian morale dissolved’’ (Mockler,
1984: 307).
The collapse of Italian morale was not confined to North Africa. The
impact on the AOI was devastating — all the more so because the
defeated Italian troops had been led by none other than Graziani:
But it was, naturally, among the Ethiopians that the defeat of
Graziani had the greatest effect. He had been their conqueror
and, as Viceroy, their tyrant. As the victorious Army of the Nile
pressed on and in its turn invaded enemy territory to capture first
Bardia and then Tobruk with fresh hordes of prisoners, George
27
Interestingly, Mockler argues that the British were embarrassed and concerned by
Haile Selassie’s return to the Middle East. Their early concern was that his presence
would provoke an attack on Sudan by Italy (Mockler, 1984: 226). After some thought,
however, it occurred to the British that after all, they were at war so then the issue became,
how best to use Selassie for their own ends. This was not clear either. Britain was
dramatically outgunned in Egypt, Sudan, Kenya, and British Somaliland: all places
where they expected Italian attacks. And attack the Italians did, though they never
pressed their advantage. It then occurred to the British that the way to take the pressure
off was to seriously foment rebellion within Ethiopia.
130
Italy in Ethiopia
Steer28 issued leaflet after leaflet with photographs of long lines

of Italian captives. (Mockler, 1984: 309)
In 1941, it was the revitalized British who took the initiative, and
Italian forces everywhere in East Africa collapsed. Haile Selassie was
returned to Ethiopia on 20 January, 1941 and, after a series of sharp
battles, British forces with Ethiopian support forced the Italians to
surrender. On May 5, 1941, five years to the day since Badoglio had
marched triumphant into Addis Ababa, the emperor once again took
his seat on the lion throne.
Outcome: Italy wins, or does it?

The Italo-Ethiopian War ended when Britain entered the war in North
Africa against Italy in 1940. At that point the Italians become obsessed
with the British and begin pursuing a war-ending strategy with the
Ethiopian resistance. The Italo-Ethiopian war is subsumed by World
War II. Had Britain not entered the war, Italy would have won in the
short term but lost in the long term. Italy did win interaction five, but
only with the promise of years of heavy bribes could it have maintained
its victory, and Italy’s economy could never have supported such a
policy for more than a year.
What emerges most clearly from this case is that Italy in 1935
possessed a supremely incompetent military. It was so ineptly led
that its effectiveness must go down in the annals of military history
as one of the poorest of the period. The Ethiopian emperor had worked
out an ideal defense strategy which took advantage of Ethiopian
strengths while minimizing its severe weaknesses, but his generals
would not listen. They insisted on opposing a conventional Italian
attack with a conventional Ethiopian defense, a defense which
included observing the peculiar (and in this case debilitating) custom
of ending all fighting at sunset. The Italian military’s incompetence,
especially evident in the early interactions of the war, forced a stark
decision: risk high losses and setbacks in a hard fight for Addis Ababa,
or cheat by using mustard gas against an adversary with no defense
against it and no capacity to respond in kind. The choice of mustard
gas facilitated Italy’s short-term military aims but — especially after
Selassie’s flight from the capital — undermined its political goal of
28
Steer had been responsible for all counter-AOI propaganda efforts, including a highly
successful Imperial Proclamation (or Awaj), in which the emperor had declared his
imminent return, and had called on all loyal Ethiopians to desert and resist the Italians.
131
establishing a viable self-sustaining colony. It stimulated rather than

reduced resistance, and it increased the costs of the subsequent bribes
necessary to keep Ethiopia’s princes from leading attacks against
Italian troops in AOI.
Analysis: competing explanations of the Italo-

Ethiopian War’s outcome
Actor interests
Italy’s interests in the conquest of Ethiopia had little to do with its
power relative to Ethiopia. It might be more accurate to argue that its
interests were explained by Italy’s power relative to other European
powers. Italians in general, and Mussolini in particular, clearly felt
that the other European powers had not accorded Italy the respect
due a European state, especially one which had been an ally during
the Great War. Italy’s interest in Ethiopia reduced to four elements.
First, a need to redress past humiliations by a decisive military victory
over the only Africans to successfully repel a colonizing European
power. Second, a need to confirm the ‘‘virility’’ of fascism as an ideology
by a hard-fought military victory: defeat or even significant delay could
undermine the legitimacy of Mussolini’s fascist regime and in this
sense, Italy’s survival was at stake. Third, a civilizing mission or duty,
which Italy sought to undertake to prove to the world its status as a true
light of civilization. Fourth, a powerful and growing sense of necessity
due to overpopulation and unemployment. The combination of these
four elements made Italian resolve to see the conflict through much
higher than that which could be predicted by a gap in the relative
material power of Italy and Ethiopia.
Ethiopia’s interests were survival, and, recognizing its weakness vis-
à-vis Italy, Ethiopia consciously sought the aid of other states in pursuit
of this interest. Foreign advisors found it difficult to understand why, in
the face of such overwhelming force, the emperor refused to back down
and accept Italian demands without a fight. But Ethiopian imagination
had been as fixated by the victory at Adowa as the Italians had been
about their defeat there. Ethiopia’s noble houses believed they could
repeat their victory again if the Italians chose to invade. Early on, the
emperor himself counted on two things. First, he counted on the inter-
vention of the League of Nations to prevent an all-out war. Second, he
hoped to fight a GWS in the event of war, which the League would
132
Italy in Ethiopia
almost certainly act to stop. Neither of these hopes were realized, but
Ethiopia never surrendered to Italy.29

In 1935, Italy had an authoritarian regime type. Italian foreign and
domestic policy were under the complete control of Benito Mussolini.
The effects of regime type on strategy are clear in this case. Because
Mussolini had complete control over the Italian press (and some control
over foreign press access to the battle space), the invasion of Ethiopia
presented strategic options which would have been more risky when
prosecuted by a democratic regime. Specifically, although Italy had
signed the Geneva Protocol prohibiting the use of chemical weapons in
war, it was clear that early on Mussolini planned to use such weapons
in Ethiopia, and, moreover, he was confident that he could get away
with it. How could this be, in 1935, when all of the European powers
now boasted mass-literate populations and extensive and sophisticated
press corps?
To get away with barbarism required managing three kinds of wit-
nesses and two audiences. The three sorts of witnesses were Italian
newsmen, foreign correspondents, and foreign aid workers. The two
audiences were Italians and other states. The fascists had a strategy for
each group.
Italian newsmen were carefully selected and briefed before going to
the Ethiopian theater:
‘‘Before we left Italy,’’ wrote Bruno Roghe . . . ‘‘we were received by the
Hon. Dino Alfieri, then Under-Secretary of State, who greeted us,
wished us luck, and gave us the exhortation that our press cards
warranted: we were to be lively and concise in our reports so as to
bring out in our style of writing the latest prerogatives of the Italian
undertaking, vibrant with youth, unshakable determination, and
single-mindedness of purpose.’’ In other words this was an invitation
to glorify the war, to exalt every incident into a shining example, to
give voice, in fact, to a paean of unqualified praise, unmarred by the
faintest shadow of doubt, the least trace of criticism.
(Del Boca, 1969: 62)
29
The fact that Ethiopia never formally surrendered to Italy had at least one crucial effect.
The laws of war prohibit the execution or mistreatment of captured soldiers, but provide
few if any such protections to ‘‘rebels.’’ Had Ethiopia officially surrendered to Italy, the
legitimacy of Graziani’s executions of rebels would have been debatable. As it was, they
were a clear violation of the laws of war as signed by Italy and the other European powers
at Geneva.
133
Foreign correspondents were often fascists in their own countries,

and those who weren’t often experienced ‘‘technical difficulties’’ filing
their reports.30 Others were intimidated, or forbidden access to key
areas or battles.
Finally, foreign aid workers were initially the target of intimidation
tactics. Italian planes were ordered to buzz Red Cross field hospitals
and to strafe near them in order to frighten the hospital administrators
into leaving. When this failed, the Italians took to directly attacking
such facilities, claiming their attacks were justified reprisals for
Ethiopian violations of the laws of war. When evidence did leak out,
and it did, the Italians had other strategies for dealing with such crises
(see also Sbacchi, 1985:63):
These were the Ethiopian charges [use of mustard gas, attack of Red
Cross units, attacks on civilians], but the Italians, anticipating them,
had already found a way to cloud them and confuse the issues. On
December 26, the Italian government had sent a letter to the League of
Nations charging the Ethiopians were using dumdum bullets. And on
January 16 [the Italians] sent the League a letter reiterating the dum-
dum charge and adding to it a charge that Ethiopian soldiers were
protecting themselves from bombardment by hiding under Red Cross
flags. Wasn’t that considered sufficient justification for bombing Red
Cross units? It was considered sufficient justification in Geneva, any-
way, for allowing the Ethiopian charges and Italian countercharges to
balance themselves out. (Coffey, 1974: 275—276)
There are two crucial points to be made here about the Italians getting
away with mass murder. First, they had a specific strategy in place for
preserving the secrecy of their barbarism, and that strategy presup-
posed an authoritarian regime type. Second, the use of mustard gas in
the particular context of Ethiopia in 1935, combined with a more vigorous
conventional attack strategy, ended the war in time to avoid interna-
tional intervention. As a result, Italy’s victory was nearly a fait accompli.
The question of Italy’s use of mustard gas was not formally raised in
Europe until March 30, 1936. At that time, it was agreed that it couldn’t
be discussed until a formal inquiry could be launched, because, as Lord
Halifax put it, it would be wrong ‘‘to prejudge a matter so grave and so
vitally affecting the honor of a great country [Italy] . . . The first step
30
As noted above, the Italian intelligence service specialized in interdicting unfavorable
evidence.
134
Italy in Ethiopia
must be to obtain the observations and comments of the Italian govern-

ment’’ (Coffey, 1974: 317).
When it became apparent to everyone that the Italians were not only
using gas, but using it liberally and against noncombatants, the ques-
tion again came up, this time at Geneva on April 8, 1936:
When a question was raised about Italy’s use of gas, most of the
delegates agreed that efforts should be made to collect reliable infor-
mation on the subject. But Pierre-Etienne Flandin, the French Foreign
Minister, objected. He asked if any inquiries had been made into
atrocities committed by Abyssinians, including the use of dumdum
bullets.
At a private meeting after the committee’s first morning session,
British Foreign Secretary Anthony Eden attacked the implications of
Flandin’s remarks. There was, Eden insisted, ‘‘a distinction between
the irresponsible atrocities of undisciplined military forces and the use
of poison gas which could not be other than a governmental act.’’
Flandin agreed. He said the Italians were ‘‘very stupid’’ to use this
form of warfare, but he doubted ‘‘the wisdom of issuing a formal
condemnation at a moment when an attempt was being made to
bring hostilities to an end, for this might disturb the negotiations.’’
The Committee of Thirteen decided at a meeting later that day not to
inquire into the Italian use of poison gas. (Coffey, 1974: 324—325)
By April 8, fully a week had passed since the last Ethiopian army in
the field had been smashed at Mai Chew. The war was already over.
In sum, Italy’s authoritarian regime type made it much simpler to
avoid the risk of public disaffection and international intervention
when Italy resorted to a barbarism strategy. Still, even authoritarian
Italy was not entirely able to hide its barbarism from the world or its
own people. The fact of nonintervention can best be explained by the
context of the day, in which the goal of European stability had come to
mean so much to Europe’s leaders that they were willing to sacrifice
every principle of morality and justice on its altar.
The impact of Italy’s regime type on its political vulnerability to
military setbacks is clear in at least one important respect. As already
observed above, Mussolini had control of all the Italian public’s access
to information about the course of the war and its conduct. Every letter,
photograph, and newsreel sent home was scrutinized and sanitized.
But among Italy’s political elites, the course of the war and the frequent
setbacks Italian forces suffered early on, should have, according to
Mack, resulted in political vulnerability due to trade-offs among elite
interests. They did not. Mussolini determined what resources would be
135
expended to conquer and develop Ethiopia and that ended the matter.
In fact, Mussolini himself makes the proposed link between authoritar-
ian regime type and political vulnerability clear:
On July 6, [1935], at Eboli, he had allowed himself to venture beyond
prudence in an address to four Black Shirt divisions which were about
to embark for Africa. Speaking from the back of a truck, he had
completely exonerated Italy’s soldiers of 1896 for their defeat at
Adowa, blaming instead the ‘‘abject’’ government in Rome at that
time. With his Fascist government in power, he assured the Black
Shirt troops, their efforts in the field would get full support at home.
(Coffey, 1974: 103)
There were trade-offs, of course, but these took the form of personnel
transfers: generals sacked or promoted. But although there were intri-
gues and arguments, the basic aim of Ethiopia’s conquest and coloniza-
tion was not questioned: not by the Italian people, and not by the fascist
political elite.
In sum, Italy’s authoritarian regime type made its resort to barbarism
more likely, and reduced the probability that it would be forced to
withdraw from the fight due to domestic political pressure or interstate
intervention.
Arms diffusion
In none of the five historical cases analyzed here was arms diffusion a
less plausible explanation for weak actor success than in the Italo-
Ethiopian War. As noted above, the Ethiopians had few rifles, and
fought for the most part barefoot. Paradoxically, this gave them greater
mobility in the tough mountainous passes leading from Eritrea to the
central plateau upon which Addis Ababa sat. For their own part, the
Italians and their Eritrean allies tended to move in motorized columns.
This gave them some advantages in terms of the artillery they could
carry but these advantages were on balance overwhelmed by the dis-
advantages of having to move and maintain communications along
poorly made roads and across innumerable ideal ambush sites.
The decisive technology proved to be airplanes. The Italians had
them, the Ethiopians didn’t. Once the Italians got organized, they
were able to use their air force to spot concentrations of Ethiopian
soldiers and thus degrade the Ethiopian ability to catch the Italians in
surprise ambushes. This was especially devastating because the
Ethiopians refused to fight at night, when the Italian air advantage
could have been nullified. When air-dropped mustard gas was added
136
Italy in Ethiopia
to the reconnaissance function of air power, the military contest quickly

became one-sided.
The Italo-Ethiopian war provides five independent tests of the strategic
interaction thesis. In this case, there were some surprises.
In the first interaction of the war, the strong actor should have won
quickly, because Italy chose a conventional attack strategy and Ethiopia
a conventional defense (same-approach). Yet not only did the Italian
offensive bog down even before coming to grips with the enemy, but an
Ethiopian counteroffensive threatened to drive the more numerous
Italians back into Eritrea. The explanation for this outcome is poor
Italian military leadership and an unwillingness to risk casualties.
These factors, in this interaction, overwhelmed the expected effects of
a same-approach interaction.
In the second interaction, the strong actor should have won, because
Italy continued its conventional attack strategy against an Ethiopian
conventional defense. Italy did win, but it won with the help of mustard
gas. In the Ethiopian theater, mustard gas proved to be a powerful force
multiplier. Although the indiscriminate nature of air-delivered mus-
tard gas makes its use barbarism under my definition, in this interaction
of the war Italy used gas to destroy Ethiopia’s military capacity, not its
will to resist. In this interaction, then, the use of gas does not indicate a
shift in strategy so much as tactics. The interaction remains same-
approach and the outcome was as expected.
In the third interaction, the strong actor should have won, because
Italy employed a barbarism strategy against an uncoordinated Ethiopian
GWS (same-approach). In this case, the use of gas was combined with the
summary executions of noncombatants, rape, and reprisal killings, all
intended to coerce the will of the Ethiopian resistance. It is difficult to
assess with any certainty the outcome of this interaction, because it was
so brief. As noted in Chapter 4, in a barbarism strategy there tends to be a
gap between when the strategy is first put into effect and when it begins
to yield military dividends. Although the Italians had been using gas
against Ethiopian troops (and, inadvertently, noncombatants) for some
time, the real initiation of a barbarism strategy dates from the accession of
Graziani to the position of viceroy in 1936, and especially after the attempt
on his life in February 1937. But, in November, Mussolini replaced
Graziani ‘‘just as he was getting started.’’ The accession of the duke of
Aosta to the position of viceroy caused a shift in strategy once again.
137
In the fourth interaction, the weak actor should have won, since Italy
pursued a conventional attack strategy against an Ethiopian GWS
(opposite-approach). And, as predicted, the Ethiopians not only sur-
vived Italian attempts to destroy them but began to contest many of the
areas formerly under Italian control. But, ironically, the very expense of
the effort to use large conventional units to chase Ethiopian guerrilla
leaders prompted yet another shift in strategy.
In the fifth interaction, the strong actor should have won, because
Italy switched from attempting to coerce Ethiopian resistance to trying
to bribe it (conciliation vs. GWS: same-approach). As predicted, Italy
did win, or might have won, had not World War II intervened. The
accession of the duke of Aosta to viceroy had prompted a major thaw in
British—Italian relations. The duke liked the British, and the British
admired and respected the duke (Mockler, 1984: 187—188, 194). Yet
Mussolini’s decision to enter the war against France and Britain
instantly changed all that.
In sum, by the logic of the strategic interaction thesis, Italy should
have won every interaction of its contest in Ethiopia save one (interac-
tion four: conventional attack vs. GWS). Yet it lost interaction one
(conventional attack vs. conventional defense). The reason it lost
when it should have won is poor leadership. The five interactions of
the Italo-Ethiopian War, the nature of the strategic interaction in each
interaction, and outcomes are summarized in Table 1.
The problem of leadership

The effects of strategic interaction were overwhelmed in the war’s first
interaction by the incompetent Italian generalship. Italy’s generals in
Africa were to become as famous for their incompetence as for their
brutality. Many Italian officers were brave and many were competent,
and none of Italy’s enemies have faulted the bravery or skill of the
individual Italian soldier. But the generals — De Bono, Badoglio,
Graziani, Cavallero, Trezzani31 — have been condemned to live in mili-
tary infamy. Mockler offers the following curious explanation:
31
Trezzani took over from Cavallero on April 20, 1940. Mockler characterizes him this
way: ‘‘He was one of the few Italian senior officers who had never served in a colonial
campaign, or indeed in the colonies at all. His career had been that of a professor at the
School of War. His interests were in tactics, his approach one of detached calculation. He
was Badoglio’s man — imbued with skepticism before he ever set foot in Africa . . . If the
Italians had deliberately decided to choose a general whose prime characteristic was his
capacity for lowering the morale of the troops he was to command, their choice could
hardly have fallen on a better man’’ (Mockler, 1984: 209).
138
Table 1. Strategic interaction and outcomes: the Italo-Ethiopian War, 1935—1940
strong actor strategy weak actor strategy strategic interaction actual outcome
Phase 1 conventional attack conventional defense same-approach Ethiopia wins

2 conventional attack/barbarism conventional defense same-approach Italy wins
3 barbarism GWS same-approach Italy wins
4 conventional attack GWS opposite-approach Ethiopia wins
5 [conciliation] GWS [same-approach] [Italy wins]
It seems to me that, consciously or subconsciously, the memory of

Adowa and of General Baratieri’s end must have been preying on the
minds of all Italian generals in Africa. Only this can explain their
reluctance ever to move forward till they had overwhelming super-
iority totally assured and lines of communication totally safe. Only this
can explain, and perhaps excuse, their inability, even then, to move
forward at more than a snail’s pace. So many missed opportunities
cannot but have been pathological. They were certainly disastrous.
(Mockler, 1984: 254)
The true test of this thesis would be to assess Italian military leadership
in other theaters of war — say Greece or Russia. Were Italian generals
more aggressive or more generally competent? They were neither.
One is tempted to argue that Italy’s top generals were merely
Mussolini’s fascist cronies with epaulettes, and the incompetence
which resulted was therefore the inevitable concomitant of an authori-
tarian regime. But this is unsatisfactory. The French military had its
share of cronyism and incompetence and it was a democracy.
Moreover, fascist Germany fielded the most effective military leader-
ship the world had ever seen.
Conclusion
The Italo-Ethiopian War was an asymmetric conflict between fascist
Italy and imperial Ethiopia. Italy won.
Italian interests reduced to revenge for the defeat at Adowa in 1896, a
need to legitimize fascism as an ideology, responsibility for ‘‘civilizing’’
a barbarous black African state in the Horn of Africa, and resources
with which to ameliorate its poverty and population problems.
Ethiopian interests reduced to survival and honor.
Relative material power did not explain relative interests in the case
of the strong actor, Italy. Relative material power may explain
Ethiopia’s interests, however. How, if at all, did regime type explain
the strategies or political vulnerability of the strong actor? Hypothesis 8 —
relative material power explains relative interests in the outcome of an
asymmetric conflict — is not supported in the Italo-Ethiopian War.
As we have seen above, Italy’s strategic choice was affected by its
regime type. Mussolini was always concerned about the possibility of
foreign intervention — economic or military — and he counted on being
able to manage information about his intentions and conduct in the war
in order to forestall such intervention. He understood the risks involved
140
Italy in Ethiopia
in his decision to make mustard gas — a prohibited weapon of war —

available to his commanders in Ethiopia. Italy’s resort to barbarism in
fact presupposed the total management and control of information on
the conduct of the campaign.
In terms of political vulnerability, Italians, regardless of political
affiliation or class, all shared the dream of having a thriving colony in
Africa.32 Mussolini demanded it as Italy’s right. Italy’s preponderance
in material power should have made it politically vulnerable to military
setbacks — if not by the weight of public opinion then at least as a
consequence of trade-offs among fascist political elites. But although
Italy experienced stunning setbacks during its Ethiopian campaign,
and although evidence of Italy’s use of mustard gas did leak out, this
didn’t happen. Regime type therefore matters more than allowed by the
interest asymmetry argument. But if nature-of-actor and interest asym-
metry arguments cannot explain the outcome, can arms diffusion or
strategic interaction?
Arms diffusion can’t. Bluntly, the Ethiopians did not receive arms of
any quantity or sophistication until the British entered the war in 1940.
For this reason, hypothesis 6 — the better armed a weak actor is, the
more likely it is that a strong actor will lose — is not tested here. Prior to
the war, Ethiopia’s emperor had been counting on diplomatic support
and even military intervention from his fellow League members in
Europe. Ethiopia was too poor and too proud to import arms as the
Boer had done prior to their fight with the British.
Strategic interaction best explains the outcome of the war, though
even strategic interaction is not proof against poor Italian leadership.
With one exception, each strategic interaction of the war contained an
interaction which favored an Italian military victory. The first interac-
tion of the war featured a conventional attack vs. a conventional
defense, and Italy appeared poised to lose this fight due to bad general-
ship. The stalling of the Italian advance, combined with effective
Ethiopian counterattacks, led in the second interaction of the war to a
shift in strategy. Italy employed a conventional attack strategy aided by
a super-weapon against an Ethiopian conventional defense. This
32
For a contrary view, see Sbacchi, 1997. Sbacchi convincingly recounts a broad Italian
opposition to Mussolini’s planned ‘‘adventure’’ in Ethiopia, but fails to provide a sense of
how widespread such opposition in fact was. In addition, Sbacchi’s account suggests that
once Italians became convinced of military success in the initial months of the war, Italian
public opinion shifted to support the invasion unconditionally.
141
strategy proved highly effective militarily, and Ethiopia’s organized

resistance collapsed within a few months. The Italians won.
But the very methods used in Ethiopia’s conquest — and especially the
post-war policy of summary executions and continued use of mustard
gas on ‘‘rebels’’ — sparked renewed resistance. The Italians might have
won this third interaction of the conflict, but before it could register a
military impact Mussolini cancelled it.
The fourth interaction of the war reverted to a conventional attack
against any organized resistance. Only where Ethiopian forces sought
to engage the Italians directly, as at Gojjam, were the Italians successful.
In the remainder of the country the Italian conventional attacks were
countered by a GWS, and the Italians lost. The costs of such operations
were prohibitive, and the combination of those costs and the perceived
need to become security self-sufficient in the event of war in Europe, led
to a final shift in Italian strategy.
This fifth interaction of the war introduced a conciliation strategy for
Italy, opposed by an Ethiopian GWS. Conciliation proved the most
effective of all the strategies the Italians had attempted since marching
into Addis Ababa in May of 1936. Only one rebel leader held out, and it
seemed likely that but for the outbreak of World War II even Abebe
Aregai would have been forced to flee Ethiopia or submit. The Italians
would have almost certainly won the war in the short term.
But World War II changed all of this. The contest in Ethiopia ceased to
be between occupying Italians and an Ethiopian resistance, and became
a fight between the British and Italians fought in Ethiopia, Sudan,
British Somaliland, and Kenya. At first, the Italians launched conven-
tional attacks against British defenses, and they succeeded almost
everywhere. However, they failed to exploit their victories, and after
the collapse of Italy’s armies in Libya and Egypt, the British launched a
series of conventional attacks against Italian defenses. The British had
hoped to coordinate these attacks with a British-sponsored ‘‘general
uprising’’ in Ethiopia. Yet although they had devoted enormous
resources to finding and funding a ‘‘Lawrence’’ for Ethiopia — Orde
Wingate — Wingate’s and Haile Selassie’s triumphant re-entry into
Ethiopia were superfluous to the outcome. The Italians fought hard in
the north, but again suffered from poor leadership. While, in the south,
a British probe toward the port city of Kassala in Italian Somaliland
resulted in a complete rout of vastly superior Italian forces. Within one
month, British forces invading from the south had rolled up the Italians
and entered Addis Ababa.
142
Italy in Ethiopia
On balance then, in four of five interactions, hypothesis 5 — strong

actors are more likely to win same-approach interactions and lose
opposite-approach interactions — was supported.
In sum, if we wish to explain why Italy lost the Italo-Ethiopian war,
we cannot rely on relative power, regime type, interest asymmetry, or
arms diffusion. With the exception of arms diffusion, these factors
clearly mattered. But what mattered most was strategic interaction
and Italian military leadership. Once it became clear that the capture
of Addis Ababa would not force the Ethiopians to surrender, the
Italians had three strategic choices. First, continue with the present
strategy, perhaps sending more troops, and equipment to the theater.
Second, offer the Ethiopians terms: something which could save face
yet end the war. Third, switch to a barbarism strategy: take the gloves
off and go after Ethiopian patriots and their families, either by holding
them hostage or by killing them outright in reprisal for continued
resistance. In the event, the Italians chose the third option. From then
on, the Ethiopians maintained a GWS. But the Italians shifted and
innovated a number of times, painstakingly improving their advantage
until World War II came to Africa and swept them away.
143
6 The United States in Vietnam:
the Vietnam War, 1965–1973
Everything rotted and corroded quickly over there: bodies, boot

leather, canvas, metal, morals. Scorched by the sun, wracked by the
wind and rain of the monsoon, fighting in alien swamps and jungles,
our humanity rubbed off of us as the protective bluing rubbed off the
barrels of our rifles. We were fighting in the cruelest kind of conflict, a
people’s war. It was no orderly campaign, as in Europe, but a war for
survival waged in a wilderness without rules or laws; a war in which
each soldier fought for his own life and the lives of the men beside him,
not caring who he killed in that personal cause or how many or in what
manner and feeling only contempt for those who sought to impose on
his savage struggle the mincing distinctions of civilized warfare — that
code of battlefield ethics that attempted to humanize an essentially
inhuman war. Caputo (1996)
US military intervention in Vietnam began soon after the French

defeat at Dien Bien Phu in 1954. The full commitment of US combat
troops, however, did not occur until eleven years later. When US
Marines came ashore at Da Nang in 1965 the US population was
about 194 million, while North Vietnam’s was approximately 19 mil-
lion.1 US and North Vietnamese armed forces totaled approximately 2.5
million and 256,000, respectively. Adding in allies who contributed
combat troops (negligible), multiplying population and armed forces,
and dividing the strong actor’s total by two results in a relative force
ratio of about 53:1 in favor of the United States. Even allowing for the
fact that the United States did not actually devote half its ‘‘power’’ to the
1
These figures are taken from Singer and Small, Correlates of War, war number 163.
144
The United States in Vietnam
conflict,2 this was an asymmetric conflict and the United States and its
allies were the strong actor.3
The Vietnamese theater

The place we now call Vietnam has been known by many names:
Cochinchina, Annam, and Indochina to name only the most prominent.
Some of these place names corresponded to countries with borders
similar to those of contemporary Vietnam and some included portions
of what are now China, Laos, and Cambodia. The rich history of this
land and its people — and in particular the strong relevance of that
history to the events that unfolded in Indochina following the end of
World War II — cannot be adequately captured here, but three key
aspects of that history bear emphasis in this introduction.
First, the terrain and climate of Vietnam has always made it a forbid-
ding place in which to conduct warfare of any type, but especially the
kind of large-scale conventional warfare for which the United States
was best prepared in 1965. Vietnam is both suffocatingly small and
intolerably porous; open to the west along an impossible-to-seal border
with Laos and Cambodia, and open to the east to the Gulf of Tonkin and
South China Sea. Vietnam’s economic and cultural life are dominated
by its two river deltas. To the south, Vietnam’s geography is dominated
by the Mekong River delta. There are vast marshes, impenetrable
jungles, mangrove swamps, countless rivers, and at times a sticky,
sweltering heat such as few can imagine without actually having
experienced it. In the Central Highlands there are thickly rain-forested
hills and valleys; and to the north, still more mountains, covered with
rain and monsoon forest, and cut by deep valleys whose rivers drain
into the second of the country’s two major river delta systems — the Red
River delta.
Second, Vietnam’s social, political, and military institutions were
shaped by its terrain and climate, as well as by its predominant form of
agriculture: wet rice farming. Wet rice farming demands cooperative
labor, and as a result the Vietnamese developed a strong collective spirit
and a militia tradition of defense against invasion (Karnow, 1983: 98—99).
2
The logic of the relative power estimate does not require calculating how much of a
given actor’s power resources were applied to the fight. What matters is the resources an
actor could have applied to win the war. On this point, see Schelling (1966: 142—143, 172).
3
Even if we assume, for the sake of simplicity, that the contest was between only the
United States and North Vietnam, the ratio shifts to 49:1 in favor of the United States.
145
Third, over the centuries Vietnam has been invaded often. Important
ethnic divisions between North and South Vietnamese were most often
shelved during foreign invasions. The Vietnamese people were there-
fore effective ‘‘nationalists’’ hundreds if not a thousand years before the
rise of nationalism in Europe.4 Frequent invasions resulted in both a
loathing of would-be foreign conquerors (see e.g., Karnow, 1983: 58),
and a strong military tradition. That military tradition often favored an
indirect defense, utilizing the advantages of terrain and climate to wear
down more powerful invaders rather than meeting them head-on
in pitched battles. Here Stanley Karnow recounts the strategy of a
thirteenth-century Vietnamese general defending against a Mongol
invasion:
The Vietnamese, commanded by the illustrious Tran Hung Dao,
repulsed each offensive. Like outnumbered Vietnamese officers before
and since, he relied on mobile methods of warfare, abandoning the
cities, avoiding frontal attacks, and harassing his enemies until, con-
fused and exhausted, they were ripe for final attack.
(Karnow, 1983: 101)
Climate and terrain abetted this strategy, but in no way mandated its
adoption. In its nineteenth-century struggles against France, for example,
Vietnamese emperors and their generals were more often brought into
direct engagements with their French attackers and in those fights
invariably lost.5
But defeating the Vietnamese in battle never proved decisive. Again,
this was partly because the Europeans had military technology specia-
lized for winning major battles, not for maintaining control of hostile
populations in difficult terrains and climates. By 1893 the French, after
much difficulty and expense, had consolidated their control of what
they came to call French Indochina: Tonkin (North Vietnam), Annam
(South Vietnam), Cambodia, Laos, and Cochinchina. World War I
soon followed, and Asian intellectuals, including Ho Chi Minh,
4
Vietnamese nationalism was not monolithic, save during foreign invasions. During the
Vietnam War, South Vietnamese communists had a national identity separate from that of
Northerners. What united them strongly during the war was the presence of foreigners on
Vietnamese soil (see Connor, 1969: 51—86; Karnow, 1983: 462, 534; and Herring, 1986: 271).
5
This happened in large measure because in earlier fights between the Northern Trinh
and the Southern Nguyen, France sold ‘‘modern’’ weapons to both sides. Modern weap-
ons (breech-loading rifles, e.g.) rewarded different tactics and strategies, and encouraged
the Vietnamese to confront their adversaries head-on. This proved disastrous, as even
small French forces could always defeat larger formations of similarly armed and trained
indigenous soldiers.
146
applauded US President Woodrow Wilson’s passionate advocacy at

Versailles of national self-determination as an antidote to future wars.
Ho and others in Asia eagerly awaited the end of European colonialism.
But when the victors — Britain and France — refused to dismantle their
Asian empires, secret societies and revolutionary groups began to form
and agitate. In 1930 the French brutally suppressed a nationalist upris-
ing in Yen Bay, just north of Saigon. But, unlike elsewhere in Asia,
French repression in Vietnam only succeeded in uniting their adver-
saries (Karnow, 1983: 58), giving them stronger incentives to resist
French rule.
Like Mao in China, Ho recognized the opportunity provided by the
Japanese conquest of Indochina. He reorganized the Indochinese
Communist Party in order to capitalize on Vietnamese nationalism
and to forge a broad alliance capable of seriously hurting the Japanese
in Vietnam. As a result, many of these underground groups became
affiliated or incorporated into the Indochinese Communist Party. These
groups sharpened their insurgency skills against the Japanese occupa-
tion, and, during this time, came to view the United States as a natural
ally. By late 1944 Vietnamese resistance led by Ho and his allies had
become troublesome enough that Japan decided to ‘‘grant’’ Vietnam’s
independence under a puppet ruler, the Vietnamese Emperor Bao Dai.
In March of 1945 Bao Dai declared Vietnam’s independence. But his
rule was short lived. In August, assuming Ho’s claims to leadership of
Vietnam had Allied support, Bao Dai abdicated and handed control of
Vietnam to Ho Chi Minh, who proclaimed the Democratic Republic of
Vietnam (DRV) on September 2, 1945.
The First Indochina War began in 1946, following the return of the
French to Vietnam, Laos, and Cambodia. The tale of France’s failed
attempt to re-conquer and pacify Vietnam, the intrigues of the 1954
Geneva Conference on the fate of Indochina, and the decade of political
and economic blunders (on both sides) that followed the Geneva
Conference make interesting reading, but there is no space for a
detailed review here. There are important points to take away from
each experience, however.
First, by 1945, Vietnamese resistance to French rule had become truly
nationalist and truly widespread. The French could not recognize that it
was already too late to win enough ‘‘hearts and minds’’ to make reoccu-
pation of Indochina a rational pursuit. Japan’s invasion had forged a
national resistance, and that resistance could not be as easily broken as
before the war. Contrary to US assessments at the time, French military
147
forces were on the whole well trained, well equipped, and skillfully led.
But they were never sufficient in numbers to quell Vietnamese resis-
tance to French rule. The French struggled in vain, and their efforts to
defeat the Viet Minh eerily prefigure later US struggles to accomplish
the even more limited goal of securing South Vietnam.6
Second, the DRV was soundly betrayed by China’s Zhou Enlai at the
negotiating table in Geneva. Far from the unified Vietnam the DRV felt
it had earned on the battlefield following its victory over the French at
Dien Bien Phu, backroom negotiations between China and France
forced the DRV to accept both partition and a delayed election to
determine Vietnam’s ultimate fate. Further negotiations — the Soviet
Union’s Molotov intervened — forced a division at the 17th parallel, and
set the date for elections in the summer of 1956.
The eve of war: North Vietnamese interests

From 1945 until the end of the second Indochina (Vietnam) war in 1975,
the DRV’s interests remained the same: drive all foreigners from
Vietnam, unifying the country under Vietnamese rule. Since the DRV
considered those working with and for the United States to be effect-
ively ‘‘foreign,’’ this also meant the ‘‘liquidation’’ — as communist doc-
trine put it — of South Vietnam’s government and its military.7
By 1959 the DRV had shifted emphasis from political agitation to
preparation for war. It sent 4000 troops to the South and began formal
construction of what would later be called the Ho Chi Minh trail: a
serpentine supply route linking the North to the South through Laos
and Cambodia. It also began smuggling arms to the South by sea. In
1960, the DRV established the National Liberation Front to coordinate
the war in the South. By 1961, Viet Cong (VC) forces had begun escalat-
ing attacks against Diem’s government in the South. Assassinations of
village leaders and provincial officials had risen from 1200 a year in
1959 to 4000, and had provoked Diem into replacing them with military
men, loyal to Diem but ignorant of the communities they were tasked to
6
French agrovilles, for example, later became South Vietnamese/US ‘‘strategic hamlets’’;
and French jaunissement later became US ‘‘Vietnamization’’ under Nixon. In 1950, the
French had to struggle with a skillfully led, perfectly dedicated foe supplied and sup-
ported by China. By 1965, US and South Vietnamese forces had to deal with the same
problem: a line of supply that extended from China through North Vietnam to South
Vietnam via the Ho Chi Minh trail.
7
Frederick Brown makes the point that to the DRV the continued existence of a South
Vietnam under foreign control was tantamount to a threat to its survival (see Brown
1980: 526).
148
serve and protect. In 1963 the DRV and VC began more aggressive and
larger-scale actions against South Vietnam (henceforth the government
of Vietnam or GVN). In January, a well-planned assault on VC forces at
Ap Bac went horribly awry, as GVN soldiers effectively refused to fight
and were severely mauled by the VC.
On the eve of Indochina’s second war, the DRV maintained its pas-
sionate — even maniacal — interest in unifying Vietnam under
Vietnamese rule. In pursuit of this objective, the DRV possessed two
distinct armed forces: a conventional military (the North Vietnamese
Army or NVA) and a nationalist guerrilla military (the VC).
The eve of war: South Vietnamese government interests

As early as 1955 the GVN was well on the way to becoming synony-
mous with the person of Ngo Dinh Diem. If the fact that the DRV was
ruled by the iron will of one man — Ho Chi Minh — is to be taken as a
criticism, then that criticism cannot be spared the GVN under Diem.
The extent to which this was true makes it much simpler to explain
why Vietnam was ultimately unified under the DRV and not the GVN;
Diem, and the nature of his leadership remain controversial subjects to
this day. A few scholars of the period claim that the picture of Diem as an
eccentric autocrat is an unfair caricature promulgated by disenchanted
US journalists, while most characterize him as a true Vietnamese nationa-
list, yet a man increasingly convinced of his own indispensability in
securing a non-communist future for Vietnam. As a result, Diem’s prio-
rities on the eve of war were in his survival as Vietnam’s sole ruler, not,
as so fervently hoped by his ally the United States, the survival of South
Vietnam as a viable, secure, and independent noncommunist state
(the two goals often diverged, and Diem appeared unable to appreciate
the difference). Diem’s interest in securing his personal control over
Vietnam (first South, then North) led to a steady restructuring of the
GVN and its military. Ministers and officers were increasingly advanced
based on loyalty to him or on a known grudge against a rival, rather than
on merit or combat experience. In the decade stretching from the Geneva
accords to the eve of the second Indochina war, this had the effect of
reducing the GVN’s military capabilities to the point where it was cap-
able only of repressing domestic opposition and securing Diem against
the machinations of his rivals in the South. This was made starkly clear to
the United States at the debacle at Ap Bac in 1963.
Yet the success of Diem’s efforts to secure his rule paradoxically
supplied its greatest threat. Under a real and ever-escalating assault
149
from the DRV, Diem’s often brutal suppression of dissent visibly wea-
kened South Vietnam’s defenses. By 1963, even loyal officers with little
interest in risking their lives in ‘‘adventures’’ against the VC in the
Mekong delta were aware that allowing Diem to continue might
mean South Vietnam’s — and by extension, their own — destruction at
the hands of the DRV.
Diem’s restructuring of the GVN had taken the better part of a
decade, and by the early 1960s US diplomats and military advisors
were well aware that it might take a decade of hard work to reconstruct
a GVN capable of capturing broad public support — a necessary pre-
condition for fighting and winning a major war against a dedicated,
well-supplied, and increasingly skilled foe. This is why the decision to
oust Diem was so controversial at the time. If, as some in the Kennedy
administration believed, Diem had to go (Diem’s repressive measures
had become the focus of increasingly irrepressible South Vietnamese
public outrage), then who would replace him and what positive differ-
ence could his replacement make?
Diem was murdered on November 1, 1963 and replaced by General
Nguyen Khanh. Instead of beginning the difficult work of repairing the
GVN and building a coalition of social and economic forces capable of
an effective defense against the VC, however, Khanh’s first priority was
securing his own rule. Corruption and incompetence continued,
prompting the United States, now under the leadership of Lyndon
Baines Johnson, to further escalate its support. That support would
begin with the president’s approval of covert sabotage and espionage
operations against the DRV, and eventually include the commitment of
US combat forces in March of 1965.
The eve of war: US interests

US interests in the outcome of the war between the DRV and the GVN
remain a source of contention to this day. In the simplest terms, how-
ever, there is broad agreement about three key aspects.
First, US interests in Vietnam were colored by the personal political
interests of one man, President Lyndon Johnson. Johnson’s true ambi-
tion as president was the passage of a package of social programs he
dubbed ‘‘The Great Society.’’ Johnson had become president in 1963
after John F. Kennedy’s assassination in Dallas, and had immediately
set about working toward his election as president in November of
1964. Although Johnson was aware that prior to Diem’s assassination
President Kennedy had planned to withdraw some 16,000 US military
150
advisors from Vietnam, the assassination of Diem put those plans on

hold while the president’s national security team reevaluated the costs
and benefits of a withdrawal from Vietnam. Moreover, Johnson was
well aware that his chief rival, the Republican nominee, Senator Barry
Goldwater, would attack him as a ‘‘liberal’’ who was soft on defense. For
this reason alone Johnson continued to support covert military action
against the DRV, including secret sabotage and espionage operations
against DRV military facilities.8 His immediate political objective was
to provide enough support to stabilize Vietnam and defend against
Goldwater’s accusation that he was soft on communism. Once elected,
he planned to reassess Vietnam: perhaps escalating a US commitment
or perhaps engineering a withdrawal.
Second, the United States was never interested in seeking the direct
overthrow of the DRV. It remained content with a Vietnam divided
across a DMZ just as Korea had been divided and for the same reason:
it feared China’s entry into the war and a possible escalation to, as
Johnson put it, ‘‘World War III.’’ US interests were in this sense, and
only this sense, accurately described as ‘‘limited.’’ In 1965 the United
States viewed the survival of the GVN as worth fighting and dying for,
under the condition that (a) it wouldn’t be too costly a war, and (b) in
support of the theory that the fall of the GVN would very likely lead to
the collapse of other noncommunist regimes in Southeast Asia, perhaps
overwhelming Japan and India and ultimately threatening more vital
US interests in Europe and the Western Hemisphere.9
Third, the deployment of US combat forces to Da Nang in 1965 was
viewed by the president as crossing a threshold, but not to the degree
that with hindsight seems so obvious today. As would happen so many
8
It was these operations, an accidental coincidence of an Op-Plan 34-A sabotage attack
against the DRV, and a top-secret DeSoto electronic espionage mission being undertaken
by the USS Maddox, that led to the first Tonkin Gulf incident of August 2, 1964, in which
three DRV patrol boats attempted to attack the Maddox with torpedoes, but were driven off
with minor damage. A second attack was alleged, and at the time genuinely believed, by
the United States on August 4. Subsequent analysis and extensive research has revealed
that this second attack, which prompted Johnson’s military retaliation and the Tonkin
Gulf Resolution (August 7, 1964), never actually happened. On this last point see
Moisie (1996).
9
After 1968 it became increasingly clear that the survival of the GVN was not worth the
cost of securing it, but by then the United States had another rationale for staying —
prestige and precedent setting. The United States had said the GVN would stand, and
even those in the administration now long convinced of the hollowness of the domino
argument could agree that a US failure in South Vietnam might endanger vital US national
interests elsewhere or in the future.
151
times later, Johnson and his advisers were attempting to use US military
forces to send careful and clear messages to the North — ‘‘we don’t seek
to overthrow you but we refuse to permit your support of ‘aggression’
against the GVN’’ — never fully comprehending how badly these ‘‘mes-
sages’’ would be translated. These force messages were invariably
thought of as necessary and limited escalations which were likely to
yield quick and positive political results. When they did not, the presi-
dent and his team tended to respond — not by questioning the process of
communicating by the use of military strikes — but by sending yet
another ‘‘clarifying’’ message.
Overall, the United States went to war in Vietnam to stabilize an ally
and prevent what it then believed to be the first step in a Soviet-
inspired, Chinese-supported, communist takeover of Southeast Asia.
In 1965, the Johnson administration believed that US military power
and technology could make it possible to secure the GVN without a
major escalation or a declaration of war. With the power to punish the
DRV from the air and destroy the VC on the ground, most in the
administration believed the only question unresolved as the Marines
waded ashore at Da Nang was, ‘‘how much punishment would they
take before calling it quits in the South?’’ (Mueller, 1980: 500).
The Vietnam War

The Vietnam War is one of the last century’s most complex military and
political engagements. US military intervention, for example, involved
four overlapping strategic interactions: (1) barbarism (ROLLING
THUNDER) against a conventional defense; (2) a conventional attack
against a conventional defense (the main-force units war); (3) conven-
tional attacks against a GWS (the guerrilla war in the South I); and (4)
barbarism against GWS (the guerrilla war in the South II).
Interaction one: ROLLING THUNDER, 1965–68

The first strategic interaction of the war began in March 1965 with a US
strategic bombing campaign, later named ROLLING THUNDER.10 Its
main goal was to destroy the willingness of North Vietnam to support
10
On the origins of the ROLLING THUNDER campaign, its purpose, and collateral
damage to noncombatants, see Karnow (1983: 397, 458). On relative US moral restraint
in ROLLING THUNDER, Karnow (1983: 653). For a dedicated analysis of ROLLING
THUNDER and its failure, see Thompson (1980); and for a theoretical treatment that
anticipates the strategic interaction thesis, see Pape (1996: ch. 6). Pape’s overall argument
152
the guerrilla war campaign against South Vietnam, and as its name
implied, the campaign was expected to take time:
Instead of a coordinated air campaign . . . which would destroy the
enemy’s ability to wage war and break their will to resist, air opera-
tions over the North were designed as a diplomatic ‘‘slow squeeze’’
signaling device. As Secretary of Defense Robert S. McNamara said on
February 3, 1966, ‘‘US objectives are not to destroy or to overthrow the
Communist government of North Vietnam. They are limited to the
destruction of the insurrection and aggression directed by North
Vietnam against the political institutions of South Vietnam.’’
(Summers, 1995: 96)
The United States sought to inflict enough pain on North Vietnam to
compel it to stop supporting the VC in the South. North Vietnam’s
defense against ROLLING THUNDER was conventional: it sought to
thwart US military attacks on its infrastructure and forces by means of
fighter aircraft and an increasingly dense radar and surface-to-air mis-
sile defense network supplied to them by the Soviets.
US Air Force generals and their civilian leaders shared a theory about
the general effectiveness of strategic bombing as a strategy. Strategic
bombing should have both hampered North Vietnam’s war effort
(interdiction) and coerced its leadership into giving up (pain). When
neither expectation was met, military and civilian elites faced a stark
choice: either reject the theory, or blame failure on some flaw in imple-
mentation. The Air Force chose to emphasize flaws, while the Johnson
administration was increasingly split: some agreed that the United
States was hitting the wrong targets, or not hitting the right targets
hard enough, whereas others — including eventually Defense Secretary
McNamara — concluded that strategic bombing against the DRV could
not work. McNamara’s reports indicated that the military value of
ROLLING THUNDER’s destroyed targets was zero.11 Bombing that
accepted collateral damage subsequent to this recognition was there-
fore barbarism: the deaths and injury to individual noncombatants
is that when air power is used to target an adversary’s armed forces, it generally wins, and
when used to target an adversary’s values (including infrastructure), it generally fails.
Insofar as punishment and denial represent indirect and direct strategic approaches,
respectively, Pape’s argument, as Mearsheimer’s (1983) before it and mine after, is an
argument for the independent causal impact of strategy on conflict outcomes, (see also
May, 2000). May explains French defeat in the Battle of France in 1940 as a function of
strategic interaction: the Germans simply had a better strategy.
11
Its political utility was less than zero: it increased international and domestic opposi-
tion to the US war effort, and although the North Vietnamese feared and hated the
bombing, they never considered altering their war aims as a result of the pain it inflicted.
153
and their property were not specifically intended, but they were
collectively deliberate and systematic.
ROLLING THUNDER continued until a week before the November
1968 US presidential election. It was an interaction in which a strong
actor (the United States) employed an indirect strategy against a weak
actor (North Vietnam) using a direct strategy, and lost.12
Interaction two: the main-force units war, 1965–1969

The main-force units war featured a series of pitched battles between
regular units of North Vietnam and of the United States and the South
Vietnamese Army (ARVN). In a series of engagements that lasted
throughout the war, US forces proved overwhelmingly successful at
destroying NVA and VC main-force units.
Examples of this interaction, which opposed a strong actor conven-
tional strategy with a weak actor conventional strategy, include Operation
STARLITE (August 1965), the Battle of Ia Drang (October—November
1965), MASHER/WHITE WING (January—March 1966), and Phase II of
Operation ATTLEBORO (October—November 1966).13 As the ground
war continued into 1968, the frequency of these interactions dwindled,
and US forces in the South focused increasingly on the problem of
counterinsurgency.
Part of the decreasing frequency of this interaction reflected
actor learning: North Vietnam’s leadership analyzed main-force unit
engagements carefully, and concluded that US forces were so adept at
combining maneuver and firepower that, unless the encounter took
place between dramatically mismatched forces (as at landing zone
Albany in November 1965), NVA and VC main-force units would
invariably be destroyed. In the context of Mao’s warning above, native
forces were fighting with inferior weapons on US terms; and as pre-
dicted, they lost.
As the loser in this interaction, the North began to innovate tactically
and strategically. In tactical terms, NVA units attempted to ‘‘cling to the
belt’’ of US or ARVN forces (Karnow, 1983: 463), getting so close so
12
No one involved in the campaign on the US side considers ROLLING THUNDER a
success, although it is fair to say that the reasons proposed to explain the campaign’s
failure have been numerous and controversial. From the military perspective, the con-
sensus is that the campaign failed because it misapplied military means to a political end.
For a more thorough discussion and analysis of ROLLING THUNDER’s target and
impact, see, e.g., Thompson (1980); Herring (1986: 147); Clodfelter (1989); and Pape (1990).
13
These examples are drawn from Summers (1995); and Marc Leepson (1999).
154
quickly that allied forces could not benefit from close air or artillery
support. In strategic terms, the North shifted more of its resources into
the guerrilla campaign in the South.
Interaction three: the guerrilla war in the South I, 1965–1973

Guerrillas in the South waged their campaign with considerable skill,
and were countered with a mix of professionalism, passion, sadism,
and incompetence. In this dimension of the conflict more than any
other, US efforts were heavily filtered through, and constrained by,
the GVN.14
In US army areas of responsibility, the first attempts to defeat the VC
insurgency involved search and destroy missions. In these counter-
insurgency (COIN) missions, regular army units, acting on local intelli-
gence that located enemy unit concentrations, would seek to make
contact with these concentrations and destroy them. Large-scale exam-
ples of this interaction include Phase I of Operation ATTLEBORO
(September—October 1965), Operation CEDAR FALLS (January 1967),
and Operation JUNCTION CITY (February—May 1967), but the inter-
action repeated itself on a smaller scale throughout the war. Forces were
killed on both sides, but the strategic balance slowly shifted to favor the
VC, largely because US forces relied heavily on indirect firepower —
tactical air and artillery support — a consequence of which was consider-
able death or injury to noncombatants (Asprey, 1994: 881).
Not all US COIN efforts failed outright or succeeded through barbar-
ism. US Marine Corps forces in the mountainous northernmost area
of South Vietnam, for example, pursued a COIN strategy combining
local and highly motivated (but poorly trained and equipped) villa-
gers with direct support from US Marine combat platoons. These
combined action platoons (or CAPs) operated on an ‘‘inkblot’’ principle:
secure a village or hamlet, then patrol out in widening circles until
intersecting with another CAPs’ secured area. The strategy had two
major disadvantages, however. First, although it protected South
Vietnamese citizens from immediate danger of terror attacks by VC
guerrillas, it could achieve success only in the long term — and time
favored the VC (Asprey, 1994: 848). Second, although militarily
14
The record of performance by ARVN units is mixed. Some units and their commanders
are extolled for their bravery, skill, and loyalty, while many are remembered only for their
cowardice, incompetence, and venality (see, for example, Karnow, 1983: 441; and
Summers, 1995: 80).
155
effective, CAP success only highlighted the inability of the South

Vietnamese government to protect its own citizens.
The United States lost this interaction. Its combat forces had been
trained and equipped to fight a uniformed regular adversary using
massive firepower, not an invisible enemy that refused to meet it in
battle.15 The indiscriminate impact of the US Army’s (and the ARVN’s)
heavy reliance on artillery and air support progressively alienated
potential allies among South Vietnam’s people. As losers, however,
US forces were not slow to innovate a strategic response.
Interaction four: the guerrilla war in the South II, 1965–1973

US strategic innovation aimed at seriously damaging the VC guerrilla
campaign in the South took two forms. The first was the Strategic
Hamlets program, and the second was the Phoenix program.16
The US Strategic Hamlets program reactivated the French agroville
program. In this version, South Vietnamese villagers were forcibly eva-
cuated from their homes and relocated to fortified hamlets.17 Where
implemented effectively, this program had the military benefit of
severely damaging VC intelligence and supply networks, at a political
cost of turning US public opinion against the war, as nightly news images
showed wailing peasants being forced to leave their villages. But the
program was rarely implemented effectively. In most cases corrupt
officials failed to deliver weapons and embezzled funds and supplies
intended to make the hamlets functioning and secure communities
(Karnow, 1983: 257). As a result, the program alienated the people
whose goodwill the United States and South Vietnam needed to win
the war: forced to leave their homes and then abandoned, many South
Vietnamese turned against their government and became active support-
ers of the VC. By decreasing the program’s COIN benefits and increasing
its political costs, the South Vietnamese government’s corruption and
incompetence rendered the Strategic Hamlets program a disaster.
The second US innovation was the Phoenix program, whose aims and
legitimacy continue to provoke sharp debate (Asprey, 1994: 910—911).
15
This is Krepinevich’s central thesis, (see Krepinevich, 1986: 4; see also Cohen, 1984:
166—167).
16
‘‘Phoenix’’ was the code name for a US assassination program that targeted VC
leadership.
17
On the Strategic Hamlets program in Vietnam, its logic, and its successes and failures,
see Karnow (1983: 255—257); Herring (1986: 85—86, 88—90); and Sheehan (1989: 310—311,
540, 687).
156
The overall military view is that Phoenix was a legitimate military

operation. It relied on special intelligence to target and destroy VC
leadership, and it proved to be the single most successful strategic
initiative pursued by US forces during the war.18 To most observers,
participants, and historians, however, the sustained effort to kill non-
combatants raised troubling questions about its legitimacy as an exten-
sion of US policy, or as a COIN strategy, regardless of its effectiveness.
Overall, the United States won this interaction. The Strategic Hamlets
program was never implemented properly, so its contribution to US
success in this interaction was negative. By contrast, the Phoenix pro-
gram, which eviscerated the VC command infrastructure in the South,
may even have provoked the North into its premature and disastrous
direct confrontation with US regular forces during the 1968 Tet
Offensive. Because both strategies systematically and deliberately tar-
geted noncombatants, both must be counted as barbarism — albeit
barbarism at the milder end of the violations spectrum.
Outcome: North Vietnam loses, then wins

The DRV won the Vietnam War, losing the military contest with the
United States by 1969, but delaying its defeat long enough (and at a
horrible cost) to force the United States to abandon the GVN (1973).
From March to April 1975 the NVA crushed the ARVN on the battle-
field. The accession to command of US forces of General Creighton
Abrams (July 1968) and his subsequent reforms and innovations —
leading to the virtual collapse of the VC/NVA military threat to South
Vietnam — proved irrelevant to the outcome, save to the extent to which
its continuing reconstruction (see e.g., Sorley, 1999) rehabilitates US
military honor.
Analysis: competing explanations of the Vietnam

War’s outcome
The outcome of the Vietnam War should stand as a puzzle in IR theory
for two reasons. First, as with all the unanticipated outcomes explored
18
There is little question that Phoenix effectively disrupted the capacity of the VC to
continue their GWS in the South. Even the North Vietnamese admit this: ‘‘Nguyen Co
Thach, Vietnam’s foreign minister from 1975, admitted that the Phoenix effort ‘wiped out
many of our bases’ in South Vietnam, compelling numbers of North Vietnamese and Viet
Cong troops to retreat to sanctuaries in Cambodia’’ (Summers 1995: 148; see also, Karnow,
1983: 534; and Herring, 1986: 232).
157
in this book, the DRV’s victory is puzzling in that the most powerful
actor, the United States, lost a war against a weak actor, the DRV.
Second, the DRV won the war even though, unlike weak actors in every
other case examined here, it lost on the battlefield.19 How can these puzzles
be explained?
Actor interests
A first explanation is that the DRV won because, as all observers came
to know, it was profoundly dedicated to the outcome it sought. As a
nation, the DRV appeared to forge itself into the living embodiment of
the sentiment ‘‘better a dead lion than a live jackal.’’ Better, in other
words, to be annihilated as a people than to live under foreign rule.
There was always, during the war, a sense of disbelief about this senti-
ment (possibly on the DRV side as well), and a reasonable concern that
it represented more the propaganda of a desperate regime than the true
aspirations of a people — a people very often divided historically. But
the balance of evidence supports the argument that the dedication of
the people of the DRV and their VC allies in the South, was nearly
unprecedented historically.20 General Abrams may have destroyed
their ability to reunify Vietnam and even set it back another twenty
years. But even under those circumstances the DRV showed no signs of
injury to its willingness to unify Vietnam, whether it took another
twenty, fifty, or a hundred years. In short, the roots of DRV resolve
had less to do with its ‘‘survival’’ as framed by IR theory (i.e., less to do
with it being the weak actor), and more to do with a two thousand year
history of nationalist opposition to foreign rule, regardless of the bal-
ance of forces.
19
To clarify, the DRV lost on the battlefield to the United States and the United States
withdrew in 1973. The DRV won on the battlefield against the GVN in 1975, but by then
few could have been surprised at the outcome.
20
If willingness to absorb casualties and continue pressing ambitious political demands
is an indicator, then the DRV is in rare company. As Mueller puts it, ‘‘the Communists lost
some 2.5—3 percent of their prewar population in the war in battle deaths. How does this
compare with other wars? It is almost unprecedented . . . scarcely any of the hundreds of
participants in the 100 international wars in the last 160 years have lost as many as 2
percent of their prewar population in battle deaths. The few cases where battle deaths
attained levels higher than 2 percent of the prewar population mostly occurred in the two
world wars in which industrial nations fought with sophisticated machines of destruction
for their very existence. In World War II . . . Germany and the Soviet Union each lost some
4.4 percent of their prewar populations in battle deaths. In World War I, Germany lost 2.7
percent, Austria-Hungary, 2.3 percent, France 3.3 percent, Rumania, 4.7 percent, and
England 2.0 percent’’ (Mueller, 1980: 507—508).
158
Compared to this the interests of the United States must at first seem
trivial. The United States sought only the survival of the GVN. But by
1965 the United States had come to view the survival of the GVN as a
necessary part of its larger overall strategy to contain communism.
Containing communism, in turn, was a vital national security interest
because if it could not be contained — if it spread — it could either directly
threaten US vital interests (stability in Europe, or oil in the Middle East,
for example), or escalate to another world war capable, in the thermo-
nuclear context, of ending all life on the planet. US interests in the out-
come of the Vietnamese civil war were therefore far higher than suggested
by the interest asymmetry thesis (Karnow, 1983: 377—378), but not as high
as those of the DRV and VC. The United States did fight in Vietnam,
initially, because many in the Kennedy and Johnson administrations
believed that if they didn’t fight the communists in Vietnam sooner,
then they’d be force to fight them in Europe or the Western Hemisphere
later. For the United States, Vietnam was a ‘‘limited’’ war not because
South Vietnam’s fate was a peripheral US interest, but because US poli-
tical elites believed that the use of force in proportion to US interests might
provoke Chinese military intervention and lead to a third world war.
Hypothesis 8 — relative material power explains relative interests in
the outcome of an asymmetric conflict — is therefore not supported in
the Vietnam case. The problem with actor interest asymmetry as an
explanation of the outcome of the Vietnam War, however, is not that it is
poorly explained by relative power, but rather that it doesn’t explain
potential outcome variation. In 1968, the United States was still com-
mitted to the idea that the GVN’s survival was a vital US national
security interest, under the theory of protecting US credibility (rather
than the earlier domino theory). In other words, both actors were
equally committed, independent of the nominal balance of coercive
forces, and on top of that the United States had won the military contest
on the ground. Given its high interests and its physical capacity to
continue escalating the use of force against its weak adversary, the
United States shouldn’t have been as politically vulnerable as it was.
Yet the United States was politically vulnerable, and that political vul-
nerability functioned just as Mack’s thesis predicts: it forced the United
States from the war.

If interest asymmetry is a poor explanation of the puzzles of DRV
victory, maybe regime type is a better one. Perhaps the DRV won
159
because its authoritarian structure made it possible to completely con-

trol information from the battle space, and enabled it to engage in more
brutal — and hence, more effective — tactics, than its democratic foe.
It is true that the DRV maintained an extraordinary degree of control
over information about the outcomes of battles, and about the degree to
which its enemies resorted to brutal conduct.21 In general, the DRV lied
to its people. It lied about nearly every aspect of the war, from the
conduct of US and ARVN forces, to casualties suffered — their own
and those of their adversaries. The DRV’s efforts to control information
from the battle space came at a significant cost in resources, and was
aided by the fact that most common soldiers in the NVA and VC were
not literate, and therefore could not write letters home in which alter-
nate versions of events could compete with the DRV line. This was
equally true of the ARVN but not of US forces, whose letters home
eventually accelerated a ‘‘credibility gap’’ already being widened by
journalists in theater.
There is also ample evidence that neither the DRV nor its VC allies
observed the Geneva Conventions regarding the treatment of noncom-
batants in war — ranging from surrendering soldiers, to prisoners of war
and enemy soldiers wounded in combat (many of whom were tortured
and then murdered, or simply murdered on the spot). The argument
that in Vietnam, in particular, a willingness to act with deliberate brutal-
ity was necessary to win, is captured nowhere better than in the words
of the fictional Lieutenant Colonel Walter E. Kurtz in Francis Ford
Coppola’s 1979 film Apocalypse Now:
It’s impossible for words to describe what is necessary to those who do
not know what horror means . . . I remember when I was with Special
Forces . . . We went into a camp to inoculate the children. We left the
camp after we had inoculated the children for Polio, and this old man
came running after us and he was crying. He couldn’t say. We went
back there, and they had come, and hacked off every inoculated arm.
There they were in a pile. A pile of little arms. And I remember I,
I cried. I wept like some grandmother . . . And I thought: my God, the
genius of that. The genius! . . . And then I realized they were stronger
than we because they could stand it. These were not monsters. These
were men — trained cadres — these men who fought with their hearts,
who had families, who had children, who were filled with love. But
they had the strength, the strength to do that. If I had ten divisions of
21
See Tin (2002: 30—32). Bui Tin supports the argument that the democratic structure of
the US government made it highly vulnerable as compared to that of the DRV.
160
those men our troubles here would be over very quickly. You have to
have men who are moral, and, at the same time, who are able to utilize
their primordial instincts to kill, without feeling, without passion,
without judgment. Without judgment, because it’s judgment that
defeats us.
The weakness of this argument is that the GVN was also authorita-
rian and fully willing to ‘‘do what was necessary’’ to win, and yet the
ARVN was always said to be a hindrance to military effectiveness
rather than an aid. In the calculation of US forces and diplomats in
theater, the ARVN were in fact the most brutal yet the least effective
forces engaged against the DRV/VC in the war. Also, US forces proved
to be just as brutal on several occasions (the My Lai massacre being only
the most famous and most public of many such ‘‘incidents’’), and as
observed above, accepted considerable collateral damage in strategic
bombing raids designed to coerce the DRV through the infliction of
pain. Yet the military effectiveness of the bombings was negligible. Had
it escalated to the point of destroying the Red River dike system (killing
up to a million DRV civilians), Bui Tin, then a colonel in the NVA,
argues that the results would have been catastrophic for the United
States:
Such a tragedy would . . . have offered Hanoi an incomparable occa-
sion for whipping up anti-American sentiments throughout the world,
for arguing for large-scale international aid and support, and for
rallying the whole socialist bloc.
My feeling is that when you take into consideration the sociopoli-
tical and psychocultural makeup of the Vietnamese, destroying the
dikes would have worked in Hanoi’s favor. The government would
certainly have been faced with many difficulties and the people would
have suffered untold hardships, but their war effort would not have
diminished . . . (Tin, 2002: 34)
But the strongest argument in favor of the effectiveness of regime

type is the structure argument; the structure of the DRV and GVN — both
authoritarian regimes — did not permit public disaffection with the war
or how it was fought to affect policy. But the structure of the US
government — a democratic regime — did permit public opinion to affect
US policy. Mack, Merom, and others correctly identify this as political
vulnerability, and specify how it operated to force the United States
from the war despite a military victory in theater.
The most important consideration here, however, is whether the
United States failed because it was a democracy. The answer is no, on
161
two counts. First, if democracies are to be associated with more

restrained conduct on the battlefield (historically not a sound associa-
tion), then the United States may, as Merom asserts, have lost because it
couldn’t ‘‘escalate the level of violence and brutality to that which can
secure victory’’ (Merom, 2003: 15). It is difficult to argue, however, that
the United States and ARVN were not brutal enough, considering that
as a proportion of pre-war population, the DRV lost from 2.5 to 3
percent — higher even than that suffered by Japan in World War II
(Japan lost an estimated 1.4 percent in that war).22 Second, it is hard
to argue that the United States in Vietnam was more politically vulner-
able, given its high interests in the outcome, than it was in the first Gulf
War of 1990, in which the rationale given for US military intervention
was only the preservation of a ‘‘new world order’’ (nothing said about
dominos). Yet the United States and its allies, who together inflicted
tens of thousands of casualties on their hapless Iraqi adversary while
sustaining only hundreds of casualties themselves, won that war
quickly and decisively, forcing the Iraqis to agree to humiliating
terms of surrender. Thus, political vulnerability drives democracies
from wars when it can’t affect authoritarian regimes, but only when
there is a gap in time between the commitment of armed forces and the
achievement of political objectives. This is what happened in Vietnam.
Thus regime type is necessary but not sufficient to explain the DRV’s
victory. A full explanation demands accounting for why some asym-
metric conflicts are over with quickly while others drag on.
Arms diffusion
Arms diffusion is not a good explanation of US defeat in the Vietnam
War for two reasons. First, the DRV’s acquisition of armaments from the
Soviet Union and China did make a difference (especially to the air
defense of key installations in North Vietnam), but only at the mar-
gins.23 Second, the battlefield impact of access to better arms was often
negative (Johnson, 1968: 442, 443). The DRV’s acquisition of new arms
22
Numbers are from Mueller (1980: 509).
23
A good counterargument would be that given ‘‘low’’ US interests in defeat of the DRV,
increasing costs at the margins was sufficient. But US interests were not low. Moreover, the
most serious costs imposed on the US were in the form of (a) outraged public opinion
following reports on the collateral damage of US bombing strikes in the North, and (b)
improvised booby traps and mines against US forces in the South. If Bui Tin and others are
correct, a lack of any conventional air defense of the DRV might have caused the United
States even greater problems by accelerating the trend toward increasing negative public
reaction to high-tech, high-altitude attacks against a ‘‘poor, defenseless, backward’’
162
tended to encourage premature escalation to direct confrontations with

US or US-supported ARVN units, and this invariably proved disas-
trous. Moreover, the lack of dependency on locals for logistical support
made it possible for the cadres to terrorize wavering South Vietnamese
peasants (summarily executing ‘‘informants’’ and so on) in direct oppo-
sition to Maoist dicta concerning the necessity of treating peasants with
respect. The legacy of this abuse would haunt the VC following the loss
of so many of their cadres in the ill-fated Tet Offensive of 1968. The DRV
was forced to send replacements from the North, but these men could
not reconstruct the intimate web of relationships with the peasants and
terrain of their predecessors. In addition, the new cadres were haughty
toward the ethnically distinct ‘‘Southerners,’’ and this, along with the
disaffection brought about through previous years of terrorization,
made it impossible to re-establish a viable insurgency in the South
after 1968. On balance then, absolute increases in armaments hurt
more than helped the DRV war effort.
Hypothesis 6 — the better armed a weak actor is, the more likely it is
that a strong actor will lose an asymmetric conflict — is therefore not
supported in the Vietnam case.
If it is true that actor interests and regime type are themselves insuffi-
cient to account for the DRV victory in Vietnam, it is nevertheless
obvious that they are necessary. But the best explanation of the war’s
outcome is strategic interaction. Strategic interaction explains how the
high resolve of the DRV/VC, along with the regime-affected political
vulnerability of the United States, combined to force the United States
from the war as and when it did.
According to my account, however, it is difficult to assess the cumu-
lative impact of the four 24 strategic interactions that made up the war.
country. Ironically then, the DRV’s use of Soviet anti-aircraft technology might count as
yet another example of advanced technology imports hurting, rather than helping, a war
effort.
24
Lewis Sorley (1999) argues that the war had yet another crucial phase after Tet, in which
US forces shifted strategies under their new commander, General Creighton W. Abrams.
In terms of the strategic interaction thesis, this would count as ‘‘Phase 5: Guerrilla War III/
indirect — indirect/US wins: abandons Vietnam/war ends.’’ The difficulty with Sorley’s
argument is measuring the effectiveness of Abrams’s talented leadership in the context of
a VC recently devastated by Phoenix and Tet, and a similarly routed NVA. If, as many
assert, the North had retreated to lick its wounds after Tet, then US military effectiveness
would appear high regardless of its strategy. For this reason, and because its inclusion
would add little in the way of a test of competing explanations, I do not include the
interaction in this analysis.
163
These interactions and their contribution to the outcome of the war are
summarized in Table 2.
The United States won two and lost two; and the impact of the
interactions on the war’s duration also seem equivocal. Shouldn’t this
imply a tie? Perhaps. But considering US public and elite expectations
prior to the fight, even a tie must count as a failure from the US
perspective, because anything other than a quick victory was going to
be politically unacceptable. Keep in mind that US strategy was aimed,
and aimed publicly, at intimidation — at coercing the DRV by force and
the threat of even more force to stop supporting the war against the
GVN in the South. In World War II, the United States had been fighting
a war many Americans believed to have been a real war: a war for
survival against a powerful, technologically fearsome, and unambi-
guously evil enemy. For the United States, that war lasted three long
years (1942—45). If three years is a long time in a ‘‘real’’ war, how much
longer must it have seemed in Vietnam, a far away place with a poor,
weak, ‘‘backward’’ and hence sympathetic enemy? Three years into its
war of intimidation against the DRV, the United States and the GVN
were attacked boldly and on all fronts by the DRV and VC. Clearly,
whatever else was working in Vietnam, and regardless of the military
outcome (NVA/VC defeat), US hopes of intimidating the DRV were
bankrupt, and the American people knew it.
Strategic interaction explains two things that Mack’s and Merom’s
theses can’t. First, as demonstrated in the Vietnam case, relative power
in an asymmetric conflict is a poor predictor of relative interests
(resolve, if you will) in the outcome of an asymmetric contest. Even
when facing weak actors, strong actors — whether superpowers or
middling powers, whether authoritarian or democratic — justify their
attacks against weak actors in terms of survival of some sort. This can be
by means of domino logic (the additive power of otherwise small defeats
could eventually constitute a direct survival threat), precedent-setting
logic (if we don’t hold here, an imagined audience will conclude we can’t
hold anywhere, and vital interests will be put in jeopardy as a result), or
by reference to some cherished identity or principle (one thinks, e.g., of
Margaret Thatcher’s justification for attacking Argentina over the
Falklands/Malvinas Islands in 1982; or George Bush Sr.’s declaration
of a need to protect a ‘‘new world order’’ by forcing Iraq from Kuwait in
1990). The point being that these causes need not be objectively exis-
tential. It is enough that they are metaphorically existential, at least
in the initial stages of conflict. If the war is over quickly, then the
164
Table 2. Summary of strategic interactions and effects in US

intervention in Vietnam, 1965—73
Strategic Duration
interaction Innovation effect effect
1 Rolling indirect— United States loses: lengthens war

Thunder direct abandons strategy
2 Main direct— North loses: withdraws, shortens war
force war direct then escalates (Tet)
3 Guerrilla direct— United States loses: lengthens war
war I indirect switches to
barbarism (Phoenix)
4 Guerrilla indirect— North loses: shortens war
war II indirect escalates, then
withdraws
metaphorical justification will likely stand. If it is protracted — a function

of strategic interaction — however, the metaphorical justification will be
challenged. In sum, strong actors are not vulnerable in small wars because
such wars are less than existential. They are vulnerable because the time it
takes to defeat an objectively weak adversary calls into question the
legitimacy of the interests, regardless of their bases in reality or metaphor.
Second, and most importantly, the strategic interaction thesis shows
why the Vietnam War wasn’t over with quickly, in say, one or two
years. As most critics of the US military in Vietnam have observed, the
initial strategic interaction was dominated by a conventional attack
(US/ARVN) against a GWS (VC). This interaction in fact best charac-
terizes the overall strategic interaction of the war; and as the thesis
predicts, one consequence of an opposite-approach interaction is that
it makes wars last longer. The failure to achieve quick results with
conventional search-and-destroy missions then led US forces to inno-
vate, leading to the CAPs program, Strategic Hamlets (again), and
Phoenix. These innovations had the effect of hurting the DRV and VC,
but the United States would not advance north of the 17th parallel with
ground forces, so the DRV could afford to wait.
If Mack’s thesis were sufficient to explain ‘‘why big nations lose small
wars,’’ political vulnerability would have to generally affect the ability of
strong actors to defeat weak actors. It doesn’t. Political vulnerability
operates only when there is an unanticipated delay between the com-
mitment of armed forces and victory.
165
If Merom’s thesis were sufficient to explain ‘‘how democracies lose

small wars,’’ he would have to support the empirical claim that strate-
gic interactions in small wars are constant, not variable. In other words,
he would have to show that, as his case selection implies, all [demo-
cratic] strong actors will attack with a conventional attack strategy, and
all weak actors will defend with a GWS, terrorism, or nonviolent
resistance (i.e., an indirect defense). But we in fact observe variation
in strategic interactions in small wars all the time; against all expecta-
tions to the contrary, the Taliban did not attempt a GWS against the US-
supported Northern Alliance in 2002. They attempted a conventional
defense and were soundly defeated. The same was true of the Iraqis a
year later. Moreover, the Russian Federation, a nominal democracy, is
still prosecuting an extremely brutal and yet counterproductive COIN
campaign in Chechnya. Even if we accept for the sake of argument that
the Russian Federation is not a ‘‘real’’ democracy, clearly the extreme
brutality (rape, extortion, looting, murder, torture, indiscriminate shel-
ling) of the Russian campaign is not winning Russia’s small war in the
Caucasus.
Conclusion
In Vietnam, the DRV (weak actor) had two entirely distinct militaries
ready to oppose US forces: one trained and equipped to fight an
indirect war (the VC), the other trained and equipped to fight a direct
war (the NVA). This meant that the DRV could be far more nimble
than the United States in shifting its strategic approach. As Eliot
Cohen, Andrew Krepinevich, Donald Hamilton, and others have
argued, with the possible exception of the US Marine Corps — which
had considerable operational experience with COIN — the US military
could never reconcile itself to the demands of a COIN war (Cohen,
1984: 165; Krepinevich, 1986; Hamilton, 1998: 155). These demands do
not imply the necessity of creating a force capable of barbarism; as the
CAPs demonstrated, it was possible to fight a GWS in the South within
the framework of the laws of war. What it was manifestly not possible
to do was defeat a people in arms quickly by such methods.
Thus the CAP example only underscores the importance of the key
causal mechanism of the strategic interaction thesis: when the power
relationship implies a quick victory, and the interaction causes a delay,
the way is clear for the operation of political vulnerability. That is, even
an ideal COIN strategy — one that destroys enemy forces without
166
destroying enemy values — takes time. If such strategies are to become a

model for future COIN operations, this implies a counterintuitive pol-
icy: when weak actors employ indirect defense strategies, strong actor
political and military elites must prepare their publics for long-delayed
victories.
In Vietnam, interest asymmetry and regime type helped the DRV
inflict costs on the United States and ARVN; and its open alliance with
the People’s Republic of China (and Soviet Union) made it impossible
for the United States to launch a ground invasion in the North. These
factors might not have mattered had the DRV escalated to an all-out
conventional assault on the South in 1965 (a strategy for which the
United States and ARVN were prepared). They also wouldn’t have
mattered had the United States taken British and French advice and
chosen the long and difficult strategy of a law-enforcement rather than
a military COIN strategy.25 To do so would have required three things:
first, reform and restructuring of the GVN (the relevant example of
success in this regard being Ramon Magsaysay’s reforms of the
Philippine government during the Hukbalahap insurgency in 1952).
Second, the politically difficult — in the US domestic political context —
step of persuading the US public of the need for a long struggle. Third,
the even more difficult job of convincing the US military to build and
deploy sufficient special operations forces to work in close conjunction
with well-trained and competently led GVN law enforcement forces.
Either same-approach interaction could, without resort to barbarism,
have secured US political objectives in Southeast Asia. Although this
strategy would suffer from the defect of taking a long time to show
results, that cost would be offset by, first, a consistent political effort to
keep expectations of a quick victory low, and more importantly, by
reducing friendly casualties and the DRV propaganda advantage of
frequent and widely publicized incidents of injury to noncombatants.
But, in the event, the United States encouraged the view that the fight
to save the GVN was a military matter and that it could be won
quickly.26 The United States attempted to coerce the DRV with
25
See Shaw (2001). Adopting this strategy would count as switching from a war-winning
to a war-termination strategy.
26
This encouragement culminated in the infamous prediction by US commander William
Westmoreland — just prior to Tet — that ‘‘the enemy’s hopes are bankrupt’’ (see Karnow,
1983: 539). Westmoreland’s proclamation mirrors that of French General Henri Navarre’s
on the eve of the battle of Dien Bien Phu, ‘‘Now we can see it clearly, the light at the end of
the tunnel’’ (Karnow, 1983: 189).
167
firepower (Herring, 1986: 226), and the DRV/VC defended their inter-
ests by means of an indirect strategy and at a tremendous cost in lives
lost. Stymied, US forces innovated new strategies in order to defeat the
DRV/VC, but the reverses they achieved on the battlefield took too long
(Karnow, 1983: 464, 480; Herring, 1986: 200). By 1968 the US public had
already begun to grasp what the Johnson and Nixon administrations —
and the military leadership — could not; by the time US Marines landed
at Da Nang in 1965 it was already too late. The war in Vietnam had
become what Michael Walzer would later term an anti-social war:
The war cannot be won, and it should not be won. It cannot be won,
because the only available strategy involves a war against civilians;
and it should not be won, because the degree of civilian support that
rules out alternative strategies also makes the guerrillas the legitimate
rulers of the country. The struggle against them is an unjust struggle as
well as one that can only be carried on unjustly. Fought by foreigners,
it is a war of aggression; if by a local regime alone, it is an act of
tyranny. The position of the anti-guerrilla forces has become doubly
untenable. (Walzer, 2000: 195—196)
Just as it had to the French public after General Massu’s defeat of the
FLN in Algeria, the thought of winning such a war soon became intole-
rable to most Americans. Better to risk a communist takeover of
Southeast Asia than to contain communism by means of winning an
anti-social war.
168
7 The USSR in Afghanistan:
the Afghan Civil War, 1979–1989
The Afghan society may now be regarded a murderous society . . . We

have become soulless and dry, no longer beings of care and love, but
brutal and fierce animals. It is not right to name a society murderous,
but the Afghan society may be called so . . . The state is a state of
killing, not only in the battlefield but also in the lanes and streets
of cities where there is no state of war. No one feels secure, and because
of this many families have fled abroad. Hassan Kakar (1995)
In Afghanistan I served at Ghazni . . . Life in the unit was fairly calm,
unless we were involved in operations, when everything was
different . . . We didn’t see any friendly Afghans anywhere — only en-
emies. Even the Afghan army was unfriendly . . . Everyone around us
was an enemy. I remember an intense feeling of anger toward the rebels
because so many of our guys were getting killed. I wanted revenge.
Then I began to doubt the goals and methods of international aid. I had a
difficult time deciding what I really believed. I just knew what I had to say
during the political instruction meeting: that we were fighting ‘‘American
aggression’’ and ‘‘Pakis.’’ Why had we mined all the approaches to the
regiment? I asked myself. Why were we aiming our machine guns at every
Afghan? Why were we killing the people we came here to help?
Whenever a peasant was blown up by a mine, no one took him to
the medical unit. Everyone just stood around, enjoying the sight of his
death. ‘‘This is an enemy,’’ the officer said. ‘‘Let him suffer.’’
Mickola Movchan (a Russian deserter in Afghanistan)
The war between the Soviet Union and resistance forces of

Afghanistan began on 27 December 1979, and ended with the with-
drawal of Soviet armed forces on 15 February 1989.1
1
The fighting did not stop, but for our purposes the asymmetric conflict between the
USSR and the various Afghan resistance groups ended when the Soviet troops evacuated
Afghan territory.
169
The Afghanistan theater

The territory of Afghanistan encompasses a mountainous region of
245,000 square miles, or roughly the size of New Mexico and Arizona
together. It is essentially the hub of a wheel which includes the former
Soviet Union to the north, Iran to the west, Pakistan to the south and
east, and China to the east. As such, it lies along a number of ancient
trade routes connecting the east to the west, and its geographic position
has been the most important determinant of its foreign policy for
centuries.
In terms of ethnicity, religion, and language, the majority ethnic
group of Afghanistan are Sunni Muslim Pushtuns (often called
‘‘Pathans’’ in nineteenth-century British writings). Pushtuns make up
about 40 percent of the population, followed by Tajiks (20 percent). The
next three largest groups — roughly equal in size — are Hazaras (Shi’a
Muslims with a distinct racial appearance), Uzbeks, and Aimaq
(Magnus & Naby, 1998: 10). The main languages of each group are
distinct from each (e.g., Pushtu for Pushtuns), but an inter-ethnic lan-
guage, Dari, is also available (Magnus & Naby, 1998: 15). Finally, 90
percent of Afghans are Sunni Muslims, while most of the remainder
follow variations of Shi’a Islam. When religious identities are most
salient, this makes most Afghans natural allies of Pakistan and Saudi
Arabia, and creates friction with Iran, which is predominately Shi’a
Muslim.
The Pushtun identity is important for two reasons. First, this group
has dominated Afghan politics and society for hundreds of years (pro-
viding all its rulers, and several for Iran and India as well). Second, the
Pushtun identity is important because the Afghan—Pakistani border —
established as the so-called Durand Line of 1893 — divides ethnic
Pushtuns almost in half (Magnus and Naby, 1998: 36). In fact, what
the Pakistanis call the North-West Frontier Province, the Afghans refer
to as Pushtunistan.
The internal political struggles which led to a Soviet decision to
intervene militarily in Afghanistan are far more nuanced and complex
than can be presented here. However its main features are as follows.
After World War II Britain allowed India to become independent in
1947, and after Muslim Pakistan split from Hindu India, the
Eisenhower administration calculated that Afghanistan was beyond
its sphere of strategic interests (Kakar, 1995: 9; Magnus & Naby, 1998:
59). As such, then President Sardar Muhammad Da’ud Khan was
170
The USSR in Afghanistan
forced to look elsewhere for development aid, which his country

needed to build roads, irrigation and power facilities, and a modern
military. The USSR stepped into the vacuum, offering billions of rubles
to the Da’ud regime, showcasing its assistance as friendly ‘‘no strings
attached’’ development aid. But, in 1963, Afghanistan’s king, Zahir
Shah, dismissed Da’ud as prime minister. During the next fifteen
years, Da’ud’s successors liberalized Afghan domestic politics while
continuing to develop Afghanistan’s infrastructure.
In 1965, political liberalization resulted in the formation of the
People’s Democratic Party of Afghanistan (PDPA) under the leadership
of Muhammad Taraki and Babrak Karmal. In 1967, the PDPA split into
separate factions: Parcham, headed by Karmal, and the more radical
Khalq, headed by Taraki. The liberalization of Afghan politics resulted
in accelerating factionalization and a growing opposition to the consti-
tutional monarchy.
In 1973 Da’ud staged a bloodless coup with Soviet and Parchami
support and declared Afghanistan a republic. But Da’ud immediately
set about slowing the pace of Afghan political reform. He arrested
prominent Islamists and Marxists, and began to purge Parchamis
from government posts while at the same time reducing Afghanistan’s
technical and economic dependence on the USSR (he signed develop-
ment pacts with Iran and Saudi Arabia which together would have
dwarfed the entire Soviet aid program of the previous two decades).
Alarmed, the Soviets requested an interview with Da’ud in Moscow in
April of 1977, and Brezhnev bluntly ordered Da’ud to stop hiring
foreign technical specialists on the grounds that many of these were
NATO spies. Da’ud retorted angrily that Moscow had no say in
Afghanistan’s internal politics, and the Soviets decided to replace him
(Magnus & Naby, 1998: 118—119, 121).
In July, the Soviets forced the reunification of the two feuding fac-
tions of the PDPA, and a year later (April 1978), the PDPA staged a
Soviet-engineered coup in which Da’ud and his family were mur-
dered.2 Taraki and his protégé Hafizullah Amin took over, and began
purging the new government of Parchamis (Karmal fled to Moscow).
Chaos ensued as the inexperienced Marxists attempted to re-engineer
Afghan society overnight. Uprisings soon broke out in major Afghan
cities, such as Herat, and were brutally suppressed. Amin eventually
2
Lloyd Rudolph argues that the April 1978 coup caught the Soviets by surprise
(Rudolph, 1985: 4).
171
ousted Taraki, who fled the country. As uprisings continued and DRA
desertions increased, Amin’s regime threatened to fly apart. Moscow
decided that the only way to salvage the situation was to depose Amin
and replace him with the more moderate Taraki.3 But their plans were
discovered by Amin, who had Taraki arrested and later murdered after
Taraki’s return to Kabul. In September 1979 the deeply angered Soviets
began elaborate plans to invade Afghanistan and install Babrak Karmal
as leader.
The eve of war: Soviet interests

There is much debate concerning Soviet interests in Afghanistan, but
the proximate causes of Soviet intervention appear to reduce to three
concerns: (1) the defense of a friendly Marxist regime in a territory
bordering the USSR (the Brezhnev Doctrine4); (2) the chance to establish
a geopolitical bridgehead in the Persian Gulf; and (3) sunk costs.5
These three rationales for intervention hardly exhaust the possibil-
ities, but with the advantage of hindsight we can make a better guess as
to Soviet interest priorities based on the mission, quality, and quantity
of armed forces they committed there.6 Whatever the disagreements
3
Rudolph suggests that Amin was too much of a Pushtun nationalist for the Soviets, and
that they decided to replace him because, among other things, his expansionist national-
ism threatened both Zia’s Pakistan and Khomeini’s Iran (Rudolph, 1985: 39). Urban
suggests yet a different rationale for the Soviet decision to oust Amin: he maintained
links with the CIA and was secretly in the West’s pocket (Urban, 1988: 40).
4
As with the United States in Vietnam, and Britain in South Africa, part of the subtext of
the Brezhnev Doctrine is simple prestige (a social component, avoiding humiliation) and
domino logic (a geopolitical component, gaining power or security). In this case, the logic
is of the form ‘‘if a Marxist regime is defeated here, Marxist regimes everywhere will be
weakened’’ (See Magnus and Naby, 1998: 63).
5
Magnus and Naby argue that in fact the Soviets were most affected by the sunk costs
argument, and by the opportunity to threaten vital US interests in the Persian Gulf: ‘‘The
Soviets, in fact, cared nothing for the revolutionary transformation of Afghanistan, but
they cared a great deal about the control of Afghanistan, which could provide them a
secure base (and valuable, well-armed allies) for further advances into more promising
areas of the Middle East and South Asia. They wished above all not to lose what they had
gained through decades of patient effort and considerable expenditure’’ (Magnus and
Naby, 1998: 122). As we will see below, the Brezhnev Doctrine and sunk costs arguments
are probably the strongest. We encountered a variant of the ‘‘sunk costs’’ argument in
Chapter 6, when Mussolini invoked it in order to justify his ‘‘solution’’ to the Ethiopian
‘‘problem’’ following the failure of his fait accompli strategy.
6
Analysis of the interests based on Soviet forces alone would be risky because these
forces — four motorized rifle divisions composed almost entirely of third-echelon Central
Asian reservists — might represent little more than Soviet arrogance or underestimation of
Afghan resistance (see a similar analysis of British interests in South Africa in Chapter 4).
172
about the USSR’s long-term objectives in Afghanistan, their immediate

objectives were clear from the beginning: (1) replace Amin with
Karmal; (2) occupy and stabilize major Afghan cities (Kabul, Herat,
Kandahar, Jalalabad); and (3) secure lines of communication between
cities (especially the Salang highway, linking Kabul with Termez to the
north and Jalalabad to the east). In short, the Soviet mission was strictly
limited. Their goal in Afghanistan was to stabilize the new regime
under Karmal’s leadership, so that Karmal could set about the difficult
political task of putting the Afghan Humpty-Dumpty back together
again (Litwak, 1992: 79). The limited mission, poor quality and small
quantity of forces supports the Brezhnev Doctrine7 and sunk costs
arguments better than the geopolitical bridgehead argument (Litwak,
1992: 79—80). The bridgehead argument is further weakened by the fact
that the alleged advantages of outright conquest (in terms of air bases
and so on) would have been available to the Soviets by treaty anyway.
But why intervene with regular armed forces in the first place? The
answer is that Amin had succeeded too well in purging opposition
from, and securing control of, the armed forces and secret police. In
essence, Amin’s efficiency and brutality forestalled the possibility of his
being replaced in the old-fashioned way, by a well-organized coup (his
discovery of the plot by the Soviets and Taraki to remove him is proof of
this). Thus the only means left to the Soviets was direct military inter-
vention with their own armed forces.
In sum, Soviet interests in Afghanistan were defensive from the
beginning. Their limited objective was to secure the regime, major
urban areas, and communications links so that Afghanistan’s new
political leadership could consolidate the revolution and pacify the
state by political, rather than military means.
The eve of war: mujahideen interests

As has been persuasively argued by an Afghan historian who lived
through these events, the Afghan state as such ceased to exist after the
7
This interpretation is also supported by Litwak. Litwak adds that, in effect, the Soviets
were concerned about precedent-setting effects (Duffy Toft, 2003) should an Islamic
regime come to power in Afghanistan, because Afghanistan borders Soviet Central
Asian Republics (Turkmenistan, Uzbekistan, Tajikistan) which were themselves formerly
Islamic states. In other words, the concern was that an Islamic Afghanistan would
destabilize Soviet Central Asia (Urban, 1988: 206; Litwak, 1992: 78). In hindsight, this
concern (whether it motivated Soviet intervention or not) has proven to be a valid one
(Magnus and Naby, 1998: 68).
173
April 1978 coup which toppled the Da’ud regime (Kakar, 1995: 125).
What remained in Afghanistan were two highly factionalized camps
gathering around two opposite poles.
On the one side were the various factions of the PDPA, which sought
to ‘‘reform’’ Afghan society while enriching themselves and eliminating
their rivals. Kakar argues that there were two problems with these
reform attempts. First, the PDPA leadership had no cadre of experi-
enced ministers who could oversee the implementation of their radical
decrees. Second, they relied on a Soviet or Marxist interpretation of
history in order to understand Afghan problems and to prescribe
proper solutions:
each [of the men who replaced Da’ud] was convinced that the PDPA
blueprint was the guideline for reorganizing both society and the state.
Thus, they relied on Soviet, not Afghan, experience, and thus, too, they
broke with the Afghan past. This may explain why, after they rose to
power, they became ever more alienated from their own people and
ever more disunited among themselves. (Kakar, 1995: 15)
On the other side were a number of mainly rural-based groups with

different ideologies, ethnicities, and agendas. Some were Sunni Muslim
and others were Shi’a. Of the Shi’a, some were Twelvers and some were
Seveners.8 Among the Sunni Muslims, some were ethnic Pushtuns
while others were Tajiks, Turkmen, or Uzbeks. Cordesman and
Wagner identify twelve separate resistance groups (Cordesman &
Wagner, 1990: 17—19). All the major resistance groups maintained a
similar organization: political leadership was located in Peshawar in
the North-West Frontier Province of Pakistan, while most field com-
mands were located in home provinces in Afghanistan.9
But however many groups or agendas there were, a key feature of the
decade-long conflict in the Afghan war was this: so long as Soviet
troops were involved in operations, resistance groups were united:
The invasion turned the civil war into a war of liberation. It gave that
war a new meaning, summed up in the word jehad, an expression
particularly moving to Muslim Afghans in such times . . . The
Russians were godless communists, and their ruthless suppression
8
Seveners and Twelvers constitute variants of Shi’a Islam. For a concise discussion, see
Magnus and Naby, (1998, pp. 84—87).
9
Kakar argues that Pakistan, fearful of another ‘‘Palestinian’’ situation in the North-West
Frontier Province, deliberately fostered the rivalries and factionalization of the resistance
groups (Kakar, 1995: 93).
174
of the Muslims of Central Asia had been related to the Afghans by the
thousands of the Muslims of Bukhara who had taken refuge in
Afghanistan. (Kakar, 1995: 111)
Furthermore, because Karmal had ‘‘arrived on the back of Soviet

tanks’’ the resistance fighters did not distinguish between DRA and
Soviet troops. To the mujahideen, they were all ‘‘the Russians’’ (Magnus
& Naby, 1998: 139).10
In sum, among the constellation of interests and identities that com-
posed the Afghan resistance, a single dominant theme emerged: get the
Soviets out (Urban, 1988: 53, 72). Beyond this simple goal there was no
consensus, and the lack of a broader consensus made the task of ejecting
the Soviets that much more difficult.
The Afghan Civil War

The Afghan Civil War contained two main strategic interactions. In the
first or enclave interaction, the Soviets limited themselves to establish-
ing and protecting key strategic areas, including logistical, urban and
industrial sites, and the roads linking them. But this proved difficult, so
in order to safeguard key areas, the Soviets launched a series of con-
ventional attacks against suspected mujahideen redoubts throughout
Afghanistan. This conventional attack strategy lasted from 1980 until
1982. The mujahideen followed a GWS strategy during this time and
throughout the war. In 1982 the Soviets switched to a ‘‘scorched earth’’
strategy. This barbarism strategy sharply escalated civilian deaths and
increased the flow of refugees to Pakistan and Iran. More importantly, it
began to severely damage mujahideen infrastructure and it took a
serious toll on their fighting capacity.
Interaction one: conventional attack vs. GWS

The Afghan war itself began with the invasion of Soviet troops on 27
December 1979. Initial Soviet invasion forces, the 40th Army, consisted
10
There appears to be some problem with this argument however. The departure of the
Soviets from Afghanistan in 1989 had two consequences whose effects are difficult to
disaggregate: (1) such mujahideen unity as there was quickly diminished after the last
Soviet combat units crossed the Amu Darya (Oxus) River, and (2) foreign logistical
support for the mujahideen dried up. Both consequences led to a diminished mujahideen
fighting capacity, but the first consequence implies that many mujahideen did in fact
distinguish between the white Russians and their DRA puppets.
175
of five divisions — four motorized rifle (MRD) and one air assault
division — supported by an additional two air assault regiments (the
103rd and 104th). The plan was for the 66th and 357th MRDs to advance
from Kushka in Soviet Turkmenistan to Herat, Farah, and then
Kandahar, while the 360th and 201st MRDs moved south from
Termez in Soviet Uzbekistan through Mazar-e-Shariff to Kabul. The
105th Guards air assault division and its two attached regiments were
already in place at Bagram and along the Salang highway passes, and
began the operation to capture Kabul at 7:15 p.m. By 8:45 p.m., the
Soviets had announced via a radio station in Termez claiming to be
Radio Kabul, that Karmal had taken over the government and
requested Soviet assistance. The ground forces now began crossing
the Oxus River on pontoon bridges.
The only serious fighting on that first day was between Soviet air-
borne and special forces troops and Amin’s personal guards at the
Darulaman Palace on the outskirts of Kabul. The Soviets had done as
much as possible to facilitate a quick takeover, including poisoning
Amin, emptying fuel from armored vehicles of the two divisions loyal
to Amin, and exchanging live rounds for blanks (for ‘‘exercises’’) in the
weapons of those same divisions (Kakar, 1995: 23). All that stood
between Amin and death were therefore 1800 personal guards who
occupied defensive positions in the palace. When the attack on the
palace by Soviet special forces units began, Jahandad, the guard com-
mander, came to Amin to ask for instructions. Due to Amin’s condition
(Kakar claims he was poisoned by the Soviets), however, Jahandad had
to make up his own mind. He decided to resist, and the attackers were
thrown back several times:
The confrontation was intense and prolonged. Both sides sustained
losses until the Afghans were finally overcome by some kind of nerve
gas . . . The invaders feared that if the Afghans were not soon over-
come, forces from the nearby military divisions of Rishkhor and
Qargha might join them. (Kakar, 1995: 25)
This passage not only references a violation of the laws of war,11 but
cites its most common rationale: it was ‘‘necessary’’ in order to increase
military effectiveness.
11
Kakar notes that none of the 1800 palace defense force survived the attack, which tends
to support the argument that something unusual was used to overcome the defenders.
176
In January 1980, Karmal announced a package of reforms and com-

promises designed to reestablish government and end resistance and
violence. Most Afghans ignored him:
The Karmal package was a comprehensive one, but it was ignored by a
large number of Afghans. To them the pledges of freedom, the release
of prisoners and the new trades unions were all irrelevant. What was
relevant was that Karmal clearly depended on the army of a foreign
power — a power that was perceived as godless and anti-Islamic.
(Urban, 1988: 53)
Most Afghans were outraged by the Soviet invasion. Many quit their
jobs and joined the mujahideen. Worse still, entire DRA units defected
to the mujahideen. The defection of whole units was a particular pro-
blem because it meant the transfer of arms — sometimes sophisticated
arms — to the resistance. Karmal’s well-intentioned attempt to reach
compromises within his government caused still more problems.
Amin’s Khalqis were still in key positions in the army and secret police,
and because Karmal refused to purge them, they remained capable of
obstructing his policy initiatives.
By mid-January DRA desertions had become such a problem that
Marshal Sokolov, in overall command of Soviet operations in the DRA,
mobilized his reserve-echelon forces. In order to seal the Iranian border,
Sokolov sent the 5th MRD to Farah and the 54th MRD to Herat. A third
MRD, the 16th, was deployed to Mazar-e-Shariff. He also sent the 201st
MRD east from Kabul to Jalalabad in order to stabilize the situation
there.12
Meanwhile, in order to relieve the pressure on the DRA 9th Division
at Jalalabad, the Soviets planned to take the offensive in the Kunar
valley. The Kunar valley perfectly illustrates the conflict topography
of Afghanistan: a long fertile valley flanked by steep mountains con-
taining numerous smaller side canyons. The valleys themselves usually
feature a river, which flows down to join other rivers from similar
valleys all through the Hindu Kush mountain range. Such roads as
there were in Afghanistan usually follow the rivers, and the intersection
of rivers thus becomes the site of urban centers and major highway
junctions. In short, Afghanistan’s valleys open onto strategic values.
12
The DRA 9th division was garrisoned at Jalalabad on the key highway connecting
Pakistan to Kabul. Urban estimates that so many of its units defected that by the summer
of 1980 its effective strength was down to no more than 1000 (Urban, 1988: 55).
177
What the Soviets quickly discovered is that they could not achieve
their minimal objective of protecting these strategic values because
there weren’t enough forces among the DRA and the Soviet army
combined to protect more than a fraction of them.13 Convoys of fuel,
ammunition, food, and replacement soldiers and parts had to travel the
full length of highways which were flanked in many places by these
valley openings. Thus, the Soviets found themselves having to mount
COIN operations outside of their garrisons simply in order to relieve
pressure on them and secure their lines of communication. The Kunar
valley operation in March of 1980 was their first major effort, and it
proved completely ineffective.
The offensive began with a sustained aerial bombardment, which
warned the mujahideen, killed civilians and livestock, destroyed fields,
houses and irrigation facilities, and initiated the depopulation of the
valley. Next, armored regiments lumbered up the valley floor, suffering
numerous ambushes. The column succeeded in relieving the garrison
at Asadabad and reestablishing government authority in Asmar, after
which it returned to base. Days later the mujahideen returned from the
mountain caves where they had hidden as the Soviets ‘‘pacified’’ the
valley, and the strategic situation returned to what it had been before
the offensive.
In June and July fighting intensified, especially in the three key
frontier provinces of Kunar, Nangrahar, and Paktia.14 The Soviet offen-
sive in Paktia proved to be a major disaster for the Soviets. There were
three problems. First, the troops themselves were poorly trained reserv-
ists, and the system of enlisted seniority in the Soviet military essen-
tially destroyed the effectiveness of NCOs.15 Second, the equipment of
the Soviet Southern Front Military District was old and obsolete.
13
By way of comparison, at the height of the US commitment in Vietnam, US troops
maintained a coverage density of 7.3 soldiers per square mile, while at no time during the
decade-long occupation of Afghanistan did Soviet soldiers achieve a coverage density of
greater than 0.7 soldiers per square mile (Cordesman and Wagner, 1990: 96). Even
restricting Soviet troops to highways, supply depots, and cities was never enough to
secure them adequately (Urban, 1988: 119—120).
14
These were key provinces because they bordered Pakistan, which throughout the war
remained the primary conduit of mujahideen logistical support (arms, ammunition, and
so on).
15
Terms of service typically run for two years, and soldiers serving their second year, the
‘‘old soldiers,’’ have license to harass and abuse new arrivals, NCO stripes or not (see
Urban, 1988: 127—128). This gap in small unit leadership meant that junior officers often
had to take on the responsibilities normally fulfilled by NCOs. This seriously degraded
the combat effectiveness of these units, especially given their COIN missions.
178
Armored personnel carriers (APCs) were 10—15 years old, and tanks
were often 20-year-old T-55s. Neither vehicle had anti-personnel weapons
which could depress or elevate sufficiently to engage targets in steep
mountain terrain. Third, the tactics used were ineffective for COIN
operations, which require infantry-heavy small units trained to act
independently, and supported by helicopters. In the summer of 1980,
there were still only an estimated 45—60 helicopters in the whole theater.
In the Paktia offensive an entire motorized rifle battalion was wiped out
in a single ambush. The first vehicle in the formation was immobilized
by a mine or grenade, and the rest of the convoy halted. Once shooting
started, the inexperienced troops began firing blindly, and continued
firing from within their vehicles until their ammunition ran out. They
were then overwhelmed. The defeat at Paktia, along with many other
setbacks that summer, prompted a massive Soviet reorganization:
‘‘From June 1980 to mid-1981, the 40th Army was restructured from a
force of seven [MRDs] to one of three [MRDs], two independent motor
rifle regiments and two motor rifle brigades’’ (Urban, 1988: 67). Tanks
were sent out of Afghanistan and helicopters were sent in. From about
60 total helicopters in 1980, by mid-1981, Afghanistan had three com-
plete helicopter regiments of 40—50 machines each. Finally, the vulner-
ability of fuel convoys prompted the construction of a fuel pipeline
from Termez to Pol-e-Khumri in Baghlan province (north of Kabul
along the Salang Highway). This pipeline was complete by August of
1980 (Urban, 1988: 68).
Offensive operations continued during the reorganization and, in
September, the Soviets launched the first of what would become the
focal point of the contest between the Soviet army and mujahideen: the
nine Panjsher pacification offensives. The Panjsher valley opened onto
a key area of the Salang highway south of the famous tunnel linking
north and south Afghanistan through the Hindu Kush mountain range.
The lower end of the Panjsher valley thus opened on the main supply
route from the USSR to Kabul. Mujahideen under the expert command
of Ahmad Shah Massud had been wreaking havoc along this route for
months. In September, the Soviets prepared to teach Massud a lesson.
It didn’t work. Although the Soviet offensive included a heliborne
landing by Soviet air assault troops, the results were the same as
at Kunar, except that this time the Soviets lost several helicopters.16
A second offensive in Panjsher in September (Panjsher II) took a similar
16
The mujahideen claimed to have shot down ten (Urban, 1988: 70).
179
form and had similar results. Meanwhile, the Soviets moved against
several mujahideen strongholds in the Logar valley, which opened on
the vital stretch of highway from Kabul south to Jalalabad. Here the
Soviets deployed heliborne troops in a new ‘‘cordon and search’’ tac-
tic.17 These new tactics, along with the deployment of much improved
APCs, led to the most successful COIN operations of the year (Urban,
1988: 72). Still, the overall results were quite meager, and the mujahid-
een still controlled most of the countryside.
In 1981 the USSR reduced its sweep operations and relied more
heavily on air power. Most of the few ground offensives which were
launched were of the Kunar and Panjsher pattern: they resulted in
sporadic desertions, lost equipment, and minor Soviet casualties,
while killing noncombatants and destroying Afghan infrastructure.18
The mujahideen were left largely unscathed.
In 1982 the Soviets returned to large-scale pacification sweeps, and
achieved moderate success. In April and May they launched Panjsher V,
and in September, Panjsher VI. Neither proved decisive, though again,
Soviet tactical (especially air-mobile) and technological innovation
made them somewhat more effective than previous Panjsher cam-
paigns. Still, Massud’s mujahideen were left with their fighting
strength intact, while the Soviets were forced to retreat to their gar-
risons. Urban gives the relative casualty figures for Panjsher VI as 2000
Soviet, 1200 DRA, 180 mujahideen, and 1200 civilians dead or wounded
(Urban, 1988: 109). In other areas, however, the Soviets were roundly
defeated. In April an independent DRA division attempted to clear the
road from Jalalabad to the Pakistan border and was almost annihilated.
What marked the Panjsher VI campaign off from previous campaigns
however, was the shift in Soviet strategy from direct attacks against
17
The tactic worked this way. A village suspected of containing mujahideen was selected
for attack. After a preliminary mortar or air strike, an air assault battalion ‘‘stop group’’ is
airlifted to a position between the village and the most likely route of retreat. Mines are
dropped to seal off others as a motorized rifle regiment slowly moves into and through the
village, closely supported by combat helicopters (Isby, 1989: 50).
18
Note that the Soviet strategy here is not yet barbarism, although it clearly had this
effect. Instead, the Soviets were attempting to ‘‘delouse’’ Afghanistan (in both ideological
and racial terms, the Soviets viewed the mujahideen as subhuman parasites). But delous-
ing requires tweezers and what the Soviets had was a tommy gun. The results of their
‘‘punitive offensives’’ (a term used by Urban to describe both the motivations and effects
of Soviet operations — see Urban, 1988: 88) were akin to attempting to delouse their DRA
ally with a tommy gun. For the remainder of the conflict the Soviets could never innovate
a way to remove the parasite without also killing its host. After 1982 they simply gave up
trying.
180
mujahideen bases, to the deliberate and systematic destruction of build-

ings, irrigation systems, crops and orchards (Rais, 1994: 102—103), as
well as the deliberate targeting of noncombatants to sow terror and
‘‘cleanse’’ disputed valleys:
Since they could not differentiate the mujahideen from the locals and
since they could not engage the mujahideen in battles, the invaders
tried to detach them from their own people. Intending to destroy the
rebels’ support among the civilian population, they also turned
against the non-combatants, destroying their villages, their crops,
and their irrigation systems and even killing them. Indiscriminate
destruction of property and human life, civilian as well as military,
thus became a feature of Soviet military expeditions. This was par-
ticularly so when the mujahideen killed Russian soldiers. In such cases
the invaders massacred civilians by the droves. By the force of circum-
stances the invaders found themselves in a situation in which they
killed hundreds and thousands of those for whose protection they had
purportedly come. (Kakar, 1995: 129)
If Kakar and Rais are right, the Soviets were pursuing the very text-
book definition of a barbarism strategy: a systematic targeting of non-
combatants in pursuit of a military or political objective. Urban himself
notes that the Panjsher VI operation had a devastating effect on
Massud’s forces:
For [Massud] this was the more serious long-term cost of Panjsher 6.
By October he was forced to appeal for food for his men. Although
Massud himself remained alive and much of his army was intact,
Panjsher 6 did cause lasting damage to guerrilla infrastructure in the
valley, undoing years of work by the mujahideen. Visitors estimated
that the population of the valley dwindled from 80,000 prewar to
45,000. (Urban, 1988: 109)
In short, Soviet barbarism appeared to be militarily effective. But was

the Soviet strategy a barbarism strategy or was it, as Urban argues, a
particularly blunt direct attack strategy which resulted in high collat-
eral damage?
Interaction two: barbarism vs. GWS

There is evidence to support both collateral damage and barbarism
interpretations, but the weight of evidence supports the barbarism thesis.
The collateral damage argument is difficult to sustain for three rea-
sons. First, as noted above, by 1987 an Afghan population of roughly
181
15 million had been reduced by almost 10 percent (Sliwinski, 1989: 39).

This means that as a percentage of their population the Afghans suf-
fered more deaths than did the Soviet Union during World War II.
Moreover, the war prompted probably the largest mass emigration in
history: almost half the civilian population — 6,000,000 Afghans — fled to
Iran and Pakistan during the war (Cordesman and Wagner, 1990: 10).
These figures by themselves cannot support the barbarism claim
because there could be other causes for the high casualties and refugees
(such as the complete lack of organized medical facilities, disease or
age, or a naturally fearful population), but they are still suggestive of an
unusually high devastation rate among noncombatants.
Second, there is widespread agreement on the deliberate destruction
of Afghan infrastructure — especially buildings, orchards, and irrigation
systems — which cannot be attributed to collateral damage, because
such destruction followed the conclusion of offensive operations:
Soviet/Afghan forces stayed in the [Panjsher] valley until 10

September, when they withdrew to Rokka once more. Before they
did so, they set about destroying houses, irrigation systems and burn-
ing crops. A major refugee flow was triggered. (Urban, 1988: 109)
It is therefore difficult to consider this collateral damage, since the

term is invariably used to refer to damage (a) not intended by the
aggressor, and (b) incidental to an attack against a legitimate target
(allowing that the definition of ‘‘legitimate’’ sometimes varies during
the course of a conflict).
Third, there are questions concerning the use of mines, chemical
weapons, and the treatment of prisoners. As noted above, the Soviets
deployed a wide variety of mines as part of their cordon-and-search
tactic. The most common (and infamous) such mine is a tiny anti-
personnel mine designated PFM-1, which Western sources referred to
as a ‘‘butterfly’’ mine. But such mines were not only used to support
direct operations, they were also dropped across wide areas along
suspected guerrilla supply transit routes, including also fields and
pastures. Not surprisingly, the majority of victims were noncombat-
ants. As to the use of chemical weapons, there is strong evidence that
the Soviets experimented with a variety of chemical agents in
Afghanistan (Isby, 1989: 76; Cordesman and Wagner, 1990: 214—218;
Kakar, 1995: 215). Yet while the issue was and remains a highly politi-
cized one, and though no hard evidence was obtained to confirm the
use of chemical agents (Urban, 1988: 219; Rais, 1994: 106—108), the
182
balance of evidence suggests they were used as a counterforce weapon

against isolated mujahideen strong points.19 Finally, Soviet troops
rarely if ever took prisoners (Kakar, 1995: 246).
In sum, although many of the atrocities cited by witnesses, journal-
ists, and historians were those normally incidental to war, the weight of
evidence suggests that the main causes of Afghanistan’s high noncom-
batant casualty and refugee rates were due to the systematic and delib-
erate targeting of noncombatants and their food, water, and shelter.
These attacks were intended to weaken the mujahideen by disrupting
their logistical and intelligence base among the broader civilian
population.
The year 1983 marked a lull in the fighting in Afghanistan. Massud
signed an armistice with the Soviets in January, and each side paused to
lick its wounds. The Soviets undertook very few sweep operations that
year, and instead increased their use of air power (again, the dispropor-
tionate impact fell on civilians, not mujahideen). They also rotated,
retrained, and resupplied their troops with better weapons; including
the lighter AK-74 assault rifle, flak jackets, cluster bombs (RBK-250), a
new 30 mm automatic grenade-launcher (AGS-17), an improved infan-
try fighting vehicle (the BMP-2, whose 30 mm cannon could elevate
50˚), and RPG-18 rockets.
After the Massud cease-fire expired in 1984, the Soviets launched
Panjsher VII in April and May. Warned of the attack, Massud launched
a pre-emptive strike on support convoys traveling the Salang highway.
His mujahideen destroyed many vehicles, but Soviet response time
caught them by surprise and many were surrounded and killed by
heliborne air assault troops. Massud launched another pre-emptive
raid on April 20, but that same day the Soviets began high-altitude
carpet bombing of the valley and mountains. Unnerved, Massud
ordered his remaining forces to withdraw to side canyons and avoid
19
Kakar’s account contains a number of similarities to those of Coffey and others
concerning the Italian use of mustard gas during its conquest of Ethiopia in 1935 (see
Chapter 5): ‘‘The Soviets used chemical agents in inaccessible areas so that others might
not know about it. For this reason, the Soviets and the regime wreaked havoc by helicopter
gunships on areas where the presence of foreigners was suspected. Apart from other
considerations, the Soviets feared the foreigners would inform the world about their use
of chemical agents in Afghanistan. They bombed the few health centers set up in certain
areas by French and other physicians. The symbol of the International Committee of the
Red Cross was anathema to the Soviets’’ (Kakar, 1995: 246). If true, the Soviet use of
chemical weapons would share the same rationale. Their method of covering it up — the
intimidation of foreign aid workers — would also be identical. Only the scale of chemical
weapons use would differ.
183
contact with the mechanized forces advancing along the valley floor.
But this time the Soviet operational plan included coordinated attacks
up the canyons. These blocking forces had been specially deployed
from opposite sides of these canyons so as to trap the mujahideen
between two heavily armed and air-supported forces. This plan, and
the new Soviet combined-arms tactics were taking a much heavier toll
than usual (Urban, 1988: 147). After the battle, both sides claimed
victory, although the Soviets clearly had the better claim: they had
captured an important mujahideen leader and established a new series
of fortified posts throughout key areas of the valley. They killed many
more mujahideen than in any previous punitive offensive, and were
later able to claim that civilians in the area were able to resume ‘‘a
normal life’’ (Urban, 1988: 148). On the other hand, the mujahideen
could claim victory because (a) they had managed to evacuate non-
combatants before the bombing; (b) they had not been destroyed as a
fighting force and they had shot down many Soviet helicopters which
had been attempting to support side-canyon clearing operations; and
(c) the Soviet decision to build fortified posts meant their job would
soon be even easier with more attractive targets close to home.
This last point proved especially important, because although the
Soviets viewed these fortified posts as a clear threat to the mujahideen,
they soon proved otherwise.20 Once built, such forts had to be manned
and constantly supplied with replacements, food, fuel, and ammuni-
tion. This meant a much smaller distance for the guerrillas to travel in
order to impose costs on the Soviets by ambushing supply convoys.21
20
Had the Soviets read their history books they’d have understood why. The French tried
the same tactic in Algeria in the 1830s against Abd-el-Kader (Asprey, 1994: 97), and later in
Vietnam in the 1950s. It was a disaster both times. The British were only able to make it
work during the South African War because they could use the entire black African
population of South Africa as a strategic resource. There are no other examples of the
successful use of a fortified line or blockhouse strategy.
21
In Seven Pillars of Wisdom, T.E. Lawrence describes the advantages of keeping the Turks
in Medina, and the terminus of a long and vulnerable line of communications: ‘‘One
afternoon I woke from a hot sleep, running with sweat and pricking with flies, and
wondered what on earth was the good of Medina to us? Its harmfulness had been patent
when we were at Yenbo and the Turks in it were going to Mecca: but we had changed all
that by our march to Wejh. Today we were blockading the railway, and they only
defending it. The garrison of Medina, reduced to an inoffensive size, were sitting in
trenches destroying their own power of movement by eating the transport they could
no longer feed. We had taken away their power to harm us, and yet wanted to take away
their town . . . What on earth did we want it for?’’ (Lawrence, 1926: 189). In a sense, what
the Soviets had unwittingly done was agree to build small ‘‘Medinas’’ all along the
Panjsher valley floor.
184
The personnel inside these posts could not exert anything other than a
negative influence on such civilians as remained in the valley:
The militia posts were also unable to influence the districts where they
were stationed. Their presence in the midst of the hostile rural people
was merely an odious symbol of the regime. When the mujahideen
attacked that symbol, the militiamen played havoc with their guns
on the villages . . . [Villagers] begged the mujahideen to leave their
villages or not to fire at the posts. A rift was thus created between
the villagers and the mujahideen. This was a victory for the regime.
A network of military posts throughout the country would have
enabled the regime to pacify the land, but the government was, of
course, unable to create such a system. (Kakar, 1995: 174)
Note that such a network was precisely what the British had been able
to create, maintain, and expand in South Africa. But the Soviets simply
did not have the mission (or the resources proportional to it) to build
such a network. Again, T. E. Lawrence’s thinking on the subject is illus-
trative. Here he calculates what it would take for the Turks to control the
100,000 square miles of territory ranging from the Hejaz to Syria:
Then I figured out how many men they would need to sit on all this
ground, to save it from our attack-in-depth, sedition putting up her
head in every unoccupied one of those hundred thousand square
miles. I knew the Turkish Army exactly, and even allowing for their
recent extension of faculty by aeroplanes and guns and armoured
trains . . . still it seemed they would have need of a fortified post
every four square miles, and a post could not be less than twenty
men. If so, they would need six hundred thousand men to meet the
ill wills of all the Arab peoples, combined with the active hostility of a
few zealots. (Lawrence, 1926: 192—193)
This then establishes the limited mission of the limited contingent as

a major explanatory factor in the strategy the Soviets pursued.22 The
territory of Afghanistan is 245,000 square miles; and even acknow-
ledging that the Soviets only needed to control some lesser portion of
that, they never had sufficient troops with which to do it (Urban, 1988:
119—120). In fact, the Soviets were so thinly spread in Afghanistan that
Massud could use threats to these posts to manipulate the pace of
operations in other provinces.
22
Kakar goes so far as to suggest a direct connection between the barbarism strategy and
the limited number of Soviet troops assigned to the theater (see below).
185
Little would change from 1985 until the Soviet withdrawal in 1989. In
1985, Mikhail Gorbachev came to power and in May, Babrak Karmal
was replaced by Mohammed Najibullah, who had been head of the
DRA secret police (KhAD) under Karmal. Although Gorbachev clearly
did not share the views of his predecessors regarding the costs and
benefits of the Afghan adventure, he gave his newly appointed
Southern Front commander, Mikhail Zaitsev, one year to engineer a
military solution to the Afghan problem. The operations of 1985 accel-
erated the Soviets’ barbarism strategy:
Soviet artillery and rocket-launchers supplemented aircraft to achieve

de-population through firepower, an approach mandated by the con-
tinued weakness of Kabul regime forces . . . The destruction of rural
agriculture was the goal of offensives in Laghman and in the Helmand
Valley. (Isby, 1989: 34)
But, after three years, the Soviets had clearly begun to hit the flat
of the curve in terms of the military effectiveness of their depopu-
lation strategy. Zaitsev intensified the use of air power and especially
heliborne-supported operations, but the mujahideen refused to unra-
vel. Why?
In 1986, Soviet operations focused more on interdiction of mujahid-
een supplies. Having destroyed almost entirely the mujahideen supply
infrastructure in Afghanistan through the cumulative effects of four
years of barbarism, the mujahideen had by this time become almost
entirely dependent on logistical support from Iran, Saudi Arabia, and
the United States (Rais, 1994: 112—113). This support arrived mainly at
Karachi, and flowed from there north through Peshawar and Quetta to
mujahideen forces throughout Afghanistan. In 1985 the SA-7 (shoulder-
fired surface-to-air missiles, or SAMs) used by the mujahideen with
moderate effect were supplemented by British-made Blowpipe SAMs.
In 1986, the mujahideen began to receive the much more effective
US-made Stinger SAMs. This prompted the Soviets to increase their
pressure on Pakistan. They did this by means of a limited number of
minor raids on mujahideen bases in Pakistan, and by attempting to
bribe non-Pushtun tribes along the Pakistani side of the border to
interdict mujahideen convoys into Afghanistan. The effort proved an
expensive failure.
But the biggest problem of 1986 proved to be the Stingers. In a major
offensive in the Nangrahar province, ‘‘three out of four Soviet helicop-
ters in a formation were destroyed in quick succession by US-made
186
heat-seeking man-portable Stinger [SAMs]’’ (Isby, 1989: 38). The

increase in attrition of Soviet helicopters had a disproportionate impact
on its military effectiveness, because for the prior six years the entire
force structure and organization of the Soviets in Afghanistan had come
to rely almost entirely on air-mobile tactics and weapons.23 In 1987
things only got worse:
[The Stinger] was the most effective [weapon] which the mujahideen
received . . . It became ‘‘a turning point of the campaign.’’ From then
on Stingers partly neutralized Soviet aerial offensives. According to
the estimates of Pakistan’s Intelligence Service (ISI), ‘‘During the sum-
mer of 1987 the mujahideen hit an average of 1.5 aircraft of varied
description every day.’’ By the end of 1987 the military situation had
deteriorated to the extent that even Najibullah admitted that ‘‘80
percent of the countryside and 40 percent of towns were outside the
control of his government.’’ (Kakar, 1995: 260)
Zaitsev, in other words, had inherited a heliborne-dependent force

in 1985, and the arrival of Stingers in quantity in 1986 and especially
1987 made it difficult for him — given the constraint in overall forces
and one-year time deadline — to engineer a military victory using
that now highly specialized force. Gorbachev called a halt, and the
Soviets began the sequential withdrawal of their forces, which ended
as the last Soviet unit crossed the Oxus River into the USSR on February
15, 1989.
Outcome: mujahideen win

Although Urban’s book was published prior to the end of the war, he
proposes some useful victory conditions:
Victory in the war in Afghanistan will be defined as the circumstances
under which the Soviet army (or the bulk of it at least) leaves. If the
PDPA regime survives and continues a slow expansion of support,
then the Soviets will have won. If the regime collapses, even within a
year or two of the withdrawal, the USSR will be judged to have
experienced a humiliation on the scale of Vietnam.
(Urban, 1988: 221)
23
This is not the same thing as saying that technology won the war for the mujahideen.
What it did do was make it impossible for Zaitsev to meet his one-year deadline. Had he
been given more time, it’s possible the Soviets could have developed and deployed
effective countermeasures without escalating.
187
The regime collapsed within three years of the Soviet departure, so by

Urban’s criteria the Soviet intervention may be judged either a success,
or a humiliation somewhat less severe than the scale of Vietnam.
In March of 1989, the mujahideen began a series of offensives
between the Pakistani border and Jalalabad, which came under siege.
In April, the mujahideen began to form regular army units in prepar-
ation for drives on Kabul. They spent most of 1990 consolidating and
reorganizing. In March of 1991 they captured Khost, taking 2200 prison-
ers. In July, the Wakhan corridor fell, opening the road transport of
supplies from Pakistan to Badakhshan. In September, the United States
agreed to halt all military aid by the end of the year and, in November,
the Soviets begin withdrawing 300 SCUD missile operators from
Afghanistan. On December 8, the Soviet Union ceased to exist, and
Soviet assistance to the Najibullah regime terminated. In April of
1992, Najibullah appealed to the United States for help in stopping
the spread of fundamentalist Islam in Central Asia, and on April 12
the Salang highway fell to the mujahideen. Two days later, Massud and
his allies captured the airbase at Bagram. On April 18, Herat, Kunduz,
and Shindand air base fell. Two days later, Najibullah was stopped
at Kabul airport as he attempted to board a plane to join his family
in Delhi. He took refuge in the UN compound.24 The mujahideen
continued to close in on Kabul, which they occupied on April 27,
1992, the 14th anniversary of the coup which had toppled Da’ud.
The costs to each side in terms of casualties are not entirely clear.
Cordesman and Wagner estimate (Cordesman and Wagner, 1990: 10;
see also Isby, 1988: 62) that the Soviets lost no more that 15,000 men
killed and about three times that many wounded.25 DRA losses are
cited as 34,000—42,000 killed or wounded, and 52,000—60,000 desertions.
The mujahideen lost about 140,000—200,000 killed or wounded. Fully
half the civilian population of Afghanistan was either killed or made
refugees (about 1,500,000 and 6,000,000 respectively).
The Soviets did not achieve their political objectives in Afghanistan.
They were not able to secure the DRA from attacks even in Kabul; and
the DRA itself could never gain enough legitimacy among the Afghan
people to mount a political challenge to the authority of the
24
Najibullah remained at the UN compound until the Taliban — a group of conservative
Sunni Islamists trained and supplied by Pakistan — captured Kabul in September of 1996,
when he was taken from the compound and hanged.
25
Borovik reports that this is a gross underestimate: according to his research, there were
already 20,000 Soviet casualties by 1981 alone (Borovik, 1990: 281).
188
mujahideen. Even more importantly, within a year, the Taliban — a

group of Islamic extremists and the very type of regime which the
Soviets most feared would come to power in Afghanistan — became
the de facto government of all but a tiny fraction of territory held by the
Northern Alliance.
Analysis: competing explanations of the Afghan

Civil War’s outcome
Actor interests
In the Afghan Civil War the USSR’s interests were not explained by its
material preponderance relative to that of the mujahideen. The USSR
invaded Afghanistan with the minimal objective of establishing and
securing a friendly Marxist regime in Kabul. They anticipated that once
stabilized, the Karmal regime would be able to successively broaden
support (by jailing extremists and initiating reforms and development
projects) until the Soviet military presence was no longer necessary. But
their presence galvanized opposition and instantly delegitimized the
Karmal regime beyond any hope of salvation. Why did they stay after it
became apparent that their original assumptions were wrong? More
importantly, why didn’t they escalate, as the United States had done in
Vietnam?
They stayed because they overestimated their own military capabil-
ities. They also overestimated the degree to which the DRA could
operate independently of Soviet support, and the degree to which the
Afghan people would consider the DRA to be independent. They also
stayed because they not only feared to suffer the humiliation of the
defeat of a friendly Marxist regime on their borders,26 but because they
feared its replacement by an Islamic fundamentalist regime in Kabul.27
The Soviets did not escalate because they believed they could innov-
ate — strategically, tactically, technologically, and politically — new
ways of increasing their military effectiveness without increasing
their costs in terms of blood and time. Their depopulation strategy,
26
A question of negative precedent setting: if the USSR could not defend a friendly and
proximate Marxist regime from a shaggy, fundamentalist rabble, how could other, more
distant allies count on it?
27
Events subsequent to their departure appear to bear this concern out: two days after the
collapse of the Najibullah regime in Kabul, civil war erupted in the former Soviet Republic
of Tajikistan.
189
heliborne-assisted cordon-and-search tactics, new equipment, and

reorganization, training and supply of DRA units are all evidence of
this.
Their decision to stay implies that Soviet interests in Afghanistan
were higher than those suggested by the interest asymmetry argument,
yet their refusal to escalate suggests that their interests in Afghanistan
were in fact limited — a term that applies equally to their forces and their
objectives in Afghanistan. Which is it?
There is no definitive way to answer this question, but the best evi-
dence supports the argument that whatever their initial objectives the
fact of the war and mujahideen resistance increased the importance of
the outcome for the Soviets far beyond that predicted by the interest
asymmetry argument. The Soviets did not leave until after they had
killed or displaced half of Afghanistan’s civilian population, and gained
the promise of a dramatic reduction of foreign support for the mujahi-
deen. Meanwhile, they continued to pour billions of rubles in develop-
ment and military aid into Afghanistan, and Najibullah’s regime was
able to survive until the collapse of the Soviet Union itself in 1991. Put
more bluntly, the real reason the Soviets did not escalate is because they
believed they had achieved their limited objectives in Afghanistan.
The mujahideen fought as though their very survival was at stake,
and the extreme brutality of Soviet and DRA operations in Afghanistan
make it clear that they were correct. Once the Soviets withdrew, how-
ever, the mujahideen began to fight among themselves. This only accel-
erated once the government which had been imposed on the Afghan
people by the Soviet Union collapsed in 1992.

In 1979 the Soviet Union was an authoritarian regime under the leader-
ship of Leonid Brezhnev. Even after the accession to power of Mikhail
Gorbachev in 1985, his policy of glasnost was not applied to the war in
Afghanistan:
Much of what we try so desperately to conceal, the Americans know

better than the Soviets. If I try to publish some information that we
consider top secret, but that appeared in the American press a long
time ago, the censor will still cross it out. An obvious question presents
itself: once the American people know something about the Soviet
Army, why don’t the Soviet people have the right to know it, too?
(Borovik, 1990: 127)
190
Its press was tightly controlled, its soldiers’ letters were censored,
and its people had no say in its foreign policy.28 Its initial conventional
attack strategy in Afghanistan was based on its mission, its force struc-
ture, and its assumptions about the quantity and quality of the resis-
tance it expected to encounter. The third-echelon mechanized forces
which swept into Afghanistan on December 27, 1979 did not expect
resistance, and they encountered very little until after they had settled
into garrison. Their initial punitive offensives into major Afghan valley
systems were little more than a show of force intended to cow the
‘‘backward reactionary peasants’’ who opposed the efforts of the
PDPA to advance the standard of socialism.
But the Afghans were not cowed, they were outraged.
The conventional attack strategy initially thought sufficient to inti-
midate the Afghans soon changed to a COIN campaign which is best
understood in Soviet historical context:
In the case of Afghanistan, one may not rule out a centrally-organized
strategy aimed at what Louis Dupree described as ‘‘migratory geno-
cide.’’ The transfer of population was a part of the Soviet counter-
insurgency doctrine which had been practiced before in its wars in
Central Asia. An effective counter-insurgency war could not be fought
against guerrillas swimming in a sea of a supportive population. The
Kabul regime, having failed to win the hearts and minds of the
Afghans, collaborated with the Soviet forces in depopulating areas of
tough resistance. (Rais, 1994: 102—103)
This is another way of saying that barbarism is a traditional COIN

strategy in Soviet military doctrine. Since the Soviet Union’s regime
type did not change from 1917 until its collapse in 1991, it is difficult to
evaluate the impact of regime type on the decision to switch to a
barbarism strategy in the Soviet case.
Difficult, but not impossible; there are two ways to understand the
relationship of regime type to strategy in the Afghan Civil War. First, the
interstate dimension was clearly a concern even for Brezhnev. Satellite
and signals intelligence data would quickly make both the build-up and
deployment of troops in Afghanistan evident to the world. But Brezhnev
calculated that just as it had in Hungary in 1956 and in Czechoslovakia in
1968, international public opinion would soon ‘‘blow over’’:
28
Soldiers’ letters were mostly self-censored: soldiers knew their families would be
worried, and most often wrote ‘‘encouraging’’ letters home so as not to worry their parents
(Tamarov, 1992).
191
This Soviet expectation eroded gradually as it became necessary to

increase the numbers of their ‘‘limited contingent’’ and to play the
leading role in fighting as the effectiveness of the Afghan military
declined precipitously as a result of increased desertions and internal
conflict between the overwhelmingly Khalqi military and the new
Parchami regime. Simultaneously, the transformation of the war
from a civil conflict to an international one aroused the Afghan people
to increased resistance, which gained more effective international
support. The reality of war, symbolized to the world by the dramatic
increase in refugees, made it impossible for the Soviets to practice
political ‘‘damage control.’’ (Magnus & Naby, 1998: 129)
Brezhnev was wrong, and he left his successors to pay the price.
Admittedly, this price was not prohibitive at first. The United States
led a boycott of the 1980 Olympic Games in Moscow, and the Carter
administration embargoed wheat to the USSR. But with the accession to
power of Mikhail Gorbachev, the closest thing to a regime change in the
USSR since Khrushchev’s denunciation of Stalin soon made the Afghan
adventure prohibitively expensive.
True, the Soviet people still did not have a direct say in foreign affairs,
and they were never told the real casualty figures in Afghanistan, but
for a new leadership attempting to bring the Soviet Union into the
twenty-first century intact, Afghanistan had become an albatross:
The war in Afghanistan . . . did not fit into Gorbachev’s overall policy
of glasnost and perestroika. He felt that it had turned into a debacle for
the Soviet Union and was too closely associated with the policies of the
Brezhnev era and the renewed Cold War. Glasnost meant that it was no
longer possible to keep the price of the war a secret . . .
(Magnus and Naby, 1998: 132)
Gorbachev was attempting to harmonize what his predecessors had

always considered to be mutually exclusive aims: nationalism, liberal-
ism, and socialism. Like Afghanistan’s Mahmud Beg Tarzi, Gorbachev
had become his own country’s ‘‘greatest modern liberal thinker.’’
Unlike Tarzi, however, Gorbachev was in charge.
In sum, the long history of authoritarian regime-types in the Soviet
Union had made it possible to prosecute a barbarism strategy without
sustaining domestic political costs.29 Barbarism had become its
29
Interstate military intervention such as that of NATO in Kosovo in 1999 was then, as
now, out of the question: Russia is still a big country and one still armed with thermo-
nuclear missiles besides.
192
standard COIN strategy. But communications and military intelligence

technology and the accession to power of a reform-minded general
secretary changed this. The Soviet barbarism strategy began to impose
international political costs almost from its outset. Domestic costs were
experienced within the USSR because many Soviet boys returned home
in zinc coffins, while many others returned shattered by mines or by the
experience of combat. These costs were not experienced by the Soviet
people as a whole until after 1986, when glasnost made it possible to
share more stories about the Afghan campaign and its costs.30
As to regime type and political vulnerability, the Soviet regime was at
no time vulnerable to the sorts of political costs anticipated by the
interest asymmetry argument. If the trade-offs mechanism anticipated
by Mack’s argument were operating, we’d expect to see major battle-
field defeats result in either pressure to withdraw or escalation. Yet
Soviet commitment in terms of men and materiel remained constant
over time. Moreover, even glasnost and perestroika combined did not
create conditions which made possible a public outcry on the war — at
least not one of a scale sufficient to force the regime to reconsider its
foreign policy.
Arms diffusion
In this case the arms diffusion argument gets its strongest support. The
mujahideen gained increasing advantages from arms they received
through Iran and Pakistan, and these increased the costs of conquest
and occupation to the Soviet Union and its DRA client. What is most
important to recognize, however, is that in no theater of war has the
crucial relationship between technology, tactics, climate, and terrain
been made clearer than in Afghanistan.
Soviet motorized infantry and armored columns were very often
disabled with technology as simple as pushing boulders onto roads
30
In fairness, however, the immense pride (in many cases deserved) and propaganda
surrounding the Soviet victory in World War II (what the Soviets called ‘‘The Great
Patriotic War’’), and the officially released casualty figures from the war, made it difficult
for many average Soviet citizens to appreciate the sacrifices of their Afghan veterans.
Vladislav Tamarov, a Russian survivor of the Afghan war, notes that ‘‘In the United States
there are 186 psychological rehabilitation centers open to help Vietnam veterans. But
where are we in the Soviet Union to go for help? We don’t even have one such center.
And so we look for that kind of help from people. That is when we run up against
misunderstandings. From these misunderstandings comes the high divorce rate among
Afghan vets, from these misunderstandings comes the turning inward, into oneself’’
(Tamarov, 1992: 7).
193
and then tossing Molotov cocktails onto the engine compartments of

immobilized vehicles. Then simple machine-gun fire — a technology
dating in its most serious use from World War I — would finish the job of
the convoy’s complete destruction. Essentially, the mujahideen used
simple technology to exploit weaknesses in Soviet tactics and strategy.
This was especially the case after 1986, when the arrival of US-made
Stinger SAMs dramatically shifted the balance of power in Afghanistan.
As observed above, it was not so much the Stingers themselves that
mattered, but rather the context: the Soviets had used heliborne infan-
try and helicopter gunships to exploit a weakness in mujahideen tactics,
only to find the tables turned once their helicopters fell prey to the
Stingers. There is no question that had the USSR been ruled by a Stalin
or even a Brezhnev, the Soviets in Afghanistan would have innovated
around this weakness. But Gorbachev called a halt to the war, wisely
reckoning that to win it would gain the USSR little, and might require
the invasion of Pakistan.
The Afghan Civil War had two strategic interactions, and during the
first the Soviet pursued a particularly blunt conventional attack strat-
egy against a mujahideen GWS. Neither side had adequate leadership,
training, or equipment, but Soviet forces were so thoroughly unsuited
to a COIN mission that their failures must be attributed almost entirely
to strategic interaction. They swung a blunt club, and the mujahideen
ducked and stabbed them in the foot with a sharp stick. The strategic
interaction was opposite-approach and the Soviets were forced to
reorganize and switch strategies.
But the second interaction of the war turned far more deadly for the
mujahideen. The Soviets consciously and deliberately targeted non-
combatants and their support infrastructure as a COIN strategy.
According to the strategic interaction thesis, barbarism should have
been the ideal counter-GWS strategy, quickly destroying the mujahi-
deen as a fighting force. It did not do this, but the reason it did not does
not refute the strategic interaction thesis. Instead, it highlights a specific
condition under which the dynamic does not apply.
The very real destruction of Afghan infrastructure and the mass
killing and forced emigration of peasants did hurt the mujahideen,
but most, like Massud’s Panjsher fighters, managed to reorganize them-
selves and their resources to compensate. After 1983 they began to rely
more for intelligence on sympathizers within the DRA, and more on
194
logistical support from foreign sympathizers.31 They were still able to

mount offensive operations, and Massud was able to do so farther and
farther from his home bases of support (a sign that they mattered less as
the war dragged on). But this made them proportionately more depend-
ent on outside support. The degree of mujahideen dependence on out-
side support is revealed in the sharp decline of their fighting capacity
after the Soviet evacuation in 1989 (Cordesman and Wagner, 1990: 97;
Magnus and Naby, 1998: 134, 159): Not only had they fallen increasingly
to fighting among themselves, but the flow of arms quickly dried up.
In sum, the Afghan Civil War makes it clear that the more independ-
ent guerrillas are of their popular support base due to outside support,
the more insulated they will be from the devastating effects of barbar-
ism, assuming they are able to gain good intelligence. However, outside aid
poses its own risks. First, it may provoke guerrillas to attempt to defeat
the strong actor in decisive engagements (such as the Tet Offensive of
1968, which proved a military disaster). Second, if the contest is fought
on nationalistic terms, as it was in Afghanistan, foreign assistance —
especially military advisors — can be used to paint its recipients as the
lackeys of the foreign powers supplying the arms and other support.
The Karmal (and later Najibullah) regimes attempted and failed to tar
the Islamic mujahideen with the foreign lackey brush. They failed
because the United States did not send military advisors to
Afghanistan in quantity. But the mujahideen had no difficulty convin-
cing Afghan and international audiences alike that the PDPA regime
was the mere puppet of Soviet interests.
Even so, the operational impact of the Soviet barbarism strategy
should not be underestimated. These depredations hurt the mujahi-
deen, and while the most successful commanders adapted themselves
to the changes, others could not, and their forces were either killed,
captured, or dispersed.
Conclusion
The Afghan Civil War was an asymmetric conflict between the DRA and
its Soviet masters (strong actor), and the Afghan mujahideen (weak actor).
The Soviets had a number of interests in Afghanistan, and the two
most important were negative: (1) defending a friendly Marxist regime,
31
They were also able to rely on the military intelligence of the ISI, Pakistan’s highly-
rated intelligence service.
195
and (2) thereby preventing the accession of a radical Islamic republic on

its southern border. To these seemingly limited objectives it was willing
to commit a force of four MRDs, one-and-a-half Air Assault divisions,
and lavish economic aid. It did not act as if its survival was at stake, but
neither can its level of interest in Afghanistan be explained by its
relative power, as implied by the interest asymmetry argument.
Mujahideen interests were in ejecting the Soviets from Afghanistan,
where the PDPA regime which controlled Kabul was not distinguished
from the Soviets. To this end they fought as if their survival was at
stake.
interests in the outcome of an asymmetric conflict — receives moderate
support. It cannot be rejected, but neither can it be accepted as an
explanation of Soviet interests. Did relative power and regime type
explain political vulnerability?
According to the interest asymmetry thesis, a major military setback
should have caused the Soviets to reevaluate their occupation regard-
less of their regime type. This didn’t happen. Every year brought a mix
of successes and failures, with the successes short lived and the failures
at times catastrophic. Yet there is no evidence that the failures ever
provoked the sort of trade-offs anticipated by the interest asymmetry
argument. Even commanding generals were not cashiered after such
failures. Instead of withdrawing, the Soviets attempted gamely to learn
from their failures and innovate around them. They often succeeded,
but never in a way which could compensate for their relatively low
numbers. In short, relative power asymmetry did not cause political
vulnerability. Instead, the Soviet Union’s authoritarian regime type
insulated it from such vulnerabilities.
Regime type also made it possible for the Soviets to avoid the three
most common defects of a barbarism strategy. Authoritarian regimes
are first of all free to construct the enemy as they see fit. Second, and
more importantly, the lack of public access to accurate information
about the costs of a conflict short-circuits domestic unrest and, hence,
political vulnerability. Third and finally, authoritarian regimes do not
by definition contain a mechanism by which popular will can be trans-
lated into foreign policy. These are, of course, ideal-type descriptions
but, in the Soviet case during the Afghan Civil War, this is the way
things worked. Thus, the decision to switch from a direct attack to a
barbarism strategy was a doctrinaire response to a recognition that the
Soviet mission had changed to COIN.
196
The mujahideen did not have a regime type, and cannot be con-
sidered as even a remotely unified actor. On the contrary, some com-
mentators argue that this lack of unity actually proved to be an asset,
because it gave the Soviets no way to decapitate resistance leadership:
Among significant features of Pushtun life are the practice of Islam,
mainly in its Sunni aspect among nearly all Pushtuns, the nonhier-
archical structure of tribal groups, and the Pushtun code known as the
Pushtunwali. All three features have come to contribute significantly
to the persistence of the Afghan resistance to the Soviet army at first,
and then to the stubborn inability of the Pushtuns either to agree
among themselves on how a new government should be formed or
to work with like-minded ideological groups to end the civil war.
(Magnus and Naby, 1998: 14)
Others emphasize the problems of lack of unity and hierarchy:

Some analysts have said that this disorganization is a strength,
because it means that the resistance cannot be decapitated. It is noth-
ing of the sort: without unity and organisation they cannot take the
initiative. If they cannot take the initiative, victory is impossible.
(Urban, 1988: 216—217)
In the context of GWS, Urban is simply wrong. The mujahideen could

win by simply not losing, and having no hierarchical structure only
made it more difficult for the Soviets to defeat them.32
Arms diffusion also receives some support in this case. But a close
examination of how arms technology interacted with tactics and strat-
egy makes it clear that the logic of the arms diffusion argument,
hypothesis 6 (the better armed a weak actor is, the more likely it is
that a strong actor will lose an asymmetric conflict), is only weakly
supported here.
Strategic interaction explains why the Soviets lost the first interaction
of the war, but it doesn’t explain the second. From 1980 to 1982, the
Soviet pursued a conventional attack strategy against a mujahideen
GWS (opposite-approach). It failed on every count. Yet after the switch
to a barbarism strategy in 1983 (same-approach), the USSR did not win,
and in 1989 it withdrew from Afghanistan completely. Why?
Barbarism should have been an ideal COIN strategy because the
Soviets were not vulnerable to domestic or interstate political costs
32
Contrast this with the Algerian resistance against France from 1958 to 1962. The FLN
did have a hierarchical structure and the French were able to exploit this by means of,
among other things, torture.
197
associated with its prosecution.33 Moreover, it should have devastated

the intelligence and logistical capacity of the mujahideen. Instead, it
only provided the mujahideen with defectors and recruits. Barbarism
wasn’t decisive in Afghanistan because the mujahideen were able to
rely on intelligence and logistical support from foreign powers from the
sanctuary of Pakistan and Iran. The mujahideen had in fact become so
dependent on this aid that when it ceased in 1989 and 1990 it shattered
their ability to topple the illegitimate government of Najibullah. Not
until the logistical supply tap was turned back on — this time for the
Taliban — did that change. Afghanistan therefore highlights one condi-
tion under which barbarism is less effective as a COIN strategy: when
the guerrillas have access to a significant source of intelligence and
logistical support (and sanctuary) beyond the reach of the strong
actor, and beyond its capacity to interdict.34
In sum, strategic interaction best explains the outcome of Soviet
intervention in the Afghan Civil War. Once it became clear that the
fight in Afghanistan was a COIN struggle, the Soviets had three stra-
tegic choices. First, continue with the present strategy, perhaps sending
more troops and equipment to the theater. Second, withdraw in some
way which could save face yet end the war. Third, switch to a barbarism
strategy. The Soviets chose the third option and their strategy suc-
ceeded in hurting the mujahideen and killing or making refugees of
nearly half of Afghanistan’s pre-war civilian population. The mujahi-
deen faced an even starker choice after the Soviets switched strategies:
either give up or continue with a strategy which would force them to
rely more and more on outside powers for support.
33
This contrasts with the Italian case in Chapter 5, because Italy was a minor power and
even the slightest real sanction by Britain (such as closing the Suez Canal to Italian
transport) or France would have been sufficient to force Italy to withdraw.
34
The same proved true of US experience in Vietnam, with Laos and Cambodia playing
unwilling host to DRV supply transits. Unlike the USSR, the United States was initially
less circumspect about violating the neutrality of Laos and Cambodia in its efforts to
interdict supplies flowing from the DRV to the VC in the South. But then neither Laos nor
Cambodia were nuclear powers or strong allies of the United States, whereas from 1979 to
1989 Pakistan was both. The KLA in Kosovo in 1999 also serves as an example of this
problem. Serb barbarism rapidly depopulated Kosovo of ethnic Albanians, and eviscer-
ated what had been a relatively incompetent (though impassioned) KLA resistance. But,
unless willing to invade Albania, the Serbs could never eradicate the KLA, and in the
event Serb barbarism only served to prompt NATO intervention and keep NATO united
in its efforts to punish Milosevic. On the relationship between Serb and Serb-supported
barbarism in Kosovo and NATO intervention, see, e.g., Daalder and O’Hanlon (2000).
198
In the end, no one really won the war. The Soviets left in 1989 and the
DRA fell in 1992. The mujahideen were almost entirely overcome by the
Taliban — a group of extremely conservative Islamists supported by
both Pakistan and Saudi Arabia. Afghanistan itself has been devastated
on a scale not witnessed since the destruction of Germany and the
Soviet Union in World War II.
199
8 Conclusion
The vast majority of wars do not come into the heavyweight range,
but are distinguished more by their duration and bitterness than their
weaponry. In a way they become more amenable to Western inter-
vention when they do develop into straightforward clashes bet-
ween regular forces. Civil wars, involving irregular fighters and
skirmishes in the streets, with political confusion rife and good
intelligence at a premium, present an appalling prospect to outsiders.
Decisive victories are few and far between. Even success can mean a
long-term commitment of troops to sustain an uneasy peace.
Freedman
This book began with a puzzle. How do the weak win wars? Through a
combination of statistical tests and the tracing of causal logic through
historical case studies, I have shown that weak actors — in this case
mainly states — win wars against much stronger adversaries when they
are able to adopt and maintain an ideal counterstrategy. Strategy, in
other words, can multiply or divide applied power.
Strong actors come to a fight with a complex combination of interests,
forces, doctrine, military technology, and political objectives, but because
armed forces are thought to be versatile in their employment, and because
strong actors are only relatively, not absolutely, strong, strong actors do
have choices in the strategies they use. Similarly, weak actors often face
constraints in their choice of strategies, but strategy is never endogenous.
The dynamics of asymmetric conflict

International relations theory leads us to expect few asymmetric con-
flicts because large gaps in relative power imply unambiguous con-
flict outcomes. The stronger you are, the less likely I am to challenge
200
Conclusion
you.1 The weaker you are, the more likely I am to provoke you,
knowing you wouldn’t dare object. More importantly, the greater
the disparity in power, the more quickly we expect the strong to
subdue the weak if it comes to blows.
In this book I examined three arguments that had the potential to
explain why strong actors lose to weak actors; using the puzzle of increas-
ing strong actor failures over time as a kind of test of the arguments’
soundness and generality. The first and strongest of these alternative
explanations was Andrew Mack’s interest asymmetry thesis, but nature-
of-actor (including Merom’s ‘‘democratic social squeamishness’’ argu-
ment), and arms diffusion arguments were also introduced and tested.
Competing explanations of asymmetric conflict

outcomes
The best alternative general explanation of asymmetric conflict out-
comes over time is Andrew Mack’s interest asymmetry thesis. Mack’s
explanation reduces to a simple explanatory model. He argues that
structure explains actor interests, which in turn determine relative
political vulnerabilities, which explain outcomes. A strong actor’s sur-
vival will not be at stake, so its interests — understood here to mean
willingness to sustain costs — will be relatively low. This vulnerability
makes it much more likely that a strong actor will be forced by domestic
political opposition or elite rivalry to withdraw from a conflict if it
suffers unanticipated costs or a military setback. This explains why
‘‘big nations lose small wars.’’
A second argument was that perhaps strong actors with democratic
regime types lost asymmetric conflicts because they were weak willed,
or too casualty sensitive, or too vulnerable to domestic criticism;
whereas authoritarian strong actors were more likely to win because
they could field more ruthless or efficient militaries, and were insulated
from domestic criticism, making it possible, in Merom’s words, to
1
It is not only the force applied that causes this effect, it is more importantly the force
available to be applied that matters most. This is why it makes sense to argue that the
dynamics of asymmetric conflict apply even though so-called strong actors are often not
overwhelmingly strong within a given conflict theater. Strong actors have choices about
their other commitments and the relative priorities of their interests. Weak actors under-
stand this, and can never rely on the existence of other commitments as a guarantee
against future strong-actor escalation. On this point see Schelling (1966: 2—6).
201
‘‘escalate the level of violence and brutality necessary’’ in order to win

(Merom, 2003: 15).
A third argument was that the diffusion of relatively advanced small
arms and other military technology to the developing world in and of
itself made conquest and occupation of weak actors more difficult;
essentially, because like strategy, technology can also be a force multi-
plier, weak actors were not really as weak as anticipated.
Strategy and strategic interaction

Mack was right about the problem of vulnerability, but his argument
overemphasizes the link between relative power and interests, and fails
to identify the key permissive condition under which such vulnerabil-
ity operates: a significant and unanticipated delay between the commit-
ment of armed forces and the attainment of pre-war military or political
objectives. If the conflict is over quickly — which is what we expect
in very asymmetric conflicts — then how can political vulnerability
operate?2
This leads to the question of why some asymmetric conflicts — such as
the USSR’s invasion of Hungary in 1956, or the US-led Gulf Coalition in
Iraq in 1991 — are quick and decisive, while others — such as the US
intervention in Vietnam in 1965, or the USSR’s in Afghanistan in 1979 —
drag on? This book has shown that the interaction of the strategies
actors employ predicts the duration of the conflict, and provides the
missing condition necessary for the operation of political vulnerability.
The importance of strategy as the key variable intervening between a
given actor’s material resources and outcomes requires us to consider
both the range of strategies available to actors and the interaction of
those choices in the conflicts themselves.
2
Mack and Merom each focus on cases featuring opposite-approach interactions, in
which a strong actor (advanced industrial country) using a conventional attack strategy
attacks a weak actor (economically backward nation in arms) using GWS or terrorism. The
strategic interaction thesis explains why strong-actor failure in these cases is overdeter-
mined. Mack offers the best general explanation of the two, erring only in overestimating
the degree to which actor motivation and regime type affects outcomes, and by omission
of the key variable (strategic interaction) that determines how long an asymmetric conflict
will last. Merom offers a more detailed account of how democratic strong actors lose, but
errs in his overestimation of the utility of brutality in winning small wars in the post-
World War II period. Democracies, for example, have won two small wars in the last half-
century without recourse to barbarism. Britain succeeded in the Malayan Emergency of
1948, and under the leadership of Ramon Magsaysay, the Philippines (with US support)
succeeded against the Hukbalahap in 1952.
202
Conclusion
Strategy
For purposes of this analysis I reduced a wide array of specific attacker
and defender strategies to four: two for attackers and two for defenders.
Attacker (strong actor) strategies are conventional attack and barbar-
ism. Defender (weak actor) strategies are conventional defense
and GWS.3
I then simplified each actor’s range of strategic options further into
two analytically distinct strategic approaches: direct and indirect. Direct
approaches — conventional attack and defense — target an adversary’s
armed forces with the aim of destroying or capturing that adversary’s
capacity to fight. Indirect approaches — barbarism and GWS — aim at
destroying an adversary’s will to fight.
These simplifications — and they are that — highlight what is most
important about the role of strategy in mediating between raw power
resources and asymmetric conflict outcomes: strategic interaction.
Once a strategy is chosen, is it the best strategy under the circum-
stances? In this context ‘‘circumstances’’ means given the other actor’s
strategy. Each strategy has an ideal counterstrategy. Actors can drama-
tically increase the effectiveness of their own strategy (essentially multi-
plying their forces) by guessing correctly about their adversary’s
strategy and then selecting and executing that ideal counterstrategy.4
Specifically, similar approaches (indirect—indirect, or direct—direct)
imply defeat for the weak actor and victory for the strong. These wars
will be over quickly, making political vulnerability (whether caused
3
These strategies are detailed in Chapter 2 . They are ideal-type constructions and by
no means exhaustive representatives of their approach categories (e.g., ‘‘conciliation’’
is another indirect-approach strategy available to strong actors). Each of these four
strategies (save perhaps barbarism) is represented by an extensive literature which need
not be reviewed here. I have instead simplified and fixed the meanings of the most
relevant strategic options, which actors may pursue independently, sequentially, or
simultaneously.
4
This said, executing an ideal strategy may be difficult for at least two reasons. First, a
given actor’s forces may have been trained, armed, and prepared for a different strategy
against a different enemy. Switching strategies — especially in the middle of a fight — can
therefore be risky. All other things being equal, it will be easier to switch to a different
strategy in the same approach (say, from terrorism to nonviolent resistance or GWS) than
to a different strategy in an opposite approach (say, from conventional defense to GWS).
Second, some of the strategies themselves cannot be quickly implemented; when guerrilla
warfare as a tactic is not supported by a previous period of social organization it tends to
fail. This was the experience of, e.g., Che Guevara in Bolivia.
203
by asymmetric interests, as argued by Mack, or by regime type, as

argued by Merom) irrelevant. By contrast, opposite approaches
(direct—indirect, indirect—direct) favor weak actors at the expense of
strong actors. They will drag on, forcing strong actors — especially
democratic strong actors — to make tough and costly decisions in
order to continue with any prospect of success.
My argument
Relative power, regime type, and political vulnerability are necessary
but not sufficient to explain variation in asymmetric conflict outcomes.
The strategies actors use are important (as are the constraints actors face
when evaluating competing offensive and defensive strategies), but
what is more important is how opposing strategies interact. I hypothe-
sized that same-approach strategic interactions would favor attackers
in proportion to their advantage in material resources. Opposite-
approach strategic interactions would favor defenders, regardless of
the attackers’ material preponderance.
Assuming the strong actor is in each case the attacker and the weak
actor the defender, the expected relationship of strategic interaction to
conflict outcomes can be seen in Figure 3 in Chapter 2. In same-
approach interactions the strong actor wins because there is nothing
to deflect or mediate the use of its material advantages in resources,
including soldiers and wealth. In opposite-approach interactions, the
strong actor’s resources are deflected (weak actors attempt to avoid
open confrontation contact with a strong actors’ armed forces) or dir-
ected at values which don’t necessarily affect the capacity of the weak
adversary to continue to impose costs on the strong actor (e.g., captur-
ing cities and towns).
Statistical evidence
The core claim of strategic interaction theory — hypothesis 5: strong
actors are more likely to win same-approach interactions and lose
opposite approach interactions — received strong statistical support.
For this relationship see Figure 4 in Chapter 2: Clearly, weak actors do
better when strategies are opposite than when they are similar. A strong
actor using a direct approach (say, a blitzkrieg or standard offensive
campaign employing infantry, artillery, armor, and motorized infantry)
is likely to lose against a weak actor employing a GWS, but likely to win
quickly against a weak actor employing a standard defense (such as
204
Conclusion
Taliban forces in Afghanistan in 2001). Weak actors were nearly three

times as likely to win opposite-approach interactions (63.6 percent) as
same-approach interactions (23.2 percent). Three other important pro-
positions were tested. First, I tested whether and to what degree exter-
nal noncombat support for weak actors might make their victory more
or less likely. Although a lack of available data made it impossible to
evaluate the positive value of this relationship, I was able to determine
that even when weak actors received no external support, they were
still more than twice as likely to win when strategies were opposite than
when they were similar. Second, I argued that the main reason opposite-
approach interactions favored weak actors is because time is an import-
ant cost for strong actors,5 and opposite-approach interactions take
longer to resolve themselves. A statistical review of all asymmetric
conflicts since 1816 strongly confirmed this proposition: opposite-
approach interactions take longer than same-approach interactions.
Finally, I explained the trend toward increasing strong actor failures
over time by recourse to a state socialization argument: actors — whether
states, boxers, terrorists, or firms — imitate the successful practices of
other actors.6 At a minimum then, the number of opposite-approach
asymmetric conflicts should increase over time in proportion to the
number of wars in which the weak won. They do. Thus, this book’s
statistical tests of the strategic interaction thesis strongly support its logic.
Evidence from cases

The causal logic of competing arguments was tested in five historical
case studies. The aim was to establish the relative impact of each actor’s
interests, regime type, weaponry, and strategy on the outcome.
The Murid War, 1830–59

The Murid War pitted the Russian Empire (strong actor) against a
coalition of Caucasian tribes under the banner of Muridism (weak
actor). Russia’s interests focused on protecting its communications
with Georgia. Murid interests were in establishing their political and
religious independence from Russia.
5
It is an important cost for both actors, but because their power advantage leads to the
expectation of quick victory, the cost of delays is effectively multiplied for strong actors.
6
Waltz has already been cited on this point. But see also, Porch (1996: xvii).
205
In terms of regime type and vulnerability, Russia simply was not

politically vulnerable. The lack of a serious security threat from the
Caucasus should have made even authoritarian Russia sensitive to
resource trade-offs, and politically vulnerable in this sense. Yet during
the Murid War, Russia faced two potentially serious interstate conflicts,
and neither caused political vulnerability. Regime type did affect the
strategies employed by each side. Russian strategy changed to suit each
new tsar’s preferences: sometimes scrupulously avoiding barbarism,
for example, and other times targeting noncombatants with vigor.
Shamil was just as autocratic about strategy. He was able to go against
his own soldiers’ training and inclinations in ordering a switch to a
conventional offensive in Kabardia.
The arms diffusion argument was not supported in the Murid War.
The Murids did not receive weapons or ammunition from abroad, and
the effectiveness of each side’s arms technology depended on the sys-
tem of its use, which in turn depended on variations in terrain and
climate. Heavy artillery proved vital to the Russian effort to conquer the
Murids, but was effective only under circumstances where Russian
forces were willing to ignore the staggering toll in casualties required
to maneuver the guns into position. The brief capture of artillery by the
Murids proved to be — contrary to expectation — a military disaster.
In terms of strategy, the war played itself out in three strategic
interactions. The Russians won the first interaction: barbarism against
a Murid GWS (same-approach).7 They lost the second interaction: a
Russian conventional attack strategy against a Murid GWS (opposite-
approach). The final interaction was again same-approach: a Russian
conciliation strategy opposed by a Murid GWS. The Russians won, and
Shamil went into exile.
The outcome of the Murid War was determined by Russia’s extreme —
almost historically unique — cost insensitivity.8 This insensitivity was
7
Since their aim was conquest and subjugation, rather than annihilation, the victory
would prove elusive. In theory, Russia’s willingness and ability to resort to barbarism
should have deterred subsequent Murid resistance. It did not. Instead, it only stimulated
and intensified resistance. If democracies lose small wars because they can’t escalate to the
level of brutality necessary to win, then Merom’s thesis will have a tough time explaining
why authoritarian regimes don’t win small wars more often than they do.
8
Recall, however, that Alexander did not approve of barbarism as a Russian military
strategy, however effective. This is worrisome for Merom’s model, which holds that in
general authoritarian regimes will be able to win small wars because they will escalate the
level of violence to barbarism. The Murid War shows that even in an autocratic regime,
‘‘normative difference’’ between state and society can exist and affect strategy (and, by
extension, costs and outcomes).
206
Conclusion
due to its regime type and to the fact that Russia’s public was almost
entirely illiterate. But the costs — half a million casualties and twenty-
nine years — are best explained by the strategic interaction thesis.
The South African War, 1899–1902

The South African War pitted Great Britain — then the world’s preemi-
nent great power — against two of the world’s smallest states — Orange
Free State and Transvaal (the weak actor).
British political elites calculated that control of Orange Free State and
Transvaal was vital for three reasons. First, they believed control of
Cape Town was key to maintaining sea communications with India and
their other colonial possessions. Second, allowing the Boer republics to
‘‘dictate’’ to the British might encourage the majority population of
Cape Colony — Afrikaners — to seize control of the Cape. Third and
finally, it might also make Britain appear weak-willed in relation to its
European great-power rivals, thus encouraging interference or inter-
vention from those rivals.9 Employing a domino-like logic throughout,
the British therefore calculated that the subjugation of Orange Free State
and Transvaal were vital interests. The Boer were interested in main-
taining their independence and their way of life (in particular, the
institution of slavery).
Two factors explain British political vulnerability in the South
African War: regime type and time to objective. As leaders of a demo-
cratic state, British policymakers could not make policy without taking
public reaction and judgments into account. In the South African War
there were two loci of vulnerability for Britain: jus ad bellum — the
justness of Britain’s resort to force in South Africa — and jus in bello —
the justness of its conduct in that war. Kruger’s ultimatum spared
British elites from difficulties over jus ad bellum: the British public
recognized the legitimacy of self defense as a justification for the resort
to arms. But what of the moral conduct of the war itself? The concen-
tration camp controversy very nearly forced Britain’s withdrawal from
the war short of achieving its political objectives.10 But had the war been
9
It might also set a bad precedent vis-à-vis other empire colonials, who might then think
of rising up to challenge British colonial rule.
10
The South African War must count as a problem for Merom’s thesis, because Britain
was a democracy, and yet had no difficulty escalating to the level of brutality (not to
mention its own costs in terms of blood and treasure) necessary to win the war. Merom’s
own account of the nature and consequences of British COIN strategy (especially under
207
over by Christmas or soon after — as widely anticipated by British elites

and the British public — neither concern would have mattered. It there-
fore makes sense to ask: what caused the delay in a war in which Great
Britain out-powered the Boer by a ratio of more than 5:1?
The delay was caused in part by the fact that the Boer possessed
modern rifles, and were skilled in their use. The British had never
fought a major engagement against an enemy so equipped, and it
took time for them to adapt, tactically, to the new situation. But the
real reason the war was not over by Christmas was because (1) the
British dramatically underestimated the Boer; (2) the Boer were much
more mobile; and (3) the Boer had better intelligence. But the battles
fought through 1900 made it clear that not even the possession of
superb infantry rifles, superior artillery, and highly mobile forces
could compensate for the overwhelming number of Britain’s forces in
South Africa. British forces simply steamrolled the Boer, moving in an
almost straight line from Cape Colony to Pretoria. The arms diffusion
argument therefore counts as a partial explanation of the South African
War’s outcome.
The real cause of the delay was a shift to a GWS by the Boer in March
of 1900. The British won the first interaction of the war (same-approach)
after overcoming their initial shock and after deploying sufficient
troops in theater. By February of 1900 the conventional war was over
and the Boer had lost. But after the fall of Bloemfontein the Boer leader-
ship chose to switch to a GWS. Strategic interaction is the best explan-
ation of the South African War’s outcome. Once it became clear that the
capture of the Boer capitals would not force the Boer to surrender,
Britain had three strategic options. First, continue with the present
strategy, perhaps sending more troops, equipment, and horses to the
theater. Second, offer the Boers terms: something which could save face
yet end the war. Third, switch to a barbarism strategy: take the gloves
off and go after Boer noncombatants, either by holding them hostage or
by killing them outright in reprisal for continued resistance. The British
chose the third option, and the war ended soon after.
Kitchener) are misleading in this regard. He suggests that Kitchener’s barbarism was
restrained by Whitehall, but no one taking the history of the war seriously could support
such a claim, except perhaps the weakest form of argument that, barbaric as Kitchener’s
farm burnings and concentration policies were, ‘‘they could have been worse’’ (see
Merom, 2003, pp. 61—62).
208
Conclusion
The Italo-Ethiopian War, 1935–1940

The Italo-Ethiopian War pitted fascist Italy (strong actor) against imper-
ial Ethiopia (weak actor). Italian interests reduced to revenge for the
defeat at Adowa in 1896, responsibility for civilizing a barbarous black
African state in the Horn of Africa, and resources with which to amelio-
rate its poverty and population problems. Italy’s interests in conquest
were higher than expected by the interest asymmetry thesis, and
because a large portion of fascism’s legitimacy lay in its alleged mascu-
line military superiority, any defeat in Ethiopia would have to be
regarded as a threat to the regime’s survival. Ethiopia’s interests (sur-
vival) were high also; but Ethiopians had a more subtle and more
dangerous interest as well: reenacting the glory of the defeat of an
arrogant Italian invader.
Italy’s regime type affected its strategic choices. Italy’s resort to
barbarism (first mustard gas and later murder and mass murder)
presupposed the management and control of information on the
conduct of the campaign.11 In terms of political vulnerability then,
Mussolini’s government remained remarkably safe throughout the
conflict. On the Ethiopian side, Haile Selassie — though nominally
Ethiopia’s emperor — had far less control over his princes and generals.
Ethiopia’s feudal regime type made it impossible for the emperor to
impose his preferred defensive strategy and the impact proved devas-
tating.12 Regime type mattered much more than suggested by the
interest asymmetry thesis.
Arms diffusion played no significant role in the Italo-Ethiopian War.
Because no weapon on the Ethiopian side acted to increase the costs of
conquest or occupation by the Italians, the arms diffusion hypothesis
was not tested.
With one exception, each strategic interaction of the war favored an
Italian military victory. The first interaction featured a conventional
attack against a conventional defense (same-approach), and Italy
appeared poised to lose this fight due to bad generalship. The stalling
11
Cf. fascist legitimacy observation above (p. 116). Even if the Italian people had no
mechanism to turn discontent into a shift in Italian foreign or military policy, the regime
still maintained a vital interest in controlling the public’s perceptions of the war.
12
The case highlights the limits of the notion of strategic choice in war; in theory, a threat
to survival should override petty organizational, cultural, or other prejudices in a choice
of strategy. Yet as this case illustrates, Ethiopians preferred to risk defeat rather than adopt
the ‘‘inglorious’’ defense strategy of striking the enemy where weak and then ‘‘running
away.’’
209
of the Italian advance, combined with the surprisingly (to the Italians)
effective Ethiopian counterattacks, led to a shift in Italian strategy. Italy
supplemented a conventional attack strategy with a super-weapon13
against an Ethiopian conventional defense. Ethiopia’s organized resis-
tance collapsed within a few months. But Italian barbarism sparked
renewed resistance. The Italians would have won this third interaction
of the conflict (same-approach: barbarism vs. GWS), but before it could
register a military impact Mussolini cancelled it. The fourth interaction
of the war reverted to a conventional attack against any organized
resistance (opposite-approach). Only where Ethiopian forces sought
to engage the Italians directly, as at Gojjam, did the Italians win. In
the remainder of the country, the Italian conventional attacks were
countered by a GWS, and the Italians lost. The costs to the Italians of
such operations were staggering, and the combination of those costs
and the perceived need to become security self-sufficient in the event of
war in Europe, led to yet another shift in Italian strategy.
This fifth interaction of the war introduced a conciliation strategy for
Italy, opposed by an Ethiopian GWS (same-approach). Conciliation
proved the most effective of all the strategies the Italians had attempted
since marching into Addis Ababa in May of 1936. Only one rebel leader
held out, and it seemed likely that but for the outbreak of World War II,
even Abebe Aregai would have been forced to flee Ethiopia or submit.
The Italians would have almost certainly won the war.
But World War II changed all of this. The contest in Ethiopia ceased to
be between occupying Italians and an Ethiopian resistance, and shifted
to become a fight between the British and Italians fought across
Ethiopia, Sudan, British Somaliland, and Kenya (it also ceased to be
an asymmetric conflict as defined here). Within a few months, British
forces invading from the south soundly defeated the Italians and
entered Addis Ababa.
Leadership
In the Italo-Ethiopian War, the problem of poor Italian leadership is an
important explanation of the war’s outcome. In the first interaction
Italian timidity threatened the entire invasion, which for political and
13
Again, the effectiveness of a given military technology can vary a great deal depending
on context. In Europe during World War I, the use of chemical weapons had not given
either side a strategic advantage. In Ethiopia, however, the use of such weapons could be
(and was) decisive during the conventional phase of the war.
210
Conclusion
ideological reasons demanded a swift and decisive victory. It was this

timidity that allowed the Ethiopians to seize the initiative, and it was
further incompetence that resulted in a series of Italian defeats which
threatened to throw them back into Eritrea. Poor Italian leadership thus
became the proximate cause of Italy’s resort to mustard gas attacks.
In sum, the outcome of the Italo-Ethiopian War is best explained by
strategic interaction and poor Italian military leadership. Once it
became clear that the capture of Addis Ababa would not force the
Ethiopians to surrender, the Italians had three strategic choices. First,
continue with the present strategy, perhaps sending more troops and
equipment to the theater. Second, offer the Ethiopians terms: something
which could save face yet end the war. Third, switch to a barbarism
strategy: take the gloves off and go after Ethiopian ‘‘rebels’’ and their
families. The AOI generally chose the third option, though it shifted
strategies and innovated a number of times, slowly and painstakingly
improving its advantage until World War II came to Africa and swept
them away.
Italy would have won the war in Ethiopia had World War II not
intervened. But after a few years, at most, Italy would have lost
Ethiopia even without the intervention of the British in North Africa
because its ideal COIN strategy — conciliation — was one it was economi-
cally incapable of sustaining. Italy’s earlier resort to barbarism — both
the use of mustard gas and mass murder — had raised the costs of a
policy of bribery beyond what it could afford to pay.
The United States in Vietnam, 1965–1973

US intervention in the Vietnamese civil war pitted the United States and
GVN (strong actor) against the DRV and VC (weak actor). US interests
were in defending the GVN from a communist takeover, furthering its
strategy of Soviet containment. DRVand VC interests reduced to survival,
but survival expanded to include the ejection of all foreigners and
foreign ‘‘influences’’ from Vietnam.
The DRV and GVN both had authoritarian regime types while the
United States had a democratic regime type. But, in Vietnam, political
vulnerability depended less on regime type and more on location. The
real fight in Vietnam was over which of the two local governments could
most credibly claim to represent Vietnamese national aspirations. The
DRV had a head start due to its organization, and its courage and
sacrifice against the Japanese in World War II and against the French
211
after 1945. The GVN began with a credible nationalist, but Diem — the
very definition of an autocrat — could not see an interest in Vietnam
separate from his own. He restructured the GVN and its military to
serve one purpose at the expense of all others: to keep him in power.
Diem’s assassination and the later introduction of US combat troops
foreclosed any possibility of the GVN winning the fight for nationalist
legitimacy. This put the democratic United States — vulnerable due to its
regime type — in the position of having to support an illegitimate (and
corrupt) GVN.
Arms diffusion did play a role in Vietnam, but the increase in costs to
the United States and GVN of Soviet and Chinese (not to mention
inadvertent GVN) support were not what drove the United States out
of Vietnam. Most US casualties during the war were not caused by
direct enemy action, either in terms of air defense or guerrilla attacks.
Most were inflicted by relatively primitive explosive devices and booby
traps — some improvised from common materials.
Strategic interaction best explains the costs of the conflict for each
side, especially in terms of time. The US military has rarely won the first
battles of any war in which it participated. Instead, its strengths have
always lain in its ability to learn quickly and adapt to a wide variety of
combat conditions. Vietnam proved a case in point. The US military
began by underestimating its adversary’s will and capacity to fight. It
ignored the considerable COIN experience of France and, to a lesser
extent, Great Britain. But it ended by innovating a number of strategies
capable of neutralizing the VC and eviscerating the NVA. By 1969 the
military contest was over and the United States had won. But the
political contest ended in a US defeat, and the strategic interaction
explains why.
No war in the twentieth century was more complex than the
Vietnam War. It featured at least five adversaries (the DRV, VC,
China, and USSR against the GVN and United States) and four over-
lapping strategic interactions. The United States lost ROLLING
THUNDER, which featured US barbarism against an NVA conven-
tional defense (opposite-approach). But the United States won the
main force units war: an NVA conventional attack against a US con-
ventional defense (same-approach). The United States lost the main
guerrilla war in the south, which featured a US conventional attack
against a VC GWS (opposite-approach), but won the ‘‘other’’ guerrilla
war when it initiated the Phoenix Program (barbarism) against a VC
GWS (same-approach). Most importantly, the US Marine Corps — the
212
Conclusion
US military institution with the most historical experience of fighting

small wars — innovated a strategy capable of defeating the VC without a
resort to barbarism. The CAP program’s only liability proved to be its
costliness in terms of time.
In sum, in most wars military victory leads to political victory, and
the duration and quality of that political victory bear some meaningful
relationship to the means used to overcome a resisting military. If
barbarism is employed to achieve military victory, any peace that
follows will be fragile and costly at best. The United States and GVN
did use barbarism and did overcome the DRV and VC militarily.
Strategic interaction explains why the war took as long as it did, how
the United States was able to defeat the DRV and VC, and why, after
doing so it was nevertheless forced to abandon the GVN. But, in the
Vietnamese civil war, military supremacy proved meaningless. The
political contest for Vietnam was what mattered and it was already
lost before the US marines came ashore at Da Nang. Even had the
United States adopted an ideal counterstrategy against each of the
two militaries it faced — no easy task — it still could not have won the
war politically (i.e., coerced the DRV into accepting a divided Vietnam).
At best, the United States would have driven its enemies into Laos,
Cambodia, and China, and then been forced to spend decades attempt-
ing — under constant duress and with little hope of success — to engineer
a viable nationalist alternative to the DRV.
The Afghan Civil War, 1979–89

The Afghan Civil War pitted the DRA and its Soviet allies (strong actor)
against the Afghan mujahideen (weak actor). Soviet interests in the
outcome of the Afghan Civil War did not rise to the level of survival,
but were far more intense than expected by the interest asymmetry
argument. They reduced to (1) defending a friendly Marxist regime,
and (2) thereby preventing the accession of a radical Islamic republic on
its southern border. Mujahideen interests were in ejecting the Soviets
from Afghanistan, where the PDPA regime which controlled Kabul was
not distinguished from the Soviets. To this end they fought as if their
survival was at stake.
According to the interest asymmetry thesis, a major military setback
should have caused the Soviets to re-evaluate their occupation regard-
less of their regime type. This didn’t happen. Instead of considering a
withdrawal, the Soviets attempted gamely to learn from their failures
213
and innovate around them. They often succeeded, but never in a way
which could compensate for their relatively low numbers.14 In short,
the Soviet Union’s authoritarian regime type insulated it from political
vulnerability.
Regime type also made it possible for the Soviets to manage the
domestic political opposition.15 Authoritarian regimes are first of all
free to construct the enemy as they see fit (democratic regimes do the
same thing but with much less freedom). In previous wars the Soviets
had shown a remarkable capacity to control the public’s access to
reliable information about the domestic costs of war, especially casual-
ties. In this war, however, Soviet efforts to control this information
failed: the combination of letters home and an unprecedented network
of soldiers’ mothers combined to make it increasingly clear that Soviet
soldiers were getting killed in numbers far larger than claimed by the
government. However, there remained no mechanism by which rising
alarm and discontent could be translated into a shift in Soviet policy or
strategy.16
In no other case does the arms diffusion argument receive stronger
support than in the Afghan Civil War. From 1979 until 1986, the Soviet
Union’s troops and its DRA allies had better weapons than the muja-
hideen but, with the exception of the combat helicopter, most of that
better technology was of dubious utility in the rugged mountains of
Afghanistan. As Soviet and DRA forces innovated better COIN tactics
they came to rely more and more heavily on combat helicopters. By
1986 their tactics — including the use of heliborne special operations
troops as blocking forces during canyon sweeps — made them both
disproportionately powerful and vulnerable. That vulnerability was
14
Numbers are not a reliable indicator of interests. As in the case of US intervention in
Vietnam, the Soviets had higher-than-expected interests (including the problem of ‘‘cred-
itability’’), but were constrained to observe limits in the number of troops deployed for
fear of escalating the conflict to another world war.
15
Domestic political opposition is one of two risks of adopting a barbarism strategy post-
World War II; the other is foreign military intervention. Here regime type plays no role
whatsoever. As in its later forays into Chechnya (1994 and 1999), Russia’s nuclear status
precluded foreign military intervention. Milosevic’s Serbia, the Taliban in Afghanistan,
and Saddam Hussein in Iraq were not as well protected and were therefore subject to
invasions justified in large measure by accurate reports of barbarism.
16
The Soviet intervention in Afghanistan strongly supports Merom’s model of an author-
itarian regime’s advantages in small wars, but presents the problem that the Soviets and
DRA together employed a strategy of exceeding brutality yet still lost the war. Brutality or
barbarism may not therefore be as effective and efficient as Merom’s model suggests (see
Merom, 2003: pp. 42—46).
214
Conclusion
made salient by a US-supplied shoulder-launched anti-aircraft missile

called the Stinger. Almost overnight what had been an unopposed
tactical advantage turned into an unmitigated tactical liability. In
1985, the accession of Mikhail Gorbachev resulted in new calculations
of Afghanistan’s political importance. Gorbachev placed a time limit on
Soviet support of the DRA. The mujahideen use of Stingers meant that
in order to maintain combat effectiveness, the Soviets would have to
innovate or dramatically increase their numbers. Gorbachev’s acces-
sion meant that there would be no time to innovate and no increase in
troops. The arms diffusion argument therefore receives strong support
in the Afghan Civil War. Combined with the shift in Soviet leadership, it
raised the costs of occupation beyond what the Soviets could afford.
They therefore left.
Strategic interaction explains why the Soviets lost the first interaction
of the war, but it doesn’t explain the second. From 1980 to 1982, the
Soviets pursued a conventional attack strategy against a mujahideen
GWS (opposite-approach). It failed on every count. Yet after a switch to
barbarism in 1983 (same-approach), the Soviets did not win, and in 1989
they withdrew from Afghanistan completely. Why?
External support for weak actors: sanctuary

According to the strategic interaction thesis the Soviets should have
won. Barbarism should have been an ideal COIN strategy because the
Soviets were not vulnerable to domestic or interstate political costs
associated with its prosecution. Moreover, it should have devastated
the intelligence and logistical capacity of the mujahideen and to a large
extent it did.17 Its primary impact, however, was to provide the muja-
hideen with defectors. Barbarism didn’t work in Afghanistan because
the mujahideen were able to rely on foreign intelligence and logistical
support and operate from sanctuaries in Pakistan and Iran. The muja-
hideen became so dependent on this aid that when it ceased in 1989 and
1990, it shattered their ability to topple the illegitimate government of
Najibullah. Afghanistan therefore highlights one condition under
which barbarism is ineffective as a counterinsurgency strategy: when
17
Mujahideen intelligence was much diminished as a result of Soviet barbarism. But
relative to the DRA and the Soviets, each of whom maintained an urban-bound and
mechanized force, it remained superior and benefited from US technical intelligence
support.
215
the guerrillas have access to a significant source of sanctuary,18 intelli-

gence, and logistical support outside the control of the strong actor.
In sum, if we wish to explain why the USSR failed in Afghanistan,
what mattered most was strategic interaction. Once it became clear that
the fight in Afghanistan was a counterinsurgency struggle, Soviet doc-
trine led them to go after mujahideen noncombatants. The strategy hurt
the mujahideen logistical and intelligence infrastructure, but sealing
the border with Pakistan proved impossible. The Soviets succeeded in
killing or making refugees of roughly half of Afghanistan’s pre-war
civilian population. The mujahideen faced an even starker choice after
the Soviets switched to barbarism: either give up or continue with a
strategy which would rely more and more on outside powers for
support.
In the end, no one won the war. The Soviets left in 1989 and the DRA
fell in 1992. The mujahideen were subsequently almost entirely over-
come by the Taliban — a group of extremely conservative Islamists
raised in and funded by Pakistan, as well as by Saudi Arabia.
Afghanistan itself was devastated on a scale not witnessed since the
destruction of Germany in World War II.
How the weak win wars

Key findings
The combination of a large-n statistical analysis and a structured,
focused comparison of historical case studies strongly supports the
strategic interaction thesis over the explanations offered by its rivals.
The book’s large-n analysis confirmed that strategic interaction is
highly correlated with asymmetric conflict outcomes. This relationship
was statistically significant, and supports my claim that same-approach
interactions favor strong actors, while opposite-approach interactions
favor weak actors. Besides testing the logic of competing explanations
of asymmetric conflict outcomes, the five historical case studies pro-
vided sixteen separate tests of the strategic interaction thesis.19 The
results of these tough tests strongly support the logic of the strategic
interaction thesis, even though in the Murid, Italo-Ethiopian, and Afghan
18
Pakistan played host to the best of the mujahideen resistance and Pakistan was both a
nuclear power and an ally of the United States.
19
As noted in Chapter 2 (fn. 47), these cases are in this sense anomalous: 77.5 percent of
asymmetric conflicts feature a single strategic interaction from beginning to end.
216
Conclusion
Civil Wars, the actual outcome of the conflict was strongly affected by extreme
cost insensitivity, incompetent leadership, and external support for the weak
actor respectively.
The expected and actual outcomes of the strategic interactions in the
five historical case studies are summarized in Table 3. 20
Alternative hypotheses
Hypotheses 5, 6, and 8 were evaluated in the individual case studies.
Hypothesis 5 — strong actors are more likely to win same-approach
interactions and lose opposite-approach interactions — was supported
in all five cases. In each case, same-approach interactions made it
possible for the strong actor to apply the full weight of its advantages
in material power to the fight, and this imposed severe costs on weak
actors (sometimes they switched strategies, sometimes they gave up).
Opposite-approach interactions made it possible for weak actors to
avoid defeat and thus made conflicts drag on. As a result, weak actors
did better in opposite-approach interactions, avoiding costs while at the
same time inflicting them on their stronger adversaries. Hypothesis 6 —
the better armed a weak actor is, the more likely it is that a strong actor
will lose an asymmetric conflict — was not supported. In each case,
including even the Afghan Civil War, the impact of technology
depended on strategy and tactics in the context of climate and terrain.
When weak actors had better technology but the wrong strategy, as in
the first year of the South African War — they lost quickly and decisively.
Hypothesis 8 — relative material power explains relative interests in the
outcome of an asymmetric conflict — was refuted. In no case was an
actor’s interest in the outcome of a fight explained by its relative power.
Hypotheses 7, 7a, and 9 can only be evaluated across cases.
Hypothesis 7 — authoritarian strong actors fight asymmetric wars better
than do democratic strong actors — was not supported. In fact, it may be
the case that authoritarian and democratic actors fight asymmetric wars
the same way; both Britain and the USSR began with direct strategies
and resorted to barbarism when the going got rough. On the other
hand, hypothesis 7a — authoritarian strong actors fight asymmetric
wars in which the weak actor uses an indirect strategy better than do
democratic strong actors — was supported here. Public reaction to
20
Table 3 features summaries of thirteen strategic interactions. I count eleven because
although an indirect strategy (targeting an adversary’s will rather than capacity to fight)
‘‘conciliation’’ is a war-termination strategy rather than a war-winning strategy.
217
Table 3. Strategic interactions and conflict outcomes in five historical cases
Phase Strong actor strategy Weak actor strategy Strategic interaction Winner
The Murid War, 1830—59
1 barbarism GWS same-approach Russia
2 conventional attack GWS opposite-approach Murids
3 [conciliation] GWS [same-approach] Russia
The South African War, 1899—1902
1 conventional attack conventional defense same-approach Britain
2 conventional attack GWS opposite-approach Boer
3 barbarism GWS same-approach Britain
The Italo-Ethiopian War, 1935—40
1 conventional attack conventional defense same-approach Ethiopia
2 conventional attack/barbarism conventional defense same-approach Italy
3 barbarism GWS same-approach Italy
4 conventional attack GWS opposite-approach Ethiopia
5 [conciliation] GWS [same-approach] Italy
The United States in Vietnam, 1965—73
1 barbarism conventional defense opposite approach DRV
2 conventional attack conventional defense same-approach United States
3 conventional attack GWS opposite-approach VC
4 barbarism GWS same-approach United States
The Afghan Civil War, 1979—89
1 conventional attack GWS opposite-approach mujahideen
2 barbarism GWS same-approach mujahideen
Italic denotes outcome not entirely explained by strategic interaction theory.

Conclusion
Britain’s resort to barbarism nearly forced it from the war, but at no time
did the USSR’s resort to barbarism threaten to do so. Thus, to the extent
that an authoritarian regime type insulates an actor from the costs and
risks of initiating a barbarism strategy, and to the extent that barbarism
is a militarily effective COIN strategy, this support for hypothesis 7a
should come as no surprise. Finally, hypothesis 9 — authoritarian and
democratic strong actors share roughly equal political vulnerability in a
prolonged asymmetric conflict — was decisively rebutted here. Britain
was at considerable risk of being forced from the war after Emily
Hobhouse brought the concentration camp controversy to light. But
the Russian Empire, the Italian Fascists, and USSR were never at risk of
similar problems. Italy ran a risk in 1935 that its fellow European states
would demand its withdrawal from Ethiopia or halt the use of mustard
gas, but in the event the weakness of Europe’s great powers allowed
him to get away with the brutal attack and profound immiseration of a
League member.
Explaining strong actor failure

Strong actors lose small wars under a number of conditions. First, when
strong actors are also great powers they have strong adversaries. They
devote considerable resources to building forces to secure their interests
against threats from those adversaries. Not only are forces built and
equipped for use against other great powers, but they are trained to
fight and react in a certain way. Doctrine, equipment, and training are
all tightly intermeshed, and building new forces with different skills
and doctrines is costly.21
In the Murid War, the first Russian forces to come up against the
Murids were trained and equipped to fight European-style adversaries
on hilly, sparsely forested, or open terrain. Their officers had been
trained on the maxims of Suvorov, and other famous Russian generals
who had engineered the total defeat of Napoleon’s Grande Armée. Their
first clash with the agile Murids in the steep mountain defiles of the
Caucasus came as a jarring shock, and it required nearly thirty years of
21
Even more seriously, officers and men trained in one context are often not useful in
another. Britain’s military ineffectiveness in World War I was often said to be due to the
fact that much of its officers’ combat experience was ‘‘colonial.’’ The reverse also holds
true: commanders who excel at conventional operations rarely do as well in COIN
operations. On the differences in leadership in conventional and unconventional settings,
see, e.g., Bowden (2000: 172—174), and Marquis (1997: 4, 8).
219
struggle, innovation, retraining, and new technology in order to defeat

the Murids.
The British came to South Africa buoyed by years of success in small
wars against the tribes and kingdoms which peppered their vast
empire. They came piecemeal, ill-equipped, and ill-trained to fight the
mounted Boer on the hot and mountainous veld. Time and again the
British lined up against the Boer for their Aldershot exercises: artillery
preparation of the enemy line, infantry advance in close order, fix
bayonets for the charge, and cavalry to annihilate the routed foe. But
what they found was counter-battery fire, and a wall of lead from well-
aimed fast repeating invisible rifles which devastated their closely
packed infantry advances. Then, after suffering this withering fire,
they charged the enemy positions only to find a few dead Boer and
the rest of the commando trotting off on their horses, kicking up dust in
the distance. The British needed a new army, new training, new equip-
ment, many more horses, and many more men in order to bring the
Boer to bay.
The Italians came to Ethiopia with precisely the quantity and quality
of troops they required in order to conquer Ethiopian tribes united
under Haile Selassie. By all accounts, with the exception of the Black
Shirt divisions, Italy’s soldiers fought well. But Italy’s military leader-
ship was so poor as to almost cause Italy’s defeat against the mostly
barefoot, poorly armed and outnumbered Ethiopian warriors.
The United States arrived in Vietnam with the conviction that its
unmatched technology and long history of military victory would
prove decisive. True, US forces had suffered a near rout in the early
months of the Korean Conflict, but US General Matthew Ridgeway’s
reapplication of basic soldiering discipline had restored US forces to
combat effectiveness, turning the tide and pushing the North Koreans
back. As a result, in Vietnam many US generals believed that talk of
‘‘new’’ or ‘‘unconventional’’ warfare was a smokescreen for a lack of
good discipline and aggressiveness. They were wrong. As in all such
conflicts lower and mid-level officers (lieutenants through lieutenant
colonels) did most of the rapid learning and adaptation; while senior
officers attempted to apply the weight of their own, often marginally-
relevant, experience to the difficult problems facing US forces in
Vietnam.22 The US military had the right forces and doctrine to defeat
the NVA, but in the years following World War II it had systematically
22
The best account of this is Sheehan (1988). See also Krepinevich (1986).
220
Conclusion
gutted its COIN capabilities (Marquis, 1997: Ch. 2). What it needed in
Vietnam was two armies,23 and what it had was one army desperately
trying to adapt to two different military universes.
The Soviets entered Afghanistan in 1979 with four motorized rifle
divisions at two-thirds strength. These reservists had outdated equip-
ment, and they had been trained to use that equipment in large-scale
operations on open terrain. They were unsuited in every imaginable
way to prosecute a COIN war in mountainous terrain. During their
decade-long struggle to defeat the mujahideen, the Soviets innovated
an almost entirely new army: new doctrine, new technology, new
training, and different troops.
All of this demonstrates that strong actors — especially great powers —
who have not recently fought small wars will tend to enter them
unprepared to fight them in terms of either doctrine or equipment.
Even those who have recently fought such wars may find their experi-
ences in one theater against one adversary do not transfer well to
different theaters and different adversaries. This doesn’t imply that
strong actors must lose asymmetric conflicts, but rather that their
initial costs will be higher than anticipated by conventional IR theory,
and by the political elites charged with committing armed forces to
combat.
After combat is joined, strong actors can lose if they adopt the wrong
strategy given their adversary’s strategy. If weak actors choose a con-
ventional defense strategy, strong actors can lose if they attempt to use
strategic air power (indirect-approach) to win. The costs in terms of
time and the collateral damage which inevitably follows such attacks
provoke outrage internationally, and often domestically as well. Either
sort of outrage can create pressure to cease hostilities short of achieving
a strong actor’s political objectives. If weak actors choose an indirect
defense strategy, strong actors can lose if they attempt to use a conven-
tional attack strategy (direct-approach) to win. GWS is specifically
designed to trade time for territory, so unless strong actors are willing
to commit millions of troops for decades, they are unlikely to win
against an adversary that avoids contact and strikes when and where
least expected.
Strong actors can use barbarism to defeat weak actors using a GWS
or nonviolent resistance militarily. But whether authoritarian or
23
See Cohen (1984: 180) on the desirability of creating two distinct forces to fight two
distinct types of war.
221
democratic the costs for strong actors of achieving such victories — even
in narrow military terms — appear to have risen steadily since the end of
World War II. More importantly, strong actors can no longer win a
subsequent peace against weak actors they’ve overcome by barbarous
means.24 And strong actors can also lose if they attempt to use a risky
barbarism strategy to overcome recalcitrant defenders employing a
GWS. Increasingly, barbarism appears to be stimulating military resist-
ance rather than deterring it. The risks of barbarism are greater for
democratic regimes than for authoritarian regimes, but both may face
the problem of international sanction. For strong actors who are great
powers or superpowers, the costs of international sanction may be
negligible, but for strong actors who are minor powers — such as Italy
in 1935, or Serbia in 1999, the risk of international sanctions can be
prohibitive.25 Democratic strong actors also face domestic risks when
prosecuting a barbarism strategy: the effects of Britain’s concentration
camp policy in the South African War, for example, would almost
certainly have forced its withdrawal from the war had there not been
such a long delay between the effects and their political impact in
Britain. In the digital age, such delays have shrunk dramatically, and
barbarism is much more difficult to conceal than at the turn of the
century.
In sum, the problem for strong actors is weak actors who pursue an
indirect defense strategy, such as a GWS or terrorism. This presents
strong actors with three unpalatable choices: an attrition war lasting
perhaps decades; costly bribes or political concessions, perhaps forcing
political and economic reforms on repressive allies as well as adver-
saries; or the deliberate harm of noncombatants in a risky attempt to
win a military contest quickly and decisively.
Limitations of the strategic interaction theory

At this point in the analysis the strategic interaction thesis can be
established as a full theory of asymmetric conflict outcomes. It has
survived two different and tough tests, and been shown to subsume
competing explanations, such as interest asymmetry, nature-of-actor,
24
Even in cases where the political aim of attacking forces is genocide, the technical
requirements for achieving genocide are, fortunately, beyond reach even for the most
advanced states. There will always be survivors and witnesses, and this creates the
likelihood of revenge attacks of varying degrees of severity following the perpetration
of barbarism.
25
Italy did not suffer from these sanctions because Britain and France turned a blind eye.
222
Conclusion
and arms diffusion arguments; strategic interaction theory brackets the

conditions under which each of these important factors operates.
Finally, strategic interaction theory explains the fact of strong actor
failure in a way consistent with the observed trend toward increasing
strong actor failures over time.
As a general theory, it will be relatively less satisfying as a complete
explanation of any particular asymmetric conflict outcome, but rela-
tively more satisfying as an explanation of all asymmetric conflict out-
comes, and in particular as a guide to strategy and policy. It has
application, for example, to counterterrorism as well as COIN strategy
(Arreguı́n-Toft, 2002), both of which will prove vital subjects for US
policymakers in the coming decades.
But as with any general theory it is not without its limitations. First,
although strategic interaction theory highlights the limitations of nar-
row or material definitions of ‘‘power,’’ it does not comprehensively
resolve those limitations. Consider strategic interaction in the Russo-
Japanese War of 1905, for example. The Russian Empire was the strong
actor and Japan the weak one. The strategic interaction was same-
approach but Russia lost. Why? Although the statistical analysis in
Chapter 2 reports the conflict this way, it is clear that another necessary
nuance of power must be a maritime/continental power distinction.
Russia was a continental power, Japan a maritime power. Japan won
because the key battles were fought at sea (even the bloody battle for
Port Arthur was determined by the superior ability of the Japanese to
deploy ground forces and reinforcements by sea).26 In demonstrating
that military effectiveness depends both on raw resources and the
interaction of plans for the employment of those resources, this analysis
has established beyond question that power is more nuanced than
realist international relations theory has to date allowed,27 yet a true
general definition of power remains beyond its scope.
Second, nothing in this analysis has made it possible to disaggregate
the effects of anti-colonialism and nationalism from those of strategic
interaction — at least not in the trend aspect of the argument. Clearly, the
26
Other examples include Britain’s successful resistance to Napoleon’s France in the
nineteenth century and to Hitler’s Third Reich in 1941.
27
This is not the same thing as allowing that the realist IR theory definition of power is a
‘‘straw man’’ in terms of argumentation. This too-simple definition of power is neither my
own nor is it a straw man. It is not a straw man because it is still relied upon by policymakers
in the real world to make calculations of the likelihood of success and failure in asymmetric
conflicts.
223
success of Mao’s resistance to Japan and the Kuomintang was due to

strategic interaction. But could Mao have knit together the mass move-
ment he did without Chinese nationalism? No. In fact, nationalism —
religious or ethnic — may count as a permissive condition of a GWS with
the capacity to evolve into a conventional sustained offensive. The Avars of
Daghestan could not have managed a coordinated offensive against the
Russian Empire alone. Something — an idea and a charismatic leader —
had to knit together the patchwork of feuding tribes into a cohesive
movement before skilful offensive operations became possible. On the
other hand, nationalism represented in a same-approach strategic inter-
action will have no impact on conflict outcomes, asymmetric or other-
wise. The Belgians were nationalistic and lost quickly to the Germans in
1914. The French were nationalistic yet lost handily to the Germans in
1940. The Arabs are nationalistic and far more numerous than the
Israelis, yet their zeal has never been translated into military victory.
In sum, the best terrorist, guerrilla, and nonviolent resistance move-
ments will be nationalistic, but not all nationalists will oppose strong
actors by means of terrorism, GWS, or nonviolent resistance.
Theoretical and policy implications

Material power is useful for theory building because it is quantifiable
and measurable in a way that courage, leadership, and dumb luck are
not. This study has demonstrated empirically that relative material
power is more than simply a methodologically useful concept; taken
alone, it explains a majority of conflict outcomes since 1800. The strat-
egic interaction thesis makes clear, however, the limitations of relative
material power by highlighting the conditions under which it matters
more or less.
This analysis suggests key policy implications for both weak and
strong actors. For weak actors, successful defense against a strong
actor depends on an indirect strategy. Because indirect strategies such
as GWS and nonviolent resistance depend on strong social support,
weak actors must either work tirelessly to gain and maintain the sup-
port (sympathy or acquiescence) of a majority of the population in
question, or risk becoming dependent on a foreign donor of logistical
support and arms.28 Given the risks involved with either aiding or
taking part in a guerrilla resistance, gaining strong social support is
28
The latter option may keep the insurgency alive, but it cannot substitute for a genuine
and locally based network of social support.
224
Conclusion
itself no mean feat. Additionally, weak actors must have or gain access
to the physical or political sanctuary necessary to make an indirect
strategy a viable choice. For strong actors, the strategic interaction
theory suggests that weak adversaries employing an indirect defense
will be difficult to defeat. Of course, not all or even most asymmetric
conflicts need follow this pattern but, when they do, and when a resort
to arms seems the only viable option, how should a strong actor such as
the United States react?
One response might be a resort to barbarism, which appears to be an
effective military strategy for defeating an indirect defense.29 But even a
cursory review of postwar conflicts reveals that at best barbarism can be
effective only as a military strategy: if the desired objective is long-term
political control — e.g., nation-building, ‘‘peace’’ keeping, or other stabil-
ity or transition missions — barbarism invariably backfires. The French,
for example, used torture to quickly defeat Algerian insurgents in the
Battle of Algiers in 1957. But when French military brutality became
public knowledge it catalyzed domestic political opposition to the
war in France, and stimulated renewed and intensified resistance by
the non-French population of Algeria (Mack, 1975: 180; Asprey, 1994:
669—671). Within four years, France had abandoned its claims in Algeria
even though it had ‘‘won’’ the war. Barbarism thus sacrifices victory in
peace for victory in war — a poor policy at best (Liddell Hart, 1967: 370).
The same must be said of counterterrorist barbarism: assassination,
torture, and random reprisal. These counterterrorist strategies have
been doggedly and expertly pursued for more than forty years by the
Israelis in Palestine, but they have brought the Israelis no closer to
peace.
An ideal US strategic response therefore demands three key ele-
ments: (1) preparation of public expectations for long wars despite US
29
This is the essence of Merom’s argument: democracies lose small wars because they’re
generally too reluctant to suffer the casualties and brutality necessary to win. Beyond the
cases he analyzes, however, are many that don’t fit this pattern. Authoritarian strong
actors with no limits on their willingness to use barbarism or suffer casualties have lost
small wars, and democratic actors have won small wars even while refusing to resort to
barbarism (e.g., Britain in the Malayan Emergency of 1948). The same logic — that barbar-
ism pays — appears at first to be true of counterterrorism and terrorism as well. Laura K.
Donohue analyzes the impact of British counterterrorist legislation in Northern Ireland
and concludes that Britain’s numerous ‘‘temporary’’ and ‘‘emergency’’ measures — which
were never temporary, and which violated civil liberties and due process — proved highly
effective in the short term. Her analysis suggests, however, that insurgents always found a
way around such measures, eventually prompting yet another round of ‘‘emergency’’
restrictions (see Donohue, 2001: 322—323).
225
technological and material advantages; (2) the development and

deployment of armed forces specifically equipped and trained for
COIN operations;30 and (3) an awareness of and preparation for the
political consequences of military victory. Without a national consensus
and realistic expectations, the United States would be politically vul-
nerable in an asymmetric conflict. Without more special operations
forces — the self-reliant and discriminate armed forces necessary to
implement an ideal COIN strategy — what begins as military operations
against isolated violent minorities will tend to escalate into wars against
an entire society (Walzer, 2000: 187). It is crucial to add that in order to
be effective these forces will need to be supported by a human intelli-
gence network that may take a state such as the United States years to
build.31 Without a well thought-out postwar plan emphasizing eco-
nomic recovery under a competent and uncorrupt administration,
even rapid military victories can degenerate into desultory insurgency.
As Eliot Cohen argued twenty years ago, the United States must be
prepared to fight and win both conventional and asymmetric wars.
Strategic interaction theory shows why the two missions demand two
kinds of armed forces: one to defend US interests in conventional wars,
and one to defend them in small wars or against terrorists. It also
highlights the importance of politics and diplomacy in combating
insurgencies and terrorists. Determined insurgents and terrorists are
difficult to defeat. But where strong actors have succeeded, they have
done so most dramatically by preceding discriminate military attacks
with meaningful political and economic reforms — reforms that effect-
ively isolated guerrillas or terrorists from their the base of social sup-
port.32 This was the pattern of British success in Malaya in 1948, and of
US and Philippine government success in the Philippines in 1952. By
contrast, military and economic support for repressive, corrupt, and
incompetent regimes led to superpower defeats in Vietnam and
Afghanistan.
Thus, strategic interaction theory explains not only how the weak
win wars, but how the strong lose the peace. Contrary to Merom’s
30
See, for example, May, in Hoffmann et al. (1981: 8, 9); Hoffmann (ibid.: 10); Cohen (1984:
166—167); and Johnson (2001: 181—182).
31
Conventional US forces are already well supplied with the kinds of technical
intelligence — satellite and electronic spectrum intelligence — necessary to achieve over-
whelming victory.
32
This will not stop all terrorism but it will make such terrorism as survives (a) less
deadly, and (b) less likely to be viewed by target audiences as much more than criminal or
irrational behavior.
226
Conclusion
thesis, following World War II the unrestrained application of brute

force rarely wins wars and can never win the peace. This is true for two
reasons. First, it has proven technically impossible to entirely destroy a
people or to destroy them sufficiently to foreclose the possibility of a
future backlash (the Hutus murdered Tutsis in Rwanda even more
efficiently than the Germans had murdered Jews in World War II, but
the Tutsis are in control in Rwanda today and the Jews have their own
state). Second, the principle of national self-determination has made it
increasingly rational for people to sacrifice their lives for the cause; the
implications of murder against a disparate group of unrelated people
differ dramatically from the implications of murder against a national.
In this nationalist context it is much more difficult to coerce a person
who would prefer to die than accede; and barbarism-provoked fear is as
likely to lead to violent resistance as to cowed compliance.
If the United States wants to win wars it must build two different
militaries. If it wants to win the peace — a far more ambitious and useful
goal — it must support its resort to arms by eliminating foreign policy
double standards and by increasing its capacity and willingness to use
methods other than violence to resolve or deter conflicts around the
world. So long as the United States continues to support corrupt and
repressive regimes in Latin America, Africa, the Middle East, and Asia,
terrorists and insurgents will continue to frustrate US interests in global
peace and prosperity. The current US government has confused mili-
tary power with state power, and by over-applying the former has
actually undermined its interests. If this policy continues and follows
the historical pattern of every previous attempt to accomplish the same ends
(peace) by the same means (the overwhelming application of military
force unsupported by political, economic, and administrative
resources), the result will be costly quagmires such as Vietnam,
Afghanistan (1979 and 2002—), and Iraq (2003—), and a future attack on
the United States or its allies that makes the terror attacks of September
11, 2001 pale by comparison.
227
Appendix
Strategic interactions and outcomes, 1816–2003

Table App. 1 A list of cases and key variables
OUTCOME
(strong
WARNAME START END actor) STRATINT
Spain—Peru 1809 1816 wins same-approach

Russo-Georgian 1816 1825 wins same-approach
Pindari War 1817 1818 wins same-approach
Kandyan Rebellion 1817 1818 wins same-approach
Greek War of Independence 1821 1828 loses/tie opposite-approach
First Anglo-Burmese 1823 1826 wins same-approach
First Ashanti 1824 1826 wins same-approach
Javanese 1825 1830 wins same-approach
Bharatpuran 1825 1826 wins same-approach
Russo-Circassian 1829 1840 wins same-approach
Albanian 1830 1831 wins same-approach
Belgian Independence 1830 1831 wins same-approach
Murid War 1830 1859 wins same-approach
First Polish 1831 1831 wins same-approach
First Syrian 1831 1832 loses/tie same-approach
Texan 1835 1836 loses/tie same-approach
Second Seminole War 1835 1842 wins same-approach
First Zulu 1838 1840 wins same-approach
First British—Afghan 1838 1842 loses/tie opposite-approach
Franco-Algerian 1839 1847 wins same-approach
Bosnian—Turkish 1841 1841 wins same-approach
Baluchi—British 1843 1843 wins same-approach
First Maori 1843 1848 wins same-approach
Franco-Moroccan 1844 1844 wins same-approach
First British—Sikh 1845 1846 wins same-approach
First Kaffir War 1846 1847 wins same-approach
228
Appendix
Table App. 1 (cont.)
OUTCOME
(strong
Cracow Revolt 1846 1846 wins same-approach

Austro-Sardinian 1848 1849 wins same-approach
First Schleswig-Holstein 1848 1849 wins same-approach
Hungarian 1848 1849 wins same-approach
Second British—Sikh 1848 1849 wins same-approach
Roman Republic 1849 1849 wins same-approach
Second Kaffir 1850 1853 wins same-approach
La Plata 1851 1852 wins same-approach
Second Anglo-Burmese 1852 1853 wins same-approach
First Turko-Montenegran 1852 1853 loses/tie same-approach
Third Seminole War 1855 1858 wins same-approach
Yakima War 1855 1858 loses/tie opposite-approach
Anglo-Persian 1856 1857 wins same-approach
Second Opium War 1856 1860 loses/tie same-approach
Kabylia Uprising 1856 1857 wins
Tukulor—French War 1857 1857 wins same-approach
French—Indochinese 1858 1863 wins same-approach
Second Turko-Montenegran 1858 1859 loses/tie same-approach
Spanish—Moroccan 1859 1860 wins same-approach
Italo-Roman 1860 1860 wins same-approach
Second Maori 1860 1870 wins same-approach
Apache and Navaho War 1860 1865 wins same-approach
Taiping Rebellion 1860 1864 wins same-approach
Nien Rebellion 1860 1868 wins same-approach
Franco-Mexican 1862 1867 loses/tie same-approach
First Sioux War 1862 1864 wins same-approach
Second Polish 1863 1864 wins same-approach
Spanish—Santo Dominican 1863 1865 loses/tie
Second Schleswig-Holstein 1864 1864 wins same-approach
Lopez War 1864 1870 wins opposite-approach
Spanish—Chilean 1865 1866 wins same-approach
British—Bhutanese 1865 1865 wins same-approach
Second Sioux War 1865 1868 loses/tie opposite-approach
First Cretan 1866 1867 wins opposite-approach
Ten Years War 1868 1878 wins same-approach
Algerian 1871 1872 wins
Second Apache War 1871 1873 wins same-approach
Second Ashanti 1873 1874 wins same-approach
Tonkin 1873 1885 wins same-approach
Dutch—Achinese 1873 1878 wins same-approach
Red River Indian War 1874 1875 wins same-approach
Balkan 1875 1877 wins opposite-approach
Third Apache War 1876 1886 wins opposite-approach
Third Sioux War 1876 1877 wins same-approach
Ninth Kaffir 1877 1878 wins same-approach
229
Appendix
OUTCOME
(strong
Russo-Turkoman 1878 1881 wins same-approach

Second British—Afghan 1878 1880 wins same-approach
British—Zulu 1879 1879 wins same-approach
Gun War 1880 1881 wins same-approach
First Boer War 1880 1880 loses/tie same-approach
Tunisian 1881 1882 wins
Franco-Indochinese 1882 1884 wins same-approach
Mahdist 1882 1885 loses/tie same-approach
First Franco-Madagascan 1883 1885 wins same-approach
Sino-French 1884 1885 loses/tie same-approach
Russo-Afghan 1885 1885 wins same-approach
Third Anglo-Burmese 1885 1886 wins same-approach
First Mandigo-French War 1885 1885 wins same-approach
First Italo-Ethiopian 1887 1887 loses/tie same-approach
Second Cretan 1888 1889 wins
Dahomey 1889 1892 wins same-approach
Second Senegalese 1890 1891 wins same-approach
Messiah War 1890 1891 wins same-approach
Congo Arabs 1892 1892 wins
Franco-Thai 1893 1893 wins
Third Ashanti 1893 1894 wins same-approach
Matabele—British War 1893 1893 wins same-approach
Sino-Japanese 1894 1895 loses/tie same-approach
Franco-Madagascan 1894 1895 wins same-approach
Balian 1894 1894 wins
Cuban 1895 1898 wins same-approach
Italo-Ethiopian 1895 1896 loses/tie same-approach
Fourth Ashanti 1895 1896 wins same-approach
Third Cretan 1896 1897 loses/tie same-approach
Druze—Turkish 1896 1896 wins
First Philippine 1896 1898 loses/tie opposite-approach
Sudanese War 1896 1899 wins same-approach
Greco-Turkish 1897 1897 wins same-approach
Indian Muslim 1897 1898 wins same-approach
Nigerian 1897 1897 wins same-approach
Hut Tax 1898 1898 wins same-approach
Second Philippine 1899 1902 wins same-approach
Second Boer War 1899 1902 wins same-approach
Somali Rebellion 1899 1905 wins same-approach
Russo-Manchurian 1900 1900 loses/tie same-approach
Ilinden 1903 1903 wins same-approach
Russo-Japanese War 1904 1905 loses/tie same-approach
South West African Revolt 1904 1905 wins opposite-approach
Maji—Maji Revolt 1905 1906 wins opposite-approach
Second Zulu War 1906 1906 wins same-approach
230
Appendix
OUTCOME
(strong
Spanish—Moroccan 1909 1910 wins same-approach

First Moroccan 1911 1912 wins same-approach
First Balkan War 1912 1913 loses/tie same-approach
Tibetan War of Independence 1912 1913 loses/tie same-approach
Second Moroccan 1916 1917 wins
Arab Revolt 1916 1918 loses/tie opposite-approach
Irish Troubles 1916 1921 loses/tie opposite-approach
Yunnan 1917 1918 wins
First Sino-Tibetan 1918 1918 loses/tie
Russo-Polish 1919 1920 loses/tie same-approach
Lithuanian—Polish 1919 1920 wins same-approach
Hungarian—Allies 1919 1919 wins same-approach
Franco-Turkish 1919 1922 loses/tie same-approach
Third Afghan 1919 1919 wins same-approach
Franco-Syrian 1920 1920 wins
Iraqi—British 1920 1921 wins
Sanusi 1920 1932 wins same-approach
Riffian 1921 1926 wins
Druze Rebellion 1925 1927 wins same-approach
US—Nicaraguan 1927 1933 wins opposite-approach
Chinese Muslims 1928 1928 wins
Chinese Civil War 1930 1935 wins same-approach
Manchurian 1931 1933 loses/tie
Soviet—Turkestani 1931 1934 wins
Italo-Ethiopian 1935 1936 wins same-approach
Sino-Japanese 1937 1941 wins same-approach
Chankufeng 1938 1938 loses/tie same-approach
Winter War 1939 1940 wins same-approach
Franco-Thai 1940 1941 loses/tie
Indonesian Independence 1945 1946 loses/tie same-approach
Indochinese 1945 1954 loses/tie same-approach
Madagascan 1947 1948 wins
First Kashmir 1947 1949 wins same-approach
Palestine 1948 1949 loses/tie same-approach
Malayan Rebellion 1948 1957 wins same-approach
Hyderabad 1948 1948 wins
Korean Conflict 1950 1953 loses/tie same-approach
Sino-Tibetan 1950 1951 wins same-approach
Philippines 1950 1952 wins same-approach
Kenya 1952 1956 wins same-approach
Tunisian Independence 1952 1954 loses/tie
Moroccan Independence 1953 1956 loses/tie
Algerian 1954 1962 loses/tie same-approach
British—Cypriot 1954 1959 loses/tie same-approach
Cameroon 1955 1960 loses/tie
231
Appendix
OUTCOME
(strong
Russo-Hungarian 1956 1956 wins same-approach

Sinai 1956 1956 wins same-approach
Tibetan 1956 1959 wins opposite-approach
Cuba 1958 1959 loses/tie opposite-approach
South Vietnam 1960 1965 loses/tie opposite-approach
Congo 1960 1965 wins same-approach
Kurdish 1961 1963 wins
Angola—Portugal 1961 1975 loses/tie
Sino-Indian 1962 1962 wins same-approach
Guinea Bissau 1962 1974 loses/tie opposite-approach
Mozambique 1964 1975 loses/tie opposite-approach
Vietnam 1965 1975 loses/tie opposite-approach
Second Kashmir 1965 1965 loses/tie same-approach
Six Day War 1967 1967 loses/tie same-approach
Israeli—Egyptian 1969 1970 loses/tie same-approach
Bangladesh 1971 1971 wins same-approach
Philippine—Moro 1972 1980 wins same-approach
Yom Kippur War 1973 1973 loses/tie same-approach
Turko-Cypriot 1974 1974 wins same-approach
Eritrean 1974 1991 loses/tie same-approach
Kurdish Autonomy 1974 1975 wins same-approach
East Timor 1974 1975 loses/tie opposite-approach
Vietnamese—Cambodian 1975 1979 wins same-approach
Western Sahara 1975 1983 wins
Chadian Civil War 1975 1988 wins same-approach
Ethiopia—Somalia 1977 1978 wins same-approach
Afghanistan 1978 1989 loses/tie same-approach
Sino-Vietnamese 1979 1979 wins same-approach
Peruvian Civil War 1982 1992 wins same-approach
Tamil Rebellion 1983 1990 loses/tie opposite-approach
Sino-Vietnamese 1985 1987 loses/tie same-approach
Gulf War 1990 1991 wins same-approach
Iraq—Kuwait 1990 1990 wins same-approach
Kurdish Rebellion 1991 1999 wins same-approach
Serbian Rebellion 1991 1996 loses/tie opposite-approach
Russo-Chechen 1994 1996 loses/tie same-approach
Kosovo I 1998 1999 wins same-approach
Kosovo II 1999 1999 wins same-approach
Russo-Chechen II 1999 same-approach
Anti-Taliban 2002 2002 wins same-approach
Gulf War II 2003 2003 wins same-approach
Number of cases ¼ 202
232
Appendix
Associated statistics for wins and losses over time

Table App. 2 Correlation Coefficients of Strategic Interaction
(STRATINT) and External Support for Weak Actors (WASUPP)
OUT2 OUT3
STRATINT 0.28 0.29

WASUPP 0.30 0.28
p < 0.01
Associated statistics for correlation between

strategic interaction and conflict duration,
1800—2003
Table App. 3 Strategic Interaction (STRATINT) by collapsed duration
of war (WARDUR2)
WARDUR2
1 (0—1 2 (2—5 3 (6—9 4 (10—14 5 (15—50

STRATINT years) years) years) years) years)
Same- 93 38 14 4 2
strategic
expected 84.7 41.9 17.5 5.2 1.7
count
row 61.6 25.2 9.3 2.6 1.3
percent
Opposite- 4 10 6 2 0
strategic
expected 12.3 6.1 2.5 .8 .3
count
row 18.2 45.5 27.3 9.1 0
percent
Column 97 (56.1 48 (27.7 20 (11.6 6 (3.5 2 (1.2)
totals percent) percent) percent) percent)
Chi-Square Value DF Pearson
Significance 17.273 4 0.002
233
Appendix
Associated statistics for interaction phase types,

1800—2003
Table App. 4 Interaction Phase Type (INTPTYPE) Distribution,
1800—2003
Valid Cum
Value Label Value Frequency Percent Percent Percent
Single interaction 0 140 69.7 78.7 78.7

Sequential interaction 1 28 13.9 15.7 94.4
Simultaneous 2 10 5.0 5.6 100.0
interaction
23 11.4 Missing
————— ————— —————
Total 201 100.0 100.0
Valid cases: 178 Missing cases: 23
234
References
Anderson, Benedict, 1991. Imagined Communities: Reflections on the Origin and

Spread of Nationalism. London: Verso.
Andreski, Stanislav, 1968. Military Organization and Society. Berkeley: The
University of California Press.
Arreguı́n-Toft, Ivan, 2001. ‘‘How the Weak Win Wars: A Theory of Asymmetric
Conflict,’’ International Security, vol. 26, no. 1: 93—128.
2002. ‘‘Tunnel at the End of the Light: A Critique of US Counter-Terrorist
Grand Strategy,’’ Cambridge Review of International Affairs, vol. 15, no. 3:
549—563.
2003. ‘‘The [F]utility of Barbarism: Assessing the Impact of the Systematic
Harm of Non-Combatants as a Strategy in War,’’ paper presented at the
annual meeting of the American Political Science Association.
Asprey, Robert B., 1994. War in the Shadows: The Classic History of Guerrilla
Warfare from Ancient Persia to the Present. New York: Little, Brown and
Company.
Baddeley, John F., 1908. The Russian Conquest of the Caucasus. London:
Longmans, Green & Co.
Bartov, Omer, 1992. Hitler’s Army: Soldiers, Nazis, and War in the Third Reich. New
York: Oxford University Press.
Bennett, D. Scott, and Allan C. Stam III, 1998. ‘‘The Declining Advantages of
Democracy: A Combined Model of War Outcomes and Duration,’’ Journal
of Conflict Resolution, vol. 42, no. 3: 344—366.
Blainey, Geoffrey, 1988. The Causes of War, 3rd ed. New York: The Free Press.
Blanch, Lesley, 1960. The Sabres of Paradise. New York: The Viking Press.
Blaufarb, Douglas S., and George K. Tanham, 1989. Who Will Win? A Key to the
Puzzle of Revolutionary War. New York: Crane Russak.
Borovik, Artyom, 1990. The Hidden War: A Russian Journalist’s Account of the
Soviet War in Afghanistan. New York: The Atlantic Monthly Press.
Boserup, Anders, and Andrew J. R. Mack, 1975. War Without Weapons: Non-
Violence in National Defense. New York: Schocken Books.
Brown, Frederick Z., 1980. ‘‘Comment on Mueller: American Misperceptions,’’
International Studies Quarterly, vol. 24, no. 4: 525—529.
235
References
Callwell, C.E., 1996. Small Wars: Their Principles and Practice, 3rd ed. Lincoln:
University of Nebraska Press.
Caputo, Philip, 1996. A Rumor of War. New York: Henry Holt and Company.
Chaliand, Gérard, ed., 1982. Guerrilla Strategies: An Historical Anthology from the
Long March to Afghanistan. Berkeley: University of California Press.
Charnay, J. P., 1994. ‘‘Strategy,’’ in André Corvisier and John Childs, eds., Chris
Turner, trans., A Dictionary of Military History and the Art of War. Cambridge,
Mass: Blackwell, pp. 768—774.
Clodfelter, Mark, 1989. The Limits of Air Power: The American Bombing of North
Vietnam. New York: The Free Press.
Coffey, Thomas M., 1974. Lion by the Tail: The Story of the Italian–Ethiopian War.
London: Hamish Hamilton.
Cohen, Eliot A., 1984. ‘‘Constraints on America’s Conduct of Small Wars,’’
International Security, vol. 9, no. 2: 151—181.
Connor, Walker, 1969. ‘‘Ethnology and the Peace of South Asia,’’ World Politics,
vol. 22, no. 1: 51—86.
Cordesman, Anthony H., and Abraham R. Wagner, 1990. The Lessons of Modern
War, Volume III: the Afghan and Falklands Conflicts. Boulder: Westview Press.
Corvisier, André, and John Childs, 1994. ‘‘Indirect Warfare,’’ in Corvisier and
Childs, A Dictionary of Military History and the Art of War. Cambridge, Mass:
Blackwell, p. 378.
Corvisier, André, and Barry Paskins, 1994. ‘‘Laws of War,’’ in Corvisier and
Childs, A Dictionary of Military History and the Art of War. Cambridge, Mass:
Blackwell, pp. 443—453.
Craft, Cassady, 1999. Weapons for Peace, Weapons for War: The Effects of Arms
Transfers on War Outbreak, Involvement, and Outcomes. New York: Routledge.
Daalder, Ivo H., and Michael E. O’Hanlon, 2000. Winning Ugly: NATO’s War to
Save Kosovo. Washington, DC: Brookings Institution Press.
Del Boca, Angelo, 1969. The Ethiopian War, 1935–1941. Chicago: University of
Chicago Press.
De Wet, Christiaan, 1902. The Three Years’ War. New York: Charles Scribner’s Sons.
Debray, Regis, 1968. Revolution in the Revolution. New York: Penguin.
Desch, Michael C., 2002. ‘‘Democracy and Victory: Why Regime Type Hardly
Matters,’’ International Security, vol. 27, no. 2: 5—47.
Donohue, Laura, 2001. Counter-terrorist Law and Emergency Powers in the United
Kingdom, 1922–2000. Portland, Or: Irish Academic Press.
Douhet, Giulio, [1927] 1972. The Command of the Air. New York: Arno Press.
1999. ‘‘Command of the Air,’’ in David Jablonsky, ed., Roots of Strategy: Book
Four. Mechanicsburg, PA: Stackpole Books, pp. 263—408.
Downes, Alexander, 2003. ‘‘Targeting Civilians in War: the Starvation Blockades
in World War I’’. Paper Presented at the 105th Annual Meeting of the
American Political Science Association, September, Philadelphia,
Pennsylvania.
Ellis, John, 1976. A Short History of Guerrilla Warfare. New York: St. Martin’s
Press.
236
References
1995. From the Barrel of a Gun: A History of Guerrilla, Revolutionary and Counter-
Insurgency Warfare, from the Romans to the Present. London: Greenhill Books.
Elman, Miriam Fendius, 1995. ‘‘The Foreign Policies of Small States:
Challenging Neorealism in its Own Backyard,’’ British Journal of Political
Science, vol. 25: 171—217.
Fremont-Barnes, Gregory, 2003. The Boer War: 1899–1902 (Oxford: Osprey
Publishing).
Galtung, Johan, 1976. ‘‘Two Concepts of Defense,’’ in Peace, War, and Defense:
Essays in Peace Research, vol. II. Oslo: International Peace Research Institute,
pp. 328—340.
Gammer, Moshe, 1994. Muslim Resistance to the Tsar: Shamil and the Conquest of
Chechnia and Daghestan. London: Frank Cass & Co.
Gast, Leon, 1996. When we were Kings. Polygram Film Productions.
George, Alexander L., David K. Hall and William R. Simons, eds., 1971, The
Limits of Coercive Diplomacy: Laos, Cuba, Vietnam. Boston: Little Brown and
Company.
Guevara, Che, 1961. Guerrilla Warfare. New York: Monthly Review Press.
Hamer, W. S., 1970. The British Army: Civil Military Relations, 1885–1905.
Oxford: Clarendon Press.
Hamilton, Donald W., 1998. The Art of Insurgency: American Military Policy and
the Failure of Strategy in Southeast Asia. Westport, CT: Praeger.
Hehn, Paul N., 1979. The German Struggle Against Yugoslav Guerrillas in World
War II: German Counter-Insurgency in Yugoslavia, 1941–1943. Boulder, CO:
Eastern European Quarterly.
Herring, George C., 1986. America’s Longest War: The United States and Vietnam,
1950–1975, 2nd ed. New York: McGraw Hill.
Hobson, J. A., 1900. The War in South Africa: Its Causes and Effects. London: James
Nisbet & Co.
Hoffmann, Stanley, et al., 1981. ‘‘Vietnam Reappraised,’’ International Security,
vol. 6, no. 1: 3—26.
Howard, Michael, 1961. The Franco-Prussian War. New York: Methuen.
Iklé, Fred Charles, 1991. Every War Must End, revised. New York: Columbia
University Press.
Independent International Commission on Kosovo, 2000. Kosovo Report:
Conflict, International Response, Lessons Learned. New York: Oxford
University Press.
Isby, David C., 1989. War in a Distant Country, Afghanistan: Invasion and
Resistance. London: Arms and Armour Press.
James, Anthony, 1996. Guerrilla Warfare: A Historical, Biographical, and
Bibliographical Sourcebook. Westport, CT: Greenwood Press.
Johnson, Chalmers, 1968. ‘‘The Third Generation of Guerrilla Warfare,’’ Asian
Survey, vol. 12: 435—447.
1973. Autopsy on Peoples War. Berkeley: University of California Press.
Johnson, Wray R., 2001. Vietnam and American Doctrine for Small Wars. Bangkok:
White Lotus Press.
237
References
Kakar, M. Hassan, 1995. Afghanistan: The Soviet Invasion and the Afghan Response,
1979–1982. Berkeley: University of California Press.
Karnow, Stanley, 1983. Vietnam: A History. New York: Viking Press.
Katzenbach, E. L., 1962. ‘‘Time, Space and Will: The Politico-Military Strategy of
Mao Tse-tung,’’ in Ltc. T. N. Greene, ed., The Guerrilla and How to Fight Him.
New York: Praeger, pp. 11—21.
Katzenbach, Edward L., and Gene Z. Hanrahan, 1955. ‘‘The Revolutionary
Strategy of Mao Tse-Tung,’’ Political Science Quarterly, vol. 70, no. 3:
321—340.
Kohn, George C., 1986. A Dictionary of Wars. New York: Facts on File Publications.
Krebs, Paula M., 1992. ‘‘‘The Last of the Gentlemen’s Wars’: Women in the Boer War
Concentration Camp Controversy,’’ History Workshop Journal, vol. 33: 38—56.
Krepinevich, Andrew F. Jr., 1986. The Army and Vietnam. Baltimore: Johns
Hopkins University Press.
Lake, David A., 1992. ‘‘Powerful Pacifists: Democratic States and War,’’
American Political Science Review, vol. 86, no. 1: 24—27.
Laqueur, Walter, 1976. Guerrilla: A Historical and Critical Study. Boston: Little,
Brown and Company.
Lawrence, T. E., 1926. Seven Pillars of Wisdom: A Triumph. New York: Doubleday.
Lea, Henry Charles, 1968. Superstition and Force: Essays on the Wager of Law, the Wager
of Battle, the Ordeal, Torture, 2nd revised ed. New York: Greenwood Press.
Leepson, Mark, ed., with Helen Hannaford, 1999. Webster’s New World
Dictionary of the Vietnam War. New York: Macmillan.
LeMay, Curtis R. with Major General Dale O. Smith, 1968. America Is in Danger.
New York: Funk & Wagnalls.
Liddell Hart, B. H., 1967. Strategy, 2nd revised ed. New York: Praeger Publishers.
Litwak, Robert S., 1992. ‘‘The Soviet Union in Afghanistan,’’ in Ariel E. Levite,
Bruce W. Jentleson, and Larry Berman, eds., Foreign Military Intervention:
The Dynamics of Protracted Conflict. New York: Columbia University Press,
pp. 65—94.
Mack, Andrew J. R., 1975. ‘‘Why Big Nations Lose Small Wars: The Politics of
Asymmetric Conflict,’’ World Politics, vol. 27, no. 2: 175—200.
Magnus, Ralph H., and Eden Naby, 1998. Afghanistan: Mullah, Marx, and
Mujahid. Boulder, CO: Westview Press.
Mao, Tse-tung, 1961. On Guerrilla Warfare, Samuel B. Griffith, tr. New York:
Praeger Publishers.
1968. ‘‘Primer on Guerrilla Warfare,’’ in Donald Robinson, ed., The Dirty Wars:
Guerrilla Actions and Other Forms of Unconventional Warfare. New York:
Delacorte Press.
Marquis, Susan L., 1997. Unconventional Warfare: Rebuilding US Special
Operations Forces. Washington, DC: Brookings Institution Press.
May, Ernest, 2000. Strange Victory: Hitler’s Conquest of France. New York: Hill &
Wang.
McCarthy, Ronald M., and Gene Sharp, 1997. Nonviolent Action: A Research
Guide. New York: Garland.
238
References
Mearsheimer, John, 1983. Conventional Deterrence. Ithaca, NY: Cornell University

Press.
2001. The Tragedy of Great Power Politics. New York: W.W. Norton & Company.
Merari, Ariel, 1993. ‘‘Terrorism as a Strategy of Insurgency,’’ Terrorism and
Political Violence, vol. 5, no. 4: 213—251.
Merom, Gil, 2003. How Democracies Lose Small Wars: State, Society, and the Failures
of France in Algeria, Israel in Lebanon, and the United States in Vietnam. New
York: Cambridge University Press.
1921. Our Air Force, the Keystone of National Defence. New York: E. P. Dutton &
Company. ‘‘Winged Defense,’’ in David Jablonsky, ed., Roots of Strategy:
Book Four. Mechanicsburg, PA: Stackpole Books, pp. 409—516.
Mockler, Anthony, 1984. Haile Selassie’s War: The Italian–Ethiopian Campaign,
1935–1941. New York: Random House.
Moisie, Edwin, E., 1996. Tonkin Gulf and the Escalation of the Vietnam War. Chapel
Hill, NC: University of North Carolina Press.
Mueller, John, 1980. ‘‘The Search for the ‘‘Breaking Point’’ in Vietnam: The
Statistics of a Deadly Quarrel,’’ International Studies Quarterly, vol. 24, no.
4: 497—519.
Muraise, Eric, 1964. Introduction à L’Histoire Militaire. Paris: Charles-Lavauzelle.
Mussolini, Benito, 1995. ‘‘The Vital Need for Empire,’’ in Roger Griffin, ed.,
Fascism. New York: Oxford University Press, pp. 74—75.
Nasson, Bill, 1999. The South African War, 1899–1902. New York: Oxford
University Press.
Nutter, John Jacob, 1994. ‘‘Unpacking Threat: A Conceptual and Formal
Analysis,’’ in Norman A. Graham, ed., Seeking Security and Development:
The Impact of Military Spending and Arms Transfers. Boulder: Lynne Rienner
Publishers, pp. 29—49.
Nye, Joseph S., Jr., 2004. Soft Power: The Means to Success in World Politics. New
York: Public Affairs.
Pakenham, Thomas, 1979. The Boer War. New York: Random House.
Pape, Robert Antony, 1990. ‘‘Coercive Air Power in the Vietnam War,’’
1996. Bombing to Win. Ithaca: Cornell University Press.
1997. ‘‘Why Economic Sanctions Do Not Work,’’ International Security, vol. 22,
no. 2: 90—136.
2003. ‘‘The Strategic Logic of Suicide Terrorism,’’ American Political Science
Review, vol. 97, no. 3: 343—361.
Paret, Peter, and John Shy, 1962. Guerrillas in the 1960s, 2nd ed. New York:
Praeger Publishers.
Parker, Christopher S., 1999. ‘‘New Weapons for Old Problems: Conventional
Proliferation and Military Effectiveness in Developing States,’’ International
Security, vol. 23, no. 4: 119—147.
Parker, Geoffrey, 1994. ‘‘Early Modern Europe,’’ in Michael Howard, George J.
Andreopoulos, and Mark R. Shulman, eds., The Laws of War: Constraints on
239
References
Warfare in the Western World. New Haven, CT: Yale University Press,
pp. 40—58.
Paul, T.V., 1994. Asymmetric Conflicts: War Initiation by Weaker Powers.
Cambridge: Cambridge University Press.
Perry, Mark, and Ed Miles, 1999. ‘‘Environmental Warfare,’’ in Roy Gutman
and David Rieff, Crimes of War: What the Public Should Know. New York:
W.W. Norton, pp. 132—135.
Pomeroy, William J., 1964. Guerrilla & Counter-Guerrilla Warfare: Liberation and
Suppression in the Present Period. New York: International Publishers.
Posen, Barry R., 2003. ‘‘Command of the Commons: The Military Foundation of
US Hegemony,’’ International Security, vol. 28, no. 1: 5—46.
Powell, Colin L., 1992. ‘‘US Forces: Challenges Ahead,’’ Foreign Affairs, vol. 71,
no. 5: 32—51.
Press, Daryl, 2001. ‘‘The Myth of Air Power in the Persian Gulf War and the
Future of Warfare,’’ International Security, vol. 26, no. 2: 5—44.
Rais, Rasul Bakhsh, 1994. War Without Winners: Afghanistan’s Uncertain
Transition after the Cold War. Karachi: Oxford University Press.
Ramakrishna, Kumar, 2002. ‘‘‘Bribing the Reds to Give Up’: Rewards Policy in
the Malayan Emergency,’’ War in History, vol. 9, no. 3: 332—353.
Reiter, Dan, and Allan C. Stam III, 1998. ‘‘Democracy and Battlefield Military
Effectiveness,’’ Journal of Conflict Resolution, vol. 42, no. 3: 259—277.
2002. Democracies at War. Princeton, NJ: Princeton University Press.
2003. ‘‘Understanding Victory: Why Political Institutions Matter,’’
Rhodes, Richard, 2002. Masters of Death: The SS-Einsatzgruppen and the Invention
of the Holocaust. New York: Alfred A. Knopf.
Roberts, Adam, 1994. ‘‘Land Warfare: From Hague to Nuremberg,’’ in Michael
Howard, George J. Andreopoulos, and Mark R. Shulman, eds., The Laws of
War: Constraints on Warfare in the Western World. New Haven, CT: Yale
University Press, pp. 116—139.
Rosen, Stephen P., 1982. ‘‘Vietnam and the American Theory of Limited War,’’
1995. ‘‘Military Effectiveness: Why Society Matters,’’ International Security,
vol. 19, no. 4: 5—31.
Rosen, Steven, 1972. ‘‘War Power and the Willingness to Suffer,’’ in Bruce M.
Russett, ed., Peace, War and Numbers. Beverly Hills: Sage Publications,
pp. 167—184.
Rudolph, Lloyd I., 1985. ‘‘The Great Game in Asia: Revisited and Revised,’’
Crossroads, no. 16: 1—46.
Russett, Bruce M., 1972. No Clear and Present Danger: A Skeptical View of the US
Entry into World War II. New York: Harper & Row.
Sbacchi, Alberto, 1985. Ethiopia Under Mussolini: Fascism and the Colonial
Experience. New York: Zed Books.
1997. Legacy of Bitterness: Ethiopia and Fascist Italy, 1935–1941. New Jersey: The
Red Sea Press.
240
References
Schaffer, Ronald, 1985. Wings of Judgment: American Bombing in World War II.
New York: Oxford University Press.
Schelling, Thomas C., 1966. Arms and Influence. New Haven, CT: Yale University
Press.
Sharp, Gene, 2003. There Are Realistic Alternatives. Boston, MA: The Albert
Einstein Institution.
Shaw, Geoffrey D. T., 2000. ‘‘Policemen versus Soldiers, the Debate Leading to
MAAG Objections and Washington Rejections of the Core of the British
Counter-Insurgency Advice,’’ Small Wars and Insurgencies, vol. 12, no. 2:
51—58.
Sheehan, Neil, 1989. A Bright Shining Lie: John Paul Vann and America in Vietnam.
New York: Vintage.
Sliwinski, Marek, 1989. ‘‘Afghanistan: the Decimation of a People,’’ Orbis, vol.
33, no. 1: 39—56.
Small, Melvin, and J. David Singer, 1983. Resort to Arms: International and Civil
Wars, 1816–1980. Beverley Hills: Sage Publications.
Smith, Iain R., 1996. The Origins of the South African War, 1899–1902. New York:
Longman.
Snyder, Glenn H., and Paul Diesing, 1977. Conflict among Nations: Bargaining,
Decision Making, and System Structure in International Crisis. Princeton:
Princeton University Press.
Snyder, Jack, 1991. Myths of Empire: Domestic Politics and International Ambition.
Ithaca: Cornell University Press.
Sorley, Lewis, 1999. A Better War: The Unexamined Victories and Final Tragedy of
America’s Last Years in Vietnam. New York: Harcourt Brace & Company,
1999.
Stigler, Andrew, 2003. ‘‘A Clear Victory for Air Power: NATO’s Empty Threat to
Invade Kosovo,’’ International Security, vol. 27, no. 3: 124—157.
Stubbs, Richard, 1989. Hearts and Minds in Guerrilla Warfare: The Malayan
Emergency 1948–1960. New York: Oxford University Press.
Summers, Harry G. Jr., 1995. Historical Atlas of the Vietnam War. Boston, MA:
Houghton Mifflin.
Taber, Robert, 1965. The War of the Flea: A Study of Guerrilla Warfare Theory and
Practice. New York: Lyle Stuart.
Tamarov, Vladislav, 1992. Afghanistan: Soviet Vietnam, Naomi Marcus, Marianne
Clarke Trangen, and Vladislav Tamarov, trs. San Francisco: Mercury
House.
Thomas, Ward, 2001. The Ethics of Destruction: Norms and Force in International
Relations. Ithaca, NY: Cornell University Press.
Thompson, James Clay, 1980. Rolling Thunder: Understanding Policy and Program
Failure. Chapel Hill, NC: University of North Carolina Press.
Thompson, Robert, 1966. Defeating Communist Insurgency: Experiences from
Malaya and Vietnam. London: Chatto & Windus.
Tin, Bui, 2002. From Enemy to Friend: A North Vietnamese Perspective on the War.
Annapolis, MD: Naval Institute Press.
241
References
Toft, Monica Duffy, 2003. The Geography of Ethnic Violence: Identity, Interests, and
the Indivisibility of Territory. Princeton, NJ: Princeton University Press.
Turtledove, Harry, 2001. ‘‘The Last Article,’’ in Harry Turtledove with Martin H.
Greenberg, eds., The Best Military Science Fiction of the 20th Century. New
York: Del Rey Books, pp. 231—262.
Tzu, Sun, 1988. The Art of War, Thomas Cleary, tr. Boston: Shambhala
Publications.
Urban, Mark L., 1988. War in Afghanistan. New York: St. Martin’s Press.
Walt, Stephen S., 1987. The Origins of Alliances. Ithaca: Cornell University Press.
2002. ‘‘American Primacy: Its Prospects and Pitfalls,’’ Naval War College
Review, vol. 55, no. 2: 9—28.
Walton, C. Dale, 2002. The Myth of Inevitable US Defeat in Vietnam. Portland, OR:
Frank Cass.
Waltz, Kenneth N., 1967. Foreign Policy and Democratic Politics: The American and
British Experience. Boston: Little, Brown & Company.
1979. Theory of International Politics. New York: McGraw-Hill.
Walzer, Michael, 2000. Just and Unjust Wars: A Moral Argument with Historical
Illustrations, 3rd ed. New York: Basic Books.
Watson, Bruce Allen, 1997. When Soldiers Quit: A Study in Military Disintegration.
Westport, CT: Praeger.
Weigley, Russell Frank, 1997. The American Way of War: A History of United States
Military Strategy and Policy. Bloomington, IN: Indiana University Press.
Wolf, Eric R., 1973. Peasant Wars of the 20th Century. New York: Harper & Row.
242
Index
Abrams, Creighton, General 157, 158, 163n Algeria see under France
Adowa, battle of (1896) 112, 112n, 116, 117n, Ali, Muhammad 23
118, 119 Allenby, General 96n
Afghan war (1979—89) 14, 19, 22 Aloisi, Pompeo, Baron 115, 116n
casualties 180, 181—182, 188, 188n, 193n Amin, Hafizullah 171—172, 176
ceasefire 183 Andreski, Stanislas 9n
(explanations of) outcome 6n, 187, Aosta, Duke of 127n, 127—128, 129—130,
188—189, 226 137, 138
international response 191—192, 192n Ap Bac, battle of (1963) 149
mujahideen aims/interests 174—175, 190, Apocalypse Now (Coppola, 1979) 160—161
196, 213 Arab—Israeli conflicts 224
mujahideen (imported) resources 178n, Ardagh, Sir John, Maj.-Gen. 80
186, 193—194, 198, 215n, 215—216 Aregai, Abebe 128, 129, 142, 210
progress of hostilities 175—181, 183—184, arms diffusion theory 10—13, 11n, 42, 202
202 applicability to case studies 193—194,
Soviet aims/interests 172n, 172—173, 197, 212, 214—215
173n, 189—190, 195—196, 213, 214n inapplicability/irrelevance to case
Soviet resources , 172n, 178—179, studies 67—68, 70, 101—102, 106,
183, 196 136—137, 141, 162—163, 206, 209
Soviet strategy/tactics 179, 180n, artillery, use of 67—68, 70, 101—102
180—183, 184—187, 189—190, 198, 221 Asian countries, common strategic
theoretial analysis 189—199, 213—216, preferences 37
214n, 217—219 Asquith, H. H., prime minister 87—88
Afghanistan asymmetric conflicts 2—3
ethnography 170 defined 43
geography 170, 177, 185 duration 27, 28—29, 35, 46, 164, 202, 205,
political history 170—172, 173—174 207—208, 213
post-1989 developments 175n, 188, 190, frequency 20n, 200—201
199, 216 initiation 30n
(reductions in) population 182 theoretical analyses 5, 6n, 20, 24—25
US invasion (2001) 19—20, 166, 204—205 (see also arms diffusion; interest
air attack 31n, 32n, 40—41, 161 asymmetry; strategic interaction)
counter-measures 186—187, 194 unexpected outcomes 21, 157—158
effectiveness 40n, 136—137 (see also weaker side, victory of)
Akhulgo, battle of 59 Athens 2
Alexander I, tsar 52 Auraris, Dejaz 128n
Alexander II, tsar 54, 64, 66, 206n authoritarian regimes
Alfieri, Dino 133 conduct of hostilities 66—67
243
Index
authoritarian regimes (cont.) Broderick, Lord 91—92n
control of information 7, 8—9, 28, Bui Tin 161
135—136, 160, 214 Buller, Sir Redvers (Lord), General 86,
decline in numbers 5, 8 89—90
demands on soldiery 7—8, 9 Bullock, Colonel 103n
local competition 211—212 Bush, George (sr.), President 164
military advantages 7—8, 27—28, 42, 68n,
196, 201—202, 209, 214n, 217—219 Calley, William, Lt. 9n
military disadvantages 8—10, 140, 214 Cape Town 72—73, 99
wartime relaxation 8 Caputo, Philip 144
Castellano, Professor 124n
Baddeley, John F. 52—53, 67 Castro, Fidel 58n
Baden-Powell, Lord 86 casualties 8, 10
Badoglio, General 121, 122—123, 122—123n Catherine the Great 51, 69
bandwagoning 58n Caucasus
Bao Dai 147 geography 49—51, 68
Baratieri, Oreste, General 111—112 political organization 55 (see also
barbarism Muridism)
aims 34 population 57
counterproductivity 17n, 58, 63, 92, 108, Russian conquest of see Murid Wars
126, 127n, 197—198, 221, 225 Cavallero, General 128, 128n
defined 31—32, 43 cavalry, use of 85n
drawbacks 9—10, 35—36, 41—42n, 213, Chamberlain, Joseph 76, 77, 99—100
214n, 215—216 (see also Chechnia
counterproductivity above) deforestation 50, 50n, 61, 62, 63, 68
effect on participants 9n geography 50
employment in case studies 53, 70, Russian attacks on 53, 58, 166, 214n
92—94, 107, 143, 153—154, 157, 175, chemical weapons 31n, 182—183, 183n, 210n
180—183 see also mustard gas
as end in itself 36n Churchill, Winston 129
international response 222 Ciano, Galeazzo 129—130
military effectiveness 27—28n, 35n, 194—195 civilians
public reactions 100, 217—219 (see also casualty figures 162—163n
concentration camps; public opinion) execution 31n
reluctance to use 15—16 targeting 8, 25
use by authoritarian regimes 16, 191, 196 see also barbarism; concentration camps
use by democracies 16 clemency see conciliation
see also air attack; chemical weapons; Cohen, Eliot 11, 12, 166, 226
concentration camps; hospitals, COIN (counterinsurgency) campaigns/
bombing of; indirect attacking strategies 31—32, 32n, 41, 102—103, 107,
strategy; mines; mustard gas 155—157, 166—167, 191, 215, 226
Bariatinsky, Prince 57, 62, 64, 66, 68 Cold War 6—7n, 11—12
Beit, Alfred 76, 79n collateral damage 181—183
Belgium, German invasion (1914) 224 colonialism
‘‘Black Week’’ (1899) 86—87 justifications 103n, 116—117, 118n
Blanch, Lesley 63 struggles against 4n, 37n, 41—42n,
Blaskowitz, Johannes 9—10 146—148
blockhouse strategy 96, 184n, 184—185 commando units, use of 83, 96—97
Blood River, battle of (1838) 73 concentration camps 31n, 32n, 91, 93—94,
Boer War see South African War 102—103
bombing see air attack conditions 95n
Botha, Louis 86, 91—92, 93, 105 death rates 94n, 103
bravery, role in battle 49 justification 91—92n
Brezhnev, Leonid 171, 190, 191—192 public reaction to 94—96, 106, 222
Brezhnev Doctrine 172, 172n, 173 conciliation 30n, 62—63, 70—71, 138, 210, 211
244
Index
Conrad, Joseph 72 Eritrea 111, 113—114, 119n, 119—120, 120n
Constantine, Grand Duke (brother of Ethiopia
Alexander I) 53—54 geography 110
conventional attack 34 political system 110—111, 118
conventional responses 38, 102, 137, see also Italo-Ethiopian War
154—155, 203—204 European countries, common strategic
defined 30—31, 43 preferences 37
indirect responses 38—39, 64, 102, 138, expanding bullets, prohibition of 103n
175, 194, 204
predicted outcomes 38, 39 fait accompli, victory as 26, 116n
conventional defense 32, 34, 43, 221 Falkland Islands 164
see also under conventional attack; Fawcett, Millicent 95n
indirect attacking strategies Flandin, Pierre-Etienne 135
Craft, Cassady 11n foreign support (for weaker side) 45—46,
Crimean War 66—67 186, 193
Cronje, General 88—89 dependence on 195
Czechoslovakia pitfalls 195
German invasion 128 France
Soviet invasion 191 German invasion (1940) 224
involvement in Algeria 42n, 168, 184n,
Daghestan 197n, 225
geography 50 involvement in Vietnam 144, 146n,
leadership 55 146—148, 148n, 184n
Russian attacks on 53 military strategies 16, 37n
Da’ud Khan, Sardar Mohammed 170—171, relations with Italy 112—113, 116—117
188 franchise, as point of dispute 75n
De Bono, General 119, 120—121, 122, 122—123n Franco-Prussian War 39n, 91
De La Rey, Koos, General 92, 93, 97, 105 Freedman, Lawrence 200
De Wet, Christiaan, General 88—89, 90—91, Freitag, General 57, 60, 61
92—93, 97, 105
Debra Libanos monastery, massacre at 127 Gammer, Moshe 57n, 61, 67
deforestation, as military tactic Gandhi, Mahatma 41n
see Chechnya Geneva Convention see laws of war
Del Boca, Angelo 110, 114 genocide 222n
democracy/ies geography, influence on combat strategies
authoritarian measures 8 12n, 178
military effectiveness 15—18, 28, 166, Georgia 69
201—202, 217—219, 225n Gladstone, W. E., prime minister 74—75
strategic choices 99—101, 161—162, gold/gems, discovery of 75, 79n
207—208, 211—212 Goldwater, Barry, Senator 151
Diem, Ngo Dinh 148—149, 149—150, Gorbachev, Mikhail 186, 187, 190, 192—193,
151, 212 215
Dien Bien Phu, battle of 144, 167n Grabbé, General 57
divine will (victory as expression of) 26—27 grand strategy, defined 29—30
domino theory 82n, 164 Graziani, General 122—123, 122—123n,
Donohue, Laura K. 225n 126—127, 128, 130—131, 137
Douhet, Gulliano 40 Great Trek 73
Downes, Alexander 35n Grenada 26
Duffy Toft, Monica 78n Grozny, construction of 53
Dundonald, Earl, Brig.-Gen. 107 guerrilla warfare 12n, 14—15n, 16, 25n, 37,
132—133
economic sanctions 40, 40n aims 34, 35n
Eden, Anthony 135 conditions (un)favorable to 41, 62n, 105,
Egypt 130 197
Eisenhower, Dwight D. 170 defined 32—33, 43
245
Index
guerrilla warfare (cont.) internment 25
employment in case studies 57, 59—61, Iraq, conflicts involving 19, 20, 162, 202
68, 70—71, 88—89, 90—92, 93, 102, see also Kuwait
106—107, 121—122, 138, 155—156, 175 Israel 225, 227
evolution into direct strategy 32n Italo-Ethiopian Treaty of Amity,
origins 17n Conciliation and Arbitration (1928)
rejection 39n, 125, 131, 166 113—114
Guevara, Che 58n, 203n Italo-Ethiopian War (1935—40) 22, 28, 198n
Gugsa, General 120 casualties 125n
conclusion 133n
Hague Convention (1899/1907) 92n, 103n Ethiopian aims/interests 118, 132—133
Haile Selassie, Emperor 109, 109—110, 113n, Ethiopian resources 120, 136, 141
113—114, 118, 119, 120, 121—122, Ethiopian strategy/tactics 121—122, 125
125—126, 130, 130n, 131, 132—133, historical background 111—116
209, 220 Italian aims/interests 116—118, 119, 132,
Halifax, Lord 134—135 140
Hamilton, Donald 166 Italian malpractice (see hospitals; laws of
Hamzad Beg 55, 64 war; mustard gas)
Hannibal 49 Italian strategy/tactics 123—125, 126—129,
Hely-Hutchinson, Sir Walter 83 140—141
Himmler, Heinrich 9—10 progress of hostilities 109—110, 119—129
Hitler, Adolf 49 theoretical analysis 132—143, 209—211, 219
Ho Chi Minh 146—147, 149 Italy
Hoare, Sir Samuel 116—117n in World War II 129—131
Hobhouse, Emily 94—95, 95n, 100, 219 see also Italo-Ethiopian War
Hobson, J.A. 79n
honor, role in military ethos 49, 118 Jahandad, Commander 176
hospitals, bombing of 124, 134 Jameson Raid 76, 101
Hungary, Soviet invasion of 191, 202 Japan, military initiatives 52, 147, 223—224
Hussein, Saddam 50n see also Russo-Japanese War
Johnson, Lyndon B., president 29n,
incompetence, role in strong-actor setbacks 150—152, 153
68, 70, 131, 137, 138—140, 210—211, 220 Joubert, General 85
India, gaining of independence 170
indirect attacking strategies Kakar, Hassan 169, 174, 176, 181
direct responses 39—41, 154, 204 Karmal, Babrak 171, 172, 173, 176, 177, 186,
indirect responses 33—34n, 41, 137, 189
203—204 Karnow, Stanley 146
predicted outcomes 40—41, 42 Kekevich, General 86
by weaker side 40n Kennedy, John F., president 150—151
see also air attack; barbarism; civilians, Khazi Muhammad 55, 58, 64
targeting of; economic sanctions Kimberley, siege of (1899—1900) 86, 88
indirect defensive strategies 27—28, Kitchener, Lord 86, 88, 89, 93—94, 96—97, 98,
221—222, 224 103n, 207—208n
see also conventional attack; guerilla Kock, General 84—85, 85n
warfare; indirect attacking strategies Korean War (1950—3) 11—12, 45n, 220
interest asymmetry, theory of 13—14, 25, Kosovo 19, 33—34n, 40n, 41n, 198n
42—43, 201 Krepinovich, Andrew 166
applicability to case studies 66, 69 Kruger, Paul, president 74—75, 76, 77—78,
inapplicability to case studies 104, 159, 80—81, 84n, 93, 101
190, 196, 213—214, 217—219 Kuwait, Iraqi invasion/expulsion 19, 49, 164
weaknesses 14—15
international relations theory Labouchère, Henry 88n
(conventional approaches) 2, 2n, 4—5, Ladysmith, siege of (1899—1900) 85n,
23—24, 221, 223, 223n 85—86, 88
246
Index
Lansdowne, H. C. K. Petty-Fitzmaurice, Muridism 55, 58, 224
marquess of 79 military organization 56, 70
Lawrence, T. E. 23, 130, 185 religious dogma 56
laws of war, violations of 31n, 103n, Mussolini, Benito 112n, 113—114, 116n,
103—104, 160—161, 176 116—118, 117n, 120—121, 122, 125n, 126,
alleged 124n 127, 128, 129, 132, 133, 135—136, 137,
attempts at concealment/justification 140—141, 172n
124, 124n, 127, 133—136 mustard gas, use of 122—125, 122—123n,
see also chemical weapons; hospitals, 125—126, 131—132, 133—136, 137
bombing of; mustard gas attempts at concealment/justification
League of Nations 113, 114, 115—116, 118, see under laws of war, violations of
132—133, 134—135 effects on victims 123—124
Lloyd George, David, PM 95 My Lai massacre 9n, 160—161
Mack, Andrew J. R. 5—7, 13, 14—15n, 16, Najibullah, Mohammed 186, 188, 188n, 190
17—18, 24, 25n, 27, 28—29, 161, 165, 201, Napoleonic wars 17n, 49, 52, 60n
202, 202n nationalism 5, 6—7n, 37n, 62n, 195,
Mafeking, siege of (1899—1900) 86, 88 223—224
Magsaysay, Ramon 167 NATO, military activities 19, 31n
Majuba Hill, battle of (1881) 75 nature-of-actor theory see authoritarian
Malayan Emergency (1948) 17, 202n, 226 regimes; democracies
Manchuria see Japan naval warfare, strategies 50n
Mao Tse-tung 12n, 32n, 33, 34, 37, 41n, Navarre, Henri, General 167n
223—224 Nazis, treatment of occupied territories
Massu, General 42n, 168 9—10, 35—36, 36n
Massud, Ahmad Shah 179, 180, 181, Nguyen Co Thach 157n
183—184, 185, 188, 194—195 Nguyen Khanh, General 150
McNamara, Robert S. 29n, 153 Nicholas I, Tsar 53—54, 59, 64, 66, 70
Mearsheimer, John 3n noncombatants, targeting of 31—32
Menelik II, emperor of Ethiopia 111—112, 118 see also civilians
Merom, Gil 15, 16—18, 29, 67, 161—162, 166, Northern Ireland 225n
201—202, 202n, 206n, 207—208n, 225n,
226—227 Orange Free State 73—74, 88n, 98
Methuen, Lord 86
Meyer, Lucas 83—84 Pakenham, Thomas 79, 82, 85—86, 87—88, 98
Milner, Alfred 76—77, 77n, 80, 83n, 98, Pakistan 170
99—100 see also Afghan War: mujahideen
Milosevic, Slobodan 40n resources
mines, use of 182 Paktia offensive (1980) 178—179
Mitchell, William 40 Panjsher offensives (1980—5)179—181,
Molotov, Vyacheslav 148 183—184
motivation (levels of) 24 Pape, Robert 40, 40n, 152—153n
Movchan, Mickola 169 Paskyevitch, General 57—58, 59, 63, 68
mujahideen see Afghan war; foreign Patriot Act (US 2002) 8
support Paul, T. V. 6n
Murid War (1820—49) 22 Paul I, Tsar 51—52
cost 49 Persia, conflicts with Russia 51—52, 66
historical background 51—54 Peter the Great 51, 69
outcome 64, 206n Philippines 202n, 226
progress of hostilities 57 Phoenix program 156—157, 157n
Russian aims/interests 65—66 Poland, German occupation of 9—10
Russian military strategy 50—51, 53, political vulnerability, impact on military
58—59, 63—64 strategy 13—14, 24, 25, 42—43
theoretical analysis 64—71, 205—207, 206n, absence of 66, 67, 69—70, 141, 206, 214
219—220 in Soviet Union 193
247
Index
political vulnerability, impact on military social structure, relationship with military
strategy (cont.) effectiveness 6—7n
in UK 100, 105—106, 207—208 see also democracies; public opinion
in United States 159, 161—162, 165, 213 ‘‘socialization’’ (of strategy) 36—37, 205
in Vietnam 211—212 Sokolov, Marshal 177
Posen, Barry 12—13 Sorley, Lewis 163n
posts, fortified see blockhouse strategy South African War 20, 22, 43—44n, 44n,
power 81—82
defined 2n, 3n, 223, 223n, 224 Boer aims/interests 80n, 80—81, 99, 104
relationship with conflict outcomes 2—3, Boer strategy/tactics 33n, 82, 84—85,
63, 65, 69, 71, 140 90—92, 107n
see also interest asymmetry British aims/interests 76—80, 99, 207
precedent, setting of (as factor in British strategy/tactics 32n, 87—88, 89,
determining strategy) 78—79, 92—94, 184n, 185, 220, 222
116—117n, 173n, 189n casualties 98 (see also under concentration
see also domino theory camps)
prisoners of war historical background 72—76
inability to hold 97 outcome 97—98, 105, 108
mistreatment 8 pre-war negotiations 80—81
refusal to take 183 progress of hostilities 82—94, 96—97
public opinion, impact on military strategy theoretical analysis 98—108, 207—208,
78, 87, 94—96, 99—101, 105—106, 207—208n, 217
161—162, 209n uniforms 103n
South Vietnam, political regime 148—150
racial issues, role in South African war 77n Soviet Union 5, 8
Ras Imru, General 123, 126 dissolution 188
Ras Kassa, General 122 involvement in Afghanistan (pre-1979)
Ras Mulugetta, General 122 171—172
Ras Seyum, General 122 military strategies/traditions 37, 38,
Rhodes, Cecil 74, 76, 76n, 79n, 86, 100 178n, 191, 192—193, 193n
Rhodes, Richard 9—10 political system 190—193
Ridgeway, Matthew, General 220 see also Afghan War; Cold War
Roberts, Lord 86, 88, 88n, 89, 90, 90n, 92—93, Sparta 2
94, 100, 107 Stalin, Joseph 8
Roghe, Bruno 133 Steer, George 130—131, 131n
Rolling Thunder, Operation (1965—68) Steyn, Marthinus, President 76, 80—81, 82,
152—153n, 154n, 156—157 93, 101, 107n
Russia Stinger missiles 186—187, 194
activities post-1992 166 Strategic Hamlets program 156
political organization 54—55, 65, 66—67 strategic interaction thesis 6—7, 18, 21,
population 57 24—25, 27, 29, 44—45n, 222—223,
see also Murid War; Soviet Union 224—227
Russo-Japanese War (1905) 223 application to case studies 63—64, 65, 68,
Rwanda 227 70—71, 102—104, 106, 137—143, 163—166,
194—195, 197—199, 206—207, 208,
Sbacchi, Alberto 141n 209—211, 212—213, 215—216
Schwarzkopf, Norman 49 central hypotheses 42, 46—47, 203—204,
September 11 attacks 227 217—219
Shamil (Murid leader) 49—50, 52, 54, 55—56, empirical evaluation 43—47, 204, 216—219
58—61, 60n, 64, 70, 206 exceptions 112n
downfall 62—63, 67, 68 limitations 223—224
Shepstone, Sir Theophilus 74 strategy/ies
Sherman, William, General 91 choice of 29n, 37—38, 200, 203, 203n, 209n,
Smith, Iain R. 79, 81 221—222
Smuts, Jan 78, 82, 91, 92, 93, 97, 105 counterstrategies 34—35
248
Index
strategy/ies (cont.) parliamentary proceedings 91—92n,
defined 29, 29n 95—96, 116—117n
limited aims 32n, 159 relations with Italy 112—113
switches 36, 36n, 43—44n, 44n, 70, 102, 116—117
106—107, 212 see also Malayan Emergency; South
types 30, 34—35, 203, 203n African War
see also conventional attack; conventional United States
defense; guerrilla warfare; indirect domestic policy/legislation 8, 25
attacking strategies; strategic foreign policy 19—21, 26, 192
interaction thesis military effectiveness 12—13
Suez Canal 99 military history/strategies 37, 38, 212
Sun Tzu 48 military strategy, ideal/future 225—227
survival, as reason for going to war 25, see also Afghanistan; Iraq; Korean war;
164—165 Kuwait; Vietnam war; World War II
Suvorov, Alexander 60, 219 unpreparedness, role in strong-actor
Sweden, war with Russia 51 setbacks 219n, 219—221
Symons, Sir Penn 79, 83—84 see also United Kingdom: military history
Urban, Mark L. 181, 187, 197
tactics, defined 29—30
Tajikistan 189n Veliaminov, General 48, 54
Talana Hill, battle of (1899) 83—84 Vietnam
Tamarov, Vladislav 193n Democratic Republic of (DRV) 147,
Taraki, Muhammad 171—172 148—149, 150, 157, 158, 159—161
Tarzi, Mahmud Beg 192 economy 145
technology (military) 4n, 5, 187n geography 145
relationship with combat effectiveness political history 145—148, 146n
11—13, 67—68, 217 see also France; Vietnam War
spread of 11, 12n Vietnam war (1965—73) 22, 44n, 45n, 178n
see also arms diffusion civilian casualties 158n
Tembian, battle of 122 (explanations of) outcome 6n, 14, 157,
terrorism 32n 158n, 226
countermeasures 225, 226n progress of hostilities 152—157, 167—168,
war on 21, 38 202
Tet Offensive (1968) 157, 163, theoretical analysis 43—44n, 157—168,
163n, 195 211—213
Thatcher, Margaret, PM 164 US aims/interests 150—152, 151n, 159,
Thucydides 2 214n
time, significance to military objectives US engagement in 14, 144—145,
106, 164—165 151—152
see also asymmetric conflicts: duration US strategy/tactics 16, 29n, 40—41n, 50n,
Tonkin Gulf incident (1964) 151n 155—157, 164, 167—168, 189, 198n,
Tran Hung Doo 146 220—221
Transvaal 73, 74—75, 98 Vietnamese aims/interests 158
Trezzani, General 138n Vietnamese resources 162—163, 166
Turtledove, Harry 41n Vietnamese strategy/tactics 154—155,
160—161
Umberto, King of Italy 111—112 Vittorio Emmanuele II of Italy 113
underestimation of opposition Voronzov, Count 57, 59—61, 60n, 62—63,
by Russia/Soviet Union 69n, 191 66, 68
by UK 79, 81, 87, 89—90, 100
United Kingdom Wal Wal, Ethiopia 114—116
involvement in Ethiopia 119n, 129—131, Waltz, Kenneth 18, 36—37
130n, 142, 210 Walzer, Michael 168
military history/limitations 37n, 79n, Watson, Bruce 9n
85—86, 101—102 Watson, C. Dale 17n
249
Index
weaker side, victory of duration 164
frequency 3—4, 43—45 impact on war on Ethiopia 129, 131, 142,
historical trends 4n, 4—5, 36, 46, 47n, 205 210
reasons 6—7n, 16—17n motives for engagement 8, 25
Westmoreland, William 167n
White, Sir George, General 83, 84—86 Yermolov, Mikhail 50n, 52—54, 63—64, 68
Wilson, Woodrow, President 146—147 Yevdomikov, General 57
Wingate, Orde, Major 142 Yohannes IV, emperor of Ethiopia 111
World War I 210n, 219n Yugoslavia 36n
World War II
aftermath 10—11, 12, 17, 18, 37, Zahir Shah, King 171
41—42n Zaitsev, Mikhail 186, 187
Allied strategies 25, 32n, 40—41n Zhou Enlai 148
commencement 128 Zulus, conflicts involving 74
250
CAMBRIDGE STUDIES IN INTERNATIONAL RELATIONS
89 Patrick M. Morgan
Deterrence now
88 Susan Sell
Private power, public law
The globalization of intellectual property rights
87 Nina Tannenwald
The nuclear taboo
The United States and the non-use of nuclear weapons since 1945
86 Linda Weiss
States in the global economy
Bringing domestic institutions back in
85 Rodney Bruce Hall and Thomas J. Biersteker (eds.)
The emergence of private authority in global governance
84 Heather Rae
State identities and the homogenisation of peoples
83 Maja Zehfuss
Constructivism in International Relations
The politics of reality
82 Paul K. Huth and Todd Allee
The democratic peace and territorial conflict in the twentieth
century
81 Neta C. Crawford
Argument and change in world politics
Ethics, decolonization and humanitarian intervention
80 Douglas Lemke
Regions of war and peace
79 Richard Shapcott
Justice, community and dialogue in international relations
78 Phil Steinberg
The social construction of the ocean
77 Christine Sylvester
Feminist International Relations
An unfinished journey
76 Kenneth A. Schultz
Democracy and coercive diplomacy
75 David Houghton
US foreign policy and the Iran hostage crisis
74 Cecilia Albin
Justice and fairness in international negotiation
73 Martin Shaw
Theory of the global state
Globality as an unfinished revolution
72 Frank C. Zagare and D. Marc Kilgour
Perfect deterrence
71 Robert O’Brien, Anne Marie Goetz, Jan Aart Scholte and Marc Williams
Contesting global governance
Multilateral economic institutions and global social movements
70 Roland Bleiker
Popular dissent, human agency and global politics
69 Bill McSweeney
Security, identity and interests
A sociology of international relations
68 Molly Cochran
Normative theory in international relations
A pragmatic approach
67 Alexander Wendt
Social theory of international politics
66 Thomas Risse, Stephen C. Ropp and Kathryn Sikkink (eds.)

The power of human rights
International norms and domestic change
65 Daniel W. Drezner
The sanctions paradox
Economic statecraft and international relations
64 Viva Ona Bartkus
The dynamic of secession
63 John A. Vasquez
The power of power politics
From classical realism to neotraditionalism
62 Emanuel Adler and Michael Barnett (eds.)
Security communities
61 Charles Jones
E. H. Carr and international relations
A duty to lie
60 Jeffrey W. Knopf
Domestic society and international cooperation
The impact of protest on US arms control policy
59 Nicholas Greenwood Onuf
The republican legacy in international thought
58 Daniel S. Geller and J. David Singer
Nations at war
A scientific study of international conflict
57 Randall D. Germain
The international organization of credit
States and global finance in the world economy
56 N. Piers Ludlow
Dealing with Britain
The Six and the first UK application to the EEC
55 Andreas Hasenclever, Peter Mayer and Volker Rittberger
Theories of international regimes
54 Miranda A. Schreurs and Elizabeth C. Economy (eds.)
The internationalization of environmental protection
53 James N. Rosenau
Along the domestic– foreign frontier
Exploring governance in a turbulent world
52 John M. Hobson
The wealth of states
A comparative sociology of international economic and political
change
51 Kalevi J. Holsti
The state, war, and the state of war
50 Christopher Clapham
Africa and the international system
The politics of state survival
49 Susan Strange
The retreat of the state
The diffusion of power in the world economy
48 William I. Robinson
Promoting polyarchy
Globalization, US intervention, and hegemony
47 Roger Spegele
Political realism in international theory
46 Thomas J. Biersteker and Cynthia Weber (eds.)
State sovereignty as social construct
45 Mervyn Frost
Ethics in international relations
A constitutive theory
44 Mark W. Zacher with Brent A. Sutton
Governing global networks
International regimes for transportation and communications
43 Mark Neufeld
The restructuring of international relations theory
42 Thomas Risse-Kappen (ed.)
Bringing transnational relations back in
Non-state actors, domestic structures and international institutions
41 Hayward R. Alker
Rediscoveries and reformulations
Humanistic methodologies for international studies
40 Robert W. Cox with Timothy J. Sinclair
Approaches to world order
39 Jens Bartelson
A genealogy of sovereignty
38 Mark Rupert
Producing hegemony
The politics of mass production and American global power
37 Cynthia Weber
Simulating sovereignty
Intervention, the state and symbolic exchange
36 Gary Goertz
Contexts of international politics
35 James L. Richardson
Crisis diplomacy
The Great Powers since the mid-nineteenth century
34 Bradley S. Klein
Strategic studies and world order
The global politics of deterrence
33 T. V. Paul
Asymmetric conflicts: war initiation by weaker powers
32 Christine Sylvester
Feminist theory and international relations in a postmodern era
31 Peter J. Schraeder
US foreign policy toward Africa
Incrementalism, crisis and change
30 Graham Spinardi
From Polaris to Trident: The development of US Fleet Ballistic
Missile technology
29 David A. Welch
Justice and the genesis of war
28 Russell J. Leng
Interstate crisis behavior, 1816–1980: realism versus reciprocity
27 John A. Vasquez
The war puzzle
26 Stephen Gill (ed.)
Gramsci, historical materialism and international relations
25 Mike Bowker and Robin Brown (eds.)
From Cold War to collapse: theory and world politics in the 1980s
24 R. B. J. Walker
Inside/outside: international relations as political theory
23 Edward Reiss
The Strategic Defense Initiative
22 Keith Krause
Arms and the state: patterns of military production and trade
21 Roger Buckley
US-Japan alliance diplomacy 1945–1990
20 James N. Rosenau and Ernst-Otto Czempiel (eds.)
Governance without government: order and change in world
politics
19 Michael Nicholson
Rationality and the analysis of international conflict
18 John Stopford and Susan Strange
Rival states, rival firms
Competition for world market shares
17 Terry Nardin and David R. Mapel (eds.)
Traditions of international ethics
16 Charles F. Doran
Systems in crisis
New imperatives of high politics at century’s end
15 Deon Geldenhuys
Isolated states: a comparative analysis
14 Kalevi J. Holsti
Peace and war: armed conflicts and international
order 1648–1989
13 Saki Dockrill
Britain’s policy for West German rearmament 1950–1955
12 Robert H. Jackson
Quasi-states: sovereignty, international relations and the Third
World
11 James Barber and John Barratt
South Africa’s foreign policy
The search for status and security 1945—1988
10 James Mayall
Nationalism and international society
9 William Bloom
Personal identity, national identity and
international relations
8 Zeev Maoz
National choices and international processes
7 Ian Clark
The hierarchy of states
Reform and resistance in the international order
6 Hidemi Suganami
The domestic analogy and world order proposals
5 Stephen Gill
American hegemony and the Trilateral Commission
4 Michael C. Pugh
The ANZUS crisis, nuclear visiting and deterrence
3 Michael Nicholson
Formal theories in international relations
2 Friedrich V. Kratochwil
Rules, norms, and decisions
On the conditions of practical and legal reasoning in international
relations an domestic affairs
1 Myles L.C. Robertson
Soviet policy towards Japan
An analysis of trends in the 1970s and 1980s

How The Weak Win Wars - A Theory of Asymmetric Conflict

Uploaded by

Copyright:

Available Formats

How The Weak Win Wars - A Theory of Asymmetric Conflict

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

How The Weak Win Wars - A Theory of Asymmetric Conflict

Uploaded by

Copyright:

Available Formats

How the Weak Win Wars

Cambridge Studies in International Relations is a joint initiative of

100 Michael C. Williams

Series list continued after index

Published in the United States of America by Cambridge University Press,

# Ivan Arreguı́n-Toft 2005

This book is in copyright. Subject to statutory exception

First published 2005

Printed in the United Kingdom at the University Press, Cambridge

Library of Congress Cataloging-in-Publication Data

ISBN-13 978-0-521-83976-1 hardback

Cambridge University Press has no responsibility for

Do not press a desperate enemy

List of figures page x

1 Percentage of asymmetric conflict victories by

This is a book about power, and how common understandings about

I have read many of these acknowledgments sections over the years.

ARVN Army of Vietnam

Why do the strong lose to the weak?

Relative power and realist international relations

100 strong actors

Figure 1. Percentage of asymmetric conflict victories by type of actor,

should always win.4 Indeed, a review of all asymmetric wars fought

Figure 2. Percentage of conflict victories by type of actor over time

authoritarian strong actors win asymmetric conflicts but democratic

Competing explanations of asymmetric conflict

will be reduced to testable propositions which will later be compared in

The nature of the actor

by threatening to execute them for ‘‘cowardice,’’ where cowardice is

Problems with the argument

their government’s wars unjust, when there’s no way to know, a priori,

This is why authoritarian regimes who resort to this strategy almost

The empirical evidence appears to support this logic, because the

Problems with the argument

advantages in open terrain, but less effective in closed, and, in particular,

asymmetry is the key causal mechanism and Mack’s argument is in this

Problems with the argument

Third, if the interest asymmetry thesis is right there should be little or

Democratic social squeamishness

Merom adds that ‘‘Essentially, what prevents modern democracies

structurally constrained in both categories. The particular relationship

Problems with the argument

charismatic representatives, to alter the state’s foreign policy or military

Why study asymmetric conflict outcomes?

Plan of the book

contains multiple strategic interactions. Thus, although only five his-

[T]aking practical account of the area we wished to deliver . . . I began

Actors in a conflict of interests each come to that conflict with three

This conventional wisdom is a problem, however. First, relying on it

Explaining asymmetric conflict outcomes

strategies (which then alter the likelihood of victory or defeat). Third,

Interests and vulnerabilities

Political vulnerability and conflict duration

with legitimacy, especially when there is a gross mismatch in relative

Political vulnerability and regime type

It is a weak argument, and Mack’s discussion previous to this comment

two advantages over their democratic counterparts. First, authoritarian

Vulnerability and strong actor defeat