Official Risks and Everyday Disasters: the Interplay of Riskscapes in Two Unplanned Settlements in Monrovia

Monrovia’s unplanned settlements provide a window into the diverse rationales and practices that go into planning and managing urban space. Even though unplanned settlements are economically, spatially, and socially integrated into cities, the desire to eliminate or formalise them persists. In the essay, I examine how everyday risks are addressed in the coexistence and co-evolution of planning and unplanned urbanisation in postwar contexts. As well as showing how expert-recommended risk management approaches cross over into local practices, I explain how and why multiple understandings of ‘risk’ and ‘resilience’ can influence practice. The unequal power dynamic suggests that the practices of state actors influence the residents of unplanned settlements, in a manner that is discernible of governmentality. As riskscapes, and by extension resilience, represent multiple realities to stakeholders, the question of coalescing the mosaic of practices into a common risk governance framework is critical. The essay emphasises that resilience-building should evolve from a thorough understanding of the dynamics of the multiplicity of riskscapes. Finally, the paper argues that an evolutionary approach to risk governance, paying attention to the interacting elements and configurations that link discourse, actors, institutions, power, and knowledge, will provide a platform for negotiating the links between risk perception and risk assessment within the emerging riskscapes. This will be the basis of a deliberative and negotiated resilience pathway that will integrate the priorities and interests of all stakeholders in the planning and decision-making process.


Introduction
Koffa recounted the incident that prompted his relocation to Peace Island, one of the ever-growing unplanned settlements in Monrovia. 1 Previously, he had lived in a small cement house he had built on 24th Street in the Sinkor suburb for 12 years. In May 2012, he watched helplessly as his house, along with other surrounding buildings, was destroyed by bulldozers of the city's government, the Monrovia City Council (MCC). He admits that he did not have the right to build on the land, but argues that the affected families were not notified beforehand. Later, Koffa moved to Peace Island, where he acquired land from another settler and immediately built a mudbrick house for his family. His hasty move to Peace Island had a knock-on effect. He was now far away from the workshop where he had once worked as a mechanic.
Small unplanned settlements, like this one in Sinkor, found throughout Monrovia on formerly unused land, are regularly demolished, while the larger and most prominent ones, such as West Point and Peace Island, home to tens of thousands of people, remain. A situation such as this illustrates the risks associated with evictions and demolitions of smaller unplanned settlements, as well as the reasons why larger unplanned settlements continue to proliferate.
The positioning of unplanned settlements as blights is rooted in the modernist thinking prevalent in urban planning in sub-Saharan Africa (Fält, 2016;Hoffmann, 2016;Kamete, 2013;Watson, 2009). In 2013, for example, in preparation for the United Nations High-Level Panel (HLP) meeting on the Post-2015 Development Agenda, the MCC launched another campaign to remove informal structures in Sinkor to eliminate eyesores. In response to the criticisms that followed, the city mayor is reported to have said 'We want to make this city the greenest and cleanest city in West Africa'. 2 Despite their similar features, unplanned settlements have a vast array of morphologies, from districts to waterfronts and from easements to enclosures (Dovey & King, 2011;Kamalipour & Dovey, 2020). These settlements are not marginal to urban life and productivity. Rather, they are economically, spatially, and socially integrated into most developing cities, and yet the desire remains to remove them (Dovey & King, 2011). Proximity to opportunities is a key driver for the emergence of these areas near or within cities. Their residents serve the formal city, where they often form a large part of the labour force. Attempts to move these settlements to the periphery, therefore, exacerbate poverty and harm the productivity of the city (Dovey & King, 2011).
The current development patterns of sub-Saharan African cities are characterised by the morphological juxtaposition of visionary modern mega-projects and the persistent unplanned urbanisation (Fält, 2016;Watson, 2014). The notion that unplanned settlements represent urban disorder is therefore at odds with the reality of many African cities, where informal processes are often the main driver of urban development (Myers, 2011;Pieterse, 2011). The gradual, unauthorised, and selforganised emergence of these settlements has become the primary method for providing affordable housing in developing cities .
Unplanned urbanisation does not necessarily mean disorganisation. However, the patterns tend to be uniquely place-specific and in some cases reflect traditional urban planning. Indigenous planning models, mostly 'unplanned' but organised, were deeply rooted before the emergence of African colonial cities (Amankwah-Ayeh, 1996;Okpala, 2009). Some elements of traditional African planning, particularly 'courtyard architecture', remain a key feature of unplanned urbanisation addressing the housing crisis in some African cities (Eguavoen, 2021(Eguavoen, , 2022Steyn, 2005).
The essay examines how everyday risks are addressed and accounted for as visionary modern planning and unplanned urbanisation coexist and co-evolve in postwar environments. To this end, it examines the interplay of riskscapes of unplanned settlements in Monrovia, focusing on the extent to which expert practices guide residents' conduct, and shape everyday risks in unplanned settlements. Furthermore, it raises the question of how the multiplicity of risks and priorities constructed by stakeholders can be integrated by authorities into ongoing risk reduction planning and policy. This paper argues that discussions on riskscapes should foreground a more holistic approach to resilience, recognising its multiplicity and emphasising a negotiated path to its realisation.
The rest of the paper is divided into seven parts. It begins with the background to the study, followed by the main theoretical thrusts. It then discusses the characteristics of the study areas and the methods of data collection and analysis. Three empirical sections follow, the first on the riskscapes of unplanned settlements, the second on governance and institutional responses to riskscapes, and the third on how residents navigate risk and opportunity landscapes. The last section contains the main conclusions.

Background
Although Liberia was not formally colonised, the development of Monrovia, beginning in 1822, followed a segregated urban planning pattern similar to colonial African cities (Amoako, 2016;Hoffmann, 2017;Home, 2015). 3 As a result, infrastructure development was concentrated in the parts of the city historically associated with the settler population known as Americo-Liberians, 4 while the local population remained on the fringes of the city in the largely unplanned slum-like communities.
3 Whether Liberia was colonised or not is disputed. Unlike other colonies in Africa that were controlled by Western powers, Liberia was founded by an organisation, the American Colonisation Society (ACS), created in 1816 with financial support from the US Congress to help repatriate freed blacks from the USA to Africa. 4 Collective term for the resettled peoples of African descent from North America and the Caribbean.
It is estimated that up to 70% of Monrovia residents currently live in an unplanned neighbourhood (Draper et al., 2018).
The exclusive planning and land rights regimes created the basis for informality to flourish. It took several decades for areas outside direct settler control to be integrated into the Liberian state after its independence in 1847 (Ellis, 2007;Gerdes, 2013;Verbrugge et al., 2015). Initially, state policy only recognised customary usufruct of land. Fee simple property rights were reserved for Americo-Liberians and acculturated groups. Planning in Monrovia also prevented the permanent settlement of the cheap labour force, which consisted largely of indigenous groups. This led to the emergence of a dark form of 'temporary urbanism' in which unplanned settlements served to provide access to labour for elites while denying workers permanent access to the city through adequate housing or land ownership. 5 Liberia is reeling from decades of civil unrest. More than a century of one-party rule dominated by the Americo-Liberian elites ended with the 1980 coup d'etat (Casper, 2011;D. Harris, 1999;Munive Rincon, 2010). Attempts to restore democracy were thwarted by the rigged 1985 elections, which had cascading effects, including a failed coup attempt in 1985 and an escalation to all-out war in 1989 (Gershoni, 1996;Kieh, 2009). The first civil war ended with the election of the main rebel leader, Charles Taylor, as president in 1997. By 1999, Taylor's rule was contested by two rebel groups, leading to the second Liberian civil war, which ended with Taylor's resignation in 2003.
The civil wars claimed the lives of about 250,000 Liberians and more than 500,000 were displaced (Johnson-Sirleaf, 2011). People were uprooted and cities, villages and rural areas were almost completely abandoned. Many flocked to Monrovia seeking shelter and settling in camps, abandoned buildings and other vacant spaces (Hoffmann, 2016;Williams, 2011). 6 Many of the displaced people who fled to Monrovia during the war still live there today. The population, which was about 600,000 before 1989, grew to about 1.2 million in 2008 during the last census (LIS-GIS, 2008). The estimated population of Monrovia for 2020 is 1.52 million. 7 There are no official current figures available, as there has been no census since 2008.
The return to peace in 2005, characterised by stability in the political and economic space, coincided with instability in the physical space for the urban poor, especially internally displaced persons (IDPs). The influx of capital into the country by Liberian returnees and international donors led to massive infrastructure development projects. Beginning in 2006, the MCC pursued an aggressive slum clearance policy and took strong measures against other forms of informality, such as street vending, but since then the approach has become more cautious (Hoffmann, 2016;Lupick, 2012;Weeks, 2012;Williams, 2011). Many IDPs were evicted from private and public properties they had occupied, while refugee camps were closed, 5 Interview with Human Rights Lawyer, Monrovia, January 6, 2020. 6 Monrovia was controlled by peacekeeping forces from the Economic Community of West African States (ECOWAS) between 1990-1997 while the rest of the country was in the hands of the warring factions. 7 World Bank (2020). https:// data. world bank. org/ indic ator/ EN. URB. LCTY? locat ions= LR prompting many to move to already overcrowded unplanned settlements such as West Point or to build new settlements such as Peace Island (Fagen, 2011;Williams, 2011).
As in many West African countries, Liberia's land tenure system is characterised by legal pluralism, which in many cases limits the scope for action of local governments. The fact that statutory and customary land tenure systems operate according to different and often mutually exclusive procedures, but are supposed to co-exist within the same land use framework, often leads to gaps, violations and uncertainty, as Frimpong-Boamah and Walker (2016) have shown in Ghana. In waterfront areas, land tenure can be even more legally fragmented, leading to multiple legal interpretations by actors, forum shopping, land conflicts, fraud and the loss of public space, as Eguavoen (2022) demonstrated for Abidjan.
The Liberia Land Rights Policy (2013) defines four land rights categories: Public Land, Government Land, Customary Land, and Private Land. 8 Government land is owned and used for government activities, public land is used and managed for the public good, customary land is owned by local communities and used or managed according to customary practices and norms, and private land is owned by individuals or private companies whose management and use decisions are secured by legal titles issued by the Ministry of Lands, Mines and Energy. The Liberian Land Authority (LLA) was established in 2016 as an autonomous body to assume the fragmented land functions previously spread across various government agencies, including the Ministry of Lands, Mines and Energy (MLME), the Centre for National Documentation Agency (CNDRA) and the County Land Commissioners of the Ministry of Internal Affairs.

Conceptualising Riskscapes of Unplanned Space
There is a growing awareness of the need to understand and address risks associated with the specificities of urbanisation in Africa. In comparison to the growing body of literature on specific risks associated with urban growth, particularly flooding (e.g. Aalders, 2018;Amoako, 2016;Amoako & Inkoom, 2018;Frick-Trzebitzky et al., 2017), there are relatively few studies on how the specificities of African urbanisation address and account for everyday risks (Adelekan, 2020;Adelekan et al., 2015;Dodman et al., 2017).
The rapid unplanned urbanisation in sub-Saharan Africa is increasingly associated with cycles of risk accumulation (Allen et al., 2017;Pelling et al., 2018). Urban risks cover a wide spectrum, ranging from everyday risks to smaller recurrent risks such as floods, to large and significant but generally rare hazards such as earthquakes (Adelekan et al., 2015;Bull-Kamanga et al., 2003;Manda & Wanda, 2017;Satterthwaite & Bartlett, 2017). The concretisation and reproduction of vulnerability through everyday risks, combined with an eroded capacity to respond, is a common feature of life in unplanned settlements in Africa. This is because vulnerable groups are constantly exposed to everyday risks at home, at work and in the community, including environmental pollution, crime and violence, lack of access to healthcare, food poisoning and heatwaves, psychological stress, poor housing conditions and the constant fear of eviction (Adelekan, 2020;Bull-Kamanga et al., 2003;Fraser et al., 2017;Ramalho, 2020;Satterthwaite & Bartlett, 2017;Songsore, 2017;van Voorst et al., 2015;Zerbo et al., 2020;Ziervogel et al., 2017;. Understanding risk accumulation is important because most risks in urban Africa are not visible and catastrophic or even episodic disasters; instead, they are everyday events that shape the lives of poor urban dwellers (Allen et al., 2017;Pelling et al., 2018;Satterthwaite & Bartlett, 2017). Unplanned spaces are shaped by different material configurations, relationships, practices and conditions, thereby creating multiple constructions of risk. According to Mol (2002), there are not only many ways to perceive a phenomenon, but also many ways to practice or enact it, resulting in different versions of the phenomenon. The notion of riskscapes emphasises the multiplicity, fluidity and overlapping nature of risks, arguing that different conceptualisations of risk can be applied to the same space with different conclusions as risks are dynamic, uncertain and contested (Aalders, 2018;Gebreyes & Theodory, 2018;Lundgren, 2018;Müller-Mahn & Everts, 2013;White & Lawrence, 2020). Furthermore, the concept considers risk as a multi-layered and intersecting assemblage that encompasses material and discursive aspects of risk (Neisser & Runkel, 2017). Riskscapes allow for a multidimensional analysis of landscapes of risk and opportunity in terms of material risks and the modes in which groups of actors interpret, experience, communicate, construct and respond to these risks (Müller-Mahn et al., 2018).
When considering the multiplicity of the riskscapes, it is important to note that reality is multiple (Law & Singleton, 2014;Mol, 2002). Just as riskscapes are multiple (Müller-Mahn et al., 2018), so too is resilience (Simon & Randalls, 2016). Following Annemarie Mol's (2002) concept of the 'body multiple', Simon and Randalls (2016) propose the term 'resilience multiple', as there is more than one resilience, but not a fragmented 'many', as resilience is realisable in multiple ways but remains singular and actionable through certain processes. The multiplicity of reality embodied in riskscapes and resilience raises the question of how to integrate them into a common governance framework. Blaser (2009Blaser ( , 2014 works on multiple ontologies is also instructive in understanding the 'multiplicity' of 'riskscapes' and 'resilience multiple'. Given the ontological multiplicity of riskscapes (Frick-Trzebitzky et al., 2017;Müller-Mahn et al., 2018;White & Lawrence, 2020) and resilience (Fraser et al., 2017;Harris et al., 2018;Simon & Randalls, 2016), adaptation practices require compromises and trade-offs for stakeholders (Harris et al., 2018). However, planning and practices can crystallise dominant political agendas (Flyvbjerg, 1996;Muchadenyika & Williams, 2017;Simon, 2015;Yiftachel, 1994) or exacerbate social exclusion (Abubakar & Doan, 2017;Parnell & Pieterse, 2016;Watson, 2009). Consequently, recent debates have increasingly favoured a more equity-oriented approach to resilience planning that emphasises inclusive and participatory processes, including those of local voices and marginalised groups, and that is receptive to different and non-traditional framings of resilience (Anguelovski et al., 2016;Harris et al., 2018;Krüger, 2019;Meerow & Newell, 2019;Vale, 2014;Ziervogel et al., 2017). This approach ensures that resilience initiatives do not systematically reinforce the status quo (Agyeman, 2013;Anguelovski et al., 2016;Harris et al., 2018;Meerow & Newell, 2019). This negotiated perspective draws attention to the contextuality of risk and resilience as phenomena that are constantly experienced and shaped in different contexts (Harris et al., 2018).
Riskscapes offer a nuanced understanding of urban risk, particularly regarding how people translate risks embedded in physical landscapes and discourses into everyday practices. In addition, it provides an insightful way to explore how practices are balanced in the governance of urban risk, ranging from expert assessments and their recommended governance measures to local perceptions and their approaches to risk management. Third, it provides insight into why certain risks are emphasised within the dominant practices and others are 'neglected'. Finally, applying the concept of riskscapes allows us to address the ontological multiplicity of the notion of resilience concerning informal spaces, particularly in terms of how risk reduction is enacted in practice by different stakeholders, thus highlighting the need for a negotiated resilience (Harris et al., 2018;Simon & Randalls, 2016;Ziervogel et al., 2017).

Study Area and Methodology
Two settlements were selected in Monrovia: West Point and Peace Island. With an estimated population of over 80,000, West Point is Liberia's most populous slum. The settlement emerged in the 1940s when fishermen settled on the sandy peninsula reclaimed from the Atlantic Ocean. West Point is recognised as a township, while Peace Island has no official status. Therefore, West Point has the presence of the state, as evidenced by the appointment of a township commissioner, a magistrate's court and a police station. Peace Island, located on a peninsula jutting into the Mesurado Wetlands, an endangered mangrove forest and Ramsar protected area, was uninhabited before the civil war. 9 The settlement was founded in 2005 by IDPs who had previously occupied the Ministry of Defence building the Congo Town suburb and has a significant proportion of former combatants. Both settlements are considered squatter settlements located on public land (Fig. 1).
The structures in the two settlements are mixed. In West Point, there are few cement houses located further from the beach, but most are built of zinc sheets and recycled materials. These materials can be easily reassembled after seasonal flooding or when the sea recedes. Buildings on Peace Island are more diverse, ranging from cement to mud bricks to zinc huts. The dynamics of development on Peace Island show that while the number of buildings at risk of flooding is increasing, the number of concrete buildings, which are less at risk of leaching, is also increasing (Fig. 2). However, the rapid spatial sprawl and increasing density of the area means that more people are exposed to additional risks.
Data were collected in Monrovia, Liberia, from August 2019 to March 2020 and drawn from five main sources: semi-structured interviews with government officials and members of non-governmental organisations (NGOs) and civil society organisations (CSOs); informal interviews with residents; focus group discussions (FGDs); review of official documents and grey literature; and field notes from direct observations (Table 1).
The interviews and FGDs were recorded and then transcribed. Interviews with government officials, NGOs and CSOs were conducted in English. The FGDs and interviews with residents were conducted in Koloqua (also known as Liberian Pidgin English), and the transcriptions were validated by local language experts from the University of Liberia's English Department.
To obtain sound and objective data/information on the diversity of the urban riskscapes, 15 officials from 12 government agencies, NGOs and civil  society organisations were purposively pre-selected, while the rest (10 officials from 9 organisations) were selected through snowballing. Participants for the FGDs and interviews were recruited primarily through a snowballing process that started with a small group that was purposively selected with the help of community leaders, the research assistant and others known to the researcher. The initial respondents helped to identify additional respondents (Table 2). The interviews with institutions took place at the premises of the respondents. Interviews with residents took place in the community, either in the participants' homes or in a central area. Individual interviews generally lasted 60 min, while FGDs averaged 90 min. The observations were non-intrusive and non-participatory and included observation of activities in public spaces, spatial features and informal interactions with residents.
Four gender-segregated FGDs were conducted with eight participants each: two were exclusively male and two were exclusively female. As the discussions on 'everyday risks' included culturally sensitive topics such as women's access to primary health care, it was decided to conduct separate FGDs so that women could speak freely. Additionally, it ensured the inclusivity and representativeness of women's voices and avoided the 'peacock effect' where men tend to dominate mixed-gender groups (Krueger & Casey, 2000;Umaña-Taylor & Bámaca, 2004). This group dynamic created an 'equal' and 'friendly' atmosphere where the research benefited from participants' shared experiences while avoiding hierarchy and power relations influencing the discussion. Participants related each other's statements to real events from their everyday lives and clarified the contradictions between their actions and perceptions (e.g. 'Didn't you ignore the community leaders' instructions not to put your container (kiosk) in the place designated for the hand pump?'). 10 The measures taken to protect and control information obtained from or about the participants and the institutions they represented included encryption or password protection of all storage devices to prevent the risk of data loss and leakage. Participants were informed of their right to withdraw their participation at any time. All personal data was fully anonymised. The data were analysed using Atlas.ti, mainly through content analysis.

Riskscapes of Unplanned Settlements
Riskscapes are socially constructed phenomena. While riskscapes are the confluence of actors' practices, one can observe the shades of objective riskscapes and subjective riskscapes, especially in terms of how risk is perceived (Table 3). These nuances are generally the result of the interplay of experts' risk assessments, which dominate the discourse on urban risk, and residents' risk perceptions. Although expert assessments for several unplanned settlements in Liberia have consistently estimated higher risks associated with conditions such as coastal erosion and flooding, residents' perceptions of risk are an ongoing process of construction and reconstruction shaped by their daily experiences. 11 A prism through which to explore practices at the local level is the recommended actions of experts and the risk management actions of residents. West Point, for example, has a community-wide sea erosion prevention committee that collects resources and solicits the interventions of the city government. 12 But while residents seek solutions for coastal defence, infrastructure and other basic services, the government advocates dismantling the settlement, mainly because of the threat of sea erosion. This position is summarised in the following comment from a former official of the Ministry of Public Works: It has always been the desire of the government to relocate West Point. Time and circumstances are forcing the government closer and closer to this decision. This settlement cannot continue to exist. (October 28, 2019) A West Point leader explained that they are habituated to the challenges of cyclical displacement due to flooding. 13 Floods can last up to 12 weeks at certain times of the year, but residents live daily with the effects of inadequate infrastructures, such as lack of sanitation and clean drinking water, as well as extreme poverty and violent crime. For example, energy scarcity increases the risk of fire outbreaks when candles, paraffin and portable generators are used in congested environments where buildings are constructed of combustible materials, or of food poisoning because many people cannot store food properly. 14 Respondents also noted that non-functioning sewers and storm drains, due to their conversion to landfills and construction sites, exacerbate the flooding problem, underlining the link between land use and everyday risks. 15 West Point is generally considered a non-viable settlement mainly because of the threat of coastal erosion and increasing severity of the seasonal flooding, and the interaction between these hazards with socio-demographic factors such as widespread poverty, vulnerability to epidemics and the prevalence of crime. 16 For example, the Monrovia Metropolitan Climate Resilience Project, funded by the Green Climate Fund to improve coastal protection in parts of Monrovia, is an example of this focus on physical geography. This points to the institutional context in which expert responses to risk often take place. While this line of thinking understandably helps to promote protectionist measures that translate into the demolition or upgrading or unplanned settlements, it does not necessarily resonate with the local population. The reaction of an official of the Environmental Protection Agency (EPA) shed some light on the matter when he remarked: People need to understand that we do not just wake up and make pronouncements. All the decisions we have made on these squatter settlements are based on science…detailed studies. But 'our people' 17 don't understand that. We must do what is right. (EPA Official, November 11, 2019).
But far from showing a 'lack of understanding', residents demonstrate a thorough appreciation of spatial dynamics, recognising both the spatial dimensions of risks and embedded opportunities. In this respect, risky spaces are also spaces of opportunity, when hedged against alternatives. This is the case for Mr. Sloh, 18 a middle-aged teacher and former IDP who moved to West Point in 2003 after the second Liberian civil war in 2004. He teaches at a secondary school just outside the community, his wife sells fresh fish on the beach and his adult daughter trades in clothes and other goods at the Waterside Market a few kilometres from the settlement. He also deals in building materials, a business that is booming because of the constant need to repair and build temporary structures. Although he indicated a high perception of risk in his responses, he elaborated on the factors influencing his continued investment in the settlement, such as the expectation that he would make money from his business. 19 So there is a risk perception paradox where residents have a high awareness of risk but decide to stay rather than leave based on trade-offs.
While West Point´s non-viability is down to the threat from coastal erosion and flooding, Peace Island on the other hand represents the threat of unplanned urbanisation to environmentally precarious ecosystems. The EPA has taken a strong stance against encroachments and sees Peace Island as the most audacious threat to the Mesurado Wetlands. 20 19

Governance and Institutional Responses to Riskscapes
To understand the dominant practices, it is important to grasp the role of power relations and knowledge production. Risk assessments not only influence the choice of risk management that maintains the prestige and objectivity of scientific research but also establish the dominant actor in urban planning. Bohle (2018), in tracing hurricane riskscapes in the Caribbean, points to the two dimensions of Foucauldian governmentality evident in them: first, directing the conduct of others using more or less strict coercive mechanisms, and second, directing the conduct of the self, amidst multiple possible actions. We can apply these two dimensions in the case study to understand the interrelationships of risk, space, and power within Monrovia's unplanned settlements, and what that means for resilience building.
The attitude of the city government towards unplanned settlements can be summarised in three approaches: first, a degree of tolerance, characterised by neglect; second, demolition and eviction; and third, legal recognition. Residents are aware of the government's attitude to unplanned settlements, particularly its reluctance to allow the expansion of new settlements and its preference to improve existing ones. This awareness effectively 'guides' the behaviour of residents, especially on Peace Island. 21 Moreover, this awareness places the onus on the residents to adopt practices to meet the demands of state actors. For example, fulfilling the government's expectations, such as preserving the Mesurado Wetlands seriously jeopardises residents of Peace Island access to fuelwood or building materials. This is part of the landscape of risk that residents must navigate, in the form of trade-offs: If the degradation of the mangrove forest attracts attention, it may hasten the government's determination to relocate the settlement, but preserving the mangrove forest also means losing access to a source of energy and building materials.
The contested riskscapes have important implications for resilience: How can resilience be achieved when state expectations and local survival strategies diverge? This question underlines the importance of a negotiation pathway that accommodates the multiplicity of riskscapes and resilience-building. For example, whilst the EPA considers Peace Island a threat to the Mesurado Wetlands, 22 other actors, such as Habitat for Humanity and Cities Alliance, as well as MCC, see Peace Island as a promising opportunity for in situ slum upgrading. 23 Therefore, resilience on Peace Island does not mean the same thing to the different actors; hence, risk reduction practices are likely to be different. However, it is important to understand what the different practices mean for residents in unplanned spaces.
This raises the question: How can governments incorporate multiple risk and resilience priorities into current plans or decisions? We start by considering these questions: 'Resilience to what?' 'For whom?' 'In what timeframe?' (Meerow & Newell, 2019). It has been argued that many cities' risk mitigation and adaptation 21 Interview with a community leader, Peace Island, September 19, 2019. 22 Interview with EPA Official, November 15, 2019. 23 Habitat for Humanity (March 18, 2021). Creating a decent place to live for residents of Peace Island in Monrovia. https:// www. habit at. org/ emea/ newsr oom/ 2021/ creat ing-decent-place-live-resid ents-peaceisland-monro via strategies exploit the poorest while benefiting the wealthier parts of the city (Agyeman et al., 2016; Anguelovski et al., 2016;Blok, 2020). Any strategy to build resilience or mitigate risk must therefore take into account the multiples of riskscapes and resilience and the multiplicity of practices they trigger (Fitzgibbons & Mitchell, 2021;Harris et al., 2018;Ziervogel et al., 2017). This recognition of multiplicity must take into account both the multi-scalar and actor's-specific interpretations of risk and resilience, as well as the priorities and strategies that meet their requirements for 'resilience' (Harris et al., 2018;Müller-Mahn et al., 2020;Raman, 2020). More importantly, building resilience procedures should take into account unequal power relations and promotes a trajectory that prioritises inclusivity and participation (Fitzgibbons & Mitchell, 2021;Harris et al., 2018;Ziervogel et al., 2017).
But how can inclusion and participation be achieved without a formal mechanism? The lack of a formal negotiation mechanism does not imply the absence of negotiations. The multi-scale governance alliances that exist in the two settlements shed light on how participatory processes work to manage complex riskscapes. Currently, NGOs such as Habitat for Humanity and CSOs such as the Slum Dwellers Association of Liberia (SLUMDAL) have established productive relationships with locals in West Point and Peace Island and mediate between them and national institutions. These governance arrangements exist along the formal and informal continuum. However, these interactions do not sufficiently address issues of justice and equity or the development of a negotiated pathway that could foreground them.
Where multi-scalar management and negotiation platforms are not prioritised, synergies appear through other modes, such as the governmentality of the state. The spatial transformation of Peace Island into a candidate for formalisation through rudimentary self-upgrading by residents is illustrative. These efforts are informally coordinated. Residents stated that awareness of the state's uncompromising stance on unplanned settlements influenced these actions. 24 From the outset, Peace Island residents sought to mimic Monrovia's 'orderly' vision as a strategy against future demolitions. 25 The most common reason respondents gave for investing in durable structures and maintaining rudimentary zoning in an area with insecure tenure was the hope that the government would interpret these measures as a first step towards upgrading, which would increase their chances of tenure. 26 However, some residents said that permanent structures also provide security for them to receive compensation if their structures are demolished by the government in the future. 27 A similar scenario was observed in Abidjan, where insecure land tenure enabled private investment and speculation with rental property (Eguavoen, 2021a-in progress). This is of strategic importance because the residents are aware that the survival of their settlement or the acquisition of land ownership depends on the position of 26 Focus Group Discussions on Peace Island on January 14, 2020, and February 8, 2020. 27 Interviews on Peace Island with an ex-combatant on December 1, 2019, and secretary for a local savings club on February 14, 2020. 25 Interview with community youth leader on Peace Island, December 12, 2019. 24 Focus Group Discussion, Peace Island, February 8, 2020. the government, which may be influenced by whether the settlement is suitable for habitation and is viable for upgrading, whether the issue of land ownership can be resolved or whether other commercial actors are interested.
Because risks are overlapping and interconnected, it is important to look at them relationally, rather than hierarchically (Fig. 3). Insecurity of land tenure and land use, for example, are linked to risks of eviction and further displacement. 28 Challenges with land tenure and land use, however, also point to a bigger challenge of poorly functioning land and housing markets which links to other risks, such as fires, landslides, or floods due to ineffective urban planning. 29 The interconnectedness of risks and their different interpretations (see Fig. 3) is directly linked to stakeholder practices, which are reflected in the different tendencies to mitigate risks. For example, residents associate insecure land tenure with the risk of eviction. The lack of modern cooking or lighting systems means using dirty fuels and candles for lighting, which has health consequences through indoor pollution and other risks such as fire outbreaks and burns, as well as crucial implications for education and productive, income-generating activities. The lack of other risk-mitigating infrastructures, such as water, sanitation and drainage systems, is also associated with everyday risks, such as the spread of waterborne diseases. These findings are consistent with several concerning studies linking infrastructure deficiencies to a range of urban risks (Baruah & Enweremadu, 2019;Butera et al., 2016;Kimemia & Van Niekerk, 2017;Kimemia et al., 2018;Zerbo et al., 2020).
The contrast between expert-recommended risk management and locals' view on risk mitigation is instructive for understanding the links between risk perception, risk assessment and risk management (Fig. 4). For example, the view that West Point should be demolished is based on risk assessments (see, e.g. Wiles, 2005) which stresses the settlement's vulnerability to coastal erosion. Residents, on the other hand, argue that the settlement's vulnerability is due to inadequate infrastructure and blame this on the government's failure to establish a coastal protection system. 30 This argument is not based on scientific research, but rather on the fact that other communities around Monrovia, such as New Kru Town and Hotel Africa, benefit from a coastal protection project. 31 This is revealing and could be interpreted as state actors' resignation to the destruction of West Point by the Atlantic Ocean. The fact that the coastal protection project launched by the government in 2018 with international partners to curb coastal erosion does not include West Point, perhaps the most at-risk settlement, was taken by one expert as a sign that saving the community is a hopeless endeavour. 32 This again raises the question of how risk reduction plans can integrate diverging priorities and interests into a risk governance framework. An evolutionary approach 30 Focus Group Discussion, West Point, February 12, 2020. 31 Coastal Defence Project, https:// www. adapt ation-undp. org/ proje cts/ enhan cing-resil ience-liber iamonts errado-county-vulne rable-coast al-areas-clima te-change 32 Interview with a former official of the Ministry of Public Works, October 28, 2019. to risk governance, focusing on the co-evolving elements and configurations that link discourse, actors, institutions, power and knowledge, will provide the platform on which the links between risk perception, risk assessment and risk management are debated and emerging riskscapes encountered (Van Assche et al., 2015). Returning to the ideas of 'multiplicity of riskscapes' and 'resilience multiple', evolutionary governance, with its emphasis on collectively binding decision-making by a diversity of actors inside and outside government, with formal and non-formal roles, can lay the foundation for inclusion and participation and thus provide the basis for a negotiated pathway to resilience building (Harris et al., 2018;Van Assche et al., 2015).

Navigating Risky Landscapes
Residents join a range of associations to expand their scope of action, build social networks and strengthen their interests. These associations range from religious to economic and recreational activities. The West Point Women's Health and Development Organisation (WPWHDO), for example, fights sexual and genderbased violence in West Point and provides education, health care, empowerment and safe spaces for rape survivors. Peace Island Community Youth (PICY) was established primarily as an advocacy group to provide opportunities for youths. These associations create spaces to shape and share ideas, strengthen and stabilise identities and develop a shared vision of a 'better future'. This resonates with Vigh's (2010) study on navigating risky conflict-affected landscapes in Guinea-Bissau, where associations in a shifting social environment provide some stability and predictability through their structure and function. An important role for these community-based associations is to negotiate visibility and assert spatial claims. These associations work closely with NGOs and CSOs, including the Slum Dwellers Association of Liberia (SLUMDAL), the Liberian branch of Slum/Squatter Settlement Dwellers International (SDI). Since 2008, SLUMDAL has been campaigning for rent control and negotiation of land rights for squatted land but has had little success in doing so. SLUMDAL has worked with government agencies on issues related to unplanned settlements but has been excluded from the process of developing the '2019 National Urban Policy (NUP)', a policy that will have critical implications for unplanned settlements.
Residents' alliances with NGOs, CSOs and international partners, particularly on the Peace Island, have led to increased upgrading activities. In March 2021, the EU and the government committed to including Peace Island in the co-funded electrification of Monrovia's suburbs. In March 2021, settlement leaders and international partners gathered to unveil new facilities completed, under the auspices of Habitat for Humanity and the Cities Alliance. The facilities include a community centre and two wells with solar pumping systems, as well as six elevated poly-tanks, three water kiosks, and two new and two refurbished toilet facilities.
Negotiations between informal actors are often the most common form of interaction concerning the use of unplanned spaces. For example, residents of Peace Island negotiate the location and use of spaces for houses, shops, temporary market stalls, etc. As mentioned elsewhere, squatting at West Point was authorised through negotiations with city officials, law enforcement and long-term residents who exercise a degree of quasi-control over the land. Residents also alleged that in some cases, bribes were paid to people posing as law enforcement representatives, and in some cases to gangs. Several factors influence the negotiation of space. The first settlers on Peace Island and West Point built their houses at higher elevations to avoid flooding and sea erosion. However, some of the first settlers on Peace Island settled near the marsh, which they used primarily for farming and fishing. The location of market stalls on Peace Island is determined by the flatness of the land, which was ideal for tables, or by physical features such as a greater distance from the marshes to avoid seasonal flooding. In West Point, fishmongers tended to locate their stalls closer to shore to interact more easily with fishermen. Some residents are involved in controlling the space, which often leads to the exclusion of less powerful residents. Community leaders who generally oppose development in the mangrove forest for fear of a backlash from the EPA, for example, have themselves been accused of exploiting the forest. This shows how ongoing negotiations at the local level can be undermined by local leaders.
Another issue repeatedly raised by the respondents is the remarkable uncertainty that pervades everyday life. They emphasised that they do not plan because they are uncertain of the future. Despite this assertion, they anticipate the future through shared positive expectations. However, everyday risk-taking is characterised not only by hedging against uncertainty through calculated practices that involve processes of self-organisation, adaptation and self-responsibility but also by the creation of risks. On Peace Island, for example, residents control run-off and erosion by digging drains to divert water elsewhere, thereby shifting the risk to other residents. This is the source of constant conflict in the settlement. 33 Mining sand on the beach is a cheap way for contractors to obtain building materials and a way for the residents of West Point to generate income. However, this practice accelerates coastal erosion and increases the risk of coastal flooding. 34 However, this everyday experience of uncertainty is not considered 'normal'. Instead, people always describe this situation as stressful. 35 This raises the question: How can one live a 'normal' life in a time saturated with uncertainty and instability? In conversations with several older residents of Peace Island, they speak of 'normal days', a time that denotes the pre-war period, compared to 'these days', which are characterised by extraordinary insecurity. This is not to say that there was no insecurity before the civil wars. But what they are experiencing now is exceptional.

Conclusion
The riskscapes of Monrovia's unplanned settlements reveal a landscape of risk and opportunity shared and managed by different actors. The multiple understandings and interpretations of 'risk' and 'resilience' inspire practices of different actors, shaped by multiple priorities, scales, contexts, or interests. However, unequal power dynamics mean that the practices of state actors influence the residents of unplanned settlements and subtly coerce them to integrate these practices into their strategies, with the resulting governmentality more about meeting the demands of the state 33 Focus Group Discussion, Peace Island, February 8, 2020. 34 Focus Group Discussion, West Point, February 18, 2020. 35 Focus Group Discussion, Peace Island, February 8, 2020Focus Group Discussion, West Point, February 18, 2020. rather than a desired course of action. The risk governance framework is thus one in which the interests and priorities of state actors prevail.
The current disaster risk reduction framework does not place sufficient emphasis on local participation. In addition, there is a lack of coordination between national authorities, a focus on key risks (rather than the conditions that lead to vulnerability) and a reliance on limited knowledge of unplanned settlement practices. Although there is no formal platform for collaboration or negotiation with government agencies, productive collaboration between residents, NGOs and CSOs in West Point and Peace Island offers the potential for co-production. Building resilience amidst the complexities of riskscapes requires collaborative governance and practical engagement based on flexibility and negotiation, involving the diverse stakeholders -from residents, state actors, civil society and NGOs -in developing a shared risk reduction strategy, rather than imposing standardised plans from above. Although government agencies are often indirectly involved in these informal collaborations or negotiations, the outcomes are often the continuous reproduction of expert knowledge in local spaces and governmentality.
Co-production can be understood as a process of shared learning and micronegotiation in which heterogeneous groups pursue a common goal (Duque Gómez & Jaglin, 2017). In Liberia's resilience initiatives, co-production has rarely been used as a method for the urban poor to develop their local organisation and build their capacity to negotiate with the state. Currently, top-down approaches dominate, accompanied by a donor-driven intervention that involves the imposition of readymade initiatives already planned by third parties. Furthermore, as riskscapes represent different realities for different stakeholders, the question of how to coalesce them into a common risk reduction or resilience strategy is relevant. This calls for a thorough understanding of exactly what kind of resilience stakeholders are seeking and how the different approaches and practices fit into the local risk governance mosaic. Riskscapes are multiple and variegated, and so is the resilience that stakeholders seek.
It is difficult for CSOs such as SLUMDAL to reduce disaster risk in a way that changes physical conditions in unplanned neighbourhoods, as they lack the necessary tools and financial resources. It can be said that residents have a good understanding of the risk situation and have developed coping strategies, but they do not have the means to change their situation in a meaningful way. While the authorities have the means, they often lack the knowledge of local conditions that would allow them to act with greater precision and certainty. In this risk-saturated environment, with multiple priorities, arguments and interests, it makes sense to set up co-production processes that lead all stakeholders to a negotiated outcome.
The lack of recognition of multiplicity in terms of riskscapes and resilience by current hegemonic risk governance means that actors' risk reduction strategies often have different emphases. For example, interventions shaped by scientific risk assessments focus on the geophysical risks such as coastal erosion or environmental degradation such as the threat unplanned settlements pose to the Mesurado Wetlands. Residents, on the other hand, are more concerned about the 'everyday risks' arising from the lack of infrastructure and services. This divergence calls for a repositioning of risk governance, with a more evolutionary approach that recognises the co-evolutionary elements and configurations that link discourses, actors, institutions, power and knowledge, to create a platform that concretises the links between risk perception, risk assessment and risk management. This complementary approach allows us to understand how residents live with these risks and, more importantly, which hazards are risks for them. Ongoing demolitions or restrictions on the use of the mangrove forest are risk reduction practices that establish the state as the master of the spatial configuration but do not address issues of rights and justice (Ziervogel et al., 2017) or emphasise a 'negotiated resilience' that takes into account the interests and priorities of the residents of unplanned settlements (Harris et al., 2018), or addresses the multi-scalar dimensions of managing complex riskscapes (Müller-Mahn et al., 2020;Raman, 2020).
Given the multiplicity of riskscapes, and by extension resilience, it is important that their multiple ontologies and multi-scalar dimensions are captured through participatory approaches that build local resilience through negotiated and co-productive processes (Harris et al., 2018;Ziervogel et al., 2017). As Harris et al. (2018) highlight, negotiated resilience does not necessarily reflect a 'formal process', but negotiations are 'often informal, uncertain, time-consuming, and may even be conducted through avoidance rather than direct deliberation and participation.' However, the possibility of informal negotiations does not necessarily preclude formal arrangements. Formal mechanisms may even be preferable to informal ones, especially for marginalised groups. While state actors, as well as powerful local actors, can easily manipulate informal negotiations to their advantage due to unequal power dynamics, formal mechanisms can create a balance that promotes compromise and trade-offs and allows marginalised voices to be heard.
The spatial heterogeneity of unplanned settlements is also discussed in this paper. There are a variety of unplanned morphologies in Monrovia, and although informal adaptations are similar across unplanned settlements, resilience-building may take different paths depending on socio-spatial characteristics. Peace Island is newer and therefore emerged during a period of heightened awareness of sustainability. Peace Island can therefore potentially be upgraded in situ. However, West Point has experienced exponential population growth, and its extreme land scarcity and vulnerability to coastal erosion make it an unsuitable candidate for upgrading. The contrast between the two settlements shows that spatial characteristics are important in designing interventions.
Despite exposure to a precarious environment characterised by geophysical hazards, the legacy of devastating civil wars, social vulnerability, among others the decision of locals to stay in these settlements vary and range from little or no knowledge of alternatives or the lack of the wherewithal to pursue alternatives. This situation also highlights the risk perception paradox, where people choose to stay in the hope of financial gain or because of other trade-offs and considerations. What is constant, however, is that they live in an environment saturated with risks, often translating these risks not only into hazards but also opportunities. But these risk translations must also interact with the multiple interpretations of spatial risks by state actors and other stakeholders, such as CSOs and NGOs, leading to multiple, and sometimes contradictory, practices; hence, the multiplicity of riskscapes.
In West Point, there is the issue of spatial and temporal uncertainty. In an environment where trust in the capability of authorities to implement coastal protection projects is low, the community's survival depends on the implementation of a coastal protection project. Here, expectations do not match reality. Coastal erosion is aggressive, but because people are attached to the area and fear for their livelihoods, and because of other intervening factors, they continue to harbour faint hopes that the settlement can be saved. Peace Island, however, is making progress towards its upgrading goals, despite a lack of official recognition as a result of collaboration with international organisations, NGOs, and government agencies. Therefore, West Point represents despair, whereas Peace Island represents hope.
In this sense, the dichotomy between hope and despair corresponds with Stokes's (1963) concept of 'slums of hope and despair, which describe the heterogeneous material and social conditions of poor neighbourhoods in developing countries. Residents of the unplanned settlements that represent 'hope', such as Peace Island, are portrayed as ambitious and positive and living in conditions that can gradually improve, while residents of the area that represents 'despair', such as West Point, are on a downward trend, are more pessimistic about the future and live in a state of uncertain stagnation. In this context, rapid urbanisation in many cases means an accumulation of risks, but it also offers promising opportunities for socio-economic development.
As we recognise the importance of the multiplicity and fluidity of overlapping and sometimes contradictory riskscapes, the gap between disaster risk and everyday risk narrows. In this context, recognising 'multiplicity', rather than a single conceptualisation of 'risk' and a single pathway to 'resilience' creates a flexibility that allows for a more inclusive process of building resilience. This ensures the integration of different agendas, and local knowledge and capacities, and creates a platform for compromise and negotiation between actors within a flexible risk governance arrangement.
Funding Open Access funding enabled and organized by Projekt DEAL.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.