Open Research Online Below the Belt? Territory and Development in China’s International Rise

China’s internationalization has been heralded by some as a new era of South–South cooperation. Yet such framings of development are pitched at an abstract space of the ‘global South’ which conceals more than it reveals. With some theory moving towards ontologies of ‘global development’, we need to capture both the connectedness and the local speciﬁcity of increas- ingly diffuse processes. This article sets out a more ﬁne-grained understanding of how political territories and processes are imagined and produced by and through China’s internationalization, focusing on infrastructure as a ‘technology’ of territorialization. Much of the focus on China’s internation- alization has been on state-to-state relations, but this obscures the ‘omni-channel politics’ that China practises. Using a critical literature review and illustrative case study, this article develops the idea of omni-channel politics to posit a view of ‘twisted’ territories in which political processes and development outcomes are more complex and contingent.


INTRODUCTION: BEYOND MONOLITHIC SPATIALITIES
The implications of China's international rise for the political economy of development are profound, but our analytical tools are relatively blunt, particularly the spatial lenses we use. As others have noted, 'China' is often treated as homogeneous (Klinger and Muldavin, 2019) which in turn produces stark assessments of China's development impact as either 'good' or 'bad' (Obi, 2019). This homogenizing tendency is compounded when we consider where China engages, with broad-brush analysis of things like 55 'China-Africa' relations that treat 'Africa' as a single entity (Chan, 2008). Allied to this is the broader framing of the world in terms of 'North' and 'South' (Horner and Hulme, 2019) which does not capture China's ambivalent position within development processes, nor the unfolding complexity of the connectedness that its rise engenders. Thus, these polarized and partial debates about China's rise are rarely helpful in illuminating the types of changes we are actually seeing in development processes and outcomes. At one level, then, this article seeks to develop the recent arguments made for a move to 'global development' (Horner, 2020) as a way of capturing this unfolding complexity.
The most visible embodiment of China's assertive international role is the Belt and Road Initiative (BRI), which was mooted in 2013 and has subsequently morphed into a multi-stranded geopolitical and geo-economic project (Flint and Zhu, 2019). China's international rise is also evolving with the BRI and other initiatives moving the country's engagement beyond state-to-state deals towards what Nederveen Pieterse (2018) characterizes as 'omni-channel politics'. For good reasons, much analysis of China's engagement with the rest of the world has focused on inter-state relations (see, e.g., Burgos and Ear, 2010;Carmody et al., 2009). However, this state-centrism has obscured other political scales, actors and processes. In the context of the emergent BRI, the political processes operating alongside, above and below the belt -hence my title -serve to buttress China's influence as a development actor.
This article is concerned with why and, particularly, how China is practising its omni-channel politics in light of the need to embrace the more 'nuanced maps' (Sidaway, 2012: 56) of development outlined above. The prefix of 'omni-' to mean 'all' is broken down into 'multi-channels' (Nederveen Pieterse, 2018: 71-72): working within existing global institutions, seeking to reform the same institutions, and creating new institutions and frameworks. This framing of what China is doing is helpful but tends to be descriptive -organizations and what they do -and pivots around a statecentrism. As a version of multi-level governance, omni-channel politics fails to capture the various ways in which the Chinese state is fragmenting (Jones and Zeng, 2019), how politics enrols multiple actors (Hameiri et al., 2019), how specific elites shape state actions (Klinger and Muldavin, 2019), and how politics is relational and shaped by other actors beyond China (Mohan and Lampert, 2013). While omni-or multi-channel politics begins to capture the emerging complexity of China's international engagement, it fails to account for these underlying processes and how they are realized through more entangled -and potentially contradictory -spatialities of power. It is here that theories of territory can add value for, as Uitermark (2015: 1) notes, considering territory 'forces analysts to consider the geographical foundations of politics as well as the particulars of governance'.
This article argues for a more spatially nuanced understanding of China's rise using ideas of territory. At its most general, territory -as I examine 56 Giles Mohan below -refers to 'techniques for measuring and controlling space' (Woon, 2019: 115) and is mainly a product of the Westphalian state system. Much of the literature on China's internationalization, including the idea of 'omnichannel politics', implies Chinese actors and institutions driving these engagements in pursuit of particular interests. In this way the Chinese state rationally 'extends' its reach in multiple ways. Recent framings of territory move us beyond this by focusing on 'territorial systems' that 'emerge from the interaction of multiple actors with different capabilities, incentives, and design strategies, and at different levels of analysis, or scales' (Atzili and Kadercan, 2017: 121). Additionally, I focus on the technologies of territory; those 'material tools used to make territory real' (Branch, 2017: 137). Much is made of cartography as a territorial technology, but for China's internationalization I am concerned with infrastructures as technologies of territorialization.
The questions I want to address are, first, what forms of territory does China create through its internationalization? And, second, what role does infrastructure play in materializing these territories? Before addressing these questions, I outline the ways in which current spatial framings of development are inadequate to comprehend China's rise -a debate that Development and Change has championed (Fischer, 2019;Horner and Hulme, 2019). The arguments presented here are based on reflections across a series of projects, most of which have focused on China's engagement with a range of African countries. I begin by reviewing the literature on development and scale; while I agree that we need to retain development studies' focus on the political economy of an increasingly globalized capitalism, this has to move beyond some of the established spatial framings. Next, I review the literature on territory and argue that we need a dynamic view of territories as folded yet largely rooted in capitalist logics. From there, I assess how multiple internationalization processes play out, and the role of Chinesebacked infrastructure projects in them. Finally, I outline a short case study from Ghana which demonstrates the complex territorialization at play, before concluding.

THE SPATIAL LIMITS OF DEVELOPMENT STUDIES
In thinking through the more complex territorialization of China's rise we need to move beyond existing spatial framings of development. In this section I outline the limitations of these current spatialities as a way to address alternative territorial framings. Development studies has been defined by a range of spatial analytical scales. At its heart is a methodological nationalism that treats the nation state as the primary unit of analysis (Currie-Alder, 2016). In the context of the wave of post-war decolonization this framing made sense since the endeavour was to stimulate development to reduce both within and between country inequalities (Fischer,Territory and Development in China's International Rise 57 2019). In cruder, and ideologically loaded, versions of modernization theory, countries were often portrayed on a continuum whereby those 'lagging' could be engineered to 'catch up' with 'advanced' countries. In the context of China, Klinger and Muldavin (2019) note that assuming the country to be coherent and monolithic serves various ideological ends rooted in Cold War sensibilities around threat and the 'other', though they argue that China also projects itself as a unified whole that in turn conceals domestic disparities.
Beyond methodological nationalism, international development is built on an aggregation of binary scales or 'meta-geographical demarcations' in Sidaway's (2007: 336) terms, which Klinger and Muldavin (2019) argue are 'outmoded'. Categories such as 'developed/developing', 'first/third world' or 'global North/South' all attempt to capture something of the global inequality that development studies seeks to ameliorate. While such binaries have an enduring role (Grovogui, 2011), because they insist that power and wealth are not evenly distributed and that proclamations about the ending of US hegemony are premature, they can also obscure. One argument around their inadequacy is empirical because data suggest a growing differentiation within countries such that both development and under-development co-exist within states the world over, rather than the latter being located primarily in the developing world (Horner and Hulme, 2019). Allied to this is awareness of planetary ecology that renders borders irrelevant, particularly with respect to climate change (Horner, 2020). Such recognitions are used to argue for the idea of 'global development' that untethers development processes from particular places and simultaneously urges analysis at a 'smaller spatial scale' (Horner and Hulme, 2019: 369). Similarly, Comaroff and Comaroff (2012: 127) point out that 'there is much South in the North, much North in the South, and more of both to come in the future'; the first point refers to the precarity experienced in the global South that is increasingly being experienced in the global North, though clearly not something to celebrate (see also Bratman, 2011). However, as Arsel and Dasgupta (2015) rightly observe, the mere existence of Third World-like conditions in a given locality does not help us to explain such situations analytically.
Another argument against spatial binaries is more theoretical insofar as spatial categories are seen as forms of governmentality that need to be challenged in order to open space for alternatives. Since much development theory relies on broadly 'North-South' framings, I have struggled to use it to discuss China's international engagements. Raghuram et al. (2014: 120) capture this tension well: 'the vectors of power … have prioritized the global South/North distinction …. The dynamism and diversity of the global South, especially its manifestation in what has come to be known as Rising Asia, ruffles commonly accepted spatialities'. This presence of Asia complicates -'ruffles' in their terms -the accepted binaries and power relations of knowledge production where the critique has focused on the colonizer as Western imperialist. China is simultaneously 'Southern', 'Northern' and neither. China plays up its history of being colonized in certain instances, often as part of the wider discourse of 'South-South' development cooperation (Mohan, 2016). Grovogui (2011) discusses 'South-South' relationships as a form of 'lateralism' in which 'moral equivalences' exist between nations. He argues 'Such equivalences symbolically suspend asymmetries among the parties in order to enable desired relationships and effects' (ibid.: 182). At the same time China projects its growing power to position other developing countries as suppliers of raw materials, much as Western powers have done for centuries. This mixing unsettles categories which infuse our work -things like 'Chinese' capitalism or 'African' agency. This unsettling also prevents us from neatly and pejoratively linking spaces to actions -West = good, China = bad, or vice versa -links that pervade public discourse around contemporary globalization.
The move towards global development should not be used to abandon the particular in the face of a new universal condition -we are all (under)developed now. Rather, the task is to hold in tension the general and the particular. As Horner (2020: 427) notes 'paying attention to geographical variation in development challenges is a must in order the challenge both flat-world claims and one-size-fits-all, universal solutions'. Hart's (2018) framing helps chart such directions by focusing on 'spatio-historical specificities as well as interconnections and mutually constitutive processes' (ibid.: 373). If we unpick China's development processes through new connections and partially decentre the West, then we need to think about scale and politics in equally complex ways. As China internationalizes, we see the creation and narration of new territorial forms as befits a power that is entering a global order which, on the one hand, may be seeing the waning of US hegemony (Fischer, 2015) but, on the other hand, is fragmenting into multiple power centres. I now turn to these questions of politics and space.

TERRITORIES, NETWORKS AND CAPITALISM
In terms of China's growing international relationships, much of the focus has been on inter-state relations, but this has obscured other political scales and more complex spatially embedded processes. Here I examine the ways in which territory has been conceptualized and how it may be transforming. Broadly speaking we have witnessed shifts from seeing territory as a bounded and state-centred control of space towards a more complex articulation of bounded spaces and fluid, cross-cutting networks. Latterly this deand re-territorialization has posited more 'twisted' territories in which a multiplicity of actors (notably, but not exclusively, global firms) co-exist in complex spatial arrangements, even as they are constrained by the 'stickiness' of pre-existing territories.

Territory and Territorialization
Territory and territorialization are a subset of wider processes regarding the control of space for securing particular interests (Jessop, 2016;Sassen, 2013). While the nation state has become the dominant mode of territory under modernity (Elden, 2013), it is by no means the only important form; moreover, other socio-spatial orders prevailed in pre-modern times. While International Relations theory deals with nation states as the territorial building blocks of the global system, it has done so relatively unreflexively and it has been human geographers that have problematized and theorized territory (Atzili and Kadercan, 2017).
Human geographer Sack's (1983) attempt at an all-encompassing theory of territory was 'neutral' in that it posited how territory in general functions, but he concedes that it can be linked to different theories of power and political economy. For Sack, territory is essentially a form of enforcement or control that involves organizing and classifying by area, and the construction of borders to communicate this classification, as well as the enforcement of these borders. Agnew's use of territoriality echoes Sack's in that '[t]erritoriality is the strategic use of territory in either the organization and exercise of power, legitimate or otherwise, over blocs of space or the organization of people and things into discrete areas through the use of boundaries' (Agnew, 2011(Agnew, : 2589. In these processes of bordering, boundaries can be both physical (walls, demilitarized zones, gated communities, immigration controls, etc.) and symbolic (national identity, terra nullius, etc.) such that the representational construction of territory is significant (Elden, 2013) for creating exclusions and, most importantly, instituting property rights (Murphy, 2013).

Territory/Network
The idea of territory assumes bounding and closure, but this closure is problematic, because territories are increasingly de-centred by flows and networks. Territory is most commonly seen as relating to the nation state (Agnew, 2011;Uitermark, 2015) which implies a 'top-down' political strategy centred on a monarch or powerful political bloc. In contrast to statebased territorial fixity, networks are diffuse and open up 'the possibility of resolutions of a more contingent, open character than those imposed from above' (Cox, 2013: 48), although '[t]erritories and networks exist relationally rather than mutually exclusively' (Agnew, 2011(Agnew, : 2589. In addressing similar issues, Sassen (2013Sassen ( , 2018 posits emergent forms of territory which break down into two main types. The first includes non-national jurisdictions inside the state's territorial jurisdiction, but often only enrolling parts of these national territories. The second is types of bordered spaces that cut across the traditional inter-state borders, either through conventional 60 Giles Mohan transport linkages or digitally. For Sassen (2018: 7), these 'bordering dynamics are partly formalized, partly emergent, and partly not necessarily meant to be formalized nor to be particularly visible'. The metaphors move from 'nested' scales seen in theories of multi-level governance to twisted, messy or tangled geographies (Jessop, 2016). Cox (2013) argues that network theories have authority, because we cannot deny that more disembodied connections are proliferating under globalization. He goes on to argue that these theories are largely descriptive and do not explain the flows and connections as having any underlying logic linked, in his case, to capitalist expansion. Cox maintains that capitalist accumulation needs 'spatial fixes' (Harvey, 1982) which necessarily require the fixity of some territories that the network theorists have attacked. Production requires immobile facilities and transportation infrastructures (Ougaard, 2018), while much economic activity is regionally organized due to post-Fordist production logics. Nation states grant access to transnational capital while also controlling and containing the contradictions of this accumulation. It is, in part, these imperatives which explain the 'remarkable tenacity' of the modern state system (Murphy, 2013(Murphy, : 1214. From the point of view of analysing territory, Harvey (1996: 53) notes: 'Processes do not operate in but actively construct space and time and in so doing define distinctive scales for their development' (original emphasis). The key is to understand the processes of connection and how they construct spatial scale as opposed to starting with those scalar units. As such, Hart urges us to start 'with what seem to be important processes and practices rather than with any sort of bounded unit -be it nation, city, village, or whatever -and engaging in an initial round of abstraction or theorizing' (Hart, 2018: 389, original emphasis). The emphasis on processes and practices also broadens the perspective away from only the logics of capitalist expansion that Cox (2013) focuses on, to look at other, potentially non-capitalist, logics.

The Technologies of Territory
For an analysis of China's internationalization, I am concerned with practices and processes, but equally with the technologies of territory. Branch (2017) argues that territory as an institutional form requires the coming together of ideas, practices and technologies. Ideas about such things as the authority of the state have long been analysed by the social sciences, while practices -notably demarcation of borders -have also been analysed in some detail. Indeed, the earlier discussion of China's omni-channel politics largely focuses on the organizations and what they practise. The third element of Branch's framework is the technologies of territory, which 'includes the tools needed (or perceived to be needed) to implement territorial practices, as well as tools that are demanded, or given new meanings and uses, by territorial ideas' (ibid.: 137). Technologies are often implicitly lumped into discussions of practices, but technologies such as mapping or surveillance methods are the manifestations of practices. Infrastructures, such as roads, ports or ICT networks, are part of the dynamics which both facilitate bordering but also enable the more complex forms of folded territories that Sassen (2018) evokes.
More recently those studying infrastructure have focused on the ways in which infrastructure plays into state-building processes (Harvey and Knox, 2012). Here, infrastructure can be a means of spreading the state's reach across its territory in highly visible ways -such as roads -or of connecting contiguous territories into new territorial forms; indeed, the BRI is an 'imagined' territory that is being brought into being through the construction of transport infrastructure (Grant, 2018;Lin et al., 2019). In these accounts state power is very much central to infrastructure systems. But power is also entwined in infrastructures through the everyday interactions that 'users' have with them. Such interactions may defy the original goals of such infrastructure -for example, where the internet is used for subversion -or be purposive acts of resistance, such as sabotage of power lines. These everyday interactions disrupt framings of infrastructure as 'standardized' (Wiig and Silver, 2019). Wiig and Silver employ the concept of standardization to argue that for infrastructure to function it has to interface with other systems across space. Spatially, these connections break down into corridors (e.g. transport routes, pipelines) and nodes (e.g. ports, cities), which parallels the discussions of territory as both fixed and networked. Wiig and Silver argue that peopling studies of infrastructure pushes us toward its 'dynamic material and spatial (re)configuration' (ibid.: 916) and the multiple embedded relations and processes that constitute spatially diffuse infrastructure networks.

China's Internationalization and the 'Infrastructure Push'
The need for China to look beyond its borders for sources of energy and other natural resources resulted in the 'Going Out' strategy from the late 1990s, whereby China encouraged outward investment and international trade (Ayers, 2013;Chalmers and Mocker, 2017). As a spatial fix it was, in part, necessary in sectors where the Chinese market was relatively saturated or where domestic sources of energy were diminishing (Downs, 2011). This internationalization strategy was enabled by huge foreign exchange reserves and was given a boost during the global financial crisis when Western sources of credit declined while Chinese banks were relatively untouched. As such the 2008-11 period saw something of an international spending spree by Chinese state-owned enterprises (SOEs) investing in resources and/or infrastructure projects. More recently, this outward investment has begun to slow and its direction has shifted somewhat towards more advanced economies and relatively less into resource extraction (OECD, 2018).
As noted, the term 'omni-channel politics' describes how China is seeking to spread its influence. Geopolitically, we are seeing two linked moves -one is to seek integration into the global order while the other is to nudge things in a new direction that serves China's ends (Cai, 2018;Zhou and Esteban, 2018). As others have argued, China has engaged in creating new multilateral organizations as well as working to boost its influence in existing ones (Alimov, 2018;Cai, 2018;King and Du, 2018;Taylor, 2012). Again, such moves are seen by some to challenge (Western) spheres of influence. Indeed, the Asian Infrastructure Investment Bank (AIIB) and the BRICS' New Development Bank were partly about frustration with the lack of reform of the Bretton Woods Institutions, while Japan responded to the creation of the AIIB by boosting its funding of the Asian Development Bank.
These geopolitical moves by China reveal the wider point that the geopolitical and geo-economic are inseparable (Cowen and Smith, 2009). For example, Chinese infrastructure projects might combine elements of grants, concessional loans and commercial lending, with the condition that construction contracts go to Chinese SOEs. Infrastructure is a key focus for China and brings together the ideas, practices and technologies that Branch (2017) sees as constituting territory. Ougaard (2018) views the 'infrastructure push' among G20 countries as part of the need to absorb surplus capital and, by so doing, reduce the costs of increasingly global transacting. As such, things like the BRI and AIIB are part of this wider 'push'. With regard to China's international strategy, Bach (2016) describes it as having an 'infrastructural foreign policy' in which '[t]he visibility of Chinese infrastructure financing is central to its global image'. But infrastructure investment is also a spatial fix for Chinese capitalism, by generating overseas markets in sectors where the large construction SOEs (Brill and Reboredo, 2019) face diminishing returns domestically while resource-seeking deals are often tied into infrastructure contracts. Thus, focusing on infrastructure necessarily involves other sectors and investments.

THE TANGLING AND TWISTING OF CHINA'S TERRITORIAL FORMS
The BRI is the flagship of China's internationalization push and can, at one level, be seen as a 'grand strategy' which is part of the country's wider global aspirations (Narins and Agnew, 2019). Authors have pointed to the twin movements that tie Chinese nation building, territory and development together (Grant, 2018;Lin et al., 2019;Yeh and Wharton, 2016). One is a national project to integrate the country, particularly connecting and developing the western regions which were de-prioritized in the reform period when investment was targeted at the eastern coastal regions. The other is  (Woon, 2019) that seeks to project Chinese state power outwards. Xi Jinping's vision echoes earlier European framings of the 'Silk Road' which would link Europe to the East by traversing the relatively 'uncivilized' central Asian region (Mueller, 2019). The underpinning rationale is based on the 'development-security nexus promoting infrastructure construction as the main booster of economic growth which is in turn perceived as necessary for peace and stability' (Benabdallah, 2019: 99). The two movements are linked in that both aim to enhance China's connectivity through transportation and communication infrastructure (Wiig and Silver, 2019), and in that the BRI has evolved out of approaches used within China's borders (Grant, 2018). Indeed, this genesis of the BRI means that despite its appearance as a grand strategy it is rather more piecemeal and emergent than official proclamations might suggest. Jones and Zeng (2019) usefully argue that in practice the BRI is a loose policy 'envelope' whose 'fuzziness' is its strength. Spatially, this fuzziness is captured in the purposive failure to develop an official map of the BRI (Lin et al., 2019) which gives policy makers the leeway to adapt the socalled strategy as macroeconomic, geopolitical and local conditions change. In Sassen's (2018) framing, territory is created across national borders but is not formalized beyond these loose policy proclamations and high-level summits.
The policies and drivers that inhabit the BRI 'envelope' are numerous. Benabdallah (2019) sees the BRI as a way of pushing new norms and practices on relatively weak states in Central Asia and parts of East Asia and Africa, while also clearly opposing the US's containment strategy for China. None of the new multilateral organizations are Western-centred, enabling China to set the rules using a 'South-South' rhetoric to build commonality. Likewise, China is keen to internationalize the renminbi and is using the financing tied to the BRI to do this and to extend some of its e-commerce platforms (Narins and Agnew, 2019). Finally, China's western regions are under-developed compared to the coastal areas and so the BRI is a vehicle for extending infrastructure into these poorly served but politically volatile regions (Klinger and Muldavin, 2019). The malleability of the BRI is necessary because in practice it is largely a retrospective rationalization of a number of previously unconnected projects across Southeast Asia, Central Asia, Eastern Europe and East Africa. As such it is evolving on a projectby-project basis, and therefore enrols other states in bilateral relations. Individual projects often legitimize their value by evoking the BRI although, in practice, success requires elite commitments that are tied to the wider development strategies of both China and the recipient state. Where such commitments are lacking, as illustrated in Brill and Reboredo's (2019) study of the Modderfontein new city in South Africa, then infrastructure projects are more likely to fail.

Bilateral Relations and Beyond
This discussion of elite commitments to projects opens up questions around the relations between the Chinese state and other states of the global South. Invariably photographs appear of handshakes between the Chinese President and the leader of the (global South) state in which a project is being delivered, against a backdrop of the two countries' flags. All the symbolism speaks of state-to-state brokerage. Such relations are constituted through familiar modalities including Memoranda of Understanding, framework agreements, concessional and/or commercial loans, trade deals, or grants. The legitimatory discourses of these relationships are, as we know, centred on 'win-win' outcomes and 'South-South' development. As Grovogui (2011) argues, the assumed 'moral equivalence' of fellow Southern nations is based on shared histories of colonization, which often conceal power asymmetries between the countries. Crucially, part of the rationale for such deals is for the recipient state to recognize the territorial integrity of China by denying Taiwan's sovereign claims (Miller, 2017).
Yet this is a relational politics in which the agency of the global South state (and occasionally civil society) is important (Kragelund and Carmody, 2016). Indeed, 'the state' on each side of the bilateral relationship is far from homogeneous or static. Even the 'classic' bilateral model of bundled concessional financing deals (Brautigam, 2011) brings together a range of major state players on the Chinese side, including the Ministry of Foreign Affairs, National Development and Reform Commission, Ministry of Commerce (MOFCOM), SOEs and state development banks. Jones and Zeng (2019) have argued for a more nuanced understanding based on the idea of state transformation. Transformation relates to the Chinese state around processes of fragmentation, decentralization and internationalization. Fragmentation refers to piecemeal reforms of party-state apparatuses which create multiple and overlapping agencies; decentralization relates to empowerment of provincial level actors; while internationalization is where formerly national actors begin to operate beyond China's borders. The outcome is that 'China's top leaders must now bargain with, accommodate and coordinate a sometimes-unruly multitude of actors' Zeng, 2019: 1417). Central authority is not absent but is exercised through cadre appointments and discretionary government control, implying that governing a 'grand strategy' like the BRI is impossible. As such, the 'omnichannel politics' is not so much a coherent multi-level governance approach but a much more emergent and messier attempt to govern across space (Jessop, 2016).
A linked way in which we need to move beyond a view of coherent states is a focus on elites located in different parts of the state. Despite the discourses of 'South-South' equivalence, such bilateral relations are often based on inter-elite bargains (Carmody et al., 2009). Klinger and Muldavin (2019) argue that much of the bilateralism is symbolic insofar as, while deals are publicized through high-level handshakes, the implementation of projects is through local and transnational actors. Echoing Hart, they go on to argue: 'The key is to be attentive to the distinct configuration of actors and scales when investigating the relationship between uneven development in China … and China's global integration, rather than proceeding with an a priori notion of states and interests' (Klinger and Muldavin, 2019: 6). Often these deals are brokered through some form of parallel government, which compromises their transparency and accountability, effectively putting them outside of democratic control. In Angola, Corkin (2013) has analysed the ring-fenced institutions that were created to manage Chinese loans which allowed President Dos Santos and his coterie to direct these investments. From the Chinese perspective, these enclaves within but also outside of the Angolan state are about controlling risk, whereby dealing directly with key elites and power brokers reduces the risk of finance passing though state treasuries, or of opposition stifling project implementation. In the oil sector, as Moreira (2013) documents, the initial forays by China into global South producer countries were based on inter-elite relations, but as deals failed and it became clear that political risks were poorly accounted for the Chinese oil companies moved into joint ventures as a means of assuaging risk.
Focusing on the specificities of actors, interests and practices reveals multiple types of economic agents including the large SOEs that are not coterminous with the Chinese Communist Party (CCP), provincial SOEs and multiple forms of private capital. While the main winners from Chinese internationalization may be the SOEs, particularly those in construction (Jones and Zeng, 2019), there have been many failed projects, which speak of somewhat reckless investment decisions on the side of Chinese firms coupled with unstable operating environments in the global South. Moreover, much of the discussion of disaggregating the state rests on complexifying the Chinese state, whereas the relationality of projects and deals means elite-based political settlements on the global South side are equally important (Hickey et al., 2015). While these elite actors 'design' territories to serve changing ideas and interests (Atzili and Kadercan, 2017), everyday interactions or subaltern resistance can circumvent these territorial designs (Wiig and Silver, 2019). Where much of this localized interaction takes place is around the porous boundaries of specific projects, which may appear enclaved but in practice are transgressed in numerous ways.

Enclaves as Networked Territories
Another way in which scholars have sought to understand the spatial politics of Chinese engagement is through enclaves (Mohan, 2013;Sidaway, 2007). As territories, these enclaves are far from uniform and echo the complex dynamics already identified in this article; they reflect the interests of a range of actors, are designed for different purposes, may be more or less formalized, can have relatively hard or relatively porous borders, and may be subnational, but equally can cut across national borders. Such enclaves include Special Economic Zones (SEZs), turnkey construction projects, Chinatowns, edge cities and natural resource concessions (Tu Huynh, 2018). While enclave thinking has long roots (see, e.g., Hirschmann, 1958) it was revived by Ferguson's (2005) discussion of African resource investment, which he argued 'has been concentrated in secured enclaves, often with little or no economic benefit to the wider society' (ibid.: 378). Ferguson explicitly linked these forms of territory to wider processes of capitalist accumulation insofar as these enclaves constitute 'lily pads' that are connected through production chains and so entrench uneven forms of development while being 'noteworthy for their ability to bypass the nation-state frame altogether' (ibid.: 379). In this sense enclaves are both locally territorialized and linked to wider networked relationships.
What the focus on enclaved forms of territory brings is a greater awareness of the multiplicity of actors engaged in the construction of territory, but also how these infrastructure nodes aimed at capturing value become embedded in multiple other social, political and economic logics. For example, SEZs became widespread in China as part of its own export-driven industrialization and are now being set up in Africa and elsewhere (Brill and Reboredo, 2019;Giannecchini and Taylor, 2018;Oya, 2018). As forms of territory these zones have been treated as 'spaces of exception' (Arnold, 2012) where different -often draconian -rules apply and through which 'variegated sovereignty' (Ong, 2006) is exercised. While Ferguson's analysis of enclaves acknowledges the transnational, capitalist connections of these spaces, he underplays the ways in which state space and enclaves work together. Foreshadowing Sassen (2018), Easterling (2016) posits the idea of 'infrastructure space' as a site of 'multiple, overlapping or nested forms of sovereignty, where domestic and transnational jurisdictions collide, infrastructure space becomes a medium of what might be called extrastatecraft -a portmanteau describing the often undisclosed activities outside of, in addition to, and sometimes even in partnership with statecraft' (ibid.: 15, original emphasis). For Easterling, then, statecraft and extrastatecraft in the form of things like SEZs can work together rather than being 'lily pads' unconnected to national state spaces.
This discussion of territory and development suggests that we need to keep in focus the connections that geopolitical and geo-economic processes bring, but that such connections are realized in particular places with their own forms of formal and informal territorialization. Focusing on individual Chinese-backed projects allows us to apprehend the range of actors, interests and practices at work and how infrastructure acts as one technology for binding such relations together in and across space. To illustrate this, I outline a case from Ghana's oil and gas sector.

CHINESE OIL COMPANIES AND GHANA'S DEVELOPMENT
China's global engagement with oil builds on its domestic experiences with oil production. The history of Chinese oil companies is beyond the scope of this article but goes back to a range of ministerial reforms and restructurings over the 1990s (Taylor, 2014) which saw the creation of three large national oil companies (NOCs) -Sinopec, China National Offshore Oil Corporation (CNOOC) and China National Petroleum Corporation (CNPC). In terms of Jones and Zeng's (2019) framing of 'state transformation' these NOCs were partially privatized so that commercial pressures began to outweigh geopolitical ones, particularly around energy security (Obi, 2019). For example, CNPC's entry into Sudan in the mid-1990s was driven by the company rather than any geopolitical directive from Beijing (Patey, 2014).
It was only from the early to mid-1990s, when China's domestic oil supplies could no longer meet demand, that oil imports increased and its NOCs began searching for investment and supply opportunities overseas, mainly in the Middle East and Central Asia, but also Africa. Chinese NOCs entered the international oil sector that was already dominated by Western international oil companies (IOCs) and so was forced to purchase small and fragmented stakes in existing fields or invest in riskier regions (Chalmers and Mocker, 2017). There was also some evidence that these Chinese oil companies overpaid for these assets given that there were few available to purchase (Downs, 2007). Those fields that they can more easily accessnotably Sudan in the early 2000s -were off limits to Western investors due to sanctions (Large, 2009). It is this lack of opportunities that has often driven Chinese investment decisions rather than a proclivity for Chinese firms to invest in dictatorships as more hawkish accounts argue.
In securing overseas deals, Moreira (2013) focuses on the role of elites in brokerage. She argues that historically Chinese firms have preferred working directly with state elites, but that the risks of this model (e.g. changes in leadership, etc.) are seeing shifts towards joint ventures by Chinese NOCs with African NOCs and/or IOCs (Obi, 2019). In addition to mergers and acquisitions, Chinese NOCs and SOEs have entered into resource swaps. The socalled 'Angola mode' of infrastructure financing (Power et al., 2012) -an arrangement where oil is bartered for low-interest loans -while overblown in its significance for Africa, is tightly controlled since the finance never leaves China, despite the client nominally being the recipient country. In these closed tendering arrangements China's MOFCOM approves the contractor and then the finances provided by the Chinese development bank pass to the Chinese SOE once the African 'client' has approved their release (Brautigam, 2011).
In Ghana a gas processing plant was built in the country's Western Region between 2011 and 2015, which reflected a resource-seeking and marketseeking logic on the part of the Chinese NOC Sinopec. The case reveals how relatively enclaved infrastructure spaces -oils rigs and major production facilities -are connected to wider political processes and embedded in particular local political economies. The relations are not only entwined with domestic elite politics in Ghana, but also already entering into spaces inhabited by the World Bank and International Monetary Fund (IMF), which have hitherto played a major role in Ghana's economy and polity.
Early attempts by CNOOC to buy stakes in a newly discovered field fell through (Phillips, 2019) but in parallel Sinopec was negotiating an oil-forinfrastructure deal. That two Chinese oil companies were competing with one another for stakes in Ghana's oil suggests that there is no singular stateled initiative by the Chinese around securing African resources. The gas plant in question treats raw gas from the offshore Jubilee Field in order to produce fuel-grade gas for the Aboadze power station. The spatiality of this project is complex. The pipeline bringing gas onshore reveals the interconnectedness of infrastructures and how a quintessentially enclaved production facility, the offshore rig, links to material and political processes (Appel, 2012) in both fixed state territories (electricity production and sales, tax revenue, etc.) but also in transnational networks (Chinese oil companies and supply chains, Chinese banks, etc.).
Since oil production started in November 2010, just three years after proven commercial reserves were found, the Jubilee Field has experienced difficulties and produces an average of 100,000 barrels per day (bpd). There are currently 17 active contract areas in upstream operations and 20 discoveries made since 2007. Most of the 20 discoveries are still being appraised but two new ones are being developed -the Tweneboa-Enyenra-Ntomme (TEN) and Sankofa Gye-Nyame. TEN began production at the end of 2016 and is projected to produce 80,000 bpd. The Sankofa Gye-Nyame field is expected to produce 45,000 bpd of oil at its peak and began production in 2017. The three fields come with associated and non-associated gas, which is critical to reducing the cost of fuelling power plants to ensure stable and reliable electricity.
On the back of the oil discovery in 2007, the Government of Ghana began a series of negotiations with Asian infrastructure providers for some sort of resources for equity arrangement. The Korean construction firm STX got quite far with a US$ 10 billion loan for housing construction before it was held up indefinitely for 'technical reasons'. In mid-September 2010 the China Development Bank offered the Ghana government a US$ 3 billion Master Facility Agreement (MFA) on a non-concessional basis (Hardus, 2017). This loan facility was dedicated to the Western Corridor Gas Infrastructure Development Project comprising 12 subprojects. The loan was split into two equal tranches of US$ 1.5 billion and was to be paid at LIBOR 1 plus 2.95 per cent with an upfront fee of 0.25 per cent and 69 commitment fees of 1 per cent per year. Sinopec won the contract for the gas processing plant while repayment is through UNIPEC Asia, which is a wholly owned subsidiary of Sinopec. UNIPEC acquired an 'off-taker' agreement to lift Ghana's share of crude from the Jubilee Field, amounting to 13,000 barrels of crude oil daily for fifteen-and-a-half years to pay for the US$ 3 billion loan. John Mahama, the then Vice-President (later President) of Ghana, spoke of the deal: 'China has emerged as a significant source of credit to Africa, traditionally our partners have been the World Bank and IMF …. The process of accessing World Bank and IMF credit has been unfortunately quite tiresome and comes with a lot of strings … we find it easier to go to the BRIC countries' (quoted in Bloomberg, 2012).
The deal was seen to be one-sided, with UNIPEC getting assured supplies while no commitments were asked of the Chinese partners. The Chinese argue that such a system is able to deliver projects quickly which certainly seems to be true, and is one reason why this form of project funding and delivery is preferred by recipient leaders who often have an eye on political longevity. The World Bank was also negotiating for a loan to develop the gas infrastructure, but as Chen (2016: 2) notes in the context of the contract offered to Sinopec, it 'provided the Ghanaians with an easy all-inclusive deal that obviated the need to conduct laborious and independent stages of tendering and financing negotiations' (see also Mohan and Tan-Mullins, 2019).
The collateralization of oil to finance infrastructure inevitably became entangled with domestic Ghanaian politics. In Ghana, two main parties vie for elections and have held power between them since the return to multi-party democracy in 1992. They are the National Democratic Congress (NDC) which professes to be social democratic and more statist in its thinking, and the more market-oriented New Patriotic Party (NPP) -although the ideological divisions between them are not huge. Shortly after oil was discovered the NDC promised to use the oil resources to develop fertilizer and Liquefied Petroleum Gas cylinder industries, as part of a long-term development plan. Around the same time the government began drafting new oil and gas legislation, notably the Petroleum Revenue Management Act (PRMA), which sought to establish how oil revenues would be used for national development (Mohan et al., 2018). When it came to the design of the PRMA, the NDC favoured immediate extraction of oil rents to finance rapid development as opposed to saving for the future. Clause 5 of the Petroleum Revenue Management Bill -the draft of the final act -prohibited the use of the Petroleum Holding Fund as 'collateral for debts, guarantees, commitments or other liabilities of any other entities'. But once the bill reached Parliament, an NDC member of parliament proposed an amendment to allow for collateralization. The proposed amendment received overwhelming support from ruling NDC MPs, and the debates that ensued followed strict party lines. The issue was eventually put to a vote, which gave the ruling NDC the green light to use future revenues from oil as collateral in accessing loans.
At one level such behaviour by the NDC was less about a vision for development and more about the short-term need to secure an election victory. Financing infrastructure is a powerful way of demonstrating political commitment to an electorate (Mohan et al., 2018) but also a tool to spread patronage resources to party-linked capitalists who secure subcontracts (ACEP, 2013). While such short-termist political incentives certainly played a role in these debates, contestations over how oil revenues should be managed were also at play. Supporting the view that Clause 5 be amended to enable oil-backed loans, ruling NDC MPs pointed to infrastructural deficits and inadequate social services as justification for the need to collateralize oil resources. The NPP opposed the collateralization of future oil revenues, seeing it as 'nothing more than eating your dinner and lunch at breakfast time'. 2 As the debates continued, the NPP opposition began to soften its stance which opened the door for oil-backed loans meant primarily for infrastructural development. Playing into this were wider negotiations around sovereign debt in which the IMF raised the ceiling for Ghana's international loans to enable the Chinese loan agreement (Odoom, 2017).
The decision to work with Chinese banks and China's NOCs was less about some 'socialist' affinity between the NDC and 'Communist' China and more to do with the willingness of the Chinese to offer an oil for infrastructure loan, which the NDC used for development of its domestic gas industry. However, the MFA loan ran into difficulties. Over the summer of 2014 it was announced that only half the loan would be used (Tranche B), and of the 12 projects only two went ahead. It transpired that the issue was the price of oil that was agreed to pay the loan. The Government of Ghana estimated it would end up paying US$ 6.4 billion for the US$ 3 billion loan while 60 per cent or more of the contracts go to Chinese companies. In August 2014 President Mahama visited Beijing, cap in hand, to renegotiate: the Chinese remained intransigent (Citifmonline, 2014). In December 2016 the negotiations were re-opened, this time using sales of gas to finance the loan although final details were never resolved and nothing more was heard of it (Citifmonline, 2016). However, a new deal has been mooted around Ghana's bauxite reserves by Chinese SOE Sinohydro, which had earlier built a large dam in the north of the country as part of a concessional loan arrangement (Foreign Policy, 2020).

CONCLUSION
I have argued that China's rise as a development actor is being organized through twisted and overlapping territories. These emergent political practices and processes are not as coherent as the idea of 'omni-channel politics' 2. Parliamentary Hansard, 9 December 2010, Col 2820. suggests since they are constituted from both formal and informal practices, some working through existing political structures and others through newly created ones or even clandestine channels. This multiplicity of territorial strategies is necessary given that China is seeking to both align with and change the existing geopolitical structures while also creating markets for Chinese goods and services. Lots of work is going into this territory making that is material, political and ideological. Crucially, despite working through different scales and across networks, the Chinese state and the regimes of other countries remain central to these forms of territory. The key is how state power works in consort with informal political and geoeconomic structures and how these are organized in networked and enclaved ways.
This territorialization excites geopolitical rivalry, particularly with the US but also some Asian and European countries. Recently it has been contested by countries like Malaysia and Sri Lanka that have withdrawn from further Chinese loan agreements due to mounting and onerous debt levels (Yamada and Palma, 2018). There is a sense that through the BRI China may have over-extended itself and mis-read some of the domestic politics it engages with. These government rebuttals as well as civil society protest have sent out warning signs, making it likely that Chinese firms will behave more like Western investors (Chalmers and Mocker, 2017).
China is now an increasingly transnational player (Klinger and Muldavin, 2019). This means that we are all implicated and, as I have argued, it moves the locus of development (studies) away from a North-South axis to more diffuse and complex relations of connectedness -but ones that are still racialized, power-laden and linked to the development of capitalism. As scholars seeking to 'do' such connected research and to theorize its meanings and implications, we need more than ever to develop ontologies and to embrace methods that tease out interwoven histories and geographies. These ontologies, such as that of multiple territories elaborated here, do not abandon concepts such as peripherality (Fischer, 2015) or uneven development, but do unpack them at finer-grained spatial scales. In turn these complex connections mean we cannot easily argue that China's rise is a 'good' or 'bad' thing for the rest of the globe but must recognize that it varies by project and place.