Participation is not the answer: epistemic violence and authoritarian practices in conservation-forced displacement

ABSTRACT

This article provides an in-depth analysis of the first 13 years of conservation-forced resettlement negotiations in the Limpopo National Park in Mozambique. It documents four strategies used by the state to disable peoples’ voices: delegitimizing their preferences, delimiting negotiations, coercing them into consent and disengaging from participatory policy. It shows how invited spaces of induced participation amplify epistemic violence and entrench authoritarian practices. It argues that, while the normative belief in the value of participation remains largely unquestioned, participation in fact legitimizes displacements. It calls for decolonial approaches to conservation that do not rely on and cause displacement.

KEYWORDS:

• Displacement
• resettlement
• participation
• decolonial
• conservation
• epistemic violence

Introduction

You see, the government here is used to being heard and paid attention to, when Frelimo talks everyone obeys. This is democracy but it is not really, and the resettlement process taught leaders to demand their rights. We taught them that they have rights over the course of the process! You see it was ‘induced participation’ and anything induced has its reactions that sometimes we are not prepared for. Like when you induce birth in a pregnant woman. Sometimes suddenly the contractions are too strong, and the woman and the baby cannot handle it. The situation becomes out of control and then the doctors decide that it is time to have a caesarean, but by the time they manage to cut her open the baby has died. Therefore, the matter becomes saving the life of the mother.¹

The quote above is reflection by a member of the resettlement committee from the Limpopo National Park (LNP) in Mozambique. Park managers decided, in 2002, one year after establishing the National Park, to resettle 7,000 residents of the park to make room for tourism development and wildlife. Initially, this resettlement project was carried out according to the World Bank Operational Policy on Involuntary Resettlement (WB OP 4.12) which states that people should participate in planning their own resettlement. The induced participation created a situation that became ‘out of [government’s] control’ and the participation ‘baby’ had to be sacrificed to try to save the mother: participatory spaces were shut down to protect the government’s authoritarian power relations with citizens and control over the resettlement project. Through a long-term case study from the LNP, we explore how participatory spaces in resettlement projects amplify epistemic violence and entrench authoritarian practices.

In the 1990s, protests erupted against forced evictions and involuntary resettlement of people from protected areas. The international response was participation: if people could participate in the planning of their own resettlement, and could be sufficiently and adequately compensated, long-term damage to displaced communities could be avoided (Cernea 1997; Guggenheim 1993). Yet, more than 20 years after the participatory clause was included in WB policy for involuntary resettlement in 2001, resettlement continues to cause harm to displaced people (Cernea and Maldonado 2018; Price and Drydyk 2023). Despite this, most authors continue to promote participation, with the caveat that it should be meaningful and equitable, and entail empowerment and agency, as the only way forward (Vanclay 2017; Cernea 2021; Price and Drydyk 2023).

In this paper, we question whether it is possible for displaced people to participate in their resettlement without being shut down or turned against with more violence than already inherent to displacement. Over the last 50 years, displacement has been justified by a global narrative that affirms that displacement is an unfortunate secondary effect inevitable for ‘development’ to occur (Milgroom, Kabra, and Wilmsen 2023). Evidence is mounting that disproves this narrative and points, rather to how displacement is a pattern of structural violence embedded within growth-focused development. Recent literature points out that most displacements are taking place in the Global South (Rogers and Wilmsen 2020) in countries with authoritarian governments, who are using dispossession as a process of unquestioned state-making (Blake and Barney 2018) and as a land management tool (Wilmsen and Webber 2015). Authoritarian regimes are displacing people to open up resource frontiers, and to extend sovereignty into hinterlands (Rogers and Wilmsen 2020; Lunstrum 2013; Lima and Kmoch 2021). Furthermore, those who are most acutely affected by displacement are groups who are already marginalized in society (Kabra 2020). They are often ‘othered’, considered to be of an inferior class, race or ethnicity (Ramutsindela, Matose, and Mushonga 2022; Chakrabarti and Dhar 2009) and cast as ‘requiring salvation from the strong arm of development’ (Wilmsen and Webber 2015).

Participation, in this context, creates a tension in which government officials are supposed to give leverage in decision making to the very people who are ‘in the way’ and perceived to be inferior and undeveloped. There are few studies that analyse in depth how participation in resettlement projects works, and how this tension plays out in practice. Those that exist point to the performative (Singh 2009), tokenistic (Cai et al. 2020) and manipulative nature of participation (Beazley 2011; Habich-Sobiegalla and Rousseau 2020) that can even make communities feel responsible for agreements made (Otsuki et al. 2017) and to its inherent contradictions and paradoxes (Aronsson 2009; Koenig 2007; Price 2015). In this paper, we build on research conducted by Dhungana and Curato (2021) which shows that participation entrenches authoritarian practices, rather than promoting citizen voice and government accountability (Dhungana and Curato 2021). We contribute to a deeper understanding of how voices are disabled in participatory processes by examining how induced participation amplifies the epistemic violence of displacement, a type of violence that is underexplored in resettlement scholarship.

This paper covers the first 13 years of resettlement negotiations in the Limpopo National Park (LNP) in southern Mozambique and offers the first in-depth history of these negotiations. The LNP case highlights many lessons on how induced participation works in practice because it was carried out with the intention to ‘get resettlement right’. Initially, time and resources were dedicated to making sure that the negotiation process was harmonious, and to including displaced people in parts of the decision-making process. Yet things went wrong. As we show in this paper, the state used four different strategies to disable people’s voices: (1) delegitimizing local ways of being and knowing through imposing a western model of development, (2) delimiting what could be negotiated, (3) coercing people through threats and manipulation, and (4) disengaging from participatory policy when participation led to uncontrollable outcomes. Our analysis shows how participation legitimizes and justifies displacement, while the normative belief in the potential for participation to improve resettlement processes and outcomes remains largely unquestioned.

In the next section, we present literature on displacement from conservation areas, highlighting its colonial dimensions, and bring together research on induced participation, authoritarian practices and epistemic violence. We then describe the methodology and provide some background on Mozambique, the LNP and its resettlement project. Our findings and discussion are organized chronologically in four parts: delegitimizing, delimiting, coercing and disengaging, each highlighting a specific way in which the state disabled people’s voices in ‘participatory’ spaces. We conclude by outlining the intractable contradictions of participation in resettlement and what lessons this case study might hold for ongoing and upcoming displacements. The number of projects that entail displacement is on the rise (Cernea and Maldonado 2018). Displacements from conservation areas are likely to continue to rise given the recently adopted Kunming-Montreal Global Biodiversity Framework that seeks to ensure that at least 30% of the earth’s surface is effectively conserved.² In this context, we posit that rather than focusing on promoting or improving participation, we should find ways to decolonise conservation and stop displacement altogether.

The coloniality of conservation and displacement

Globally, conservation areas have proliferated in the last 40 years (Environment and Climate Change Canada 2023), and with them the numbers of displaced peoples (Cernea and Schmidt-Soltau 2006; Brockington and Igoe 2006; Kabra 2018). This is largely because of the ‘fortress conservation’ model of nature conservation that emerged in the United States in the early nineteenth century and extended throughout the world (Nash 2001), which excludes people from protected areas. The separation between humans and nonhumans, the notion of ‘nature’ as pristine and pure, and the idea that people are a threat to the natural world are tenets of colonial logic that drive this model (Ramutsindela, Matose, and Mushonga 2022). The creation of conservation areas void of people was also supported by narratives that portray certain inhabited landscapes as empty (Brockington 2002; Neumann 1998) without considering the ways in which landscapes as we know them have been shaped by people for centuries (Adams and Mulligan 2012). More recently, neoliberal approaches to conservation have reconceptualized the biophysical environment as a container of ecosystem services and ‘natural capital’, relying on financial instruments and private investments to finance conservation through the market (Fletcher 2020). Colonial histories continue to resonate throughout market-based conservation initiatives, which portray local people as villains and blame them for environmental degradation while disregarding their epistemologies in overcoming existing challenges (Apostolopoulou et al. 2021). Where violent evictions used to be the norm in the past, governments now increasingly use the slow violence of neglect, systemic discrimination and cultural delegitimization to make land ‘grabbable’ (Weldemichel 2022; Mahalwal and Kabra 2023).

Southern Africa has a long history of displacement of people for nature conservation since colonial times when people were evicted to establish the parks such as Kruger National Park. Today, nature conservation continues to be driven by neo-colonial (Adams and Mulligan 2012; Domínguez and Luoma 2020; Ramutsindela 2004) interests. Despite political decolonization and independence, in many African countries, governance over land still retains a colonial structure reinscribed through internal colonialism characterized by accumulation of land in the hands of the elite (Carlos Bezerra and Paphitis 2021; Chitonge 2018). Ramutsindela, Matose, and Mushonga (2022) illustrate how violence is a permanent feature of conservation in Africa; it reproduces colonial hierarchies, racism and exclusion that continues to be justified by colonial ‘othering’. Furthermore, apolitical, technocratic rhetoric such as participation masks the deeply political nature of conservation.

Amidst these neo-colonial exclusions, the plight of displaced people came to the forefront of conservation debates (Brockington and Igoe 2006). It was clear that displacement causes long-term harm and intergenerational poverty (Cernea 1993). The World Bank drafted its first policy to protect people from forced evictions and to mitigate impoverishment risks in 1980, mainly driven by the massive displacements of people caused by the building of dams. Rights-based approaches to displacement followed, most notably the World Commission on Dams which incorporated participation as a central component of its framework (WCD 2000). In 2001, the WB added the participatory clause to their operational policy for involuntary resettlement, also applied to conservation forced displacement (WB OP 4.12). Fletcher usefully describes successive waves of conservation, from fortress conservation (1860-1960) to community-based conservation (1970-2000) to neoliberal conservation (2000-), pointing to the beginning of a fourth wave of post-neoliberal conservation characterized by state-led developmentalism. The rhetoric of promoting participation is at the heart of neoliberal conservation, which does not entail a dismantling of the state but rather a new role for the (authoritarian) state as guardian of neoliberalism’s operational logic (Fletcher 2020). The World Bank has been a primary agent of neoliberalization, whose interventions have benefited the elite and intensified socio-spatial and social-environmental inequality. Today, conservation projects are on the rise and most projects are formally bound to participatory policies.

The paradox of participation in resettlement

In conservation-forced displacement and resettlement (CFDR), just as in development-forced displacement and resettlement (DFDR), participation is often imposed or ‘induced’ as a requirement by donors. Compensation tends to be overly focused on the visible, material components such as infrastructure including houses, schools, electricity or other modern services (Vanden Berg and Todd 1999; Witter and Satterfield 2014). It is measured in primarily economic terms and often overlooks cultural, non-tangible losses (Downing 2002; Mathur 2013). The main policy goal of participation is to design more adequate compensation that would lead to ‘development’ (for an in-depth reading of the policy text see Koenig 2007; for a critique of this notion of development see also Milgroom, Kabra, and Wilmsen 2023).³ Induced participation is meant to contribute to post-resettlement conditions better tailored to people’s needs, cultural habits and to the local context through co-creation of resettlement plans in order to mitigate impoverishment risks (Cernea 1997; World Bank 2004). It is also meant to be beneficial for the resettlement project by pre-empting large-scale, costly public protest, and reducing the risk of human rights abuses that would tarnish the reputation of donors funding the project.

This points to a paradox of participation in resettlement: participation is meant to ease resistance, but it often opens space for resistance. Induced participation is intended as a technocratic exercise and is not meant to give rise to political contestation. Regardless, ‘invited spaces’ for participation (Cornwall 2002) can lead to unintended effects including political awareness and social learning because participation, closely linked to power, is inevitably political (Hickey and Mohan 2004; Williams 2004). Displaced people’s lives and livelihoods are on the line and given the chance affected people will express their grievances (Oliver-Smith 1991). The WB-induced participation policy clause, firmly rooted in a belief in linear planning, is based on a seemingly unproblematic belief that resettlement can be involuntary and that despite this, people will voluntarily proceed to participate in the planning of their own dispossession (Koenig 2007; Aronsson 2009). The equal right of displaced people to veto resettlement decisions or the development project would be an important condition for participatory spaces to address deeper grievances by displaced people (Chakrabarti and Dhar 2009) yet this, in practice, is very rare. The cards tend to be already stacked against displaced people and ‘participation’ tends to remain a symbolic gesture of one-way transmission of information to satisfy donor requirements (Wilmsen and Webber 2015), or a covert mode of coercion and manipulation.

Authoritarian practices and epistemic violence

Dhungana and Curato (2021) show how participation in post disaster governance entrenches authoritarian practices. They use Glasius’s (2018, 527) definition of authoritarian practices as ‘patterns of action that sabotage accountability over whom a political actor exerts control, or their representatives, by means of secrecy, disinformation, and disabling voice’. They find that participation (a) controlled rather than democratised information; (b) silenced rather than promoted citizen voice; and (c) distorted rather than responded to grievances. While post disaster governance differs from CFDR mainly because of the urgency of the situation, the framework remains useful because it highlights how participation has a net effect of disabling people from having a say in the interventions shaping their lives. This paper focuses specifically on the authoritarian practice of disabling voice. Disabling voice is a disruption of the dialogical flow of information from the citizens to the state (Glasius 2018). By bringing epistemic violence into this framework, we explore how participation leads to the disabling of people’s voices.

Epistemic violence is violence exerted against or through knowledge; violence inflicted by discourse. In her canonical essay, Spivak illustrates the way in which knowledge held by marginalized people was dismissed due to privileging colonial, western epistemic practices. She describes the way groups were silenced by damaging a group’s ability to speak and be heard (Spivak 1988). Spivak develops the concept of epistemic violence as ‘the remotely orchestrated, far-flung, and heterogeneous project to constitute the colonial subject as Other’ (Spivak 1988, 280). Although epistemic violence can take many forms, when the ‘other’ is deemed unintelligible, they are prevented from having a voice altogether, or experience reduction of credibility and denial of epistemic agency (Dotson 2014). Importantly, epistemic violence is located in a specific origin (the primacy of western/European knowledge based in modernity) and a specific history of colonialism and capitalism (Brunner 2021).

Epistemic violence often leads to distributive injustice in the material realm in which a certain group determined as the inferior ‘other’ is allocated fewer resources (Fricker 2017). Epistemic violence is a form of slow violence that is inherent in dispossession (Chakrabarti and Dhar 2009; Lesutis 2023). Pérez (2019) argues that epistemic violence is a structural form of slow violence that is gradual, cumulative and almost imperceptible in nature: ‘it subsists without being detected, either because it is presented in a context affected by other, more evident forms of violence, or because it is framed within a ‘well-meaning’ proposal (Pérez 2019, 9). The ‘well-meaning’ proposal, in this case, is the participation of people in their own resettlement.

Methodology

The first author carried out transdisciplinary research in and around the Limpopo National Park for a period of 4 years, from December 2006 to June 2010, and then conducted follow-up visits in 2014 and 2016. The bulk of the fieldwork informed her PhD dissertation: she lived intermittently in the village of Nanguene for two years before resettlement and in the post-resettlement location for 18 months. With the intention of documenting the resettlement process, she moved to the village of Nanguene because this was the first village to be resettled as part of a pilot project intended to develop a resettlement action plan that would be used for the resettlement of the other villages. Resettlement negotiations started in 2002 for Nanguene and in 2006 this resettlement was to be imminent. However, it took another two years for the resettlement to take place (2008) and during this time (2006–2008) she collected in-depth ethnographic data on the process of negotiations that unfolded as village leaders, park staff and donors designed the compensation package and ironed out the plan for resettlement.

In addition, she participated in 12 meetings consisting of negotiations about resettlement conditions between village residents and park officials, as well as between donor representatives and government officials from 2006 to 2010 and one in 2014. She conducted more than 200 open and semi-structured interviews with residents in 13 villages as well as with village leaders, district, provincial and national government officials, park staff, donor representatives, private consultants, and NGO staff. She also analysed unpublished park documents, consultancy reports and meeting minutes, including those documenting the negotiation process starting in 2002. For the duration of the research, from 2006 to 2016, she worked with two research assistants and interpreters to add precision and subtle cultural interpretation to her comprehension of Shangana. She transcribed and translated meeting records and interviews from Shangana or Portuguese to English. She analysed data manually using inductive coding. Ethics approvals were granted through the University of Wageningen and the government of Mozambique through the Limpopo National Park. Consent was given orally before any data collection.

We have come back to this data today, in 2023, because we observed, through following the development of literature on displacement and resettlement, that patterns of failed participatory processes continue to be repeated. In this context, the long term, in-depth data collected 15 years ago is valuable for understanding why. Furthermore, this formative period in the history of the LNP has been insufficiently documented. With a view to including recent developments on resettlement in the LNP, the first author conducted three follow-up conversations in 2022 with people who continue to work in the LNP.

Resettlement in the Limpopo National Park

In 2001, the government of Mozambique established the Limpopo National Park in an area called Coutada 16 that had previously been Portuguese colonial hunting grounds. The new national park was created as a steppingstone to the development of the Great Limpopo Transfrontier Conservation Area (GLTFCA), composed of national parks in Mozambique (Bahnine, Zinhave and Limpopo), South Africa (Kruger) and Zimbabwe (Gonarezhou) and the areas in between and around these parks (Figure 1). The GLTFCA was proclaimed in 2002 to promote, in theory, regional cooperation, conservation across national boundaries, and economic development (Wolmer 2003). It was one of a series of ‘Peace Parks’ developed in the late 1990s and early 2000s as part of a trend to make conservation areas that straddled national borders that speak to regional ‘peace’ efforts (Büscher 2010). The Peace Parks Foundation (PPF) is a South African NGO that has driven the creation of most of the transfrontier conservation areas (TFCA) across Southern Africa and is an important player in pushing exclusionary conservation models in Southern Africa that benefit both the economy and the power of South African elite (Draper, Spierenburg, and Wels 2004). TFCAs represent a shift toward neoliberal conservation in which non-state actors, such as NGOs and international donors, have taken on important roles such as instigating, managing and funding the initiatives (Büscher 2010; Ramutsindela 2004; Duffy 2006).

Figure 1. The Limpopo National Park, in Mozambique, borders Kruger National Park in South Africa. The original location of the first village resettled (Nanguene) and its post-resettlement location next to the host village of Chinhangane are indicated. Eight other villages located along the Shingwedzi River in the center of the LNP were slated for resettlement to areas outside the park boundaries. (Map credit: author)

The park was managed by what was called the Project Implementation Unit. It was made up of the LNP project director who was a South African PPF employee and the Mozambican park administrator. The park was institutionally supported by the World Bank, and had funding from the German Development Bank Kreditanstalt für Wiederaufbau (KfW). Because the LNP falls within Gaza Province, the provincial governor also was an important person for decision making in the resettlement process (Figure 2). At the time of data collection, management of conservation areas was under the jurisdiction of the National Directorate of Conservation Areas (DNAC) within the Ministry of Tourism (MiTur). The transfrontier parks, including the LNP, were managed by the TFCA unit, located within DNAC.

Figure 2. Main actors relevant to the LNP resettlement process during the main period of fieldwork (2006–2010). Abbreviations stand for: Ministry of Tourism (MiTur), National Directorate of Conservation Areas (DNAC), Transfrontier Conservation Area Unit (TFCA), District Administrator (DA), Project Implementation Unit (PIU), Peace Parks Foundation (PPF), South Africa National Parks (SANParks), World Bank (WB) and Kredietanstalt für die Wiederaufbau (KfW).

The area was home to 26,000 people. A year after the establishment of the park, park managers in Mozambique and South Africa considered the presence of people in parks to be highly problematic for the development of tourism (Spierenburg, Steenkamp, and Wels 2006). In 2002, park managers, including those employed through PPF in South Africa, decided that approximately 7,000 residents, those living in the centre of the park, would have to be resettled to areas outside the park’s boundaries (Figure 1). The daily implementation of the resettlement project was carried out by the resettlement program made up of park staff who were Mozambican employees of the Ministry of Tourism. The resettlement program staff was made up of 2–4 people over the course of the research period.

The resettlement was labelled ‘voluntary’ by both the Mozambican government, as well as the foreign donors, to ease resistance and uphold international reputations, even though it was not voluntary because residents were not given a choice to refuse the resettlement (Milgroom and Spierenburg 2008). It was decided jointly by the German Development Bank Kreditanstalt für Wiederaufbau (KfW) and the WB that the process would adhere to the World Bank’s Involuntary Resettlement Operational Policy 4.12 (WB OP 4.12) as a condition for funding (Milgroom and Spierenburg 2008). This is because KfW, the sole external donor funding the resettlement initiative, did not have the policies or expertise to manage a resettlement project. One of KfW’s main concerns was to avoid unjust treatment of project-affected people to protect their reputation in Germany. The World Bank had launched the transfrontier conservation area concept in Mozambique in 1998 and had funded the establishment of three transfrontier conservation areas (Limpopo, Lubombo and Chimanimani). Given its experience and policy for resettlements, and its involvement in the TFCA projects in Mozambique, the WB agreed to serve as backstoppers for the resettlement project, responsible for approving and monitoring the implementation of the resettlement pilot project according to the WB OP 4.12. An external KfW consultant also regularly evaluated and assisted the resettlement project. The fact that the resettlement was called ‘voluntary’ yet it was guided by an involuntary resettlement policy caused confusion with respect to how to proceed in the initial phases of negotiations (Milgroom and Spierenburg 2008).

The Anglo-Saxon, ‘fortress’ model of conservation described above, was the predominant model of conservation in South Africa. However, in Mozambique, until the case of the Limpopo, people lived inside all of the national parks in the country. There had been processes of displacement from other parks, most notably from Gorongosa National Park, but none carried out according to World Bank policy standards. Legislation governing conservation areas in Mozambique claims a modernist perspective with little regard to local culture or people’s use of resources in proposed conservation areas, including the globally renown 1997 land law, famous for its recognition of communal land use and its rejection of private property (Lunstrum 2008; Matusse 2019). The land grabs of the last decade made it evident that Mozambique’s modernist agenda facilitates the power of elites to procure benefits from land acquisitions (Fairbairn 2013). Mozambique, and specifically the area of the LNP, has a long history of displacement carried out for various motives from colonial times of the Portuguese through the socialist policies of Frelimo to neoliberal trends of resource commodification.

Before the establishment of the park, residents depended on agriculture, hunting, gathering and seasonal work in South Africa for their livelihoods. They do not self-identify as Indigenous peoples, and they are ethnically and culturally connected to South Africa. Although the case of the LNP has received considerable attention from academics, there has not been much transnational advocacy to defend the rights of local residents to stay on the land and oppose their displacement case. Local NGOs were restricted from supporting communities before resettlement but have assisted in some post-resettlement livelihood reconstruction, and in publicly denouncing some injustices. Five villages have been resettled to date, but this case is the longest standing resettlement case on the continent (Lunstrum 2016).

Ampliflying epistemic violence and entrenching authoritarian practices

We now turn to the strategies used by the state to control participation and its outcomes: delegitimizing, delimiting, coercing and finally disengaging from participation. We analyse how these strategies were used by government officials to disable the voices of the displaced, amplifying epistemic violence and entrenching authoritarian practices over 13 years of resettlement negotiations in the LNP.

Opening space for participation: a pretense of enabling voice

Donor imposition of the WB OP 4.12 as a guiding policy for resettlement altered the way the Mozambican state authorities usually interacted with the people they were displacing. A KfW representative commented:

When we first heard about resettlement, the government was ready to come in here on trucks and cart people away. We said, well, in that case we will not fund resettlement. We knew resettlement would be difficult, but we could not allow people to be taken away on trucks.⁴

When the park was established, committees were set up in each village in the park, at the insistence of KfW, to inform people that they lived in a national park. Yet, neither the district administrator nor the LNP park staff were equipped with experience or skills for facilitating participation and failed to do so, creating considerable distrust among villagers (DeMotts 2017). To the question, ‘do you want to be resettled?’, many residents replied ‘yes, because we have to. The government says so’.

In December 2003, the first meeting to discuss the conditions of resettlement was held at a beach resort in Bilene far from the LNP. The meeting was attended by eight village leaders, one from each of the villages to be resettled, the provincial director of tourism, the park administrator from Mozambique, the director of the LNP project (representative of PPF), the district administrators of the three districts in the LNP, provincial government representatives from three ministries, two members of the LNP resettlement program staff and three foreign consultants. The village leaders were outnumbered by government officials of multiple ranks, creating an atmosphere of formality and accentuating the imbalanced power relations. In this meeting, consultants presented (in English and translated into Shangana) WB OP 4.12’s twelve principles of resettlement⁵ (Huggins et al. 2003, 8). Despite translation from English, being outside of their geographic territory, in an institutionally unfamiliar location for village leaders (a beach resort) and outnumbered by government officials, village leaders were able to add two new principles to the list:

1. Communities/households should be resettled as a whole, not as individuals

2. There must respect for cultural habits. There needs to be a budget for ceremonies at the new site including community level ceremonies and household level ceremonies (to inaugurate corrals, etc).

These two new principles reflect the values of the people being displaced: the importance of collectivity rather than individuality, and the recognition of traditional cultural practices. This openness to incorporating local values into the international policy framework set the scene for what appeared to be a recognition of the validity of displaced people’s knowledges, and a willingness to listen and adapt to their concerns.

During this meeting, a structure for information sharing and decision making about resettlement was created. The Consultative Committee on Resettlement (CCR) was the main space in which the formal participatory process was to take place. It was ceremoniously presided over by the provincial governor, the most powerful government official in the group, and included the same leaders from the resettling villages as were present in the first meeting (some of whom were FRELIMO officials and some of whom were traditional leaders recognized to be the de-facto leaders), as well as the leaders of the two host villages already chosen for the pilot project (Banga and Chinhangane), a representative from the NGO ORAM (Associação Rural de Ajuda Mútua – Rural Organization for Mutual Aid), park staff from the resettlement program and provincial and district administrators from various ministries. In the first CCR meeting in 2004 in the beach town of Xai Xai, a few months later, the park director specified that these meetings are for ‘(1) each person to express what they feel, (2) for each leader to present any concerns from people in their village and (3) to work together to find ways to overcome any difficulties that emerge in the resettlement program.’⁶ It was emphasized that together they would decide on the steps and strategies for implementing resettlement.

In this initial phase of resettlement negotiations, it appeared as though some structures were being set up to enable a space for meaningful participation in which displaced people would have a voice in shaping the resettlement. This echoes a similar experience by people displaced from a conservation area in India who also had a short-lived period of apparent political influence that lasted until they gave their consent to being resettled (Beazley 2009).

It should be noted, however, that the conditions for meaningful participation were not met in the invited spaces of participation described in this paper: meetings were too infrequent, they resembled one-way information delivery, excluded women and other marginalized people, and involved only the village leaders without incorporating mechanisms for representation of the other village residents (see also Otsuki et al. 2017). Furthermore, mechanisms of elite capture were evident but an analysis of the intra- and inter-village power dynamics is outside of the scope of this paper. There was also a lack of information about what was going on, misinformation and lack of transparency throughout the entire process. As stated above, early meetings took place out of the local environment, in a beach resort with a formal context of imbalanced power relations and were framed by foreign consultants speaking in English. These conditions already set the scene for epistemic violence, despite what was being said in the meetings about making decisions together.

Delegitimizing: the epistemic violence of imposing a pre-determined model of development

At the very same, first meeting in 2003, discussions ensued about the type of houses that displaced people would receive. This topic would dominate the resettlement negotiations for the next 10 years. The focus on infrastructure development rather than livelihood rehabilitation is one cause of long-term harm in resettlement (Downing 2002). Four options for the compensation houses were presented, the first three being based on traditional, mud house models. The consultants recommended the fourth, a brick house with a zinc roof, wooden doors and windows (Huggins et al. 2003). In the second CCR meeting one village leader rejected this modern model saying: ‘the houses are too small and go against the local culture. A couple cannot share walls with their children.’⁷ Some of the other leaders supported him saying that they could not house two wives under one roof. No KfW representatives were present at the CCR meetings – they depended on the LNP staff to manage the negotiations and communicate the village preferences to donors in reports. This quickly led to manipulation of information.

At the third CCR meeting in 2005, representatives of the provincial offices of the Ministries of Tourism, Agriculture and Rural Development, and Environmental Action decided on the cement house model. Despite having added the principle of respect for cultural habits in the list of principles of resettlement, the government officials delegitimized the testimonies of leaders who spoke in favour of traditional houses, overriding their request as inadequate. They stated that cement houses arranged in rows on square plots of land would improve their lives and ensure development. The village leaders’ request for culturally appropriate housing was rejected and looked down upon. A KfW representative explained:

We wanted to provide houses, but local houses of the same kind that they have now, an exact replacement of what they have, but slightly improved, for example with cement floors. However, the government turned this into a debate about ‘development’. They said that we need to improve their lives by building them cement houses.⁸

The leaders’ opposition to the cement houses was recorded in the CCR meeting minutes, but the park director deceived the KfW representative by telling him that the residents preferred the cement houses in order to attain approval because the KfW representative was eager to approve what the displaced people preferred.⁹ In the name of ‘development’ the Mozambican government inflicted epistemic violence by delegitimizing people’s testimonies and disregarding the village leaders’ request to respect their cultural practices, specifically their preference for individual huts, rather than a shared house. This led to a model of resettlement that entailed a loss of a traditional way of living, a very blatant example of epistemicide (de Sousa Santos 2015).

The infliction of epistemic violence was partly masked by two factors: the paternalistic good intentions of ‘development’ and the good relations that the resettling committee kept with the displaced people. An important part of the justification for displacement was that people would benefit from so-called ‘development’, therefore it was incongruous for the LNP to concede to the needs and preferences of villagers if what they desired did not align with the government’s notion of development. In that context, treating people as backwards, uncivilized and unintelligible was a way to justify the displacement itself.

Ironically, later in the process, the cement houses became one of the most attractive parts about resettlement. This could be a result of one type of epistemic violence called testimonial smothering in which people truncate their own testimony out of fear of the way it will be received by the listeners (Dotson 2011). In other words, people did not dare to claim to want something that was not ‘modern’, out of fear of being judged as ‘uncivilized’. It could also be that people came to desire what had been denied to them over a long history of neglect and marginalization as described in a case of CFDR in India (Kabra and Das 2022).

Delimiting the scope of the negotiations

In 2007, another structure was created for participation of leaders in the resettlement planning, called the Resettlement Working Group (Grupo de Trabalho de Reassentamento) (GTR). This was different from the CCR because it consisted of several village residents from the villages to be resettled from the pilot project (Nanguene and Macavene) as well as the corresponding host villages (Chinhangane and Bingo). True to the intended participatory nature of the space, the decision to involve three people from each village rather than just the leader was proposed by the group in the first GTR. Only one or two staff members from the LNP resettlement program and the external consultant hired to help develop the Resettlement Action Plan (RAP) participated in the meetings, creating a much more relaxed space for village members to express themselves. The intention was to create a space for making decisions together about the compensation package that was being defined for the first time for the pilot project.

In the GTR, displaced residents expressed that they wanted large extended family homesteads rather than individual nuclear family plots. They wanted a spatial organization that allowed for empty plots between houses so that their future children and grandchildren would not have to settle far away from them. They requested manual hand pumps because of the ease of repair and their distrust in who would have to pay for the diesel (when powered by a generator) or the electricity when this eventually came, if ever, for the electric ones, and fear of theft of solar panels. They wanted to keep their cattle near their houses and rejected the idea of common holding pens. They also were concerned about there not being enough fields to cover the needs of the resettling villages, as well as the future needs of the host villages. Distrustful of the proposal to substitute their dryland fields for small plots of irrigated agriculture, the infrastructure for which would have to be collectively managed and could break down, they requested more land for dryland farming. They also requested to be compensated monetarily for their time spent in planning meetings because of the injustice of the expectation that they would do it for free on behalf of the village while the LNP staff were paid to be there, and because of the time lost in their fields needed for securing food for the year.¹⁰ Although there were differences among the leaders, there were no objections made to the requests made by any of the leaders.

None of these issues were taken into consideration except for the rejection of the common holding pens for livestock. Long before this discussion, external consultants, Mozambican government representatives from relevant ministries and the resettlement program at the LNP had already decided on the model for resettlement behind closed doors. This echoes other cases that have been documented (Nikuze, Sliuzas, and Flacke 2020). People would be resettled as nuclear families, the village would be organized in square plots, and there would be no room left for future generations of extended family. Each family would get 1ha of rainfed land despite the number of hectares they had in their original location (up to 50 has in the families with the most land) or the amount of land they would need to produce enough food to feed their families (Milgroom and Giller 2013). Any land holdings over 1ha would be compensated for in cash. The livestock would be kept outside of the village and grazing land was to be shared with the host village. Despite requests by displaced residents, negotiating access to more land for housing, agriculture and grazing was politically very sensitive and out of the question for local government officals (Milgroom 2015).

In the two spaces for formal participation, the CCR and the GTR, the resettlement committee invited displaced people to participate in making decisions about their futures only to discuss unimportant details, such as the door handles of the compensation houses. Government officials portrayed village representatives as unintelligible and their requests for more land, for example, as unnecessary, impossible or impractical. In our view, the harm caused by eliciting opinions and then disregarding them is a form of slow violence hidden behind the well-meaning proposal of participation.

Some government officials, especially the members of the resettlement committee who interact regularly with the residents of the park and had established relationships with leaders, recognized the displaced people as very intelligible. Yet, they were stuck as employees expected to carry out the desires of the state (Lunstrum 2013; Rew, Fisher, and Pandey 2006). The ‘success’ of epistemic violence was precisely in these good relations that the resettling committee kept with the displaced people. They acted and spoke as if they wanted the best for the displaced people and were working hard to get the best compensation possible. When things didn’t work out, they were able to save face and blame the higher up government officials or the donors for the decisions made.

By delimiting the negotiations and focusing the ‘participatory’ process only on the houses, the resettlement committee managed to uphold an image, a performance, of participation for the donors. But participation in decision-making on key issues impacting the lives of residents did not take place. As one leader said, ‘Decisions are made by people coming from the outside and no one listens to our suggestions.’¹¹ This charade did not address the important livelihood needs of displaced people and prepared the ground for resistance and contestation.

Coercing: manufacturing consent through ‘proving participation’

KfW had found out that the Mozambican government officials had deceived them when they said that villagers preferred cement houses. Therefore, to avoid this occurrence again, KfW insisted that the park staff prove village residents’ approval of the compensation house before they were built. To this end, KfW requested that two model houses be built so that residents could see, comment and ultimately provide their consent to the houses before the remaining houses were built. Village leaders and residents from close-by villages visited the model houses (see Figure 3, photo 1 in which the village of Nanguene is visiting the model house to assess it accompanied by an English consultant and two LNP staff) and discussed their views at the fifth CCR meeting. Government officials understood the donor’s ‘no-objection’ procedure to mean that the park staff had to collect signatures from community leaders proving that the village residents had approved of the houses and the pressure was on to attain them. The resettlement project in the LNP was a highly visible initiative with considerable political implications, represented by the houses (see Figure 3, photo 2 in which the president of Mozambique Armando Gebuza was visiting the model houses). One LNP official noted: ‘The provincial administrator (DA) was sent here and told not to come back until she had signatures of the leaders agreeing to the houses. She didn’t even change clothes for three days because of it.’¹² In the meeting, all of the leaders were asked to sign a document that was written in English, not in their language, that stated that they approved the model houses when, in fact, they had not approved. As a unified front, all of them refused to sign the documents.

Figure 3. Photo 1: Residents of the village of Nanguene evaluated the model house with the external technical advisor and two LNP staff members from the resettlement committee. Photo 2: The president of Mozambique Armando Guebuza visited the model houses. Both photos are from 2007, taken by J Milgroom.

The traditional form of governance of the villages living in the LNP entailed making decisions with a group of elders. The leaders very clearly expressed their discomfort in the way they were being asked to make decisions because they needed to be able to take a question back to their village, decide together with elders and bring the final decision to the next CCR. Although the traditional forms of governance were upheld in different ways in different villages, none of the leaders wanted to be blamed for making decisions singlehandedly. Village leaders requested not to be the messengers, but this request was not heeded. As one leader said, ‘The government wants us out, so they need to do what we say. We are having problems with conflicts in our communities – we have to bring the news back and forth from the communities.’¹³ In this case, their ‘participation’ began to resemble resistance.

The park staff, including the PPF park manager who only got involved in daily resettlement decisions when it was important, decided to change strategies: to divide and conquer, to meet with one village at a time.¹⁴ As one report mentioned, ‘because the CCR puts together the willing and the opposing leaders, there is a negative influence on those who want to see the process advancing.’¹⁵ The resettlement committee staff held an open village meeting the next day in the village closest to the park headquarters, Macavene, the second of the villages slated for resettlement. The intention of the meeting was to provide enough information to the village members to convince them to agree to the model houses. Although some people were eager to be resettled and move on with their lives, they all wanted larger houses, or became wrapped up in the leverage they found through refusing to accept the model houses. Not only did they unanimously refuse to agree to the model houses, but some also requested they have a second door, others complained about the process of resettlement, for example, of building houses before opening the fields, and still others complained about the content of the proposed compensation package and accused the park staff of making promises that they are not keeping. The resettlement committee tried to get the village members who were present to sign the agreement at the end of the meeting, but no one agreed to sign it.

The park staff left with long faces to ‘inform the bosses’ of the results of the meeting.¹⁶ In a document written in English – for the donors in Germany – summarizing the process of deliberation about the model houses it was stated that there was a need for ‘superior decision making to overcome this impasse.’¹⁷ In order to overcome the ‘impasse’, LNP staff called in reinforcements to attend meetings with Macavene and Nanguene, each in their separate villages: a World Bank representative and higher-level politicians. LNP staff explained: ‘We needed to just draw the line somewhere or it [community resistance] will never stop.’¹⁸ In the meeting in Nanguene, a ‘white man’, the WB representative, told a story about people living in a newly established park in Cameroon who had refused to move out. As the wildlife intensity increased, people from the village were killed. He mentioned that they would be restocking the Limpopo park with animals too. The story deeply impacted the residents of Nanguene. A villager described the meeting like this:

Everyone was refusing to leave. We were sad and upset and the white man said that the houses are very nice and that they are well-made and he would like to have one for himself. He said they were going to release more animals here. When we understood that they were going to release more wild animals close to our village we saw that it was not worth it to stay here. We decided to accept [to be resettled] because of fear that our children will get killed by animals.¹⁹

That day the leader and other elders of Nanguene signed the document, agreeing to the resettlement, and to the model houses. KfW accepted the ‘proof’ of resident ‘participation’ and the rest of the resettlement houses were built for the village of Nanguene.

When delegitimizing and delimiting gave rise to resistance by leaders and village residents, government staff proceeded to use coercion. The political pressure from high up in the Mozambican government was intense to move the resettlements along. It is important to mention that PPF was also instrumental in encouraging this shift towards the more forceful tactic of coercion. Although the PPF park director tended to stay in the shadows, PPF is well-known for their low regard for local communities (Draper, Spierenburg, and Wels 2004) and their influence on the Mozambican government’s position on participation is not to be underestimated.

As Arnall (2019) described from a case in Northern Mozambique, the officials used deception, pressure to acquiesce and playing upon people’s emotions as forms of manipulation to get people to provide consent, but they also used more explicit threats (releasing lions) and bribes (nice houses), creating an overall effect of disabling voice through coercion. Nanguene was finally resettled in November of 2008, but each successive village learned from the first resettled villages both about the hardships of resettlement and that they could successfully resist certain conditions by refusing to consent to resettlement. One man refused to resettle from the village of Macavene and remained there with his family alone, causing a headache for park managers (Otsuki et al. 2017).

Disengage: change of guiding policy, reinterpretation of participation and discourses of sovereignty

Tired of having to deal with petitions from displaced people, and tense relations with donors, representatives from the provincial and national governments proposed jointly²⁰ that the WB OP 4.12 would be replaced with the National Institute for Disaster Management (Instituto Nacional de Gestão de Calamidades – INGC) resettlement policy, one used to resettle people from flooded areas almost every year. In the INGC model, residents are given materials to build their own houses while living in tents provided by the government, and there is no participation of displaced people in the decision-making process. This idea was presented to village leaders in 2008, in the sixth CCR meeting. The Governor, district administrator (DA), a representative from INGC and administrators from neighbouring districts were all present.

Government officials claimed three main reasons for switching policy frameworks. They argued that resettlement was taking too long and that in the meantime people were suffering from wild animals and the park project could not proceed. Ironically, the second argument was that WB OP 4.12 was a foreign imposition. As one LNP staff member explained: ‘Mozambique has experience with resettling people – we should adopt the INGC model because it is something that comes from here – a home-grown model.’²¹ Narratives of separation from foreign donors were used to try to convince people that the new model was for their own benefit: home-grown, African and self-made. Thirdly, they said that if people constructed their own houses, their ownership of the outcome would increase. The district administrator said: ‘People were consulted, but us Africans, we take much better care of something we made ourselves over something that we bought or was given to us.’²² Her conception of participation was that people would ‘participate’ by constructing their own houses.

Village leaders rejected the proposal, claiming that other resettled villages had their houses made for them, and that this was not fair. One village leader insisted, as in previous meetings, by saying, ‘If you want us to leave our land, you must resettle us as we say.’ This infuriated the governor who cut the meeting short, insulted the leadership capacity of the village leaders and insisted that they had all agreed upon the approval of the INGC model for resettlement houses.²³ Despite the fact that no one had agreed to anything, the governor ended the meeting by saying, ‘Thank you for the patience. We have on our agendas that we should get tents because you accepted the INGC model.’ This was a crackdown on participation and evidence of entrenchment of authoritarian practices. However, it was also an extension of party-state sovereignty. Lunstrum (2013) illustrates how the Mozambican government was able to articulate state sovereignty into the borderland of the LNP through the resettlement project by gestures such as this one of expecting the village leaders to be extensions of the state apparatus and obey orders.

KfW could not agree to fund resettlement under the INGC model. Therefore, in 2010 the Mozambican government decided that they themselves would temporarily pay for resettlement so that they didn’t have to abide by the KfW guidelines.²⁴ This strategy was approved by the council of ministers in April 2011. As a result, the Massingir district government together with INGC was given the responsibility, previously that of the LNP, to implement resettlement. Despite this shift of control over the resettlement process, the South African director of the park who worked for and represented PPF still remained involved with the resettlement process, as did the resettlement committee at the LNP, but the decisions were made jointly with the district administrator and the provincial governor. This strategy of switching to a policy that is less beneficial for displaced peoples has been documented in a conservation area in India, where Kabra and Das (2022) show how abiding by a voluntary resettlement scheme was a strategic response to avoid other laws with more safeguards to protect displaced people.

Displaced villagers and leaders continued to refuse to agree to building their own houses and government officials could not force them to do so. Authoritarian practices failed to control outcomes in this context. Resentful of the strength of the village leaders’ resistance, the district government came up with a third strategy for the construction of houses in which teams of villagers from the area, including the host-villages, were taught the necessary skills and were hired to build the houses for the second and third villages to be resettled, Macavene in 2013 and Massingir Velho in 2015. INGC only built houses but did not take care of the reestablishment of livelihoods, provide fields for agriculture or consider the sustainability of the resettlement conditions.²⁵

While the INGC model disengaged from participation, government officials continued to use the narrative of participation to ease resistance and uphold a performance of involving displaced people in the process of resettlement. However, the term was depoliticized and reinterpreted to suit their needs. ‘Participation’ was reduced to committees of people from the displaced villages who were transported to the new village regularly before resettlement to visit and sweep the new houses and report back to their village. A staff of the resettlement program from the district of Massingir government said, ‘Negotiation is that conversation in which both parties reach consensus about them leaving their land. You can’t do anything in the communities without negotiations. We can’t do anything without consulting the communities. It is necessary to negotiate … but they end up understanding that they must leave their homes.’²⁶ This quote indicates how the concept of participation entails a slow way of forcing people out. This stripped the already threadbare tapestry of participation to a minimum: to avoid resistance and move people as quickly as possible.

No more villages were resettled until houses were finalized for the resettlement of Makandazulu in 2021, managed primarily by the Mozambican government. In 2022, the village of Bingo was resettled.²⁷ This new wave of resettlements was ushered through in part because of a rhino poaching crisis (Witter and Satterfield 2019). Current residents of the park have been living under a situation of considerable vulnerability. Some of the remaining population have resisted resettlement knowing that the post-resettlement conditions were sub-optimal. Some, however, are still waiting to be resettled and stuck in what has become a 20-year limbo in which they are not provided with basic services or opportunities or communication with park staff (Witter and Satterfield 2019).

The resistance of village leaders and members in negotiations showed considerable agency potentiated through the process of induced participation. However, a careful look at what they resisted indicates that they focused their protest on potentially attainable material gains within the learned epistemic frame of the project. This is one impact of epistemic violence; they did not officially protest the fundamental basis of the compensation that most deeply affected their livelihoods and future generations, such as not getting enough land, fields for dryland agriculture, or extra residential space for extended families. Rather, they contested having to build their own houses, and the size of the houses, all details already sanctioned by the project.

Despite this, as the quote at the beginning of the article states, the contractions of induced participation were too strong and the participation baby was sacrificed to save the mother. Governmental pushback against village resistance caused an entrenchment of authoritarian practices, infliction of epistemic violence and almost total shut down of participatory spaces. When the imposed participation began to infringe upon the nation state’s sovereignty and cause an upset of the state-citizen power relations, the government disengaged from participatory policy, returning as much as possible to status quo authoritarian rule while upholding the persuasive mask of tokenistic participation. This was the most drastic form of epistemic violence, an abandonment of the attempt or intent, however performative, to listen.

Conclusion

Most scholars and conservation practitioners that look for ways to minimize the harm caused by displacement conclude that participation must be a central part of resettlement projects. This paper questions this widespread, normative assumption that participation can lead to better post-resettlement conditions. According to theories of (unattainable) ‘good’ resettlement practice, initially the LNP resettlement project had everything in place – the political will, resources, policies, laws and enforcement – to enable a meaningful participation of leaders and village residents in shaping their own resettlement. However, induced participatory spaces exposed displaced people to epistemic violence and led to entrenchment of authoritarian practices. This paper documents four strategies used by Mozambican government officials in the LNP to disable displaced people’s voices: delegitimizing local people’s ways of being and knowing, delimiting what could be negotiated, coercing people into providing consent and finally disengaging from participatory policy. These strategies enabled the government to pushback against the political implications of participation.

Government officials initially encouraged displaced people to express their needs and desires in invited spaces of induced participation. They did so because of donor requirements: according to World Bank policy, displaced people had to be invited to participate in decisions about their resettlement. Participation created an intractable tension in which government officials were supposed to open up decision-making spaces for the very people who were ‘in the way’. In the ensuing participatory process, government representatives ignored displaced people’s petitions, casting them off as irrelevant and inappropriate. Displaced people were delegitimized, inadvertently or purposefully deemed unintelligible and their perceived capacity as valid knowledge holders was questioned. On one hand, the dismissal of their demands was made possible by the othering process that depicted displaced people as ‘poor’, ‘uneducated’ and ‘backwards’, and some government officials treated them as if they did not know what was best for themselves. On the other hand, their opinions were dismissed because they did not align with the government’s notion of ‘development’. Upholding the image of development as represented by cement houses and square plots was part and parcel of the continued justification of the displacement itself. While the creation of induced participatory spaces exposed displaced people to epistemic violence, this violence was masked by the good intention of bringing them (a predefined, western notion of) ‘development’ as well as the well-intended proposal of participation.

As displaced people were expressing considerable agency and resistance, government authorities perpetuated the participation rhetoric but delimited decision-making to only include minor details. People were invited to participate in decisions about the type of door handles but could not discuss or influence the design of what was actually important for their livelihoods, such as access to land for agriculture, grazing and the spatial organization of their intergenerational households. Controlling what was possible to be discussed and deeming key topics off limits was another form of epistemic violence – there were invited spaces for ‘participation’, but not about the things that matter. Even so, nominal participation fostered resistance rather than compliance. This led government officials to coerce residents into consent through bribes, manipulation, and threats to attain signatures from village leaders to prove to the donors that the village residents had in fact ‘participated’. In an attempt to quell any future resistance, the state disengaged from participatory policy as much as possible, while reinterpreting participation to mean building the houses, engaging in menial tasks or as a part of the negotiation process that would inevitably end in displacement one way or another. Government officials used a completely depoliticized narrative of participation to ease resistance. Displaced residents resisted government pressure, but their protests were, for the most part self-constrained to relatively minor material gains they perceived to be achievable within the learned epistemic frame.

This exploration of the limits and downsides of ‘participation’ in resettlement planning points to two conclusions. First, the inherent tensions of participation in conservation-forced displacement are intractable because of the assumptions upon which displacement is based: the non-negotiable superiority of hegemonic development and the imagined unintelligibility of people slated for displacement. It is precisely the ‘othering’ that is inherent in displacement that negates the possibility for meaningful participation to occur. We posit that it is not possible for displaced people to ‘participate’ without becoming victims of more violence. The epistemic violence of participation is piled on top of the violence of displacement. Second, participation does not only expose people to epistemic violence but legitimizes epistemic violence as an acceptable practice. Participation masks the infliction of epistemic violence under the good intention of asking people about how they want their lives to be, even if these desires are not heeded. Unquestioned, participatory rhetoric legitimizes the harmful practice of displacement and resettlement. Participation is not the answer. Rather than focus on how to arrange conditions for improved participation, we must work towards decolonizing conservation and exploring radical alternatives to the neoliberal and capitalistic development model that causes and relies on displacements in the first place.

Authors’ contributions

Data collection and analysis: JM. Conception and design of the paper: JM. First draft of the paper: JM. Revisions and edits: JM and PC.

Acknowledgements

Jessica is grateful to the LNP staff and especially the residents of Nanguene and Chinhangane for their trust and many hours of long conversations. She would like to thank Brooke Wilmsen, Chris de Wet, Janice Jiggins, Jens Anderson, Cees Leeuwis, Petra Derkzen, Niels Roling, and Steve Sherwood and for input on earlier iterations of this paper. Jessica and Priscilla would also like to thank two anonymous reviewers for their constructive feedback.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by HORIZON EUROPE Marie Sklodowska-Curie Actions [Grant Number 844637]; National Science Foundation; and Wageningen University and Research Centre.

Notes on contributors

Jessica Milgroom

Jessica Milgroom (she/her) is an Assistant Professor at the Centre for Agroecology, Water and Resilience of Coventry University (UK). Her research focuses on understanding how systemic oppression and exploitation has shaped, and continues to shape, food systems and people’s relationships with nature. She works on agroecology, food sovereignty and settler colonial studies, as well as ‘development’-forced displacement. She explores what decolonial and feminist perspectives reveal for collective construction of powerful, generative narratives.

Priscilla Claeys

Priscilla Claeys is Associate Professor in Food Sovereignty, Human Rights and Resilience at the Centre for Agroecology, Water and Resilience of Coventry University (UK). Her research areas include the right to food and food sovereignty, the right to land, agrarian movements and global food governance. Priscilla is also passionate about feminism, gender equality and ways to encourage inclusion and diversity in policy-making spaces. She is the President of FIAN International.

Notes

1 LNP staff, Maputo, 27 May 2010.

2 Kunming-Montreal Global Biodiversity Framework, UN Convention on Biological Diversity, CBD/COP/DEC/15/4, 19 December 2022. https://www.cbd.int/doc/decisions/cop-15/cop-15-dec-04-en.pdf

3 WB policy, including its participation clause, quickly became the globally accepted standard upon which other polices were modelled such as the Asian, African the Inter-American Development Bank’s policies, International Finance Corporation (IFC) Performance Standards, as well as national resettlement policies around the world. The current policy language in the ESF is stronger, indicating the requirement for affected communities to be engaged from early in the process of project design through to post resettlement.

4 Interview with KfW representative, Massingir, April 2007.

5 See the original text of the WB OP 4.12 2001 for the full list of principles (World Bank 2001).

6 Minutes from first CCR meeting 2004.

7 Minutes from the second CCR 2004, page 9.

8 Interview with KfW representative, Massingir, June 2007.

9 Interview with KfW representative, Massingir, April 2008.

10 GTR meeting May 2 2007.

11 GLTCA meeting, village leader, March 6 2007.

12 LNP staff, Massingir, September 2007.

13 CCR meeting, March 2007.

14 Conversation with park staff, Massingir, March 2007.

15 LNP’s 2007 Report on the Consultation Process for Model Houses Approval, page 3.

16 Meeting in Macavene March 2007.

17 LNP’s 2007 Report on the Consultation Process for Model Houses Approval, page 3

18 Technical advisor, Massingir, 15 May 2007.

19 Nanguene resident, Nanguene, 24 April 2007.

20 CCR Meeting and interview 2010.

21 LNP staff, Massingir, 5 August 2008, and MiTur official, Maputo May 2010.

22 Massingir District Administrator, Massingir, 14 June 2010.

23 CCR meeting 2008.

24 CCR 6 November 2012, meeting minutes.

25 LNP resettlement committee September 2016.

26 District resettlement committee staff, June 2014. Part of this interview can be seen in the documentary film Orphans of the Land (Milgroom 2017).

27 https://www.rm.co.mz/gaza-inicia-reassentamento-de-familias-residentes-no-parque-nacional-do-limpopo/