10
Body count and religion in the Boko Haram crisis: Evidence from the Nigeria Watch database
p. 213-236
Résumé
This chapter tackles the issue of victims and religion in the Boko Haram crisis through a quantitative, critical examination of the deaths it has resulted in among Christians and Muslims in the affected states of Nigeria. After a brief discussion of the sectarian movement’s changing ideology towards Christians, 317 Boko Haram-related events, identified in the Nigeria Watch database from July 2009-December 2012 data, are analysed to provide the most reliable possible body count for the period. The study assesses the status – civilian, Boko Haram member, JTF member, etc. – and the religious affiliation of all reported individuals who died as a result of the conflict. As it proved impossible to determine the religion of more than 60 per cent of the fatalities (mostly civilians), available demographic data were used to provide estimates. Despite an obvious increase in deadly attacks against Christian civilians, the combined data suggest that approximately two civilian victims out of three were Muslims. This challenges the commonplace temptation to understand the crisis in terms of a religious confrontation between Muslims and Christians. It suggests that such a representation may be fed by an overreporting of Christian victims compared with Muslim victims, in both Nigerian newspapers and international reports.
Texte intégral
Introduction
1A small Islamic sect created in the early 2000s by Mohammed Yusuf with a few dozen followers in Borno and Yobe states has become, in less than ten years, a high-profile terrorist organisation known as Boko Haram, strong enough to challenge the Nigerian state and threaten the stability of bordering countries such as Niger, Chad, and Cameroon. The essence and goals of the radical group, as well as its methods, targets, and victims, have evolved over time and are not easily retraceable. In the Western narrative, the conflict is frequently presented as a religious war opposing Islamic extremists to Christian communities. Christian killings have indeed been widely reported, in both the Nigerian and international media, strengthening the idea that Christian communities are the primary victims of the crisis in terms of fatalities. In 2013, declarations by Nigerian President Goodluck Jonathan, however, stressed that Muslims are the most numerous victims of the ongoing struggle (Daily Trust 2013). The issue is undoubtedly sensitive as the Christian Association of Nigeria (CAN) immediately replied to the statement by accusing President Jonathan of “distorting facts”, and recalling “the statistics” that “are there for everybody to see how Christians have been massacred by the terrorists” (The Nation 2013a). The debate is not meaningless; it gives the conflict a particular definition and chooses to analyse it exclusively in terms of religious confrontation. This perspective, which is undoubtedly an important characteristic of the current crisis, disregards a number of other factors. Nevertheless, even when tackling the issue of casualties in terms of religion, a major problem remains: statistics are not there. Indeed, data presented to support both President Jonathan’s and CAN’s affirmations are nonexistent at worst, and rather weak at best.
2In this article, we use the Nigeria Watch database to build a statistical basis for a discussion on the religious affiliation of the fatalities in the conflict and the preferred targets of the radical group. The aim is to empirically examine the identity of the victims, through their religion (basically Islam and Christianity) and also their status (civilians, Boko Haram members, Nigerian security force members, other officials, etc.). The Nigeria Watch database provides material on violent deaths in Nigeria from 2006 onward, and our analysis covers the case of the Boko Haram crisis from July 2009 to December 2012. Religious affiliation as a category has been excluded from public censuses in Nigeria since 1962; but since this question is inherently linked to our study, we also tackle the issue of demographics and religion in the Nigerian states where Boko Haram has been active. Our results are used to better assess the reality of the ongoing violence and debunk some preconceived ideas.
3First, we recall the ideological background of Boko Haram and attempt to provide an account of its public stance towards Christians. We then present the Nigeria Watch database, the nature of the data used, and the way it was extracted. Finally, we propose a step-by-step analysis of these data, leading to estimates of the numbers of Muslims and Christians who fell victims to the crisis during the period under consideration.
Ideological background and attitude towards Christians
4Until the beginning of the 2000s, before becoming the spiritual leader of his own spiritual community, Mohammed Yusuf was a follower of Sheikh Ja’far Adam, a renowned and charismatic Islamic scholar in the Izala movement. The two men shared the same Salafi/Wahhabi religious ideal inspired by the Saudi model. The Izala doctrine, which appeared in Nigeria in the late-1970s, advocates for the establishment of an Islamic society that will correspond to the values and practices of the pious ancestors (Salafs) but has never prevented members from being active participants in the state system or attending public schools.1 The movement is characterised by its active proselytism (dawah), and Yusuf, being an excellent preacher, subsequently became a significant Izala figure in Maiduguri (Brigaglia 2012a).
5From 2003 onwards, Yusuf’s ideology gradually moved away from the Izala ideal, radicalised, and developed into a discourse characterised by the systematic rejection of all secular aspects of Nigerian society ([Anon.] 2012).2 He built his argumentation around the concept of boko (Higazi 2013), a Hausa word that brings together the notions of sham, fraud, deceit, and lack of authenticity, but also refers to Western education (Newman 2013). According to Yusuf, the federal state of Nigeria represents a boko model in which Muslims cannot participate. In his preaching, Yusuf used to repeat that “Boko haram da aïki’n gomenati haram”, which means that receiving a secular education, as well as working for the government, was forbidden for Muslims – hence the nickname Boko Haram given to the movement by outsiders.3
6Such divergences led to an open conflict with Sheikh Ja’far and Yusuf’s departure from the Izala movement. Secluded in Maiduguri, he was prevented from preaching in Izala mosques and was directly attacked in sermons given by his former mentor ([Anon.] 2012; Brigaglia 2012a). At about the same time, in 2002–2003, a group of presumed Yusuf students calling themselves Al Sunna Wal Jamma (“Followers of the Prophet’s Teaching) decided to create a distinct community near Kanama, a small town of Yobe State, in order to live outside of any secular framework. Members of this community were the first to engage in violent confrontations with state authorities. The process of Yusuf’s ideological escalation is, in a way, consistent with the Izala philosophy, which was born out of the contestation of the traditional Muslim elite’s authority. By refusing the supremacy of Tijanyyia scholars, the proponents of the Izala doctrine encouraged the acquisition of an individual religious knowledge that promotes a direct relation with God and facilitates a certain disconnection from Islamic clerics. Yusuf’s dissidence also arose from a desire for emancipation from the control of Izala’s scholars as he developed a radical discourse in his preaching for the rejection of the secular state. By stepping outside of the Izala doctrine conveyed by the great figures of the movement, Yusuf created his own school of thought. Likewise, members of the Kanama community wanted to literally break away from the federal secular state and consequently began a new experiment which was consistent with the message of Yusuf (Higazi 2013). The radicalisation process of the Kanama community may have encouraged Yusuf to further strengthen his own rhetoric.4 It is worth noting that Yusuf’s sect began committing violent acts against police stations and other symbols of the Nigerian state authority after the repression of the Al Sunna Wal Jamma in 2003, although such actions long remained sporadic.
7A tipping point was reached in July 2009, after a week-long battle in north-eastern Nigeria between government forces (both police and army) and members of the sect. A series of attacks on police stations by Boko Haram members in Bauchi, Borno, and Yobe states – indicating that the group had prepared for an armed confrontation with the secular state (even in a rather amateurish way in several instances) – was followed by harsh military repression. The uprising resulted in hundreds of casualties and led to the extrajudicial assassination of the sect’s leader.5 The first deliberate attacks against Christians were witnessed at this time (Amaza 2013). Subsequently, after more than a year of restructuring, and seemingly under the subdued leadership of Yusuf’s second-in-command, Abubakar Shekau, followers became true activists and diversified their actions in a more politically focused way, with improved weaponry, logistics, and organisation. Large-scale terrorist attacks began in December 2010, systematically targeting people – such as Christians – and institutions that had not been strategic targets for Boko Haram before then.6
8Unlike most terrorist groups, which formulate a discourse and define political goals before striking, Boko Haram’s ideology seems to have developed gradually alongside its violent actions. The official name of the movement also evolved. Originally known as Ahl al-sunna wa’l jama’a ala minhaj al-salaf (Association of the People of the Sunnah for the Implementation of the Salafs’ Model), it was renamed around 2010 as Jamā‘atu ahl al-Sunna li’l-Da‘wa wa’l-jihād (Association of Sunnah People for Proselytisation and Armed Struggle) (Higazi 2013). The reference to jihad seems to have appeared after the repression of July 2009 and anchored the group in a warmongering logic. Despite the ‘Ahl al-Sunna’ mention, which refers to the Wahhabi school of thought, the new name confirms the split from the Izala movement, which never encouraged the use of lethal violence. If the official name provides interesting information, it is rarely used by commentators. The nickname Boko Haram, catchy and easy to remember, is preferred. Unfortunately, it is also a simplistic ideological shortcut that fits some classical stereotypes about Muslims being fanatics opposed to modernity (Brigaglia 2012a). Yet, since 2009, the group has shown significant abilities to adapt, innovate, and develop sophisticated means of action and propaganda. Over the past few years, Boko Haram has become a serious security threat for the Nigerian state, acquired a transnational influence, and redefined itself in the broader regional framework by networking with Salafi groups in Mali, Niger, and possibly Algeria and Somalia (Higazi 2013).7
9The evolving and opportunistic nature of Boko Haram’s ideology must be taken into account as we are attempting to clarify the rhetoric of the sect towards Christians. When Boko Haram took advantage of inter-community conflicts in Plateau and Kaduna states to spread violence in new territories, they essentially highjacked the long-standing tensions that had built up more around land access than religious issues (ibid.). Also, the rise of a dissident group, Ansaru, which has blamed the leaders of Boko Haram for killing Muslims and has introduced new practices such as the kidnapping of Westerners, signals the existence of internal debates and confirms our impression of the group’s rather fuzzy and constantly evolving ideological substance. At the sect’s beginning, it seems there was no particular animosity vis-à-vis other religions in Yusuf’s doctrine, while Islamic clerics were clearly targeted and many of them assassinated (Brigaglia 2012b).8 This seems to have changed surreptitiously from 2009. Between 26 July and 31 July, approximately twenty churches were destroyed and more than fifty Christians killed in Maiduguri. It was also reported that Boko Haram’s members forced people to convert to Islam under the threat of death (Amaza 2013). For Christians living in Borno State, it was beyond dispute that the uprising was religiously motivated. According to a witness of these events, who made public a detailed daily account from 23 July to 30 July, the objectives of the sect were both “crippling the government and terrorizing Christians in Borno” (ibid.). Nevertheless, he was very surprised when Boko Haram began targeting Christians. On 27 July, he writes:
news began filtering in that some churches had also been destroyed including the Eklessiyar Yan’uwa a Nigeria …, which was the largest indigenous church in Borno State. ‘Impossible!’ was my first reaction. The sect had assured us that the fight was only against the government. (ibid.)
10In fact, for the first time in the history of the radical group, a somewhat clearer anti-Christian rhetoric appeared in the discourse of Yusuf right in the middle of the crisis. Asked by the press about Muslim and non-Muslim peaceful co-existence during the time of the Prophet, he stated that the latter came to a miscreant land and gradually turned it into an Islamic state when, on the contrary, Borno “was an Islamic state before the colonial masters turned it to a kafir land” (Daily Trust 2009).9 So, in Yusuf’s vision, the restoration of an Islamic state in north-east Nigeria required the suppression of others religions, especially Christianity, which is clearly assimilated with colonialism and its avatar, the post-colonial Nigerian state. The troubled days of July 2009 somehow illustrate a vain attempt by the radical group to launch an Islamic reconquista of Maiduguri and Borno, and Christians began at this point to be visibly included among the ‘enemy’, along with the Nigerian security forces and the secular state as a whole. This vision has been put into practice over the past few years despite the fact that Abubakar Shekau seems not to have ever reiterated such an anti-Christian agenda in his public videos.
11We believe that a statistical, comparative study of lethal violence perpetuated against both Christians and Muslims during the conflict can help us move away from the difficult analysis of an elusive ideology to a survey of the actual number of victims it generates in practice.
Note on the methodology: Using Nigeria Watch to identify victims, targets, and religious implications of the Boko Haram crisis
12The Boko Haram crisis fits into the category of conflicts where the collection of accurate data is a fundamental problem, and the general lack of relevant information leaves considerable space for bias and speculation in the analysis. For a number of reasons, it has been impossible to determine the exact total number of victims resulting from the crisis and to precisely assess their characteristics: Muslims, Christians, civilians, officials, soldiers, Boko Haram members, etc. Furthermore, the clear identification of Boko Haram’s targets, a necessary condition to apprehend the group’s agenda, if any, has also been found to be problematic. Among the factors accounting for this situation, we can underline the following.
13First, the blurred nature of Boko Haram itself: since the death of Mohammed Yusuf in July 2009, the group has had neither a uniform leadership (despite the progressive affirmation of Abubakar Shekau) nor an entirely coherent agenda. Most attacks reported by the government forces or the press incriminate “suspected” Boko Haram because no claim is made and the identity of the perpetrators remains uncertain. Sometimes, attacks are clearly claimed by Boko Haram’s ‘original’ leadership, or are consistent with its ideological line, or answer to obvious practical needs (such as assaults on prisons detaining Boko Haram members). In many cases, however, the perpetrators could just as well be common criminals using the name of Boko Haram or could be conveniently designated as such by the police as an easy answer to complex investigations. These ‘usurping’ practices have developed to the point that, in some instances, Boko Haram members have led punitive raids against their ‘fake’ counterparts (Daily Post 2012).
14Second, the ideological and media war: the fight between government forces and Boko Haram can be described in terms of control of the truth and manipulation of public opinion. For example, the Joint Task Force (JTF) – composed of the Nigerian army for the most part, and also the police and the State Security Service (SSS) – seems to be extremely reluctant to release figures for the number of victims among its ranks and, as a result, the military personnel reported killed in action, even after major crises, may appear low compared with the level of violence reported. Furthermore, government forces rarely admit killing civilians, even after clear evidence,10 and tend to report civilians as Boko Haram members. A similar strategy is undertaken by Boko Haram, which often denies figures given by the government and claims more deaths from the JTF than official reports do. Nevertheless, the media power is clearly on the government’s side in this case, and it is significant that Boko Haram started targeting what it considers to be “biased” media, as shown by the attacks on the newspaper ThisDay in Abuja and Kaduna.11
15Third, the technical difficulty of reporting deaths: there is no well-tested mechanism of covering all the deaths related to the violence of the Boko Haram crisis. Government reports, press coverage, and individual testimonies are incomplete, selective, and potentially biased. Especially in the states where the level of violence has been particularly high, such as Borno and Yobe, it can be assumed that the breakdown of deaths entails severe distortions.
16It is thus striking to observe how NGO reports and press and academic articles, which aim to provide accounts of casualties, are characterised by an undeniable vagueness in their sources and methodology and a lack of critical analysis in their breakdowns. In this regard, the manner in which data were collected for this article, using the Nigeria Watch database, has the merit of being transparent and of admitting limitations.
The Nigeria Watch database
17The database12 monitors and compiles violent deaths occurring in Nigeria since 1 June 2006. It relies principally on ten Nigerian daily newspapers (Daily Champion, Guardian, Punch, ThisDay, Vanguard, Independent, Daily Trust, The Nation, PM News and New Nigerian – the last two were replaced by Leadership and Nigerian Tribune in 2013), which are analysed by information retrieval specialists on a daily basis. All events reported in these newspapers and involving at least one violent death are listed and described in the database. Furthermore, other sources of information, such as the police, the judiciary, hospitals, human rights organisations (mainly Human Rights Watch and Amnesty International), as well as private security firms, companies, and embassies, are used to cross-check data and mitigate methodological biases. As stated on Nigeria Watch’s website:
[The database] deals with fatalities resulting directly from intentional or unintentional violence. In a country where civil registration hardly exists, estimates on the number of deaths are not very reliable and can vary a lot from one source to another, especially during violent events. The press in Nigeria is one of the most developed in Africa …. Yet journalists often tend to overestimate casualties and do not always cross-check their information. Moreover, the press is mainly based in the Christian South, especially Lagos, and does not cover properly the Middle Belt and the predominantly Muslim North. As for the Nigeria Police, it underreports crime. It does not publish detailed crime statistics, while people do not trust the police and avoid reporting violence …. Such discrepancies are the reason why this database includes as many sources as possible, in order to cross-check information and compute averages.
18An event is defined as one or more violent deaths happening in one or more contiguous local government areas (LGAs) and ends after at least seven continuous days with no death reported. Events are described briefly and recorded by date (start date / end date), location (state, LGA, city, town, neighbourhood), protagonists, type of conflict, cause of violence, and number of deaths. The database provides scanned copies of each source used, and several documents can be scanned for the same event if they bring additional or contradictory information.
19Obviously, the Nigeria Watch project does not assert it can cover all violent deaths in Nigeria, but it has proved to be a useful and consistent tool to measure trends in violence in a systematic manner. In regard to a study on the casualties of the Boko Haram crisis, the database enables one to work with concrete and verified numbers, and helps show the inherent limitations of this type of research.
The collection of data on victims and targets
20This study looks at all the deadly events related to the Boko Haram crisis, from 24 July 2009 (beginning of the so-called Boko Haram insurgency) to 31 December 2012.13 The term ‘crisis’ is used in order to obtain as comprehensive a spectrum of the victims as possible. We have considered any death related to the phenomenon: Boko Haram activities (suspected and confirmed), government forces’ response and repression measures, and other connected events that do not belong to the first two categories. Included, for instance, is the killing of two suspected Boko Haram members by an angry mob in Maiduguri on 6 October 2011 (ThisDay 2011). Each event was therefore considered relevant or not on a case-by-case basis.
21The first set of data was gathered through a search of the words “boko” and “haram” in the database. Approximately 360 events comprising about 650 scanned documents were screened as a result, and 317 remained after the screening process; indeed, several events had no connection with the crisis even if the description bore the name Boko Haram or the acronym “BH”. In some cases, especially those involving a large number of deaths, further research was conducted beyond the scanned documents (online or in newspaper archives). The events were then classified according to the date, location, total number of deaths, perpetrators, security involvement, number of deaths per category, and targeted location if any. The perpetrators are divided into 4 basic groups: Boko Haram, suspected Boko Haram, government forces (JTF, police, SSS, etc.), and others (e.g. mob). Security involvement, either 0 or 1, indicates whether or not government forces killed at least one person during the event. The categories of victims are comprised of the following: Christians (total), Muslims (total), civilians of unknown faith, Boko Haram members, JTF, police, SSS, other official personnel (prison wardens for the most part), Christian civilians, Christian clerics, Muslim civilians, Muslim clerics, traditional leaders, politicians, media workers, and expatriates. The categories of targets encompass the following: churches, mosques, police stations, army barracks, JTF convoys, JTF checkpoints, beer gardens and other recreational places where alcohol is sold, markets, banks, schools, prisons, personal residences, immigration and customs offices, SSS buildings, and other official buildings. The categories were decided on the sample of the first 30 events and adjusted along the way.
22In regard to the categories of victims, several obviously overlap. When it was possible to determine, we included both status and religion – in the case of a policeman explicitly presented as Christian, for instance. The faith of the victims, if not stated clearly, was derived from obvious Christian and Muslims names or titles reported in the source of information. Furthermore, people killed in a mosque were classified as Muslims while people who died in a church were assumed to be Christians. Boko Haram militants were obviously considered Muslims, although we created a distinct category comprising the total number of Muslims minus the number of Boko Haram militants. We attempted to distinguish as much as possible between civilians and fighters on each side of the conflict. In addition, any time there was a doubt about religious affiliation, victims systematically entered the category “faith unknown” in the final calculation. From our breakdown, we found a total of 2,993 deaths (including 766 Boko Haram militants) from July 2009 to December 2012, which is comparable with, but slightly lower than, numbers generally found in the literature about the crisis.
Methodological challenges
23Our approach faces a number of obvious challenges. As previously stated, the search on the database was conducted through the words “boko” and “haram”. This term, increasingly given by the national and international media and used by the Nigerian federal government, has become the generic appellation of the group, although it has never called itself so.14 A first question is whether the search may have missed some events where Boko Haram militants were involved but not yet designated as such. In most of the 2000s, articles involving suspected followers of Mohammed Yusuf refer to “Nigerian Talibans”. For instance, in the reporting of the confusing events of April 2007 that led to the deaths of Sheikh Ja’afar Mahmud Adam as well as civilians, militants, and policemen, connections are made with the group of Yusuf’s students responsible for several attacks in Yobe in 2003: the protagonists are, most of the time, called “Talibans” in the diverse reports (ThisDay 2007; Vanguard 2007). Since our study begins with July 2009, this issue may have been mitigated. It seems indeed that from the 2009 insurgency onwards, the name Boko Haram has been recurring in the press and only a few articles may have avoided using the name. This leads to another problem: the identification of what is Boko Haram and what is not.
24As previously underlined, many articles and reports talk about “suspected” Boko Haram. As a result, we classified events and dismissed those that were very unlikely to be among Boko Haram activities, although we kept several over which there was a legitimate doubt. The frontal attacks against the police, the JTF, prisons, and other state institutions, when not claimed by Boko Haram, are obviously likely to be part of the group’s activities. Attacks against Christians, as well as targeted assassinations of divergent Muslim clerics, are also very plausible and common in the Boko Haram agenda. Offensives on bars and other places dispensing alcohol also fall into a coherent strategy. Robberies and attacks on banks can be perpetrated by the radical group for understandable economic reasons but are of course also conducted by common criminals, and the distinction between the two can be difficult. Finally, some assassinations imputed to Boko Haram do not appear to belong to any logical strategic framework and are therefore particularly difficult to classify.
25Furthermore, another difficulty arose from the sometimes contradictory breakdown of deaths from one newspaper to another. When sources gave different numbers, we generally used the highest figure, considering that deaths are likely to have been underreported in the states where the crisis has been the most violent. However, we always gave weight to the most specific source – for instance, a source citing a witness who stated having seen a certain number of bodies – over more approximate sources.
26Focusing now on a few case studies will illustrate better our methodological approach.
Case studies
27One instance of crime vaguely imputed to Boko Haram and that we decided not to retain in our breakdown is the slaughter of two teenagers in Damaturu, Yobe State, in July 2012 (Vanguard 2012). The only source reporting the event was the Lagos-based Vanguard newspaper. The article states that “the state Commissioner of Police, Patrick Egbuniwe, who confirmed the incidence, said the command was suspecting members of the Boko Haram Islamic sect” but does not give any information justifying this suspicion, and it later adds: “the Commissioner, who wondered why the attackers will slaughter 15-year-old set of twins said the incident occurred on Saturday”. The article also asserts that a primary school was burnt down in the same city – which fits into Boko Haram activities – but does not relate the two events. The religion of the twins is not mentioned. Given the extreme vagueness of the assumption, reported in a South-based newspaper, about a crime which does not make sense in terms of the Boko Haram agenda – at least from the information provided – the event was not counted as part of the Boko Haram crisis. We do not imply that such crimes cannot be committed by Boko Haram members. Some instances show they can; they are not numerous, however, and at least one source of motive can usually be roughly identified.
28On the other hand, we retained cases where the involvement of Boko Haram is definitely not stated or not even clearly assumed in the source but makes sense in the broader context: for example, the killing of an Islamic cleric in a mosque in Biu, Borno State, in August 2012 (The Nation 2013b). The relevant report explains that, on the same day the JTF was distributing “thousands of audio tapes of Islamic messages criticising the Boko Haram”, one Muslim cleric and one worshipper were shot during prayers at two different mosques in Biu by unidentified gunmen. The article mentions that “the distributed materials were mainly from the preaching of the slain Muslim cleric, Sheik Mahmud Ja’afar”, previously mentioned. It specifies that Boko Haram’s activities have been deadly in Borno State, but it does not clearly state that the gunmen are suspected to be Boko Haram members nor does it cite any officials suggesting this. However, we considered this event as part of the Boko Haram crisis because attacks against Muslim clerics and communities professing a different vision of Islam and condemning Boko Haram have been a recurrent feature in the sectarian group’s strategy. Furthermore, the fact that the JTF was propagandising against Boko Haram through Islamic messages the same day as the murders occurred can hardly be a coincidence.
29The choices that we had to make in reviewing the violent events encompassing the Boko Haram crisis led inevitably to some level of subjectivity in our methodology. Nevertheless, as a meticulous work, our breakdown is available for scholars to assess,15 and every source used in the process can be checked in the Nigeria Watch database and the bibliography.
Generating and interpreting statistical data on body count and religion
30Figures collated from our use of the database show that the faith of victims of the crisis is inconsistently reported by journalists and other sources, probably because they do not have this information or because it does not strike them as worth mentioning. As we processed the press cuttings and complementary sources related to 317 Boko Haram-related events that occurred from July 2009 to December 2012, we attempted to identify the religious affiliation of all the fatalities. The results of this investigation are presented in Table 10.1, which shows the total number of victims reported in our breakdown, including Boko Haram militants (who appear among Muslims).16
31Different facts emerge from this table. Firstly, the faith of a majority of the victims remains unknown. More precisely, we do not have religious indications for about 61 per cent of the victims, representing 1,527 alleged civilians and 309 members of the Nigerian armed forces or other officials. Clearly, in this case, the civilian category may include unidentified members of Boko Haram and government agents, for in several instances available data did not make it possible to differentiate them from civilians. Such is the case with the July 2009 crackdown on the sect in Maiduguri. Although gross estimates of the number of casualties circulate, the conditions in which the massive repression of the sect’s uprising was conducted, the difficulty of differentiating civilians from Boko Haram fighters among the many corpses recovered in neutral clothing, disturbing videos showing members of the federal forces assassinating disabled young men who could not seriously be considered as combatants, and the expedient way victims were buried in mass or individual graves by a variety of actors, prevent the determination of precise figures. We shall see below how we propose to overcome such a difficulty and establish facts about the numbers of Muslim and Christian victims recorded in this conflict.
32Table 10.1. Faith affiliation of deceased victims in the Boko Haram crisis (2009-2012)
Recorded victims according to faith | n= | Sub-categories of victims | n= |
All Muslims | 897 | Muslim civilians | 56 |
|
| Muslim clerics | 21 |
|
| Boko Haram militants | |
|
| Nigerian forces/officials | 766 |
Muslims minus Boko Haram militants | 131 |
|
|
Christians | 260 | Christian civilians | 245 |
|
| Christian clerics | 6 |
|
| Nigerian forces/officials | 9 |
Faith unknown | 1,836 | Civilians | 1,527 |
|
| Nigerian force/officials | 309 |
Total | 2,993 |
|
|
33Secondly, Table 10.1 shows a clear difference between the established numbers of Muslim and Christian civilians killed. At first glance, this could lead us to think that for each Muslim civilian killed, five Christian civilians fell victim to the conflict. This interpretation, however, may not reflect the reality, as two factors need to be taken into account: 1) the relative disparity with which Christian and Muslim deaths are reported by the media; and 2) the relative size of the statistical populations. In fact, Christian victims are probably more likely to be reported in national newspapers because this type of information coincides with the interests – or anxieties – of the bulk of their readers, who are located, in the main, in the southern part of the country, the stronghold of most of the large Christian communities. In a way, we can say that an event involving a Christian civilian who fell victim in northern Nigeria to the Boko Haram crisis is far more likely to find its way into the columns of the Nigerian (and, for that matter, international) printed media than a similar event involving a Muslim civilian. The death of a Christian in northern Nigeria or the Middle Belt resonates with the well-established underlying fear among many Nigerians of seeing their country getting gradually engulfed in another civil war motivated by an explosive combination of ethnic and religious strife. This general statement, however, is not always verified in practice, and there are some cases where violence directed against Christian targets, although reported by the press, did not allow for a comprehensive breakdown of victims, and the number of Christians killed is likely to have been underestimated. This is, for instance, the case with the uprising of July 2009 in Borno as well as the series of bomb attacks claimed by Boko Haram at Jos on Christmas Eve 2010, both events being particularly deadly. Finally, we need to recognise that the statistical samples of Christian and Muslim civilians are small, respectively 13.4 per cent and 3.1 per cent of the entire population of alleged civilians documented in our database (n=1,828).17 Basically, we cannot derive strong conclusions from a sample that does not take into account 83.5 per cent of the civilian victims of the crisis who themselves represent over 51 per cent of the total number of recorded victims.
34On the other hand, the strong predominance of Muslim over Christian non-civilians killed during the conflict, at the rate of six to one, may well be representative of the fact that the Nigerian military and police forces fighting Boko Haram predominantly comprise Muslims, although no statistical data exist to support such a statement – and questions surrounding the reality of the Islamisation of the security forces remain an extremely touchy topic in Nigeria.18 Once again, our statistical sample does not allow us to argue beyond conjecture.
35Similarly, the fact that Muslims clerics seem to have paid a higher price than pastors and other Christian clerics mirrors the reality of the campaign of terror led by Boko Haram against strongly discordant voices among the Islamic scholarly community, through planned assassinations typically conducted by two men on a motorcycle firing an automatic weapon at close range at their targeted victim in the street.
36Table 10.1 suggests that the key to a discussion of the religious factor in the death of thousands as a result of the crisis lies in our capacity to speculate in an informed way on the faith of the 1,836 unknown faith victims. Without resolving the question of the religious affiliation of unknown faith civilian and non-civilian victims, representing respectively approximately 51 per cent and 10.3 per cent of the total number of casualties, one wonders how analysts could discuss the Muslim-versus-Christian nature of the conflict.
37Clearly, there is no satisfactory and indisputable way of attributing a religious affiliation to these 1,836 victims. Whatever strategy is used to do so, we need to acknowledge its experimental nature, expose clearly how figures have been reached, and remain open to criticism. In this paper, we have opted for a method based on available demographic data. Although the latter remain fragile, disputed, and scarce, they seem to be the only quantitative sources that can assist us in building a case. Our core hypothesis is simple: on a state-by-state basis, to correlate our data on civilian19 victims of unknown faith affiliation with existing information about the relative demographic weight of Muslims and Christians across the area affected by the Boko Haram crisis. However, simple concepts are seldom translated into practice with ease, and the following paragraphs will expose choices that were made to circumvent a number of methodological obstacles. In any case, we understand that our assumption that all civilian victims are random collateral damage of actions by Boko Haram and Nigerian security forces – and therefore equivalent to random demographic data – is a mere theoretical position in which we are constrained by the limitations of our data.
38First, what demographic data are we talking about? Nigeria is well-known for generating contested or even ‘made-up’ census data. This is attested by a vast literature that criticises its alteration and falsification for political reasons (Bamgbose 2009). Added to this general concern about the value of the census data, we must also deal with the fact that religion – as a discrete category on questionnaires used by enumerators – was excluded from all recent national surveys for being too sensitive. As a result, it is impossible today to reliably assert the relative percentage of Muslims and Christians who live in the different states of Nigeria. As we go back in time, however, we find that the first and last censuses incorporating ‘religion’ as a category were conducted in 1952 and 1963. Despite many controversies, these two censuses – especially the one conducted in 1952 – seem to be considered much more genuine (ibid.) than all those that followed, including the last one conducted in 2006. The downside of this is that they provide a snapshot that reflects a fifty-year old reality far from that of present-day Nigeria. Furthermore, fifty years ago the administrative divisions of Nigeria were quite different from now, and it takes considerable patience and a pinch of boldness to attempt a projection of the figures recorded then onto the new administrative framework made up of 36 states and the Federal Capital Territory (FCT). Such an approach was adopted by Philip Ostien, a former lecturer in law at the University of Jos in Plateau State and an independent scholar since 2008. His projections were published in different forms in 2007 and 2012 (Ostien 2007; 2012a; 2012b). In Table 10.2, we reproduce the results obtained by Ostien for the 13 states where we recorded at least one victim of the Boko Haram crisis without a recognised religious affiliation.
Table 10.2. Relative proportion of Muslims and Christians recorded during the 1952 and 1963 censuses and projected onto the 2013 administrative map of Nigeria
States | % Muslims | % Christians | % Muslims | % Christians | % Muslims |
| Census 1952 (Ostien 2007) | Census 1952 (Ostien 2012b) | Census 1952 Ostien (2012b) | Census 1952 (Ostien 2012b) | Census 1952 (Ostien 2012b) |
Adamawa | N/A | 3.8 | 32.9 | 16.0 | 34.6 |
Bauchi | 74 | 1.1 | 76.3 | 1.6 | 83.4 |
Borno | 84 | 0.6 | 80.9 | 2.7 | 88.3 |
FCT | N/A | N/A | N/A | N/A | N/A |
Gombe | 74 | 2.5 | 70.1 | 6.2 | 75.0 |
Jigawa | 98 | 0.4 | 98.4 | 0.8 | 98.0 |
Kaduna | 61 | 10.2 | 57.3 | 25.1 | 55.7 |
Kano | 98 | 0.5 | 97.8 | 1.1 | 97.0 |
Katsina | 95 | 0.3 | 95.2 | 0.4 | 94.6 |
Niger | 44 | 3.5 | 46.8 | 4.0 | 62.4 |
Plateau | N/A | 11.4 | 25.5 | 23.2 | 26.1 |
Sokoto | 94 | 0.5 | 96.3 | 0.4 | 98.9 |
Yobe | 84 | 0.5 | 87.3 | 1.0 | 94.8 |
39Figures computed by Ostien differ slightly in his 2007 and 2012 studies, as the author refined his methodology. In addition, his 2007 work focused on Shariah states only, which left out Adamawa and Plateau on our list as well as the FCT. Interestingly, in 2007 he was able to obtain and publish comparative estimates of the relative percentage of Muslims per state from the World Christian Database (WCD), dating to 2002. Aware of potential biases, and “assuming the truth lies somewhere in between (the 1952 census and the WCD data)” (Ostien 2007), he proposed averaging the two percentages to produce a new estimate. Table 10.3 reproduces these data and proposes an update of Ostien’s estimate on the basis of his revised 2012 interpretation of the 1952 census. The result, in column 5, is a theoretical percentage of Muslims per state, which we will consider to be our low-end estimate, as we are averaging the 1952 census – giving a lower percentage of Muslims than the 1963 census – with a Christian-affiliated source which provides the lowest estimates known. Following Ostien, we can agree that “there are many pitfalls here, but for many reasons, it is not easy to do better” (ibid.).
Table 10.3. Low-end estimate of percentage of Muslims in selected states
State | 1 | 2 | 3 | 4 | 5 |
| % Muslims | % Muslims | % Muslims | % Muslims | % Muslims |
Adamawa | N/A | N/A | N/A | 32.9 | N/A |
Bauchi | 74 | 61 | 68 | 76.3 | 68.7 |
Borno | 84 | 49 | 67 | 80.9 | 65.0 |
FCT | N/A | N/A | N/A | N/A | N/A |
Gombe | 74 | 49 | 62 | 70.1 | 59.6 |
Jigawa | 98 | 70 | 84 | 98.4 | 84.2 |
Kaduna | 61 | 51 | 56 | 57.3 | 54.2 |
Kano | 98 | 69 | 84 | 97.8 | 83.4 |
Katsina | 95 | 74 | 85 | 95.2 | 84.6 |
Niger | 44 | 52 | 48 | 46.8 | 49.4 |
Plateau | N/A | N/A | N/A | 25.5 | N/A |
Sokoto | 94 | 74 | 84 | 96.3 | 85.2 |
Yobe | 84 | 49 | 67 | 87.3 | 68.2 |
40To provide low-end estimates for Adamawa, Plateau, and the FCT, we needed to resort to other data, as no WCD estimates were available for them. For FCT, the only recent estimate in the literature is the informed opinion of Medugu (2012), who suggests that there should be roughly as many Muslims as Christians. This is not really satisfactorily, but for lack of a better assessment, we shall adopt it in this study. For Adamawa, we decided to adopt the percentage of Muslims obtained during the 1963 census (34.6 per cent, after Ostien) (Ostien 2012b), which is slightly higher than the one obtained in 1952 (32.9 per cent, after Ostien) (ibid.) and therefore probably more credible. Finally, in the case of Plateau, we retained the percentage proposed by the Ostien in his 2012 study of Plateau State (2012a), which we found to be the lowest figure available (16 per cent). All final estimates are listed in Table 10.4.
Table 10.4. Minimum and maximum estimated % of Muslims in states affected by the Boko Haram crisis
State | Estimate | Estimate % |
Adamawa | 34.6 | 65.0 |
Bauchi | 68.7 | 97.0 |
Borno | 65.0 | 97.0 |
FCT | 50.0 | 50.0 |
Gombe | 59.6 | 73.0 |
Jigawa | 84.2 | 98.0 |
Kaduna | 54.2 | 55.7 |
Kano | 83.4 | 97.0 |
Katsina | 84.6 | 94.6 |
Niger | 49.4 | 91.0 |
Plateau | 16.0 | 26.1 |
Sokoto | 85.2 | 98.9 |
Yobe | 68.2 | 94.8 |
41On the other hand, to obtain high-end estimates, we relied on recent scholarship when available. For instance, several recent studies developed within the framework of the Nigerian Research Network and commissioned by the Oxford Department of International Development suggest a link between the overall percentage of Muslim population and the percentage of elected Muslim politicians (Alkali 2012). Figures provided by these authors were adopted in the case of Adamawa, Bauchi, Borno, and Gombe because no other data was available. For all remaining states, we adopted results obtained from the 1963 census, which were the highest available. Although 100 per cent of elected politicians in Yobe State are Muslims (Alkali 2012), Christians are present in this state, and we therefore preferred to rely on the percentage obtained from the 1963 census (94.8 per cent, after Ostien) (Ostien 2012b). Table 10.4 presents low-end and high-end estimates of the relative percentage of Muslims in all the states selected in this study.
42Based on these figures, we can now extrapolate how many of the 1,527 civilian victims recorded as “faith unknown” may actually have been Muslims. Results are presented in Table 10.5. Overall, we estimate that between 64 and 90 per cent of the 1,527 civilian victims were Muslims, while between 10 and 36 per cent of them professed another religion, most probably Christianity. Clearly, the religious landscape of northern Nigeria is not as dualistic as we tend to believe, and other forms of worship and belief coexist with Islam and Christianity. However, social pressure has taken its toll on such belief systems to the extent that they tend not to be publicly and explicitly acknowledged by actors. For the purpose of this study, we therefore simplified the reality and considered that Islam and Christianity were the two main acknowledged religions in the area under scrutiny.
Table 10.5. Estimation of the percentage of Muslim and Christian believers among the civilian victims labelled as “faith unknown” in the Boko Haram conflict
State | Number of victims recorded as “faith unknown” | Number of assumed Muslims (Min.) | Number of assumed Muslims (Max.) | Number of assumed Muslims (Mean) | Estimate % Muslims (Min.) | Estimate % Muslims (Max.) |
Adamawa | 62 | 21 | 40 | 31 | 34.6 | 65.0 |
Bauchi | 15 | 10 | 15 | 12 | 68.7 | 97.0 |
Borno | 837 | 544 | 812 | 678 | 65.0 | 97.0 |
FCT | 41 | 21 | 21 | 21 | 50.0 | 50.0 |
Gombe | 15 | 9 | 11 | 10 | 59.6 | 73.0 |
Jigawa | 1 | 1 | 1 | 1 | 84.2 | 98.0 |
Kaduna | 53 | 29 | 30 | 29 | 54.2 | 55.7 |
Kano | 183 | 153 | 178 | 165 | 83.4 | 97.0 |
Katsina | 1 | 1 | 1 | 1 | 84.6 | 94.6 |
Niger | 16 | 8 | 15 | 11 | 49.4 | 91.0 |
Plateau | 53 | 8 | 14 | 11 | 16.0 | 26.1 |
Sokoto | 2 | 2 | 2 | 2 | 85.2 | 98.9 |
Yobe | 248 | 169 | 235 | 202 | 68.2 | 94.8 |
Total | 1,527 | 976 | 1,373 | 1,174 |
| |
Total % | 100 | 63.9 | 89.9 | 76.9 |
| |
Estimated total % of Christians | 36.1 (Min.) | 10.1 (Max.) | 23.1 (Mean) |
43For the 309 members of the Nigerian armed forces and officials killed in the same conflict and labelled as “faith unknown”, we decided not to apply the same methodology because soldiers, policemen, SSS operatives, and other categories of personnel can be drawn from many different parts of the country. There are no available data on the ‘state of origin’ of these victims, nor do we have general statistical data on the religious affiliation of the different components of the armed forces, police, etc. On this basis, we decided to consider hypothetically significant the ratio we recorded in Table 10.1 between identified ‘Muslims’ (n=54) and ‘Christians’ (n=9) in the category ‘Nigerian forces/officials’. On this basis, we propose that 85.7 per cent of our group of 309 victims could have been Muslims (n=257), while the remaining victims could have been Christians (n=52). Due to the lack of comparative data, we tentatively adopted this percentage as our high-end hypothesis, and introduced a ratio of circa 50:50 as a low-end hypothesis, giving an estimate of 155 Muslims and 154 Christians.
44Using the methodology presented above, we are finally able to provide estimates of the alleged total number of Christians and Muslims killed during the Boko Haram crisis between 2009 and 2012 (Table 10.6). This table clearly indicates that whatever configuration is used (low-end or high-end estimate), the number of Muslim victims is always higher than the number of Christian victims. Boko Haram members aside, the number of Muslim victims is expected to be two to three times higher than that of Christian victims.
45Returning to Table 10.1, it is striking to realise that figures derived from our database and showing known religious affiliation of victims provide us with a reversed perspective, with 260 Christian against 131 Muslims casualties, i.e. two times more Christians that Muslims. Clearly, as we discussed above, this is an artefact of the much stronger visibility of Christian casualties in the media and in other open sources. Such a visibility sustains the popular idea in some circles that Christians are the primary victims of the Boko Haram insurgency. However, such a view needs to be challenged, as we know that most of the terrorist and counter-terrorist attacks that resulted in massive casualties produced a large number of collateral victims among civilians, whose faith is usually not reported. Our methodology helps us to reframe the debate by including the large majority of the anonymous, collateral victims whose voice is ordinarily suppressed from scholarly and non-scholarly discussions alike. There is no reason to believe that the bulk of these collateral victims are not demographically representative of faith distribution in north-eastern Nigeria, and, if this hypothesis is correct, we should therefore accept the idea that the conflict has resulted in the majority of its victims being among civilian Muslims populations.
Table 10.6. Estimated faith affiliation of deceased victims in the Boko Haram crisis (2009-2012)
Recorded victims according to faith | n= | % | Sub-categories of victims | n= | % |
Muslims | Muslim civilians | ||||
(Minimum)a | 2,028 | 67.8 | (Minimum)b | 1,032 | 56.5 |
(Maximum)c | 2,527 | 84.4 | (Maximum)d | 1,429 | 78.2 |
Muslims minus Boko Haram militants: | Muslim clerics | 21 | |||
Boko Haram militants | 766 | ||||
(Minimum) | 1,262 | 42.2 | Nigerian forces/officials | ||
(Maximum) | 1,761 | 58.8 | (Minimum) | 209 | |
(Maximum) | 311 | ||||
Christians | Christian civilians | ||||
(Minimum)e | 466 | 15.6 | (Minimum)f | 399 | 21.8 |
(Maximum)g | 965 | 32.2 | (Maximum)h | 796 | 43.5 |
Christian clerics | 6 | ||||
Nigerian forces officials | |||||
(Minimum) | 61 | ||||
(Maximum) | 163 | ||||
Total | 2,993 |
Notes:
a) This figure was obtained by adding the total number of known Muslims (n=897, see Table 10.1), the estimated minimum number of Muslims among civilian labelled as “faith unknown” in Table 10.1 (n=976, see Table 10.5), and the estimated minimum number of Muslim victims who were either members of the armed forces or officials (n=155).
b) This represents the minimum % of Muslims among civilian victims only.
c) This figure was obtained by adding the total number of known Muslims (n=897, see Table 10.1), the estimated maximum number of Muslims among civilian labelled as “faith unknown” in Table 10.1 (n=1373, see Table 10.5), and the estimated maximum number of Muslim victims who were either members of the armed forces or officials (n=257).
d) This represents the maximum % of Muslims among civilian victims only.
e) This figure was obtained by adding the total number of known Christians (n=260, see Table 10.1), the estimated minimum number of Christians among civilian labelled as “faith unknown” in Table 10.1 (n=154, see Table 10.5), and the estimated minimum number of Christian victims who were either members of the armed forces or officials (n=52).
f) This represents the minimum % of Christians among civilian victims only.
g) This figure was obtained by adding the total number of known Christians (n=260, see Table 10.1), the estimated maximum number of Christians among civilian labelled as “faith unknown” in Table 10.1 (n=551, see Table 10.5), and the estimated maximum number of Christian victims who were either members of the armed forces or officials (n=154).
h) This represents the maximum % of Christians among civilian victims only.
Conclusion
46Establishing a precise number of victims in this crisis, as well as identifying both their religion and status, has proven a perilous enterprise. Our breakdown – which stops on 31 December 2012 although Boko Haram-related violence has continued since – has inevitable flaws, and it is assumed that even within the time-frame studied some important events may have been missed. Nevertheless, the data presented here, made available to the academic community, is the result of a transparent approach. What is most significant in our contribution is probably not the number of casualties we arrive at, but the percentages resulting from the data analysis. Indeed, beyond the religious divide, one striking conclusion is that 62 per cent of the victims of the Boko Haram crisis are civilians (inclusive of clerics), reinforcing the idea that modern conflicts mainly affect non-combatant populations (Ramsbotham, Woodhouse & Miall 2007). Our data suggest that roughly 25 per cent of the victims are Boko Haram activists and 13 per cent are Nigerian security force members and other state officials. It should not been forgotten that, as discussed above, a number of civilians may have been presented as Boko Haram members in government reports and that the army has been very careful not to disclose too many of its casualties.
47As for the religious affiliation, we estimate that a minimum of 42.2 per cent and a maximum of 58.8 per cent of the victims are Muslims (excluding Boko Haram members). The proportions in regard to the Christian populations are respectively 15.6 per cent and 32.2 per cent (Table 10.6). Therefore, and most likely, Muslims are particularly hit by the ongoing crisis. However, our data also show that attacks against Christians have been continually increasing, both in scale and frequency, from Christmas Day 2010 and the bombings in Jos. Yet, we need to warn against engaging carelessly in a debate over which community suffers the most, as it may serve the strategy of Boko Haram’s leadership aimed at fomenting antagonism between Christians and Muslims. By targeting Christians, the sect tugs at Nigerian identity’s heart-strings and deeply challenges the fragile basis of the nation’s unity. Boko Haram sees the secular state as a colonialist reminiscence and Christians are fully integrated in this colonial picture. Nevertheless, the state and its security prerogatives remain a privileged target, as our data accounts for more than 100 official buildings (police stations, army barracks, prisons, etc.) attacked, with subsequent fatalities, over the period examined. The Nigerian state itself is greatly accountable for the aggravation of the situation. The long-lasting flaws in public administration, the fluctuating strategies, and the blind and deadly repression that reminds us of the darkest pages of the colonial wars have left civilians from both faiths with very little means of protection.
48Our paper aims to trigger debate on the issue of casualties and religion in the Boko Haram crisis. How should data be collected and presented? How should such data be interpreted? To what extent is the religious approach pertinent in identifying the victims? Our initiative is thus far unique in regard to the Boko Haram phenomenon, even if it also demonstrates clearly the limitations and difficulties of such a task. The Nigeria Watch database, moreover, is a pertinent tool in analysing the current crisis.
Vanguard (2012), “15 years old twins slaughtered in Yobe”, 23 July.
Bibliographie
Des DOI sont automatiquement ajoutés aux références bibliographiques par Bilbo, l’outil d’annotation bibliographique d’OpenEdition. Ces références bibliographiques peuvent être téléchargées dans les formats APA, Chicago et MLA.
Format
- APA
- Chicago
- MLA
Alkali, Muhammad Nur et al. (2012), “Overview of Islamic actors in Northeastern Nigeria”, Working Paper No. 2, Nigeria Research Network, University of Oxford.
Amaza, I.P. (2013), “The Boko Haram tragedy”, published on Christian Solidarity Worldwide http://cswng.org/the-boko-haram-tragedy/#comment-922 (Accessed July 2013).
Anonymous (2012), “The popular discourses of Salafi radicalism and Salafi counter-radicalism in Nigeria: A case study of Boko Haram”, Journal of Religion in Africa 42(2): 118-144.
Bamgbose, J.A. (2009), “Falsification of population census data in a heterogeneous Nigerian state: The fourth republic example”, African Journal of Political Science and International Relations 3(8): 311-319.
Brigaglia, A. (2012a), “Jaʿfar Mahmoud Adam, Mohammed Yusuf and Al-Muntada Islamic trust: Reflections on the genesis of the Boko Haram phenomenon in Nigeria”, Annual Review of Islam in Africa 11: 35-44.
Brigaglia, A. (2012b), “A contribution to the history of the Wahhabi Da’wa in West Africa: The career and the murder of Shaykh Ja’far Mahmoud Adam (Daura, ca. 1961/1962-Kano 207)”, Islamic Africa 3(1): 1-23.
Daily Post (2012), Abiodun Badejo, “nine killed, as Boko Haram fights fake members in Damaturu”, 29 August.
Daily Trust (2009), Ahmad Salkida, “Muhammad Yusuf interview”, 27 July.
Daily Trust (2013), Wakili, Isiaka, “insurgents killed more Muslims than Christians – Jonathan”, 21 June, p. 4.
Higazi, A. (2013) “Les origines et la transformation de l’insurrection de Boko Haram dans le nord du Nigeria”, Politique Africaine 130: 137-164.
10.3917/polaf.130.0137 :Medugu, N.I. (2012), “Muslims of the federal capital territory: A survey”, Background Paper No. 5, Nigeria Research Network, University of Oxford.
Newman, P. (2013), “The etymology of Hausa Boko”, mega-Chad research network. http://lah.soas.ac.uk/projects/megachad/publications/Newman-2013-Etymology-of-Hausa-boko.pdf
Ostien, P. (2007), Sharia implementation in Northern Nigeria 1999-2006: A sourcebook, vol. I-V. Ibadan: Spectrum Books.
Ostien, P. (2012a), “A survey of the Muslims of Nigeria’s North Central geo-political zone”, Working Paper No.1, Nigeria Research Network, University of Oxford.
Ostien, P. (2012b) “Percentage by religion of the 1952 and 1963 populations of Nigeria’s present 36 States”, Background Paper No.1, Nigeria Research Network, Oxford University.
Ramsbotham, O., T. Woodhouse & H. Miall (2007), Contemporary conflict Resolution (2nd ed.). Cambridge: Polity Press.
The Nation (2013a), Akowe, Tony, “Jonathan distorting facts on Boko Haram victims- Northern CAN”, 23 June. http://thenationonlineng.net/new/jonathan-distorting-facts-on-boko-haram-victims-northern-can/ (Accessed September 2013).
The Nation (2013b), Joseph Abiodun, “Gunmen kill two in Borno mosques as JTF distributes anti-Boko Haram tapes”, 22 August.
ThisDay (2007), Ibrahim Shuaibu, “Talibans Kill 13 Policemen in Kano”, 18 April.
ThisDay (2011), Michael Olugbode, “Mob kills 2 suspected Boko Haram members”, 8 October.
Vanguard (2007), Tina Anthony, “Soldiers kill 25 fundamentalists in Kano”, 19 April.
Notes de bas de page
1 In their sermons, the Izala mallams such as Sheikh Mahmoud Ja’far, Sheikh Abubakar Giro, Sheikh Kabiru Gombe, and Sheikh Albani Zaria attach considerable importance to education. The deep social and religious reform they advocate for requires the involvement of their followers at every level of the state. These speeches, made in Hausa and recorded on DVDs manufactured by small local production companies (such as An-Noor Islamic Production, Sawtul-Hikmah, Sautu Rijaalis Sunnah, Darul Islam Production), are on sale in street shops and markets all over northern Nigeria.
2 In support of his argument, Yusuf referred to Abu Zayd, a Saudi Islamic scholar who claimed that modern secular education was “totally lacking any Islamic legitimacy”.
3 From 2005 onwards, the popularity of Mohammed Yusuf began to grow in the Republic of Niger, where his sermons recorded on DVDs were widely disseminated. The Nigerien version of his name was Mahamadou Issoufou, and he was known for the numerous references to the concept of boko in his preaching; thus, people began to call him Mahamadou Issoufou ‘Boko Haram’ (interviews with DVDs sellers in Niamey, Maradi and Zinder, 2012–2013).
4 Although most of them are thought to be former students of Mohammed Yusuf, the nature of the relation between members of the Kanama community and the Boko Haram leader have not been fully clarified. It seems, however, that most of the survivors integrated into Yusuf’s group after the government repression in 2004.
5 Borno in general and Maiduguri in particular were by far the two places most affected in terms of casualties. Our breakdown accounts for about 900 victims. However, other estimates propose more than a thousand people killed. See Brigaglia (2012a: 35) and Higazi (2013: 4). The latter cites the Report of the Administrative Committee of Inquiry into the Boko Haram Insurgency in Borno State. Main Report (5 volumes), October 2009. This report has never been made public and records 1,387 victims.
6 Including the bomb attack on the UN building in August 2011 and the series of attacks on Christian targets at Christmas 2010 and 2011, the coordinated attacks in Kano in January 2012, the bombing of ThisDay newspaper headquarters in April 2012, and the destruction of several telecommunication facilities.
7 Boko Haram’s connections with Niger Republic are particularly strong, since victims of the repression of the Nigerian government in northern Nigeria have been crossing the border. These connections are also grounded in very dynamic trans-border trade and kinship-based relationships among groups on both sides of the border with shared ethnicity, language, and long-term history. In Niger, groups based on a similar ideology are known to have been formed, such as Sake gueme Haram (Hausa: “shaving the beard is forbidden”), which appeared in 2007 in the region of Diffa.
8 The most symbolic assassination was the murder of Sheikh Ja’far, perpetrated in April 2007 at the Juma’at Mosque of Kano, when he was leading the morning prayer.
9 Kafir is an Arabic term which means “miscreant” or “infidel”.
10 See for instance the controversy involving government forces during the 2009 uprising, when the Nigerian police were filmed perpetrating extrajudicial killings on civilians: “Nigeria killings caught on video”, Al Jazeera, 10 February 2010.
http://www.aljazeera.com/news/africa/2010/02/20102102505798741.html (Accessed July 2013).
11 “Nigeria’s ThisDay newspaper hit by Abuja and Kaduna blasts”, BBC World News, 26 April 2012, http://www.bbc.co.uk/news/world-africa-17856362 (Accessed August 2013).
12 The project was created in 2006 by Dr. Marc-Antoine Pérouse de Montclos, researcher at the Institut de Recherche pour le Développement (France), with the support of private firms. The database and more information are accessible online at http://www.nigeriawatch.org/. Since 2013, the analysis of newspapers has been conducted from the French Institute for Research in Africa (IFRA-Nigeria), on the University of Ibadan’s campus, with funding from DFID through the Nigerian Stability and Reconciliation Programme (NSRP) and the British Council.
13 We began our analysis from this date because, aside from the death of Sheikh Ja’afar in 2007 and the subsequent deadly events (whose perpetrators remain unconfirmed), no data was available prior to July 2009.
14 As previously stated, the proper name is Jamā'a Ahl al-sunnah li-da'wa wa al-jihād. See also the interesting discussion on the meaning of boko, in Newman (2013).
15 The final spreadsheet can be accessed at http://ifra-nigeria.org/IMG/xls/table-boko-haram-conflict.xls
16 As explained in this chapter, there are legitimate doubts over certain official reports, which may have amalgamated civilians and Boko Haram members.
17 This figure was obtained by adding Muslim civilians, Christian civilians and civilians of unknown faith.
18 A recent petition circulated by a Nigerian NGO was widely debated over the Internet. See for instance “Group drags IGP Abubakar to President Jonathan over ‘Hausanisation & Islamisation’ of Nigeria Police”, News Express, 6 May 2013. http://www.newsexpressngr.com (Accessed October 2013).
19 Members of the Nigerian government forces and officials are treated separately, using a different method. See later discussion.

Le texte seul est utilisable sous licence Creative Commons - Attribution - Partage dans les Mêmes Conditions 4.0 International - CC BY-SA 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.
Anti-corruption campaign in Nigeria (1999-2007)
The politics of a failed reform
David U. Enweremadu
2012
Boko Haram: Islamism, politics, security and the state in Nigeria
Marc-Antoine Pérouse de Montclos (dir.)
2014
Violence in Nigeria
A qualitative and quantitative analysis
Marc-Antoine Pérouse de Montclos (dir.)
2016
Transnational Islam
Circulation of religious Ideas, Actors and Practices between Niger and Nigeria
Élodie Apard (dir.)
2020