Introduction

Conflict-Induced Urbanization in Cabo Delgado, Mozambique

In October 2017, an armed insurgency started in Cabo Delgado, the northernmost province of Mozambique, with an attack on the town of Mocímboa da Praia. In the following years, attacks increased throughout the province against state institutions, private companies and the civilian population in rural areas, spreading material destruction and human atrocities. The Mozambican army began a retaliatory campaign, supported by private and public international security forces, to regain control of the territory and dissolve the insurgents (Centro para Democracia e Desenvolvimento, 2022; Matsinhe & Valoi, 2019). Among many interpretations of this insurgency, the official discourse of terrorism, external destabilization, and Islamic jihadism is recognized as part of its complexity, within the history of Islam in Mozambique (Bonate, 2013; Chichava, 2020; Morier-Genoud, 2021). But existing research points to multiple drivers for the violent discontent against the state: the historical political favouring of some ethnic groups; the eviction of local populations from their lands by extractive industries, and their exclusion from the economic benefits; the structural lack of investment in socio-economic development, consolidating poverty and illiteracy, and depriving the youth of opportunities (Macalane & Jafar, 2020; Maquenzi & Feijó, 2019; Weimer, 2020). This multidimensional phenomenon has attracted increased academic interest in Cabo Delgado from national research centres such as Observatório do Meio Rural and Instituto de Estudos Sociais e Económicos; and foreign institutes integrating Mozambican scholars, such as the International Growth Center, the Chr. Michelsen Institute, and the London South Bank University; inter alia.

This violence has so far led to more than 4000 deaths (Cabo Ligado, 07/02/2023), and a massive displacement of civilians in Cabo Delgado and neighbouring provinces, estimated at more than one million in November 2022, when at least 220,000 were living in the provincial capital Pemba (International Organization for Migration, 2022). This is more than double its 201,846 residents in 2017 (Instituto Nacional de Estatística de Moçambique, Censo 2017), a dramatic demographic growth for a “secondary city” (a settlement with less than 250,000 inhabitants (UN-Habitat, 2022)), which accelerates urbanization processes prompting challenges of space, resources, infrastructure and governance; but also opens opportunities for sustainable development if the impacts of rapid urbanization are effectively tackled (Büscher, 2018). The scientific literature on conflict-induced urbanization in Africa focuses on two main perspectives: the human experience — conflict-led displacement and resettlement in cities; and the urban dimension — the material and immaterial effects of conflict and violence on cities. From a human standpoint are relevant the struggles and motivations of displaced persons who choose to resettle in urban centres, to find work and ensure livelihoods, to avoid movement restrictions and assistance dependency, and to access secure shelter and basic service provision. Host communities are the concurrent side, carrying the load of socioeconomic structures and urban governance. Interactions between international actors, local authorities, host communities and newcomers are often characterized by unequal and violent relations of space and power (Kobia & Cranfield, 2009; Baconyi et al, 2019). From the urban stance, cities are often the first line of assistance to persons displaced by war, in the face of slow responses from governments and institutions. Dramatic population influx results in pressure on urban resources (land, food, water, livelihoods), disordered urban sprawl, neighbourhood densification, environmental degradation, social stress, and economic disruption, with a consequent rise of inequality (Beall & Goodfellow, 2014; Muggah & Abdenur, 2018). Simultaneously, new patterns of space, flows of mobility, forms of economy and cosmopolitan identities emerge in conflict-urbanized settlements — the opportunities for reconstruction and development. War shapes relationships between urban dwellers, municipal authorities and state institutions (Beall et al, 2011; Büscher & Mathys, 2018; Oldenburg, 2018).

This article presents the theoretical framework for an inquiry about conflict-induced urbanization in Cabo Delgado, located within a contemporary critique of mainstream urban studies. From phenomenon to theory, the article posits that this phenomenon can be critically analysed through a participatory visual methodology; and from theory to practice, it proposes a design for a comparative case study in the city of Pemba, which was tested in an exploratory field trip in September 2022.

From Phenomenon to Theory

An Epistemic Critique to the Study of Cities

Formal urban scholarship first emerged from urbanization processes observed in Europe and North America since the 1800s, consequent to industrialization, technological innovation, economic growth, rural exodus, and the rise of the bourgeoisie and proletariat as new social classes. Euro-American urbanization became the paradigm of “development” and “modernity”, and rationality, efficiency, and more recently, “greenness”, and “smartness” became conditions of urban “success” against which to compare “the rest of the world” (Robinson, 2006). This Eurocentric worldview based on difference, separation and hierarchization translated into the idea of Euro-American cities as canon for urban theory, and normative assumptions about “other” cities as places of absence and neglect, refusing their intrinsic features and alternative modernities as valid for scientific knowledge (Simone, 2004). Positivist sociology aims to be “pure science” by merging economics and spatial analysis, using empirical constructs, mathematical models, and statistical analysis to test hypotheses, formulate generalizable theories and predict future developments for urban societies (Koch & Latham, 2017, Eds.). This quantitative mindset sees cities as networks of infrastructure, bureaucracy and technology, missing their qualitative, sensorial, and lived dimensions, overlooking the structural forces shaping cities — capital, class and politics — in their territorial and historical contexts; the fragmentation and hierarchization of urban space results in social injustice, especially for poor inhabitants and ethnic minorities (Koch & Latham, 2017; Mbembe & Nuttal, 2004, Eds.). However, viewing cities as a result of economic and institutional interactions still ignores other sorts of division beyond class, such as race, religion and gender, and devalues the agency of urban dwellers to create change, as victims of structural forces; formal governance is only one side of urban life, as informal networks of interaction play a role in providing services and goods, solving problems and creating opportunities (Simone, 2004). Moreover, formerly colonized societies cannot be fully studied without addressing the impacts of colonialism, especially in knowledge production. Thus, interdisciplinary scholars reclaim “other cities” as equally valid for urban theory, as in every city are observable common features of urbanization and globalization, at different levels and scales (Robinson, 2006). Contemporary urbanists propose horizontal comparisons instead of vertical hierarchizations to apprehend the drivers of urban (trans)formation — how cities are made and lived by their inhabitants in their material lives, subjective meanings, and collective interactions, with their daily challenges and strategies to overcome them (Myers, 2001; Parnell & Pieterse, 2016; Patel, 2016). This critique questions research methods from Euro-American contexts which apply poorly to unstable governance, irregular data, great human need, and changeable security found in many cities around the world. Collaborative, interdisciplinary, comparative, and mixed modes of knowledge co-production can grasp the complex quantitative and qualitative dimensions of the urban and bridge gaps between academics, politicians, practitioners, and inhabitants, for positive change. Non-academic voices from cinema, art, photography, journalism, and literature shall be accepted as valid to analyse, theorize, and communicate the city (Parnell & Oldfield, 2014, Eds.; Marrengane & Croese, 2020, Eds.). Contemporary scholars-activists are pushing for the “decolonization” of research by moving from academic “extractivism” to collaboration, aware of issues of power and trust between researchers and participants, and the conventional academic outputs reinforcing them. To see participants as equals enables the humanistic potential of social research (Gubrium & Harper, 2016). One example of this critique is the construct of Southern Urbanism (Pieterse, 2015; Schindler, 2017) emerging from rapid urbanization in Africa and Asia: engaged theory and interdisciplinary methodological experimentation, grounded in these diverse realities, are crucial to producing useful knowledge for positive urban change, through participatory tools and endogenous propositions for urban governance and sustainability. Other decolonial constructs are Indigenous Knowledge (Owusu-Ansah & Mji, 2013) and Relational Research (Gerlach, 2018), connecting knowledge to history and culture against the “neutral objectivity” of Eurocentric positivist science; in most indigenous societies, knowledge is experiential and collective. These constructs query mainstream urban theory and are useful to critically analyse cities, for which some methodologies are presented hereafter.

Visual Methodologies to Critically Research Cities in conflict

Cities are objective spaces where inhabitants carry their material lives in built and natural environments; simultaneously, cities are subjective places of individual and collective dynamics between institutions and citizens; cities are spatial, social, economic, political and environmental terrains of human interaction, privileged contexts of conflict due to agglomeration and heterogeneity (Kallus, 2016; Menezes et al, 2012). This interface between humans and the environment is embedded in values, beliefs and attitudes, imprinted by sensorial impressions and historical memories; to research them, it may be useful to use research methods that grasp these material and immaterial urban realities. In sites of conflict, complexity is often made invisible through the marginalization of groups and the negligence of policy; visual methods can contribute to reversing this invisibility (Moore et al, 2008). Visual methods of data collection and representation can portray these multisensory and conflictual urban conditions; conduct qualitative inquiries beyond quantitative analysis; bring forward excluded groups and produce socio-spatially grounded knowledge, to usefully contribute to urban change and justice (Lambert & Allen, 2016; Portela & Errandonea, 2017). Also, visual methods can approach cities from the viewpoint of the main actors: the urbanites — to depict their lived realities, identify daily challenges, and propose suitable solutions; participation in research can activate feelings of agency and citizenship, evermore significant in conflicted territories. Participatory methods recognize the active role of “research subjects” as “research participants” and “knowledge co-producers”, which in urban research can bridge gaps between citizens, academia, and policymakers for positive development and urban sustainability (Cervantes, 2019; Marrengane & Croese, 2020, Eds.). In summary, a methodology of visual techniques can be suitable for the apprehension and representation of the urban space, with participatory methods for the expression of the experiences of the urban residents, to critically analyse cities in contexts of conflict. This asks for a deep knowledge of the research context through literature and direct contact, efficient communication and organization skills, and a flexible design to encompass unexpected challenges arising from collaboration (Gubrium & Harper, 2016).

According to the cited authors, participatory visual methodologies include a few steps: (1) initial discussions among participants; (2) co-production of visual material; (3) participatory analysis of the visual data collected; and (4) validation and publication of results in collaborative outcomes. (1) Initial discussions may be individual or collective (if safe and relevant) with diverse urban stakeholders — municipal officers, traditional authorities, political agents, community representatives, ordinary citizens of varied genders and ages, long-term residents, and recently arrived persons; the goal is to stimulate productive discussions about the phenomena of interest, the challenges around them, and the places where they happen, how to make them visible, with what expected outcomes. (2) The participatory production of visual data is done with images (photography), videos (filmmaking), or maps (cartography). Photographs and videos can be produced with cameras and cellphones in individual or collective city walks, or individually during daily life, to access places of difficult or dangerous admittance to outsiders, or events in which their presence may change existing dynamics and affect a truthful grasp of the phenomena. Similarly, collaborative maps can render visible what is otherwise “invisible”, as they have the power to conceal and reveal the existence of places; maps can materialize representations of social practices in the urban space through hand-drawn or open-source online cartography. (3) The participatory analysis of visual data happens when participants observe the photographs, videos, and maps produced, and discuss in interviews or focus groups through the technique of visual elicitation. Visual elements trigger memories, reflections, emotions, and opinions without direct orientation or structured questions, in organic conversations at a personal pace; this can unlock data otherwise sensitive to reveal, especially around trauma and vulnerability, balancing research relationships where participants are experts rather than just informants. Spatial elicitation can also happen in walking tours, where direct observation of places can trigger the sharing of information. (4) The final step is the validation of results among participants and the publication of the visual outcomes, to share the knowledge acquired with urban stakeholders, academics, and the public, through reports, presentations, storytelling events, visual exhibitions, paper and digital publications, social media, scientific writings, and conferences. Consented recordings of the participatory process, observation notes and audience feedback give contextual information to triangulate the primary data.

Challenges, Ethics, and Positionality in Participatory Visual Methodologies

When designing a participatory visual methodology, it is essential to consider its relevance for the theoretical framework of the project, the spatial and socio-cultural context, and the ethical challenges that may arise, especially in cities amidst conflict. Participatory visual methods must be tailored to the research questions and the participants to usefully collect and analyse data and obtain valid results; the strengths and weaknesses of each method shall be considered before selection. Photography requires low technical equipment and skills, being effective for participatory research almost everywhere, and especially appealing to youth with access to social networks; however, it may not be appropriate if images are problematic for cultural or political causes. Filmmaking is a powerful tool for raising awareness, and the global spread of audiovisual platforms enables easy sharing of videos with low-speed internet, making it an efficient research tool; still, some technical skills are needed, as is software for video edition (editing the co-produced video clips should be participatory as well), which renders its usage somewhat complex. Map-making can be a top-down process with participants trained in specific tools, depending on their interest, skills, time and funding for the project; or it can be a bottom-up approach with tools selected according to the baseline conditions of participants and context. This flexibility is a plus, yet some geographic literacy is needed to read and produce maps, which are subjective to perceptions of space and connections to places, thus needing triangulation (Gubrium & Harper, 2016).

A realistic selection of methods must consider the technological literacy of participants and logistical issues such as time, access to electricity (to charge devices) and the internet (to share data). It calls for a reflexive mindset to grasp and integrate cultural perceptions about visual media and digital technologies in the research context. As for any social inquiry, it is essential to weave a network of contacts with local non-governmental organizations, academic institutions and civil associations, and to establish respectful relations with institutional authorities and traditional leaders, to ensure personal safety, access to relevant data, accurate analysis and results. Ethical concerns of social research are imperative in participatory visual methods: transparency, confidentiality, anonymity, and voluntary participation and withdrawal prevent potential harm and ensure respect for participants; it is essential to discuss sensitive issues such as goals and expectations, the implications of showing faces and identifying contested places, and with whom to share the outcomes, revisiting these issues regularly throughout the process. Truly participatory methodologies move from “informed consent” to “informed decision making” (Caldarola, 1985; Cervantes, 2019; Gubrium & Harper, 2016). Further levels of difficulty add to situations of conflict and suspicion, as institutional authorities tend to filter data, complexify bureaucracy, condition access and watch over researchers tackling sensitive issues. It is also a challenge to keep constancy in participants among displaced persons who may move further or return home. Attention must be paid to traumatic topics, underlying dynamics of power in group discussions, and potential consequences of collaborating in research — awareness, sensitivity, respect, and empathy are essential skills for participatory inquiries (Feijó, 2020; Idris, 2019).

Validity-wise, there is a risk of content selection bias in visual methods: what to show in a photo, to which scale to define a map, for how long to shoot a video, and the data collected and discarded with these choices. It is useful to mention that photographs, videos, and maps produced by participants are not more biased, and thus less valid, than oral or written data provided in interviews or questionnaires, and thus, to be considered accurate expressions of the participants’ realities. To minimize bias, visual data shall be triangulated with other methods and consolidated with observation and fieldnotes (Moore et al, 2008; Pauwels, 2019).

A final reflection shall be given to the author’s positionality in the research — in this case, a European female scholar aiming to critically study African cities based on endogenous scholarship; an outsider with inevitable cultural limitations and an unconscious privileged attitude, the result of an imperialist education; and stepping in a conflicted city where profit and injustice coexist. These confines are wittingly in deconstruction, for which a critical theoretical inquiry and a participatory methodology with local partnerships can contribute, consolidating the direct experience from previously living in the research context.

This section posited a link between conflict-induced urbanization and participatory visual methods for a critical urban inquiry; hereafter, it is translated into a comparative case study in Pemba, for which a pilot study was conducted in September 2022.

From Theory To Practice

A Comparative Case Study in a City Amidst Conflict

Pemba is a coastal city confined on the southern peninsula of its bay, a natural harbour surrounded by fertile plains (Fig. 1). Rising from sandy beaches to a central plateau southward, the present-day Municipality of Pemba includes thirteen neighbourhoods: Paquitequete, Cimento, Ingonane, Natite, and Cariacó, the oldest quarters in the city centre; Josina Machel, Eduardo Mondlane, Maringanha, and Gingone, the earliest suburbs; Mahate, Chuiba, Muxara, and Metula, the most recent periurban sprawl (Fig. 2).

Fig. 1
figure 1

The limits of Pemba city (in solid line) and its location on a peninsula within Pemba Bay (GoogleEarth)

Fig. 2
figure 2

Map of Pemba with its neighbourhoods (in solid lines) (Município de Pemba/Fundação E35 2021)

The scarce available literature about the history of Pemba and Cabo Delgado mentions four broad, sequential yet intertwined stages of development: the formation of the native ethnic cultures in this region within the Swahili world; the colonial occupation; the national liberation and the ensuing 16-Year War (the “civil war”); and the new millennium with the “oil and gas boom and bust”.

Medeiros (1997) traces the ethnology of Cabo Delgado in the mid-19c., with Swahili-Makhuwa and Mwani people on the islands and the coast from Palma to Pemba Bay; Makhwa-Mettho settlements in the low hinterland plains in present-day Ancuabe, Chiure, and Montepuez; and Makonde groups in the east inland plateaus of Mueda and Macomia. The Makhuwa-Mettho may descend from Bantu people who migrated from central Africa; they controlled the routes of the trade caravans from the hinterland to the coast and practiced agriculture. Their mix with Arabs, Persians, Hindus, and Malagasy arrived via the Indian Ocean may be the origin of the Mwani, who live mostly from sea-related activities; later, the Zanzibari Sultanate extended its influence on coastal present-day Mozambique with trade, religion, and cosmopolitanism until the I Great War. The KiMwani language relates closely to KiSwhaili and eMakhuwa with lexicon from ShiMakonde, Arabic and Persian. The Makonde may not have been, originally, a homogenous ethnicity, but a growing group of people who escaped throughout the centuries from the enslaved caravans and took refuge in the hard-to-reach, forest-dense plateaus, gradually consolidating a culture of warriors, hunters, and crafters.

Alvarinho (1992) states that Pemba Bay was visited since ancient times by Swahili and Malagasy fishers and traders. By the 19c. local chieftains had converted to Islam, adopted Swahili habits and ruled over semi-feudal societies: sultan Mugabo in present-day Ancuabe, Muaria in Montepuez, Said Ali in Mussange, and Macesse in Metuge. The Portuguese attempted to establish a colony in Pemba in 1857, then known as Pampira, near present-day Paquitequete, as an extension of Ibo island, then the colonial capital of Cabo Delgado and Niassa districts, under the “Overseas Province of Mozambique”; but it lasted only for 5 years and was eventually abandoned. Around 1880, clans from the Muaria dynasty moved towards the coast and established settlements throughout the peninsula, in places with available freshwater and access to the sea, as in present-day Paquitequete, Natite, Ingonane, Wimbi and Nanhimbe, Maringanha, Mahate, and Chuiba, founding the old cores of the contemporary neighborhoods.

Following the Conference of Berlin in 1885, the Portuguese government, lacking resources to effectively occupy the territories of “Portuguese East Africa”, transferred its administration to private companies with royal authorization to explore and develop; the then Cabo Delgado and Niassa districts came under the administration of the Niassa Company in 1891, based in Ibo island. Due to Pampira/Pemba’s favorable location as a port in the trade routes from the hinterland, the Company installed a new military post and commercial port in 1899, changing its name to Porto Amélia in honor of the Portuguese queen, then moved its headquarters from Ibo in 1902. In the decades after the town would grow and develop with the migration of Makhuwa-Mettho and Makonde from inland, Makhuwa and Mwani from the northern coast, European, Indians, Malagasy, and Comorian from the islands, forming the heterogenous population existing today; there were 18,604 registered inhabitants in 1908. The concession to the Company ended in 1929 with its dissolution. The new fascist dictatorship Portuguese government of “Estado Novo” established the Municipality of Porto Amélia in 1932 as the capital of Niassa Province; it was elevated to city status in 1958.

Months before the independence of Mozambique in 1975, the city became the capital of the new province of Cabo Delgado and Pres. Samora Machel changed its name to Pemba. With the departure of the colonial structures, the consolidated centres in the provincial capitals were virtually devoid of residents and then reoccupied by the freedom fighters and citizens related to FRELIMO (the Mozambican Liberation Front); many municipal roles were occupied by staff with no experience of urban management (Maloa, 2016). In Pemba, the unplanned neighborhoods in the suburbs expanded to receive migrants from the hinterland, mostly peasants bringing with them a rural and collective lifestyle to the city. Concurrently, the urban population grew as the city became a place of refuge for people fleeing the rural guerilla during the 16-Year War when RENAMO (the Mozambican National Resistance) reached Cabo Delgado in 1979; after the end of the war, rural populations seeking opportunities continued migrating. Registered residents numbered 43,000 in 1980, 84,897 in 1997, and 138,716 in 2007 (Instituto National de Estatistica). Rodrigues (2022) mentions that this growth was not followed, however, by relevant public investment or infrastructure development. Only by the early 2000s, with new touristic interest from private operators in the coast of Cabo Delgado, did the city of Pemba start to be the object of attention. The subsequent discovery of submarine reservoirs of liquefied natural gas off the coast of Cabo Delgado in 2009, the construction of the necessary infrastructure and the installation of associated services for their exploration originated a “boom” of private companies and human resources in Pemba from 2010 onwards, resulting in the consolidation of existing neighborhoods and informal expansion beyond the periurban limits. Then, a global international crisis related to “mega-projects” from 2014, allied to national scandals of corruption in 2015 led to critical withdrawals from foreign donors to the Mozambican state budget; in addition to the deterioration of the security due to the insurgency in Cabo Delgado starting in 2017, culminated in a standstill in the gas exploration, halting the dreams of wealth and development for the province and its capital.

The largest settlement in Cabo Delgado with 201,846 people in 2017 (Instituto Nacional de Estatística de Moçambique, Censo 2017), Pemba is a “secondary city” in Mozambique’s urban system, with less than 250,000 inhabitants; half of Africa’s urbanites live in these agglomerations, small municipalities which thus become crucial sites for improved living conditions, sustainable urban planning and regional development (UN-Habitat, 2022). Pemba has been the top destination for people fleeing the violence, especially since the insurgents’ takeover of Mocímboa da Praia in 2020; in November 2022, there were 220,835 displaced persons registered in the city — 21% of the total in the province (International Organization for Migrations, 2022); this is more than double the urban population in just 5 years, which raises questions about space and resources in a city confined to a peninsula.

Departing from this context, the research questions for this study relate to the dimensions of conflict-induced urbanization mentioned in the introduction of this article: what are the impacts of armed conflicts in the city of Pemba? What are the challenges and opportunities posed by these impacts on urban sustainability and governance? To answer critically, a comparative case study was designed in two levels (Fig. 3): a longitudinal comparison will be conducted between the impacts of the ongoing insurgency (identified through empirical data collected with participatory visual methods and interviews) and the impacts of the 16-Year War (via secondary data — archives and existing research); and a sectional comparison of these impacts in two selected neighbourhoods in the city. These impacts will be operationalized through analytical variables of the spatial, environmental, social, economic, and political features of urbanization, to be identified in each neighbourhood with participatory visual methods; the most salient will then be longitudinally and sectionally compared. Examples of these features identified in the reviewed literature are (spatial) morphology, (environmental) change, (social) heterogeneity, (economic) diversity, and (political) participation. Implications of the impacts on urban sustainability and governance will be theoretically transferred from existing literature and policy.

Fig. 3
figure 3

Conceptual diagram of the comparative case study designed for this inquiry (by the author)

As an inquiry about a city shall involve its inhabitants, residents and displaced persons will be selected in each neighbourhood by “snowball” sampling, departing from personal contacts, to decide upon and participate in the spatial and visual production and elicitation activities for data collection; semi-structured interviews will also be conducted with key informants in the municipality, local associations and humanitarian organizations. This means interacting with people with varied degrees of education, economic capabilities, social status, political power, cultural perceptions, religious beliefs, and personal agendas. Participatory photography and cartography can be suitable for this specific study for their simplicity of use and quick results with little preparation — using cellphone cameras to capture images and mobile maps to pin locations during walking tours, and then printed photos and maps for visual elicitation, the goal is to detect how Pemba is changing due to war, the impacts on daily life, desirable improvements, and the perception about state institutions. In situations where taking photos may be sensitive, maps can depict the identified phenomena. Video or audio recordings of conversations may be sensitive in this specific context and condition the participants’ disposition; thus, collection and elicitation activities will be consolidated with participant observation and field notes. A preparatory talk about goals, challenges, ethics, and techniques will take place in each neighborhood, and the semi-structured interviews with key informants will contribute to triangulating the visual data collected. Finally, to facilitate, enrich, and validate this inquiry in its participatory critical nature, a collaboration will be established with a local university, whose terms are described hereafter. The results of the study will first be validated with the participants, and then shared with municipal stakeholders, in academic outputs and the public, in an exhibition co-organized by the participants and the local university.

A Pilot study in the City of Pemba

This design for a case study about conflict-induced urbanization in Cabo Delgado was tested in an exploratory trip to Pemba in September 2022, preceded by meetings in Lisbon with experts in Mozambique, urban studies, displacement, and research methods. The fact that the context was already familiar to the author from previous direct experience eased the contacts, the preparations, and the actual fieldwork. The trip consisted of one week in Maputo for archival research and interviews, and 2 weeks in Pemba for methodological test. In Maputo, interviews were conducted in the Faculdade de Arquitectura e Planeamento Físico da Universidade Eduardo Mondlane, the research centre Observatório do Meio Rural, and the insurgency observatory Cabo Ligado. This allowed the author to collect secondary data about Pemba’s history; to be updated on the conflict, the movement of displaced persons and the ambiance in the city; to validate the idea about the neighbourhoods to select for the study; and to get more contacts.

In those interviews, it was stated that Pemba is a destination for people fleeing mainly from the coast of Cabo Delgado, while those inland tend to travel upcountry. This influx became notable after August 2020, following the attack on Mocímboa da Praia, and increased in 2021 and 2022. People come by sea or road, which determines the entry point in the city: vessels dock on the beach of Paquitequete in the north; vehicles enter from the south towards Muxara and Mahate. Further movement within the city follows family ties and acquaintances, leading to prevalent origins and ethnicities in different neighbourhoods.

Rapid population growth has several impacts: in consolidated areas, it leads to densification and overcrowding, construction in unsuitable areas, overload of already scarce infrastructures, increasing waste accumulation and sanitation issues. In periurban areas, empty plots without infrastructure are occupied with self-construction, increasing disorganized sprawl and related issues of waste and sanitation mismanagement. Socio-economic impacts are akin throughout the city: unemployment, aid dependency and the surge of informal activities for income generation, often desperate as criminality and prostitution. Overloaded public services fail to respond to all the needs, and public discontent arises due to the inability of the Municipality and the State to manage the worsening situation. These accounts were later corroborated in Pemba by residents, displaced persons, a historian, staff from international organizations, a reporter, the founder of a local association, a taxi driver and a foreign priest living there for 10 years.

Based on the personal previous knowledge about the city validated by these initial interviews, the selected neighbourhoods for this inquiry are Paquitequete and Mahate, for their analogous processes of change due to the conflict, but distinct morphology and historical development, which may lead to some similar and other contrasting data, hopefully enriching the sectional comparisons (Fig. 4).

Fig. 4
figure 4

Map of Pemba with the location of Paquitequete and Mahate neighbourhoods (dotted lines) (adapted from Município de Pemba/Fundação E35)

Facing the bay, Paquitequete is Pemba’s oldest settlement, emerged in Swahili times; crossed by the coastal road, it lays on a sand strip separated from the peninsula by a canal once covered by a mangrove. Historically it has remained outside of urban policy: densely occupied with winding, unpaved streets, many buildings are self-constructed and not serviced by supply networks. Descendants from the northern coast of Cabo Delgado, of ethnical majority Mwani and professing Islamic religion, its residents live mostly from the sea: fishers, craftworkers, and traders. A solidary community with a strong cultural identity, their interactions with the formal city are minimal: for work, retail shopping, and bureaucratic issues. As there are no “machambas” for subsistence farming within the neighbourhood, the historical informal trade of fresh produce and basic goods from the hinterland around the bay (Miéze, Metuge, Mussange) is vital for the household economy (Bruschi et al, 2005; Pereira, 2012; Del Bianco, 2020)

Paquitequete presents interesting features for this inquiry: spatially dense and consolidated, it allows to question where newcomers find shelter, how public services and the environment are impacted; ethnic-culturally homogeneous, incoming outsiders change the social fabric; increasing consumer demand and available labour affects the local economy; and it allows to inquire how hosts and guests perceive the state and their own participation in urban governance (Fig. 5).

Fig. 5
figure 5

Paquitequete (solid line) with the coastal road (dashed line) and the canal (dotted line) (adapted from GoogleEarth)

Mahate is a diverse context. No secondary data was found about this neighbourhood; hence these accounts result solely from the conducted interviews. It was the first periphery of Pemba, when people fleeing the 16-Year War settled near the airport, at the time just outside the city. Eventually absorbed by the urban fabric, it became a municipal neighbourhood. The main road to the city splits it in “Mahate-bay”, the old core westward, mostly equipped with infrastructure, a health centre, a market and schools; here a Catholic Mission was installed in the 1940s. To the east, “Mahate-sea”, the newest part with seaview, recently urbanized with many plots not yet serviced by networks. In the limit of Mahate with Muxara is Pemba bus station, a vital place for the arrival of people, fresh produce and processed goods. Ethnic-culturally descendants of Makhwua from the southern hinterland of the province, Mahate’s residents are mainly farmers, traders and craftworkers, and predominantly Muslims.

Mahate contrasts with Paquitequete in its urban features: a sprawled territory with low density, developed more recently and in a short period, socially heterogeneous and economically diverse. In common there is a deficiency in infrastructure, self-construction, absence of “machambas”, and a sense of disconnection from the formal State. These variances and similarities between the neighbourhoods will hopefully enrich the comparison of the impacts of armed conflicts in the city of Pemba (Fig. 6).

Fig. 6
figure 6

Mahate (solid line) with the airport (to the north) and the main road (dashed line) (adapted from GoogleEarth)

A research partnership was launched with the Universidade Católica de Moçambique (UCM) in Pemba. Two students were chosen from the Graduate Program in Environmental Management, by the criteria of (1) being from Pemba and speaking local languages, (2) interest to collaborate, and (3) planning to choose related graduation topics. This teamwork will enable local knowledge and translation for the author and support the students in their monograph research and writing. The partnership also includes lectures by the author in the university in a following visit, and the co-organization of the final exhibition at the end of the project.

The activities of this pilot study in each neighbourhood consisted of walking tours with key informants and focus groups with long-time residents and recently arrived displaced persons selected through “snowball sampling”. The goals were (1) to know the participants’ circumstances, (2) to identify important places in the neighbourhood and (3) changes happening due to the conflict; (4) to assess their disposition to partake in the study and (5) the suitability of participatory photography and cartography.

In Paquitequete, two walking tours were taken on consecutive mornings, with the co-founder of a local association (resident in the neighbourhood) and a national staff from a humanitarian organization (no longer resident but born and raised there). These tours were crucial for the author, an outsider, to get a deeper sense of the place beyond the previous knowledge of having lived in Pemba. Information was shared about urban life and the conflict; the appropriateness of taking pictures with cell phones and using maps was discussed, with both techniques validated due to their participatory nature. The week after a focus group took place with ten women. The idea to select only female participants was suggested by the two entry contacts, to enable an open talk as opposed to a mixed-gender group, due to cultural, religious, and gender dynamics. Six participants were long-time residents and four displaced, all identified as Mwani and Muslim. The moderator was the co-founder of the local association, supported by a colleague from the humanitarian organization and the students from UCM, and the talk was held in KiMwani.

Two residents were born and raised in Paquitequete; the others were from Mussange, Quissanga, and Ibo, and arrived young in the 1990s and 2000s. Their main occupations are childcare and the informal trade of cooked food, charcoal, and fresh produce. Among the displaced, one fled from Quissanga during its first attack in 2020, and three escaped from Mocímboa da Praia in 2021 during the town’s occupation by the insurgents. All arrived by boat to Paquitequete. Back home, they lived from informal trade, mostly food. Here they struggle to find shelter and generate income, living from the neighbours’ charity, the support of local associations, and sporadic vouchers from humanitarian organizations; there is alleged corruption among local authorities in the registration of displaced persons and distribution of these vouchers. Inquired as to why remain in the city instead of the resettlement camps in the outskirts, all referred to insalubrious conditions and deficient food assistance; in “Paquite” they find solidarity and can explore ways to generate income — “Some neighbour always helps”. All expressed the intention to return home, but none knew when.

During the talk several important places in the neighbourhood were mentioned: the old mangrove, a natural leisurely area destroyed for the construction of the canal, now covered with stagnant water and waste; the football pitch, an unpaved field at the entry of the neighbourhood, facing the oldest mosque; a square with an old fountain (out of order for many years), popular to buy breakfast; the beach, where fishers arrive in their boats and sell directly to customers and sellers, where boats depart to and land from other coastal towns; the old mill, where women used to work, providing autonomy and conviviality, but “unfortunately” recently closed.

When inquired about observable changes due to the war, the residents referred to an increase in population and social heterogeneity, with many new people and the distrust that stems from unfamiliarity — “Before we were all cousins, now we don’t know each other”; still, they are solidary with newcomers — “We are all Mwani people”. There is densification and overcrowding, with many inhabitants in each house due to the absence of space for new construction; increased demand for basic goods, subsequent shortages, and rising prices; increased demand for water supply and longer waiting times in public fountains; accumulation of unmanaged waste, an old problem made worse; and increased competition over informal economic activities, as more labour is available. Despite this, there does not seem to be an increase in criminality or feelings of insecurity — “We sleep outside, in our courtyards”.

The last question regarded their interest and availability for this study, taking photos with their cellphones and discussing printed photos and maps in the next visit. Not all of them have smartphones but all gave positive feedback and expressed interest.

Due to the unfeasibility of arranging a male focus group because of time constraints, a fisherman from Quissanga was interviewed to counterpart the female accounts. He fled in 2021, losing his house and boat. In “Paquite” he stays in his daughter’s house, where once lived seventeen people. Without tools he cannot fish independently, and it is difficult to work for others due to competition — “There are many fishers and few fish in Pemba”. Aid vouchers are scarce and wrapped in corruption and power abuses.

These pilot activities delivered satisfactory results: the initial contacts were useful entry points and knowledgeable contributors for a deeper sense of the place; the use of online maps for location pinning was positive (Fig. 7); the focus group was fruitful, with spontaneous participation; the use of cellphones for photos was tested (Fig. 8) but not the visual elicitation; yet, the participants’ feedback consolidated with observation validated this methodology for extended data collection. In the next visit, individual chats shall be prioritized, focus groups arranged by gender, and maps and photos must be printed for visual elicitation. Separate interviews shall be conducted with authorities, religious leaders and health and education staff.

Fig. 7
figure 7

Online map with locations in Paquitequete (pins with flags) (adapted from GoogleEarth)

Fig. 8
figure 8

Assemblage of photos from Paquitequete as taken by a participant: the football pitch with the oldest mosque, the beach, the health centre, and the square of the old fountain

In Mahate it was a contrasting experience: due to the extension, the scouting tour was done by car with a national staff of a humanitarian organization, who, although not native or resident, knows the neighbourhood well because of their work; it was possible to get a deeper sense of the urban sprawl and the variations between both sides of the main road.

The focus group took place the week after, with five participants of each gender and from different parts of Mahate, selected through “snowball sampling” departing from the staff of the Catholic Mission. Three were native/long-time residents, six had recently arrived, and one was a native — the moderator — who moved young to Mocímboa da Praia and returned after fleeing the insurgency in August 2020. The talk was held in eMakhwua. All natives identified as Makhuwa; two work in the Mission, the third takes care of children, and the moderator was a lab technician in Mocímboa, now working in Pemba Provincial Hospital. All newcomers were displaced from Mocímboa da Praia. Two Makonde fisher-farmers came in 2019 following family links; in Pemba, they do not fish but farm a plot near Miéze for consumption and surplus trade. The other four identify as Makhuwa and arrived in 2020. Two fisher traders are unemployed, stating the difficulty of working for others in Pemba; a carpenter refers to the same struggle and has no funds to restart his own business; one was an administrative officer in Mocímboa Secondary School, now assisting displaced students in Pemba Secondary School — “We are civil servants in exile”.

Significant places referred to were the schools, where students from displaced families are integrated; the health centre, recently renovated by a humanitarian organization to accommodate the increasing number of patients; the market building, where fresh produce and basic goods are sold; the public fountains, where many inhabitants access water; the mosques and the Catholic Mission, two faces of religion and assistance; and the beach, although a bit far from the neighbourhood, which is an important place for leisure and subsistence fishing — the displaced carpenter declared “I like to go there to get some fish and look at the sea”.

References were also made to recent observable changes: the rapid sprawl in an already large neighbourhood, precarious self-construction in empty plots without infrastructure, shortage of basic goods and rise in prices; but also, a new school built by the Municipality, the inclusion of students in the education system and the general acceptance of newcomers — “Education is very important for the children, also for the ‘refugee’ children” (displaced persons are often mentioned as “refugees” by the residents).

This talk was challenging: participants were suspicious and responded only when interpellated, without spontaneity; men spoke more confidently while most women remained silent. It was not possible to test visual elicitation but after the discussion locations were pinned and photos were taken with a cellphone in a walking tour (Figs. 9 and 10). Participants of both genders expressed interest in this study despite the general closed attitude; one stated “We need to have our struggles known outside of Cabo Delgado”. Contextual information on this attitude was given later by a key informant: radical groups in Mahate, connected to the insurgency, feed suspicion among neighbours and families; local authorities control the distribution of humanitarian vouchers, deciding “who eats and who doesn’t”; people fear retaliation. But that is exactly why they deserve to have their stories conveyed safely, which this methodology can do, as photos of places and maps seem to be useful where words may harm. Hence, for extended data collection, it is advised to allow time for personal interactions, to prioritize individual interviews and walking tours over focus groups, to avoid conversations of political matter and to simplify visual elicitation by printing the produced photos and maps. Separate interviews shall be conducted with state and religious authorities and health and education staff.

Fig. 9
figure 9

Online map with locations in Mahate (pins with flags) (adapted from GoogleEarth)

Fig. 10
figure 10

Assemblage of photos from Mahate as taken by a participant: the elementary school, the market, and the health centre

The time in Pemba allowed also for general observation and comparison with the previous time when the author lived there: visible military presence in strong demonstrations such as walking patrols and low helicopter flights over the city; many foreigners and vehicles from international organizations and security companies; new high-end restaurants, leisure, and entertainment venues frequented by both Mozambicans and foreigners; new accommodation buildings in premium locations; and more available touristic activities such as boat rides, diving, and craft trade. Inversely, visibly unoccupied persons gather on street corners, squares, and along roads; more street vendors; long waiting lines in banks; and children sellers and beggars. According to a taxi driver, prices have been rising since 2019. These may be unsurprising expressions of a war industry — the rise of humanitarianism, the emergence of a local middle class, the consolidation of regional elites, and the impoverishment of low-income urbanites — to be asserted in the extended data collection. Armed conflict deepens existing challenges of inequality, urban management and governance in Pemba.

Conclusion

This article presented the theoretical framework for an inquiry about conflict-induced urbanization in Cabo Delgado, positing that this phenomenon can be critically analysed through a participatory visual methodology, to reveal the impacts of armed conflicts in the city of Pemba from the perspectives of its diverse urban actors. The first section presented a contemporary critique of urban studies, its translation into methodologies suitable to research cities in conflict, and reflections on challenges, ethics, and positionality in research. The second section proposed a design for a comparative case study with participatory photography and cartography, interviews, observation, and archival research. Lastly, it described the pilot study conducted in Pemba to test this methodology. A partnership was established with the Universidade Católica de Moçambique; Paquitequete and Mahate were the selected neighbourhoods, for their analogous processes of change due to the conflict, but distinct morphology and history; in both were conducted walking tours and interviews with key informants and focus groups with residents and displaced persons.

In Paquitequete, the focus group consisted of six residents and four displaced persons, who arrived in 2020 and 2021. They prefer to stay in the city instead of going to the resettlement camps in the surroundings because of insalubrious conditions and aid dependency, even if they struggle to generate income and face corruption among local authorities in the distribution of food assistance. The residents mentioned the most observable changes happening in the neighbourhood due to the war: population growth and social heterogeneity, household densification, increased demand for basic goods and rising prices; increased demand for water supply and accumulation of unmanaged waste; increased competition over informal economic activities, as more labour is available. The activities produced good results even if some techniques were not tested, and the feedback from participants, consolidated with participant observation, validated the methodology for extended data collection. In Mahate, the focus group was challenging, as participants were afraid to speak for fear of retaliation from local authorities. But it is exactly for this reason that they deserve to have their stories told; and despite this sensitivity, photos of places and maps seem to be innocuous, so this methodology can succeed. The group consisted of four residents and six displaced persons, arrived throughout 2019 and 2020. Those with farming skills managed to find machambas to farm in the outskirts of the city, but fishers and craftworkers struggle to generate income. Displaced civil servants continue to perform their duties “in exile”. The residents referred the most observable changes in the neighbourhood due to the ongoing conflict: rapid urban sprawl, precarious self-construction and deficient infrastructure, shortage of goods and rising prices; but also, the inclusion of students in the education system and the general acceptance of newcomers. In parallel, general observation in the city allowed to notice a developing war industry, manifesting in increased military and humanitarian presence, and new related infrastructures; it was also evident the upsurge of unoccupied people and informal activities. Armed conflict deepens existing challenges of inequality, urban management and governance in Pemba.

For extended data collection, it will be useful to have interviews and walking tours before focus groups, which shall be arranged by gender, and to avoid political conversations; separate interviews shall be conducted with local authorities, religious leaders, and health and education staff. The results will then be validated with the participants, presented to the Municipality and publicly shared in an exhibition co-organized with the university. With this inquiry, the author aims to contribute to engaged urban studies in Mozambique and Portugal, and to transform traumas of war into opportunities for sustainable development and prosperity.