FEMALE COMBATANTS AND THE PERPETRATION OF VIOLENCE: Wartime Rape in the
Sierra Leone Civil War
Author(s): DARA KAY COHEN
Source: World Politics , July 2013, Vol. 65, No. 3 (July 2013), pp. 383-415
Published by: Cambridge University Press
Stable URL: https://www.jstor.org/stable/42002215
REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/42002215?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access
to World Politics
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND THE
PERPETRATION OF VIOLENCE
Wartime Rape in the Sierra Leone
Civil War
By DARA KAY COHEN*
MUCH assumes of that the women current scholarship are victims and on violence men are in perpetrators. conflict settings Evenassumes that women are victims and men are perpetrators. Even
in studies of female fighters, the tendency is to take for granted that
women do not participate in acts of violence, whether by choice or
because women are assumed to play merely supporting roles to their
male combatant counterparts. As one scholar notes, women in armed
groups may more often be referred to as camp followers or dependents,
whereas men are viewed as fighters, combatants, or soldiers.1 What
roles do female combatants serve in armed groups? In particular, to
what extent do female combatants participate in perpetrating violence
against noncombatants? Does the presence of women fighters prevent
men from abusing noncombatants, as is often argued in policy circles?
Or do women, when given the opportunity and facing similar social
constraints and pressures, tend to perpetrate violence alongside their
male peers?
*The author thanks Jana Asher, Patrick Ball, Michael Barnett, Jeffrey Checkel, James Fearon,
Amelia Hoover Green, Leah Knowles, Ronald Krebs, Michele Leiby, Meghan Foster Lynch, Rose
McDermott, Fionnuala Ni Aolain, Laura Sjoberg, Jeremy Weinstein, Elisabeth Wood, the editors of
World Politics , and the anonymous reviewers for helpful comments and advice. Matthew Valerius and
Rebecca Olson provided excellent research assistance. Special thanks to Jana Asher and the Human
Rights Data Analysis Group, as well as to Macarían Humphreys and Jeremy Weinstein, for sharing
survey data, and to pride-SL and Ibrahim Bangura for arranging interviews. Earlier versions of this
article were presented at the 2009 International Studies Association annual convention and at the
United States Institute for Peace in 2010. 1 gratefully acknowledge the support of the National Science
Foundation (SES-0720440), the United States Institute for Peace, and the Center for International
Security and Cooperation (ciSAc) and the Freeman Spogli Institute for International Studies
at Stanford University. This study received Human Subjects approval from the Stanford University
Institutional Review Board.
1 MacKenzie 2009.
World Politics 65, no. 3 (July 2013), 383-415
Copyright © 2013 Trustees of Princeton University
doi: 10.1017/S0043887113000105
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
384 WORLD POLITICS
Despite entrenched views th
there is an increasing awarenes
tion as more than just cooks, c
have shown that women may b
issued guns and who perpetra
I focus on the involvement o
in which the male perpetrato
strong: the wartime rape of no
cent advances in raising aware
wartime rape,2 the convention
violence are entirely male.3 In
for understanding women's inv
women are participating in th
academics and policymakers ma
nations and solutions.
In this article, I maintain that
become perpetrators of wart
"perpetrator." As discussed be
actual rape of victims (that is,
other women were involved in
In line with other studies of ga
pants to be perpetrators of rap
Sierra Leone civil war, includ
ants and newly available surv
batants perpetrated rape alon
population-based survey data
perpetrated nearly one in four
Sierra Leone.
A growing body of research from recent conflicts suggests that
the phenomenon is not limited to Sierra Leone. Evidence from other
cases indicates that women were implicated in the commission of
2 Carpenter 2006.
3 One important exception is Sjoberg forthcoming. While I examine both male and female perpetrators
of rape, I focus only on female victims of rape. In the Sierra Leone survey data used in this
article (ABA/Benetech Sierra Leone War Crimes Documentation Survey (slwcd)), there was exactly
one report of a male rape victim. This should not be interpreted as evidence that men were rarely
raped but should be taken to mean only that men rarely reported rape. Additionally, in interviews, the
issue of male rape victims was seldom broached. Thus there are precious little data on the topic in the
context of Sierra Leone - or, indeed, in many other contexts. The research on male victims of wartime
rape is still nascent; for detailed analyses of the issue, see Sivakumaran 2007; and Stemple 2008.
4 Groth and Birbaum 1979. However, it should be noted that as a legal matter, restraining a victim
may not constitute responsibility for an act of rape. Thanks to Fionnuala Ni Aolain for this clarification.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 385
acts of brutal, including sexual, violence.5 A pop
conducted in 2010 in the Democratic Republic
found that 41 percent of female sexual violence
they were victimized by female perpetrators, as d
sexual violence victims.6 In Liberia, female fighter
the rape of women, including rape with objects
sexual crimes against men, such as cutting off the
in armed criminal gangs, paramilitary, and self-de
are reported to have committed sexual violence,
against other women and members of enemy ga
genocide in Rwanda, women were active perpetr
and sexual violence.9 A report from a nongover
cites examples of women involved in rape, inclu
ordering rape and turning over victims to be rape
the prisoner abuse scandal at Abu Ghraib, in w
Iraqi prisoners being sexually abused and humil
were broadcast by American media outlets, reve
abusing men during wartime, albeit short of rap
one of the women involved in the scandal describe
in encouraging and participating in the violence
perpetrators featured in the photographs were
quently highlighted parts of the prisoner abuse
that the victims of the sexual violence were male se
found common expectations of victim-perpetrator
lectively, these cases indicate that the involveme
time sexual violence in Sierra Leone is not an anom
there is increasing evidence of the involvement of
lations, there are currently few explanations for w
5 Vikman 2005, 42, notes that "female soldiers appear equally capable
sexual violence" and cites both Rwanda and Abu Ghraib as instances in
wartime sexual violence. Sjoberg and Gentry 2007 focus on female perp
their case studies include Abu Ghraib, Chechnya, Yugoslavia, Sudan,
6 Johnson et al. 2010.
7 Specht 2006; Advocates for Human Rights 2009.
8 Faedi 2010.
9 Jones 2002; Sharlach 1999; Wood 2009.
10 African Rights 1995; Landesman 2002. Sharlach 1999 argues that
raise serious doubts about existing theories on women and violence. Ho
has tended to assume that Rwandan women are unprecedented in their
for example, Jones 2002.
11 See McKelvey 2007 for a collection of essays on feminist response
12 Gourevitch and Morris 2008.
13 Bourke 2007; McKelvey 2007.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
386 WORLD POLITICS
While recent research on fem
may sometimes be active figh
women in wartime rape has
perpetration of wartime sex
den may be due, at least in par
not asked about the sex of p
the aforementioned 2010 su
well-known population-based
lence in conflict settings, in
one, and Uganda.15 With th
none of these studies did th
of the perpetrator.16 The su
the first to question victims
global extent of the phenome
dence suggests that the part
violence may be more commo
What explains female partici
male perpetrators of wartime
not all - of the same reason
fighters face similar social pr
and, given a similar set of cir
forms of violence. That wom
prising only because of the
policymakers often make ab
One of the main puzzles of w
across a wide variety of conte
during conflict than during p
a robust explanation for war
increased frequency during
also for the form that the vi
to the intragroup social dynam
with low levels of internal co
abduction to recruit fighters
14 E.g., MacKenzie 2009.
15 United Nations General Assembly 20
UN report is not the same survey used in
16 Although the original survey instrum
based on the surveys were reviewed for w
the articles report data based on question
assumed to be male in all cases.
17 For related arguments about the similarities in men's and women s participation in suicide terrorism,
see, for example, O'Rourke 2009; and Bloom 2011.
18 E.g., Mezey 1994; Theidon 2007.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 387
coherent armed unit. Gang rape, an especially cos
violence, is one method for creating bonds betwe
ment implies that in cases of abduction where
fighters, more women will participate in wart
there are very few women fighters, then there sh
of female perpetrators.
Understanding the role of women in combatant
cal implications for several arguments concerned
timization, and violence. First, the evidence prese
blind spot in the current research on war violenc
which are armed groups. Without correctly ident
who constitute the armed groups and, more sp
counting for the large numbers of female comba
is impossible to understand motivations for extre
conflict. Second, many of the common explana
are dependent upon the assumption that combata
tle of the previous scholarship has acknowledge
may have in perpetrating rape against noncombat
also has implications for the literature on cohesio
the effect women may have on unit cohesion. Wh
has emphasized the ease with which social cohe
that among fighters who have been abducted by
fear each other and have no basis for trusting on
unlikely to develop spontaneously. Finally, the
for arguments that are centered on the ways that
affect violence. The findings presented here conc
the bonding role of male aggression in groups;21
also suggests that many men and women are lik
participate in - rather than rebel against - m
ior. To be clear, the argument is not that wom
ral .perpetrators of violence, but rather that und
19 Exceptions include Sharlach 1999; Jones 2002; Alison 2007;
and case-specific studies of selected conflicts, such as Rwanda (for
However, only the first two articles, which both focus exclusively on
briefly mentioning the existence of female perpetrators.
20 Kier 1998. Kier argues that previous research has not identif
cohesion and military effectiveness. I make no argument about th
and military (often defined as battlefield) effectiveness; I maintain on
cohesion among fighters, which in turn has an effect on the ability o
structure of the group, and to function in only the most basic of se
serve a purpose similar to the role that basic training plays in organiz
nate individuality" (Kier 1998, 22).
21 Hudson et al. 2009.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
388 WORLD POLITICS
combatants of both sexes may f
violence and that both sexes are
similar ways.22
The incomplete view of the i
petration of violence has unde
where women are frequently e
armament, demobilization, an
countability for perpetrating v
By analyzing the role of wom
growing chorus of critiques b
the programming of governm
tions related to sexual violations
The remainder of the article
summarize three arguments dra
has implications for female com
rape committed against nonco
search in economics, sociology
ternative argument about inter
socialization. In Section II, I brie
Leone and outline the data and
sequent analysis. In Section III
the support for each of the arg
ant socialization best accounts f
female perpetrators. In Section
for future avenues of research a
I. Wartime Rape and Female Combatants
A common contention in the literature is that the presence of women
in a combatant unit will be negatively correlated with the tendency to
abuse noncombatant women.24 Wood, acknowledging that the exact
mechanism remains unclear, repeatedly notes the apparent relationship
between the large number of female insurgents in Sri Lanka and
El Salvador and the relatively low levels of sexual violence perpetrated
22 This argument thus serves both to challenge and to uphold conventional wisdom. As discussed
later, female fighters were rarely reported to perpetrate any form of violence alone or in groups of only
women. The gendered patterns in the perpetration of violence suggest that masculinity may be an
important factor regarding the types of violence that are selected (for example, rape) and who commits
it (mostly men and some women). It is most accurate to describe violence, including gang rape, as
male-led, with both female and male followers.
23 Carpenter 2003; Carpenter 2006; MacKenzie 2009; Annan et al. 2011.
24 Card 1996; Wood 2008; Wood 2009; MacKinnon 2006.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 389
by these groups. Here, I outline a set of four di
ing female combatants and the rape of noncom
drawn from the previous literature and one ori
the four arguments yields observable implica
female fighters on rape against noncombatants.
Traditional Perspective
The traditional perspective - a set of arguments that features common
themes and assumptions about gender roles and violence - reflects the
widely held belief that women are more nurturing and less bellicose
than men, either by their nature or though socialization.25 In conflict
settings, women are usually not perceived to be combatants or even to
be actively involved in the violence of war. 26 Women are more typically
conceived of as victims of war, regardless of their relative risk of death
or injury as compared with men.27 Indeed, as one scholar argues, the
need to specify the phrase "female combatant" itself implies that the
conventional understanding is of a male-only fighting force.28 While
rooted in age-old ideas about femininity, the notion that women are
inherently more peaceful than men and, furthermore, that women may
pacify male violence continues to inform current policy debates. Historian
Gerard DeGroot, an advocate for the inclusion of more women in
peacekeeping operations, argues that women in armed groups appear
to have a "civilizing effect"29 on men by preventing undesirable male
behavior, including the rape of noncombatants. In a 2010 keynote address
to UN officials, DeGroot said: "Women can improve the effectiveness
of peacekeeping operations for the simple reason that they are
not men. Women, it seems, are less inclined toward violence."30 While
not stated explicitly in DeGroot's remarks, the central mechanism is
that women essentially shame men into behaving more appropriately.
The traditional perspective implies that groups with female fighters
should be very unlikely to commit rape. The argument yields two
observable implications. First, perpetrators of sexual violence should
be almost entirely male, regardless of the proportion of women in the
25 Bourke 1999; Sharlach 1999; Alison 2004. See van Creveld 2000, 441, as an example of what I
term the traditional perspective; he argues that in recent civil wars, women are not active fighters but
rather are either victims or supporters of mens violence: "Young or old, in or out of uniform, women's
involvement in these conflicts is overwhelmingly as eggers-on, camp followers, and victims."
26 Carpenter 2003 argues that men are often viewed as "presumptive combatants," while women are
seen as "presumptive noncombatants."
27 Carpenter 2003.
28 Alison 2004.
29 Carvajal 2010.
DeGroot 2008.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
390 WORLD POLITICS
group, and, second, there sh
fighters feel shame committin
Substitution Argument
The substitution argument suggests that the presence of women in
armed groups diminishes the "need" for the rape of noncombatants. 31
The argument is founded on the basic assumption that male combatants
require sexual gratification through intercourse with women and
that this desire overrides all other factors that might serve to prevent
rape. According to this perspective, female fighters can serve as substitutes
for the rape of noncombatants in one of two ways. First, female
combatants may be raped by or forced to marry male combatants.
Or second, the fighters may be allowed to form consensual intimate
relationships with each other. Scholars have suggested that one reason
behind the institution of forced marriage in the Lord s Resistance
Army in Uganda was that forcing male combatants into intimate partnerships
with abducted women obviated the rape of noncombatants.32
The Liberation Tigers of Tamil Eelam (ltte) in Sri Lanka is another
such example. Wood explains that after an initial period of prohibiting
marriage, the ltte allowed consensual marriage within the group
and strictly prohibited all other sexual relations, including those with
noncombatants.33 Although Wood argues that these rules are reflective
of norms within the organization, it is reasonable to question how
and why these norms begin. One plausible origin of such norms is that
access to sexual relations within the group decreases the likelihood of
rape outside the group. A final example is the Revolutionary Armed
Forces of Colombia (farc) in Colombia. Gutiérrez theorizes that having
a substantial number of women in the FARC may have prevented
that group from committing acts of sexual violence against civilians.34
He notes that while sexual violence has been common against female
members of the FARC, the group has generally refrained from acts of sexual
violence against noncombatants, despite its otherwise violent history.
The argument yields a set of observable implications. First, as with
the traditional perspective, the substitution argument suggests that
groups with more female fighters should be less likely to commit rape
of noncombatants. Second, perpetrators should be almost all male,
regardless of the number of women in the group. Finally, there should
be numerous reports of rape of female combatants, forced or consen-
31 This section draws from a description of the substitution argument in Wood 2009.
32 Annan et al. 2011.
33 Wood 2009.
34 Gutiérrez 2008.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 391
sual marriage, or frequent sexual relations betwee
as relatively fewer reports of rape against civilian
Selection Argument
A third argument suggests that belligerent types of people choose to
join combatant groups in order to commit violence. Mueller argues
that the violence during the wars in the former Yugoslavia and Rwanda
was caused by a small number of criminals. These bands of people,
described as "criminal and hooligan opportunists" and "common, opportunistic,
sadistic . . . marauders,"35 Mueller argues, were loosely organized
by political elites to use coercive violence for political ends.
The combatants who behave violently are thus understood to be rough
types who enjoy violence, and war provides an excellent excuse to commit
it without the constraints of peacetime. Weinstein makes a related
argument: insurgent groups with access to material resources attract
more violence-prone recruits than do groups that rely on ideology
alone to recruit fighters. Weinstein argues that these initial conditions
predict the types of recruits that are attracted to join a group and in
turn determine whether civilian abuses will be committed on a mass
scale.
Both Weinstein and Mueller maintain that armed groups may attract
violent types, whether unwittingly through the offer of material
inducements or actively by encouraging especially violent people
to commit more violence. Research on group sexual violence in other
contexts lends credence to the selection argument. In studies of college
campus violence, scholars have suggested a selection mechanism for
why members of fraternities and sports teams seem more likely than
those in other types of campus organizations to engage in gang rape;
they argue that sexually aggressive people are more likely to join fraternities
or sports teams.36 Groups that commit rape, then, should attract
the types of people, whether male or female, who seek to perpetrate
acts of sexual violence. Observable implications include, first, evidence
of both female and male volunteers to armed groups and, second, reports
of both female and male perpetrators of sexual violence.
Combatant Socialization
A final argument is based on the idea that wartime violence - and especially
rape - results from internal group processes.37 Combatant groups
"Mueller 2000, 42-43.
36 O'Sullivan 1991.
37 1 present a version of this argument - tested on a cross-national sample with a brief case study of
Sierra Leone - in Cohen forthcoming.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
392 WORLD POLITICS
that recruit new members t
central dilemma; namely, such
force out of a collection of s
order to compel them to joi
sociology, and criminology, I
enables armed groups with fo
of loyalty and friendship from
mistrust. Researchers have established that violence can serve an important
function in organizing the structure of groups with continual
influxes of new members, and they have shown that performing acts of
brutal violence can be an important part of the process of integrating
new membership and maintaining social order among existing members.38
Violence is also believed to be useful for cutting ties to a combatant
s previous life, making it more difficult for an individual fighter
to desert, as well as creating a sense of loyalty to the group and a collective
responsibility for violent acts.39
Others have also noted the role that gang rape can play in bonding
together males in military units.40 These scholars typically trace
the source of sexual violence perpetrated by armed factions to group
norms - a militarized sense of masculinity imparted to combatants
through the training process. Wood counters that military training is
not sufficiently different across groups to account for the variation in
which armed groups rape and which do not.41 I also argue that rape
and socialization within the military unit are inextricably linked, although
for a different reason: I maintain that rape is a result of the
practical needs of combatant groups faced with the challenge of forming
a coherent fighting force of veritable strangers.
In one of the few studies of "nonconventional methods for promoting
unit cohesion," anthropologist Donna Winslow examines the practices
of the Canadian Airborne Regiment (car).42 She argues that the need
for unit cohesion is especially strong in the CAR because the members
must rely on each other for jumping out of airplanes. Winslow documents
how the CAR engaged in a variety of degrading and sometimes
sexualized rituals whose ultimate purpose was the creation of bonds
of loyalty and friendship within the group; she cites research that has
38 Kaminski 2003; Humphreys and Weinstein 2008.
39 The literature on child soldiering frequently mentions similar reasons for forcing child combatants
to commit acts of violence in their home communities.
40 Card 1996; Goldstein 2001. See also Coulter 2009 for a discussion of how rape may have been
used as an initiation rite by rebel groups in Sierra Leone.
41 Wood 2006.
42 Winslow 1999.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 393
found that the more extreme these rituals, the st
unit. Bonding within armed groups, she conclu
of inappropriate or harmful practices. Similarly,
may be greater in groups that have abducted the
in these groups must almost immediately begin
for basic survival despite having little foundation
trust. Along a continuum of violence, gang rape
that may be used to create these bonds.43
Research on group rape in peacetime has repe
increases esteem among perpetrators,44 and it can
in increasing cohesion between fighters in armed
social processes found to be important in gang
trast to rape committed by a lone offender. Ga
to be more normal and less pathological than a
difference is attributed to the fact that pressures
individuals to behave in ways that they would ne
The argument is not that individual combatan
to be more cohesive with the group that has f
or her but rather that exit from the armed grou
option - and that participation in group violenc
tinued estrangement from ones peers. The exis
of social cohesion is of great importance for th
groups to survive as groups - in order to avoid
may result in desertion, fractionalization, or u
powerful means of creating cohesion. In this w
ated through gang rape is functional for armed g
not in the sense of increasing "military effective
which social cohesion has typically been studied.4
The argument predicts that forcibly recruit
more likely than voluntary fighters to engage
in particular - against civilians. As with the se
observable implications include both male and
rape. The argument also implies that the most
should be gang rape. The observable implications
are summarized in Table 1.
43 An alternative view is that gender violence is worse than other forms of nonlethal violence; see,
for example, Sharlach 2000 on rape as a form of genocide.
44 Scully 1990; O'Sullivan 1991; Franklin 2004.
45 Bijleveld and Hendriks 2003; Brownmiller 1975.
46 Groth and Birnbaum 1979; Porter and Alison 2006; Franklin 2004.
47 Again, I make no claims about the influence of gang rape on military effectiveness and would
argue there is little correlation between the social cohesion created through acts of gang rape and the
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
394 WORLD POLITICS
Table 1
Female Combatants and Wartime Rape against Noncombatants:
Theoretical Predictions
What Effect Do
Women Fighters Observable Implications
Have on the Rape
Argument Mechan
Traditional women fighters mostly male evidence of
disrupt male shaming
bonding ~ of men by
women
Substitution women fighters mostly male rape of, or
serve as sexual partnerships
partners/rape ~ with, female
substitutes combatants
Selection fighters, both male and male and female
women and female volunteers
men, seek +
groups to commit
violence
Combatant forcibly recruited male and gang raPe 1S
Socialization combatants of female common form
either sex more neutral of rape
likely to gang
rape
II. Data, Methods, and Background
Sources of Data
I draw on three main sources of data in the following analysis. First, I
conducted thirty-four in-depth interviews with ex-combatants, twelve
women and twenty-two men, in Sierra Leone between 2006 and 2008,
including male and female representatives from all of the major fighting
factions. These interviews of perpetrators are significant because
they are among the first of their kind; most previous research has focused
on the victims of sexual violence.48 All of the interviews were
ability of armed groups to be successful fighters. Available data do not suggest that there is a correlation;
by one estimate, the RUF in Sierra Leone, which reportedly committed the vast majority of the
reported rape, lost almost two-thirds of the battles they fought over the course of the war; see Bellows
and Miguel 2009.
48 For example, Coulter 2009, 133, notes that most of what is known about rape in Sierra Leone
focuses on victims, and, more broadly, there is a need for more research on perpetrators. Baaz and Sterns
research based on interviews with perpetrators in the DRC (e.g., 2009) is an important exception.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 395
conducted on the condition of anonymity, and al
on-one. Most of the interviews were conducted with a translator in
Krio in Freetown, Sierra Leone, at the offices of pride-sl, a local NGO
that advocates for the rights of former fighters. The NGO arranged interviews
for me, drawing on its extensive network of ex-combatants.
I also recruited interviewees through informal networks of taxi drivers,
many of whom are ex-combatants. Each interview typically lasted
about two hours. 49
Every effort was made to draw interviewees from each of the major
fighting factions, and interviewees included male and female combatants,
former child fighters and adult combatants, rank-and-file
fighters and commanders. Nevertheless, it should be noted that this
analysis suffers from biases that plague any study based on a smallN,
nonrandomly selected sample. The interviews should be understood
not as representative but rather as illustrative examples of the
patterns described in the survey data. While it may be argued that
ex-combatants are unlikely to be truthful about their involvement
in wartime atrocities, the interviewees were forthcoming about their
personal experiences, as well as the experiences of their faction as a
group.50
Beyond interviews, I use two additional sources of population-based
data. For information on patterns of wartime rape, including details
on the perpetrators, I rely on a newly available data set from the aba /
Benetech Sierra Leone War Crimes Documentation Survey (slwcd), a
nationally representative survey of randomly selected households.51 For
data on the demographics of the combatant groups, including information
on recruitment patterns and the sex of combatants, I use data
from a nationally representative survey of ex-combatants conducted by
Humphreys and Weinstein.52
Before reviewing the evidence for each of the competing arguments,
I provide a brief overview of the basic patterns of violence and armed
group demographics. A broader discussion of the causes of wartime
rape in Sierra Leone is beyond the scope of this article; I focus here
49 For additional details on the interview process, see Cohen 2010.
50 Perhaps most importantly, ex-combatants would seem to have an incentive to claim they were
ordered to rape, thereby passing responsibility to their superiors - not, as is the case, that they committed
rape in the absence of orders. See the section on competing mechanisms for more detail on
this point.
51 Asher and Human Rights Data Analysis Group 2004.
52 Humphreys and Weinstein 2004.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
396 WORLD POLITICS
exclusively on which argument
combatants in gang rape.53
The eleven-year-long civil co
noted for its brutal forms of v
ing amputation, cannibalism
of child combatants. It is est
people - or about 2 percent o
served as combatants in one or
the war.54
Female Fighters in the Sie
Among the major armed gro
(ruf), Armed Forces Revolut
fense Forces (cdf), and the Sier
icant variation in the groups in
on the ex-combatant survey, f
part of the ruf, which was 24
minority of the SLA and the
tively. Overall, about 5,200 w
ment and demobilization prog
true number of women fighter
both sexes were similar, with a
groups as the war wore on.
Patterns of Rape in Sierra Leone
Rape was reported to have been widespread during the conflict, and
the most frequently reported form of rape was gang rape.56 Much of
the rape in Sierra Leone occurred in the late 1990s, at the height of
the war. Survey data from the slwcd show a spike in rape in 1998 and
1999, during which almost 70 percent of the rape for the entire course
of the war was committed.57 These two years of the war encompass the
most intense years of fighting, including the January 6, 1999, invasion
of the capital city of Freetown.58 The SLWCD survey data show that
53 See Cohen forthcoming for a brief case study that offers a more general examination of the
causes of wartime rape in Sierra Leone and draws on the same set of survey data and interview evi-
dence.
54 Humphreys and Weinstein 2008.
55 Humphreys and Weinstein 2004.
56 Asher and Human Rights Data Analysis Group 2004.
57 Data calculated from Asher and Human Rights Data Analysis Group 2004.
58 Other data sources confirm that rape reached its height in the final years of the 1990s; e.g.,
Physicians for Human Rights 2002, 48.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 397
the remaining reports of rape occurred at a fairl
course of the conflict.
Patterns among Perpetrators of Rape
Most of the wartime rape in Sierra Leone was reported to have been
committed by the ruf rebels. Based on evidence from surveys and human
rights reports, as well as the data collected by the Truth and Reconciliation
Commission, the overwhelming majority of rape victims
in Sierra Leone identified their perpetrators as members of the ruf.59
It is important to note that this finding is not an artifact of the number
of combatants in the ruf. Among the population of demobilized
fighters, only about one-third of these, or about twenty-four thousand
people, were ruf combatants. According to combatant demobilization
data, the CDF was the largest armed group, with about 50 percent of
the total combatants, and the SLA/afrc was the smallest of the major
fighting factions with around 12 percent.60 Thus the RUF committed a
disproportionate amount of the wartime rape in the conflict.
III. Evaluating the Evidence: Causes of Female Perpetration
of Wartime Rape
Traditional Perspective
The belief that women are "less inclined toward violence," as reflected
in the speech quoted earlier, is an example of how the traditional view
of women often goes unquestioned. However, a more detailed analysis
indicates that the traditional view is insufficient. Women can and do
perpetrate violence in both peacetime and conflict settings, and theories
of violence must be able to account for its existence. Recent literature
rejects the notion that women are inherently nurturing and grapples
with the reality that female combatants may be equally as prone to
brutality as men.61
Interviewees in Sierra Leone reported that women were rumored to
be especially vicious fighters and had a reputation for encouraging excessive
violence. As one ex-combatant reported: "We would sometimes
59 Physicians for Human Rights 2002; Truth and Reconciliation Commission (trc) of Sierra Leone
2004. In interviews, villagers would often identify their attackers simply as "rebels." Observers
have noted that soldiers would commit crimes out of uniform, sometimes as collaborators of the RUF,
sometimes as subterfuge; see Smith, Gambette, and Longley 2004; Keen 2005; TRC of Sierra Leone
2004. There is no way to correct for these potential forms of bias. My assumption is that, on average,
people are correct in their identifications of their attackers.
60 Humphreys and Weinstein 2004.
61 McKelvey 2007.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
398 WORLD POLITICS
put women in front when
could be the fiercest fighter
were much more hot tempe
lot of people - they were fi
insurgents in other settings r
with female combatants in
Ireland and the ltte in Sri L
in these conflicts were repute
perhaps because female com
recognition in a traditionally
The traditional perspective a
ing cannot happen in the pres
of this literature is that the
into better behavior. Resear
Sociological studies of mixed-g
act more aggressively in the p
women may become more b
female-only group.65 This fin
by a former ruf combatant w
side women: "The men felt
'Let's go!' to the men. That m
ence of women in mostly mal
influence over the way an org
in positions of power to make
they are in such positions, w
sions that are different from
The literature suggests that w
places accept the status quo an
a process known as "cooptatio
women's aggressive behavior t
62 Interviewee 26, former AFRC fighter, M
63 Interviewee 27, former RUF fighter,
were reported to be more violent than m
64 Alison 2004, 457. See also Sharlach 1
65 Bystydzienski 1993.
66 Interviewee 12, former ruf fighter, M
67 Additionally, it is uncertain whether
present in a mostly male group significan
of women in a given group may increase
numbers are irrelevant and that there m
group behavior.
68 Bystydzienski 1993.
69 Bystydzienski 1993.
/u Rabbie and Lodewijkx 1995.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 399
more, studies conducted by the US Army found that
did not gready disrupt unit cohesion;71 nor was g
factor in terms of which students were subjected to
hazing violence. 72 Gendered norms are still powerfu
but not in the manner anticipated by the traditional
The hypothesized correlation between the large
fighters and low levels of civilian rape does not hold
fact, data from Sierra Leone indicate that the propo
an armed group is positively associated with the s
mitted by the group. The estimated percentage of w
major fighting factions in Sierra Leone is display
with the estimated proportion of rape perpetrated b
data suggest that the conventional wisdom that fem
a stifling effect on violence may ultimately be misg
least, the data contradict one of the central observ
the traditional perspective: groups with more women
ted rape but actually committed more rape than did
women. At least in Sierra Leone, the mere presence o
did not appear to lessen, let alone eradicate the incid
A second observable implication of the traditio
that the perpetrators of rape should be male. This to
by the evidence from Sierra Leone. Using the SLW
percent of the reported rape was gang rape; of the g
percent was committed by male-only groups and 2
sex groups.74 In other words, the survey data indica
ticipated in one in four of the reported incidents of
71 Goldstein 2001, 201. Additionally, Armed Forces and Society has publis
years on the issue of whether women have an effect on military group co
that women do not disrupt cohesion; however, other scholars (for examp
they do.
72 Pershing 2006.
73 While this may still imply that only all-male groups within each armed group were committing
wartime rape, data indicate otherwise.
74 Using the sampling weights provided by Asher and Human Rights Data Analysis Group 2004,
there were a total of 31,759 estimated reported rapes during the course of the war. Of these rapes, 75.8
percent (24,090) were gang rapes. Of the reported gang rapes, 17,928 or 74.4 percent were committed
by male-only groups. Mixed-gender groups of perpetrators thus committed 6,163 gang rapes, or 25.5
percent, which constitute 19.4 percent of the total reported rape.
75 The potential for several types of errors or biases that may have resulted in mistakenly reporting
mixed-sex groups of perpetrators in the SLWCD survey was considered. First, many enumerators
coded the existence of female perpetrators and in all regions of the country, reducing the possibility
that the recording of the mixed perpetrator code was an enumerator error. Second, I hand checked all
the original survey forms with a mixed perpetrator code for gang rape, eliminating the possibility that
the mixed perpetrator codes were the result of a data entry error. Third, interviewee fatigue is not a
concern for potential bias, given the structure of the survey, during which enumerators were instructed
to allow the interview subjects to tell a narrative, and when needed, to ask open-ended questions for
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
400 WORLD POLITICS
Table 2
Major Combatant Groups by Proportion of Women and
Reported Rape Perpetration
Percentage of Female Percentage of All Reported
Faction Fighters in Group Rape Perpetrated by Group
ruf 24.4 85.6
sla/afrc 8.8 8.3
CDF 1.9 0.1
Sources: Female
culated from As
perpetrators of r
sla, sla/ruf), cdf
Percentage of re
groups not shown
That one-q
mixed-sex g
tion that th
unlikely. Bu
had the mo
the most ra
part of the R
neither sin
trators of r
form of viol
Finally, the
batants that
peers. The f
that shame
respondent
in front of
the sex was
clarification. The
a set of options w
odology is that vi
and that had atyp
perpetrators of g
simple way to cor
76 The fact that
only groups sugg
to the contrary,
ure 1 for a summ
77 Interviewee 17
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 401
Substitution
The substitution argument implies that female combatants may absorb
the sexual needs of their male peers, thereby reducing the likelihood of
noncombatant rape. Several problematic assumptions undermine this
perspective. First, as Wood maintains,78 the argument takes for granted
that sexual gratification can be found only through acts of sexual intercourse.
Additionally, it assumes that the central purpose of the rape of
noncombatants is the gratification associated with the sexual act. These
assumptions cannot account for the form that wartime rape takes and, in
particular, why gang rape is so common in wartime; nor can they account
for the persistence of rape with objects, a common form of violence and
an act not obviously associated with gratification for the perpetrator.
An important implication of the substitution argument is that female
fighters should be subjected to rape by male combatants or that
there should be a system of coerced partnership, such as forced marriage,
or the allowance of consensual sexual relationships. Although
there are no statistical estimates on forced marriage in Sierra Leone,
the SLWCD survey inquired about the incidence of "sexual slavery," a
separate survey item from rape and a violation that may be seen as
roughly equivalent to forced marriage.79 About 70 percent of the reported
cases of sexual slavery in the SLWCD survey were committed
either by the RUF or by "rebels." In contrast, about 4 percent of the reported
sexual slavery was committed by the SLA and less than 1 percent
was committed by the CDF. Interview evidence also suggests that rape
of female fighters was common, at least upon their initial introduction
to the RUF. "Most of the ruf women were raped when they first
joined," reported a group of female former ruf. 80 Counter to expectations,
the RUF, which had the most female combatants and the majority
of the reported sexual slavery, nonetheless committed the most rape
of noncombatants.81 A final implication of the substitution argument,
like the traditional perspective, is that perpetrators of rape are male. As
78 Wood 2009.
79 Legally, however, sexual slavery is a separate crime from forced marriage. See Mackenzie 2010
for a history of how these crimes were viewed by the Special Court in Sierra Leone. Although trial
judges initially ruled that sexual slavery and forced marriage were the same crime, the Appeals Chamber
determined that forced marriage involves harms beyond forced sex and constitutes a separate crime
against humanity.
80 Interviewees 30, 31, and 32, former ruf fighters, March 31, 2008. See also Peters and Richards
1998; Coulter 2008. More generally on sexual violence against fighters in African wars, see McKay
and Mazurana 2004.
A similar pattern is apparent in Baaz and Stern 2009, where the authors found that soldiers in
the DRC did not consider the presence of women in their units to have any effect on male fighters'
desire to rape noncombatants.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
402 WORLD POLITICS
described in the previous sectio
where women were reported to
Selection
The selection argument implies that violence-seeking women and men
join groups that permit them to perpetrate violence. The ex-combatant
survey data show that women did volunteer to fight in Sierra Leone
but in far lower numbers than did men.82 Of the fighters who served
with the ruf, 93 percent of the women reported that they were abducted,
compared with 85 percent of the men. The selection argument is
centered on explaining the incidence of wartime rape as a result of the
desires of especially violent volunteers. However, in the ex-combatant
survey none of the male or female ruf volunteers reported that they
were offered access to women or men as an enticement to join, making
it unlikely that ruf volunteers of either sex sought to join in order to
commit rape or to gain access to sexual partners.
Additionally, because the ruf consisted mostly of abducted people,
the population of any given unit was a random collection of people
within a given age range, almost none of whom could be said to have
chosen to join.83 Importantly, interview respondents who self-reported
being abducted into a fighting faction also self-reported that they had
committed wartime rape. It is not likely, therefore, that mostly ruf volunteers
committed rape. While it may be the case that respondents are
not being truthful about their decision to join, the general pattern is
likely still true - the majority of the fighters in the RUF were kidnapped
into the group, and those who were kidnapped were also perpetrators
of rape. In sum, rape was not committed only by "opportunistic joiners";
rather, there is strong evidence that rape was committed by both
volunteers and abductees.
Female perpetrators of gang rape provide strong evidence against selection
theories of wartime rape. Peacetime studies find that individual
rape perpetrators are almost always male,84 and female sex offenders
82 Humphreys and Weinstein 2004. In the Sierra Leone war, 30 percent of male ex-combatants
across all armed groups reported that they were abducted (85 percent of these served with the ruf),
while 84 percent of all female fighters reported that they were abducted (88 percent of these served
with the ruf).
83 Although evidence of how random abduction may be is limited, Blattman 2009 uses the random
nature of abduction in the Ugandan civil war to provide the basis for a natural experiment comparing
ex-combatants and noncombatants.
84 However, as Bourke 2007 shows, although only 1 percent of imprisoned rapists are women,
women who commit acts of sexual violence are likely to be charged with lesser crimes than men who
commit a similar act - thus, she argues, it is difficult to make accurate comparisons of male and female
perpetrators of rape.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 403
typically act with at least one male coperpetrator
indicate that while it is unlikely that women are
groups in order to commit rape, once they are pa
they may be coerced to participate in gang rape w
Combatant Socialization
A final argument is that rape against noncombatants may result from a
socialization process that takes place among the rank and file of certain
combatant group units. The specified mechanism applies equally to al
combatants; there is no expectation that male or female fighters will be
relatively more or less likely to participate.
As discussed in the previous section, victims reported women as perpetrators
of rape in Sierra Leone. An important observable implication
of the combatant socialization theory is that the most common form of
rape should be group rape. The combatant socialization argument can
account for why wartime brings an increase in rape by multiple perpetrators
when such violence is - perhaps universally - relatively rare
during peacetime.86 As previously described, women were reported as
perpetrators only in acts of gang rape; men, by contrast, were reported
as perpetrators in both single perpetrator rape and gang rape.
Interviews with ex-combatants shed light on how women participated
in gang rape. Members of the RUF were the only ex-combatants
interviewed who had knowledge of women in their respective factions
participating in acts of gang rape. According to ex-combatants, the
women in the ruf were involved in gang rape in a variety of ways. First,
women in the RUF not only acted as liaisons to locate potential victims
but also restrained victims while they were being raped. One woman
said: "Women would tell the men that 'I found a beautiful woman for
you.' We would help capture her and hold her down."87 Other interviewees
repeated a similar description: "Women fighters would hold
down unarmed women for men to rape."88 The fact that women would
seek potential victims provides evidence that rape was seen as a way for
the combatants to pursue intragroup acceptance as part of an organized
process among combatants. One scholar made a similar observation
about the phenomenon of gang rape in college fraternities, where providing
a victim to be raped was a means for increasing cohesion and
85 Bourke 2007.
86 In one of the few surveys on peacetime gang rape conducted in an African country, Jewkes et al.
2006 found that approximately 14 percent of respondents in a randomized survey of 1,370 young men
in rural South Africa reported participating in gang rape.
87 Interviewee 28, former ruf fighter, March 29, 2008.
88 Interviewee 24, former ruf fighter, March 28, 2008.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
404 WORLD POLITICS
solidarity among perpetrators.
on types of participation, restr
to have been the most typical w
trators, and it likely comprises
survey.
Beyond holding victims down, interviewees witnessed female combatants
raping other women with objects: "Women rebels would often
be involved in rapes - some of the women RUF would hold onto the victims'
hands, some women ruf would rape the women with bottles and
with sticks."90 Fighters noted that the women fighters did not seem to
empathize with their victims. "The RUF women did not feel any mercy
for the women [victims]," reported one former fighter.91 "Women
fighters were under the influence of drugs so they did not even think
about mercy for women being raped," one ex-combatant said.92 These
replies suggest that the group pressures that compel men to participate
in gang rape affected women in similar circumstances.
Evidence from Sierra Leone supports the idea that rape served a
bonding function. Ex-combatants reported experiencing feelings of
belonging in the aftermath of gang rape. Many of the fighters reported
that rape was an activity they viewed as "fun"93 or "entertainment."94
As one ex-combatant said: "The rebels felt pleased that they were having
so much sex, and we would brag to each other about enjoying it so
much."95 Others confirmed that rebels would often discuss their sexual
prowess with each other, recounting the number of women they had
raped during a particular raid. An ruf ex-combatant said: "[After raping]
we would make fun by saying, 'That girl was sweet.'"96 "[During
the war] it was a jungle system' and watching each other was fun, in
a way,"97 said another fighter. Another noted, "We would watch each
other and joke about how some guys were not doing it correctly."98
Finally, another confirmed, "The entire unit watches. Everyone laughs
and is jubilating."99 One of the most revealing interviews was the
following reply, to an open question about what activities the group
89 O'Sullivan 1991, 146.
90 Interviewee 24, former ruf fighter, March 28, 2008.
91 Interviewee 24, former ruf fighter, March 28, 2008.
92 Interviewee 26, former afrc fighter, March 29, 2008.
Interviewee 33, former SLA fighter, March 31, 2008.
94 Interviewee 20, former SLA fighter, June 19, 2007.
95 Interviewee 7, former ruf fighter, August 1, 2006.
96 Interviewee 12, former RUF fighter, May 29, 2007.
97 Interviewee 11, former ruf fighter, May 28, 2007.
98 Interviewee 12, former ruf fighter, May 29, 2007.
99 Interviewee 14, former ruf fighter, May 31, 2007.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 405
would do together: "The group rape of women
terward, we would feel good and talk about it a
ourselves, and laugh about it."100 These intervie
dence that rape was a cohesive activity, not somet
group. Additionally, they reveal common understa
relevant to sexual violence against civilians; simi
findings from the DRC, these interviews reveal th
the uncertainty of war and the chaos of the "jung
izing gang rape as a bonding tool.101
As additional evidence of the cohesive elements o
ants reported admiration, not disgust, for those
many rapes. In interviews, they described a cu
who had raped many women achieved a legendar
peers - one interviewee spoke with awe about a
had raped more than two hundred women.102 Exthat
those who participated in rape in Sierra L
more courageous, valiant, and brave than their pee
mitted rape were respected by their peers as "b
virile warriors.103
One of the standard measurements of milita
whether veterans stay in touch in the postwar per
vey of ex-combatants in Sierra Leone, members of
likely than members of the CDF to stay in touch w
faction after the war.104 This finding is especially
CDF members were mainly recruited by friends
members of the RUF were overwhelmingly abd
presented here suggests that social cohesion was st
in part, by participation in group rape.105
Women Soldiers versus "Females Associated with the War"
Although not the central focus of this analysis, it is important to note
that women perpetrated many common types of violence in Sierra
Leone. MacKenzie argues that despite substantial involvement in the
100 Interviewee 27, former ruf fighter, March 29, 2008.
101 Baaz and Stern 2009 note a difference in soldiers' descriptions in the DRC between "lust rape"
and "evil rape," the latter resulting from rage and including object rape and rape with the intention to
kill. The interviews presented here do not indicate such a distinction between the intentions of rape,
but the normalization of sexualized violence by the fighters is clear.
102 Interviewee 18, former RUF fighter, June 4, 2007.
103 Interviewee 18, former ruf fighter, June 4, 2007.
104 Humphreys and Weinstein 2004.
105 The temporal variation in abduction of fighters is also roughly correlated with the level of wartime
rape; see Cohen 2010.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
406 WORLD POLITICS
armed groups in Sierra Leone
female fighters are often dis
the war."106 As illustrated in F
tors committed a significant p
lence. These data provide star
active members of the fightin
in Sierra Leone achieved infam
ers. Several interview responde
such as Adama "Cut-Hand," a
to beat infants to death in ri
ditionally, the ex-combatant
women in the RUF reported th
a combat soldier (compared w
cent of female RUF reported t
upon entering their unit (com
While women may have served
groups, the evidence demonstr
many forms of violence with t
Competing Mechanisms
Here, I briefly consider several competing mechanisms for the observed
patterns of violence in Sierra Leone. One alternative argument
is that a feminist ideology in a combatant group may result in both a
higher percentage of female combatants and a decreased propensity to
rape noncombatants.109 The underlying hypothesis is that there exists
an inverse relationship between the number of women in an armed
group and the propensity to rape civilians.110 This hypothesis can be
tested empirically, and I have presented evidence that suggests the relationship,
at the very least, does not hold in Sierra Leone.
Furthermore, an argument in which an ideology both attracts female
recruits and discourages rape can apply only in conflicts where fighters
have been persuaded to join a group voluntarily.111 This was not
106 MacKenzie 2009.
107 Interviewees 30, 31, and 32, former RUF fighters, March 31, 2008.
108 All data calculated from Humphreys and Weinstein 2004.
109 Wood 2008, 346, argues that a feminist ideology both encourages girls and women to join and
discourages sexual violence." Similarly, Alison 2004 argues that women are more likely to be combatants
in "liberatory" rebel movements with long-term goals of social change.
110 Wood 2006; see hypothesis 6, where Wood notes this pattern in Sri Lanka, El Salvador, Peru,
and Colombia.
111 Wood 2008, 332, notes that the LTTE also has had policies of forced recruitment of child soldiers,
which may make it inappropriate for an ideologically based case study.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 407
Figure 1
Perpetration of Violence, by Group Size and Sex"
Source: All values calculated from Asher and Human Rights Data Analysis Group 2004.
aGroups comprising only women were not reported to have committed amputation, forced cannibalism,
destruction of property, or rape. 0.9 percent of the reported assault/beating was committed
by a group of women (not shown).
the case in Sierra Leone, where the vast majority of fighters in the RUF
were forcibly recruited and where female ruf fighters reported a higher
rate of abduction than did men.112 The dynamic may be quite different
in groups where women are being actively recruited to take up arms
or in groups with an explicit feminist liberation ideology - admittedly,
however, such groups are rare.113 Additionally, even in cases where
women have been motivated by a liberation ideology, such as in Mozambique,
scholars have found that women behaved similarly to men.114
There is little research on why armed groups may recruit women,
but one might surmise many nonideological reasons, including as a last
112 By contrast, none of the women in the CDF reported being abducted, while 2 percent of the men
reported being abducted. All data calculated from Humphreys and Weinstein 2004.
113 In Cohen forthcoming, I estimate that about 45 percent of insurgent groups between 1980 and
2009 recruited their fighters by force.
114 Coulter 2009, 138.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
408 WORLD POLITICS
resort to fill ranks when bodie
other cases supports a despera
age of military-aged men as on
recruitment of women since th
writes that desperation was a f
the LTTE.116 Ultimately, the
equality should not be assume
armed groups.
A second alternative argume
women - were ordered by their
of a broader top-down milita
ten cited as a conflict in which
But interviews revealed little evidence that combatants were ordered
to rape. Of the thirty-four ex-combatant interview subjects, about 75
percent said that they had never seen or heard a commander give an
order to rape; these responses run counter to the expectation that excombatants
may want to claim that they were ordered to commit such
violence.
Were the women forced to participate in acts of gang rape? The brief
answer is yes, women were likely pressured to participate in much of
the violence against noncombatants.118 Women were themselves often
victims of violence; several interview subjects reported becoming bush
wives to their attackers once they joined the RUF.119 Any analysis of
the relationships between women in the RUF and their bush husbands,
however, should not ignore the agency women had in the process - being
a bush wife often ensured greater security for the wife, as well as
other benefits.120 Furthermore, by participating in and encouraging the
rape of other women, the women fighters may have believed they made
themselves safer from sexual violence within the group. Rape for these
women may have in part served a self-protective purpose; this was not
mentioned in interviews, however.
1,5 Bourke 1999, 342.
116 Wood 2009, 151.
117 For example, a 2001 documentary on wartime rape in Sierra Leone is entitled, Operation Fine
Girl: Rape Used as a Weapon of War in Sierra Leone.
118 On agency and female fighters in Sierra Leone, see McKay 2005; and Coulter 2009.
119 Coulter 2009 has a similar finding; she writes that "almost all abducted women were raped" (p. 152).
bome women chose to remain with their bush husbands arter the war; some male and remale
interview subjects reported forming genuine relationships with their former victim/attacker that lasted
into peacetime. See also McKay and Mazurana 2004; Coulter 2009.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 409
The fact that women combatants were themse
lence must be considered alongside the fact th
jected to nearly identical forms of violence. With
women alike were abducted into the group and
as members of the group. Both men and wome
tions within the faction and made similar cho
with the dynamics of the group.121 Coulter main
tant not to lose sight of the fact that men who w
as fighters were also victims of violence122 and t
also limited by difficult circumstances, albeit in d
the case for women. She writes: "... one does n
petrator just because one is also a victim."123 A
female combatants rarely reported having hea
the rape of a civilian, both men and women parti
of gang rape of noncombatant women.
Lastly, some observers argue that combatants w
to compel them to commit acts of violence.124 If
difficult to ascribe rational motivations to combat
ever, evidence shows that despite rumors of w
minority of combatants reported drug use. Ov
of the combatants from both the ruf and th
ing drugs whenever they were available." Wom
reported similar rates of drug use (35 percent
tively).125 The variation in drug use is not suffic
variation in the perpetration of wartime rape a
the ruf perpetrated such a large proportion of th
IV. Conclusion
The evidence presented here supports the idea that female combatants
in Sierra Leone were active members of their fighting factions. Women
121 Jones 2002, 88, comes to the same conclusion about the Rwandan genocide: "The evidence
presented here suggests that when women are provided with positive and negative incentives similar
to those of men, their degree of participation in genocide, and the violence and cruelty they exhibit,
will run closely parallel to their male counterparts."
122 Coulter 2008.
123 Coulter 2008, 69.
124 Coulter 2009.
125 Calculated from Humphreys and Weinstein 2004. The authors collected the variable to determine
whether drugs may have been used as a material incentive in the process of "eliciting and
sustaining participation." The way the question was asked does not reveal whether using drugs was
optional, a reward, or forced.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
410 WORLD POLITICS
in the ruf perpetrated group v
rape. The participation of wom
the power of internal group dy
analysis has implications that e
Multiple examples of female pe
contexts indicate that the socia
also be prevalent in other conf
As scholars begin to grapple w
perpetrated violence, new exp
the full reality of women's wa
here is one plausible explanat
alongside men. Women and m
from within armed groups an
expected to commit similar
plication of the combatant so
of female recruits in a group s
comes. This may be especially t
and when women constitute a
in armed groups. The results
tions for how scholars underst
perpetrators of violence, as w
violence, and masculinity.
Gendered assumptions about
perpetrators of violence have h
resulted in women being largel
conflict policy processes. For e
ing investigated for war crime
tried.126 The view that women
norms about the inherent no
female fighters from benefit
and reintegration programs
largely excluded in Sierra Leon
change regarding the roles tha
atrocities. In June 2011, Paulin
minister of family affairs, bec
by an international tribunal fo
her responsibility as a superior
Finally, the data presented in
care in research on violence a
126 Husejnovic 2011.
127 MacKenzie 2009.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 411
As scholars begin to answer the call by the Un
more rigorous numerical data on wartime sexua
should be aware of gendered assumptions - oft
the conventional wisdom - when they set out to d
view protocols, and other data-collection tools.1
wartime rape should be explicit in gathering e
details of perpetrators, including the sex of the a
128 UN Security Council Resolution 1960, passed in December 201
lection on sexual violence.
References
Advocates for Human Rights, The. 2009. "A House with Two Rooms: Final Report
of the Liberia Truth and Reconciliation Commission Diaspora Project."
June 19: 245.
African Rights. 1995. Rwanda: Not So Innocent: When Women Become Killers."
London, UK: African Rights.
Alison, Miranda. 2004. "Women as Agents of Political Violence: Gendering Security."
Security Dialogue 35, no. 4 (December): 447-63.
tions of Masculinity." Review of International Studie
Annan, Jeannie, Christopher Blattman, Khristop
zurana. 2011. "Civil War, Reintegration, and Gende
Journal of Conflict Resolution 55, no. 6 (December):
Asher, Jana, ed., and Human Rights Data Analysis G
Sierra Leone War Crimes Documentation Survey (s
.org/ sierraleone/.
Baaz, Maria Eriksson, and Maria Stern. 2009. "Wh
culinity, Violence, and Sexuality in the Armed For
International Studies Quarterly 53, no. 2 (June): 495
Bellows, John, and Edward Miguel. 2009. "War an
Sierra Leon e." Journal of Public Economics 93, no.
Bijleveld, Catnen, and Jan Hendriks. 2003. Juvenile
between Group and Solo Offenders." Psychology , C
tember): 237-45.
Blattman, Christopher. 2009. "From Violence to Votin
ipation in Uganda ? American Political Science RevieA U ' * J '
Bloo
ada:
Bour
tiet
Brownmiller, Susan. 1975. Against Our Will: Men , Wome
N.Y.: Simon and Schuster.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
412 WORLD POLITICS
Bystydzienski, Jill. 1993. "Women
the Use of Force." In Ruth Howe
Use of Military Force. Boulder, C
Card, Claudia. 1996. "Rape as a W
5-18.
Carpenter, R. Charli. 2003. Women and Children First : Gender, Norms and
Humanitarian Evacuation in the Balkans, 1991-1995." International Organization
57 , no. 4 (November): 661-94.
Boys in Conflict Situations." Security Dialogue 37, n
Carvajal, Doreen. 2010. "A Female Approach to Pe
Times. March 5. At http://www.nytimes.com/2010/
-ffpeace.html, accessed October 26, 2011.
Cohen, Dara Kay. 2010. "Explaining Sexual Violence
diss., Stanford University.
dence, 1980-2009." American Political Science Review
Coulter, Chris. 2008. "Female Fighters in the Sierra Le
Assumptions?" Feminist Review 88 (April): 54-73.
Sierra Leone. Ithaca, N.Y.: Cornell University Press.
DeGroot, Gerard. 2008. "Wanted: A Few Good Women." Ke
Gender Issues in Peacekeeping Missions conference, Unit
ment Fund for Women (UNIFEM) and the Rwanda De
Rwanda, March 28-29.
Faedi, Benedetta. 2010. "From Violence against Women to
Haiti." Ph.D. diss., Stanford University.
Franklin, Karen. 2004. "Enacting Masculinity: Antigay Vio
as Participant Theater." Sexuality Research and Social Po
25-40.
Goldstein, Joshua S. 2001. War and Gender : How Gender Shapes the War System and
Vice Versa. Cambridge, UK: Cambridge University Press.
Gourevitch, Philip, and Errol Morris. 2008. "Exposure: The Woman behind the
Camera at Abu Ghraib." New Yorker , March 24. At http://www.newyorker
.com/reporting/2008/03/24/080324fa_fact_gourevitch, accessed October 26,
2011.
Groth, Nicholas A., and H.Jean Birbaum. 1979. Men Who Rape: The Psychology of
the Offender. New York, N.Y.: Plenum Press.
Gutiérrez, Francisco Sanin. 2008. "Repertoires of Violence: Colombia 1930-
2006." Manuscript.
Hudson, Valerie M., Mary Caprioli, Bonnie Ballif-Spanvill, Rose McDermott,
and Chad F. Emmett. 2009. "The Heart of the Matter: The Security of Women
and the Security of States." International Security 33, no. 3 (Winter): 7-45.
Humphreys, Macarían, and Jeremy M. Weinstein. 2004. "What the Fighters Say:
A Survey of Ex-Combatants in Sierra Leone, June-August 2003." At http://
www.columbia.edu/~mh2245/SL.htm, accessed May 17, 2013.
American Journal of Political Science 52, no. 2 (April)
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 413
Husejnovic, Merima. 2011. "Bosnian Wars Wicked
Balkan Insight. February 7. At http://www.balkan
bosnian-war-s-wicked-women-get-off-lightly, acce
Jewkes, Rachel, Kristin Dunkle, Mary P. Koss, J
Nduna, Nwabisa Jama, and Yandisa Sikweyiya. 2006
Young, Rural South African Men: Prevalence, Patte
cial Science and Medicine 63, no. 11 (December): 294
Johnson, Kristin, Jennifer Scott, Bigy Rughita, Mich
Ricardo Ong, and Lynn Lawry. 2010. "Association o
man Rights Violations with Physical and Mental He
Eastern Democratic Republic of the Congo ? Journ
Association 304, no. 5 (August): 553-62.
Jones, Adam. 2002. "Gender and Genocide in Rwanda
search 4, no. 1 (March): 65-94.
Kaminski, Marek M. 2003. "Games Prisoners Play: Al
a Total Institution." Rationality and Society 15, no.
Keen, David. 2005. Conflict and Collusion in Sierra L
grave: 242.
Kier, Elizabeth. 1998. "Homosexuals in the U.S. Military: Open Integration and
Combat Effectiveness." International Security 23, no. 2 (Autumn): 5-39.
Landesman, Peter. 2002. "A Womans Work." New York Times Magazine , September
15. At http://www.nytimes.com/2002/09/15/magazine/a-woman-s-work
.html?pagewanted=all&src=pm, accessed May 17, 2013.
MacKenzie, Megan. 2009. Securitization and Desecuntization: Female Soldiers
and the Reconstruction of Women in Post-Conflict Sierra Leone." Security
Studies 18, no. 2 (May): 241-61.
ual Violence." International Feminist Journal of Politics 12,
MacKinnon, Catherine. 2006. Are Women Human? And Oth
logues. Cambridge, Mass.: Belknap Press.
McKay, Susan. 2005. "Girls as 'Weapons of Terror' in North
erra Leonean Rebel Fighting Forces." Studies in Conflict an
5 (September-October): 385-97.
McKay, Susan, and Dyan Mazurana. 2004. Where Are the Gir
Forces in Northern Uganda , Sierra Leone and Mozambique
and after War. Montreal, Canada: International Centre for
Democratic Development.
McKelvey, Tara, ed. 2007. One of the Guys: Women as Aggr
Emeryville, Calif.: Seal Press.
Mezey, Gillian. 1994. "Rape in War." Journal of Forensic Psy
cember): 583-97.
Mueller, John. 2000. "The Banality of 'Ethnic War.'" Intern
no. 1 (Summer): 42-70.
UKourke, Lindsey. 20U9. What s bpecial about remale
Security Studies 18, no. 4 (December): 681-718.
O Sullivan, Chris. 1991. Acquaintance Gang Rape on Cam
rot and Laurie Bechhofer, eds., Acquaintance Rape: The H
York. N.Y.: John Wiley and Sons: 140-57.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
414 WORLD POLITICS
Pershing, Jana L. 2006. "Men and
Dominated Elite Military Institu
470-92.
Peters, Krijn, and Paul Richards. 1998. '"Why We Fight': Voices of Youth Combatants
in Sierra Leone." Africa 68, no. 2 (March): 183-210.
Physicians for Human Rights. 2002. "War- Related Sexual Violence in Sierra Leone:
A Population-Based Assessment." At https://s3.amazonaws.com/PHR
_ReportsZsierra-leone-sexual-violence-2002.pdf.
Porter, Louise, and Laurence Alison. 2006. 4 Examining Group Rape: A Descriptive
Analysis of Offender and Victim Behavior." European Journal of Criminology
3, no. 3 (July): 357-81.
Rabbie, Jacob M., and Hein r. M. Lodewijkx. 1995. Aggressive Reactions to
Social Injustice by Individuals and Groups as a Function of Social Norms,
Gender, and Anonymity." Social Justice Research 8, no. 1 (March): 7-40.
Scully, Diana. 1990. Understanding Sexual Violence: A Study of Convicted Rapists.
Boston, Mass.: Unwin Hyman.
Sharlach, Lisa. 1999. "Gender and Genocide in Rwanda: Women as Agents and
Objects of Genocide." Journal of Genocide Research 1, no. 3 (November): 387-
99.
Rwanda." New Political Science 22, no. 1: 89-102.
South African and Peruvian Counterinsurgencies." J
(December): 441-60.
Sivakumaran, Sandesh. 2007. "Sexual Violence against
European Journal of International Law 18, no. 2: 253
Sjoberg, Laura. Forthcoming. Rape among Women: W
Subordination . New York, N.Y.: New York University
Sjoberg, Laura, and Caron Gentry. 2007. Mothers , Monst
lence in Global Politics. London, UK, and New York,
Smith, Alison L., Catherine Gambette, and Thoma
Mapping in Sierra Leone: Violations of Internat
from 1991 to 2002." No Peace without Justice. At ht
Conflict-Mapping-Sierra-Leone- Violations-Inte
-Law-1991-2002.html, accessed October 26, 2011.
Specht, Irma. 2006. Red Shoes: Experiences of Girl Com
Switzerland: International Labour Office.
Stemple, Lara. 2008. "Male Rape and Human Rights." Hasting Law Journal 60:
605-46.
Theidon, Kimberly. 2007. "Gender in Transition: Common Sense, Women, and
War .n Journal of Human Rights 6, no. 4 (July-August): 453-78.
Truth and Reconciliation Commission (trc) of Sierra Leone. 2004. Witness
to Truth: The Final Report of the Truth and Reconciliation Commission of
Sierra Leone." At http://www.usip.org/publications/truth-commission-sierra
-leone, accessed October 26, 2011.
United Nations General Assembly. 2006. "In-Depth Study on All Forms of Violence
against Women: Report of the Secretary-General." Document A/61/122/
Add.l, July 6.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms
FEMALE COMBATANTS AND VIOLENCE 415
Van Creveld, Martin. 2000. "The Great Illusion: Women
lennium 29, no. 2: 429-42.
Vikman, Elisabeth. 2005. "Modern Combat: Sexual Viole
Anthropology and Medicine 12, no. 1 (April): 33-46.
Weinstein, Jeremy. 2007. Inside Rebellion : The Politics o
bridge, UK: Cambridge University Press.
Winslow, Donna. 1999. "Rites of Passage and Group
Airborne." Armed Forces and Society 25, no. 3 (April):
Wood, Elisabeth Jean. 2006. "Variation in Sexual Violen
and Society 34, no. 3 (September): 307-41.
tion." In Ian Shapiro; Stathis Kalyvas, and Tarek Masoud,
and Violence. New York, N.Y.: Cambridge University Pr
Rare?" Politics and Society 37, no. 1 (March): 13 1-61
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 16:07:22 +00:00
All use subject to https://about.jstor.org/terms