Abstract
Azar’s framework for analysing protracted social conflict is used to examine the roles played by religious and other actors in episodes of inter-religious violence and their aftermath in two cities in India (Ahmedabad and Mumbai) and two in Nigeria (Jos and Kano). Although the violence was apparently related to differences in beliefs and/or religious expansionism, explanations were more complicated. Religious organizations provided relief to victims from their own tradition, but played only limited roles in the longer term aftermath. Several features of the violent episodes and their aftermath help to explain why the inter-religious conflicts studied have become protracted.
Introduction
Religion (in the context of this paper, Christianity, Islam and Hinduism) is said to be associated with and even a motivation for violence. It appears that inter-religious conflict and violence has increased in recent decades, much of it occurring in urban areas. Although religious organizations may also provide help to the victims of conflict and seek to improve understanding between members of different faith traditions, conflict and violence that appears to have a religious basis tends to be more headline grabbing than the other roles religious groups play.
Conflict is a universal feature of human societies, but its escalation into violence is not, and indeed it can play a creative role. The question is why some differences are not addressed in peaceful ways, resulting instead in violent clashes between groups, including those with different religious affiliations, which often take on the features of protracted social conflict – episodic violence interspersed with periods of relative calm, which may be disrupted by further outbreaks of violence rather than evolving into lasting peace. This paper is concerned with conflict that is intra- rather than inter-state; apparently over religious beliefs or identity rather than territory, resources or political ideology; and marked by periodic clashes rather than one-off, continuous or chronic violence. There is increasing concern with pervasive everyday (and multiple forms of) violence in many cities worldwide, often linked to criminality (e.g. Moser, 2004; Moser and Horn, 2011; Moser and McIlwaine, 2006; Pieterse, 2010; Rodgers, 2010), and this is certainly a major issue in both Nigeria and India, but it is not the main focus of this paper. Rather, the research aimed to untangle the role played by religion and religious organizations in the inter-religious clashes that have occurred in the four case study cities in India and Nigeria in recent years and their outcomes, both immediate and longer term. The cities chosen for study were those in which inter-religious clashes had occurred rather than contested or polarized cities per se (see, for example, Bollens, 1998, 1999, 2007; Varshney, 2002). The research did not undertake an explicitly spatial analysis (Chatterjee, 2009), examine attempts by urban planners and policy makers to address urban conflict and violence (unlike Bollens), or seek to understand why violence conflict occurs in some multi-religious cities while others remain peaceful (unlike Varshney). Instead it sought to disentangle the roles played by religion in the clashes and their aftermath.
The research on which the paper is based was carried out as part of a wider programme examining the links between religion and development. Because local context is important in understanding inter-religious conflict and violence, research into major clashes that had occurred in the 15 years prior to the study was undertaken in 2008–9 in two cities in each country – Mumbai and Ahmedabad in India and Jos and Kano in Nigeria. 1 Existing research was almost entirely confined to what had happened during successive violent episodes in these cities and elsewhere – their characteristics, triggers, explanations and immediate effects, so for our understanding of the episodes themselves, we were able to rely largely on reports of government, commissions of enquiry, the accounts of other observers (for example, Human Rights Watch) and academic analyses, although these were validated during the study. Reviews of the available material showed, however, that the aftermath of the violence was under-researched, so the empirical work undertaken by the local teams in 2008–9 concentrated on this, largely through extensive semi-structured interviews with key informants, both individuals and groups. The informants included government officials, staff and volunteers associated with non-governmental organizations (NGOs) and religious organizations that had been involved, observers and individuals who had been affected. Care was taken to assemble local research teams that were balanced in religious and gender terms, had local knowledge on which they could draw, and were able and willing to take a neutral, non-judgemental stance.
As might be expected, the research was difficult. First, both religion and violent conflict are sensitive and emotive areas, not just for the victims of violence and those with strong religious affiliations, but also for observers. What happens and why are always highly contested – multiple interests are at play, not always overtly, the events themselves are chaotic and confusing, and the explanations offered depend in part on informants’ own position in relation to the other actors. Second, asking questions about violence can be dangerous for the researchers concerned, especially if the clashes are recent and a climate of fear prevails. This was particularly the case in Jos, where at the time of the research the most recent violence had occurred in 2008 and it flared up again while the findings were being written up in 2010. In India, the findings were discussed and their implications for policy and action explored in a workshop involving many of the relevant actors, but plans to do the same in Nigeria in 2010 had to be abandoned, although feedback was obtained from the audiences of three other presentations.
Framing the Research
Intra-state conflict includes civil wars, insurgencies, coups and revolutions, as well as the episodic violence that occurs at the local level, often in the context of longer drawn out conflicts. Clashes occur between two (or more) distinct groups within a state, but may also involve the state, directly or indirectly. The groups are generally defined (and mobilized) around distinct communal identities, based on real or perceived religious, ethnic, cultural (and sometimes class) differences. This raises questions about the nature of identity, in particular what is meant by ‘religious’ or ‘ethnic’ identity. Often, analysts use these terms as if they are understood, without providing definitions – what Green (2004) terms the ‘ordinary language approach’. However, not only does this conceal more than it reveals, it also undermines any attempt to make sound comparisons.
Earlier categorical approaches to identity have given way to an understanding that individuals and groups construct and draw meaning from multiple identities, which are influenced by cultural and historical factors, vary over time and both reflect and constitute reality (Brubaker and Cooper, 2000; Puttergill and Leildé, 2006). People do identify themselves with certain categories (such as men or women; members of families and ethnic or caste groups; children, youth or adults; or adherents of specific religious traditions). However, the meanings they give to these categories, the ways in which their perceptions of their own identity influence their relations with others and how their sense of identity changes over time vary, implying that identity is socially constructed. An interpretivist methodological approach is therefore needed to develop an accurate understanding of the identities in play in particular local contexts.
Religion has substantive attributes, particularly belief in a transcendental reality or spiritual being(s) (the sacred), religiosity (signified by adherence to certain beliefs and practices) and informal or formal affiliation with one or more religious congregations (mosque, church, temple) or organizations. It also has social functions: it plays a role in the construction of people’s world views, in social relationships and in wider socio-political organization. In most instances, it can be characterized as a set of core myths, teachings and beliefs that, among other things, specify (or suggest) how to live in accordance with the teachings of the faith tradition and how society should be ordered (Rakodi, 2011). Religious teachings are nevertheless constantly reinterpreted and individually or jointly acted upon to a greater or lesser extent in the context in which adherents live. Especially in multi-religious contexts, there are real differences between and within religious traditions, competition between which may result in a hardening of religious identities, but there is also considerable syncretism of belief and practice, which may lead to an individual or group’s religious identity softening over time.
Religious and ethnic identities are often conflated, with religion, culture and language seen as different aspects of the umbrella term ‘ethnicity’. However, although religious and ethnic identities can overlap, it is more analytically useful to distinguish them because the links between ethnicity, religion and violence vary from context to context (Appleby, 2000). In the ‘ordinary language’ approach, ethnicity tends to be used to denote either minorities or all identity groups (for example, Fox, 1999, 2004). Such imprecision has led some to attempt a universal definition. These include, at the extremes, primordialists, who regard ethnic groups as ‘immemorial and thus given facets of social life,’ and modernists, who argue that ‘ethnic groups are a product of modernity and, more specifically, the creation of modern elites for their own purposes’ (Green, 2004: 7). Commonly, in such universal definitions, the term refers to ‘collectivities that share a myth of origin [and] … a ‘culture’, the most notable aspect of which is language’ (Bates, 2006: 167; see also Horowitz, 1985; Young, 2007). However, the extent to which ethnicity is conceived of ascriptively, the dimensions included, the balance between objective attributes and subjective feelings, and the extent to which ethnic categories are believed to be fixed or malleable and boundaries between them clear or porous varies between analysts and contexts. In India, for example, inter-religious conflict is often referred to as ‘ethnic’, but this does not imply that the opposing groups have a common area of origin or share a single language (Gupta, 2011; see also Bollens, 1999: 17). In contrast, in Africa ethnicity is regarded as being characterized by common ancestry, territorial claims, culture and language. Today, the primordialist view that takes ethnicity as a fixed characteristic of individuals and communities has largely been discredited in favour of constructivist approaches that emphasize its relational nature, implying that it may or may not be called upon or activated in specific circumstances or by particular actors (Agbu, 2011).
Both religion and ethnicity are related to national identity and nationalism in complex ways, with constitutional arrangements, access to power, experience of discrimination and struggles for recognition and equality influencing both people’s sense of identity and the incentives for political actors to instrumentalize it, in turn fuelling discourses of inclusion and exclusion (Dorman et al., 2007).
There are a wide variety of explanations for conflict and violence, which can be broadly grouped into four: grievance, greed, state incapacity, and failure to satisfy human needs, each of which will be briefly reviewed later. The grievances arising from economic inequality have long been popular explanations of political instability and, in more extreme cases, violent conflict, although recent research suggests that horizontal economic and political inequalities between socio-cultural groups provide a better explanation of contemporary civil wars and communal conflicts than vertical class-based inequalities (Langer and Brown, 2007; Ostby, 2004; Stewart, 2004). Cultural status is often associated with economic and political exclusion and inequality, with cultural status inequalities linked to recognition of and respect for ethnocultural practices, language and religion. Religious restrictions and perceived religious insults are particularly potent for group mobilization because of their symbolic value (Langer and Brown, 2007). Critics of grievance theories attribute the escalation of conflict into civil war to perceived tradeoffs between opportunities for predation (through control over the exports of primary commodities), the availability of finance, the cost of rebellion, military advantage, risk and occasionally ethnic tension – the greed hypothesis (Collier and Hoeffler, 2000).
Many analyses stress the importance of the nature and role of the state, including its role in generating and manipulating conflict, the extent to which it is accountable, and its capacity to manage political and social difference and competition. Human needs theorists have contributed additional insights into the causes and dynamics of communal violence, suggesting that needs such as safety/security, belongingness and recognition, freedom, participation and justice are essential to human wellbeing. Because such needs are so strongly felt, it is suggested, denial of them can cause conflict to escalate into violence (Azar, 1990; Burton, 1990; Goodhand and Hulme, 1999; Miall, 2004). Moreover, because such needs are considered to be ‘non-negotiable’, human needs-based conflicts are perceived to be particularly challenging to resolve and more likely to become ‘deep-rooted’, ‘intractable’ or ‘protracted’ (Appleby, 2000; Azar, 1990). Increasingly, analysts call upon a combination of all four approaches to understand the complexity of conflict and violence (see, for example, World Bank, 2011).
There are also contending explanations for inter-religious communal violence more specifically. Hasenclever and Rittberger (2000) distinguish between primordialist, instrumentalist and constructivist views. Primordialists argue that differences in religious cosmologies are the basis for tension and conflict. Instrumentalists suggest that religious and cultural markers are used by political entrepreneurs to mobilize their followers in their own or wider political interests (e.g. Brass, 2003; Hansen, 2001). Available analyses emphasize the causes of mobilization rather than when and how it occurs (but see Brass, 2003, for an exception). Constructivists believe that social conflicts are embedded in cognitive structures such as religion, ethnicity or nationalism, which give meaning to power and interests. Hasenclever and Rittberger themselves favour a constructivist approach based on an assumption that leaders and followers each weigh up the potential and risks of using the mobilizing force of religion to provoke violence (see also Appleby, 2000; Fox, 1999). The likelihood that either will succeed in achieving their aims, they suggest, is related to whether identity is value- and not merely interest-based, the willingness of followers to make sacrifices and the nature of the relationship between the parties (see also Varshney, 2002).
Growing concern with protracted intra-state conflict has focused attention on the periods between violent episodes, when relative calm may be translated into lasting peace or, alternatively, failure to address the roots of conflict and the effects of violence sow the seeds for further outbreaks. Explicit consideration needs to be given, Azar (1990) suggests, to the covert, latent or non-violent conflicts that persist during these periods. A broad distinction can be drawn between the immediate aftermath of a clash or riot, when the priority is to restore order and provide relief; the subsequent period, during which those affected attempt to resume everyday life and physical reconstruction occurs; and the longer term, during which social, economic and political relationships are renegotiated and the underlying causes and immediate grievances may or may not be addressed. These are not, of course, distinct chronological periods – they are likely to overlap, the likelihood and speed of progressing from one to another varies, and the discourses adopted and actions taken (or not taken) at one stage influence the others. Moreover, memories and interpretations of the time before a violent episode, the episode itself and the aftermath change as time elapses, affecting both people’s accounts and the ability of researchers to assemble detailed and reliable accounts. Existing studies vary in the extent to which they differentiate between and probe the experiences, perspectives and actions of different actors, including victims and perpetrators, survivors, state agencies, civil society organizations and religious actors (see, for example, Borer, 2003; Brass, 2003; Das, 1990; Jasani, 2008; Mehta and Chatterji, 2001).
Analysts’ perspectives on the roles in peace-building and conflict transformation of revenge, reconciliation, forgiveness and justice differ (e.g. Auerbach, 2005; Lambourne, 2004; Lederach, 2001). Although interest in the role of religious actors in peace-building is growing, the literature is limited, especially in terms of the everyday roles that religious actors and organizations may play. Ideas from the broader literature on social capital, including the negative implications of segregation and the positive role of social ties between adherents of different communal identity groups, are only just beginning to be applied to analysis of the aftermath of violence, including inter-religious violence (e.g. Colletta and Cullen, 2000; Goodhand et al., 2000). While some accounts put forward a positive view of the contribution of religious actors in the aftermath of violence, others note that they can be reactive, inexperienced, unwilling to cooperate with each other, self-serving or manipulative (e.g. Alger, 2002; Jasani, 2008).
People from different religious, ethnic and cultural backgrounds live together in close proximity in urban areas. While in many cities for much of the time, mutual tolerance and adaptation enable differences to be accommodated, in some instances urban areas may become a stage for the overt expression of intergroup antagonisms. Sometimes this is attributed to the mere proximity of people from different backgrounds, but Bollens suggests that other preconditions for violent conflict must also be present, in particular weak and illegitimate governance incapable of managing communal differences and allocating resources equitably. Typically, responsibility for urban governance, including the maintenance of security, is split between multiple levels of government, agencies and jurisdictions which, especially at the local level, lack capacity and legitimacy. In addition, if opposing sides see the other as a threat to their physical, cultural and/or social survival and the politically dominant group advances its own interests at the cost of others, disadvantaged groups may come to see violence as the only way of protecting themselves and changing discriminatory structures and processes. The situation is exacerbated by the process of segregation that often occurs in societies experiencing identity-based conflict (Bollens, 1999).
The term ‘protracted social conflict’ was used by Edward Azar to describe the intra-state conflicts on which he worked during the better part of two decades, from 1970 to his death in 1991. His understanding of such conflicts and approach to peace-building drew on analyses of over 60 intra-state conflicts between the 1960s and the 1980s, in countries as diverse as Lebanon, Sri Lanka and Northern Ireland, as well as reflecting (and indeed prefiguring) many of the evolving international ideas about conflict and violence discussed earlier. At the time, his work did not make much impact on mainstream conflict analysis, for a variety of reasons (Ramsbotham, 2005). However, the analytical framework in which his work culminated, while quite broad and general, is valuable for the comparative analysis of protracted identity-based conflicts and will be adopted in this paper. 2
Azar was concerned with understanding not just violent episodes themselves, but also the periods between the episodes (referred to here as the aftermath) in order to understand why in some instances lasting peace is established, whereas in others conflict becomes protracted and violence recurs. As noted earlier, he defined protracted social conflict as long drawn-out intra-state conflict revolving around religious, cultural or ethnic communal identity, characterized by intermittent episodes of violence, but with no clear beginning or end. He identified three aspects of episodic violence that need to be understood: its genesis (or preconditions), the dynamics of specific episodes and the outcomes of outbreaks of violence (the aftermath). The analytical framework suggests a number of hypotheses about the explanations, characteristics and implications of each. These hypotheses are used in this paper to consider whether the inter-religious conflicts in the case study cities were in fact about doctrinal differences and religious expansion or whether alternative explanations are needed to understand their causes, characteristics and outcomes. These are elaborated at the beginning of the relevant sections of the analysis but, first, a brief introduction to each of the violent inter-religious clashes that were studied is provided.
Inter-Religious Violence in Urban India and Nigeria
In all four of the cities studied there had been a history of intermittent violence, not all of which was inter-religious. However, this research focused on significant episodes of inter-religious violence during the 15 years prior to the study. It concentrated on Ahmedabad in India and Jos in Nigeria, considering earlier violence in Mumbai and Kano for comparative purposes. Brief descriptions of the main episodes are given in the following sections.
Hindu–Muslim Violence in Mumbai and Ahmedabad
The research examined Hindu–Muslim violence in Mumbai in 1992–3 and Ahmedabad in 2002 (Gupta, 2011). The Mumbai riots began in December 1992 on the day that an ancient mosque (the Babri Masjid), which had allegedly been built by the Mughal Emperor Babar on the site of the birthplace of the Hindu god, Lord Rama, in Ayodhya, Uttar Pradesh, was destroyed by Hindu activists. Muslims throughout India protested, most violently in Mumbai, where they attacked Hindu homes, temples and places of business. Within a month, Hindu activists retaliated, targeting Muslims in areas where Hindus were in a majority. The Srikrishna Enquiry Committee estimated that about 900 people (including 575 Muslims) died and over 75,000 people were affected by the violence.
The aftermath of this violence rumbles on. In 2002, at Godhra, Muslims allegedly attacked a train carrying Hindu activists returning from working on the construction of a Hindu temple on the site of the destroyed mosque, killing about 60. Hindus retaliated in Gujarat, especially in the Ahmedabad-Baroda-Godhra conurbation, where over three months of violence ensued. 3 Muslim homes and establishments were targeted, with the result that, in the state as a whole, 174,000 people sought refuge in camps and an estimated 250,000 people were displaced (Chandhoke, 2009; Gupta, 2011). In Ahmedabad, the Shah Alam relief camp alone gave refuge to about 2,200 families (about 10,000 people).
Christian–Muslim Violence in Kano and Jos
In 1999 violence broke out in the northern Nigerian city of Kano between Hausa indigenes and Yoruba in-migrants in apparent revenge for an earlier attack on Hausa in-migrants in the south-western city of Shagamu, which had been triggered by disrespect supposedly shown by Hausas for Yoruba traditional practices. The attacks spread from those by Muslim Hausa against Yoruba (who are roughly half Christian and half Muslim) to violence against non-Yoruba Christian groups. Further incidents of Christian–Muslim violence occurred in 2001 and 2004, when between 200 and 250 people died (Human Rights Watch, 2005).
Jos, a religiously plural city in Nigeria’s Middle Belt, did not experience large-scale violence until 2001, when conflicts between ethnic groups took on a religious dimension and widespread Christian–Muslim clashes resulted (Best, 2007). According to Human Rights Watch (2001), over 1,000 people were killed during six days of violence and 50,000 displaced people were living in eight relief camps. In 2008 Muslim protests over the alleged rigging of local government elections in Jos North led to violence against Christian homes, churches and businesses. There were further violent episodes in January and April 2010 (Best and Rakodi, 2011). 4
Understanding Protracted Social Conflict
The Preconditions for Violence
The first element of Azar’s analytical framework identifies four linked clusters of preconditions that determine the likelihood of conflict occurring and its characteristics:
communal content
the extent to which human needs are satisfied
the role of the state
international linkages
The first refers to the existence of cultural identity groups which are politicized as a result of a country’s colonial experience and historical rivalry between communal grups. Azar suggests that people seek to satisfy their needs for security (mainly basic material needs), access to socio-political institutions and, to enable them to participate in such institutions effectively, recognition and acceptance. Deprivation in respect of one of these needs hinders people’s ability to satisfy others and is related to the role played by the state in regulating social, political and economic interactions. The dominance of a state by one or more communal groups, its legitimacy and its capacity for effective governance are key to the satisfaction of needs. They are influenced by a state’s economic and political/military international links, which may induce it to pursue not only foreign but also domestic policies ‘disjoined from, or contradictory to, the needs of its own public’ (Azar, 1990: 11). Analysis of the case study cities confirms that these factors have considerable explanatory power, although there is insufficient space here to examine them in the detail they deserve.
In India, Hinduism is the majority religion (80%) and Islam the largest minority religion (13%). Although partition occurred in the 1940s, its effects are still very much alive – they include the memories of those affected by the large-scale killings and displacements that accompanied it, the perceived ongoing threat from Pakistan and the rise of Hindu nationalism, expressed and mobilized by organizations such as the Rashtriya Swayam Sevak Sangh (RSS) and Shiv Sena. Such organizations receive funds from members of the Hindu diaspora and use the tensions with Pakistan to portray Muslims in India as potential traitors (Chandhoke, 2009; Gupta, 2002, 2011; Mander, 2004). Although there are great economic inequalities in India, these do not seem to have been the main motive for the violence. In Mumbai, although some prosperous Muslim businesses were targeted by Hindu rioters, Muslims are generally poor, as they almost all are in Ahmedabad, so economic envy was not a motivating factor for Hindu attacks. Rather, attacks by activists from the Hindu majority against the Muslim minority in both cities were not only prompted but also sustained by the Hindu nationalist agenda, as demonstrated by the differences between Mumbai and Ahmedabad.
In Mumbai, the Maharashtra state government and security services did their best to end the violence, 5 supported by concerned politicians, officials and secular citizens. For example, committees were established in sensitive neighbourhoods (mohallas) to improve rapport between the police, residents (especially Muslims) and government agencies. In contrast, in Ahmedabad, the Gujarat state government was in the hands of the Hindu nationalist Bharatiya Janata Party, which openly backed anti-Muslim sentiments, failed to restore order and let the attacks by Hindus on Muslims carry on unchecked, with the result that they lasted much longer than in Mumbai and had a more devastating impact (Gupta, 2011). 6
In Nigeria, roughly half the population claim to be Christians and half Muslims. 7 Islam arrived during the 11th and 12th centuries from the north, becoming well established by the 15th century. It was initially associated most closely with one of the three dominant ethnic groups, the Hausa-Fulani, who established Islamic states and actively sought to expand Islam, often by violent conquest and slavery, most recently in the early 19th century (Harnischfeger, 2004; ICG, 2010). Since the 20th century, rivalry for political domination between the three main ethnic groups, the Hausa-Fulani, Igbo and Yoruba, has had a religious dimension, with the Hausa-Fulani being almost entirely Muslim, the Igbo Christian and the Yoruba mixed. 8 Christianity first arrived in Nigeria in the 15th century but did not take root until the 18th and 19th centuries, when it spread from the south with colonial trade and administration. Many of the missionaries concentrated their activities in the south of the country but a few targeted the north, despite the restrictions placed on their activities by the colonial government as part of its agreements with Muslim rulers (Kastfelt, 1994). Minority groups that had resisted Hausa-Fulani expansionism and Islam (especially in the country’s Middle Belt) were predisposed to convert to Christianity rather than Islam.
Hausa-Fulani Muslim domination of both the northern region (later subdivided into states) and the federal government from independence in 1966 through the period of military rule was resented by other groups. The poverty of the northern states exacerbated competition over federal resources, largely from revenues from oil extraction in the south of the country. Educational disadvantage and economic difficulties, coupled with ineffective economic development policies, have given rise to significant groups of disadvantaged young people, especially young men, many of whom have migrated to urban centres throughout Nigeria in search of economic opportunities (ICG, 2010). In northern Nigeria they include almajirai, followers of itinerant Islamic scholars, who in the past could maintain themselves by an established system of begging and alms giving but can no longer do so (Watts, 1999). Such young men can be easily recruited by rival politicians to bully and intimidate both their opponents and voters at large (Human Rights Watch, 2007). The restoration of democracy in 1999 paved the way for 12 northern states to reintroduce Shari’a criminal law and attempt to establish Islamic states. Protests by Christians sometimes turned violent, leading many to flee, for example to Jos.
In Kano, the 1999 violence took on a religious dimension when the initial violence against all Yoruba (Christian and Muslim) spread to non-Yoruba Christians. In 2001 clashes between Igbo in-migrants and mostly Hausa-speaking youths, triggered by a confrontation over the failure of some young Igbo women to cover their heads, erupted into attacks on Christians later in 2001 and in 2004 (Human Rights Watch, 2005). However, the underlying conflicts are predominantly linked to struggles over the control of resources, which have another significant dimension, that of citizenship and indigeneity.
At national and local levels, citizens are defined as those born in or with ancestral ties to a particular ancestral community/local area within Nigeria. Those indigenous to an area claim that their rights to land and government resources should take precedence over those of in-migrants, even of several generations standing, claims that are often contested, sometimes violently (Harnischfeger, 2004; Osaghae and Suberu, 2005). Where indigenous and in-migrant groups have different religious affiliations, the conflict may take on a religious dimension. For example, in the colonial tin mining city of Jos, local government elections are contested between indigenous Afizere, Anaguta and Berom groups and settler Hausa-Fulani. The latter claim that they founded the city at the beginning of the 20th century and so are indigenes, while the former suspect them of trying to rig the elections to gain control of Jos North local government which can distribute resources and allocate land (Danfulani and Fwatshak, 2002; Harnischfeger, 2004; Human Rights Watch, 2001). As the indigenous groups are Christian and the Hausa-Fulani Muslim, the conflict can be seen as inter-religious. In this instance it is not, although claims of doctrinal truth, the desire to evangelize (which is regarded as a religious duty in both Christianity and Islam) and international influences (both ideological and funding, largely from American evangelical Christianity and Wahhabi Islam) have fuelled the increasing assertiveness of both faith traditions in the Nigerian context, so that some of the clashes have had a significant religious content. General state failure to address inequality, institutionalize peaceful democratic means of resolving conflicts, maintain security and resolve constitutional issues related to religious freedom and citizenship rights for all also contribute to the continuing occurrence of violent conflicts (Best and Rakodi, 2011).
Thus none of the specific cases of inter-religious violence studied here were about doctrinal differences per se. Neither were they about religious expansion, although this is more of an issue in Nigeria. In all the examples analysed, religion has been used as a mobilizing tool in politically motivated conflict, but the relative privilege and disadvantage of the groups concerned and the role of government in addressing inequality or failing to do so, as well as its ability to maintain security in an even-handed fashion, are also important, as are the international links of the rival groups concerned.
The Dynamics of Ongoing Conflict
The second element of Azar’s analytical framework focuses on the dynamics of the process. Violence is often triggered by what appear to be insignificant events. However, Azar suggests that these are often turning points when individual victimization is collectively recognized and expressed. 9 When a protest is met by retaliation or repression, concern with the immediate trigger spills over into multiple issues and leads to increased communal organization and mobilization. Thus both political actions by communal groups and the nature of government responses are relevant to understanding the dynamics of conflict and violence. The political motives for many of the conflicts and violent episodes in the case study cities and the ways in which religion was used as a tool to justify the violence and mobilize supporters have already been discussed. Feeling threatened, Azar suggests, the state is more likely to employ coercion or cooption than to seek accommodation with communal groups. The response of the security agencies is important in determining whether violence is brought to an early end and order restored. Their performance also has a lasting effect on people’s perceptions of the state’s ability and desire to safeguard law and order. In practice, the security forces are often unwilling and/or unable to restore order. In addition, they are sometimes seen as favouring one group over another. The contrasts between Mumbai and Ahmedabad demonstrate this.
In the former, despite the inefficiency and partiality of some police officers, the desire of influential politicians and senior police officers to restore order brought the violence to a speedy end. In Mumbai, many Muslims living in the mixed residential areas where most of the violence occurred took refuge at Musafirkana, where Haj pilgrims congregate before they leave for Mecca, in mosques, or in Muslim-majority neighbourhoods. Non-governmental organizations (NGOs), some prominent Muslim businessmen and clerics, and concerned officials were praised by interviewees for making relief supplies available promptly and assisting victims to claim the compensation to which they were entitled. In contrast, in Ahmedabad, Narendra Modi, the state’s Chief Minister, and other Bharatiya Janata Party (BJP) politicians explicitly backed the attacks against Muslims. Not only did the state police stand by, they prevented people fleeing the violence from taking refuge in the state reserve police camp. The state government established relief camps only reluctantly and temporarily, leaving most Muslims to flee to camps established by faith-based organizations (FBOs) or Muslim-majority neighbourhoods (Gupta, 2011).
In Nigeria, although the police provided sanctuary for the victims of attacks, they were seen as incompetent and often, if officers’ ethnic and religious identity reflected that of the perpetrators, as biased. When they were unable to restore order, the army was called in, often after a delay. The army’s greater capacity (and generally greater impartiality) was effective, if brutal, in restoring order and when it maintains an ongoing presence in a city to pre-empt further violence, this generally seems to be welcomed by residents (Best and Rakodi, 2011). Relief was provided primarily by family and friends, although religious and secular organizations and government agencies also provided humanitarian assistance, especially in the camps, with government arrangements improving over time. For example, the establishment of the Plateau State Emergency Relief and Management Agency after the 2001 crisis improved the state government’s ability to respond after the 2008 clashes (Best and Rakodi, 2011).
The ineffectiveness and partiality of the security agencies have exacerbated mistrust of the state, especially in Nigeria, and led to the emergence of self-help security arrangements at the community level (Best and Rakodi, 2011). These are a double-edged sword in that they may challenge state authority and result in the militarization of society, with groups taking the law into their own hands and the numbers of weapons in circulation increasing (see also World Bank, 2011: 136). In addition to the actions and strategies of communal actors and state agencies, the patterns of interaction between contestants influence both the characteristics of ongoing conflict and the nature of the episodic violence, tending to reinforce and consolidate antagonistic attitudes.
Azar’s framework for analysing protracted social conflict does not distinguish between the instigators and perpetrators of violence. The latter are typically young disadvantaged men, who are easy to mobilize and tend to be blamed, while those behind the violence, seemingly politicians and sometimes religious demagogues, mostly remain in the shadows. This especially seems to be the case in Nigeria, where political violence is endemic and investigating the identity of instigators both difficult and dangerous. In practice also, perpetrators and victims are not distinct groups, as retaliation and revenge attacks occur or targeted communities attempt to protect themselves (Best and Rakodi, 2011; Gupta, 2011). While in the short term, particular communal groups or politicians may appear to win advantages from violent episodes, there are, as noted by Azar, often no clear winners and losers.
Understanding the Outcomes
The outcomes or aftermath of violent episodes must be analysed to understand whether and why conflict is resolved and replaced by lasting peace, or is reinforced and becomes protracted. Azar identifies four aspects of the aftermath that, if the outcomes are negative, tend to perpetuate conflict and increase the likelihood of further outbreaks of violence.
whether public order is restored and physical security maintained
whether the state and other actors have the capacity to tackle the underlying factors, particularly satisfying the basic needs of those involved, especially as resources may be diverted from development to the maintenance of internal security
to what extent a vicious cycle of fear and hostility is created or counteracted, both amongst the communities involved and more widely
linked to the above, the ability of individuals and communities to control their own lives and situations in the face of continued marginalization and lack of confidence in the state.
These have implications for both those who live and work in cities and city governments.
Starting with the last of the factors listed above, the priority for victims in all four cities was in the short term to find a place of safety, followed by shelter and basic necessities, to resume earning, and for their children to start attending school again. In the longer term, victims have to decide whether they can return to their homes and businesses or not, may attempt to obtain government compensation, and expect that their attackers will be brought to justice (Best and Rakodi, 2011; Gupta, 2011). The strategies open to the victims and the roles played by different actors varied. Here it is only possible to identify some of the key similarities and contrasts between the case study cities.
Comparing Mumbai and Ahmedabad
In India, Mumbai’s greater prosperity, active city government, diverse economy and large Muslim business class contrasts with Ahmedabad’s dependence on a declining textile industry, lack of strong radical trade unions, weak city leadership dominated by a BJP-controlled state government, and lack of a prosperous, self-confident Muslim elite, influencing the strategies open to victims (Gupta, 2011).
In Mumbai, the Muslim business community was able to provide employment and economic opportunities to victims who had lost their jobs, premises and stock. Similar opportunities were not available to the Muslim victims in Ahmedabad, who were largely artisans, micro-entrepreneurs or unskilled labourers. In the former, many of those displaced during the violence were able to return to their homes and businesses in their former areas of residence, move to nearby Muslim-majority areas (driving up the price of property in these areas) or relocate to new suburbs with mixed income populations, public transport connections and economic opportunities (Gupta, 2011).
In Ahmedabad, in contrast, the Muslim-owned businesses targeted in the riots lacked the extensive support networks of Mumbai’s Muslim business community and, although most Muslim victims were able to find low-paid work similar to that in which they had been engaged prior to the violence, often elsewhere in the city, many were impoverished. Thus 80% of the 70 families interviewed complained that they had been worse off since 2002, even though in two-thirds women had increased their economic participation, generally working at home sewing or making incense sticks. State government connivance in the violence, government neglect of the victims, the virtual absence of a prosperous Muslim business community and the orientation of civil society organizations to social welfare rather than defending human rights meant that victims had to rely on assistance from Muslim religious organizations (and a few NGOs). When the killing began on 28 February 2002, Muslims reported that they had had nowhere to go but relief camps set up by FBOs, especially the Jamaat-i-Islami sponsored Islamic Relief Committee (IRC), the Gujarat Sarvajanik Relief Committee, an arm of the Jamaat-i-Ulema, and the Urban Economic Forum. While secular NGOs assisted in the camps and claimed much of the credit, it became clear during the research that their contribution had been small in comparison to the Muslim organizations, apart from providing legal assistance to many victims (Gupta, 2011).
Victims were entitled to government compensation, although claimants experienced considerable difficulties in obtaining it and the amounts were meagre. The FBOs helped them by providing money and materials, repairing or rebuilding houses and, for those too scared to return to their previous areas of residence, establishing resettlement colonies. For example, in areas developed by the IRC, single room dwellings with shared water supply, toilets and bathroom facilities were provided. Practices with respect to identifying recipients, the type of tenure provided and whether or not victims’ compensation was supplemented by FBO funding varied between the Muslim organizations. In addition, FBOs installed facilities in areas experiencing an influx of new residents, including a number of schools. There was, however, little evidence that they had radicalized the beneficiaries, some of whom described difficult relationships with the organizations running resettlement colonies and criticized their ideas and methods of operation. For example, the new housing areas were often in distant locations that lacked services, public transport connections, economic opportunities and good quality housing and some of the organizations were suspected of diverting funds received from overseas to the construction of lavish mosques or for their own benefit (Gupta, 2011).
In Mumbai, informants reported that children in Muslim-majority areas started attending school again soon after the riots had ended, but that it took up to a year for children in other areas to resume going to school. In Ahmedabad, the climate of fear was such that parents were reluctant to let their children use public transport, and it took much longer for Muslim children, especially girls, to resume their education. In Mumbai, most Muslims seek education for their children in English or Urdu, feeling little need to develop fluency in the local Marathi language. However, Muslim parents in Ahmedabad saw education in Gujarati as being crucial to both their children’s future work prospects and their ability to blend in and be accepted, needs recognized even by Muslim organizations operating schools (Gupta, 2011).
Finally, in both cities, victims expressed the view that the psychological scars left by the trauma they had suffered would only heal once the perpetrators had been brought to justice. While some are wary of frontally addressing this issue lest it hinder the healing process (see, for example, Oommen, 2008), it is a priority for the vast majority of the victims, who complained that the failure of the court system to bring the perpetrators to account breaches their rights as citizens (Gupta, 2011).
Comparing Kano and Jos
In Nigeria, the victims were displaced to the homes of friends or relatives, camps or other locations. Short-term assistance was provided by local congregations, wider denominational networks and FBOs, in addition to secular NGOs. With repeated violence, individual organizations became better organized and were able to mobilize more resources (including from government agencies). Those displaced came to different conclusions about whether to return to the areas in which they had previously lived and worked, or to relocate to an area in which their ethnic or religious group was in a majority. Fear and the reluctance of landowners to lease land or accommodation to non-indigenous groups deter people from returning to their original area of residence. In Jos, both Christians and Muslims relocated following the 2001 and 2008 crises. As non-indigenous or minority groups moved into more limited areas within the cities (such as the Sabon Gari in Kano), property prices rose, densities increased, less suitable land was built on, and pressures on infrastructure and services worsened in these neighbourhoods (Best and Rakodi, 2011).
Relocation changes the pattern of demand for services. Religious groups preferred to repair or rebuild damaged places of worship and facilities such as schools, because they have both symbolic and practical significance. When such buildings were repeatedly destroyed despite increased security measures or quieter worship practices, some religious bodies ultimately decided to relocate them to the areas in which their users now live. Religious organizations responded more quickly than the state, for example, opening new schools that, while in theory open to all, are rarely attended by children of the other faith. While some of the businesses attacked (both large and small) reopened on the same site, some were unable to resume operating and others relocated, with the result that produce markets gradually became more segregated along religious and ethnic lines (Best and Rakodi, 2011).
Religious organizations played an important role in providing relief for their own adherents, but have only played limited roles in providing longer term support for restoring livelihoods or rebuilding houses. While many religious, as well as secular and government organizations, preach tolerance and co-existence, and there have been a variety of reconciliation or conflict prevention initiatives, including attempts to encourage Christian–Muslim dialogue, these seem to be rather ad hoc, limited and superficial in the face of deeply entrenched antagonism and government ineptitude and bias (Best, 2007: 80–100; Best and Rakodi, 2011). Informants universally regarded the police as inept, leading many communities to maintain the vigilante arrangements they had established at the time of the violence.
Lastly, as in India, commissions of enquiry were set up after each of the violent episodes, although in Nigeria the commissions’ reports were not published. Victims in Nigeria, as in India, were critical of the government for not implementing the recommendations made by successive commissions, as well as failing to bring the perpetrators of violence and members of the security forces suspected of extra-judicial killings to justice (Best and Rakodi, 2011).
In both countries, there has been increased residential segregation along religious lines (see also Arif, 2009; Bollens, 1998). In Mumbai, even in religiously mixed neighbourhoods, Hindus and Muslims are increasingly segregated into different lanes. In Ahmedabad, the process of residential segregation has a long history, starting before the development of the textile industry, continuing during industrialization, and reinforced by the 1969 riot, following which nearly half the Muslim population ended up living in Juhapura, outside the municipal boundary. Fragmented housing areas, politics, workplaces and trade unions, Chandhoke (2009) argues, prevented the emergence of effective pro-democracy civil society organizations (CSOs) and facilitated the rise of Hindutva politics. In addition, Chatterjee (2009) details how the violence has reordered both neighbourhoods and common spaces, replacing ‘democratic imaginations with xenophobic perceptions, which then institutionalize difference and segregation’ (p. 1011). Increased numbers of Muslim-dominated suburbs have emerged in both cities, although more of these are middle or mixed income in Mumbai than in Ahmedabad (Gupta, 2011). Although residential differentiation and segregation in Nigeria has pre-colonial and colonial roots, the inter-religious violence has exacerbated it in both Kano and Jos (Best and Rakodi, 2011). The re-naming of ‘ethnically cleansed’ neighbourhoods (for example by Christians as Jesus Zone or New Jerusalem and by Muslims as Jihad Zone or Saudi Arabia has symbolically reinforced the segregation; Harnischfeger, 2004: 446). Flight to other cities spreads accounts of the violence, sometimes exaggerated, fuelling anger and the desire for revenge (see, for example, Smith, 2006). Increased residential segregation reduces contacts between members of different faith communities, making it easier to manipulate their perceptions of each other and reinforcing their sense of a threatened communal identity.
A second result of residents’ experience of violent episodes and their aftermath in both countries has been decreased confidence in the state’s neutrality, capacity to maintain order and ability to address the underlying problems. In Gujarat, the state government lost its legitimacy in the eyes of Muslims when it permitted, even encouraged, the majority community to attack them in the name of ‘the people’, who were claimed to have one history, culture and religion. The more it has prevaricated on the award of compensation and the pursuit of justice, the less credible it has become with members of the Muslim minority (Gupta, 2011). In Nigeria, government agencies are often accused of having paved the way for violence by giving permission for protests, demonstrations or elections despite high levels of tension and are criticized for their failure to respond promptly, neutrally and effectively when violence occurred (Best and Rakodi, 2011). The ongoing crisis of confidence in the ability of states to maintain order, provide services and achieve development objectives is exacerbated by their failure to ensure justice and take effective initiatives to build peace.
Conclusion
The analytical framework developed by Azar (1990) provided a useful conceptual starting point for the analysis of protracted inter-religious conflicts in the cities studied in this research. Not only does it reflect the most convincing insights from the international literature that existed at the time it was developed, many of his ideas have been confirmed by more recent analyses. That is not to say that the framework can be adopted without critical appraisal: it was developed more than 20 years ago and was set out in broad terms. Theoretical explanations and empirical analyses of conflict and violence have proliferated since 1990 and the framework does not necessarily enable a full understanding to be reached of conflicts in particular societies and localities. Gupta (2011), for example, draws on other theory for his analysis of Hindu–Muslim violence in India. However, the hypotheses included in the framework have mostly been confirmed by this comparison of episodic communal violence in Mumbai, Ahmedabad, Kano and Jos, despite the contrasts between the countries and cities studied. First, some of the main conclusions from the comparative analysis will be presented and finally, some tentative implications for policy and practice will be outlined.
In the context of a protracted social conflict, violent episodes have a tendency to occur in cities because people with different identities live in close proximity to each other, competition over resources and opportunities can be intense, news and rumours can reach large audiences quickly, and it is relatively easy to organize and mobilize large numbers of people. Inter-religious clashes are often attributed to differences in religious beliefs or evangelization efforts. There is some truth in this in Nigeria, enabling religion to be used as a justification and potent mobilizing tool for what are really inter-ethnic struggles for power and resources. Indigene/in-migrant and majority/minority status are also important factors influencing access to resources and inequality, especially in Nigeria but also in India – religious differences may be aligned with different groups, so that they can be played upon, rather than being the causes of conflict. It is often difficult to determine the identity of the instigators of violence, but generally they seem to be political rather than religious leaders and organizations, although religious actors may influence relations between members of different faith traditions because of the symbolic and emotional salience of religion.
Violent episodes have immediate costs to cities and their populations in terms of deaths and injuries, destruction of property (houses, business premises, government and religious buildings) and disruption of lives, livelihoods and social networks. They also have longer term economic and social costs, of which four are highlighted here. First, many victims are impoverished, although some people may benefit from the reconstruction process and rising property prices. Second, the cost of maintaining effective security is likely to increase, while the emergence of vigilante arrangements to compensate for security shortcomings potentially unsettles relationships between city residents, government agencies and the security services. Third (repeated) violence almost always seems to lead to increased residential and community segregation, damaging the prospects for mutual understanding, conflict resolution and the prevention of further violence. Finally, new investment both within and from outside a city where violence has occurred may be deterred. 10 This constrains the opportunities for residents (especially young people) to earn adequate incomes, so that inequality persists and may even increase, as well as limiting the ability of local government to raise revenue to improve services.
The research found that religious organizations often played important roles in the provision of immediate relief to those affected by the violence, but a mixed role in the longer-term aftermath, because of their limited capacity and inability to foster improved understanding between faith groups in a context of inter-religious conflict and violence. While NGOs can and do play roles in both relief and development, the informants in this research were firmly of the view that they are no substitute for an effective state. However, government agencies often fail to fulfil their responsibilities for restoring order, maintaining security, supporting reconstruction and ensuring justice for the instigators and perpetrators of violence. Although civil society organizations, including religious organizations, may play important roles in relief, restoring normalcy and building bridges between groups, much of the responsibility must lie with government. The likelihood of conflicts being sustained and reinforced, as well as periodically erupting into violence, is increased if governments fail in their five key tasks: to maintain security that is neutral and even-handed, to address poverty and inequality (especially the lack of opportunities for disadvantaged groups and young people), to foster links between communal and religious identity groups, to provide access to justice for those harmed, and to uphold the citizenship rights of all.
It is possible to identify five implications for policy and practice arising out of this comparative analysis of inter-religious conflict and violence in the urban context, although there is insufficient space to do more than provide a basic list here. 11 First, the main priority is for the citizenship rights of all to be recognized and protected by the law and the courts and for perpetrators (and instigators) of violence to be brought to justice. Second, governments and security services must operate in accordance with the law and constitution, with their accountability ensured through the democratic process. However, security sector reforms are also likely to be needed, to prevent the proliferation of weaponry and the emergence of vigilantes, to ensure the neutrality of the security services, and to improve their capacity. Third, government organizations (and sometimes the security services) can play positive roles in providing relief and establishing mechanisms to improve coordination and communication when violence occurs. Fourth, the needs of disadvantaged groups (especially the young, and especially young men) need to be recognized and addressed, to safeguard or improve their wellbeing (for example Muslims in India) and reduce tensions (for example between poor Muslims and poor Christians in Nigeria). Finally, while increased segregation in the aftermath of violence cannot be prevented, because people’s need to feel safe must be respected, appropriate policy responses are needed to prevent it worsening the economic position and wellbeing of those who are forced to relocate and widening the divide between religious communities. 12 Both government agencies and civil society groups, including religious organizations, have roles to play in this regard.
Footnotes
Funding
The research was funded by UK Aid from the UK Department for International Development for the benefit of developing countries. The views expressed are not necessarily those of DFID.