ARTICLES

 

Between Resilience and Radicalisation: Reassessing the Trajectory of Internally Displaced Populations in Cabo Delgado, Mozambique

 

 

Egna Sidumo^I; Bjørn Enge Bertelsen^II

^IDept. of Government, University of Bergen
^IIDept. of Social Anthropology, University of Bergen. https://orcid.org/0000-0002-3194-3664

 

 

---

ABSTRACT

Displacement is an endemic phenomenon that affects those uprooted, the communities that feel the impact of those arriving, governments, and the international agencies which are increasingly engaged in organising the displaced. The current war in the north of Mozambique, which has caused a massive displacement of people from 2017 onwards, may be related to a number of factors, including economic, social and even political. Although some actors and analysts include ethnicity as part of the causes, this has more often than not been analytically downplayed when grappling with the dynamics of, particularly, the groups that oppose the Mozambican government. This article analyses the dynamics of relationships between internally displaced people and host communities in Cabo Delgado, especially underlining, firstly, the resilience of communities in the face of extremist violence and, secondly, the distrust that typically shapes conflict. The latter dimension is aggravated by a historical past based on ethnic, political and social differences - cleavages that are accentuated and reproduced within the centres of displaced people and between these and host communities. We argue that factors such as poverty, hunger, lack of jobs or work opportunities, as well as poor access to arable land for family food production, exacerbate the relationship dynamics and create an environment conducive to the outbreak of small-scale conflicts that can, in the medium and long term, open spaces for radicalisation and more violence.

Keywords: Mozambique, displacement, war, extremism, Cabo Delgado, resilience

---

 

 

Introduction

In Mozambique, thousands of people were forced to move from their areas of origin due to the violence and conflict that broke out in the northern province of Cabo Delgado in October 2017.¹ As we deploy the term 'displaced' here to denote this process, it is intended to transcend a singular focus on a certain category of people and thereby encompass wider societal dimensions and populations.² This inclusive understanding of displacement assumes importance if we are to understand the current war in Northern Mozambique, which caused a massive dislocation of people, and where any analysis of its causes must recognise multiple factors, including economic, social, religious and political ones.³

Focusing on Cabo Delgado, this article presents and analyses the dynamics of relationships between internally displaced people and host communities, emphasising the relevance of humanitarian aid in centres; the question of land; political affiliation; biases based on place of origin; ethnicity; religion; and perceptions of violence. Crucially, we also underscore how, during the ongoing war, forms of resilience are commonly exerted in the face of wide-spread violent extremism and mistrust.⁴ Arguably, the multiple forms of resilience reflect a more profound and long-standing history of civil or non-violent everyday life in the context of war and upheaval in Mozambique, as also shown by analyses of the civil war (1976/77-1992) in the country.⁵ Furthermore, as others have shown for both northern and central Mozambique, various modalities that may be labelled resilient, reconstructive, or mediating in the face of extreme violence, fundamentally shape socio-cultural worlds, ritual, perceptions of politics and, even, time.⁶

While such analyses of present and past periods of warfare and extreme violence in Mozambique often situate a socio-cultural dynamic as central, some of this research often lacks a more nuanced understanding of the on-the-ground and in situ dynamics between various categories of non-combatants. For example, many analyses do approach the conflicts in either a dialectical fashion - such as in the dualism of a singular local population versus a category of perpetrators of violence, be they the forces of the state or non-state actors - or in a triangular form where the civilian population is often represented in analyses as caught in the crossfire between state and nonstate perpetrators of violence.⁷

Furthermore, as Salvador Forquilha and João Pereira have shown in their analysis of long-standing (but also war-induced) patterns of mobility in Northern Mozambique, there is also an argument to be made for taking in both deeply embedded migratory patterns related to fishing and mining and the porosity of the border between Mozambique and Tanzania.⁸ In addition, and as they also helpfully point out, the recent war in Northern Mozambique is also fuelled by migratory patterns from other African war zones (for instance the Democratic Republic of Congo) and patterns of mobility from what they call 'Salafist circles' in Tanzania, Kenya and Somalia. Thus, their analysis underlines a general argument which we also agree with, namely for the need to de-nationalise analyses of the conflict in Northern Mozambique. In order to do so, it is also necessary to pay attention to the narratives made by those embroiled in the violence - hereunder IDPs and host communities.

 

Enter the mashababe: A select genealogy of Cabo Delgado violence

The insurgent group operating in Cabo Delgado has been given many names, including Al-Shabab, Ansar Al-Sunna, Islamic State Mozambique, machababo or mashababe. In this text we will use the latter term in keeping both with emerging agreements among researchers, as well as its resemblance to local perceptions and pronunciations.⁹ Emerging in 2017, the rebel group has, since then, conquered territory and inflicted significant casualties on the Defence and Security Forces of Mozambique (FDS) and contributed massively to the widespread displacement of populations. The unfolding violent conflict has led to the emergence of a serious humanitarian situation, indexed by the accommodation centres created to receive displaced populations, and the intervention of regional forces through SAMIM-SADC¹⁰ and Rwandan forces in 2021.

After one year of these regional interventions there was an improvement in security in select northeast parts of the province with the Rwandans creating a gradually expanding security belt in Mocímboa da Praia. In contrast, SAMIM-SADC faced strong resistance in the areas around Nangade district, where the villages became completely abandoned. Later thousands also fled villages in Macomia and Muidumbe.¹¹

At the time of Sidumo's fieldwork, in 2022, Cabo Delgado had almost 85 centres housing nearly a million people designated as displaced.¹² Until 2022, the city of Pemba, with around 220,000 displaced people, was the district with the largest number distributed - many with host families in the Paquitequete neighbourhood.¹³ Metuge, Chiure and Balama were home to a total of 219,279 displaced persons distributed in 24 centres - the majority being in Metuge with 15 centres.¹⁴ As the number of displaced is comparatively high in Cabo Delgado, the province is apt to study, both due to its level of internal displacement and the significant spatial and temporal variation in Internally Displaced Persons flows.¹⁵

These two examples both underline how displacement grew significantly since the outbreak of the first attacks in October 2017. Widely documented, this rise in numbers was accompanied by an increase and geographical spread of violent events that induced hundreds of thousands of people to flee.¹⁶ According to data available at the important observatory and website Cabo Ligado, between January and December of 2020 there were on average 30 violent events in Cabo Delgado per month with the peak in June with over 55 attacks and the lowest number of reported attacks in March, which was around 10.¹⁷

In 2021, however, the figures changed dramatically. For, despite the number of reported attacks dropping on average, one event changed the entire dynamic of the war - both analytically and strategically: the attack on Palma. This unfolded on 24 March 2021 and was doubly one of the most violent as well as most spectacular events so far during the war, crucially involving the incursion of insurgents into the town of Palma, home of massive investments in the oil and gas sector, where the group carried out a large and well-coordinated attack. By the end of this attack, an estimated 1,3 million people were estimated to need assistance and more than 900,000 faced crisis or emergency levels of food insecurity.¹⁸

During 2021, peaks of violence were recorded in July - the month which coincided with the arrival of Rwandan troops - and in November, even with the presence of regional forces. While we will not go into details about the causes and effects of the emergence of these forces, we would like to underline that all these events contributed significantly to increasing the number of displaced people - both as clashes between those designated as 'terrorists' and these forces increased, as well as the civilian population being caught between those fighting.¹⁹

Interspersed increases and decreases in violent events, combined with a more generalised trend of more attacks from 2017 onwards, impacted the flow of internally displaced people. At the beginning of the attacks, in 2017 and 2018, the number of internally displaced people was below 50,000, mostly from areas close to Mocímboa da Praia, growing slowly until the first half of 2020, when it rose exponentially from 235,000, to the late 2023 where there is an estimate of around one million IDPs. This number has remained stable since 2022, with only small variations due to the very slow return of displaced people that has occurred in recent times due to a (seeming) stable security situation.²⁰

This overview of the sharp rise in violence and the concomitant increase in numbers of displaced underscore how domestic management practices might, with an emphasis on the potential, endanger the safety of the communities and facilitate the emergence of new spaces, forms and potentials for what we term 'radicalisation' in Cabo Delgado. A caveat is now in order. When we use the term 'radicalisation' in this article, we do not mean to imply that we would like to single out highly vulnerable and impoverished poor populations of IDPs or their host communities as potential insurgents, terrorists or threats to any national or local political order. Instead, we approach radicalisation as reflecting these hardships and as a way in which to grapple with the potential hardening of political identities, as well as the kind of work of establishments of political imaginaries and cosmologies that are constantly in the process of being constructed in Mozambique - a reflection of already existing political violence and conflict involving the Mozambican state, armed groups and the opposition, as well as deprivation and halted possibilities for future aspiration, especially for the youth.²¹ As for instance Liisa Malkki has shown for Tanzania already three decades ago, analysing emergent political cosmologies and dynamics in camp settings is crucial to attain a deeper understanding of dynamics of radicalisation than the far too simplified blueprint statements of the situation in Cabo Delgado constituting a case of 'Islamic Terrorism' as alleged by early observers, such as Bussotti and Torres.²² Thus, and put differently, we understand that it is not necessarily the flow of displaced people that has brought or can bring challenges to people's security but the management of displaced people in centres characterised by great lack (in most senses of that term).

It should also be noted that since the start of the conflict in 2017, the years between 2021 and 2022 have been exceptionally challenging in terms of managing the effects of the conflict in the province. Crucially, the significant increase in March 2021 of the number of displaced people after attacks to the town of Palma, when people hailing from different places, as well as as diverse historical, ethnic and social origins found themselves immersed in a common context of survival. Although this mixing and coexistence in the centres predated the Palma attack, this feature became increasingly prevalent as attacks increased in several districts following it.

In the first years of managing displaced people, people were organised according to the neighbourhood and/or community to which they belonged. As the attacks grew, this organisation, which continued to be a management goal, was in practice diluted reflecting the number of displaced people who flocked to the IDPs centres and neighbourhoods around seeking assistance.²³ In the following section we turn to analysing the effect of such a significant shift in the composition and dynamics in the IDP centres and surrounding communities.

 

Locating sites for empirical research: some considerations

To gather empirical evidence that served as the basis for this article, a total of six IDP centres in Cabo Delgado were selected in three districts: Metuge, Chiure and Balama. Empirical field observations and data gathering on the ground was executed by Sidumo for this, while Bertelsen has conducted brief fieldwork in Pemba in November 2023, providing supplementary perspectives.

While not largest in numbers of IDPs, these districts all have an intersectionality of factors that allow analysing, in addition to what we identify as 'resilience' factors. While we will return to the notion of resilience below, we hold crucial intersectional factors for Metuge to be increasing forms of tension, as well as restrictive forms of security. As will be developed below, this entailed the imposition of restriction of freedom of movement, particularly of displaced people. These were, then, subject to high levels of surveillance and, in some cases, violence by the military generally. Crucially, such violence followed information shared by community members with military and police authorities about their fellow displaced and such revelation of information was, we argue, often based on what we will call their values. Values-driven relations aim to transform political and social behaviour of others and include history, tradition, cultural/religious background and ethnicity which, combined with interests (security, prosperity and well-being), shape the external and internal policies of most countries, and for our purposes, also communities.²⁴

Balama and Chiure, close to the province of Nampula, were targeted by the mashababe, who managed to expand their actions in 2022. Also here, issues of trust and the revelation of information from host communities were key issues.²⁵ During fieldwork by Sidumo on site in 2022, a young man who benefitted from professional technical training resettled in the Ocua IDP Center in the Chiure district, said:

They [in reference to the receiving community] distrust everyone and consider us all terrorists and that is why we are distant from them and many times they do not accept that we provide services. They accuse us of being informants, friends or family members of those who attacked us in our villages.

Although a very short excerpt from a longer conversation, the young man's frustration illustrates how many host communities believe that their proximity to displaced people's centres and/or living with displaced families make them vulnerable. Such vulnerability is not only to potential attacks by the mashababe but, moreover, also to surveillance and control by police and military authorities. This composite sense of being at risk due to the presence of IDPs creates gaps and distrust between the displaced communities and the recipients. The actual on-site dynamics is important here:

Metuge borders the Quissanga district, which registered the first attacks in March 2020 and was the first district to implement a model of assistance, in addition to the setting up of IDP centres. Balama borders Niassa and Nampula - the latter through Lalaua that, together with Mecuburi are districts with borders with Cabo Delgado that has not yet suffered any attack. Chiure borders Nampula through the districts of Memba and Erati, and was attacked for the first time in 2022, and is one of the pioneer districts in the planning and ordering of centres for displaced persons with a land distribution system for creation of neighbourhoods and streets and with minimum conditions of access to water and basic services.²⁶

 

Fieldwork: Methodological and ethical considerations

Fieldwork for this article draws mainly on work carried out by one of the authors, Egna Sidumo, between August 2020 and October 2021 but is supplemented by Bjørn Enge Bertelsen's fieldwork in Pemba in November 2023. We will first provide an overview of Sidumo's work before, in brief, describing the work of Bertelsen.

Sidumo focused on semi-structured interviews with individuals and focus groups, especially in the IDP centres. These were conducted by two male and female researchers who covered both sexes in resettlement centres in Pemba-Metuge, Chiure and Balama districts. Interviews were undertaken in what can only be described as contexts of high distrust and vulnerability, the level of distrust reflecting linguistic and ethnic diversity as well as a scarcity of livelihood that oftentimes position researchers as targets of either distrust or being construed as potential benefactors - in spite of our efforts to inform that the encounters were exclusively for research purposes. Crucially, there is also the risk of being identified as a representative of the state and/or person with an undisclosed agenda - challenges that Sidumo and her team, and Bertelsen later on, were acutely aware of during fieldwork.

Building on work that began in 2020, Sidumo and team carried out interviews between 5 and 25 October 2021 with stays of 3 to 4 days. Due to linguistic diversity, it was necessary to elicit support to translate statements from people who spoke neither Emakhuwa nor Shimakonde (the main languages of the province²⁷) but Kimwani and Kiswahili. While one may argue that such participatory multiplicity may have skewed the information received, such a potential weakness was counteracted by a systematic checking of responses and translations.

Direct observation of the individuals and different groups in the centres was also privileged, with special attention to women, and youth due to their specific role in the communities, especially regarding resilience and mistrust. While configured differently among Makhuwa, Mwani and Makonde, the attention to women is not coincidental as in Northern Mozambique's matrilineal communities, women, especially mothers and sisters, have authority in their families and communities beyond the domestic work and their roles as wives.²⁸ Hence, in an analysis of our kind, there is a need to also understand gendered dynamics of behaviour and relationship in the camps for displaced people.²⁹

For the semi-structured interviews with individuals, the samples were based on their responsibilities and/or roles at the level of local administration and centres or reception areas, and for the focus groups, the sample was based on the snowball technique, due to the nature of the centres and to the hierarchical relationships within the structure of the IDP centres. Local leaders of centres were asked to identify age-differentiated members of focus groups of men and women.³⁰

Another group that participated in the focus group discussions separately, was the local and traditional leaders of the regions of provenance. Due to their image of authority, there was an understanding that they would be more predisposed to assume and defend a more authoritarian position. They could also even impose a specific way of analysing the concerns of displaced populations based on their interests and/or narratives that could interest a series of stakeholders who finance the subsistence of local leaders in the centres and reception areas influencing the narrative of the displaced.³¹

As local leaders tended to choose people in their own network, a selection was requested that included at least one representative from each district and representation of different groups (ethnic, religious, linguistic and gender). Furthermore, we opted for this differentiation as participants would receive a symbolic monetary compensation that we did not want to be interpreted as favouring certain groups to the detriment of others.³² The information that participants would benefit from a monetary offer and the respective distribution was only made after the focus groups had been carried out to prevent communities from feeling obliged to omit or add information about their experience, and avoid riots that even occurred in some centres when displaced people thought, for example, that people identified to participate in focus groups and/or interviews benefited from additional social services.³³

One of the greatest, if not the main, challenges for researchers during fieldwork, especially in contexts of violence and conflict, is directly linked to negotiating access to the places where they intended to collect data. As a Mozambican researcher with work in Cabo Delgado, Sidumo had previously collaborated with the Cabo Delgado provincial level to obtain necessary permits. During fieldwork, Sidumo also worked with local CSOs (civil society organisations) that assist displaced people in the places where data collection took place, always maintaining impartiality and transparency about the nature of the data to be collected, as well as the methodology.

Bertelsen's short and Pemba-based fieldwork in November 2023 was to a great extent informed both by long-term collaboration with Sidumo and built, in part, also on preliminary analyses of material made available by her. Having had no previous experience with fieldwork in Cabo Delgado but having undertaken research in Manica province and Maputo since 1998, Bertelsen also benefitted greatly from being able to access the research infrastructure provided by João Feijó from Observatorio do Meio Rural during his fieldwork in Pemba. During this time, Bertelsen oscillated between direct observation in various parts of Pemba city and formal as well as informal interviews and conversations with displaced people from Cabo Delgado, as well as residents in the city.

Several scholars³⁴ recommend that researchers explicitly prepare for both anticipated and unanticipated ethical dilemmas and in this case, measures were taken to inform local leaders, respondents, and participants of focal groups about the objectives of the research and their rights as participants. Although care was taken to choose safe locations for fieldwork, a prior assessment of local security was also made for both Sidumo and Bertelsen's cases regarding the vulnerabilities of specific population groups in each of the centres. Finally, for neither Sidumo or Bertelsen, no questions were asked about the (potential) involvement of interviewees, conversation partners or focus group participants in acts of violence.

 

A divided community

As already stated, many of those who have survived conflicts could arguably be described as 'resilient', in the commonsensical meaning of persevering and surviving in the face of extreme adversity. Simultaneously, however, as our material shows, the IDPs may also be characterised as taking (or may take) an extended period to initiate and cultivate relationships of trust in the areas where they settle. In many cases, this follows exposure to violence, whether among their peers (who may also very often be victims of violence), or with the communities that are commonly labelled as 'receiving' IDPs. Such reception, as we will develop, is a complex and sometimes contested process and may be both peaceful and integrative or more fraught with conflict and tension.

A note should now be made about the use of the term 'community'. At one level, the notion of 'community' may be approached as a political-cum-ideological fiction more than as an analytical concept. As Brint has pointed out, the notion has, for instance, fallen out of use in sociology - although an argument is made for it still being relevant (albeit in a reformatted fashion).³⁵ Moreover, as Barrios makes evident in a critical analysis of, precisely, the concept of 'community resilience', the notion of 'communities' must neither be seen as static nor bounded. We should therefore be careful in assuming coherence, solidarity or other values to be operative in IDP centres as well as other entities that one may label 'communities'.³⁶ Nevertheless, when we choose to retain the term in this article, it is based on the pragmatic stand of contributing to those studies that use the term comunidade (community), as well as to highlight what we see as aspirations of development or retaining relational and other pro-social dynamics on the ground - aspects also relevant for understanding resilience. Let us be more specific about the links we make between resilience and the potential for radicalisation through five factors³⁷:

1. Cultural identity and connectedness: This is familiarity with one's own cultural heritage, practices, beliefs, traditions, values and norms (can involve pluri- or multicultural dimensions).

2. Bridging capital: This relates to trust and confidence in people from other groups; support for and from people from other groups; strength of ties to people outside one's group; having the skills, knowledge and confidence to connect with other groups; valuing inter-group harmony; active engagement with people from other groups.

3. Linking capital: Trust and confidence in government and authority figures; trust in community organisations; having the skills, knowledge and resources to make use of institutions and organisations outside one's local community; ability to contribute to or influence policy and decision making relating to one's own community.

4. Violence-related behaviours: The willingness to speak out publicly against violence; willingness to challenge the use of violence by others; acceptance of violence as a legitimate means of resolving conflicts.

5. Violence-related beliefs: The degree to which violence is seen to confer status and respect; degree to which violence is normalised or well tolerated for any age group in the community.

Furthermore, as shown by Hopwood and O'Brien³⁸, analyses of those labelled resilient in the context of conflict, demonstrate a diversity of heightened forms of awareness. Resilient people are more prone to recognise individual or group characteristics or behaviours, or perhaps external or environmental factors, that are useful when it comes to facing existential threats. This has also been shown by several studies of other conflicts across both African as well as non-African contexts: Henrik Vigh's analysis from Guinea-Bissau and Northern Ireland on being 'sussed out',³⁹ Carolyn Nordstrom on the importance of rumours for survival during civil war,⁴⁰ and Danny Hoffman's analysis of reading the landscape and socio-spatial contexts among young fighters in Sierra Leone are among these.⁴¹

Existing research has also highlighted negative interactions between displaced people and host communities. Tricia Hynes' work on refugees highlights that trust and mistrust are central for studies of displaced.⁴² Others have explored how host populations are affected by the arrival of numerous refugees or displaced, as well as how these communities deal with trust issues when security and protection needs are not met.⁴³ Drawing on these studies we here consider that trust in communities of both migrants and IDPs rests on forms of reciprocity and the expectation of a shared future.⁴⁴

Government institutions and humanitarian agencies face serious challenges in helping populations and ensuring access to basic goods for everyone's survival. Often, these efforts inadvertently increase inequalities and may foster a sense of injustice among displaced people and a risk for a propensity for violence through radicalisation - the process of developing extremist ideologies and beliefs - as a precursor to political extremism. We note this despite the recognition that most radicals did not (and do not) engage in extreme political violence, and that many political extremists did not (and do not) 'radicalise' in any traditional sense.⁴⁵ We also underline that 'radicalisation' is a concept that stresses the evolving and fungible character of identity formation.⁴⁶

These senses of inequalities and injustice fit into Borum's conceptual model that attempts to explain how grievances and vulnerabilities are transformed into hatred against a target group and how hatred is transformed into a justification or impulse for violence.⁴⁷ Fundamentally, the process begins with framing some unsatisfactory event, condition, or situation. This leads to a complaint (this is not right) because it is understood to be unfair (this is not fair).

The injustice is attributed to a target policy, person or nation (it's your fault). The party responsible for the situation is then defamed and demonised (you are evil), facilitating the justification or impetus for aggression. This is illustrated in Figure 1. While originally developed as a training heuristic for law enforcement, we suggest it may also illuminate some aspects of the cycles of violence in Cabo Delgado.

 

 

Radicalisation is a multi-determined process, driven and sustained by multiple causes, and there are, of course, limits to such a graphic representation above. For instance, causal factors often include broad grievances that 'push' individuals toward a radical ideology, and narrower, more specific 'pull' factors that attract.⁴⁹ Authors such as Habibe et al.,⁵⁰ as well as Maquenzi and Feijó,⁵¹ argue that while the main causes of the emergence of violent extremism in Mozambique were domestic, these were linked to a set of political grievances as well as social, economic, cultural and religious differences - in addition to a lack of public security and the almost non-existence of government control in several areas of the province.

Push factors can be defined as conditions conducive to violent extremism at the structural level (high rate of unemployment, general lack of education, absence of governance), and pull factors refer to the personal and socio-economic incentives that may contribute to the decision of joining a violent extremist group. This may include a sense of power, economic freedom, ethnic or religious reward.⁵²

 

Assessing intra- and inter-community dynamics

In 2018⁵³ and 2020⁵⁴, two studies on the community resilience capacities and vulnerabilities in the context of violent extremism were undertaken in the Northern Mozambican provinces of Nampula and Cabo Delgado by employing the above-mentioned BRAVE questionnaire. These studies revealed that while both provinces presented strong resilience capabilities, Cabo Delgado presented significant vulnerabilities compared to Nampula.

In the Nampula study of 2018, communities presented strengths in the overall five resilience factors (Cultural identity and connectedness; Bridging capital; Linking capital; Violence-related behaviours; and Violence-related beliefs), showing signs of community engagement and connectedness; showing trust in the government authorities; exhibiting strong cultural identity and solid non-violent behaviours and beliefs. The major risk factors in Nampula were linked to the context of cyclones (considering that the research was conducted right after the cyclone Kenneth, which affected both Nampula and Cabo Delgado); to poor involvement in group activities; and to overall socioeconomic inequalities in the country.⁵⁵

In contrast, Cabo Delgado communities showed moderate levels of resilience to violent extremism, to the extent that, for example, political, ethnic and religious issues were raised among displaced communities and were concluded to be the basis of distrust between them.⁵⁶ The low to moderate levels of resilience for Cabo Delgado were, generally, manifested through what was argued to be poor connectedness between the communities and their culture and in low levels of engagement within the communities, especially by young men. Young men are identified as most vulnerable while, conversely, women who were more involved within the community nonetheless had weak participation in decision-making. Also, the perception of political autonomy and equal treatment remained poor, especially amongst the youth population. Crucially, these also showed dissatisfaction in relation to the government authorities, which has also been widely documented in other studies.⁵⁷

Interestingly for our research, while violent norms were generally rejected, the willingness to effectively oppose violent behaviour and beliefs were argued to be underdeveloped.⁵⁸ Inter-community connectivity was found to be moderate to low as individuals proved to be more positive towards engagement with similar communities rather than with different ones. In fact, cultural tensions became apparent amongst people originating from different districts - especially, Mocímboa da Praia, the first district to suffer terrorist attacks in Mozambique on 5 October 2017, and Quissanga.

Ethnic tensions were also identified between the Mwani, Makhuwa and Makonde groups, as Mwani are considered as being intolerant of other ethnic groups while the Makonde are perceived to hold feelings of superiority over other ethnic groups.⁵⁹ Scholars such as, for instance, Eric Morier-Genoud,⁶⁰ Edward Alpers,⁶¹ Sérgio Chichava⁶² and many others have argued and shown that historically derived and composed ethnic tensions in Northern Mozambique has been important to the conflict. Morier-Genoud suggests that the predominantly coastal Mwani have been recruited for the insurgents' groups due to experiencing socio-political marginalisation, a lack of economic development, and estrangement to the political preferences of their neighbours, Makhuwa and Makonde. The latter also has historical connections to the party in power and the leadership of the current president of Mozambique, Filipe Nyusi, as documented by several researchers.⁶³

Muslim communities were also found to be expressing intolerance towards Christian diet, styles of worship and interfaith marriages, and showed to be more receptive to violent rhetoric than Christians. This can, perhaps, in part be explained by what Eric Morier-Genoud⁶⁴ has argued, namely that religious divisions are superimposed onto ethnic, social, political and power divisions in the province where the coastal Makhuwa (Makhuwa-Meto) and mainly Mwani communities which had a 'glorious' pre-colonial Islamic and Swahili past, form a Muslim majority whilst the pre-dominantly Christian minority Makonde hold the social, economic and political power in the province.

In fact, the difference between religious groups proved significant, simply as Muslims make up a majority of Cabo Delgado society and as violent extremism in the province is strongly linked to extremist Islamic ideologies. Politically, these three groups have different preferences which affect their confidence (or lack thereof) in the Mozambican government. For instance, data on the electoral processes in 1994 and 2014, demonstrated that, although the major opposition party Renamo failed to win against the ruling party (Frelimo) in the province, it achieved more significant support in coastal areas where much of the population is Muslim and predominantly Makhuwa and Mwani. This contrasted with the complete hegemony of Frelimo in the plateau districts and inland areas of the province.⁶⁵

In sum, we argue, these features are salient not only throughout Cabo Delgado but not least in the centres for the displaced and may go some way to explaining the moderate levels of resilience to violent extremism manifested in the poor connectedness between the communities in the origin areas and in the displaced centres. Further, the dynamics presented above are consistent with the general development in these centres where people from different contexts are practically forced to share space despite their differences.

 

 

 

 

 

 

As many other similar types of refugee camps or camps for displaced people across Africa⁶⁶, also the IDP centres visited could be characterised as spaces cordoned off from their surroundings and with clear internal structures and forms of leadership separate from the surrounding communities.

The centres in Figures 2 and 3 were funded by United Nations but also host or involve the World Food Program (WFP), the International Organisation for Migration, the United States Aid Agency (USAID) and several other UN Agencies and NGOs. Further, international institutions in the field of health offer psychosocial and other support to the displaced, and CSOs in different clusters generally lead the assistance in close coordination with the local governments where they are located.

The displaced populations installed in this centre mainly receive a basic monthly food basket from the WFP for their survival, as well as medical assistance, especially for children. For various reasons, however, many of these families have not received food regularly, nor medical care, not to mention that the means and possibilities of subsistence through work are also scarce, creating tensions of various kinds.

 

 

The distribution of food is primarily undertaken by the WFP with the support of CSOs contracted for this process. However, the registration of beneficiaries, a key prerequisite for the distribution of food, is exclusively done by government authorities at the local level. Crucially, these often reflect patterns of corruption or patronage to benefit family members or other people based on ethnic and/or religious affinities, mainly in reception neighbourhoods and to a lesser extent in accommodation centres where people are registered based on time of arrival. Under these circumstances, food distribution and other support is somewhat biased despite local authorities insisting that there are absolutely no criteria based on origin in the distribution of support.

Furthermore, support is inadequate to cover all IDPs in the centres leading to severe critique and protest, creating, we argue, a fertile environment for the outbreak of conflicts.⁶⁷ Let us provide one example. The distribution of aid in Cabo Delgado is led by humanitarian agencies of the United Nations supported by CSOs into thematic clusters and local authorities.⁶⁸ These identify, register and make lists of those displaced people who are deemed vulnerable and may, therefore, be eligible to receive support. However, support is irregular and in some centres the food is sometimes inadequately supplied, let alone the utensils. In this regard, a woman displaced from a centre in Balama stated the following in an interview:

Here we receive food from the World Food Program, but we never received pans, bowls, plates or backpacks for our children to go to school, but we heard that in Metuge and Ancuabe people have everything and even have a hospital. When we get sick, it is hard to go to the hospital. Why don't we also have the right to all that? Why do others have everything and us nothing?

 

Potentialities for tensions and instability in Cabo Delgado's IDP Centres

Since the start of the war, while several NGOs and CSOs have been installing themselves to assist the victims of the conflict, they have done so in a scattered and disjointed way, according to Guivala and Chingotuane.⁶⁹ For instance, the location of specific centres makes assistance difficult to reach, subjecting the communities to greater subsistence problems than others that are closer to the district headquarters and/or more accessible to humanitarian agencies. Exacerbating the situation, in these centres there is a significant absence of government officials and representatives.

The centres are generally organised according to the origin of the displaced. Ostensibly for security reasons, people from different districts, villages and/or areas are identified according to their areas of origin to control access and guarantee that all people who access the centres are, or can be, duly recognised as members of their communities, whether through language, religion or even recognition by part of the members of that region with kin or family members. Although there are no reports of specific segregation by language, origin or religion, a sense of abandonment and difficulties in accessing resources and aid is quite often interpreted by IDPs as a willed act to make their lives difficult.

This impression is also supported by macrotendencies as, between November 2022 and February 2023, the flow of humanitarian aid has suffered limitations. This particularly affected WFP's distribution of food, in addition to having had to deal with an inflation in the number of people registered to receive food.⁷⁰ Reflecting the sentiments outlined above, in the centres, families and groups that did not have access to food felt discriminated against, reinforcing their convictions that only certain groups had access to support. The situation was further aggravated, we hold, by two factors.

For one, this form of ordering, despite preventing unknown individuals from accessing the centres and allowing the effective recognition of the populations to receive support, may not only be seen as a simple distribution tool. Instead, it also cements displaced people based on their origin, religion and, further, makes these vulnerable to inferences which may be made about their political preferences - regardless of these being the least or most likely to initiate conflicts within centres. This type of structuring along ethnic and/or sectarian lines may be harmful for people's safety.

Second, this division prevents different people to coexist and to share profound experiences of the war and their common fate in the camps. They are, thus, unable of building or bridging capital and end up as increasingly isolated communities fostering feelings of marginality which are deepened when opportunities, whether economic or social, or even in the distribution of aid, are not universal. Let us provide an example from fieldwork undertaken in October 2021.

At 7 am, the team of researchers arrived at a centre for displaced people in Ocua, Chiure District, for another day of work. As it commonly happens in all the centres where we work, we asked that people be grouped by age (adolescents and young people) and gender (men and women), so that the focus groups could start. Due to the level of lack of jobs and opportunities, communities always think that anyone who has paper and pen in hand comes to register people so that they can have social and/or economic benefit. Our team clearly explained that it was not a matter of registration, but the populations gathered and created a situation of near riot that did not allow us to work. We immediately had to take shelter in one of the huts in the centre where we decided that there were no security conditions to manage the focus groups. A group approached us and started shouting. These are examples of what was expressed:

We want to talk... it's the fault of the people of Mocímboa da Praia... they are troublemakers and they always do that… They always have complaints and use violence as a form of pressure.

Made aloud by both young men and women who wanted to participate in the focus groups, these expressed that the attitudes of the members of their communities were wrong. They also accused them of being destabilising elements, thereby demonstrating a clear lack of trust between them who believed that part of their difficulties in accessing support, such as food and/or employment, were compounded by having to be mixed with people who did not share their values. The example challenges assumptions about the uniformity of camp dwellers and it, also and crucially, starkly demonstrates the flaws in the territorial-populational organisation of the camp - an example of the security-territory-population continuum in operation.⁷¹ However, in relation to our main argument, it also serves to demonstrate a weakness in the bridging capital factor, in that coexistence and cultural exposure in the last years failed to build or was, simply, not promoted by authorities working with displaced populations. In such a perspective, values of trust and respect for differences, however defined, are essential to mitigate the probability of emergence of radicalisation networks in centres. Evidence from the field shows, however, that the organisation of the camps entails fundamental challenges to inter-community and inter-personal communication to the point of experiencing very high levels of distrust - especially between people categorised as belonging to different ethnic groups.

Together with experienced marginalisation, a sense of abandonment and a depletion of material, social and economic resources, these make the centres more likely to have the potential to develop forms of contestation which may, in the long term, turn into violent forms of demonstrating strength and demanding responses to their demands targeting government officials and humanitarian workers. It is important here to underline the 'may' as we would like to stress that distrust does not necessarily entail forms of radicalisation.

Another difficult factor and cause of potential conflict relates to (material) relations beyond the camp. For example, between 2020 and 2021, a land distribution policy for displaced populations was initiated in some centres in Cabo Delgado and by 2022 and according to data from the government of Chiúre district, the district had demarcated 3,396 and distributed 2,648 plots of equal number of families.⁷² Under construction were, until November of 2021, also one health centre, two classrooms, as well as other ongoing rehabilitations in Marrupa. For instance, in Balama, displaced families received land meant not for housing but for agricultural food production, supported by the distribution of seeds and technical assistance from agronomist technicians.

This distribution of land to displaced people - which sometimes includes land far from the main village - has not been peaceful and relations have in some instances become strained and fraught with fear. It has also had the direct effect that displaced populations, fearing to suffer attack from the so-called host communities, will abandon cultivating great distances from the village, opting instead to encroach on lands closer to the housing areas. Also, they may attempt to pay fees to be able to cultivate closer to their homes. A group of women who participated in the focus groups explained to us why they have this preference:

Here in the district, we have land, and the government has given us land to produce food, but they are far from the centre, and we are afraid... I'd rather pay 500MT to the population here to lend me land than to go produce far away... My whole family died because they were ambushed on the way to the machamba [plot of agricultural land] and we don't know when or where the mashababe will come to attack us again…

Because of these disputes, so-called host communities, in for instance Chiure, have shown great resistance to ceding their land to displaced people - the latter nevertheless engaging in exchanging food received from humanitarian agencies for plots of land for rent. The justification for not agreeing to cede land derives from the fact that the host communities do not want to live with the displaced populations for a long time and hope that they will soon be able to return to their areas of origin.

Importantly, the dispute for arable land close to inhabited areas is, however, peaceful, manifested in speeches, with no record of violence occurring. While this demonstrates low levels of bridge capital between them, it also, importantly, underlines the belief that violence is not an acceptable situation and/or valid means of resolving differences. This finding shows that, although populations do not benefit from specific programmes to prevent the spread of violent extremism in their areas, their beliefs, and behaviours in relation to violence are strong enough to avoid outbreaks of violence in their communities.

Furthermore, the fact that the violence that has lasted for 6 years has not spread to the centres where thousands of people live, may indicate that the experience of indiscriminate attacks, of which they were victims, has raised their awareness of the use of non-violent means as a way of a claim - a pattern of resistance to violence itself also concurrent with research on earlier periods of large-scale political violence in Mozambique.⁷³ Furthermore, in all centres and neighbourhoods where people had the opportunity to share their experiences, it was unanimously stated that the violence did not target people based on their colour, language, ethnicity or religion - with only minimal variations in the way women and girls were treated by the insurgents in the incursions through the villages.⁷⁴ Nonetheless, concurrent with other reports, it was also expressed how, depending on their age, appearance, religion and in some cases ethnicity, girls and women were chosen to serve the displaced as wives or slaves, abducted, raped and/or killed in extreme cases - reflecting a gendered pattern of violence.⁷⁵ A woman recounted this experience to us about what the first attacks were like, and showed her position in relation to violence:

When they arrived in the village in Muidumbe, they didn't ask anything… They killed everyone, burned everything, and took the women they wanted, especially the most beautiful ones... Violence doesn't solve anything... I tell my children every day not to think about killing people for cause of war. We cannot kill our brothers.

In addition to these gendered patterns of violence, there were also particular groups of people that faced difficulties after fleeing attacks - for instance those having lived in fishing communities, especially those that could not be in areas close to the sea (Metuge, Pemba and Mecufi), received arable land for agriculture, which made adaptation and survival difficult. Most of these from fishing communities, known as excellent traders, opened small businesses selling essential products in various centres (see Figure 6).

 

 

Through the implementation of a programme geared towards providing technical training and paid internships, young people from fishing communities are served by mobile training units.⁷⁶ However, despite the implementation of training in areas such as carpentry, complaints and adaptation challenges persist. Generally, host communities complain about high labor costs and pay well below the market, alleging inexperience of these new service providers.

Commonly, host communities avoid contact with displaced populations due to stigma they may have or on suspicion, sometimes verbalised, that these harbour links to the mashababe - despite such links having not in any way been proven. In this regard, a young man who benefited from professional technical training but who was unable to find work, said the following:

I was a fisherman in Mocímboa da Praia but here I have nowhere to fish. Here the government taught us how to make tables and chairs and other things... I already have another knowledge and the materials but here people also don't have money to buy things and sometimes they don't want us in their homes... They say we are terrorists.

Probably due to these (as well as, potentially, other) tensions, those interviewed, regardless of their origin and/or ethnic group, unanimously condemned violence and reiterated that they did not know who was committing acts of violence in their communities. Furthermore, they also reiterated their desire to return to their areas of origin expressing that they could bear to live with difficulties and without means of subsistence. An elderly man pointed out that, under the circumstances in which they live in the centres, they would prefer to return to their areas of origin. However, not everyone shares this view - here in the voice of a middle-aged man:

We ask the government to stop the war so that we can go back to our homes… we are tired of staying here without food and without machamba… if they tell me to come back tomorrow, I'll run again… I don't want to stay here…

As of late 2022, the absence of adequate aid services, the desire for a better quality of life, and perceptions of improved security conditions has motivated almost 352,437 displaced to return to their homes in 238⁷⁷ locations in Mocímboa da Praia, Muidumbe and Palma district.⁷⁸ New resettlement areas include land distribution programmes and systems for building and organising more resilient houses, with basic services such as water and/or hospitals (fixed or mobile). Despite these returns, sometimes spontaneous and sometimes with the support of the government, humanitarian agencies continue to show concern for the safety of the returnees due to the lack of security guarantees.⁷⁹ This still very limited return of the populations, despite supposedly reducing the pressure of the displaced people centres, also represents a danger for the returned populations. This is while, despite the improvement of the security situation, sporadic attacks do continue. This could in the short and medium term facilitate the process of attracting young people and women to the groups, strengthening their ability to operate not only in the areas where the populations have returned but also in areas where displaced populations decide to settle.

 

Resilience against violent extremism in a context of mistrust

Based on our discussions above, we can surmise that the factors driving radicalisation may occur at various levels - from the individual to the community - and displaced communities in all the centres reflect such differences. However, above all, the centres bring together different individual experiences that may have the potential for radicalisation. If we return to the resilience factors presented above and link them with the areas of origin of the displaced people, we can easily conclude both that, although there was no record of violence, tension is present, and differences are engaged and expressed at community level. This was illustrated in some of the cases and quotes above.

It is imperative to underline that despite the frictions typical of coexistence between different people in a new environment, displaced populations do attempt to orient their lives around socio-cultural values at both an individual and group level. Such orientation may entail exclusion or integration of members of the community. Importantly, however, following a policy that has persistently been promoted by government authorities, they are only allowed to focus on their immediate demands and interests and therefore very rarely refer to what could have been seen as their main differences and challenges in neither living together in the centres of displaced people nor those tensions that arise between them and the host communities. The potential that unresolved tensions can have in the outbreak of new conflicts is therefore largely muffled.

Exacerbating this situation, while local authorities are generally absent in the management of the centres, they are represented in the local leaderships - but these often have no or limited capacity to grapple with the dynamics of potentially conflicting relationships between the populations. This inability has the potential to create major frictions that are currently manifested in the desire of many displaced people to return to their areas of origin. For despite all those displaced (those interviewed for this work) receiving food and in some cases even medical assistance and/or education for their children, they felt profoundly neglected and at risk. Many understand that the best option for them is to return to their perceived areas of origin.

In sum, as we have attempted to empirically provide some basis for and without in any way allocating blame or forms of negative agency to the IDPs, displaced people may (often inadvertently) have the potential to destabilise host communities. This may occur directly by changing the demographic composition (ethnic or sectarian) of communities and, indirectly, by imposing a heavy burden (economic and social) on local communities that already struggle to access scarce resources. The centres for displaced persons therefore represent an important focus of the conflict because they are areas of a high concentration of people of diverse ethnicities, languages and religions, living in a context of poverty and scarcity where they reproduce power relations based on their history and vision of the circumstances that preceded the violence.

Although there are no studies or reports that document that there is propensity for community violence in displacement centres, these may have the potential for violence and conflict - particularly because the socio-economic conditions that partly contributed to the outbreak of violence continue to occur in the centres. In some cases, the conditions of the populations have deteriorated significantly in the displacement centres. Furthermore, among the complaints of displaced people, it was possible to see that not only do they feel somehow abandoned and/or unjustified, but they also clearly identify the actors who, in their view, make their living conditions difficult, therefore making appeals for them to have the opportunity to return to their areas of origin. Some of the actors identified by the communities include the receiving communities, the government and some humanitarian agencies.

Because of this deterioration in economic and social conditions, in some areas minor tensions emerge between the receiving communities, and the displaced communities for various reasons. Experiences or feelings of favouritism toward the displaced felt by the host communities in relation to humanitarian agencies, and the refusal to share their land so that the displaced can produce their food, confirm the moderate inter-community connectivity.

The dynamic of the war is changing continuously. As of early 2024, the spread of violence to other areas not previously attacked, as well as what may be seen to be changes in the strategy of the mashababe, including having a less hostile attitude towards the returned populations, indicate novel contours of violence on the horizon. In our view, this would demand a greater commitment of the government and partners to not managing only displaced communities but also paying attention to the needs of host communities. Supporting and capitalising on pro-social sentiments shared among the diverse groups would be a good place to start.

 

 

1 This work would not have been possible without the valuable contribution of the reviewers who, from the first versions, made important comments and selflessly guided our work. The same should be said for the guest editors of this issue - thank you! We would also like to thank the researchers who supported part of the data collection. In Cabo Delgado between 2020 and 2021: They are Carlos Abudo, Joselino Sualé and the Social Action Technique assigned to the District of Chiure, Flávia Nicolau. In addition, we would like to extend our thanks to João Feijó who generously supported and facilitated Bertelsen's fieldwork in November 2023 in Pemba. We would also like to acknowledge that research for this article, including fieldwork, has been generously supported by the Norwegian Research Council-funded project God, grievance, and greed? Understanding Northern Mozambique's new Islamist war (RCN project number 316070) hosted by Chr. Michelsen Institute and with Dr Liazzat Bonate as project leader.
2 E. Colson, 'Forced Migration and the Anthropological Response', Journal of Refugee Studies, 16, 1, 2003, 1-18.         [ Links ]
3 See also E. Morier-Genoud, Towards Jihad? Muslims and Politics in Postcolonial Mozambique (London: Hurst Publishers, 2023).         [ Links ]
4 J. Feijó, Assimetrias Sociais - Pistas para Entender o Alastramento do Jihadismo Islâmico em Cabo Delgado (Senegal: Friedrich-Ebert-Stiftung - Gabinete Paz e Segurança Centro de Competência África Subsariana, 2020).         [ Links ]
5 B. E. Bertelsen, Violent Becomings: State Formation, Sociality, and Power in Mozambique (New York: Berghahn Books, 2016);         [ Links ] C. Nordstrom, A different kind of war story (Philadelphia: University of Pennsylvania Press, 1997);         [ Links ] C. Geffray, La cause des armes au Mozambique. Anthropologie d´une guerre civile (Paris: Karthala, 1990);         [ Links ] E. Morier-Genoud, M. Cahen, & D. M. d. Rosário (eds), The War Within: New Perspectives on the Civil War in Mozambique, 1976-1992 (London: James Currey, 2018).         [ Links ]
6 B. E. Bertelsen, 'Civil war and the non-linearity of time. Approaching a Mozambican Politics of Irreconciliation', Journal of the Royal Anthropological Institute, 28, S1, 2022, 50-64, DOI: https://doi.org/10.1111/1467-9655.13753 ;         [ Links ] N. Nyachega & W. Mwatwara, 'On Renamo "War", Entrepreneurial Synergies and Everyday Life in the Honde Valley Borderlands, c.1980s-2020', Journal of Southern African Studies, 47, 6, 2020, 973-991, DOI:10.1080/03057070.2021.1985315 ;         [ Links ] N. Wiegink, Former Guerrillas in Mozambique (Philadelphia, PA: University of Pennsylvania Press, 2020).         [ Links ]
7 There are, of course, multiple and important analyses that also provide a more nuanced reading, including P. Israel, In step with the times. Mapiko masquerades of Mozambique (Athens, OH: Ohio University Press, 2014);         [ Links ] S. C. Lubkemann, Culture in chaos. An anthropology of the social condition of war (Chicago: University of Chicago Press, 2008);         [ Links ] D. Trentini, At Ansha's. Life in the spirit mosque of a healer in Mozambique (New Brunswick, N.J.: Rutgers University Press, 2021).         [ Links ]
8 S. C. Forquilha & J. Pereira, 'Migration Dynamics and the Making of the Jihadi Insurgency in Northern Mozambique', e-Journal of Portuguese History, 20, 2, 2023, 132-154.         [ Links ] For overviews of long-standing patterns of migration, also related to trade and artisanal fishing, see also A. R. F. d. Conceição, Entre o mar e a terra. Situações identitárias do Norte de Moçambique (Cabo Delgado) (Maputo: Promédia, 2006).         [ Links ] As Liazzat Bonate has also shown in much of her important work, there are also long-standing patterns of mobility and contact that can be traced through Islam and trade networks. See, for instance, L. J. K. Bonate, 'O jihadismo transnacional e a insurgência em Cabo Delgado, Moçambique', Afro-Ásia, 65, 2022, 519-553.         [ Links ]
9 'Mashababe' is a term used locally for those insurgents invoking the Islamic faith. For a discussion on terminology, see also J. Feijó, 'Caracterização e organização social dos Machababos a partir dos discursos de mulheres raptadas', Observador Rural, 109, 2021. Available in https://omrmz.org/wp-content/uploads/OR-109-Caracteriza%C3%A7%C3%A3o-e-organiza%C3%A7%C3%A3o-social-dos-Machababos.pdfização-social-dos-Machababos.pdf (accessed 6 November 2023).         [ Links ]
10 Southern African Development Community Mission in Mozambique (SAMIM-SADC), deployed on 15 July of 2021 as a regional response to support Mozambique to combat terrorism and acts of violent extremism. More information available in SADC Mission in Mozambique (SAMIM) in Brief | SADC.
11 J. Feijó, 'A Return to a New Future, or a Return to the Past? - The Population Return to the Northeast of Cabo Delgado', Destaque Rural, 195, 2022. Available at https://clubofmozambique.com/wp-content/uploads/2022/11/DR-195-A-return-to-a-new-future-or-a-return-to-the-past-1.pdf , (accessed 4 November 2023).         [ Links ]
12 See International Organization of Migration (IOM), 'Assessment of Displacement Dynamics' (Maputo: IOM Mozambique: August). Available here https://dtm.iom.int/product-series/displacement-report-8?page=0 (accessed 6 November 2023).         [ Links ]
13 Secretaria de Estado da Provincia de Cabo Delgado (SEPCD), 'Evolução da situação dos deslocados, Gabinete do Secretario de Estado' (GSE, 2023).         [ Links ]
14 SEPCD, GSE.
15 Let us be more specific in terms of IDPs: The district of Chiúre had around 13,989 displaced families in 2023, equivalent to 55,786 displaced people, coming from the districts of Mocímboa da Praia, Quissanga, Palma, Nangade, Macomia and Muidumbe. It has 6 reception centres, namely the Marrupa Centre, Katapua, Meculane, Ocua, Mazeze and Chiúre-Sede, with plans for other centres to be built depending on the evolution of the displaced situation. Balama District has 21,851 displaced people housed in three centres located in Mphiri, Ntawanne and Angalia, from the districts of Mocímboa da Praia, Palma, Muidumbe, Quissanga, Macomia, Nangade and Meluco, victims of occasional attacks in the last 3 years when the violence began to spread geographically.
16 See, e.g., J. Feijó, J. Maquenzi & N. Balane, 'Do we stay or do we return? Poverty, social tensions and displaced population's perspectives of return', Destaque Rural, 171, 2022. Available at https://omrmz.org/destaque_rural/voltamos-ou-ficamos-pobreza-tensoes-sociais-e-perspectivas-de-regresso-das-populacoes-deslocadas-english-version-available/ , (accessed 4 November 2023).         [ Links ]
17 ACLED. Cabo Ligado, 'Cabo Ligado Monthly: December 2020', 2021. Available at https://acleddata.com/acleddatanew/wp-content/uploads/2022/01/Cabo-Ligado-Monthly_Dec-2021.pdf, (accessed 4 November 2023).         [ Links ]
18 A. Doctor, 'After Palma: Assessing the Islamic State's Position in Northern Mozambique', Program on Extremism at George Washington University, 2022.         [ Links ]
19 See here, again, J. Feijó, 2021.
20 Although there is no space for a full comparison with earlier Mozambican conflicts here, we would like to note that such a slow pace of return not only reflects the lack of a sense of peace and stability but also mirrors earlier patterns of partial and hesitant return after conflict - for instance following the end of the civil war and its multiform social, material, spiritual, political, gendered and cosmological upheavals. See, e.g., these works: Bertelsen, Violent Becomings; S. C. Lubkemann, Culture in chaos. An anthropology of the social condition of war (Chicago: University of Chicago Press, 2008);         [ Links ] Wiegink, Former Guerrillas; A. M. Honwana, Espíritos vivos, tradições modernas. Possessão de espíritos e reintegração social pós-guerra no sul de Moçambique (Maputo: Promédia, 2002).         [ Links ]
21 See, e.g., J. Sumich and B. E. Bertelsen, 'Just out of Reach. Imminence, Meaning, and Political Ontology in Mozambique', Current Anthropology, 62, 3, 2023, 287-308;         [ Links ] J. Pearce, 'History, legitimacy, and Renamo's return to arms in central Mozambique', Africa, 90, 4, 2020, 774--795.         [ Links ] For an instructive analysis of youth and radicalisation in Cabo Delgado, see J. Feijó, J. Maquenzi & A. R. Rachide, 'Ingredients for a youth revolt - poverty, consumer society and frustrated expectations', Observador Rural, 121, 2022. Available at https://omrmz.org/wp-content/uploads/OR121-Ingredients-for-a-youth-revolt-poverty-consumer-society-and-frustrated-expectations-English-Version.pdf (accessed 19 October 2023).         [ Links ]
22 L. Bussotti & C. Torres, 'The Risk Management of Islamic Terrorism in A Fragile State: The Case of Mozambique', Problems of Management in the 21st Century, 15, 2020, 9-22.         [ Links ] See also L. Malkki, Purity and exile. Violence, memory, and national cosmology among Hutu refugees in Tanzania (Chicago: University of Chicago Press, 1995).         [ Links ]
23 This finding is based on direct observation by Sidumo and her interviews with local leaders of the first centres created in Metuge in 2020, and subsequent work in these centres and others within the province of Cabo Delgado. Other sources that address the organisation and management of displacement centres can be found at https://response.reliefweb.int/mozambique/camp-coordination-management-cccm/assessments.
24 T. Obrien, 'Interests and values in international relations', New Zealand International Review, 38, 4, July/August 2013, 16-21.         [ Links ]
25 On trust and distrust in relation to refugees, please see J. C. Knudsen & E. V. Daniel, Mistrusting refugees (Berkeley: University of California Press, 1995).         [ Links ]
26 M. Guivala & É. Chingotuane, Causas, Determinantes, Agravantes e Impacto do Conflito no Norte de Moçambique: Uma Análise na Perspectiva de Género (Maputo: ONU Mulheres, 2022).         [ Links ]
27 Cabo Delgado Province Language Map available in Cyclone-Kenneth-in-Cabo-Delgado-Mozambique---Crisis-language-map.png (709×1024) (translatorswithoutborders.org).
28 Aphwiamwene: a figura da Rainha nas sociedades matrilineares Moçambicanas (tsevele.co.mz).
29 See also, as a background text, E. Sidumo, 'O papel das mulheres na prevenção da violência extrema: subsídios para uma estratégia de prevenção do extremism violento sensivel ao género em Moçambique' in A. R. Agy (ed), Genero, Cidadania e Desenvolvimento em Moçambique (Maputo: Escolar Editora, 2022), 41-62.         [ Links ] For works documenting key gendered dimensions in the North, see for instance Signe Arnfred's highly useful retrospective account of analyses of women and power over almost five decades: S. Arnfred, 'Re-visiting "African Tradition", Re-thinking Gender and Power: Learning from Fieldwork in Northern Mozambique', Nordic Journal of African Studies, 32, 3, 2023, 286-306, DOI: 10.53228/njas. v32i3.1089.         [ Links ]
30 The purpose of organising groups by age was to preclude the possibility that young women, for example, would feel intimidated in telling their experiences and challenges in the centres for displaced people in front of other adult women who could be guardians and/or direct family members. As for the separation between men and women, the objective was to prevent women, due to cultural issues, from not actively participating in the discussions because they considered that men had priority in speaking to the detriment of women.
31 The scope of this article as well as the space allotted, does not provide opportunity for any full analysis of the political, social and cultural roles of the vast field of what is sometimes called tradição (tradition) in Mozambique. See these sources for an introduction: B. E. Bertelsen, Violent Becomings; H. M. Kyed, J. P. B. Coelho, A. N. d. Souto and S. Araújo (eds) A dinâmica do pluralismo jurídico em Moçambique (Maputo: Centro de Estudos Sociais Aquino de Bragança, 2012).         [ Links ]
32 All focus group participants received a fixed amount regardless of age, gender or position for participating in the research. The decision to give monetary compensation was made to reflect the precarious position of the IDPs and to compensate them for the time they made themselves available.
33 Food distribution riots were witnessed by the main author of the article in the districts of Chiure (where there were gunshots) and Montepuez where displaced people organised to demand more food. For more information on irregular food distribution and tension among displaced people, see also J. Alberdi et al., 'War, forced displacement and response to the crisis in Cabo Delgado - Mozambique' (Pemba: Ayuda en Acción, 2023). Available at ayudaenaccion.org/uploads/2023/06/War-forced-displacement-and-responses-to-the-crisis-in-Cabo-Delgado.pdf (accessed 5 November 2023).         [ Links ]
34 S. P. Campbell, 'Ethics of Research in Conflict Environments', Journal of Global Security Studies, 2, 1, 2017, 89-101, DOI: https://doi.org/10.1093/jogss/ogw024;         [ Links ] L. A. Fujii, 'Shades of truth and lies: Interpreting testimonies of war and violence', Journal of Peace Research, 47, 2, 2010, 231-241, DOI: https://doi.org/10.1177/0022343309353097;         [ Links ] D. Kostovicova & E. Knott, 'Harm, change and unpredictability: The ethics of interviews in conflict research', Qualitative Research, 22, 1, 2022, 56-73, DOI: https://doi.org/10.1177/1468794120975657.         [ Links ]
35 S. Brint, 'Gemeinschaft Revisited: A Critique and Reconstruction of the Community Concept', Sociological Theory, 19, 1, 2001, 1-23, DOI: https://doi.org/10.1111/0735-2751.00125.         [ Links ]
36 R. E. Barrios, '"Here, I'm not at ease": anthropological perspectives on community resilience', Disasters, 38, 2014, 329-350, DOI:https://doi.org/10.1111/disa.12044.         [ Links ]
37 The Building Resilience Against Violent Extremism (BRAVE) is a measure of risk and protective factors for young people's resilience to violent extremism, through a 5-factor analysis. For more details, please check Background - BRAVE (resilienceresearch.org).
38 B. Giddings et al., 'Environment, economy, and society: fitting them together into sustainable development', Sustainable Development, 10, 4, 2002, 187-196, DOI:https://doi.org/10.1002/sd.199.         [ Links ]
39 H. Vigh, 'Vigilance: On conflict, social invisibility, and negative potentiality', Social Analysis, 55, 3, 2011, 93-114.         [ Links ]
40 Nordstrom, A Different Kind.
41 D. Hoffman, The war machines. Young men and violence in Sierra Leone (Durham: Duke University Press, 2011).         [ Links ]
42 T. Hynes, 'The issue of "trust" or "mistrust" in research with refugees: choices, caveats, and considerations for researchers (London: UNHCR Working Paper no. 98, 2003).         [ Links ]
43 A. H. Richmond, Global Apartheid: Refugees, Racism and the New World Order (Oxford: Oxford University Press, 1994).         [ Links ]
44 Colson, Forced Migration.
45 R. Borum, 'Rethinking Radicalization', Journal of Strategic Security, 4, 4, 2012, 1-6.         [ Links ]
46 I. V. Starodubrovskaya, 'Islamic Conflict and Violence in Local Communities: Lessons from the North Caucasus', Perspectives on Terrorism, 14, 2, April 2020, 80-92.         [ Links ]
47 R. Borum, 'Radicalization into Violent Extremism II: A Review of Conceptual Models and Empirical Research', Journal of Strategic Security, 4, 4, Winter 2011, 37-62, DOI: http://dx.doi.org/10.5038/1944-0472.4.4.2.         [ Links ]
48 Although the original model refers to what the author terms 'the terrorist mentality', the notion of terrorism is not relevant for the present analysis.
49 R. Borum, 'Radicalization into Violent Extremism II: A Review of Conceptual Models and Empirical Research', Journal of Strategic Security, 4, 4, Winter 2011, 37-62, DOI: http://dx.doi.org/10.5038/1944-0472.4.4.2.         [ Links ]
50 S. Habibe, S. Forquilha & J. Pereira, 'Radicalização Islâmica no Norte de Moçambique: O Caso de Mocímboa da Praia', Cadernos IESE nº17 (Maputo: Instituto de Estudos Sociais e Económicos, 2019).         [ Links ]
51 J. Feijó, 'Assimetrias Sociais - Pistas para Entender o Alastramento do Jihadismo Islâmico em Cabo Delgado' (Dakar: Friedrich-Ebert-Stiftung - Gabinete Paz e Segurança Centro de Competência África Subsariana, 2020);         [ Links ] Habibe et al., Radicalizaçã Islâmica.
52 USAID Policy, 'The Development Response to Violent Extremism and Insurgency: Putting Principles into Practice' (New York: USAID, 2011). Report available at https://pdf.usaid.gov/pdf_docs/Pdacs400.pdf (accessed 6 November 2023).         [ Links ]
53 ALPS Resilience, 'A Survey of Communities at Risk of Violent Extremism in Tanzania and Mozambique: Rapid Risk Assessment Concerning the Resilience of Vulnerable Communities against Violence' (Cape Town: Alps Resilience, 2019). Available at alpsresilience.com/_files/ugd/ae1dfd_758282fe793444ad81a84aa2cd5b7fd5.pdf.         [ Links ]
54 ALPS Resilience, 'Identifying Resiliencies in Cabo Delgado, Mozambique', 2020.         [ Links ]
55 ALPS Resilience, 'A Survey of Communities at Risk of Violent Extremism in Tanzania and Mozambique: Rapid Risk Assessment Concerning the Resilience of Vulnerable Communities against Violence' (Cape Town: Alps Resilience, 2019). Available at alpsresilience.com/_files/ugd/ae1dfd_758282fe793444ad81a84aa2cd5b7fd5.pdf (accessed 6 November 2023).         [ Links ]
56 ALPS Resilience, 'A Survey'.
57 See here, for instance, J. Feijó, 'Asymmetries in Access to the State: A Fertile Land for the Penetration of Islamic Jihadism?', Observador Rural, 93, June 2020. Available at Observado https://omrmz.org/wp-content/uploads/Observador-Rural-93-Asymetries-in-the-Acess-to-the-State.pdfr-Rural-93-Asymetries-in-the-Acess-to-the-State.pdf (omrmz.org) (accessed 6 November 2023).         [ Links ]
58 ALPS Resilience, 'Identifying Resiliencies'.
59 ALPS Resilience, 'Identifying Resiliencies'.
60 E. Morier-Genoud, 'The jihadi insurgency in Mozambique: origins, nature and beginning', Journal of Eastern African Studies, 14, 3, 2020, 396-412. DOI: 10.1080/17531055.2020.1789271.         [ Links ]
61 E. A Alpers, 'Ethnicity, politics, and history in Mozambique', Africa Today, 21, 4, 1974, 39-52.         [ Links ]
62 S. Chichava, 'Por uma leitura sócio-histórica da etnicidade em Moçambique', IESE discussion paper01, 2008.         [ Links ]
63 See, e.g., Israel, In Step.
64 Morier-Genoud, 'The jihadi insurgency'.
65 L. D. Brito, 'Electoral behavior in the first multi-party elections in Mozambique' in B. Mazula (ed), Mozambique: Elections, democracy, and development (Maputo: Inter-Africa Group, 1995), 473-499;         [ Links ] L. D. Brito, 'O sistema eleitorial. Uma dimensão da reprensentação política em Moçambique' in L. D. Brito, C. N. Castel-Branco, S. Chichava & A. Francisco (eds), Desafios para Moçambique, 2010 (Maputo: Instituto de Estudos Sociais e Económicos, 2010), 17-29;         [ Links ] M. A. Pitcher, 'Mozambique elections 2019: Pernicious polarization, democratic decline, and rising authoritarianism', African Affairs, 119, 476, July 2020, 468-486;         [ Links ] J. Hanlon, 'Collapsing electoral integrity in Mozambique', Journal of African Elections, 20, 1, 2021, 44-66;         [ Links ] Morier-Genoud, 'The jihadi insurgency'.
66 See, again, Malkki, Purity and Exile.
67 H. B. Danesh, 'Human Needs Theory, Conflict, and Peace: In Search of an Integrated Model' in D. J. Christie (ed.), Encyclopedia of Peace Psychology (Hoboken, New Jersey: Wiley-Blackwell, 2011).         [ Links ]
68 CIP, 'Número de deslocados'.
69 M. Guivala & É. Chingotuane, 'Causas, Determinantes, Agravantes e Impacto do Conflito no Norte de Moçambique: Uma Análise na Perspectiva de Género' (Maputo: ONU Mulheres, 2022).         [ Links ]
70 Nações Unidas Moçambique, 'Fundos Secam à Medida que a Fome Cresce em Cabo Delgado', report available at https://mozambique.un.org/pt/207306-fundos-secam-%C3%A0-medida-que-fome-cresce-em-cabo-delgado (accessed 6 November 2023).         [ Links ]
71 M. Foucault, Security, territory, population. Lectures at the Collège de France, 1977-1978 (Basingstoke: Palgrave Macmillan, 2007 [2004]).         [ Links ]
72 In detail, these included 956 in the Administrative Post of Chiúre-Sede (1,040 Marrupa and 310 Chiote), 950 in Katapua-headquarters, 1,012 in Meculane (of this 948 for housing and 64 for services), 301 in Chiúre-velho, and 433 in Ocua-sede. 3,620 houses were built, 1,350 in the Administrative Post of Chiúre-Sede (1,040 Marrupa and 310 Chiote), 950 Katapua (Headquarters) 586 Meculane, 433 Ocua (Headquarters) and 301 Chiúre-Velho (Headquarters).
73 See, e.g., Bertelsen, Violent Becomings; Wiegink, Former Guerrillas, V. Igreja, 'Trauma e limpeza ritual de veteranos em Moçambique', Cadernos de Estudos Africanos, 21, 43-69;         [ Links ] Honwana, Espíritos vivos.
74 This view of violence as in a sense supra-individual, has also been documented and analysed in Bertelsen, Civil war; Nordstrom, A Different Kind.
75 See also here the important report by J. Feijó, 'Caracterização e organização social'.
76 Marcela & Chingotuane, Causas, Determinantes.
77 ReliefWeb Mozambique - Conflict displacement, Cabo Delgado (DG ECHO, IOM DTM), 'ECHO Daily Flash of 13 January 2023'. Accessible here https://reliefweb.int/report/mozambique/mozambique-conflict-displacement-cabo-delgado-dg-echo-iom-dtm-echo-daily-flash-13-january-2023 (accessed 6 November 2023).         [ Links ]
78 S. Ballard & E. Culombo, 'Enhancing Humanitarian Aid and Security in Northern Mozambique' (Center for Strategic & International Studies: 8 September 2022). Available here: https://www.csis.org/analysis/enhancing-humanitarian-aid-and-security-northern-mozambique (accessed 6 November 2023).         [ Links ]
79 ReliefWeb, UNHCR Mozambique Cabo Delgado - Update, Internal Displacement Response - January 2023. UNHCR Mozambique Cabo Delgado - Update, Internal Displacement Response - January 2023 [EN/PT] - Mozambique | ReliefWeb.         [ Links ]