World Politics 54 (October 2001), 99–118
Research Note
“NEW” AND “OLD” CIVIL WARS
A Valid Distinction?
By STATHIS N. KALYVAS*
THE decline of interstate armed conﬂict and perceived rise in the
frequency of civil wars since the end of the cold war1
—especially in
Europe2
—have contributed to a new wave of interest in civil wars.3
This
interest focuses on ethnic competition as a source of conﬂict and widely
regards civil wars of the post–cold war era (“new” civil wars) as fundamentally
different from their predecessors (“old” civil wars); “new” civil
wars are distinguished as criminal, rather than political, phenomena.
Since the issue of ethnic competition has been effectively tackled by
recent research,4
this article instead challenges the distinction between
“new” and “old” civil wars by arguing that the tendency to see fundamental
differences between them is based on an uncritical adoption of
categories and labels grounded in a double mischaracterization. On the
one hand, information about recent or ongoing wars is typically incomplete
and biased; on the other hand, historical research on earlier wars
tends to be disregarded. This is compounded by the fact that the end of
the cold war has robbed analysts of the clear categories that had made
possible an orderly, if ultimately ﬂawed, coding of civil wars. Accordingly,
the distinction drawn between post–cold war conﬂicts and their
predecessors may be attributable more to the demise of readily available
conceptual categories than to the existence of profound differences.
* The author thanks Pierre Hassner, Soﬁa Pérez, Roger Petersen, Scott Straus, Libby Wood, and
participants in the May 2000 CERI/IEP conference on “La guerre entre le local et le global,” for their
comments.
1
Recent research shows that the prevalence of civil wars in the 1990s is attributable to a steady accumulation
of conﬂicts since the 1950s, not the end of the cold war. See James D. Fearon and David
D. Laitin, “Ethnicity, Insurgency, and Civil War” (Paper presented at the Laboratory in Comparative
Ethnic Processes, Duke University, 2000).
2
Rogers Brubaker and David D. Laitin, “Ethnic and Nationalist Violence,” Annual Review of Sociology
24 (1998).
3
Steven R. David, “Internal War: Causes and Cures,” World Politics 49 (July 1997).
4
Fearon and Laitin (fn. 1); Nicholas Sambanis, “Partition as a Solution to Ethnic War: An Empirical
Critique of the Theoretical Literature,” World Politics 52 (July 2000).
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 99
This article traces the origins of this distinction and then disaggregates
it along three related dimensions: causes and motivations,
support, and violence. I show, through the use of recent, mostly ethnographic
research, how incomplete or biased information on recent civil
wars taints our interpretation; by using mostly recent historical research
on a large number of old civil wars, I demonstrate how inadequate attention
to this kind of research affects our understanding of past civil
wars. This article concludes with methodological suggestions on the
study of civil wars.
ORIGINS OF THE DISTINCTION
Most versions of the distinction between old and new civil wars stress
or imply that new civil wars are characteristically criminal, depoliticized,
private, and predatory; old civil wars are considered ideological,
political, collective, and even noble. The dividing line between old and
new civil wars coincides roughly with the end of the cold war.
The tendency to denigrate recent or ongoing wars—particularly
when other nations’ civil wars are compared to one’s own—is not new.
Consider the argument put forth in 1949, by F. A. Voigt, a British journalist
covering the Greek Civil War:
In the English and American civil wars, there were high-minded patriots on either
side. In these conﬂicts, the people were so evenly divided and the issues
were of such depth, scope, and variety, that it is not possible for the historian to
condemn one side utterly and to attribute exclusive righteousness to the other,
even if he may have the conviction that the triumph of one side was a national
calamity or the reverse . . . Such considerations do not apply to the Greek Sedition
which attained the magnitude but not the nature of an indigenous revolutionary
civil war. The Sedition is not to be explained in terms of any popular
grievances or of any failure on the part of the State.5
The post–cold war manifestation of this type of argument can be
traced in part to best-selling “lay” authors who articulated graphic accounts
of recent civil wars in places like Liberia, Bosnia, and Sierra
Leone.6
In addition, a number of scholars in security studies and international
relations have also advanced various versions of this argu-
100 WORLD POLITICS
5
F. A Voigt, The Greek Sedition (London: Hollis and Carter, 1949), 68–69.
6
See Hans Magnus Enzensberger, Civil Wars: From L.A. to Bosnia (New York: The New Press,
1994); Robert D. Kaplan, Balkan Ghosts: A Journey Through History (New York: Vintage, 1994); idem,
“The Coming Anarchy: How Scarcity, Crime, Overpopulation, and Disease are Rapidly Destroying
the Social Fabric of our Planet,” Atlantic Monthly 44 (February 1994); Michael Ignatieff, The Warrior’s
Honor: Ethnic War and the Modern Conscience (New York: Henry Holt and Company, 1998).
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 100
ment.7
Even some economists have adopted a related analytical distinction—between
“justice-seeking” and “loot-seeking” civil wars—
and are building models based on the assumption of rebellion as a
criminal enterprise.8
The adoption of the distinction is not a mere academic
exercise insofar as it motivates speciﬁc policy demands, including
“humanitarian law-enforcement.”9
For example, the 1999 agreement
ending the civil war in Sierra Leone met with opposition from many
human rights activists, journalists, and opinion makers who believed
that the rebels were violent criminals and not political revolutionaries
and that it was therefore immoral to grant them amnesty and invite
them to participate in the new government.10
THREE DIMENSIONS
In most accounts, old and new civil wars vary along three related dimensions.
These broadly stylized categories are summarized in Table 1
as follows:11
“NEW” & “OLD” CIVIL WARS 101
7
Edward N. Lutwack, “Great-powerless Days,” Times Literary Supplement, June 16, 1995; Kalevi J.
Holsti, The State, War, and the State of War, (Cambridge: Cambridge University Press, 1996); Chris
Hables Gray, Post-Modern War: The New Politics of Conﬂicts (London: Routledge, 1997); Mark
Dufﬁeld, “Post-modern Conﬂict: Warlords, Post-adjustment States and Private Protection,” Civil Wars
1, no. 1 (1998); David Keen, “The Economic Functions of Violence in Civil Wars,” Adelphi Paper 320
(1998); Mary Kaldor, New and Old Wars: Organized Violence in a Global Era (Stanford, Calif.: Stanford
University Press, 1999); Mats Berdal and David M. Malone, eds, Greed and Grievance: Economic Agendas
in Civil Wars (Boulder: Lynne Rienner, 2000).
8
Herschel I. Grossman, “Kleptocracy and Revolution,” Oxford Economic Papers 51, no. 2 (April
1999); Paul Collier, “Rebellion as a Quasi-Criminal Activity,” Journal of Conﬂict Resolution 44, no. 6
(2000); Paul Collier, “Economic Causes of Civil Conﬂict and their Implications for Policy,” in Chester
A. Crocker, Fen Osler Hampson, and Pamela Aall, eds., Managing Global Chaos (Washington D.C.:
U.S. Institute of Peace, forthcoming); Paul Azam and Anke Hoefﬂer, “Looting and Conﬂict between
Ethno-Regional Groups: Lessons for State Formation in Africa” (Paper presented at the World Bank
Center for International Studies Workshop on “The Economics of Civil War,” Princeton University,
March 18–19, 2000); Paul Collier and Anke Hoefﬂer, “Justice-Seeking and Loot-Seeking in Civil
War,” Manuscript, World Bank, 1999); idem, “Greed and Grievance in Civil War,” World Bank Policy
Research Paper 2355 (Washington, D.C.: World Bank, 2000).
9
Kaldor (fn. 7), 66.
10
A United Nations ofﬁcial described the population’s desire for amnesty in exchange for peace as
representing a peculiarly African understanding of justice. See Remy Ourdan, “Le Prix de la Paix,” Le
Monde, December 2, 1999. Interestingly, the publication of this article coincided with the announcement
of a peace agreement in Northern Ireland. Critics of the Irish agreement were in turn criticized
by the same media that condemned the Sierra Leone deal, on the exact opposite grounds. For example,
the French newspaper Le Monde (December 4, 1999), which condemned the amnesty agreement in
Sierra Leone praised the British journalist Hugo Young, who supported the participation in the new
government of a former IRA commander suspected of murders, since without him, “there would be no
peace agreement.” The peace agreement in Sierra Leone was also condemned on pragmatic grounds.
It was pointed out that “from the rebels’ point of view, why have peace when it is the absence of law
and order that enables one to loot?. . . In fact the rebels never had any intention of honoring the peace
accord; they were only interested in waging war and looting the country.” William Reno, “When Peace
Is Worse than War,” New York Times, May 11, 2000. Yet could not the same argument be made about
the peace agreement in Mozambique, which has since been widely hailed as a success story?
11
Some scholars collapse many of these dimensions into one, while others emphasize some dimensions
at the expense of others. Kaldor (fn. 7) seems to compare new civil wars with old conventional
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 101
1. Old civil wars were political and fought over collectively articulated, broad,
even noble causes, such as social change—often referred to as “justice”. By contrast,
new civil wars are criminal and are motivated by simple private gain—
greed and loot.
2. At least one side in old civil wars enjoyed popular support; political actors
in new civil wars lack any popular basis.
3. In old civil wars acts of violence were controlled and disciplined, especially
when committed by rebels; in new civil wars gratuitous and senseless violence is
meted out by undisciplined militias, private armies, and independent warlords
for whom winning may not even be an objective.
COLLECTIVE VERSUS PRIVATE CAUSES AND MOTIVATIONS
Taking into account the broad causes of civil wars and the individual
motivations of their combatants, many scholars implicitly hold that old
civil wars were motivated by broad, well-deﬁned, clearly articulated,
universalistic, ideologies of social change,12
whereas, new civil wars tend
to be motivated by concerns that often boil down to little more than
simple private gain. Recent work by economists is premised on a
dichotomous distinction between grievance and greed—rebels are
either bandits motivated by private greed or are political actors seeking
to ameliorate collective grievance.13
UN Secretary-General Koﬁ Annan
recently pointed out that “the pursuit of diamonds, drugs, timber,
102 WORLD POLITICS
wars. Keen (fn. 7) argues that looting generates “rational” rather than gratuitous violence. The claim
that new civil wars are motivated by looting is sometimes made in contradistinction to their purported
ethnic motivation—while sometimes ethnic motivations and looting are merged. For the former view
see Koﬁ Annan, “Facing the Humanitarian Challenge: Towards a Culture of Prevention,” UNDPI (New
York, 1999); for the latter view see John Mueller, “The Banality of ‘Ethnic War,’” International Security
25, no. 1 (2000).
12
Kaldor (fn. 7), 6.
13
Collier and Hoefﬂer, 2000 (fn. 8), 2–3; Collier (fn. 8); Collier and Collier and Hoefﬂer produce a
number of mixed greed-grievance models in which rebellion begins as a collective grievance and is sustained
by greed. All models, however, presuppose this dichotomous distinction. According to a World
Bank Press Release: “New World Bank research suggests that civil wars are more often fueled by rebel
groups competing with national governments for control of diamonds, coffee, and other valuable primary
commodities, rather than by political, ethnic, or religious differences…‘Civil wars are far more
likely to be caused by economic opportunities than by grievance, and therefore certain rebel groups
beneﬁt from the conﬂict and have a very strong interest in initiating and sustaining it,’ says [Paul] Collier.”
World Bank, “Greed for Diamonds and Other ‘Lootable’ Commodities Fuels Civil Wars” (News
Release 2000/419/S, http://wbln0018.worldbank.org/news/pressrelease.nsf, accessed April 20, 2001).
TABLE 1
OLD AND NEW WARS
Old Civil Wars New Civil Wars
Causes & motivation collective grievances private loot
Support broad popular support lack of popular support
Violence controlled violence gratuitous violence
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 102
concessions and other valuable commodities drives a number of today’s
internal wars. In some countries the capacity of the State to extract
resources from society and to allocate patronage is the prize to be
fought over.14
The criminal metaphor takes many forms. For Enzensberger,
competing factions in new civil wars are “warrior gangs.”15
Kaplan
describes civil wars in Africa as criminal actions by bandits and
disenfranchised soldiers, teenage hooligans, and child-soldiers on
drugs.16
Some even argue that new civil wars lack purpose entirely. As Enzensberger
argues: “What gives today’s civil wars a new and terrifying
slant is the fact that they are waged without stakes on either side, that
they are wars about nothing at all.”17
Further, “there is no longer any
need to legitimize your actions. Violence has been freed from ideology,”
and the combatants have an innate inability to think and act in terms of
past and future.18
Such arguments, however, are often based on incomplete or biased
evidence derived from journalistic reports that tend to quote uncritically
city-dwellers and members of progovernmental organizations.
Fieldworkers have described such views as paying “scant regard to the
insurgents’ own claims concerning the purpose of their movement . . .
and [preferring] instead to endorse a view widespread among capital
city elites and in diplomatic circles.”19
Gourevitch points out that “by
denying the particularity of the peoples who are making history, and
the possibility that they might have history, [such arguments] mistake
[their] failure to recognize what is at stake in events for the nature of
these events.”20
More generally, the concept of looting is analytically problematic
because it is unclear whether it refers to the causes of war or the motivations
of the combatants (or both). The ﬁrst problem is the direction
“NEW” & “OLD” CIVIL WARS 103
14
Annan (fn. 11).
15
Enzensberger (fn. 6), 22.
16
Kaplan (fn. 6).
17
Enzensberger (fn. 6), 30. Emphasis in original.
18
Ibid., 20–1, 29.
19
Paul Richards, Fighting for the Rain Forest: War, Youth, and Resources in Sierra Leone (Oxford:
James Currey, 1996), xvii. In his study of the war in Mozambique, anthropologist Christian Geffray
castigates “journalists who cannot investigate [the war] on the ground,” and international media that
reproduce “information and analyses” reﬂecting the views of “urban elites, national intellectuals, and
foreigners.” Christian Geffray, La cause des armes au Mozambique: Anthropologie d’une guerre civile
(Paris: Karthala, 1990), 19.
20
Philip Gourevitch, We Wish to Inform You That Tomorrow We Will be Killed with Our Families: Stories
from Rwanda (New York: Farrar, Straus, and Giroux, 1998), 182.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 103
of causality—do people wage war in order to loot or do they loot to be
able to wage war?21
If the latter is the case, then looting may be no different
from the widely accepted practice of “revolutionary taxation”.
Second, it is not always clear who is doing the looting—elites, autonomous
militias, armed peasants? Third, the linkages between looting
and grievances are complex and ﬂuid.22
Can we seriously reduce the
1992 Los Angeles riots to a phenomenon of “looting” even though
much looting—among many other things—did take place? Finally,
there are serious empirical problems. The signiﬁcance of empirical indicators
proxying for “lootable” resources raises important questions of
internal validity—beyond failing to address problems of causality. To
say, in short, that the civil war in Sierra Leone is mainly about diamonds
appears to be a gross oversimpliﬁcation.23
Civil wars in Colombia,
Somalia, and Sudan are even less amenable to such simpliﬁcation.24
Researchers who have studied new civil wars by conducting lengthy
ﬁeldwork in war zones—as opposed to interviewing victims and government
ofﬁcials—provide very nuanced accounts that fail to support
the grievance/looting dichotomy. They ﬁnd rebel motivations are diverse
and include concerns that go beyond mere banditry.25
Peters and
Richards have shown about Sierra Leone, for example, that many rankand-ﬁle
members of the African rebel movements that have been stigmatized
as lacking any ideology appear in fact to have had a
sophisticated political understanding of their own participation.26
Their
ideological motivations are simply not always visible to observers looking
for “Western” patterns of allegiance and discourse. They make the
ﬂawed assumption that organizations using religious idioms and local
104 WORLD POLITICS
21
Although the direction of causality may be irrelevant for predicting the likelihood of civil wars, it
matters when deriving empirical, theoretical, and normative implications about civil wars.
22
Collier and Hoefﬂer (fn. 8) acknowledge the complexity of the possible connections between
“greed” and “grievance.”
23
Richards (fn. 19).
24
Mauricio Romero, “Changing Identities and Contested Settings: Regional Elites and the Paramilitaries
in Colombia,” International Journal of Politics, Culture and Society 14, no. 1 (2000); Isabelle
Duyvesteyn, “Contemporary War: Ethnic Conﬂict, Resource Conﬂict or Something Else?” Civil Wars
3, no. 1 (2000); Catherine Besteman, “Violent Politics and the Politics of Violence: The Dissolution of
the Somali Nation-State,” American Ethnologist 23, no. 3 (1996).
25
A psychologist who treated hundreds of ﬁghters in the Liberian Civil War drew the following
proﬁle: “He is someone usually between 16 and 35 years of age, who may have decided to become a
combatant for several reasons: to get food for survival, to stop other ﬁghters from killing his family and
friends, was forced to become a combatant or be killed, sheer adventurism etc.” E. S. Grant, quoted in
Stephen Ellis, The Mask of Anarchy:The Destruction of Liberia and the Religious Dimension of an African
Civil War (New York: New York University Press, 1999), 127.
26
Krijn Peters and Paul Richards, “‘Why We Fight’: Voices of Youth Combatants in Sierra Leone,”
Africa 68, no. 2 (1998).
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 104
cultural practices to mobilize people27
—rather than easily recognizable
universalistic appeals—lack any ideology. The use of traditional
processes of initiation, for example, is central in African rebel organiza-
tions.28
Chingono’s study of Mozambique emphatically argues that Renamo,
“by resuscitating and defending peasant outlooks of the world,
which had been suppressed by Frelimo . . . was articulating peasant
ideologies.”29
To understand modern rebel leaders—often referred to pejoratively
as warlords30
—a useful source of insight is the relevant historical literature
on warlordism (focusing on China), which argues that the key
feature of warlordism is rule rather than looting. Warlords are never
mere bandits; they are lords of a particular area by virtue of their capacity
to wage war.31
Whereas bandits—in China and elsewhere—must hit
and run in order to survive, warlords levy taxes, administer justice,
maintain some degree of order, and generally assume the burdens of
government in the areas they control.32
They are state builders. Saint
Augustine observed this very phenomenon: “If by accessions of desperate
men this evil [brigandage] grows to such proportions that it holds
lands, establishes ﬁxed settlements, seizes upon states and subjugates
peoples, it assumes the name of a kingdom.”33
Rebel organizations in
Africa, often dismissed as mere criminal gangs, develop a complex apparatus
of rule in the areas they control—which is less visible but not
very different from the order implemented by “justice oriented”
rebels.34
These organizations also engage in organized, systematic, and
sophisticated economic interactions with foreign ﬁrms, which buy raw
materials and sell weapons,35
an activity at odds with the extreme fragmentation
implied by many views.
“NEW” & “OLD” CIVIL WARS 105
27
Tom Young, “A Victim of Modernity? Explaining the War in Mozambique,” in Paul B. Rich and
Richard Stubbs, eds., The Counter-Insurgent State: Guerrilla Warfare and State-Building in the Twentieth
Century (New York: St. Martin’s Press, 1997), 136–37; Stephen L. Weigert, Religion and Guerrilla
Warfare in Modern Africa (New York: St. Martin’s Press, 1996); Ellis (fn. 25); Thomas H. Henriksen,
Revolution and Counterrevolution: Mozambique’s War of Independence, 1964–1974 (Westport, Conn.:
Greenwood Press, 1983), 76.
28
Richards (fn. 19), xix.
29
Mark F. Chingono, The State, Violence, and Development: The Political Economy of War in Mozambique,
1975–1992 (Aldershot: Avebury, 1996), 55.
30
See, for example, Reno (fn. 10).
31
James E. Sheridan, Chinese Warlord:The Career of Feng Yü-hsiang (Stanford, Calif.: Stanford University
Press, 1966), 1.
32
Ibid., 19.
33
Saint Augustine, The City of God, trans. John Healey (London: J. M. Dent; New York: E. P. Dutton,
1931), IV:iv.
34
Stephen Ellis, “Liberia 1989–1994: A Study of Ethnic and Spiritual Violence,” African Affairs 94,
no. 375 (1995), 165–197; Dufﬁeld (fn. 7); Geffray (fn. 19).
35
William Reno, Warlord Politics and African States (Boulder, Colo.: Lynne Rienner, 1998).
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 105
The typical picture of “ideologically oriented” actors in old civil wars,
for its part, is often misrepresented as well. Such actors have often
engaged in criminal activities, large-scale looting, and the pronounced
coercion of the populations whose grievances they claimed to represent.
Indeed, looting is a recurring element of civil wars, including the most
ideological ones such as the Russian and Chinese Revolutions36
and
anticolonial rebellions, such as the one in Indonesia in the 1940s.37
Even Lenin entered into agreements with “criminal elements” during
the Russian Civil War. The behavior of the Red Army in Kharkov and
Kiev in 1919, as it emerges from Soviet records, led the historian
Vladimir N. Brovkin to assert that “in plain English, the Bolshevik
rulers were thieves and rapists.”38
“Taxation” is a key rebel activity in all
civil wars, and incumbents do not shy away from outright looting.
During the Vietnam War, one could ﬁnd among South Vietnamese
militiamen former criminals “who preferred ﬁghting to sitting in jail,”
while American advisers often allowed the members of the CIAsponsored
Provincial Reconnaissance Units to “keep money captured
during their operations.”39
The paradigmatic ideological political
actors, the members of the French Revolutionary armies, were described
by their contemporaries as “highwaymen,” “vagrants,” “robbers,”
“vagabonds,” and “vicious, bloodthirsty hooligans.”40
Nor should one
forget that their adversaries, the counterrevolutionaries, resorted to
banditry as well.41
Furthermore, the importance of ideological motivations in old civil
wars has been greatly overstated. To begin with, there is a clear epistemic
bias in favor of the assumption that old civil wars (as well as most
individuals participating in them) were motivated by grand ideological
concerns. Because intellectuals tend to be primarily motivated by ideology,
they tend to assign overwhelmingly ideological motives to both
106 WORLD POLITICS
36
Lincoln Li, The Japanese Army in North China, 1937–1941: Problems of Political and Economic Control
(Tokyo: Oxford University Press, 1975), 229; Odoric Y.K. Wou, Mobilizing the Masses: Building
Revolution in Henan (Stanford, Calif.: Stanford University Press, 1994), 154; Orlando Figes, A People’s
Tragedy: The Russian Revolution, 1891–1924 (New York: Penguin, 1997), 666–67.
37
Robert Cribb, Gangsters and Revolutionaries: The Jakarta People’s Militia and the Indonesian Revolution,
1945–1949 (Honolulu: University of Hawaii Press, 1991), 54.
38
Vladimir M. Brovkin, Political Parties and Social Movements in Russia, 1918–1922 (Princeton:
Princeton University Press, 1994), 121.
39
Mark Moyar, Phoenix and the Birds of Prey:The CIA’s Secret Campaign to Destroy the Viet Cong (Annapolis,
Md.: Naval Institute Press, 1997), 168.
40
Richard Cobb, The People’s Armies (New Haven: Yale University Press, 1987), 5.
41
Charles Tilly, The Vendée (Cambridge: Harvard University Press, 1964), 6.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 106
participants and civilians in civil wars.42
Moreover, when not crudely
“disguising” ethnic or local claims, universalistic ideological appeals
were propagated by means of traditional cultural idioms often not unlike
those used by movements in new civil wars. For example, Lan has
shown how the “progressive” Zimbabwean rebels who fought against
the country’s racist regime used traditional religion (and its practitioners)
to mobilize peasants.43
In addition, it is a grave mistake to infer the
motivations of rank-and-ﬁle members from their leadership’s articulation
of its ideological messages.44
Microlevel historical studies consistently demonstrate how superﬁcial
was the adoption of ideological claims (typically expressed in
bafﬂing acronyms) across a range of civil wars. A common ﬁnding in
numerous studies of old civil wars is that at the mass level, local
considerations tended to trump ideological ones. Dallin et al. make this
point about German-occupied Soviet Union, where an individual’s
decision to side with the Germans or the partisans was not determined
by “abstract considerations and evaluations of the merits and demerits
of the two regimes, nor even by likes and dislikes or experiences under
the Soviet regime before the occupation.”45
Swedenburg’s subtle analysis
of Palestinian collaboration with the British during the 1936–39
Palestinian rebellion makes the same point.46
Likewise, McKenna’s
focus on the “unauthorized narratives” of Muslim rebels and supporters
in the Southern Philippines reveals “that ordinary Muslims’ perceptions
and representations of the war were often conspicuously independent
of the ideological inﬂuences of any separatist leaders or, for that matter,
“NEW” & “OLD” CIVIL WARS 107
42
It turns out that political violence is not directly caused by individual, radical ideologies even in
urban environments, as Della Porta shows in her study of Italian and German terrorist organizations.
See Donatella Della Porta, Social Movements, Political Violence, and the State: A Comparative Analysis of
Italy and Germany (Cambridge: Cambridge University Press, 1995), 196. As Barrington Moore puts it:
“The discontented intellectual with his soul searchings has attracted attention wholly out of proportion
to his political importance, partly because these searchings leave behind them written records and also
because those who write history are themselves intellectuals.” Barrington Moore, Social Origins of Dictatorship
and Democracy: Lord and Peasant in the Making of the Modern World (Boston: Beacon Press,
1966), 480
43
David Lan, Guns and Rain: Guerrillas and Spirit Mediums in Zimbabwe (London: James Currey,
1985). See also Henriksen (fn. 27), 76, for Mozambique.
44
Paul Jankowski, Communism and Collaboration: Simon Sabiani and Politics in Marseille, 1919–1944
(New Haven: Yale University Press, 1989), ix, xii.
45
Alexander Dallin, Ralph Mavrogordato, and Wilhelm Moll, “Partisan Psychological Warfare and
Popular Attitudes,” in John A. Armstrong, ed., Soviet Partisans in World War II (Madison: University of
Wisconsin Press, 1964), 336.
46
Ted Swedenburg, Memories of Revolt: The 1936–1939 Rebellion and the Palestinian National Past
(Minneapolis: University of Minnesota Press, 1995), 169–70.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 107
of any elite group.”47
Observing effective insurgent performance in
combat has often led to the erroneous inference that rebels are highly
dedicated to an ideological cause. However, numerous studies have
concluded that men in combat are usually motivated by group pressures
and processes involving: (1) regard for their comrades, (2) respect for
their leaders, (3) concern for their own reputation with both, and (4) an
urge to contribute to their success of the group.48
Recent sociological
research on religious conversion, a “choice” even more amenable to ideological
considerations than politics, shows that doctrinal appeal does
not lie at the heart of the conversion process: most people do not really
become very attached to the doctrines of their new faith until after
their conversion.49
Usually, processes of joining are rooted in network
dynamics. Stark, Wickham-Crowley, and Petersen argue that social
network ties (especially friendship and kin ties) are the best predictors
of joining a movement.50
As Hart points out about the Irish Revolution
and Civil War:
The most important bonds holding Volunteers together were those of family
and neighborhood. Indeed, IRA companies were very often founded upon such
networks . . . Twelve of the thirteen veterans I interviewed had fought on the
republican side. None could remember making a speciﬁc choice to do so. “I
hadn’t a clue”; “It was very confusing altogether.” Judging by the recollections of
Cork veterans, the Treaty itself and republican ideology were rarely discussed
within their ranks. “The politics of it was second place at times.” Most couched
their decisions in the same collective terms they used to describe their joining
the organization.51
In short, the handy presence of coherent conceptual categories along
the familiar left-right axis, which blinded casual observers to the complexity
and messiness of civil wars, appears to have led to a signiﬁcant
overstatement of the ideological content of old civil wars via unwar-
108 WORLD POLITICS
47
Thomas M. McKenna, Muslim Rulers and Rebels: Everyday Politics and Armed Separatism in the
Southern Philippines (Berkeley: University of California Press, 1998), 194–95. Collective grievances
tend to be expressed only under restrictive conditions. See Elisabeth Wood, “Pride in Rebellion: Insurrectionary
Collective Action in El Salvador” (Manuscript, New York University, Spring 2001).
48
Dave Grossman, On Killing: The Psychological Cost of Learning to Kill in War and Society (Boston:
Little, Brown, and Company, 1995), 89–90; Walter Laqueur, Guerilla Warfare: A Historical and Critical
Study (New Brunswick, N.J.: Transaction, 1998), 272. Obviously, this does not answer the question
of how and why an organization capable of providing such training and leadership emerges.
49
Rodney Stark, The Rise of Christianity: How the Obscure, Marginal Jesus Movement Became the
Dominant Religious Force in the Western World in a Few Centuries (New York: Harper Collins, 1997),
14–17.
50
Ibid.; Timothy Wickham-Crowley, Exploring Revolution: Essays on Latin American Insurgency and
Revolutionary Theory (Armonk, N.Y.: M. E. Sharpe, 1991), 152; Roger Petersen, Resistance and Rebellion:
Lessons from Eastern Europe (New York: Cambridge University Press, 2001).
51
Peter Hart, The I.R.A. and Its Enemies: Violence and Community in Cork, 1916–1923 (New York:
Clarendon Press, 1999), 209, 264.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 108
ranted inferences from the elite to the mass level. In this regard, the end
of the cold war seems to have caused the demise of the conceptual categories
used to interpret civil wars rather than a decline in the ideological
motivations of civil wars at the mass level. Ironically, detailed
research about these wars conducted years later tends to be ignored by
analysts of contemporary civil wars who keep relying on the ﬂawed depictions
produced when the old civil wars were ongoing.
POPULAR SUPPORT VERSUS LACK OF SUPPORT
Since old civil wars were supposed to grow out of accumulated popular
grievances, they were assumed to be based on considerable popular support—at
least for the rebels. By contrast, new civil wars appear to be
fought by political actors who lack any such support. According to
Kaldor: “Whereas guerrilla warfare, at least in theory as articulated by
Mao Tse-tung or Che Guevara, aims to capture ‘hearts and minds,’ the
new warfare borrows from counterinsurgency techniques of destabilization
aimed at sowing ‘fear and hatred.’”52
Similarly, Nordstrom described
the Mozambican rebels of the Renamo as “a particularly lethal
rebel movement that has virtually no ideology or popular support,”
formed by foreign powers intent on destabilizing the country, and responsible
for “over 90 percent of all atrocities committed.”53
Likewise
Pécaut argues that the war in Colombia is not a civil war because the
population does not support any side at all.54
Such statements are often based upon incomplete or biased information.
Nordstrom’s account, for example, relies exclusively on interviews
with refugees in areas “recently liberated from Renamo control by
government forces” and information provided by progovernmental organizations
(such as the Organization of Mozambican Women), relaying
the government’s view of the rebels. She reported that “in
Mozambique, [the Renamo] is generally referred to as bandidos armados
(armed bandits),” ignoring that incumbents in all civil wars use such
terms to describe insurgents.55
Recent studies based on evidence that
“NEW” & “OLD” CIVIL WARS 109
52
Kaldor (fn. 7), 8.
53
Carolyn Nordstrom, “The Backyard Front,” in Carolyn Nordstrom and JoAnn Martin, eds., The
Paths to Domination, Resistance, and Terror (Berkeley: University of California Press, 1992), 271–72.
54
Daniel Pécaut, “En Colombie, une guerre contre la société,” Le Monde, October 10, 1999. Similar
statements are commonly made about Sierra Leone. See, for example, Reno (fn. 35).
55
In a subsequent account, Nordstrom provided a more nuanced portrayal of the situation in
Mozambique. Carolyn Nordstrom, “War on the Front Lines,” in Carolyn Nordstrom and Antonius
C. G. M. Robben, eds., Fieldwork under Fire: Contemporary Studies of Violence and Survival (Berkeley:
University of California Press, 1995), 142.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 109
was hard (if not impossible) to collect while the civil war was ongoing
indicate that Renamo enjoyed a considerable level of popular support.56
This support was present in rural areas controlled by Renamo, where
researchers and journalists rarely traveled, rather than in the cities
under governmental control.57
Conversely, the perception that rebellions in old civil wars were
based on widespread popular support has been repeatedly called into
question. To begin with, the view that leftist rebellions, in Latin
America and elsewhere, were based mostly on widespread and consensual
popular participation has been questioned by careful microoriented
research.58
Likewise, the Vietcong relied on extensive coercion
against the civilian population.59
Contrary to what Kaldor argues,60
mass population displacement is nothing new—as suggested by such
classic wars as the Russian, Spanish, and Chinese Civil wars.
Furthermore, individual loyalties in old civil wars, as in new civil
wars, are often informed less by impersonal discourses and more by
ﬂuid, shifting, and often locally based cleavages. Many studies describe
messy processes, often characterized by a disjunction between underlying
cleavages on the one hand, and violent conﬂict and identities on the
other. For example, Hart’s analysis of Cork County in Ireland from
1916 to 1923 unearths a high level of variation in political attitudes at
the microlevel, an “array of—often conﬂicting—local loyalties [which]
turned every part of Cork into a political patchwork.”61
When, in 1923,
Irish nationalists fought a civil war, the decision about which side to
join was “shaped, as always, by group loyalties and rivalries. Factional
divisions became political battle lines.”62
110 WORLD POLITICS
56
Young (fn. 27); Chingono (fn. 29). Chingono also points out that “while Renamo would not have
survived without external support, exclusive focus on external factors equally distorts the reality and
denies the Mozambicans’ own history; they are reduced to mere passive victims of manipulations and
machinations of powerful external forces.”
57
Similar observations have been made about Liberia and Sierra Leone. See Ellis (fn. 34); Richards
(fn. 19).
58
See, for example, David Stoll, Rigoberta Menchú and the Story of All Poor Guatemalans (Boulder:
Westview, 1999); Carlos Iván Degregori, “Harvesting Storms: Peasant Rondas and the Defeat of
Sendero Luminoso in Ayacucho,” in Steven J. Stern, ed., Shining and Other Paths: War and Society in
Peru, 1980–1995 (Durham and London: Duke University Press, 1998), 128–57. The same is true for
anticolonial wars in Africa. See Norma Kriger, Zimbabwe’s Guerrilla War: Peasant Voices (Cambridge:
Cambridge University Press, 1992).
59
Douglas Pike, Viet Cong:The Organization and Techniques of the National Liberation Front of South
Vietnam (Cambridge, Mass.: MIT Press, 1966).
60
Kaldor (fn. 7), 8.
61
Hart (fn. 51), 220.
62
Ibid., 265–66.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 110
Family and faction dictated the course of the IRA split in units all over Ireland,
often in highly predictable fashion. Once again, it was the Brennans against the
Barretts in Clare, the Hanniganites against the Manahanites in east Limerick,
and the Sweeneys versus the O’Donnells in Donegal as all the old feuds were
reignited.63
Similar dynamics are observable in most old civil wars. For example,
the South Vietnamese village of Binh Nghia displayed a “lukewarm attitude
toward the Viet Cong” because the local communist movement
had originated across the river, in the Phu Long hamlets, with which
whom they had a long-standing and hostile feud over ﬁshing rights.64
As Manrique describes, in the central Peruvian valley of Canipaco the
population enjoyed a “kind of honeymoon” with Shining Path, which
ended when a dispute erupted between two communities over the distribution
of lands previously usurped by haciendas.65
Because the meaning of rebellions is often articulated by elites in the
language of national cleavages, many observers erroneously code them
as actually mobilizing popular support along those cleavages. Fieldworkers
disagree. In his analysis of the Cultural Revolution in one Chinese
village, Hinton reports that warring factions used the language of
class struggle, with each faction claiming that the other represented
landlords and counterrevolutionary elements. Hinton, however, found
that the conﬂict was polarized around competing clans: the Lu family,
which dominated the northern and larger section of the village, and the
Shen family, which played a major role in the southern section of the vil-
lage.66
The same discovery was made by the writer of a report on the
1927 Haifeng uprising in South China, a region polarized into com“NEW”
& “OLD” CIVIL WARS 111
63
Ibid., 266.
64
“The hostility between the Phu Longs and Binh Nghia was generations old, focused on a feud
over ﬁshing rights. It was natural that the Phu Longs assumed economic as well as political power
when the Viet Cong were on the rise and this was done at the direct expense of ﬁshermen from Binh
Nghia. So later when the Viet Cong came across the river to spread the gospel, there were many in
Binh Nghia who resented them and any cause they represented. The police chiefs had fed this resentment
with money and had built a spy network.” See F. J. West, Jr., The Village (Madison: University of
Wisconsin Press, 1985), 146–47.
65
“The participation of armed Shining Path cadres on the side of one of the communities in a
massive confrontation against a confederation of rival communities provoked a rupture with the latter,
who decided to turn over two senderista cadres they had captured in the scufﬂe to the authorities in
Huancayo. This action provoked Shining Path reprisals, which culminated in the execution of thirteen
peasant leaders. The victims were kidnapped from their communities and assassinated in the central
plaza of Chongos Alto.” Nelson Manrique, “The War for the Central Sierra,” in Stern (fn. 58), 204–5.
66
William Hinton, Shenfan: The Continuing Revolution in a Chinese Village (New York: Vintage
Books, 1984), 527.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 111
peting alliances of villages known as Red Flag and Black Flag, which
had grown out of lineage struggles: “When the Red Army arrived ﬂying
red banners, the troops were greeted by landowners and peasants
alike from Red Flag villages who thought they were allies in the struggle
against the common enemy, the Black Flag villages.”67
Moreover, locally segmented cleavages often aggregate in misleading
ways—wealthy peasants may support one political actor in one region
and its rival in a neighboring region;68
wealthy merchants can be
targeted by poor right-wing death squad members in an otherwise
class-polarized conﬂict;69
sets of diverse (overlapping or not) regional
and local cleavages, such as socio-economic, factional, lineage, clan,
tribal, gender, or generational cleavages, combine to produce misleadingly
uniform aggregate cleavages; vertical relationships (patron-client)
and vertical ties (communities, neighborhoods, townlands, parishes,
corporations, factions, clans, or kin) often trump horizontal cleavages.70
Group interests are often “localistic and region-speciﬁc;”71
individual
motivations are not necessarily informed by impersonal cleavagerelated
grievances, but often by local and personal conﬂicts,72
even by
common crime.73
As Tilly has observed about the Vendée: “The most
microscopic information we have on communal politics in southern
Anjou resists forcing into categories of class and locality alone, and calls
for hunches about kinship, family friendships, the residues of old feuds,
and the like.”74
The same applies for societies that are sharply polarized
in terms of class75
and ethnicity.76
Social relations and the connections
that formed identities before the war become a matter of “constant
112 WORLD POLITICS
67
Robert Marks, Rural Revolution in South China: Peasants and the Making of History in Haifeng
County, 1570–1930 (Madison: University of Wisconsin Press, 1984), 263.
68
David H. Close, “Introduction,” in David H. Close, ed., The Greek Civil War, 1943–1950: Studies
in Polarization (London and New York: Routledge, 1993), 1–31; Geffray (fn. 19).
69
Benjamin D. Paul and William J. Demarest, “The Operation of a Death Squad in San Pedro la
Laguna,” in Robert M. Carmack, ed., Harvest of Violence: The Maya Indians and the Guatemalan Crisis
(Norman: University of Oklahoma Press, 1988), 128, 150.
70
Hart (fn. 51), 177; McKenna (fn. 47), 162; Kriger (fn. 58), 8; Hinton (fn. 66), 527; Marks (fn.
67), 264.
71
Young (fn. 27), 138–42; Chingono (fn. 29), 16; Swedenburg (fn. 46), 131–33; Wickham-Crowley
(fn. 50), 131.
72
See, for example, McKenna (fn. 47); Swedenburg (fn. 46); Paul and Demarest (fn. 69).
73
Mueller (fn. 11); Paul and Demarest (fn. 69); James D. Henderson, When Colombia Bled: A History
of the Violencia in Tolima (University, Ala.: University of Alabama Press, 1985).
74
Tilly (fn. 41), 191.
75
Stoll (fn. 58).
76
Richards (fn. 19), 6; Mohand Hamoumou, Et ils sont devenus Harkis (Paris: Fayard, 1993); Jan T.
Gross, Revolution from Abroad: The Soviet Conquest of Poland’s Western Ukraine and Western Belorussia
(Princeton, N.J.: Princeton University Press, 1988).
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 112
reformulation.”77
In many ways, civil wars provide a medium for a
variety of grievances to be realized within the space of the greater
conﬂict and through the use of violence. As Lucas argues about the
Revolution in southern France, “the revolutionary struggle provided
a language for other conﬂicts of a social, communal, or personal
nature.”78
In short, micro-oriented studies of old civil wars offer a ground-level
view of civil wars as “welters of complex struggles,”79
rather than as
simple binary conﬂicts between organizations crystallizing popular support
and collective grievances along well-deﬁned cleavages. In old civil
wars, popular support was shaped, won, and lost during the war, often
by means of coercion and violence and along lines of kinship and locality;
it was not purely consensual, immutable, ﬁxed, and primarily ideological.
In this respect, old civil wars are not as different from new civil
wars as they appear to be.
CONTROLLED VERSUS GRATUITOUS VIOLENCE
Violence in new civil wars is consistently described as both horriﬁc and
senseless, meted out by assorted militia and paramilitaries, mercenaries,
and independent warlords for whom winning the war may not even
be an objective.80
Human rights organizations and the press described
the gruesome massacres that took place in Algeria in 1997 as “senseless,”
“wanton,” and “incomprehensible” instances of “random butch-
ery.”81
Such descriptions often come with a culturalist shade. In the last
days of September 1998, when Serb soldiers massacred twenty-one
women, children, and elderly people near the village of Gornje Obrinje
in Kosovo, a detailed journalistic account concluded it was an instance
of “the practice of taking violent revenge [which] is a time-honored tradition
in the Balkans.”82
These descriptions are often complemented by
arguments that attempt to explain acts of violence by simply stating
their effects. A psychologist who treated the maimed victims of the
“NEW” & “OLD” CIVIL WARS 113
77
Mary Elizabeth Berry, The Culture of Civil War in Kyoto (Berkeley: University of California Press,
1994), xxi.
78
Colin Lucas, “Themes in Southern Violence after 9 Thermidor,” in Gwynne Lewis and Colin
Lucas, eds., Beyond the Terror: Essays in French Regional and Social History, 1794–1815 (Cambridge:
Cambridge University Press, 1983), 152–94.
79
Susan F. Harding, Remaking Ibieca: Rural Life in Aragon under Franco (Chapel Hill: University of
North Carolina Press, 1984), 59.
80
Kaldor (fn. 7), 93.
81
Stathis N. Kalyvas, “Wanton and Senseless? The Logic of Massacres in Algeria,” Rationality and
Society 11, no. 3 (1999), 243–85.
82
Jean Perlez, “Kosovo Clan’s Massacre Stands as Gruesome Evidence of Serb Revenge,” International
Herald Tribune, November 16, 1998.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 113
Revolutionary United Front (RUF) in Sierra Leone pointed out that “it
was the goal of the rebels to take away their role as men, fathers and
husbands.”83
Nordstrom states that “Renamo, with its tactics of severing
the noses, lips, and ears of civilians, seems to reclaim the original
sense of the absurd.”84
Enzensberger points to “the autistic nature of
perpetrators, and their inability to distinguish between destruction and
self-destruction.”85
A book quoted ad nauseam is Joseph Conrad’s Heart
of Darkness.86
Such senseless violence just was not as prevalent in old civil wars if
we are to believe Enzensberger, who argues that in the American,
Russian, and Spanish Civil Wars “there were regular armies and fronts;
the central command structures attempted to carry out their strategic
objectives in a planned way through strict control of their troops. As a
rule there was political as well as military leadership, following clearly
deﬁned goals, and ready and able to negotiate when necessary.”87
Yet a
quick perusal of the evidence from old civil wars conveys a quite different
image.
To begin with, the perception that civil wars are particularly cruel
predates new civil wars—it is one of the most enduring and consistent
observations,88
stressed by observers and participants alike, ever since
Thucydides’ depiction of the civil war in Corcyra.89
While the violence of ethnic conﬂicts has received sustained attention
lately, violence is in fact the central component of all kinds of civil
war, ethnic and non-ethnic alike. For example, a nineteenth-century
French counterrevolutionary leader remarked that “excesses are inseparable
from wars of opinion.”90
Likewise, Madame de Staël observed
that “all civil wars are more or less similar in their atrocity, in the upheaval
in which they throw men and in the inﬂuence they give to vio-
114 WORLD POLITICS
83
Norimitsu Onishi, “Sierra Leone Measures Terror in Severed Limbs,” New York Times, August
22, 1999.
84
Nordstrom (fn. 55), 142.
85
Enzensberger (fn. 6), 20.
86
See, for example, Ignatieff (fn. 6), 5.
87
Enzensberger (fn. 6), 15.
88
Nico H. Frijda, “The Lex Talionis: On Vengeance,” in Stephanie H.M. Van Goozen, Nanne E.
Van de Poll, and Joseph Sergeant, eds., Emotions: Essays on Emotion Theory (Hillsdale, N.J.: Lawrence
Erlbaum Associates, 1994), 267.
89
Thucydides depicted the civil war there as a situation in which “there was death in every shape
and form. And, as usually happens in such situations, people went to every extreme and beyond it.”
Thucydides, History of the Peloponnesian War, trans. Rex Warner (Harmondsworth, N.Y.: Penguin
Books, 1977), 241. One easily ﬁnds in the historical literature sentiments emphasizing the inherent
cruelty and ravaging nature of civil war. See Claude Petitfrère, La Vendée et les Vendéens (Paris: Gallimard/Julliard,
1981), 50; and John Gunther, Behind the Curtain (New York: Harper, 1949), 129.
90
Roger Dupuy, Les Chouans (Paris: Hachette, 1997), 237.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 114
lent and tyrannical passions.”91
Latin America has been a privileged
setting of very violent but mostly non-ethnic civil wars.92
Descriptions
of extreme violence in such old civil wars as the Russian and Spanish
are abound.93
The practice of using local semi-independent militia is
widespread among many “ideologically” oriented actors.94
Likewise, the
abduction of children in order to turn them into ﬁghters may be associated
with new civil wars in Africa but it was consistently practiced in
many “ideologically motivated” rebellions, such as the Afghan insurgency
following the Soviet invasion95
and the Shining Path insurgency
in Peru.96
Many children became ﬁghters in Guatemala, El Salvador,
and Nicaragua.97
During the (supremely “ideological”) Chinese Cultural
Revolution, the most violent groups were composed of young Red
Guards, ranging in age from eight to ﬁfteen.98
Turning to new civil wars, it is important to begin by pointing out
that our understanding of violence is culturally deﬁned.99
Killings by
knife and machete tend to horrify us more than the often incomparably
more massive killings by aerial and ﬁeld artillery bombings. As Crozier
put it forty years ago: “The violence of the strong may express itself in
high explosives or napalm bombs. These weapons are no less discriminate
than a hand-grenade tossed from a roof-top; indeed, they will
make more innocent victims. Yet they arouse less moral indignation
around Western ﬁresides.”100
Moreover, the “senseless” violence of new
civil wars is often not as gratuitous as it appears. The massacres in Algeria
were often highly selective and strategic,101
as was the violence
“NEW” & “OLD” CIVIL WARS 115
91
Madame de Staël, Des circonstances actuelles qui peuvent terminer la révolution et des principes qui
doivent fonder la république en France, ed. Lucia Omacini (1798; Geneva: Librairie Droz, 1979), 10.
92
Tina Rosenberg, Children of Cain: Violence and the Violent in Latin America (New York: Penguin,
1991), 7.
93
See, for example, Julio de la Cueva, “Religious Persecution, Anticlerical Tradition, and Revolution:
On Atrocities against the Clergy during the Spanish Civil War,” Journal of Contemporary History
33, no. 3 (1998); Figes (fn. 36); Brovkin (fn. 38).
94
Kalyvas (fn. 81), 265–77; Paul and Demarest (fn. 69); Degregori (fn. 58).
95
Artyom Borovik, The Hidden War: A Russian Journalist’s Account of the Soviet War in Afghanistan
(London: Faber and Faber, 1991), 25.
96
Ponciano Del Pino H., “Family, Culture, and ‘Revolution’: Everyday Life with Sendero Luminoso,”
in Steve J. Stern, ed. Shining and Other Paths: War and Society in Peru, 1980–1995 (Durham and
London: Duke University Press, 1998), 171.
97
Ilene Cohn and Guy S. Goodwin-Gill, Child Soldiers:The Role of Children in Armed Conﬂict (Oxford:
Clarendon Press, 1994); Ariel C. Armony, “The Former Contras,” in Thomas W. Walker, ed.,
Nicaragua without Illusions (Wilmington, Del.: Scholarly Resources), 207.
98
Lynn T. White III, Policies of Chaos:The Organizational Causes of Violence in China’s Cultural Revolution
(Princeton, N.J.: Princeton University Press, 1989), 280–81.
99
Joseba Zulaika, Basque Violence: Metaphor and Sacrament (Reno and Las Vegas: University of
Nevada Press, 1988).
100
Brian Crozier, The Rebels: A Study of Postwar Insurrections (Boston: Beacon Press, 1960), 158.
101
Kalyvas (fn. 81).
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 115
used by RENAMO. Young found that the most extreme atrocities were
part of a carefully drawn—and largely successful—plan to battle harden
young, mostly forcibly conscripted young guerrillas. Likewise, atrocities
committed against the population at large were concentrated in southern
Mozambique, where the FRELIMO government had a strong base.102
Paul Richards, an anthropologist who studied the civil war in Sierra
Leone, provides a nuanced analysis of the strategically motivated rebel
violence in this country:
Take, for instance, a spate of incidents in villages between Bo and Moyamba, in
September-October 1995 in which rebels cut off the hands of village women.
What clearer instance could there be of a reversion to primitive barbarity? Images
ﬂood into the mind of hands cut off for the manufacture of magic potions.
But behind this savage series lay, in fact, a set of simple strategic calculations.
The insurgent movement spreads by capturing young people. Short of food in
the pre-harvest period, some captives, irrespective of the risks, sought to defy
the movement and return to their villages where the early harvest was about to
commence. How could the rebels prevent such defections? By stopping the
harvest. When the news of rebel amputations spread in central Sierra Leone
(the rice granary of the affected region) few women were prepared to venture
out in the ﬁelds. The harvest ceased . . . Having decided not to take part in the
February 1996 elections the rebels then started to use the same tactic to scare
away would-be voters—cutting off the hands that might otherwise cast a
vote.103
Indeed, the European commissioner for humanitarian affairs described
the atrocities committed in Sierra Leone as carefully planned and centralized
rather than gratuitous and random.104
To summarize, both the perception that violence in old civil wars is
limited, disciplined, or understandable and the view that violence in
new civil wars is senseless, gratuitous, and uncontrolled fails to ﬁnd
support in the available evidence.
116 WORLD POLITICS
102
“By virtue of the numbers involved, the elimination of its supporters could not be achieved by
simply picking off a handful of local party ofﬁcials. Such violence was less evident in areas where
 inﬂuence and presence had been eliminated and  was relatively well established. In
the Gorongosa region there was reasonably good and co-operative coexistence with the civilian population
and little apparent fear. The  presence in the Zambezia seems to have been less brutal
and better organised from its ﬁrst arrival in the area.” Young (fn. 27), 132–33.
103
Richards (fn. 18), xx.
104
“Such atrocities are not part of traditional warfare in Africa. They are the result of an orchestrated
strategy to terrorize civilians, carried out by troops trained in such barbarous techniques. The
systematic pattern of these crimes, as well as the scale of the terror, do not support claims that the
rebels are retreating, isolated and beyond control. Field reports indicate that rebel movements could
not take place without communication, control and supplies from the outside. Crimes on this scale are
usually orchestrated.” Emma Bonino, “No Court to Deter the Barbarity in Sierra Leone,” International
Herald Tribune, July 8, 1998.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 116
CONCLUSION
The parallel reading of emerging research on new civil wars and overlooked
historical research on old civil wars suggests that the distinction
between them should be strongly qualiﬁed. Civil wars undoubtedly differ
from each other in a number of respects. However, the available evidence
suggests that differences tend to be less pronounced than usually
argued and that they may not array themselves neatly and dichotomously
around the end of the cold war.
The demise of the cold war potentially affected the way in which
civil wars were fought, if not their frequency. Clearly, the disappearance
of external sources of legitimation and funding provided by competing
superpowers puts a premium on local resources. Yet, the exact mechanisms
that link funding and war—from diasporas to lootable resources—
and how they affect the ways in which civil wars are fought remain
inadequately speciﬁed.
At the same time, it is often overlooked that the end of the cold war
has decisively affected how civil wars are interpreted and coded by both
participants and observers. By removing coherent, if ﬂawed, political
categories and classiﬁcatory devices, the end of the cold war has led to
an exaggeration of the criminal aspects of recent civil wars and a concomitant
neglect of their manifold political aspects. It is highly possible
that interpretations of recent civil wars that stress their depoliticization
and criminalization are attributable more to the demise of the conceptual
categories generated by the cold war than to the end of the cold
war per se.
Nonetheless, the demise of the conceptual categories engendered by
the cold war is an opportunity rather than a handicap; it allows us to
probe the core of civil wars unhindered by the constraints of externally
imposed lenses. The wrong research path would be to again coin conceptual
categories grounded in current events rather than good theory.
The study of violence is particularly vulnerable in this respect. As
Horowitz points out, it “has been characterized by considerable reactivity
to the occurrence of violent events of various classes. Theory has
twisted and turned in response to events and the changing identity of
the protagonists.”105
Flawed categories and the assumptions derived
from them undermine even the most sophisticated modeling exercises.
In turn, good theory requires sound conceptual categories and reli“NEW”
& “OLD” CIVIL WARS 117
105
Donald Horowitz, The Deadly Ethnic Riot (Berkeley: University of California Press,
2001), 33.
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 117
able empirical indicators. Such categories can only be generated by a
process of parallel analytical and empirical research. For example, patterns
of looting may or may not covary with levels of war centralization,
ethnic polarization, ideological commitment, or levels of violence. We
need to specify the key mechanisms carefully, identify the relevant empirical
indicators, and collect appropriate and accurate data. Furthermore,
the importance of historical research cannot be overemphasized.
Clearly, research on civil wars must be grounded in sustained, systematic,
and long-term observation or ethnographic reconstruction at the
mass level coupled with archival research. Such research is essential because
civil wars are particularly vulnerable to the trade-off between visibility
and signiﬁcance. Highly visible information, such as elite
discourses or widely advertised atrocities, can be outwardly misleading
and is less signiﬁcant than hard-to-collect evidence about crucial but
undertheorized and underresearched aspects of civil wars, such as the
type of warfare and actors, the forms of resource extraction, and the
patterns of violence. By illustrating the potential pitfalls of failing to do
so, this article argues that a research program for the study of civil wars
must embrace such approaches.
118 WORLD POLITICS
v54.i1.099.kalyvas 12/5/01 10:03 AM Page 118