The anti ‐ politics of sustainable development: Environmental critique from assemblage thinking in Bolivia

In this paper I argue that assemblage theory provides an innovative way to extend critique of sustainable development as it is being remade by the 2015 United Nations Sustainable Development Goals (SDGs). Drawing on recent fieldwork in Bolivia, I examine the early take ‐ up and implementation of the SDGs in a site of intensifying resource extraction and struggles for radical development alternatives. I foreground the assemblage of institutions, discourses, landscapes, and infrastructures that are at once disciplined and held together to materialise and legitimise particular interpretations of sustainable development. This helps highlight what I term the “ lost geographies ” of the assemblage. Based on this analysis, I argue that the SDGs as assemblage act as a form of anti ‐ politics by rendering neutral and apolitical the conflictive politics of extractivism. As global momentum to combat climate crisis and environmental crisis grows, such assemblage work helps explain how powerful, extractivist development logics are nevertheless being maintained and reworked. power right, and rise in state vio-lence and in the dogged promotion of the Bible – all for how indigeneity will be recast in the coming months. This paper provides insight into the legacies of the MAS administration for the politics of extractivism, sustainability, and territory, including an analysis of how disparate national political agendas connect with contemporary constellations of global development.


| INTRODUCTION
In 2015, sustainable development was launched as the guiding concept behind the United Nations Agenda 2030 and their Sustainable Development Goals (SDGs). Seventeen Sustainable Development Goals (SDGs) form a part of Agenda 2030, which unites development banks, multilateral and bilateral institutions, states, international non-government organisations (INGOs), civil society organisations, and, increasingly, private sector actors in balancing social, economic, and environmental arenas within a "plan of action for people, planet and prosperity" (UN, 2015, np). The shift to "Sustainable Development" constitutes the response of mainstream development actors to global warming and intensifying climate change. The concept of sustainable development, however, is not new and has already been much critiqued as ineffective within critical social science (see Adams, 2001;Fletcher & Rammelt, 2017;Redclift, 2005;Svampa, 2015). In this paper, I draw from recent fieldwork in Bolivia to argue that an assemblage reading of the early take-up and implementation of the SDGs offers something new to how we appraise the environmental remit of sustainable development. While global development goals are often studied as a global policy framework (Hulme, 2009) or evaluated in terms of their effectiveness (Hulme, 2009), data capture (Jerven, 2013), and North-South development politics (Bond, 2006), an analysis of how an assemblage of institutions, development discourses, landscapes, and energy infrastructures is being disciplined and held together reveals through modulating topologies of relations between human and non-human agents across space and timealways in flux, in formation, and being undone (Guattari & Delueze 2000). In geography, this thinking has proved an important way to analyse the production of the social and interrogate the relations between stability and transformation (Anderson et al., 2013;. In this paper, it is used to analyse the production of the socio-environmental and examine the relationship between socio-environmental stability and transformation as sustainable development is materialised, institutionalised, and legitimised. Assemblage thinking is well-suited to an analysis of the SDGs, which constitute a consensus-based political technology to promote a common ethos for all. An assemblage lens enables a deeper understanding of how sustainable development is being remade from existing practices, partnerships, and relationships of power, as well as how the categories of "development" and the "environment" are being reworked as the assemblage comes into being. Following Li, this enables analysis of how certain problems and solutions "become thinkable" as the SDG assemblage becomes "stabilized as discursive formation" (2005, p. 386).
In so doing, this paper contributes to work that analyses how global development agendas operate in Bolivia specifically, as a site of radical agendas, contentious political practices, and attempts to create political spaces and practices that go "beyond capital and beyond the state" (Aguilar, 2014). In her work on global development programmes for conflict resolution in El Alto, Bolivia, anthropologist Susan Ellison argues that imported ideas about citizenship and democratic participation diverge from local social relations and modes of political engagement and conflict to narrow political-economic imaginations (2018, p. 29). For the consensus-focused SDGs, this line of enquiry exposes what is lost as agreements on "sustainable development" are secured, specifically what agendas are being disciplined and diluted.
In geography, assemblage thinking has driven an empirical focus on "how spatial forms and processes are themselves assembled, are held in place, and work in different ways to open up or close down possibilities" (McFarlane & Anderson, 2011, p. 172). In this paper, this plural, complex, and relational mode of thinking is extended to the environment-development nexus to question the ways in which hegemonic interpretations of sustainable development emerge in relation to particular (and contested) landscapes, development discourses, institutions, social movements, histories, energy megaprojects, and infrastructures. To do this, I draw on the work of Li (2005Li ( , 2007 and Foucault (2012). Li, to foreground the "hard work required to draw heterogeneous elements together, forge connections between them and sustain these connections in the face of tension" (2007, p. 2) and Foucault, to reveal how power operates within the assemblage. Disciplinary power operates as a mix of self-regulation, self-governance, and repression (Foucault, 2012;Rabinow & Rose 2003;Fraser, 2003), useful for understanding social ordering beyond legislation. Though Foucault mapped the contours of disciplinary power within the nation-state, it remains relevant to analysis of transnational governmentality in this era of global development (Fraser, 2003). First, an analysis of the capillary modes of disciplinary power (analysed through the early implementation and uptake of the SDGs on the ground) informs our understandings of how the systemic dynamics of sustainable development are created and maintained. Second, an analysis of how NGOs encounter and treat territorial movements reveals how diverse NGOs self-regulate to support hegemonic interpretations and practices of sustainable development, despite disparate organisational agendas (see Fisher, 1997). The subsequent analysis reveals that the emergent SDG assemblage in Bolivia arises from "an existing repertoire" of linkages between development institutions, actors, and discourses and is a process begun from "habit, accretion, and bricolage" (Li, 2007, p. 4). However, it also reveals the disciplining involved in assembling sustainable development in Bolivia, to close down more radical interpretations.
Critically, the non-human elements of an assemblage "do not simply provide a context in which unequal social processes occur; rather, they are agential actors in their own right" (Ranganathan, 2015(Ranganathan, , p. 1304. In urban geography, the work of Deleuze and Guattari has been used to reveal and analyse cities as complex relational "topologies of many intersecting entities, networks, and flows whose combined force shapes urban development" (Amin, 2015, p. 243). This combats theoretical framings that filter out "much of the richness of place … [the] plurality of things happening, the changes occurring over the course of the day, the many different pursuits and affects of the people gathered, the resonances of the still architecture and silent infrastructure, the amplifications of the amassed bodies and entities" (Amin, 2015, p. 241). Here, this approach includes the materialities of sustainable development, specifically the ways that particular geographies direct the content and transformative potential of the SDGs. In a context of intensifying resource extraction, this extends our analysis to include the physical manifestations of post-neoliberal neo-extractivism (the roads, bridges, infrastructure, and megaprojects) and consider how these co-constitute what is seen and recognised as sustainable development. In doing so, this dynamic of assemblage thinking reveals how development and the environment are brought together in an uneasy union within sustainable development discourse.
Together, work on assemblage and Foucault's theories of disciplinary power reveals how the SDGs work as a form of anti-politics, meaning as a process of translating social and political struggle into neutral technical problems that are subsequently bureaucratised and drastically depoliticised (see Ferguson, 1990). This theoretical frame reveals that it is politics, partnerships, power dynamics, and context that drive and inform the early take-up and implementation of the SDGs and foregrounds the critical importance of sustainable development discourse in guiding the take-up of specific goals and targets.

BOLIVIA
The 2015 UN SDGs combine a response to climate change with development targets, comprised of 17 goals, quantified by 169 targets, and measured by 230 indicators (UN, 2015). The SDGs are a timely response to calls to greatly reduce the human environmental footprint, found in debates about climate change (IPCC, 2018), biodiversity loss (Apostopolopoulou & Adams, 2015;Rockström et al., 2009), and the Anthropocene (Bonneuil & Fressoz, 2016;Crutzen, 2006). The concept of sustainable development, however, is old and much critiqued. Since its mainstreaming by the 1987 Brundtland Commission, the concept has been rejected as an insufficient and ineffective resolution to tensions between development and the environment, with recent critics arguing that continuing environmental degradation is proof that pursuing sustainable development has not worked (Asare et al., 2015).
While the Brundtland Report defines sustainability as "development that meets the needs of the present without compromising the ability of future generations to meet their own needs" (World Commission on Environment & Development, 1987, np), there remains no consensus on how to measure or operationalise such a definition (Liverman, 2018). Seen as "woefully vague," it has been cast as an unsuitable basis "for either theory or action" (1994, pp. 541-556) and as an empty signifier (Brown, 2016;Laclau, 1996). The vagueness of the term, however, makes it politically malleable and "attractive like apple pie" (Terborgh in Adams 2004, p. 176) precisely because there is no technical definition. It can thus be appropriated by groups and practices that are neither environmentally sustainable nor particularly developmental (Luke, 2005;Redclift, 2005). Further, its widespread applicability casts everyone as victims of depoliticised environmental decline, concealing the specific people, practices, and places that are causing socio-environmental degradation (Swyngedouw, 2010). This depoliticising dynamic of sustainable development is analysed throughout this paper, for its effects on a site where demands for resource extraction threaten Amazonian forested territories and struggles for territorial sovereignty.
A core rationale for the goals (and Agenda 2030) is to promote growth-led development. This is reflected in the goals themselves, which include Goal 8 for "Decent Work and Economic Growth" and Goal 9 for "Industry, Innovation and Infrastructure" (UN, 2015). It is also reflected in the deepening of partnerships between the private sector and development institutions. For Mawdsley, the SDGs help to "normalize a radical shift in development finance," as private finance is being invited to guide, shape, and fund Agenda 2030 (2018, p. 191). Private sector representatives are being asked to advise global development governance and policy by the United Nations, other multilaterals, and national development agencies (Mawdsley, 2015(Mawdsley, , 2018. The SDG Mining Atlas, for example, is addressed to the mining industry and promotes the "shared belief that the mining industry has an unprecedented opportunity to mobilize significant human, physical, technological and financial resources to advance the SDGs" (UNDP, 2016, np). For Mawdsley, deepened partnership with the private sector supports and consolidates a return to the central analytic of economic growth over poverty reduction (2018). Despite these shifts, however, the SDGs are promoted as newly global in orientation and the result of a much extended consultation process across the global North and South (UN, 2015; see Horner & Hulme, 2019). They are promoted as for the state, civil society, and the private sectorwith little (if any) acknowledgement of sites of conflict between these groups (see Hope, 2020).
In the early 2000s, moves towards pluri-nationalism and post-neoliberalism marked Bolivia and Ecuador as the most radical countries in Latin America's Pink Tide (move left) (Goodale & Postero, 2013;Gudynas, 2011;Kohl & Farthing, 2006). In the 2009 Constitution, the Morales government enacted a new development discourse as the guiding ideology of the state, one drafted from multiple indigenous cosmologies and articulated in critique of mainstream development discourse. Vivir Bien (Buen Vivir in Ecuador and elsewhere) decentres economic growth to instead prioritise wellbeing and harmony within communities and with the non-human (see Gudynas, 2016;Walsh, 2011). Such a significant shift, to recognise indigenous ontologies of nature and development within state apparatus, strengthens regional debates about plurality and the decolonisation of Nature in Latin America (see de la Cadena, 2010;de la Cadena & Blaser, 2018). In its implementation, however, Vivir Bien has been curtailed by the government's neo-extractivist development model.
In Bolivia, the profits from resource extraction have funded welfare programmes and brought significant reductions in extreme poverty and inequality (ONU Bolivia, 2018, p. 9). For Álvaro García Linera, radical Marxist scholar-activist turned Vice-President, "extractivism is not a goal in itself but it can be the starting point for overcoming it" (2012). The Morales government committed to a Process of Change (proceso de cambio) to strengthen the state's capacity to mediate global HOPE capitalism for the benefit of Bolivian citizens (Cavooris, 2015;Feldman, 2015). Public spending increased following a new hydrocarbon tax and the royalties were distributed to the military, to state universities, to newly founded indigenous universities, to pensions, and to cash transfer programmes (Fabricant & Gustafson, 2016). For these reasons, extractivism has been understood as a defining feature of post-neoliberalism (Veltmeyer & Petras, 2014;Gudynas, 2016). The vast majority of public investment, however, was allocated to energy and hydrocarbon development (40%) and transportation infrastructure (29%), accounting for nearly 70% of the 2016 public investment budget for 2016 (MEFP, 2015in McKay, 2017. This revealed that social capital investments prioritised further accumulation, particularly in the extractive sectors (McKay, 2017). Redistributive policies also failed to reflect a serious commitment to diversifying the economy and the control and distribution of royalties remained highly centralised, with minimal contributions made to activities that directly supported indigenous groups (Radhuber, 2012). Rural class structure similarly contradicts claims of significant, egalitarian reform, as the agro-industrial elite have retained power and land (Webber, 2017;McKay, 2019).
The Morales government, however, was dismissive of opposition to their development model. Linera, for example, drew on Bolivia's conflictive history with foreign development institutions (see Petras, 1997) to dismiss anti-extractive protest as the "hyper-environmentalism" and meddling of foreign NGOs (Linera, 2012), despite evidence to the contrary (Hope, 2016). Instead, the Morales government invoked a discourse of resource nationalism, "where the imaginaries of resources intersect with notions of rights, identity and citizenship" (Childs, 2016, p. 540). In Bolivia, resource nationalism has underpinned expanding extractive frontiers and stifled dissent, as well as shaped the frames of social movement responses (Pellegrini, 2016). In certain cases, where territories overlap with the country's biggest reserves, hydrocarbon extraction has led to forms of hydrocarbon citizenshipwhere citizenship claims are intertwined with the political economy of hydrocarbon extraction (Anthias, 2018).
Bolivian commitments to neo-extractivism form a part of an unprecedented rise in resource extraction across Latin America since the early 2000s, when a resource boom was exploited by both neoliberal and post-neoliberal states (Bebbington 2009;Bebbington & Humphreys-Bebbington, 2011). Latin American extractivism is defined by the Argentinian sociologist Svampa as a "pattern of accumulation based on the overexploitation of generally nonrenewable natural resources, as well as the expansion of capital's frontiers toward territories previously considered nonproductive" (2015, p. 66). For Svampa, there has been a switch from "the Washington Consensus, with its focus on finance, to the Commodities Consensus based on the large-scale export of primary products," constituting "a new economic and political order" sustained by a surge in global demand for raw materials (Svampa, 2015, p. 117). The impacts of resource extraction and global market demand go beyond commodity pricing and geographies of available capital. For Arsel, Hogenboom, and Pellengrini, international cycles of boom and bust "constantly affect national politics and policies as well as local dynamics around extractive activities" (2016, p. 1). As extractive activities become central to state development models they are prioritised over all other agendas to constitute an "extractive imperative" (2016, p. 1). Consequently, extractivism has significant socio-political impacts, for example influencing the recognition and political power of indigenous environmentalism in the TIPNIS (Hope, 2016).
Contemporary dynamics of extractive capitalism cause environmental degradation (Svampa, 2015). The environmental costs of extractive infrastructure across the Amazon, namely roads, railways, and ports, are also detrimental and cause deforestation (Bebbington et al., 2018). Such environmental costs have motivated new oppositional movements across Latin America, constituting a new eco-territorial turn that contests the impacts of the commodities consensus on land, territorial rights, and conservation (Gonzales, 2014;Svampa, 2015). In Bolivia, the socio-environmental consequences have been severe and geographically specific. In 2011, the government opened up conservation areas for extraction, eroding environmental protections that preceded the election of Morales. Since 2015, the government promoted the country as the "energy heart of Latin America," extending commitments to resource extraction with new hydropower dams and plans for fracking (Ministerio de Hidrocarburos, 2015). Where new sites of extraction or related infrastructure cross or border territorial boundaries, territorial rights have been undermined. The interim government shows no signs of lessening commitments to extraction and have already commissioned scoping studies on fracking (Pagina Siete March, 2020).
As a site of hydrocarbon and resource wealth (entangled with both transnational extractive capital and regional consumers), Bolivia is thus instructive for the much broader challenge of fossil fuel production, commodification, and consumption. How this challenge is addressed by global environmental agendas, namely the SDGs, has been researched in two periods of fieldwork between September and November 2017 and in April 2019. These built on the knowledge and networks gained during 9 months researching the TIPNIS conflict between 2011 and 2012. I carried out semi-structured interviews and participatory mapping with each participant (snowballing from existing networks), to ascertain the organisational and campaign networks through which territorial movements operated (Fontana et al., 2005;Longhurst, 2010). In addition to the 57 interviews during 2011-2012 on the TIPNIS conflict, in 2017 and 2019 I carried out a total of 23 interviews with indigenous leaders, activists, and civil society organisations and completed 19 maps of interviewee activist networks. I then interviewed eight international and national development NGOs, one bilateral institution, and three government ministries about their uptake of the SDGs and engagement with territorial movements. As participant observation, I attended three protest events and two meetings of lowland indigenous organisations, and stayed for two weeks with an activist in La Paz, which enabled observation of local activist practices and their links to the selected case studies.
The two cases studied were the conflict over road building to access hydrocarbon pools in the Territorio Indígena y Parque Nacional Isiboro Sécure (TIPNIS; Isiboro Sécure Indigenous Territory and National Park) and resistance to hydropower dams in Chepete and Bala. The TIPNIS spans 1.2 million hectares of tropical forest and two Amazonian tributary riversthe Rio Sécure and the Rio Isiboro. It was one of Bolivia's first national parks, only recognised as the historical territory of Tsimané, Yuracaré, and Mojeño-Trinitario indigenous communities after the 1990 lowland indigenous March for Territory and Dignity. Since the 1990s, it has been co-managed by SERNAP, the state ministry for conservation, and by the indigenous government. In 2011, the Morales government started building a road through the territoryplanned to run past two large hydrocarbon pools (with concessions already granted to a quarter of the territory). TIPNIS communities were not consulted, in adherence with ILO 169 Right to Free, Prior and Informed Consent. In 2011, the TIPNIS government voted against the road and started a month-long march to La Paz, demanding the road be stopped and the territory be protected.
Arriving in La Paz, the TIPNIS leaders met with Morales, who agreed to stop building the road and to protect the park. Celebrated as a victory, the marchers returned home (Bautista et al., 2012). However, a number of U-turns have prolonged the conflict and muddied its effects. The Morales administration's plans for the TIPNIS, viewed as a violation of indigenous and territorial rights by TIPNIS supporters, caused the disintegration of the 2005 Unity Pact (an alliance of five indigenous social movement organisations that was created to inform and guide governmentunderpinning Morales' claim to head a government of social movements; see Postero, 2010;Webber, 2017). The unified position of TIPNIS communities has also been undone over time, (see Hirsch, 2019) undermined by the gifts given by Morales to some communities (televisions and canoe motors), by the creation of duplicate indigenous organisations to support Morales, and by a government consultation discredited by an independent review (for bypassing collective decision-making structures and misleading participants; Caritas Boliviana, 2013). As part of this project, I re-interviewed activists and leaders who opposed the Morales administration in the 2011 TIPNIS march.
The Chepete and Bala hydropower dams are being built on the River Beni in the lowland Department of La Paz and constitute Bolivia's largest infrastructure project to date. Bala, to the north, will flood an area of 94 km 2 . Chepete is further south and will flood 677 km 2 of forest (five times the size of urban central of La Paz; Fundacción Solón, 2016). The River Beni also forms the perimeter for both Madidi National Park and the Pilon Lajas indigenous territory and the dams are thus contentious. An access road and bridge is already being built alongside Madidi National Park and the Pilon Lajas territory. This road will be the first paved road of its size in the region, significant for new claims on land, new settlements, and deforestation (Bebbington et al., 2018). Opposition to Chepete and Bala is well-networked with other anti-extractive territorial movements in Bolivia and the TIPNIS conflict is used to garner publicity and raise awareness of similar conflicts (interviews TIPNIS activists). Cross-cutting themes of extractivism, territorial rights, and environmental degradation unite the two cases, as do dynamics of sustainability and development.

| ASSEMBLING THE SDGS -EMERGENCE, HABIT, AND BRICOLAGE
Agenda 2030 offers flexibility to states to interpret and implement the SDGs in line with national settings (Sexsmith & McMichael, 2015) seen as symptomatic of a "looser script" for global development than the MDGs (Horn & Grugel, 2018, p. 75). In Bolivia, the early take-up of the SDGs is guided by existing development agendas and partnerships, supporting Li's argument that assemblages emerge as "a matter of habit, accretion, and bricolage" (Li, 2007, p. 4).
In-country, the implementation of the SDGs is authorised by an agreement between the UN and the central government. Despite the creation of Agenda 2030 being newly participatory, in practice the UN works primarily with governments to adopt the SDGs. In Bolivia, this process was complicated by the reluctance of government ministers to work with the UN or adopt "foreign" development agendas (interview Org2). Although Morales has used the UN as a platform, his administration has simultaneously rejected dominant development paradigms and institutions as the imperialism of the Global North. Bolivia's relationship with global development has long been conflictive, for example because of the devastation caused by the structural adjustment policies of the 1990s (see Ellison, 2018;Kohl & Farthing, 2006). To reach the consensus required by the SDGs, the UN required "friendly ministers" to agree to meet and support the goals (interview Org3). Although partnership was secured, the SDGs have been matched to pre-existing national development targets. This has meant aligning the 17 SDGs to Bolivia's 13 pillars for development (themselves upholding constitutional commitments to HOPE | 213 Vivir Bien). This was explained as an agreement "at a conceptual level" (Org3), with the consequences for implementation to be determined by specific projects. The broad concept of sustainable development, and the new flexibility for states, has enabled partnership between the Bolivian government and the UNsignificant given Bolivia's recent history with international development institutions. In light of contemporary commitments to "progressive extractivism," however, this disciplines the take-up and implementation of sustainable development on the ground to weaken those opposing extractivism and to dilute the transformative power of Vivir Bien.
The merging of Vivir Bien and the SDGs by the UN and the Bolivian central government is contradictory, when viewed from the perspective of wider activist and academic debates. Sustainable development's explicit commitments to growth are at odds with how Vivir Bien/Buen Vivir has been conceptualised and advocated by activists and scholars. In their critical reading, Vivir Bien/Buen Vivir provides an alternative to sustainable development, as it decentres growth and instead moves toward a more holistic measure of wellbeing (including how communities live with and treat nature; Gudynas, 2011;Radcliffe, 2012;Walsh, 2011). It is because of this that Vivir Bien/Buen Vivir has engaged post-development scholars (Escobar, 2010), degrowth scholars (D'Alisa et al., 2014), and those seeking alternatives to development rather than alternative forms of development.
Deference to state-led development is particularly relevant to the environmental remit of the goals because post-neoliberal extractivism depends on contractual partnership with transnational firms and extractive capital. Included in Bolivia's sustainable development assemblage are the 2006 contract renegotiations with global hydrocarbon firms, when the Morales administration renationalised the sector. This gave Bolivia a higher percentage of royalties from private gas companies (amounting to US$5.5 billion in 2014 and US$3.7 billion in 2015; Fabricant & Gustafson, 2016;Kaup, 2010, p. 273). Following nationalisation the government still worked closely with transnational firmsrelied on for their expertise and technologies (Kaup, 2010) and the state now contracts transnational firms within a nationalised hydrocarbon framework. Bolivia contracts work, for example, from Brazil's state-oil giant Petrobras, Spain's Repsol, the UK's British Gas, and France's Total (Fabricant & Gustafson, 2016). Shell returned to Bolivia in 2015 after eight years' absence (CEDIB, 2015). Commitments to neo-extractivism have subsequently created "new alliances between an aspiring consuming class, more conservative elements within the MAS, and domestic and foreign capital" (Fabricant & Gustafson, 2016, p. 272)all related to Bolivia's national development plan and strategy. Beyond Bolivia, many of these hydrocarbon firms, such as Shell, have set out their commitments to specific SDGs (see Shell, 2018). However, they play a role in shaping the wider sustainable development assemblage, as a regulatory force that influences the uptake of the SDGs (to fit within a model of extractive-led development).
State-led development rests on partnerships with the global private sector and partnerships with transnational actors thus indirectly shape national policies on environment and development. As an example of how disciplinary power operates in the capillaries of the SDG assemblage, the state-led conservation ministry SERNAP cannot seek funding from those who promote neoliberalised conservation mechanisms but do work around (and support) the infrastructure of extractivism. Partly this responds to direct policy and partly this is self-regulated within SERNAP offices. In the TIPNIS, for example, SER-NAP was clear that its conservation mandate would continue "road or no road" (interview M1). Further, SERNAP wardens helped convinced communities about the benefits of the road as part of the controversial TIPNIS consultation (interviews). SERNAP maintains its commitments to biodiversity conservation and economic development for communities within comanaged spaces, such as Pilon Lajas, without challenging large-scale dam building in the region. Thus, in different ways and at different scales, state-led conservation plays a supporting role for extractive-led development.
In summary, the partnerships of the SDGs emerge from existing development agendas, networks, and politics, which pre-date the environmental goals for Agenda 2030 and discipline its uptake. The impacts of this on anti-extractive activism is twofold. First, seeking dialogue and partnership can come at the expense of contentunderplaying differences in interpretation. Second, the dynamics of particular partnerships are hidden, including the terms on which partnership has been agreed.

| TRANSPOSING MEANING AND DISCURSIVE REALIGNMENTS
Within Bolivia, the central government aligned Vivir Bien to the national development plan, associating Vivir Bien with economic growth, productivity, and industrialisation. The Minister for Hydrocarbons and Energy, Luis Alberto Sánchez, explained that making Bolivia the energy heart of Latin America was a key step in generating the funds to ensure Vivir Bien, defined by economic wellbeing and growth (Ministerio de Hidrocarburos, 2015). Activists opposing extractivism and its infrastructure (within the selected case study conflicts) explained how this transposed meaning (Li, 2007) stopped them from using the language of Vivir Bien: It is a discourse of living well but the one that the government proposes is nottheirs is a discourse of hydroelectricity, mining, extraction, hydrocarbon … (Berta) [W]ell we have not used that term so far (in this campaign) because it is the term used by the government, who interpret the lowest form of Good Living (Vivir Bien). (Carlos) These shifts in the capillaries of the SDG assemblage show that Vivir Bien is lost as a mode of articulation (following Hall 1986) for those demanding alternative development trajectories, as it has been redeployed to fit extractive-led development. Crucially, the global sustainable development agenda is not providing openings for the contentious politics of extractivism. Instead, the links between Vivir Bien and resource extraction are being consolidated by the sustainable development agenda, as small, self-regulating actions on the ground exclude anti-extractive agendas, meaning those opposing extractivism find themselves with less and less tools to contest it. The take-up of the SDGs strengthens the government's interpretation of Vivir Bien and smooths over its most contested impactsits environmental impacts.
Assemblage thinking is here insightful for including material components as agential actors to better understand how sustainable development is being materialised. In government discourse, the infrastructures and megaprojects of "hydroelectricity, mining, extraction, hydrocarbon" are implicitly brought into discussions of Vivir Bien and revealed as material components of sustainable development. In the posters and materials of government ministries, roads, bridges, mines, and energy projects are representations of development. The Ministry for Roads, for example, explained road building as a "first step in securing development," echoing ideas about development as modernisation or industrialisation more than reflecting the contested politics of road building and development within Bolivia (interview M2). The TIPNIS conflict demonstrated that extractive infrastructure prioritises accessing resources over benefiting communities. Although some community members opposed any road within their territory, for others the key issue was that the road was being built over 60 km away from their communities and was thus of little benefit (interviews). In Chepete and Bala, activists pointed out that though the access road is being promoted as bringing development, it threatens existing eco-tourism projects.
The impacts of extractivism go beyond specific sites of resource exploitation, requiring new infrastructure to access natural resources and enable transportation to global markets (Bakker & Bridge, 2006;Perreault, 2018). Bridges and roads have been analysed for their "affective force" as symbols of development and progress, despite their often contradictory effects (Harvey & Knox, 2012, p. 521). Their role in sustainable development is particularly contradictory due to the environmental impacts that roads have on surrounding landscapes, forest cover and biodiversity (see Bebbington et al., 2018). In the Western Amazon, this threatens high rates of biodiversity, as well as forests that are vital for fighting global climate change (Finer et al., 2015). In Bolivia, these material components consolidate interpretations of both sustainable development and Vivir Bien (both tied to the dominant model of extractive-led development), raising questions about the extent to which both self-regulating and transformative behaviours are shaped by the wider socio-material assemblage.
The SDGs are emerging from existing development agendas and partnerships, uncovering the roads and material infrastructures of extractive-led development as part of the sustainable development assemblage. Subsequently, questions arise for future work on the materiality of sustainable development and the role that infrastructure plays in shaping how sustainable development is seen, recognised, practised, and constituted. Following the next section on the disciplining power of extractive capital, I consider how the SDG assemblage informs the (uneasy) union of environment and development within sustainable development discourse.

| DISCIPLINING
Direct repression is a mechanism of disciplinary power and in Latin America this is visible in the treatment of those who protest the negative consequences of extractive-led growth. New dynamics of contention have been documented by the Latin American Observatory of Mining Conflicts, who note the criminalisation of those opposing mining in 143 out of the 245 mining conflicts documented; the Environmental Justice Atlas, which maps extractive-related social conflicts in the region (https://ejatlas.org); and Global Witness, which documents those attacked or killed for defending land and the environment from encroaching frontiers of capital. It documented 207 deaths in 2017, the highest yet recorded. The majority of these (almost 60%) were in Latin America (Global Witness, 2017). One strand of this paper contributes an analysis of how the SDGs, as a consensus-based discourse and initiative, work in this context. In Bolivia, conflicts over extractivism have fragmented the unified indigenous movement of the early 2000s. As already introduced, indigenous politics have become very divided, with MAS supporters keen to utilise resource extraction for a welfare state and economic development, while others (often led by lowland groups) contest and negotiate its effects.
Globally, the relationship between indigenous groups and international development actors has been seen as crucial in helping minority groups negotiate with states (see Andolina et al., 2009;Keck & Sikkink, 1999) but the uptake of the SDGs in Bolivia reveals a drastic reworking of the "boomerang effect" and development infrastructure is being disconnected from anti-extractive indigenous territorial politics (Hope, 2020). Disciplining is here examined for the more explicit and repressive ways that power is felt in the SDGs assemblage (Foucault, 2012). It is used to both analyse how the work of NGOs is modified to suit extractive interests and to explain the wider holding-together of the sustainable development assemblage.
In Bolivia, it is evident that NGOs are disciplined to ensure that the contentious politics of extractivism are not a part of their work. INGOs were clear that they could not get involved in disputes between the state and civil society, despite previously supporting those leading Bolivia's eco-territorial turn. This was explained as undermining the sovereignty of Bolivia and the government's right to determine the country's national development agenda. However, reticence to publicly engage with conflicts over extractivism was also explained as a response to the legislation and bureaucracies of government. INGOs described a climate of "fear and mistrust" (INGO1) when working in the countryone that challenged their work and created a sense of insecurity. NGOs referred to government attempts to pass new legislative frameworks for their work and to the ever-changing administrative process for gaining authorisation to work within the country, with several INGOs experiencing unexpected U-turns. Several bureaucratic and administrative shifts were cited as creating this new work culture and one NGO worker said they "spent most of (their) time doing bureaucracy" (INGO2). It was also widely cited that the government ejected an international NGO from Bolivia in 2010 for their support of the TIPNIS opposition. The MAS government accused them of political meddling (injerencia política), though it published no specific allegations. In this context, NGOs expressed difficulty in knowing how to best respond to rising social conflict, in ways sensitive to the responsibilities they had to their Bolivian partners and staff. They explained that: Not that you're looking to be in situations that are conflictive with the government but if we're funding something or funding a partner that did a report that the government didn't like I would then have to be thinking about risk to us, the risk to us physically, the risk to the organisational reputationall these types of considerations you have to consider first …. You would then worry about risks to the partner. (INGO2) One small Bolivian NGO worked in the immediate vicinity of one of the case studies. Yet they were clear that their work, though explicitly environmental in its remit, could not include or address the conflict. Despite their geographical proximity, they did not make any links with those contesting the megaproject and said that "any declaration (of support) would be problematic for our work" (NGO4). Several prominent Bolivian organisations have reported experiencing threats of closure or criminalisation in response to criticising government policy. CEDIB (Centro de Documentacion e Informacion en Bolivia; The Centre for Documents and Information in Bolivia), a research organisation and public archive, has published research that is critical of the government, for instance foregrounding the new forms of social resistance to extractivism and related socio-environmental degradation. Since 2016, the organisation has reported facing pressure to close, interpreted as government intimidation. Pablo Solón, ex-Ambassador to the United Nations and Director of Fundación Solón, has also publicly reported being criminalised for opposing the hydro-electric dams at Chepete and Bala (Fundacción Solón, 2017). Civil society organisations that engage with the conflictive politics of extractivism (and support those contesting extractive mega-projects) face significant challenges to their work. CEDIB described independent NGOs as experiencing "state penetration," lamenting the consequences for an independent and vibrant civil society (NGO1). For these actors, Bolivian civil society was being weakened and undermined by the Morales administration.
Referring to the government's disciplinary practices, one Director argued that of "the significant [development] networks that exist in Bolivia, there are many networks of NGOs with these [same] issues that are absolutely silenced. They do not want to talk about the fact that there is harassment towards NGOs" (NGO1). The organisations that do speak publicly about these dynamics feel threatened for "having objectives that are explicitly political, which [in the eyes of the government] should be prohibited" (NGO1). An envisioned role for civil society organisations as politically neutral and, following Li, as working on the "technical" problems of development was echoed in meetings I attended with UK-based NGOs. During this meeting, I observed a conversation in which NGOs referred to UK policy constraints, which stipulate that they must not work as political actors. This reveals another way in which disciplinary power works through NGOs, via the supporting discourses and legislative frameworks that keep charities tentative about undertaking work that may be deemed too "political" (for more on development and politics, see Hickey & Mohan, 2005).
Conflictive relationships with the central government also jeopardise international funding flows. One INGO explained that Bolivia's "hostile environment" shaped what funders were willing to support and meant that they avoid anti-extractive movements and "are just … keen to fund classic sustainable development … and steer away from anything more meaningful" (INGO2). Another NGO explained how their engagement with anti-extractive activism was constrained by funding decisions further afield: yesterday we presented a project to the XXXXXX … to defend the human rights defenders of the Amazon who are affected by mega projects and other types of violations, and even though they had approved a previous project [of ours] this time they rejected us, because they want to work in dialogue with the government. (NGO3) The global framework of Agenda 2030 offers no respite from this disciplining, as it promotes consensus and partnership between states, civil society, and the private sector. Neither are there obvious routes into Agenda 2030 for those marginal to governments. There are no legally binding accountability mechanisms through which to mediate disputesa problem for those trying to contest or hold governments to account, debate what is or is not sustainable, or bypass the national context. Although there are key categories for civil society engagement with the SDGs, namely the nine Major Groups that secure participation to the SDGs at the global level, interviewees (both representing institutions and activists) suggested that gaining access would be hard, if not approved by the national government. Beyond the Sustainable Development Knowledge Platform (and unlike the COP), there is no international forum or court within which groups can challenge how the goals are being interpreted and implemented. In Bolivia, the subjectivities of development workers (from different organisations) were both disciplined and disciplinary, directing the SDG agenda away from the conflictive politics and contested landscapes of extractivism. The avoidance of politics, when seen within the institutionalising practices outlined in the earlier section, reveals how the SDGs enact a neutralised, bureaucratised, and apolitical version of sustainable development.

| LOST GEOGRAPHIES
I argue that assemblage theory enables a closer analysis of how sustainable development is constituted, materialised, and disciplined, foregrounding how the particular partnerships and modes of consensus-building across the SDG assemblage secure complementary interpretations and practices of global sustainable development. An assemblage lens clarifies how particular issues (here, extractivism) and geographies (here, territories) are excluded, which I refer to here as the lost geographies of sustainable development. In this final empirical section, relational assemblage thinking is used to identify the particular territories and geographies that the SDG assemblage excludes, consider the agency and counter-assembling within these lost geographies, and finally, define the (uneasy) union of environment and development consolidated by the sustainable development assemblage.
In its emergence and disciplining, the SDG assemblage works to exclude the contentious politics of extractivism in Bolivia, which have specific territorial and protest geographies. The benefits and consequences of extractive-led development in Bolivia are unevenly distributed and lowland geographies have been particularly affected by the recent boom in hydrocarbon extractivism because of large pools of natural gas and different histories of territory and land (Linera, 2012;Fabricant & Postero, 2015). Anthropologist Andrew Canessa demonstrates that contemporary shifts within Bolivia (most prominently the creation of an indigenous-led state) necessitate "a conceptual distinction between inclusive national indigeneity for the majority which seeks to co-opt the state and a concept of indigeneity for a minority which needs protection from the state" (Canessa 2014, p. 153). Yet, the lost territorial and protest geographies of sustainable development are not without agency and can be read, in part, as a counter-assemblage.
Those opposing the Chepete and Bala dams in solidarity and, at times, in coordination with the continuing TIPNIS opposition remain politically active but without the support of international NGOs, as outlined above. Participant mapping revealed new networks forming between TIPNIS activists and emergent activist groups (collectivos) who are challenging extractivism in Bolivia. These collectivos are small and environmentalworking on localised environmental issues and in response to Bolivia's contemporary environment/development politics, as already defined. In explaining this counter-assemblage, Graziela, an activist within new collectivos, said: In response to being divided, silenced, corrupted and manipulated, we are now small …. We call ourselves from the left … we are inclusive, we are twenty collectivos [small collectives], some tiny. We talk. It is small and it is fantastic. (Graziela) An assemblage lens clarifies how particular interpretations, actors, practices, and materialities are stabilised as sustainable development discourse. This lens also reveals its omissionsin this case the conflictive politics of extractivism. The lost geographies of the SDGs, namely the territorial and protest geographies that form a part of the wider eco-territorial turn in the Latin American region, confirm continuing opposition to the socio-environmental impacts of extractive-led development, show how this opposition is weakened, excluded, and marginalised, and reveal the environments (and environmental politics) that are brought in to the uneasy environment-development union of sustainable developmentas well as those that are excluded.

| CONCLUSION
In this paper, I have made two core arguments and contributions. First, that assemblage theory, as used with Foucault's theory of disciplinary power (2012), offers an innovative and productive way to conceptualise the reinvigorated sustainable development agenda. Viewing the early take-up and implementation of the SDGs through an assemblage lens reveals how certain actors, subjectivities, policies, discourses, landscapes, and infrastructures form the dominant discourse of sustainable development, which in turn guides the take-up of specific goals and trajectories of transformation. Combining this with Foucault's theories of disciplinary power reveals how authority operates within the assemblage, through both self-governance and direct repression, to secure common interpretations and practices of global sustainable development. This exposes sustainable development's lost geographies (in this case extractive and protest), enabling us to connect, analyse, and explain contradictory inclusions and omissions within a global agenda for socio-environmental transformation. Extending assemblage work on the production of the social , assemblage has here been used to understand the production of the socio-environmental and to define sites of stability and transformation, as sustainable development is institutionalised, materialised, and consolidated.
Combining theories of assemblage with disciplinary methods of power (Foucault, 2012) is particularly well suited to an analysis of the SDGs (and Agenda 2030 more broadly), as a consensus-based political technology that promotes a common ethos for all. A conceptual framework that combines theories of assemblage and disciplinary power foregrounds the specific partnerships behind the SDGs (both institutional and between the human and non-human) to reveal the intricacies of consensus-building. The findings here support past critiques of sustainable development (namely that it is vague, depoliticising, and easily appropriated by interests that are neither sustainable nor developmental). I have found, however, these qualities to be instrumental to the rationale and functioning of the SDGs. In Bolivia, the broadness of sustainable development (as a concept for all) has enabled the SDGs to be taken up within the parameters of extractive-led development, narrowing Bolivian agendas for socio-environmental transformation.
Second, the subsequent analysis demonstrates that, following Ferguson (1990), the SDGs as assemblage act as a form of anti-politics by excluding anti-extractive political and territorial geographies to render neutral and apolitical the conflictive politics of extractivism. The sustainable development assemblage is being disciplined to suit state and private sector interests in ways that limit significant changes to extant unsustainable partnerships and practices. As a site of hydrocarbon and resource wealth, Bolivia is entangled with both transnational extractive capital and regional market demand and is therefore relevant to wider concern for how fossil fuel production, commodification, and consumption are addressed by global environmental agendas. In documenting how the SDGs neutralise the contested politics of extractivism, I hope to "expand possibilities for thinking critically about what is and what might be" (Li, 2007, p. 2). Despite calls from development geographers that academics should get behind the SDGs, as a pragmatic step to ensuring their effectiveness (see Liverman, 2018), the findings in this paper problematise this stance. Rather, and echoing Ferguson (1990), they support the importance of documenting and examining the stifled and excluded politics of the counter-assemblage.

| EPILOGUE
At the time of writing, Bolivian elections have been postponed until September 2020. Whether caused by government fraud or a coup (Londoño, 2019), to view Morales' descent as happening in one concise moment erases the key sites of contention within his administrationincluding those that dogged their politics of extractivism and indigenous territorial rights (Rivera, 2019). Nevertheless, what has followed is the increasing power of the Christian far-right, and a rise in state violence (including a massacre in El Alto), in racism, and in the dogged promotion of the Bibleall worrying for how indigeneity will be recast in the coming months. This paper provides insight into the legacies of the MAS administration for the politics of extractivism, sustainability, and territory, including an analysis of how disparate national political agendas connect with contemporary constellations of global development.