"Go Back and Tell Them Who the Real Men Are!" Gendering Our Understanding of Kibera's Post-election Violence

Abstract

Using a gendered analysis, this article examines the post election violence in Kibera, Kenya, between December 2007 and February 2008. Through indepth interviews with Kibera residents, the article interrogates how gender influenced violent mobilizations in Kenya's most notorious slum. Most scholarly analyses have tended to understand the post-election violence as a result of politicized ethnic identities, class, and local socio-economic dynamics. Implicitly or explicitly, these frameworks assume that women are victims of violence while men are its perpetrators, and ignore the ways in which gender, which cuts across these categories, produces and shapes conflict. Kibera's conflict is often ascribed to the mobilization of disaffected male youths by political Big Men. But the research findings show how men, who would ordinarily not go to war, are obliged to fight to save face in their communities and how women become integral to the production of violent exclusionary mobilizations. Significantly, notions of masculinity and femininity modified the character of Kibera's conflict.

Headnote

Using a gendered analysis, this article examines the post election violence (PEV) in Kibera, Kenya, between December 2007 and February 2008.Through in-depth interviews with Kibera residents, the article interrogates how gender influenced violent mobilizations in Kenya's most notorious slum. Most scholarly analyses have tended to understand the post-election violence as a result of politicized ethnic identities, class, and local socio-economic dynamics. Implicitly or explicitly, these frameworks assume that women are victims of violence while men are its perpetrators, and ignore the ways in which gender, which cuts across these categories, produces and shapes conflict. Kibera's conflict is often ascribed to the mobilization of disaffected male youths by political "Big Men." But the research findings show how men, who would ordinarily not go to war, are obliged to fight to "save face" in their communities and how women become integral to the production of violent exclusionary mobilizations. Significantly, notions of masculinity and femininity modified the character of Kibera's conflict. Acts of gender-based violence, gang rapes, and forced circumcisions became intensely entwined with ethno-political performances to annihilate opposing groups. The battle for political power was also a battle of masculinities.

If there are no elections we are friends, kama ndugu na dada, like brothers and sisters.

With elections we are enemies.

Stall owner in Laini Saba, Kibera

Broaching the subject of the 2007/8 post election violence (PEV) in Kibera with residents who lived through it inescapably changes the mood of the conversation. Heads shake, eyes drop, and a silence engulfs what might have been a lively and vigorous discussion. When people begin to talk again, it is in halting whispers. "Mimi siwezi tamani kua hivyo tena": I would not wish to go back to that again, a businesswoman in Laini Saba, one of Kibera's thirteen villages said to me.1 I heard this refrain again and again.

According to the United Nations High Commissioner for Human Rights, about 1,200 Kenyans were killed, 300,000 displaced, and millions of Kenyan shillings worth of property and goods destroyed during the 2007/8 PEV. Kibera, which lies five kilometers from Nairobi's city centre, was one of the worst-affected areas. A survey undertaken in the slum showed that as many as sixty people lost their lives in the violence - almost half the total for the whole of Nairobi (de Smedt 2009).2 Even more live with scars of rape and forced circumcision (Musau 2011). Millions of shillings worth of property was looted and burned as ethnic militias rallied behind their leaders: Raila Odinga, the Luo leader of the Orange Democratic Movement (ODM), and Mwai Kibaki, the Kikuyu founder of the rival Party of National Unity (PNU). The violence only ended after a peace agreement between the two leaders in February 2008.

After February 2008 Kibera's households organized themselves into ethnic enclaves, with each of the slum's villages becoming even more ethnically homogenized (Waki et al. 2008). One interviewee explained it as follows:

Mashimoni is considered a Luo place, Kisumu Ndogo is Luodominated but also has other tribes like Luhyias. Makina is where you find a majority of Nubians, and Laini Saba has Kikuyus and Kambas. Luos live in Gatwekera and Olympic. It didn't used to be like this, we were all mixed. But since the elections, we have developed ethnic strongholds in the slum.

The ethnic enclaves were practical for those who lived in Kibera, and made business and safety sense. If a Luo landlord had rental housing in Kikuyu-dominated Laini Saba, they sought Kikuyu tenants to safeguard their investments in the event of political violence. Across the settlement, landlords and tenants negotiated swaps to move people to areas where their ethnic group was the majority. This formula worked for everyone: landlords protected their investment and tenants and their neighbors were safer. As one Luo, a single mother who rents out eight rooms in Kikuyu-dominated Laini Saba, put it:

I rent my rooms out to Kikuyus, that way I know my business is safe. No-one pushed me to go and live in Mashimoni and pay rent there even when I own houses in Laini Saba. But I did it because I knew I would be safer and my property would be safer in case violence broke out again. It was our only guarantee, to be in a place where the majority are your tribe.

This article uses a gendered lens to understand the conflict that led to the ethnic enclaving of Kibera after the PEV. It explores the ways the social construction of masculinities and femininities shaped the production of the violence. It describes how gender necessarily intersects with ethnic, political, and class identities during times of conflict. The modes of violence manifested in Kibera after the elections cannot be explained by ethnic/political identities alone. By looking at gender, we can understand why men, who would not ordinarily go to war, are obliged to fight to "save face" in their communities. Similarly, the narratives of Kibera residents illustrate how integral women are to the production of violent exclusionary mobilizations. I argue that if we ignore gender, we cannot fully understand why forced circumcisions and rapes became part of the machinery of violence. Indeed, we cannot disentangle gender norms from ethnic and political identities in Kibera's 2007/8 PEV. The ways in which gender-based violence occurred illustrates the desire to annihilate the ethnic and political integrity of opposing groups.

The research is based on in-depth interviews with Kibera residents conducted between February 2012 and April 2013. The interviews were held with men and women from different ethnic, class, and educational backgrounds. The author also sat in meetings of organizations that were active in preventing a repeat of the PEV in the 2013 presidential race: Peacenet, Community Housing Finance, and Kenya Tunauwezo.

1. Kibera - the Background

Established around one hundred years ago, Kibera began as a settlement for aging Sudanese Nubian askaris, members of the British army's King's African Rifles.3 The Nubians were settled on land that was then a military training ground, as a reward from the British government for their loyalty and service protecting the railway line that linked Uganda to the Kenyan coast (Parsons 1997; de Smedt 2011). What started out as a settlement for aging Sudanese veterans and widows soon expanded with the migration and integration of local populations. Although the Nubians were the original settlers of Kibera they never acquired legal title to the land. The British were reluctant to grant them legal ownership because of the value of the land, and racial tensions with neighboring Europeans.

As Nairobi city grew, so too did Kibera. As one of the few places where Africans could live close to the city, it attracted ethnic groups from all over the country because of its proximity to the industrial area, city center and neighboring middle-class housing estates. Kibera's population figures are highly contested. Kenya's 2009 census counted 170,070 inhabitants, significantly lower than ear- lier estimates of 700,000 to 1,000,000 (Research International 2005; Desgroppes and Taupin 2011). Although Kibera's population is ethnically mixed, it is easy to see how group imbalances fuel ethnic and class tensions within the settlement (De Smedt 2009). Recent statistics show that Luos comprise 36 percent, Luhya 27 percent, Kamba 15 percent, and Nubians only a small fraction (Marx, Stoker, and Suri 2015). Although Kikuyus are the majority ethnic group in the country, they form only 6 percent of the population of Kibera. Nevertheless, they wield enormous economic power as one of the main structure-owning groups (together with Nubians they represent 55 percent of landlords) (Marx, Stoker, and Suri 2015; Joireman and Vanderpoel 2010).

To add to the precariousness generated by ethnic and class inequalities, neither tenants nor landlords possess legal title to land. Although there is a robust land and property market (Syagga, Mitullah, and Karirah-Gitau 2002), Kibera's land is legally owned by various Kenya government departments and parastatals (Williams 2011). Any land transactions and ownership claims in the settlement are therefore outside of state laws and regulations. This uncertainty only exacerbates ethnic tensions. When questions of indigeneity arise, the fissures between groups intensify: Nubians as the "original" settlers versus all "other groups"; long-standing communities like the Luo and Luhya versus "recent" migrants; landowners versus tenants, and so on. These structural insecurities only worsen what is already a fragile social, political, and economic situation.

1.1. The "Stolen" 2007 Elections

Witness accounts of the PEV violence across Kenya describe a series of phases (Waki et al. 2008). Interviewees in Kibera distinguish two. Violence first broke out after the electoral commission announced the victory of Mwai Kibaki on 30 December 2007. "People didn't believe it", a woman in Mashimoni village told me.

Kenya's 2007 elections were indeed different from previous ones. In an effort to promote transparency, the electoral commission broadcast the results live on television as they came in from constituencies across the country. The televised counting showed Kibaki's rival Raila Odinga in the lead in many constituencies. "Something didn't add up" said another respondent, "the elections showed Raila winning, and [this] we could see live on the television. So when Kibaki was sworn in, people got angry, that is when the violence started."

Kibera is a Raila stronghold, and the initial unrest was related to the frustration of Raila supporters who believed the election had been stolen. The violence began in Toi market at the northern end of Kibera, where discontented Raila supporters destroyed three thousand stalls belonging to traders from different ethnic backgrounds (Waki et al. 2008). Although accounts differ slightly between different parts of Kibera, the majority of those I spoke to said that the early unrest seemed "random". In other words, businesspeople of all ethnic backgrounds suffered losses and damage to property when what people saw as a legitimate protest against election rigging turned into vandalism and theft.

1.2. The Ethnic Turn

It was what respondents called the second phase that introduced a violent ethnic aspect into the conflict and resulted in the systematic carving out of ethnic enclaves. Although the ethnic partitioning of Kibera does not easily fit the xenophobic framework of discrimination against foreign nationals, it points to the multiple repertoires of violent mobilization against "the other" in contemporary African cities. After a week of violence triggered by the announcement of the election results, people began to rebuild their businesses and prepare for the beginning of the school year believing that the unrest was over (Kihato 2013). Then, in mid-January, what respondents describe as ethnically-driven revenge attacks began. According to interviews, this round of violence was triggered by the killing of Kikuyus in a church in the Rift Valley a few hundred miles away. "That is when Luos and Luhyias in Laini Saba [a Kikuyu stronghold] were targeted by Kikuyu militia, the Müngiki and flushed out", said one Luo respondent who owns rooms for rent in Laini Saba and was herself "flushed out".4 Barely a kilometer south of Laini Saba, in the Luo strongholds of Gatwekera and Kisumu Ndogo, Kikuyus were flushed out of their homes in apparent revenge for the evictions in Laini Saba. The murder of a prominent member of Raila Odinga s Orange Democratic Movement later in the month only escalated the ethnic character of the violence (de Smedt 2009; Waki et al. 2008). So a process triggered by a national political event took on a causality of its own, producing multiple forms of violence that became deeply contextualized to Kibera's local socio-economic and political dynamics.

1.3. Kibera; A Microcosm of Kenya's Gendered Ethno-politics

Kenya's history of politicized tribalism and patrimonial politics dates back to colonial rule (Lonsdale 1992). Decades of colonial and post-colonial rule have encouraged a system which links the distribution of national resources to ethnicity. During elections, as the quote introducing this article indicates, ethno-political identity is heightened as national politicians exploit ethnic differences. In the PEV there is both firm and anecdotal evidence that "Big Men" paid youth groups and gangs in Kibera to fight opposition groups there (Waki et al. 2008; Kenya National Commission on Human Rights 2008; Odallo 2010). One interviewee stated: "I know some guys who were getting paid one hundred shillings to kill people. Can you imagine what can you do with one hundred shillings? Yet these guys would take people's lives".

Nestled tightly between middle-class estates, Kibera has attracted significant scholarly attention because of its position as an opposition stronghold and its patron, the prominent political leader Raila Odinga. To explain Kibera's PEV and ethnic "flushing", some scholars point to the failure of the Kenyan state and the patrimonial nature of its politics (Branch 2008; Klopp and Kamungi 2008). These statecentered analyses show how the state's long history of oppression, its inability to address regional (ethnic) inequalities, its links to gangs, and its failing countervailing institutions foment violence and exclusionary politics at a local level (Klopp and Kamungi 2008; Cussac 2008; Maupeu 2008; Lafargue and Katumanga 2008). Yet while the Kenyan state is complicit in the production of an exclusionary and often violent politics, state-centered analyses fail to take into consideration the emergence of insurgent local political groups that seek to command and control local resources through networks of patronage and violence (Médard 2008). Although these groups may have links to leaders at the national level for example Mungiki to Kikuyu leaders, Taliban to Luo leaders (Anderson 2002), their goals are rarely to take over the state, but rather to control local resource allocations. Other scholars point to the unequal impact of global economic and development processes. Tutzer (2010) argues that while a weak state and patrimonial politics may have kindled ethnic conflict, the negative economic effects of structural adjustment programs resulting in income inequalities have left national leaders little choice but to compete for ever-dwindling resources along ethnic lines.

1.4. Intersecting National and Local Political Agendas

While broad structural processes can provide an understanding of the context within which violence can occur, it is the ways in which these intersect with local dynamics that explains why a place like Kibera might be particularly susceptible to violence. One group of authors looks at how class, historical processes of exclusion, the mobilization of gangs, and local governance structures provide a context for violence in Kibera (Shilaho 2006; Dimova 2010; Dercon and Gutiérrez-Romero 2012). Shilaho argues that the unequal nature of land allocations in Kibera fed into the violent conflict (2006) while Médard points to the emergence of local militias (2008). De Smedt highlights the limits of national patronage systems and the importance of local class dynamics in fuelling the violence in Kibera (2009).

Perhaps one of the clearest examples of the way local and national struggles intersect is the landlord/tenant relationship. Kibera is a political stronghold for Raila Odinga, and his Orange Democratic Movement (ODM) has a significant following there. Kikuyus in Kibera are in the minority, yet are seen as a wealthier business community. These class differences increase local ethnic and political tensions. When, in a bid to gain votes for the 1992 presidential election, Raila Odinga supported a Luo rental boycott, Kikuyu structure owners and Nubian landlords were hard-hit (de Smedt 2011). With the "Big Man's" protection, Luo tenants felt emboldened to stop paying rents to their largely Kikuyu landlords. While this helped Raila garner votes in Kibera, it also suited local actors who felt exploited by the high rentals of poor-quality housing - providing an outlet for class ani- mosity. The tension between tenants and landlords continues to play a part in the local struggle for power and resources, and a race for national political office often exacerbates it.

Whatever their merits, the analyses of PEV in Kibera remain gender-blind. They ignore the ways gender roles shape the nature of conflict and conversely, how gendering conflict shifts the assumptions we make about gender roles in society. Most analyses fail to tell us how the distribution of power, resources, and access to governance structures at global, national, and local levels differentially affects men and women. So while we have knowledge of how inequalities between ethnic, class, and political groups influence the nature of conflicts, we have far less insight into how gender, which crosscuts all these groups, produces and shapes conflict (Byrne 1996). The following sections explore this dimension in the context of Kibera and seek answers to the following questions: How did being a man or woman in Kibera during the PEV impact upon ethnic and political identities? How did discourses of masculinity and femininity as understood in Kibera s context influence the way women and men participated in (or abstained from) the conflict? How did gender modify the nature of the violence?

2. The Art of War and the Politics of the Mundane

Kibera s violence illustrates how decisions that are often considered personal and apolitical become highly significant in the context of unrest. Aside from the fear and the brutal nature of the violence, what people remember most is the hunger they experienced. At the time, walking through Kibera - whether by day or night - was almost literally a matter of life and death. Militias from different groups guarded the entrances to the slum and cordoned off the settlement. A man in Gatwekera said:

You walked through Kibera holding your ID high in your hand. If you were in dangerous territory you would be pleading for your life. If you were in a place with your own tribe, you spoke the language loudly so that the gangs would know who you were.

In Laini Saba, Kikuyu militias set up roadblocks to vet who came in and out. Luo or Luhyia were likely to be assaulted or murdered. The Luo Taliban militia manned entrances into Mashimoni, and Kikuyus passing through would be assaulted or murdered. These roadblocks meant that few people could leave the slum to buy food and no supplies could come in. Those who had food supplies in their shops were too frightened to open them for fear of their lives and property. When they did open, they sold their produce through a small window for no more than ten or twenty minutes at a time. "We had money but it was useless because we could not buy food", said one woman in Olympic neighborhood."Cabbage was ten shillings [approximately 14 US cents] before the violence" a second-hand clothes seller in Laini Saba related, "but during the violence it went up to seventy bob [seventy Kenya shillings, approximately one US dollar at the time). Cooking oil was too expensive for me to afford at that time."

With growing hunger, decisions about food - finding it, cooking it, and who would eat it - became the main preoccupation of Kibera s residents. The fixation on food was important not only for physical survival, but also for the survival of the factions fighting in the unrest. We know from contexts of war in Africa, South-East Asia, and Latin America that women play an indispensable part in its production (Cock 1991; Enloe 2004; Afshar 2004; Thompson 2006; El-Bushra 2004). In Kibera food, and with it women, became an integral part of the violence because the militias needed to eat in order to the secure their territories. Women in Kibera were expected to cook for the gangs who secured the perimeters of their enclaves. Some of the women I interviewed said that they were forced into this role and had no choice. Others understood it as a kind of barter: "Our men were out there fighting for us, we needed to feed them and make them strong so that they could protect us and our property," said a businesswomen in Laini Saba. In one section of Laini Saba, where business owners had millions of shillings worth of goods to lose in the violence, neighbors organized around tasks like cooking:

One person would volunteer their house to cook and we would contribute what we had to cook for the neighbors and the men. Let me make it clear, there was no planning that we are going to fight. It just happened randomly after the announcement of the results from the media.

There were areas in Kibera that were not as organized, where cooking was less communal. Nevertheless, militia youth in these areas would demand food from houses where women were cooking. "You understand that you cannot cook in Kibera without your neighbor knowing?" said a woman living with her three children in her onehundred-square-foot room in Gatwekera:

Even though we have our own homes, we live on top of each other. If I had little food to cook for my children, I could not hide it from others, and the gangs would always know where the food was. We were forced to feed them. What could I do? They were keeping us safe.

"Sometimes it was a matter of choosing whether you would go hungry or you would feed the men," said another woman who lives in Kibera's Soweto section. "I remember many nights when I had to decide who to feed - my hungry and crying child, or my husband who was out all night fighting? Can you make that choice?" The presence of the militia was also important for those with business and property in Kibera who had managed to leave after the violence broke out. "The business people paid youth to pro- tect their property. You have to be preparing food for these people to make sure they were able to protect." Kibera's women are not alone: women's roles in war are far-ranging, from supportive roles as care-givers, nurses, cooks (that reinforce their gendered roles) to combat fighting alongside men. In Thompson's words,"[women] are an integral part of the political economy of war and the financing of war" (2006, 348).

Once food supplies ran out in a household, someone had to try to find sustenance. That task sometimes fell on women because they seemed more likely to be able to plead for sympathy if they encountered opposing gangs. One Luhyia woman described her fearful encounter with Mungiki in Laini Saba while carrying vegetable oil.

I had to cross Laini Saba to go to Mashimoni where, as a Luhyia, I was safer ... I met a group of Mungiki. They were a few men, I can't remember how many, but I recognized one of them. He is elderly and lives down the road here. I know him and his wife well. I was so scared. I had bought cooking oil in town and was carrying it in a paper bag. The younger Müngiki man, the leader, took my cooking oil and said to me: "These are the women who cook for their men so that they can beat us up. We need to teach you people a lesson". I pleaded with him, crying to let me go. "I am a mother", I said to them, "I have no problems with anybody."

The woman eventually escaped unharmed after her Kikuyu neighbor vouched for her, and the gang allowed her to continue to Mashimoni with her cooking oil.

If women's decisions around cooking and food provision were politicized during the violence, so too was sleep. In Laini Saba, groups of neighbors organized to protect their lives and property. At night, women and men would form security groups on their streets and take turns sleeping. "Sleep became a luxury" one Kikuyu businesswoman said to me. "First of all you were scared to go into your house to sleep in case the Luo gangs came. Second, if you were seen to be sleeping too much, your own people would begin to say that you were not helping to keep property safe". Another businesswoman in Laini Saba said: "We were not sleeping at night, we would huddle together and protect our property."

Commonplace assumptions tend to make women invisible in war times. Even where they are acknowledged as actors in a conflict, they are understood as playing "supportive" roles to men who fight or design war strategies. This perspective not only devalues women and validates certain acts of war over others, but is blind to the ways in which conflict is produced and reproduced. If we understand conflict and its production as broader than the moment of violence, as incorporating spaces beyond the battlefield, and as integrating a diverse cast beyond soldiers, generals, or gangs - we begin to see its long production line and the multitude of actors who facilitate its progression (Enloe 2004).

3. Scrambling Gender Roles in Kibera's Conflict

If taking a step back from the battlefield allows us to see how women play a part in facilitating the manufacture of violence, zooming out even further allows us to look at the ways male and female socialization shapes conflict. Consciously or unconsciously, society tends to stereotype women as violence-loathing - the peace-loving weaker sex in need of protection - while men are seen as aggressive with an appetite for war and violence (White 2007; Enloe 2004; Giles and Hyndman 2004; Cock 1991).

Kibera's conflict both reinforced and disrupted these stereotypes. Byrne suggests that conflict creates highly volatile and fluid spaces which allow us to problematize the "unquestioning and fixed notions of masculinity and femininity at a time when gender identities and relations are, as a result of conflict, in a considerable state of flux" (Byrne 1996,31).

When I asked groups of men and women in Kibera what social expectations they experienced, they gave standard responses: people should be God-fearing and hard-working. Men are responsible for supporting their families financially and women for taking care of the home, preparing meals, and nurturing children. Yet respondents were quick to admit that even in "normal times" these idealized notions of masculinity and femininity rarely stood up to the realities of everyday life. A casual walk through the slum reveals women hard at work in stalls and hair salons while young men hang out on street corners. When asked about violence, the stock response was that men were prone to violence, but not women. "Women are peaceful. They don't like fighting," was a typical response.

But these commonly held beliefs about male and female roles became scrambled during the conflict. In fact, contrary to political and scholarly assumptions that women are peace-loving and generally absent in war, there is evidence that women actively took part in Kibera's 2007 postelection conflict in a variety of ways. Many interviewees said that women acted as spies, identifying "enemy homes" which would then be looted, burned, or forcefully occupied. One Luhyia woman in Mashimoni said:

Women were generally the ones who pointed out Kikuyu homes in this area. It is because as women we are the ones who know our neighbors and who lives where in the area. They would point to the homes of other tribes.

A Luo woman in Gatwekera related how:

We women would find stones and fill buckets. By this time, we had left our children [safe] with relatives outside the slum. Women and men - we would all sleep outside. When we heard [the call] mawe! stones! we would get the stones from our compound and supply the men who were guarding us.

3.1. "Men Who Behave Like Women"

I was sitting at a makeshift stall along the train tracks in Laini Saba talking to a group of women gathered in heated conversation. The six women were from diverse ethnic groups, and were talking about "men who behave like women", referring to men who stayed at home during the unrest. There was laughter as they talked about some men's cowardice, and the way some of them had taunted their partners to go out and fight. The conversation continued with women making jibes that so-and-so's husband is not a real man because he would not go out to fight.

My neighbor was telling me how she tried to chase her husband out of the house. Other men were fighting and hers was just sitting in the house complaining he was hungry.

Another added:

Even me I told mine to stop sitting in the house like a coward, "go and fight like a man" I said to him, "stop being a woman!"

A gendered approach allows us to analyze these conversa-tions and understand how aggressive militarized masculinities become the ideal in times of violent conflict. The jibes at "men who behave like women" are aimed at men considered too weak and cowardly to fight. These "feminizing" insults not only draw attention to how women contribute to fuelling violence even when they are not actively involved in it themselves; they are intended to emasculate these "cowardly" men. The slights go right to the heart of understanding how gender influences the production and reproduction of violence. Jacklyn Cock describes how white women in apartheid South Africa socialized boys into aggressive masculinities, all the while maintaining their gendered roles as peace-loving mothers (1991). When we look at gendered relationships in a society, and expectations of men and women, we see how men, who might ordinarily not go to war, can be pressured to do so. Coulter points out how: "Men who refuse to fight will often be ridiculed, jailed, or even killed for their cowardice, or lack of manliness" (2008,57).

Masculinist notions also serve as a powerful tools for making men into soldiers because military forces encourage aggressiveness and competitiveness while censuring emotional expressions and denouncing physically weak soldiers as effeminate. (White 2007,866)

In a militarized context, being a "real man" is associated with aggression and a capacity for violence. Yet the meaning a "real man" is not static. It shifts during times of conflict and in "normal" times. When discussing the 2007/8 viol- ence, few interviewees fail to mention the "idlers", the young men with no jobs who spend their days sitting around street corners. These idlers, I was told, are violent and dangerous - vagrants in the day, muggers and thieves by night. The idlers' aggression is criminalized in normal times, when gendered roles in Kibera are normalized, because it is targeted at ordinary citizens trying to make an honest living. However, their violence is a celebrated resource during times of conflict because it protects families and neighborhoods from opposing ethnic factions. This distinction between "good" and "bad violence" seems contradictory. Wamucii and Idwasi point to the distinction between legitimate security and delinquent violence in Nairobi slums (Wamucii and Idwasi 2011). When I posed the question why aggression was rejected during times of peace and demanded, even celebrated, in times of conflict, the responses from men and women of all ethnicities and social groups were similar: while violence was unwarranted during times of war or peace, many argued that wartime aggression was socially sanctioned. In the words of a mother whose son participated in the violence, "the men were fighting for the tribe .. .to protect the community, they were doing good. If they were not there, the Kikuyu gangs would have come here." As Coulter (2008,55) points out: "War is not a-social, but it creates its own social orders ..." In a militarized context, new norms and values determine acceptable and unacceptable modes of action and interaction. In times of conflict the ideal model of manhood is a militarized one - intricately intertwined with violence and obligations to protect the community and fight the enemy other.

3.2. Masculinized Women

If "feminized" men were ridiculed for being cowardly and effeminate, "masculinized" female killers instilled horror. A Luhya woman who owns a hair salon in Laini Saba told me she witnessed Kikuyu women kill a young man.

It was very difficult for me, very difficult. It is hard to see someone, a young boy, being killed. But let me tell you, it is harder when it is women who are doing the killing. A woman like me. That is not how women are supposed to be.

These destabilizing acts allow us to see masculinity and femininity as characteristics that can be attributed to both male and female biological categories. The Kikuyu womens machete attack on the Luo boy can be seen as a gendered act, one that projects domination over the emasculated and subjugated other. Conflict scholars have written about the importance of separating sex from gendered notions of masculinity and femininity. Bohan makes the point that "the factors defining a particular transaction as feminine or masculine are not the sex of the actors but the situational parameters within which the performance occurs" (1997, 39). In other words women can perform masculinity in the sense of power, domination, and violent subjugation. West and Zimmerman suggest that gender is an act, a social interaction that is separate from sex or sex category (1987).

Kibera's post election conflict reveals the fluidity of gender roles and obligations. The conflict illustrates the fiction of ideal gender types, because it disrupts commonly held assumptions that men are inherently war-loving and women inherently peaceful. Gender roles and responsibilities are socially constructed, localized, and subject to change depending on the context. The male/war female/ peace binaries are not fixed to a biological category. Paying attention to gender in times of conflict allows us to think of gender as actions - as enactments of masculine and feminine traits that are independent from biological sex. As such, women can take on masculine aggressive and domineering traits and men stereotypically feminine - caring and submissive - characteristics. Kibera's example allows us to understand masculinity and femininity as linked more to power (or its absence) than to biological attributes.

To acknowledge men and women's complicity in conflict, however, is not deny that they are victims. To be sure, both women and men can be victimized and empowered by war (Zarkov 2001). At a policy level, gendering conflict is important in devising appropriate responses that recognize both men and women's agency and victimhood in times of war and peace.

4. "Go Back and Tell Them Who the Real Men Are" - the Battle for a Hegemonic Ethno-masculinity

The previous section unsettled the idea of fixed binaries between men and women. The scrambling of genders allows us to understand that characteristics ascribed to men and women are social constructions and not inherent in their biology. Women can enact "masculine" traits just as men can perform "feminine" ones.

This section extends the analysis by arguing that notions of masculinity and femininity attach not only to individuals but to political, ethnic, or religious groups (Skjelsbaek 2001). This proposition provides a compelling framework for explaining the nature of Kibera's conflict and the reasons it manifested in widespread incidents of gang rapes and forced circumcision. Although there is legal contestation as to whether acts like forced circumcision comprise sexual violence, I agree with Skjelsbaek that acts such as "rape, forced prostitution, forced marriage, forced circumcision and forced nakedness" are sexual violence (Skjelsbaek 2001,212-13; also Carpenter 2006).

At a women's hospital in Mashimoni, Kibera, a nurse recounted how widespread forced circumcision was during the PEV.

We knew that women could be raped and many, many women I know were raped during that time. But in Laini Saba, men were being killed. They said they were circumcising them, but they would tell our men to put their penis' on the railway line. Then they would cut them. They called it circumcision, but how do you cut a man's penis in half and say that is circumcision?

Witnesses told the Waki Commission (established by the Kenya government to investigate the PEV) of similar incidents of forced circumcision and mutilation. A doctor testifying to the commission said that what he witnessed was "pilary amputation" where his patient had his "whole penis actually cut" (Waki et al. 2008,258). A Luo woman in Huruma estate, north-east of Nairobi city center, said:

I heard many people outside saying that "even here there are some ODM people we want to circumcise" They were many and were making a lot of noise. They pushed the door saying that "Kihii [Kikuyu for uncircumcised man] you are the ones troubling us"(Waki et al. 2008,259)

Why were forced circumcisions part of the ethno-political battle in Kibera? What social relationships and processes made the performance of this act such a significant part of an ethnic and political conflict? Within the context of political and ethnic strife, gender allows us to see the forced circumcision of Luos by Kikuyu gangs as an act of domination.

To provide some context: Unlike the Kikuyu, Luo men do not traditionally get circumcised but have other comingof-age-rituals. Conversely, male circumcision is an important marker of adulthood in Kikuyu tradition, signifying the transition from boy to man. Within Kenya's political context, circumcision has been appropriated as a symbol of political power and wealth and an assertion of Kikuyu superiority over Luos (Ahlberg, Njoroge, and Olsson 2011). The political rivalry between the Luo and Kikuyu, and the accompanying ethnic chauvinism, has a long colonial and post-colonial history (Atieno-Odhiambo 2002). Kenya's first president, Kenyatta, extolled the virtues of circumcision, and the Kikuyu elite frequently insist that they cannot be ruled by "boys" (Ahlberg, Njoroge, and Olsson 2011; Kamau-Rutenberg 2009). This hegemonic masculinity is not simply an elitist attitude towards the other, but an endemic feature of Kikuyu discourse (Mucheru-Oyatta 2007; Corey-Boulet 2011). Popular Kikuyu musician Kamande wa Kioi's song, Uhuru ni Witu (Uhuru is ours, referring to Kenya's fourth president Uhuru Kenyatta) puts it this way:

Wegutha githuri na hiiki kai kiigi iriya nyukwa?

You thump your chest about Hague/' is Hague your mother's?

Ni kuri kirumi kia Jehova, riria ombire thi na iguru

There is a curse from Jehova when he created heaven and earth

Abiristi mataruaga matigaathe Isiraeli

Philistines who do not circumcise cannot rule Israel

Iburahimu agia Jehova, erirwo athii akagirimwo

When Abraham gave Jehova trouble, he was told to get circumcised,

Nawe General wa Migingo ruhiu no ruraria thio

And you General of Migingo [referring to Raila Odinga] your knife is being sharpened.

Wa Kioi was charged for hate speech, but the idea that Luos cannot lead Kenya because they are uncircumcised remains a feature of Kenyan political discourse. Seen in this context, the penis mutilations asserted the dominance of Kikuyu masculinity - intricately intertwined with politics - over Luos as a group. The accompanying spectacle and performance - demanding that a man strip, jeering at his penis, pulling his foreskin, ordering him to place his penis on the railway line, and mutilating it with a panga (machete) - was a message, not just to the uncircumcised man, but to the ethnic and political group he represented. Forced circumcisions projected a hegemonic Kikuyu masculinity and this gendering act symbolized the Kikuyus assertion of power and domination over their Luo rivals.

If we treat ethnic and political groups as gendered, we can understand the character of this violence. Here, the perpetrator's ethnic group is perceived as masculine, powerful, and composed of "real" men. The victim, the uncircumcised man and his group, are emasculated through the forced circumcision. The victim is not considered a real man because he is not circumcised, and his manhood is further devalued in the domineering and subjugating act of forcibly "circumcising" him. He and his ethnicity are feminized in this gendered act. Hague (1997) examines gendered acts of feminizing and masculinizing group identities in the Bosnia-Herzegovina war. He shows how the masculine attributes of domination, violent subjugation, and power were attributed to the national identity of "Serb" and "Bosnian Serb" (Hague 1997). Similarly, Skjelsbaek points out that "the victim of sexual violence in the war-zone is victimized by feminizing both the sex and the ethnic/ religious/political identity to which the victim belongs, likewise the perpetrator's sex and ethnic/religious/political identity is empowered by becoming masculinized" (Skjelsbaek 2001,225).

The nature of the beatings and circumcisions in Laini Saba prompted retaliation in Luo-dominated areas of Kibera. Luo members of the Taliban admitted that they used rape to send a message to their Mungiki rivals. Kikuyu factions were as guilty of rape as other factions in Kibera, but the following story, told by a woman of Kamba origin (seen as Kikuyu allies), illustrates particularly sharply how the strategy of rape became a way of recovering Luo masculinity threatened or weakened by the conflict:

I was coming back from Ukambani where I had gone to vote. I didn't know that there was any problem in Kibera so me I was just passing through to get home. I had not walked very far into Kibera when I met a group of youth. At first I did not even notice that they were all Luo. "Where are you going?" they asked me. They started to behave very aggressively towards me. I don't remember what happened next, but I was raped by all of them. You think we are not men? You think we are not men? That is what they said to me. When they finished they said to me "go back and tell them who the real men are".

Conflict authors write about the use of spectacle in violent acts in times of war as separating everyday sexual violations from wartime acts (Sharlach 2000; Carpenter 2006; Gerecke 2010). Although there is gender-based violence in Kibera during "normal times," it is important to note that the incidences of gang rapes and forced circumcisions increased dramatically during the political crisis (Waki et al. 2008). These sexual acts of violence are deeply embedded in the ethnic and political conflict.

Gang rapes like the one described by the Kamba woman were not simply an act against the woman, but against the ethnic group she represents. In times of conflict rape has a significance that moves beyond the subjugation of women: it becomes a deliberate strategy to decimate the opposing side (Holmes 2013)."Rape is used to rip apart the fabric of society not only by undermining women but also their men" (Afshar 2004,48). Indeed rape had multiple consequences on individuals, families, and communities in Kenya during the 2007/8 PEV. The Waki Commission found that rape had resulted in men leaving their wives. Women were traumatized not only by the physical violation and their exposure to sexually transmitted diseases including HIV/AIDS, but also the disintegration of their families and the hostility they faced in some of their communities (Waki et al. 2008). And this is precisely why rape became such an important strategy in the conflict: it became a means of breaking down, feminizing, the other side. The ringing words of the rapist gang "go back and tell them who the real men are" encapsulate this battle of masculinities. Through the spectacle of the rape, the Luo gang members were asserting their hegemonic masculinity over the Kikuyu and their Kamba allies.

Conclusion

Gender and accompanying notions of masculinity and femininity help explain aspects of Kibera's conflict that remain inexplicable in the dominant political, class, and ethnic frameworks. Using a gendered analysis reveals how gender intersects with political affiliation, class, and ethnicity to produce a violent exclusionary politics. Looking closely at what happened in Kibera disrupts fixed binary conceptions of gender roles and reveals how the hegemonic construction of masculinities and femininities fueled the conflict and shaped its outcome. As the violence unfolded it became clear that the battle for political power was also in fact a battle of masculinities between the Kikuyu and Luo groups and their allies. What was at stake was not only the loss of a national election, but the loss of an ethnic machismo, the loss of manhood. We see this in the acts of violence and the accompanying discourses of "real men" versus "boys". Indeed Kibera's PEV shows us how gender was at the very heart of the violence - its inscription into Luo and Kikuyu political identities reveals a frightening yet illuminating aspect of the conflict.

Sidebar

I would like to thank Aurelia Segatti, Laurent Fourchard, and Peter Geschiere for comments on earlier drafts of this article.

Footnote

1 Kibera's villages include Kianda, Soweto West, Raila, Gatwekera, Makina, Kisumu Ndogo, Kambi Muru, Kichinjio, Mashimoni, Laini Saba, Lindi, Soweto East and Silanga

Footnote

2 Athough the word "slum" is sometimes considered derogatory and inappropriate amongst development practitioners and academics, I use the term in this article because it is how Kibera residents refer to their settlement.

3 For a comprehensive historical analysis, see (Parsons 1997; Williams 2011; de Smedt 2011)

Footnote

4 For more on the Mungiki and its formation see Anderson (2002) and Wamue (2001).

Footnote

5 Referring to the International Criminal Court in the Hague where Uhuru Kenyatta had been charged for crimes against humanity and his role in the PEV. The charges were dropped.

References

References

Afshar, Haleh. 2004. Women and Wars: Some Trajectories towards a Feminist Peace. In Development, Women and War: Feminist Perspectives, ed. Haleh Afshar and Deborah Eade, 43-59. London: Oxfam Publications.

Ahlberg, Beth M" Kezia Njoroge, and Pia Olsson. 2011. We Cannot Be Led by a Child: Forced Male Circumcision during the Post-Election Violence in Kenya. Uppsala, Sweden: Department of Women's and Children's Health, Uppsala University and Karaborg Institute for Research and Development. http://aegis-eu.org/archive/ecas4/ecas-4/panels/41-60/panel-44/AhlbergNjoroge-01sson-Full-paper.pdf.

Anderson, David. 2002. Vigilantes, Violence and the Politics of Public Order in Kenya .African Affairs 101:531-55.

Atieno-Odhiambo, Elisha Stephen. 2002. Hegemonic Enterprises and Instrumentalities of Survival: Ethnicity and Democracy in Kenya. African Studies 61 (2): 223-49.

Branch, Daniel. 2008. At the Polling Station in Kibera. London Review of Books 30 (2): 26-27.

Byrne, Bridget. 1996. Towards a Gendered Understanding of Conflict. IDS Bulletin 27 (3): 31-40.

Carpenter, R. Charli. 2006. Recognizing Gender-Based Violence against Civilian Men and Boys in Conflict Situations. Security Dialogue 37 (1): 83-103.

Cock, Jacklyn. 1991 .Colonels and Cadres: War and Gender in South Africa. Cape Town: Oxford University Press.

Corey-Boulet, Robbie. 2011. In Kenya, Forced Male Circumcision and a Struggle for Justice. The Atlantic, August 1. http://www.theatlantic.com/international/ archive/2011/08/in-kenya-forced-male-circumcision-and-a-struggle-for-justice/242757/.

Coulter, Chris. 2008. Female Fighters in the Sierra Leone War: Challenging the Assumptions? Feminist Review 88 (1): 54-73.

Cussac, Anne. 2008. Institutional Shortfalls and a Political Crisis. Les Cahiers D'Afrique de L'Est 38:232-58.

Dercon, Stefan, and Roxana Gutiérrez-Romero. 2012. Triggers and Characteristics of the 2007 Kenyan Electoral Violence. World Development 40 (4): 731-44. doi:10.1016/j.worlddev.2011.09.015.

Desgroppes, Amélie, and Sophie Taupin. 2011. Kibera: The Biggest Slum in Africa? Les Cahiers de l'Afrique de l'Est 44:23-34.

de Smedt, Johan. 2009. "No Raila, No Peace!" Big Man Politics and Election Violence at the Kibera Grassroots. African Affairs 108 (433): 581-98. doi:10.1093/afraf/adp043.

Dimova, Margarita. 2010. Bullets Beyond Ballots: Re-Examining the 2007-2008 Post-Election Violence through Longer-Standing Youth Mobilisation Patterns in the Slum Area of Kibera, Nairobi. M.A. thesis, Utrecht University.

El-Bushra, Judy. 2004. Fused in Combat: Gender Relations and Armed Conflict. In Development, Women and War: Feminist Perspectives, ed. Haleh Afshar and Deborah Eade, 152-71. London: Oxfam Publications.

Enloe, Cynthia H. 2004. The Curious Feminist Searching for Women in a New Age of Empire. Berkeley: University of California Press.

Gerecke, Megan. 2010. Explaining Sexual Violence in Conflict Situations. In Gender, War, and Militarism: Feminist Perspectives, ed. Laura Sjoberg and Sandra Via, 138-54. Santa Barbara: Greenwood.

Giles, Wenona Mary, and Jennifer Hyndman. 2004. New Directions for Feminist Research and Politics. In Sites of Violence: Gender and Conflict Zones, ed. Wenona Mary Giles and Jennifer Hyndman, 301-15. Berkeley: University of California Press.

Hague, Euan. 1997. Rape, Power and Masculinity: The Construction of Gender and National Identities in the War in Bosnia-Herzegovina. In Gender and Catastrophe, ed. Ronit Lentin, 50-63. London: Zed.

Holmes, Georgina. 2013. Women and War in Rwanda: Gender, Media and the Representation of Genocide. Vol. 39. IB Tauris.

Joireman, Sandra, and Rachel Sweet Vanderpoel. 2010. In Search of Order. In World Bank Conference on Land, Policy and Administration. Washington, D.C: World Bank. http://siteresources.worldbank.org/EXTARD/Resources/336681 -1236436879081 /5893311-1271205116054/Joireman.pdf.

Kamau-Rutenberg, Wanjiru. 2009. Watu Wazima: A Gender Analysis of Forced Male Circumcisions during Kenya's Post-Election Violence. Wajibu 24 (3): 3-5.

Kenya National Commission on Human Rights. 2008. On the Brink of the Precipice: A Human Rights Account of Kenya's Post-2007 Election Violence. New York: Human Rights Watch.

Kihato, Caroline Wanjiku. 2013. The Bookseller of Kibera. The M&G Online, August 28. http://mg.co.za/article/2013-08-28-the-bookseller-of-kibera/.

Klopp, Jacqueline, and Prisca Kamungi. 2008. Violence and Elections: Will Kenya Collapse? World Policy Journal 24 (4): 11-18.

Lafargue, Jerome, and Musambayi Katumanga. 2008. Kenya in Turmoil: PostElection Violence and Precarious Pacification. Les Cahiers D'Afrique de L'Est 38:11-32.

Lonsdale, John. 1992. The Political Culture of Kenya, Occasional Paper 37. Edinburgh University: Centre of African Studies, http://rossy.ruc.dk/ojs/ index.php/ocpa/article/viewFile/3906/2080.

Marx, Benjamin, Thomas M. Stoker, and Tavneet Suri. 2015. There Is No Free House: Ethnic Patronage in a Kenyan Slum," paper presented at the Working Group in African Political Economy, Columbia University, http://www.mit. edu/~tavneet/Marx_Stoker_Suri.pdf.

Maupeu, Hervé. 2008. The Role of Religious Institutions. Les Cahiers D'Afrique de L'Est 38:311-40.

Médard, Claire. 2008. Elected Leaders, Militias and Prophets: Violence in Mount Elgon (2006-2008). Les Cahiers D'Afrique de L'Est 38:373-95.

Mucheru-Oyatta, Njeri. 2007. Mind Boggling: Chapter 75. Mind Boggling blog. December 29. http://njeri-nm.blogspot.com/2007/12/chapter-75.html.

Musau, Nzau. 2011. Kenya: Silent Victims - the Case for Kibera and Kisumu Shootings Investigation By ICC. Star (Nairobi), December 29. http://allafrica.com/stories/201112300207.html.

Odallo, Beatrice Nyawanda. 2010. An Analysis of the Post-2007 General Election Conflict Mediation Process in Kenya. LL.M thesis, University of Pretoria, South Africa.

Parsons, Timothy. 1997."Kibra Is Our Blood": The Sudanese Military Legacy in Nairobi's Kibera Location, 1902-1968. International Journal of African Historical Studies, 87-122.

Research International. 2005. Kibera Social and Economic Mapping: Household Survey Report. Nairobi: Government of Kenya and UN HABITAT.

Sharlach, Lisa. 2000. Rape as Genocide: Bangladesh, the Former Yugoslavia, and Rwanda. New Political Science 22 (1): 89-102.

Shilaho, Westen Kwatemba Godwin. 2006. Ethnic or Class Conflict? The Politics of Conflict in Kibera, Nairobi, Kenya. M. A. thesis, University of the Witwatersrand, Johannesburg. October 21. http://wiredspace.wits.ac.za/ handle/10539/5776.

Skjelsbaek, Inger. 2001. Sexual Violence and War: Mapping out a Complex Relationship. European Journal of International Relations 7 (2): 211-37.

Smedt, Johan Victor Adriaan de. 2011. The Nubis of Kibera: A Social History of the Nubians and Kibera Slums. Ph.D. thesis, Leiden University, https://openaccess.leidenuniv.nl/handle/ 1887/17638.

Syagga, Paul, Winnie Mitullah, and Sarah Karirah-Gitau. 2002. Nairobi Situational Analysis Supplementary Study: A Rapid Economic Apparisal of Rents in Slums and Informal Settlements. Nairobi: Government of Kenya and UNHABITAT.

Thompson, Martha. 2006. Women, Gender, and Conflict: Making the Connections. Development in Practice 16 (3-4): 342-53. doi: 10.1080/09614520600694976.

Tutzer, Mirjam. 2010. Causes for Post-Election-Violence 2007/2008 in Kenya Reconsidered.

Waki, Philip, Gavin McFadyen, Pascal Kambale, David Majanja, and George Kegoro. 2008. The Kenyan Commission of Inquiry into the Post-Election Violence - Waki Report. Nairobi: Commission of Inquiry into Post-Election Violence. http://www.kenyalaw.org/Downloads/Reports/Commission_of_In quiry_into_Post_Election_Violence.pdf.

Wamucii, Priscilla, and Peter Idwasi. 2011. Social Insecurity, Youth and Development Issues in Kenya. In Africa in Focus: Governance in the 21st Century, ed. Kwandiwe Kondlo and Chinenyengozi Ejiogu, 188-99. Pretoria: Human Sciences Research Council, http://www.hsrcpress.ac.za/downloadpdf. php?pdffile=files/PDF/2283/eBook_Africa_in_Focus.pdf&downloadfilename=Africa%20In%20Focus%20-%20Entire%20ebook*page=230.

Wamue, Grace Nyatugah. 2001. Revisiting Our Indigenous Shrines through Mungiki. African Affairs, 453-67.

West, Candace, and Don H. Zimmerman. 1987. Doing Gender. Gender and Society 1 (2): 125-51.

White, Aaronette M. 2007. All the Men Are Fighting for Freedom, All the Women Are Mourning Their Men, But Some of Us Carried Guns: A Raced0Gendered Analysis of Fanons Psychological Perspectives on War. Signs: Journal of Women in Culture and Society 32 (4): 857-84. doi:10.1086/513021.

Williams, Christopher. 2011. Slum Dwellers, Bankers and Bureaucrats: A Historical Comparative Analysis of State Formation, Urbanization and the United Nations in East Africa. Ph.D. thesis, New School, New York.

Zarkov, Dubravka. 2001. The Body of the Other Man. In Victims, Perpetrators or Actors, ed. Caroline O. N. Moser and Fiona C. Clark, 69-82. London: Zed.