Imagining the Highway: Anticipating Infrastructural and Environmental Change in Belize

ABSTRACT This article examines the social and political, as well physical, construction of infrastructure, by attending to the implications of a highway yet to be built. In southern Belize, where the development of rural road networks figures strongly in historical narratives of political and environmental change, the recent paving of a major domestic highway has had distinctive implications for livelihoods and land rights among the predominantly Maya population of rural Toledo district. At the time of research, a plan for a new paved highway to the Guatemalan border animated longstanding debates over territoriality, environment and development, even as the details remained elusive. Bringing political ecology into conversation with attention to the perception of sensory environments, and the affective power of anticipation, I argue for extending anthropological conversations about infrastructure to encompass the meanings and consequences of imagined infrastructures for the ways people encounter, experience and enact social and environmental change.

here have a lease but it hasn't become official yet, the paper is in process stillit is all locked in the office! I had returned to Belize in 2007 in part to follow up on the implementation of a project to pave a new transnational highway to Guatemala (during my previous stay in the Toledo district in 2006, developments were expected imminently). Yet the more I attempted to learn about the new highway, the more unstable this vision of the future seemed. Paper trails were obstructed, official word unforthcoming. Residents of the areas through which the highway was expected to pass were both hopeful and anxious, offering wry laughter and rolling their eyes when discussing the possible road; lamenting at once the neglect of their 'forgotten district' and the unwanted attentions of state and commercial interests towards livelihoods and lands that already felt insecure. The alcalde's ambivalent account highlighted concerns about the politicaleconomic implications of the highway for livelihoods and human-environment relationships, and the sense of vulnerability and frustration felt by rural Maya villagers in the face of foreign capital, political patronage and labyrinthine bureaucracy. Throughout my time in Toledo, my attention was drawn to how the consequences and meanings of infrastructural and environmental change were bound up with village residents' emotional and embodied lives within their environments, and with the political contours of the region's history.
This article explores the effects of this anticipated infrastructure development in rural southern Belizea geopolitically sensitive area that has been subject to territorial claims and counter-claims through colonial and postcolonial periods. By attending to the anticipation of a highway yet to be built, I examine the shifting perceptions and consequences of imagined infrastructures for rural residents' encounters, experiences and interventions in changing social and environmental contexts. The shifting figure of the imagined highwayconjured through rumour, conjecture and elusive documentationwas a lightning rod for social/political relationships, animating long-running negotiations which I analyse here under the headings of territoriality, environment and development, in acknowledgement of the interconnected themes that emerged from the ethnography: the negotiation of boundaries, the sensory and speculative aspects of environmental change, and the affective politics of development interventions. The shift to an anticipatory mode provokes questions: who decides what is desirable and what is possible, and for whom? To what extent can project outcomes be foreseen and influenced, and by whom?
In the villages of Toledo, there was widespread discussion and speculation about this potential paved link to the border from the Southern Highway (a recently upgraded domestic road which connects Toledo's principal town of Punta Gorda to the country's main north-south route). At the time, the rural roads running through Toledo were unpaved, periodically impassable and stopped short of providing navigable vehicular access to Guatemala. The land border in Toledo is almost entirely unmarked on the ground; there is no immigration/customs post. Formal crossing requires travelling by sea via Punta Gorda, or by road further north through Cayo district. In the absence of international highways, people have long traversed the region by other means. People regularly informally cross the border back and forth for a variety of purposes: petty traders and personal visitors use boats, or well-trodden footpaths and mule tracks, navigating a river on stepping stones or horseback; xateros (xate palm harvesters) and others carrying contraband are more likely to cross under forest cover.
The potential of a paved connection to draw southern Belize into a major international shipping routethe physical manifestation of the Plan Puebla-Panamá (PPP)/Mesoamerica Project regional economic integration programmeis complicated by a territorial disagreement with roots in disputes between Britain and Spain, and subsequently between Britain and Guatemala. This remained unresolved at Belize's independence in 1981; Guatemala upholds a claim to the southern part of Belize. 2 Of interest here is the 1859 Anglo-Guatemalan Treaty, which describes the boundaries of British Honduras (now Belize) and provides for the construction of a new trade route between Guatemala City and the Atlantic. In support of its claim, Guatemala has cited Britain's failure to build such a route (Zammit 1978: 15-16;Young & Young 1988). Ironically, it seems that building a highway that could improve Guatemala's trade linksas per the treatyhas been stymied by associated political sensitivities: to formalise a crossing, Guatemala would presumably need to recognise this section of the border. Thus, a project ostensibly concerned with mobility became mired in inertia an outcome of what Jeremy Campbell (2012) calls the 'incompleteness of colonial geographies', whereby unresolved political legacies fuel uncertainties and tensions about frontiers, governance, and the material realities of life near borders and highways.
In 2002, an environmental impact assessment (EIA) had been undertaken for the extension from the Southern Highway to the border village of Jalacte, along the most northerly of the existing routes (Halcrow 2002). With loans from OPEC and Kuwait secured in the early 2000s, the Government of Belize proposed that the two transportation projects (the Southern Highway and Border Highway) would facilitate muchneeded economic development of the south. However, while the Southern Highway progressed (albeit haltingly), the border highway project was not implemented. Domestic and international sensitivities about land rights and borders made stable and reliable information on the highway project difficult to pin down, and during fieldwork I frequently found myself entangled in the same uncertainties as my interlocutors.
Although the Toledo Strategy and Action Plan 2006-2009 alluded to the project (Hutchinson 2005: 24-25), the timescale and intended route remained unclear, until roadside signs appeared in 2008, announcing the delayed final stage of the Southern Highway paving project and the Belize/Guatemala border road (Figure 1). Following a general election in February of that year, works to complete the final section of the Southern Highway continued, yet solid plans for the border road remained elusive: infrastructure funds were needed to repair storm damage elsewhere; global economic circumstances changed; the border issues remained unresolved; and political priorities shifted. (The United Democratic Party (UDP), which had regained power after two terms in opposition, has been historically less oriented towards Central America than its main opponent (Payne 1990: 124)as one young man opined; the UDP thought the road 'pointless, as it leads to nowhere'.) Eventually, following an EIA update in 2010, and a loan from the Central American Bank for Economic Integration, a Belizean construction company began work in 2011 along the northern route; '[t]his time however, without the immediate plans to incorporate a border crossing' (BEC 2010: 0.10), though government press releases acknowledged this as an ultimate goal. At the time of writing, construction of the border highway is all but complete.
This article examines the period of uncertainty that preceded the 2010 EIA update, drawing on 18 months of ethnographic field research between 2006 and 2009. As such, it seeks to analyse the anticipatory aspect of infrastructure-in-the-making, and to encourage reflexive thinking about the political spaces between imagination, planning and implementation.
Research focused on three villages I will refer to as San Carlos (a predominantly Mopan village, to the north of the possible border highway routes); Crique Jawan (a predominantly Q'eqchi' (Kekchi/Ketchi) village, near the border with Guatemala); and Mahogany Bank (a mixed Q'eqchi' and mestizo village near the market town). I lived with local families and supplemented participant observation, meeting attendance, informal discussions and semi-structured interviews with 'ethnography on the move', travelling extensively using local transportation to better understand village residents' road use in rural Toledo district. Predominant productive activities were small-scale agriculture and/or fishing; some individuals were employed as soldiers, teachers or plantation workers. Excepting the mestizo residents of Mahogany Bank, the majority of my interlocutors identified themselves as (Mopan or Q'eqchi') Maya. 3 Maya groups have long been involved in legal struggles with the government of Belize. At the time of research, a claim for indigenous customary rights to land was progressing through the courts on behalf of the Maya communities of Southern Belize. These claims are crucial context for development in Toledo; decades of mistrust between the parties underpinned expectations and responses to the road proposals (see also Wilk 1997;Harrison et al. 2004;Wainwright 2008). 4

Infrastructural Futures: Politics and Perception
A growing body of work across anthropology, geography, and science and technology studies addresses the extensive social dimensions of infrastructure and increasingly explores its analytic productivity (e.g . Star 1999;Barry 2013;Larkin 2013;Appel et al. 2015). Recent ethnographic accounts of infrastructure and environment have demonstrated how attention to the interactions of technology, politics and ecology can aid understandings of the patterning of social difference, resource distribution and political violence (Carse 2012;Rodgers & O'Neill 2012;Harvey & Knox 2015). Roads constitute but one variety of infrastructure, though perhaps an archetypal one. From Roman roads to Maya sakbe'ob, the Silk Road to Route 66, roads hold pivotal places in cultural imaginaries. They have been explained as channels and technologies of economic or political power (Fairhead 1992;Thomas 2002;Harvey 2005), and linked to modernity, progress, and the compression of time and space (de Pina-Cabral 1987;Roseman 1996;Harvey 2005;Virilio 2006). Large infrastructure projects attract expectations and dreams (Löfgren 2004). Although roads are sometimes characterised as 'non-places', which in Marc Augé's definition 'cannot be defined as relational, or historical, or concerned with identity ' (1995: 77-78), this article acknowledges the relations, histories and identities bound up with roads as provocative sites for anthropological investigation of social and environmental change (Dalakoglou 2010;Dalakoglou & Harvey 2012;Orr 2016;Harvey & Knox 2015).
To understand how negotiating divergent views of potential infrastructural transformations of the environment may generate meaningful and material outcomes, I address the historical and political-economic contexts of environmental controversies, an approach familiar to the field of political ecology. I do this, however, mindful that the field has attracted calls to move beyond deterministic 'chains of explanation' (Robbins 2004;Coombes et al. 2012), and towards more nuanced consideration of place, agency, meaning, 'webs of relations' and 'socio-ecological lives' (West 2005: 633;Biersack & Greenberg 2006;Rocheleau & Roth 2007). These provocations inform my inquiry into how imagination and power are brought to bear together with materials in everyday lives. This requires situating contemporary planning in political context, and considering sensory aspects of how people perceive, anticipate and enact changes in their lived environments (Ingold 2000). The potential highway in Toledo provides an exemplary case for such an enquiry into the emergent socio-environmental effects of imagined infrastructures, in a situation where resource and development politics have long been contested, and where the sensations and perceptions of current transportation practices and materials are mutually constituted with hopes and fears for future infrastructures.
How are political relations manifested effectively and affectively, not only in their structuring by physical materials, but also in historically situated, value-laden and future-oriented planning, portent and promise? Examining infrastructures in an anticipatory modeenquiring how things come to matter before they become 'matter'is, I suggest, an important contribution to an anthropology of infrastructure; one that can counter the potential displacement of attention to human agency by a focus on the 'structuring effects of materials themselves' (Knox 2015: 105), and can address concerns that new materialist approaches may 'blind us to that more complicated world of relations' that lies between self-evident things (Fowles 2010: 25), precisely by reflecting on the productivity of imaginative as well as material forms in generating socio-ecological futures. Focusing on expectations, emotional responses and protests before infrastructure is materially present calls attention to the agency and constraints of different actors as they consider and respond to the potential implications of the road. This is not, then, an ethnography of concretely existing infrastructure, but of the social relations that pattern and crystallise around its imaginings and implementations. The paved highway's absence in the present invites engagement beyond the material to consider questions of people's emotional and political experiences and their orientation to the world to come (Bille et al. 2010). The perception of its absence is heightened by rumours and plans that draw attention to what is not, but could be. Thinking about imagined infrastructures opens space to deliberate the potential structuring effects of materials, while foregrounding less tangible social fields that connect past, present and future. Anthropologists have in recent years explored such fields with reference to planning (Abram & Weszkalnys 2013), rhetoric (B Campbell 2010), preparedness (Lakoff 2008) and hope (Reeves 2016, Weszkalnys 2016). These works have examined how expectations are imagined, discussed, institutionalised and emotionally experienced, and they have stimulated my thinking about how social and political dynamics are realised in physical artefacts yet also transcend them, shaping the kinds of future-oriented knowledge and action that are seen as possible and legitimate.
Recent anthropological attention to anticipation has started to analyse and theorise its forms and its affective implications, for example, in fields of biotechnology (Adams et al. 2009), climate change (Whitington 2013), precarious labour (Molé 2010), special economic zones (Cross 2014) and resource extraction (Weszkalnys 2016). According to Adams et al., 'Anticipatory modes enable the production of possible futures that are lived and felt as inevitable in the present, rendering hope and fear as important political vectors. Parsing anticipation means exploring the politics of affect as much as speculative epistemologies ' (2009: 248). Focusing my analysis on the anticipation of infrastructures, which in Brian Larkin's formulation have a 'peculiar ontology' that dwells in their being at once political and poetical -'things and also the relation between things' (2013: 329), I emphasise that these relations do not only connect spatially but also across time, influencing the present through the very expectations and sensations conjured by their absence or incompleteness. Ethnographically exploring processes with outcomes that are not fully determined is a way to place responsibility, ethics and political action at the centre of enquiry.
In the next section, I frame the discussion with a brief history of road development in southern Belize. Discussing their political ecology alongside their sensory associations, I show how roads are temporal markers in memories and life histories, harbingers of social and political change, and palimpsests of state-public relations. I then organise my analysis of the perceptions and effects of the imagined highway with reference to territoriality, environment and development, each of which bridges concerns of political ecology, perception of environment and infrastructural anticipation.
Territorialityconceived as the drawing of social and spatial boundariesbroaches notions of property, belonging and place-making, and access to spatially fixed resources (Casimir & Rao 1992). The impact of road building on indigenous peoples' land rights and livelihoods has been a major concern for Maya groups in southern Belize. The potential for a new highway connecting Guatemala and Belize also raises anxieties about international territoriality, given the ongoing border dispute. As Penny Harvey argues, roads 'transform particular territorial spaces into sites of fantasy and projection for politicians, planners and local people ' (2005: 131). Imagining the highway thus entails hopes and fears about the future of territorial boundaries at multiple levels. Environmentwithin which humans are engaged participants rather than detached observers (Abramson 2000: 23;Ingold 2000) can enable and constrain the forms and materials of infrastructures such as roads, which in turn alter ecological and sensory surroundings, and motivate speculative changes in land use, settlement, and resource access before projects are confirmed, let alone ground broken. Development efforts in Toledo, supported by governments, NGOs and international donors, have been aspirational, but implementation has often failed to meet expectations, fuelling cycles of dependency and underdevelopment. Infrastructure planning in this context realises a tension between what Rodgers and O'Neill (2012) term passive and active infrastructural violence: on one hand, the absence of infrastructure marginalises those left out of connective networks; on the other, its imposition can herald unwanted surveillance, intrusion and expropriation. Paying attention to the spaces between proposals and implementation grants insight into how and when passive infrastructural violence can give way to active infrastructural violence. 5 Anticipating new interventions through the lens of past experiences, people in Toledo raise questions that go beyond the costs and benefits of particular ventures to ask who has, or should have, the capacity to shape the lives and environments of the people of southern Belize.
To conclude, I propose that exploring the politics and perceptions of infrastructure in anticipatory mode is a way to extend understanding of how promises of 'development' bear cultural, sensory and material effects for people living in its midst. It thus contributes to understanding the dynamics of social and environmental change, by illuminating the mutual constitution of resources and meanings in situ and in process.

Toledo Roads: Tracing the Past
Roads are not a new phenomenon in this region. From the Preclassic period, the Maya built broad sakbe'ob (white roads) within and between ceremonial centres and outlying areas. In Yukatek and Mopan, beh (b'e in Q'eqchi')road, trail or pathcan also refer to one's life course or destiny, and is often incorporated in references to measuring time and movement (Keller 2011). For the Q'eqchi' of Verapaz, the powerful tzuultaq'a spirits are 'roads' that travel the mountain valleys (Adams 2001). Anthropological accounts from the broader region highlight how roadbuilding distils the cultural politics of socio-economic development at different scales. In 1950, Robert Redfield reported on the 'road to the light' from Chan Kom to Chichén Itzá in Yucatán, describing it as starting out 'toward Chicago rather than toward Mexico City ' (1950: 153-154) to indicate what he saw as the village's progress towards North American rather than Latin culture. Critiquing Redfield, Quetzil Castañeda identifies the predominant view of road construction as the 'Euro-Yucatec symbol par excellence of the intrusion of Modernity and the diffusion of its Light into the Primitive Darkness of Other Folk ' (1995: 135). Peter Hervik described the 1965 opening of the Merida highway as a key driver of a 'crucial break in the continuity of the use of Maya language and dress patterns ' (1999: 29). Across the Americas, highway construction has been arguably the most visible manifestation of regional integration projects in the era of free trade, including the PPP/Mesoamerica Project, and the Initiative for the Integration of the Regional Infrastructure of South America.
Under colonial rule, road infrastructure in British Honduras was largely neglected until the 1930s (Grant 1976: 69-70). Vehicular access to Toledo's villages was limited until rural roads were improved during the 1970s and 1980s, supported by the government and USAID. The paving of the Southern Highway in the early 2000s, coordinated by the Government of Belize and the Inter-American Development Bank (IDB), provided Toledo with its first 'all-weather' link to the rest of Belize. Previously, journeys were made on foot or horseback, by air, sea, orfrom the mid-1960svia a rough road prone to dustclouds and floods. The US$32.4 million 'rehabilitation' project covered 149 miles of highway and feeder roads, with donors including the IDB, UK, Taiwan, Kuwait, OPEC and the Caribbean Development Bank. After a delay, the final 10-mile section was completed in 2010 ( Figure 2 shows these works underway).
Today, the rural roads branching from the Southern Highway towards the Guatemalan border are rough and often narrow, winding or steep (Figure 3). In dry months, dust can reduce visibility to a few metres. In the wet season, creeks breach the low bridges, stranding their main traffic of repurposed US school buses, operated by villagers as market transport.
In the Toledo villages where I stayed, the arrival of a road was an important temporal motif, distinguishing now from before in oral histories. Echoing the respect they expressed for early settlers who cleared paths through the forest, residents recounted local lobbying for roads to link them more effectively with networks of transportation, goods and knowledge. Past successes were emphasised, though relationships with state representatives and other bodies were ambiguous. In San Carlos, some attributed the building of their access road to timber or oil companies; others claimed more local control. San Carlos resident Roberto's emotive account illustrates the tension between asserting autonomy and requesting support: [The people] see that they need a chairman, a village council, alcalde system … they worked together, they sent a letter to Prime Minister George Cadle Price. They cleared this road by themselves, no? They wanted to open this road so they wrote a letter to the prime minister, way to Belize City … They went to the lands minister, just went their own selves. And in the way they are coming togetherbecause they really want the road to go to the villagethey are not playing, they are serious! This reflects how successful navigation of state procedures enabled the road, echoing Roseman's (1996) description of rural Galicians' negotiation of compromise and resistance to reconstruct development histories.
Across Toledo, the arrival of a village road was often associated with institutional change, most explicitly the introduction of village councils since the mid-twentieth century. Supporting a view of road development as a mechanism by which states render marginal places more legible (Scott 1998), village councils have gradually supplemented or replaced alcaldes with an authority more closely tied to national political structures (Moberg 1992). In San Carlos, the village elder related the opening of the village road to the installation of the council and increased contact with government. The establishment of Mahogany Bank's alcalde and council in the mid-1990s coincided with the arrival of US missionaries, Guatemalan Q'eqchi' speakers and the track road. Crique Jawan shared its council with a neighbouring village until it got its own road in the late 1990s. Studies from around the world show how road projects can evoke memories of corvée labour and resettlement under colonial and other regimes (Fairhead 1992;Roseman 1996;Thomas 2002;Harvey 2005). In 1930s Guatemala, President Ubico's road building programme extended state power over lands, economies and bodies through forced indigenous labour (Grandia 2012: 43). Some of those labourers fled to what is now Belize, where national goals have also been influential, if less directly coercive: the government's pre-independence development push in Toledo can be viewed at least in part as an attempt to garner the loyalty of those living near the margins of the emerging postcolonial nation-state (Wainwright 2008: 206).
Throughout the 1990s, the potential negative impacts of paving the Southern Highway underpinned arguments put forward by the Toledo Maya Cultural Council (TMCC)a key organisation campaigning for indigenous rights to lands occupied and used by Maya people but classified as government property: Paving this road will be a blessing and a curse to the people of Toledo. It will bring much revenue to the district and tie in this 'forgotten district' with the rest of the nation … However, the incorporation of Toledo into the rest of the country may further disadvantage indigenous communities. The Maya are the poorest people of the nation, unable to tap into the benefits of increased commerce. The land tenure question, regarding the control of the land on which we live, must be resolved before the highway is completed. We want to make it clear that we are not against development. But neither do we support development which would have The enthusiasm of governments and donors for roadbuilding and accompanying projects to formalise individual property rights in rural Belize may in effect continue the colonial authorities' efforts to 'pin down the wandering milpa'. 6 Past studies in the region have shown that improving rural roads encourages shifting cultivators to become more sedentary (Wilk 1984). Farmers living along the Southern Highway told me that its paving (along with the impacts of Hurricane Iris in 2001) led many to reduce subsistence production and pursue income diversification. Better transportation has supported a burgeoning market for cacao, which now provides an important income for many Toledo farmers. Nonetheless, I encountered some farmers agonising over whether to plant the permanent tree crop: unlike temporary subsistence crops such as maize, beans and rice, the long-term investment required limits mobility for work or family travel, and can affect tenure arrangements where land is collectively held. The vagaries of seasonal employment, costs of staple goods and vulnerabilities of monocropping make reducing a subsistence base risky (Wilk 1997;Zarger 2009).
Despite concerns highlighted in the report of the Environmental and Social Technical Assistance Project (ESTAP) (GoB & IDB 2000), designed to mitigate the effects of the Southern Highway upgrade, ongoing conflicts over land tenure were not resolved prior to commencement. Many years later, and notwithstanding rulings affirming Maya rights (in the Belize Supreme Court in 2007 and 2010, Court of Appeal in 2013, and Caribbean Court of Justice in 2015), the issue remains unresolved in practical terms. Although Maya groups negotiated a 10-year moratorium on land sales within two miles of the Southern Highway, Grandia (2009: 157) found that this was not necessarily upheld by the Lands Department. Maya people have responded in diverse ways to threats of land loss: some have attempted to secure leases; others have joined groups working to demarcate community lands, file lawsuits and oppose cadastral surveys.
As well as channelling state interests, roads in southern Belize can also be sites for popular mobilisation. Farmers' cooperatives, women's groups and land rights organisations use improved transport connections to forge links between spatially distributed groups with common interests. Community leaders can more easily petition ministerial offices. During the land claims of the late 2000s, one particularly compelling image was of convoys of chartered buses on the Southern Highway, carrying Maya villagers to Belize City to hear court proceedings. In Stann Creek district, residents of Maya Center village protested their loss of representation in the management of the Cockscomb Basin Wildlife Sanctuary by blocking the park's access road (Wainwright & Ageton 2005;Medina 2015).
Temporal as well as spatial interferences can disrupt the expectation that accessibility, let alone socio-economic progress, proceeds smoothly from road construction. Despite the Southern Highway's new 'all-weather' designation, floods washed away a major bridge in 2008. It took four years to install a replacement, leaving Toledo once more periodically cut-off to road vehicles each time the temporary causeway eroded. In some parts of rural Toledo where road sections had once been paved, lack of maintenance has resulted in deep compacted ruts in the asphalt, rendering them harder to navigate than never-paved sections. The deterioration testifies materially to the sense of neglect so often articulated in rural Toledo, popularly referred to as 'the forgotten district'. Timing is also significant in the widespread acknowledgement that roads are often repaired prior to elections in return for political favours. Such politicisation renders Toledo's roads productive sites from which to read an archaeology of political relations.

Imagining the Border Highway
I now shift focus from existing roads to the effects of then-ambiguous plans for a paved highway through Toledo to the Guatemalan border. My aim in the rest of this article is to explore ethnographically the social effects of uncertainties about community and personal security, land rights and human-environment relationships, and the promises and dangers of development and regional integration, as rumours of the new highway provided a conduit for reflecting, articulating and in some cases acting on expectations of future environments and infrastructures.

Territoriality: The End of the Road
The planned border highway has implications for social and spatial boundaries at local and international levels. The three villages discussed here are located at the 'ends' of earth or gravel roads, and thus well nested within the popular geography of Toledo as a place where roads terminate. For their residents, the current lack of throughtraffic is significant for 'community': the accessibility and surveillance that roads grant to outsiders are at least partially deflected by local people's ability to keep watch over their roads and 'protect community autarky ' (de Pina-Cabral 1987: 718). Hearing a vehicle approach, people living near the roadside would peer through gaps in wall panels and from behind half-closed doors, remaining hidden as they determined the visitor's identity and motive. Women are warned not to walk alone on roads, to avoid meeting 'drunken men'. 7 Walking roads alone, one might also encounter the legendary forest guardian Tata Duende, or other apparitions from Amerindian and Caribbean folklore. One young man from a Mopan village recounted a friend's macabre vision of a procession of coffins crossing a road, bringing to mind Masquelier's (2002) accounts of Nigerien road spirits that articulate the dangers of high speeds, poor lighting, negligent driving and insufficient maintenance. These narratives show how existing and future roads play into culturally inflected ways of relating to forest environments and shifting boundaries between life and death, human and spirit worlds.
As noted in the introduction, many Toledo residents living near the border maintain genial links of kinship and trade across it. Nonetheless, Belizean nationalist discourses, and popular association of Guatemala with structural and physical violence, have stirred anxieties that new traffic into Belize would increase violent encounters. Coupled with memories of the Guatemalan war and knowledge of regional narcotics trafficking, tales of cross-border organ and child theft circulate in Toledo. As in other places, vehicles are key motifsmodes of transporting victims, or desired goods that make trafficking an appealing income source (Scheper-Hughes 2000;White 2000: 122-147). Through-traffic and cross-border traffic thus entwine folkloric tropes of unregulated mobility with contemporary anxieties about the security of Maya livelihoods, land rights, cultures and bodies. Installing an immigration post was not expected to help; indeed people in Belizean villages near the border were anxious that this would only constrain their own mobility, mapping onto national tensions about ethnicity that could lead to persecution of those identified as Maya or mestizogroups often associated in the popular imagination with migration from neighbouring countries. 8 Geronimo, a farmer and tour guide from San Carlos, evoked uncertainty about what the highway might bring, as he asserted the need to take a defensive stance in terms of home security: Some are saying that whenever a highway is completed it's good … other people say that if the highway is completed a lot of bad people will come indrugs people … We in the villages are not well prepared with our houses. When the highway comes you have to be prepared with your house because you don't know what's coming.
Many women, too, explained they would have to start locking windows and doors because of the anticipated breakdown in trust and social regulationit is harder to observe and respond to vehicles travelling at the speed afforded by asphalt. They feared they would have to keep their children and yard animals indoors. Previous highway upgrades have contributed to spatial reconfigurations of socio-economic relationships within roadside villages, as concerns about safety, noise and comfort, as well as economic factors, influenced decisions about housing locations, building materials and land use (see also Wilk 2006: 162). An engineer who worked on the Southern Highway recalled villagers adjacent to the new road moving their thatched timber houses 'back into the bush' or replacing them with cement and zinc (he surmised that residents felt the paved road had made their thatched houses look 'out of place'inconsistent with the asphalt highway's modernity). Enhanced access may thus increase physical and social 'closure' of some communities, by encouraging residential fortification, movement away from the roadside and village-imposed limits on in-migration, as recently introduced in Crique Jawan. That these measures were being considered and in some cases implemented in anticipation of the international highway indicates the effects of as-yet-intangible infrastructures in the face of current territorial insecurities (local to international), and the articulated need to manage boundaries carefully in preparation for an even more uncertain future.

Environment: Sense and Speculation
Concerns about territory and security were often framed in terms of changing environmental perceptions. Residents expressed anxiety that increased accessibility brought by a paved highway would disrupt their village's valued 'coolness': a sensory reference to its social attributes of security, morality and tranquillity (Haines 2012). The IDB (2012) reports that traffic on the Southern Highway has increased by over one hundred percent since its paving; this is proclaimed as a measure of economic efficacy, but the increased speed and volume of traffic also raised concerns about the safety of pedestrians, cyclists and livestock.
Not all infrastructural changes were feared. In Mahogany Bank, one man described how, in the mid-1990s, he and fellow villagers dismantled their timber homes, paddled them across the river in canoes, and reassembled them on the opposite bank, following news that a track road to the market town was to be opened from the new site. As I have described elsewhere, according to my interlocutor the translocation occurred 'because they say the light is coming'. His reference to 'light' alludes to the anticipated (and still-awaited) extension of electricity provision in concert with the road, as well as to sunlight penetrating cleared forest canopy, thus exemplifying how development projects are multisensory experiences for people living in affected areas (Haines 2012: 109). Wilk (2006: 162) notes that many Q'eqchi' people describe roadside settlements as bright places, where access to hospitals, schools, churches and markets cultivates an aura of aspiration (recall also Redfield's (1950) 'road to the light'). The narrative of the reorganisation of the village precipitated by the anticipation of 'light' and, by implication, 'development', demonstrates the material effects of the sensory, affective and political aspects of imagined infrastructures, manifested in the physical movement of people and structures towards promised opportunities.
Rising land values associated with paved roads have implications for environments, economies and social relations, particularly in frontier zones and places where land rights are at stake. The potential implications of the border highway for agricultural practices in Toledo's Maya villages have been examined by Wainwright et al. (2015), using remote sensing analysis and historical sources. They conclude that the new highway could lead to greater deforestation in Belize by encouraging more intense, non-customary agricultural practices oriented towards Guatemalan markets. However, they note that this is only one possibility: the outcome will depend on asyet-uncertain dynamics concerning the future of the border, the trajectory of regional integration and the negotiation of Maya land rights.
Faced with these uncertainties, Toledo residents' expectations of the paved border highway have been informed by the afore-mentioned shifts in land use, tenure and livelihoods along the recently paved Southern Highway, and also by knowledge and experiences of Guatemala's Petén department (adjacent to Toledo) and the Franja Transversal del Norte, where road projects supported by state colonisation agencies and military development plans attracted military officers, commercial plantations and cattle ranchers, forcing out smaller-scale farmers (Schwartz 1995;Grandia 2012;Solano 2013). Studies of the Maya biosphere reserve have argued that building roads in this region will be detrimental to both economy and environment (Ramos et al. 2007). Given the route of the proposed highway, the impacts of PPP, the Central America Free Trade Agreement and increased accessibility of Guatemalan markets are anticipated to expand further into Toledo. Ranching already straddled the border through informal land leasing arrangements in Crique Jawan, where locally identified 'rich men' were deemed to be capturing the benefits. In San Carlos, too, Geronimo articulated fears of land loss to more powerful interests, framed in terms of insecurity and inevitability: A lot of people are trying to complain … saying if the highway is completed a lot of important people [will come] to buy the land, because they heard the land was here. The company will come and buy the land, and grow some citrus or do whatever they want.
While it was not widely thought that the border highway would pass directly through Mahogany Bank, repercussions were foreseen throughout the district. Rumours intensified speculative tensions about land security. Some residents thought that the existing flood-prone track road might soon be paved or extended. Juan explained that he and his fellow villagers needed to prepare for the inevitable by securing property rights immediately: We have to get [the land] … because everybody will rush, because this is a highway and a riverside. That's why people are getting the land, because they are getting ready. And that's what you call development! Some good and some bad thingsmore people, more problems, more work.
Those with the resources to navigate uncertain and clientelistic processes were expected to benefit disproportionately. Some villagers had obtained leases only to lose their land, unable to avoid defaulting or using the lease as collateral; others found their documents lost or denied when political regimes shifted.
Paved highways will make land more accessible for resource extraction ventures involving heavy machinery and productsindeed, recent seismic explorations in Toledo have heightened the urgency of land claims. Speculative projects substantially affect environments, economies and sociopolitical organisation as resource frontiers are actively brought into being, inspired by desires and promises, inscribed in documents, legal instruments and environments (Tsing 2000). In São Tomé and Príncipe, Gisa Weszkalnys has argued that oilrather than being a lone determining force -'causes certain things to matter ' (2011: 365). I contend the case is similar for the figure of an imagined highway in Toledo. Applying the lens of anticipation reveals how expectations and fears can affectively animate social and political patterns, emphasising particular versions of history and visions of the future, in processes of shaping resources and negotiating access and rights.

Development: Mobility and Infrastructural Violence
A major impetus for the Southern Highway upgrade was the premise that poor transport infrastructure has inhibited economic development in southern Belize. 9 For many Toledo residents, the lack of paved roads was a vivid reminder of state neglect, signifying the perceived impotence and corruption of party-affiliated 'politicos', and the failure to deliver meaningful development. In the rural tourism sector, for example, unsold crafts and empty guesthouses exemplified unfulfilled promise, and the passive infrastructural violence (Rodgers & O'Neill 2012) inflicted on those excluded from networks of mobility and opportunity. People in San Carlos and Crique Jawan were optimistic about increased access to education, health care, and markets for labour, agricultural produce and tourism. They noted that electricity and better water systems would be slow to arrive without a paved road. Maps in the ESTAP report illustrate the association of such facilities with more accessible roads. Developments (including a new clinic) had started to appear along possible highway routes. The bare poles lining one of these routes promised future electricity provision, while emphasising its current absence.
Although the potential border highway offered hope for some, I encountered much scepticism (how many tourists would stop along itlet alone venture further?); some described it as an existential threat. Even now, with works underway, benefits remain contingent on undetermined factors including land security, and migration of powerful and disempowered people. Road projects in rural and border regions are buttressed by promises of improved circumstances for 'marginal' populations, yet oftenas Donna Flynn observed in Bénin: ' … it is not so much that the road is the border residents' pathway to the city … as that it is the city's pathway to the border' (1997: 320).
Using Rodgers and O'Neill's (2012) terms, this reflects concern that passive infrastructural violence may give way to active infrastructural violencevia surveillance, regulation and expropriation, hindering rather than enhancing freedom of movement for certain groups. In Toledo, where infrastructure projects are framed by colonial and postcolonial resource extraction, forced labour and civil war in neighbouring countries, regional narcotics trafficking, narratives of organ theft, experiences of border-related violence, and the parallel threats of state abandonment and interference, they attract uneasiness about appropriation by nation-states, criminals and corporations. The dual lens of passive and active infrastructural violence links sensory and structural effects of material change. The attention to both absence and presence resonates with my argument that an imagined paved highway is affectively encountered as historical/contemporary absence and anticipated future presence.

Conclusion
In this article, I have examined how a planned border highway animated debates about territoriality, environment and development. I have drawn out material, sensory and imaginative facets of development planning and implementation, and explored how these shape and are shaped by the histories, politics and expectations of the physical environment, in an area where land is at stake, identity politics are sensitive, and a prominent sense of place is associated with living in the 'forgotten district'. At once agents and indices of change and controversy, roads can re-orientate and reinforce relationships among people, institutions and environments, and stand for tensions between development plans and political realities, aspirations and anxieties. They can be sites of hope and mobilisation, promising the light of education, employment, markets and health care; they also bring the heat of challenges to livelihoods, resources and values.
Extending beyond the materiality of existing roads, an anticipated paved highway towards a contested border commanded imaginative force. Pre-empting physical construction, its anticipation bore consequences for practices and perceptions of territoriality, from international and community boundaries to concerns for the integrity of collective and individual homes and bodies. Expectations of changing environments were reflected in perceptions and narratives of sensory change as well as negotiations of opportunity and precariousness, as speculation added to the perceived urgency of securing land for livelihoods. Promoted as both means and ends of development, the road bore the potential to overcome the passive violence of exclusion. However, many Maya peopleand other rural residentsdistrusted another project anticipated to disrupt socio-ecological relations, bypass locales without bringing meaningful benefits, and erode local control of resources and meanings through active violence of surveillance, extraction and appropriation.
The anticipated highway provided a hook on which local people hung aspirations, reservations and curiosities (see also Hirsch 1994: 706). However, the lack of certainty in the proposals at the time of research also inhibited specific contestations by local residents and/or social movements. Since my main fieldwork period, technical visions have become evidently unsettled as construction progresses: the border road willfor nowstop short of crossing the border, thus interrupting smooth integration; construction was also delayed through at least one village owing to protests by local people and Maya organisations. Worldwide, planners and engineers encounter not only physical obstacles such as topography or weather, but also the friction generated through geopolitical disturbance of plans, and productive tensions of the cultural and the technical (Tsing 2005;Khan 2006;Harvey & Knox 2008), through which alternative futures may be imagined and questioned.
While 'imagination' seems somewhat antithetical to the concreteness of roads, this article reveals how transformations wrought and promised by infrastructures-in-themaking become part of navigating the present, raising questions about (im)possible or (un)desirable roles of states, institutions and local people. By virtue of its sensory and emotional implications (light, aspiration, frustration, noise, insecurity), imagining a highway is an affective as well as political-ecological concern. This article's anthropological attention to imagined infrastructure foregrounds questions of care for what may be to come, exploring a politics of hope, fear and uncertainty, as well as of materiality, speed, access and obstacle. Turning attention to anticipation foregrounds the social and political construction of transportation infrastructure and its implications for lives and livelihoods. It thus moves toward a better understanding of how imagining new configurations of infrastructure also entails imagining and enacting new ways of living within changing environments. Anticipatory moments, when physical transformations are on the cusp of being materialised, cast social and environmental relations in dynamic suspension: oriented by existing resources, meanings and power balances; emergent through negotiation of outcomes that are not fully determined. At stake are not only the benefits and costs associated with material artefacts of development, but also the rights and capacities to influence the present and the future. Yukatek groups are often collectively referred to as 'Maya'. However, 'Maya' is also used locally to refer specifically to Mopan people and language. In this article I use the word in its more encompassing sense.

Case details are available at the University of Arizona's Indigenous Peoples Law and Policy
Program website: http://law2.arizona.edu/Depts/iplp/outreach/mayaBelize.cfm 5. Issues of maintenance, access and entitlement may also influence changes in the opposite direction. 6. The quotation, from A.C.S. Wright's (1959) colonial report, Land in British Honduras, refers to the drive to 'settle' Maya cultivation practices (Wainwright 2008: 87, 153-240). 7. Concerns about alcohol abuse were often linked with the increased availability of commercial rum in village bars since the 1990s (also attributed to improved road access). 8. In 1997, it was reported that a number of Q'eqchi' farmersamong them Belizean citizenswere coercively deported to Guatemala from a Belizean border village by the Belize Defence Force (Wilk 1997: xvii). 9. World Bank research in Belize has, however, questioned the assumption that roads are always effective for poverty alleviation (Chomitz & Gray 1996).