Networks do not float freely: (Dis)entangling the politics of Tamil diaspora inclusion in development governance

Scholarship on diaspora engagement strategies has suggested that such strategies are embedded either in binary state-diaspora relations, or global structures of domination. This paper builds on the idea that diaspora engagement is contextually embedded but complicates the understanding of this context, by moving beyond structuralist or state-centric models. It draws on a range of relational theories, to suggest that diaspora engagement strategies in the development field are contextually embedded in complex entanglements of power relations. Data from a multi-method study of the Tamil diaspora in Toronto, from 2009 to 2018, reveals that inclusion in these diaspora engagement strategies is shaped by an entanglement of power relations, which include social networks, and legitimacy claims in overlapping cultural fields, but also spatial relations, whereby geography and material resources are often-overlooked dimensions of this space.


INTRODUCTION
In the summer of 2016, about a half-dozen representatives of non-profit organisations working in international development gathered in the Centre for Social Innovation (CSI) on Spadina Avenue in downtown Toronto to form what would become the Diaspora Engagement Networking Group (DENG). The meeting was facilitated by the Ontario Council for International Cooperation (OCIC), a Canadian government-funded organisation whose mission it is to connect and build the capacity of its member organisations working in the international development sector. The meeting was attended by representatives of organisations who were working with diaspora that had 'diaspora engagement' written into their strategy, as well as some who self-identified as members of a diaspora. Amongst them were international NGOs (INGOs) like Cuso International, with many years of experience implementing diaspora and development programmes, as well as local diaspora-run organisations still in the process of applying for charitable status.
Whilst the OCIC representative facilitated the session, they made clear that members should take charge. Participants revealed mixed reasons for attendance and also a broad range of understandings of what 'diaspora' meant to them. Most had only a vague understanding of the term but felt like it could be a useful frame to get their community to rally behind the UN sustainable development goals (SDGs), and most revealed an urge to leverage their privileged position in Canada. After this initial meeting in 2016, the group proceeded to meet semi-regularly, to exchange knowledge on how best to engage other diaspora individuals and organisations, and to discuss the functioning of the networking group itself. The group also organised a joint panel discussion at a global conference on the SDGs in Toronto in 2017, and designed and conducted a survey of organisations working with or through diaspora on any aspect of the SDGs.
Far from an epiphenomenal event, I suggest that the Toronto-based DENG is of broader global political importance in that it presents an example of a diaspora engagement strategy. In this paper, I define diaspora as individual and collective agents that mobilize socially, politically or economically towards a homeland in which they do not permanently reside. Further, I define diaspora engagement strategies as the policies and practices to manage diaspora involvement in political processes. From existing research, we know that there has been a rapid rise amongst migrant sending states in the designing of strategies to engage their diaspora populations (Délano, 2014;Délano & Gamlen, 2014;Gamlen et al., 2013;Mylonas, 2013;Ragazzi, 2009Ragazzi, , 2014Ragazzi, , 2017; e.g. through extraterritorial voting rights or dual citizenship schemes. But there has also been an increase in practices that specifically target diaspora mobilization in the field of development. In this paper I use the term development to denote a global 'field' structured around the problem of 'development' , with actors struggling over what development is, and how it can be achieved, albeit from different positions of power. 1 Development governance then describes one outcome of such a struggle whereby it has emerged that development needs to be managed, i.e. via rule setting, or the promotion of particular practices. Diaspora contributions to development then need to be equally managed; diaspora strategically 'engaged' . For example, actors such as the EU, the World Bank, NGOs and the private sector, have devised strategies to manage global remittance flows (Ratha, 2003) and their impact on development, through implementing micro-banking schemes and taxation. Diaspora engagement strategies in international development may also include holding policy consultations with diaspora on international aid policy, creating knowledge exchange and volunteering programmes, building diaspora organisational capacity to deliver programmes in home countries, and providing opportunities for networking in the diaspora's new country of residence, as outlined in the vignette.
Why is this important? Besides leveraging diaspora remittances for economic growth, the stated purpose of diaspora engagement in the development field is often to 'include diasporas in decision-making' . As the above example of the DENG illustrates, diaspora engagement strategies are described in such a way that emphasizes transparency and participation in line with dominant norms in the international development field. Yet, in doing so they often establish different kinds of barriers to participation in development governance. Diaspora engagement strategies do not simply provide space for localised and non-hierarchical cooperation and knowledge exchange. They create new hierarchies amongst diaspora organisations, foster certain behaviours and practices over others, leading to a reshuffling of the international development field as a whole, or simply reifying already existing global power dynamics.
How do we make sense of the DENG and diaspora engagement strategies more broadly? Why was the DENG formed and what was its purpose? At this point in time, Sri Lanka did not have an official 'diaspora strategy' , and was not a country systematically 'embracing' or 'tapping' its diaspora, for development purposes or otherwise. Besides, the phenomenon described here takes place in a host country setting. But neither in Canada could a formal engagement policy be identified. After the election of 2015 had brought Justin Trudeau's Liberal Party into power, Global Affairs Canada (GAC), had just replaced the Department of foreign affairs, trade and development (DFATD), and the new government had not yet formulated an official international assistance policy. Scholarship on 'diaspora engagement strategies' that are grounded in the home or host states, thus do not seem fully equipped to deal with the phenomenon under investigation. But neither do critical structuralist approaches that see diaspora engagement as a function of global capitalism (Ragazzi, 2014;Varadarajan, 2008Varadarajan, , 2010, or the maintenance of liberal order in the global South (Laffey & Nadarajah, 2012). Whilst the emphasis on structural dynamics and constraints is compelling, such macroapproaches tend to brush over local variation in the political struggles that might inform diaspora engagement strategies, and states remain the primary units of analysis. That the formation of the DENG was fairly ad hoc, its operation driven by micro-interactions between actors at multiple scales, only becomes evident when we move away from a topdown view of already formalised diaspora policies. With this in mind, some scholars of national diaspora strategies have called for increased attention to the micro-foundations of these strategies (Mylonas & Delano, 2017). I will follow the lead of these scholars, but will apply the micro-lens to diaspora strategies in the development field, rather than centering single state bureaucracies. A micro-perspective will also complicate the picture of who engages and governs diaspora, as there is evidence to suggest that diaspora should be conceptualized not just as subjects that are engaged, but as governance actors in and of themselves, with the capacity to influence global political processes (Craven, 2018).
Evidently, the nature of diaspora inclusion in development governance needs to be better understood, but in a way that eschews the binary of being either purely state driven vs. a symptom of the expansion of neoliberal governmentality practices across the globe. This paper seeks to rise to that challenge, by building on data collected during nine months of fieldwork for my PhD amongst members of the Tamil diaspora in Toronto, Canada. By assuming an ethnographic perspective on diaspora engagement strategies, this paper will deepen our understanding of what constrains and enables diaspora inclusion in development governance. By drawing on relational theories, I will disentangle a number of power relations that shape instances of diaspora inclusion in development governance, namely connectivity, legitimacy, and spatiality. I will argue that the inclusion of diaspora in development governance processes and practices is always conditioned by the interrelationship between these dimensions, rather than just one or the other. This paper will proceed as follows. It will first situate this study in relation to existing literature on diaspora engagement. It will examine explanations for the emergence and proliferation of such mechanisms within (emigration) states, at the global level, as well as taking a look at theories that emphasise the contextual embeddedness of diaspora mobilization. Next, the paper will sketch an alternative analytical framework by engaging with relational theories and the spatial turn in IR, including theories on socio-spatial positionality and assemblage thinking, to broaden our understanding of the power relations that diaspora engagement strategies are embedded within. This section will also outline the research design. The following section will then discuss research findings. I will introduce historical data from my study of the Toronto-based Tamil diaspora and their encounters with and inclusion in development governance, centering on the period between the end of the civil war in 2009 until immediately after the 2015 Canadian elections. This will set the scene for a more in-depth analysis of ethnographic data collected at various events centering diaspora and development in Toronto. Ultimately, I will show that each instance of purported diaspora 'engagement' or inclusion in development governance entails that diaspora actors must carefully navigate their position in professional, organisational and diasporic networks, balance questions around legitimacy (specifically effectiveness and representation), and secure economic capital to ensure material access to spaces of decision-making.

THE POLITICS OF 'DIASPORA ENGAGEMENT'
A rich and substantive literature has emerged in recent years that explains the formation and spread of diaspora engagement strategies (Délano, 2014;Délano & Gamlen, 2014;Gamlen, 2014Gamlen, , 2019Gamlen et al., 2013;Ragazzi, 2014Ragazzi, , 2017. This literature takes as its conceptual starting point the 'deterritorialization of the nation state' (Basch et al., 2005), empirically manifesting in the increase in mechanisms, policies, institutions and practices that have emerged to channel the (potential for) economic and social remittances of diaspora populations. Scholarship has tended to focus on formal policies, for example the extension of voting rights, special visa programmes or the creation of diaspora engagement institutions by home states. Indeed, most of this literature suggests that diaspora engagement politics are firmly embedded in relations between the diaspora and its home state (Gamlen et al., 2013;Koinova & Tsourapas, 2018;Kuschminder & Siegel, 2011;Mylonas, 2013;Tsourapas, 2018Tsourapas, , 2020, and that states 'govern' or manage their diaspora according to national interests. Differing reasons have been given to explain why states might have an interest in 'engaging' their diaspora populations. Diaspora engagement strategies have been conceptualized as exercises in nation-building, or economic calculation. Gamlen et al. (2013) have created a typology of reasons for the proliferation of diaspora engagement institutions. They suggest that some states seek to 'tap' their diaspora for economic gain, or 'embrace' them for symbolic gain, whilst others create diaspora engagement policies due to 'norm diffusion' . One criticism that has been levelled at this literature is that it has tended to focus on the extraterritorial practices of democratic states. These practices are, on the whole, designed to encourage transnationalism, rather than suppress it. In contrast, building on ground-breaking work by Laurie Brand (2006), which precedes the turn to 'diaspora engagement policies' , a growing group of scholars has begun to study the extraterritorial practices of authoritarian states (Dalmasso et al., 2017;Glasius, 2018;Tsourapas, 2015Tsourapas, , 2018Tsourapas, , 2020. Authoritarian states, these scholars argue, show us that there exists a darker side to diaspora engagement, one that is perhaps less enshrined in formal policies and institutions and that is driven by state's security concerns. As such they may constrain transnational political action (Chaudhary & Moss, 2019), by repressing and intimidating expatriate dissidents (Moss, 2016;Öztürk & Tas, 2019).
Further, whilst the bulk of literature on diaspora engagement has centered empirically on the (liberal or authoritarian) emigration state, some scholars have concerned themselves with decentering the state in the study of diaspora engagement strategies because "both state and non-state agents are implicated in these projects" (Ho 2011, see also Craven 2018). Similarly, others have noted that studies of diaspora engagement strategies "need to focus more on actors and spaces" beyond the state (Delano & Gamlen 2014;Adamson 2016). Heading these calls, Gamlen sheds light on how diaspora policies relate to wider global governance processes. In his 2014 article, he draws on the literatures on policy diffusion and epistemic communities, as well as governmentality theory, to make the argument that 'the rise of diaspora institutions is driven by efforts to form a coherent but decentralized system of global migration governance' (Gamlen, 2014, 192). In his follow-up book, Gamlen (2019) then suggests that this increased effort to build a global migration governance regime is fostering 'strategic competition over populations rather than territories' , or what he calls Human Geopolitics.
In the wake of the increased attention payed by both states and other global governance actors to diaspora, authors have pointed their analytical lens at structural forces that drive state−diaspora relations, and thus explain the rise of diaspora engagement strategies. In contrast to the more policy-oriented literature, they make explicit the oftenexploitative power relations, which underlie diaspora engagement practices. Some have argued that through engaging diaspora, states seek to increase their power and build their capacities. However, accounts differ in terms of the nature of the structure or logic that shapes diaspora engagement. For example, Varadarajan (2008) suggests that contemporary state attitudes towards diaspora must be understood as part of the hegemony of a neoliberal global political economy, whereby diaspora communities offer opportunities for expanding the global capitalist system, in favour of powerful Western states. Meanwhile, Ragazzi (2014) uses Foucauldian governmentality theory to explain why states are increasingly interested in engaging their respective diaspora, in an effort to reproduce the global political economy. Similarly, Laffey and Nadarajah (2012) suggest that diaspora engagement forms part of a liberal government logic, albeit this time as part of a larger effort to securitize and generate liberal order in the global South. These accounts illuminate how diaspora engagement strategies are embedded in a discursive or political economy structure, dominated by state capitalism and hierarchies of global exploitation. Whilst the emphasis on structural dynamics and constraints is compelling, such macro-approaches tend to brush over local variation in the political struggles that inform diaspora engagement strategies. States remain the primary units of analysis and are assumed -a priori -to be the main locus of power in global politics. Further, whilst, these approaches allow us to conceptualize constraints to diaspora engagement beyond single state interests, they limit our thinking about possibilities for counter-hegemonic action and diaspora agency, and do not take into account variation or interpretations of policies in particular social or political contexts.
Other authors have sought to decenter the state by looking inside, rather than beyond it. For example, Délano and Mylonas' (2017) JEMS special issue explores the micro-foundations of diaspora politics. I intend to build on this work, but apply the micro-lens not to single state bureaucracies, but rather to diaspora strategies in the development field.
Such a micro-perspective will also complicate the picture of who engages or governs diaspora, as there is evidence to suggest that diaspora should be conceptualized not just as subjects that are engaged, but as governance actors in and of themselves, with the capacity to influence global policy processes from the agenda setting to the delivery stage (Craven, 2018). The lines between governors and governed thus become increasingly blurred. Diaspora are disciplined and managed, but they also organise and engage in collective action. Thus, in order to understand the nature of diaspora inclusion in development governance, specifically the micro-foundations of diaspora engagement strategies, it is essential that we look at the rich literature that exists on diaspora mobilization, in the area of development, but also in international politics more generally.
Building on early empirical studies that explored the role of diaspora in conflict and post-conflict reconstruction (Adamson, 2002(Adamson, , 2013Baser & Swain, 2008;Collier & Hoeffler, 2004;Demmers, 2002;Koinova, 2018;Lyons, 2007;Mohamoud, 2005;Orjuela, 2008;Shain, 2002;Van Hear & Cohen, 2017;Zunzer, 2004) the role of diaspora in (postconflict) development has been increasingly explored (Bakewell, 2009;Brinkerhoff, 2008Brinkerhoff, , 2011Newland & Patrick, 2004;Pellerin & Mullings, 2013;Sinatti & Horst, 2015;Turner & Kleist, 2013;Van Hear et al., 2004). Empirically, studies have shown that diaspora mobilization for development ranges from indirect contributions to development like remittance sending (Helweg, 1983;Ratha et al., 2007), innovation (Minto-Coy, 2016) and knowledge transfer (Tejada in Chikanda et al., 2016), to more direct contributions such as volunteering locally (Darieva, 2017), contributing to infrastructure (re)development, especially in post-conflict settings (Kleist, 2018), and even institutional redesign and restructuring of public services in the homeland (Craven, 2018). Meanwhile, theoretical explorations of the link between diaspora and development have focussed on conceptualizing diaspora capacity and agency to contribute to development. An early study by Brinkerhoff (2008) explored empirically how diaspora contribute to development in their homelands, whilst in later work she conceptualizes diaspora as brokers or norm entrepreneurs with an 'in-between advantage' . The concept of the in-between advantage suggests that agency/power is in some way intrinsic to diasporic individuals, who are uniquely positioned and thus prone to entrepreneurialism. The idea of the entrepreneurial diaspora resonates with earlier studies in IR that have sought to make sense of the power of non-state actors in global governance. Crucially, Keck and Sikkink (1998, 1999 applied the concept of the norm entrepreneur to the actions of transnational advocacy networks. For them, it is connectedness across national and international scales/actor-bridging networks that enables norm-driven global advocacy. Keck and Sikkink (1998, 4) argue that network thinking bridges an 'increasingly artificial divide between international and national realms' .
However, the political realities of the 21st century have shown that ideas or norms do not travel that freely across scales or spaces. At least since Brexit it is evident that global ideas and norms do not diffuse evenly or uniformly across rural or urban spaces, even in countries as small as the United Kingdom. So, whilst ground-breaking at the time of writing, the agency-centrism of the above-mentioned approaches do not fully capture the ways in which relations in networks can both enable and constrain. Whilst networks matter, there has perhaps been a tendency amongst both governance actors and liberal constructivist scholars to conceptualize them as flat, non-hierarchical structures that enable connection, but do not constrain it (Jones et al., 1997; for a critique see Sørensen & Torfing, 2016).
Evidently, whilst it is unhelpful to think of diaspora engagement strategies as entirely suppressing diaspora agency in global politics, as some of the earlier cited accounts tend to do, an overemphasis on diaspora agency equally obscures the complexity of what actually unfolds during such processes; top-down diaspora management and bottom-up mobilization are often closely intertwined. As will become evident, transnational connectivity alone does not guarantee that diaspora will be included in development governance or decision-making. In recent years, significant theoretical advances have been made in the study of diaspora mobilization. Scholars have begun to emphasise the importance of thinking of this mobilization as 'contextually embedded' . This is important, as it suggests a more nuanced understanding of diaspora agency and the opportunities and constraints to mobilization in different circumstances.
But there is variation in how contextual embeddedness is thought about. Whilst some scholars think of diaspora mobilization as embedded entirely within the hoststate (Kuschminder & Siegel, 2011, others highlight the interplay of multiple hoststate contexts (Karabegović, 2017). Adamson and Koinova (2013) have written about how diaspora's embeddedness in (global) cities gives them power through access to networks and other forms of social, political, and economic capital. Horst (2018) argues that civic engagement by Somalian diaspora youth is contingent upon multi-sited embeddedness, whilst simultaneously also being a practice through which they create belonging (to a multi-sited identity/community). In other words, mobilization is shaped by belonging to multiple places, and viceversa. The importance of contextual embeddedness in shaping diaspora power and capacity to mobilise is explored in a recent 2018 JEMS special issue, with an empirical focus on spaces of contested sovereignty and conflict. The editors compellingly argue that (b)eyond considering diaspora agency as an actor in world politics, mainstream literatures need to be further aware that diasporas are not free-floating individuals, groups, or networks, but are embedded in contexts -local, national, supranational, and global -that shape their activism and are eventually shaped by it. This paper agrees with the above studies on the importance of thinking about contextual embeddedness for diaspora mobilization, albeit with a different empirical focus on international development governance, rather than conflict dynamics. Here, the contribution by Koinova (2017) is of particular importance. She deploys the concept of 'socio-spatial positionality' in her analysis of diaspora power and advances the argument that, whether diaspora should be considered an asset in post-conflict state building is dependent upon their linkages to different global contexts.
Rather than think of positionality only in terms of identity and social relations, socio-spatial positionality encompasses also a spatial dimension of power. Diaspora power in global politics is thus shaped by (multiple) linkages to particular places. Importantly, she theorises more precisely the nature of the context, namely as one characterized by different but interrelated forms of power. This is important. What does such an understanding of diaspora power as 'sociospatial positionality' offer to the study of diaspora engagement strategies in development? It suggest that we need to take seriously the situatedness of diaspora in different contexts; contexts in which social, and spatial forms of power are entangled. In the next section, I will develop the analytical framework through which to better understand the nature of diaspora inclusion in development, as well the methods used to generate and subsequently analyse the data.

Conceptual discussion
One aim of this paper lies in questioning the assumption that diaspora engagement strategies in the field of development can be explained by focusing either on top-down structural constraints or diaspora agency asserted from the bottom-up. As discussed, most existing accounts of diaspora engagement strategies look at structural factors, such as global governance norms, the global political economy, or geopolitics. This is somewhat presumptive, a priori excluding the possibility for multiple forms of power to be in operation at any one time. It also risks reification of state centrism and methodological nationalism. Meanwhile, relational theories make room for the notion that no one form of power exists on its own; that diaspora can be both enabled and constrained in multiple ways in the context of a diaspora engagement strategy. Building on such theorizing, I suggest that the politics of diaspora engagement strategies are bound up in complex entanglements, e.g. of agency and structure, global and local, and so must be gently and carefully untangled. For example, when a diaspora organization moves from being primarily focused on attaining charitable status and hustling to get 'a seat at the table' of decision-making on development governance, to actively shaping the norms around diaspora engagement in development governance, the lines between governed and governor become increasingly blurred. It is my intellectual ambition to be able to hold this sort of complexity, whilst also advancing knowledge on diaspora engagement, more broadly.
How can we theorize about diaspora engagement in a way that takes full account of the complexity of relations in which it is embedded? The next section will discuss the theories and conceptual building blocks that underlie my research design, followed by the methods used to generate and subsequently analyse my data. This study adopts a broadly relational approach, emphasising relative positionality, embeddedness and entanglements, over universals, absolutes, or binaries. Importantly, I wish to foreground relational theories with a materialist dimension. After all, diaspora engagement strategies, or diaspora governance is not just a virtual or ideational phenomenon, even if it is exercised through networks. Specifically, this notion that networks are embedded in space/place, echoes some of the recent forays of IR scholars into the study of Actor-Network-Theory. For example, Nexon and Pouliot (2013, 344) point out the false divide between ideational and material forces, suggesting that 'relations are simultaneously semiotic and material' . The material embeddedness of Actor-Networks is of central importance to this paper. Further, in her 2017 study, Koinova draws on Bourdieu's 'field theory' to theorise about socio-spatial diaspora power. For Bourdieu, 'fields' are arenas of production and circulation that are structured by hierarchical power relations. Specifically, he theorizes the interrelated forms of power that structure these fields. Rather than focus purely on economic capital or material power (as political economy approaches would have us do), fields are structured by social, cultural and economic capital (Boudieu, 1984). Thus, in this paper, I will use the term field in the Bourdieusian sense, to denote a hierarchically structured arena (e.g. centered around development) in which different forms of capital are simultaneously at play. But I will not rely entirely on Bourdieu for my analysis of diaspora engagement strategies. Whilst a useful to concept think from, in this empirical setting it is unlikely that we find a discrete 'field' (in the IR and in the Bourdieusian sense). Governance will be more fractured, situated in multiple fields that often overlap (e.g. local, national, global). A relational concept which can help us think about how governance is fractured or situated in overlapping fields is the assemblage.
Assemblage thinking has recently become popular in IR as a way of making sense of the operation of power in international and global politics (Acuto & Curtis 2014). Assemblages are conceptualized as a system of relations between human and non-human agents (Abrahamsen & Williams, 2009. Similarly, DeLanda (2016, 2) emphasises that assemblages are characterised by two things: 'that the parts that are fitted together are not uniform either in nature or in origin, and that the assemblage actively links these parts together by establishing relations between them' . Dickinson (2017) has employed assemblage thinking in the analysis of diaspora engagement, albeit from a macro human geography perspective. Unlike traditional network thinking, assemblages allow us to expand our analysis of relations to include relations between agents that traditionally exist on different ontological planes (individuals, organisations, states), or in overlapping fields of power, and also relations between these actors and their material environment.
For example, Dittmer's (2017) work on Diplomatic Material has been ground-breaking for using assemblage theory to understand the agency of material buildings, like the UK Foreign Office, or stacks of paper found inside office buildings, in shaping global political processes.
Ultimately, what all of the above relational theories share is a concern for entanglements between different forms of power, as well as a materialist dimension. Thinking through these theories with the empirical data from my case study, I have operationalized a relational analytical frame through which I disentangle the different forms of power that play a role in the politics of diaspora engagement. Specifically, I will show how connectivity, legitimacy, and spatiality all work together to shape the nature of diaspora inclusion in development governance. Ultimately, this paper argues that the inclusion of diaspora in development governance (such as consultation processes/capacity building) is conditioned by entanglements and complex interrelationship of social networks, cultural fields, and the material environment.

Data collection and analysis
The argument brought forward in this paper builds primarily on data collected during fieldwork for my PhD, from June-August 2016 and subsequently from January-June 2017 in Toronto, Canada. I chose Toronto as a site where I expected that the 'global politics' around Tamil diaspora engagement in development governance would be revealed. By this I mean struggles that go beyond the diaspora-homeland relationship, for example around global development norms, or, related to this, the geopolitical concerns of both diaspora host-and homeland. I draw conceptual and methodological inspiration from anthropologist Tsing (2011), who has pioneered the study of global-local connections, and whose work on Friction lends itself to the study of global-local entanglements.
I thus conducted multi-method fieldwork amongst members of the Tamil diaspora that were mobilizing towards their homeland in various ways, as well as the practitioners and policy makers that were involved in governing this mobilization. Centrally, I participated in and observed events during which I recorded ethnographic fieldnotes of diaspora transnational mobilization and governance in action. For example, within the first few days of arriving in the city in the summer of 2016, I was invited to participate in the first meeting of the above mentioned DENG. I subsequently attended 3 more of its in-person meetings from January to June 2017, became part of an email group where information on diaspora and development was shared, as well as participated in two larger multi-stakeholder fora where the DENG was either discussed or its members were present. I also conducted (in English) over 40 open ended interviews with the diasporic and non-diasporic elites that I encountered in these spaces, and completed desk research, by collecting and analysing secondary academic sources and a collection of policy documents, news articles, government documents, and think tank reports on Tamil diaspora repression/governance and diaspora governance more broadly.
I recorded both my secondary, interview and ethnographic data in NVivo. I subsequently analysed my field data in light of the literature on diaspora engagement and governance practices, iteratively establishing a relational analytical framework around dimensions of connectivity, legitimacy, and spatiality, as discussed above.
I will now outline the emergence of Tamil mobilization for development and the governance of such mobilization in Toronto, before delving deeper into an analysis of the governance practices I encountered during my fieldwork.

A brief history of Tamils and development governance
This section will lay out some of the key historical conditions and junctures that have shaped Tamil mobilization for development and have made Tamil diaspora inclusion in practices of development governance, such as the DENG, possible. It begins by introducing us empirically to the phenomenon of diaspora inclusion in development governance.
Linking diaspora and development: From the United States to Canada Whilst the connection drawn between migration and development was discussed throughout the 20th century, usually mirroring dominant political economic theories (de Haas, 2010), the specific linkage of development to diaspora took off only in the early 2000s. Here, a driving force was the publication of a World Bank report on remittances, which illuminated just how large a proportion of global financial transactions was made up by money sent by migrants to their countries of origin (Ratha et al., 2003). Word amongst DC's policy-making community travelled fast and soon the diaspora and development nexus was expanded beyond remittance sending (Newland & Patrick, 2004). In Canada, federal interest in engaging diaspora communities was off to a slow start, despite the country's high immigrant density and significant remittance sending. 5 Whilst, under Stephen Harper's conservative party leadership, funding of the Canadian International Development Agency (CIDA) was cut, 6 the Standing Committee on Foreign Affairs and International Development in 2012 began calling witnesses for a report on the role of the private sector in international development. 7 Invited were a number of experts from the US, who made a point to inform the parliamentary committee about the importance of diaspora as private sector actors in the development field. Whilst some MPs in the committee became interested in exploring this 'diaspora option' (Pellerin & Mullings, 2013), largely because they themselves had large migrant constituencies, at the federal level interest largely fizzled out. Further, at this time diaspora groups themselves seemed reluctant to be seen merely as remittance senders or money givers by the Canadian state. 8 At the municipal level, things looked slightly different. Driven by the talks on diaspora and development in Ottawa, in 2012 the Toronto-based peace-building NGO the Mosaic Institute hosted an event in cooperation with Washington DC-based IdEA, titled Diasporas@Toronto. This event was also one of the earliest occasions that saw Toronto-based Tamil diaspora organisations engaged in a wider inter-ethnic development network. But Tamil participation in this event was met with disapproval, in particular by the more established Tamil diaspora community long engaged in political activism (Amarasingham, 2015). The following section will explore why this was the case.

Tamils in Canada and their relationship with Sri Lanka
Of the approximately one million strong global Tamil diaspora, estimates suggest that up to 180,000 of those currently reside in Canada, 9 most in the Greater Toronto Area (GTA). Converging push and pull factors in the 1980s make it possible to account for the size of the Tamil community in Canada today. Almost in conjunction with the outbreak of civil war in Sri Lanka in 1983, a number of political shifts occurred in Canada affecting also Canadian immigration policy.
As Amarasingham (2015, 78) has argued, 'Tamil migrants (. . . ) benefited from the period of "turbulence" that marked Canadian immigration and refugee determination in the 1980s' . Of course, Canada was also a preferred destination for Tamils based on their familiarity with the English language. Despite a large proportion of Tamils settling in Canada as political refugees in the late 1980s and early 1990s, today the Tamil diaspora in Canada is very heterogeneous, in terms of immigrant generation, religion, caste, and class (Gunasingam, 2014; for discussions of the relevance of caste in the Tamil diaspora see Paramsothy, 2018;Troulliet, 2020). Perhaps the most central cleavage amongst diaspora Tamils in Toronto, but also globally, is found in their position towards homeland politics. While, it is difficult do justice to the complexity that characterises the relations of the Tamil diaspora with their homeland, both during the conflict and after, I will highlight a number of dynamics that have conditioned the engagement of the Toronto-based Tamil community in development governance today.
The engagement of the Tamil diaspora community in homeland politics (including development governance) has overwhelmingly been regarded as contentious. Primarily, this is because the Tamil diaspora, particularly those residing in the UK and Canada, were viewed as 'peace-wreckers' or spoilers during the civil war (Baser & Swain, 2008;Eichhorst, 2007;Newman & Richmond, 2006;Zunzer, 2004). There is evidence to suggest that the diaspora was the main financial source of the insurgent group fighting the Sri Lankan state government for national self-determination, the Liberation Tigers of Tamil Eelam (LTTE). Some liberal advocacy organisations have highlighted that many of the financial contributions to the LTTE's war-effort were violently extorted from the Tamil diaspora community (Becker 2006). However, this does not negate the fact that diaspora Tamils were also voluntarily sending large amounts of money to their families and relatives, to ensure their survival and also to further the Tamil struggle back home. Whatever the 'true' reason behind the high number of remittances reaching the LTTE-controlled areas during the Sri Lankan civil war, it made the Sri Lankan state government, as well as the majority Singhalese population on the island extremely suspicious, nay fearful, of the Tamil diaspora overall. This suspicion continues to this day. 10 Thus, one reason for the dearth of Tamil diaspora inclusion in development governance is that 'the government (has viewed) the diaspora primarily through the lens of security, and not development' (Amasingam & Poologaindran, 2016).
But suspicion swings two ways. Throughout my research amongst the Tamil diaspora community in Toronto, inquiries about engagement in development were met with bewilderment. Indeed, the concept of 'development' was firmly associated with repressive activities that the government had been pursuing in the North and East of the island since the defeat of the LTTE. 'Development' was essentially understood by Tamils in the diaspora to signify Singhalese military occupation and land grabbing, for example, it was the 'Presidential Task Force for Resettlement, Development, and Security in the Northern Province' which took over the governance of majority Tamil areas after the war (Guriybe & Tharmalingam, 2017, 182). Finally, a large proportion of the Tamil diaspora in Toronto today views engagement in development as a distraction from more pressing areas of Tamil post-war diaspora mobilization, such as lobbying for accountability and human rights at the UN or with the Canadian government.
Certainly, the end of the war in 2009 must be considered a window of opportunity for Tamil diaspora engagement in homeland development. Not only did the extremely violent fighting in the North and East of Sri Lanka, killing countless Tamil civilians, lead to an increase in the perceived need for external humanitarian and development assistance. 2009 also presented an important turning point for international geopolitical engagement with the Sri Lankan state, and by extension the Tamil diaspora community (Brun et al., 2012;Guyot, 2018). As evidence of war crimes committed by the GOSL against Tamil civilians began to mount, international perception of the Tamil diaspora also began to shift.
Whilst in Canada, and in Toronto in particular, Tamils had made headlines through their involvement in gang violence, 11 and subsequently as terrorist-sympathisers, in 2009 they were able to garner the support of the federal government (Godwin, 2018). An important milestone for this was the Tamil occupation of the Gardiner Expressway in downtown Toronto (Jeyapal, 2013). In early 2009, during the most violent fighting of the entire civil war, a huge protest was staged outside of the Ontario provincial government building, later spilling on to one of the major traffic arteries of the city. 12 Whilst some criticized the event for portraying the Tamil diaspora as a nuisance and assign it no political importance, it is used by many Tamils to signal a shift in diasporic consciousness. For example, the 2009 events awakened many Tamil youth to the plight of their ethnic kin in the homeland, which many of them had never stepped foot on, thus leading to a renewed interest in seeking relations with their 'homeland' . 13 On the other hand, it certainly affected perceptions of the Tamil diaspora held by fellow Canadians, who first considered the disruptions a nuisance (Krishnamurti 2013), but gradually became more receptive toward Tamil ethnic grievances (Bradimore & Bauder, 2012;Jeyapal, 2014). 14 After 2009, Tamil diaspora attention, also that of younger generations, broadly began to focus on the international level, specifically the UN Human Rights Council (UNHRC), where claims to transitional justice and accountability were being negotiated. In many ways, the post-war period saw the relationship between the Tamil diaspora and the Sri Lankan state strained even more.
During this time the opportunities for formal Tamil diaspora involvement in development governance were minimal, and there were no Sri Lankan state-led efforts to harness Tamil diaspora contributions to development governance.
Similarly, in the early years after the end of the war the topic of 'development' was barely raised amongst Tamils in Toronto, at least not publicly or in a formalised manner.

2015: The stars align for Tamil diaspora inclusion
The year 2015 presented another important milestone for Toronto-based Tamil diaspora organisations working in the development field. A number of globally interconnected events occurred that set the scene for the inclusion of Tamil diaspora in relevant development governance networks and spaces.
On the one hand, in Sri Lanka the election of Maithripala Sirisena to the office of president, in a surprise defeat of incumbent Mahinda Rajapaksa, stirred hope amongst factions of the Tamil diaspora that progress towards justice was on the horizon. Whilst Rajapaksa's reign was defined largely by his Singhalese chauvinism, corruption, and violent decimation of the Tamil population during the final stage of the civil war, 15 Sirisena ran on a platform of 'good governance' and promises to ensure accountability for Tamil victims. The election outcome also briefly made it safer for Tamil diaspora members to physically travel to Sri Lanka. Ultimately, a window of opportunity for diaspora Tamils to engage in development governance in their homeland opened up. Meanwhile, Canada also held elections in 2015. Replacing conservative Prime Minister Stephen Harper, the election of Justin Trudeau, leader of the liberal party, signalled a move towards economically liberal, socially progressive, and a more internationally oriented Canadian politics. Further, both national elections coincided with the official launch of the UN SDGs. A central innovation from the Millennium Development Goals (MDGs) was their insistence that development was not just the responsibility of the global South; rather it concerned everybody. Developed countries in the global North were now equally responsible for the SDGs' success, for example in the fight against climate change, and the eradication of world poverty. This argumentation rests on a global systems approach which sees the world as integrated; any efforts to affect change must thus be holistic and sensitive to this global interconnectedness (Zhang, Prouty, Zimmerman, & Mihelcic, 2016). Articulated in this manner, the SDGs quickly found resonance amongst Canadian development professionals, and the newly minted ministry of 'Global Affairs Canada' wholeheartedly embraced the new international development agenda. Whilst remaining a national body, the renaming of the ministry indicated what Tsing (2011, 57) has referred to as 'scale-making projects' , implying that the mere invocation of a global scale can indicate a profound shift in meaning. However, the 'global'positioning of the ministry responsible for development governance did not translate to a larger budget to back up this position federally. This became evident in 2017, during the release of the new budget for the Canadian Feminist International Assistance Policy. Indeed, lack of additional funding provided for the implementation of the highly ambitious SDGs placed additional pressure on civil society organisations working in the development sector. 16 Meanwhile, immediately before the federal elections in 2015, the then-DFATD, had tasked the OCIC with holding a consultation on the engagement of 'cultural communities for sustainable development' . 17 Even though the term 'diaspora' was avoided in the title, in my discussions with participants and organisers it became clear that the consultations were largely borne from the 2012 report on the role of non-state actors in international development by the Standing Committee of FAAE. 18 Whilst the OCIC was put in charge of organising the event, they turned to the Mosaic Institute for support, explicitly for their expertise on all things diaspora, implicitly for their extensive network of organisations (specifically in the field of diaspora and peace-building, see above). The consultation gathered a number of academic experts on diaspora and civil society organisations, like WUSC and Cuso International, both with a long history of running diaspora volunteering programmes in the global South. As suggested in the vignette, this event would turn out to be highly significant for the emergence of a Toronto-centered group of organisations and individuals focused on engaging diaspora in development. Indeed, it was here that the idea for the Toronto-based DENG was conceived. Importantly, the executive director of the Mosaic Institute had extended invitations to a select number of recently formed Tamil diaspora-run development and peacebuilding organisations. This was a significant step towards the inclusion of Tamil diaspora in development governance. For the invited Tamil organisations, the 2015 consultations provided a significant opportunity to broaden their organisational networks and consolidate their relationship with the OCIC, and the broader Canadian (and global) development community.
What becomes evident in these preceding paragraphs is that the emergence of strategies to engage diaspora in development governance was not linear, and far from resembled a national roll-out or international downward diffusion of policy. Rather, it was characterised by contingency, and the formation of unlikely alliances, sometimes out of chance, often out of convenience. The story told here encompasses entanglements between organisations, spaces, ideas, and agents that a state-centric account of diaspora engagement in development governance would entirely miss.
But acknowledging the existence of complexity, contingency, and chance is only the first step towards a better understanding of the politics of diaspora inclusion in development. I will now disentangle the conditions that make possible this seemingly chaotic engagement. Specifically, I will show how diaspora inclusion in development governance, has been conditioned by social connectivity, legitimacy claims, and spatial embeddedness.

Disentangling conditions for diaspora inclusion in development governance
'Keeping the connection alive': Fostering social connectivity and networks of communication The Diaspora Engagement Networking Group, described in the introductory vignette, presents an example of a diaspora engagement strategy that seeks to foster diaspora inclusion in development governance. As mentioned, the DENG was founded on the back of another diaspora engagement strategy, an event titled 'Cultural Communities for Sustainable Development Consultations' , which took place in the spring of 2015. In the lead up to its International Assistance Review, DFATD had made the top-down decision that the OCIC should organize these consultations as a matter of urgency. Indeed, I learned from interviews with the organisers that the consultations were planned very last minute, leaving little room for a well-thought out strategy on whom to invite or engage in the consultations. Consequently, existing networks were leveraged. The OCIC and Cuso International already shared an office in the Centre for Social Innovation, which meant that Cuso was a shoe-in for the consultation event. In fact, much of OCIC's focus on diaspora had emerged from close cooperation with Cuso International and its diaspora partner organisations. But the ad hoc creation of guest lists was not unique to this particular event. Other evidence found in my data suggests that decisions to include or engage diaspora in development governance or peacebuilding followed a similar pattern.
Below is a response from the Mosaic Institute's executive director to the question of how the decision was made on whom to invite to an engagement event held in 2012, titled Diasporas@Toronto: Mosaic has an amazing network. (. . . ) Mosaic is about people. We were the only game in town that did not have a connection to a specific diaspora community, which meant that we were in a position to partner with a lot of ethno-specific organisations, as well as with non-ethno-specific orgs, academic, business. We knew a lot of people. And I think we generated the list through our own networks, and we also had a significant in-kind contribution from my former employer, which was the Ontario Ministry of Economic Development and Trade. They loaned us the Ontario Investment and Trade Center -of which I had been the inaugural director -which is a beautiful international trade presentation centre on Yonge Street (. . . ) so I think it might have been a bit quid pro quo in that they provided a lot of names to investors as well. Which was great because they wanted to be seen as actively engaged in diaspora communities and conversations about business networks. So that worked for them. The peacebuilding part worked for us, the development piece worked for us, and for Cuso International, and for IdEA. And IdEA was able to say "look, we have this partner event happening in Toronto that extends the region, the footprint of IdEA, but is organised through a particular Canadian lens" and . . . everyone was happy. 19 This excerpt confirms that the decision-making process around whom to include in the event is based, in large part, on existing social networks and communication channels. My data suggest that because of the ad hoc nature of many such events, a crucial dimension of diaspora inclusion in government-led consultations or civil society capacity building events is whether or not an organisation or individual has a strong network. Accordingly, a large proportion of practices within the development field are directed towards the creation and maintenance of social relations. The founding of the DENG can then be explained as a response to this demand/need for social connections and does indeed foster the inclusion of diaspora in decision-making around development governance. This also has an impact on the importance ascribed by diaspora members to their own networking practices. I asked one of my Tamil  of what I was trying to do with (the organisation), and he was very happy to see that kind of thing. And so, he asked the OCIC to invite me to be a participant in the round table that they had in 2015. And since then we kind of kept that connection alive. So, I sometimes exchange emails with the Executive Director at the OCIC and we just kind of stayed close to OCIC. And then they had this Global Citizenship week and we participated and I think our poster won second place in the competition. So, all of this was great. But I think I want to get more involved. So, this is why we decided a couple a months ago to actually apply for membership at the OCIC. And get access to the resources, but also get access to the policy formulating tables that OCIC has in Ottawa.
In this excerpt, he mentions the importance of repetition in maintaining social connections ('We kind of kept that connection alive' , 'I sometimes exchange emails . . . and so we kind of stayed close'), but he also emphasises the importance of more formal relations to create and maintain access to resources and 'the policy making tables' where decision around international development governance were being made.
But my data suggests that even beyond the development field diaspora actors need to spend a large proportion of their time and energy building social capital. This is illustrated in excerpts from an interview with the national spokesperson of the Canadian Tamil Congress (CTC). It was his response to my question on how the CTC began its operations in Canada in the mid to late 2010s.
(. . . ) we at CTC, we believe in engagement. And engagement will always work. Initially, sometimes people don't engage with you, but they don't know you. When you go through people -when you get closer to them -it's all the same. We are the same, you and me. We may have different colours, maybe we speak different languages. But, fundamentally, people are the same. Humans are the same (. . . ) the only thing is: you have to get to know them. We believe in that very firmly. Unless you get to know somebody, you cannot pass on remarks to them. So, we went and we sat with those people and started explaining to them. And they got to know us.
(. . . ) Locally we met with MPs and nationally we met with leaders of the parties. We went to their conventions, we met with their delegates and met with ministers, you name it. And when we go to parliament, we make it a point of meeting all different parties and leaders of the parties, at least the person who has some say in the party.
Here he describes some of the struggles that the organisation experienced after its inception, and the strategies that were used in trying to establish itself in the Canadian public sphere as 'The Voice of Tamil Canadians' . What emerges from the excerpt is that my interlocutor is convinced that CTC's success is ultimately linked to 'engagement' , and that 'engagement will always work' . But what exactly does 'engagement' mean in this context? He seems to employ the term to describe a range of practices, from reaching out to journalists for media coverage, meeting with politicians to communicate concerns and issues affecting the Tamil community, to attending events and seeking out opportunities for face-to-face interactions with public representatives and decision-makers. He also emphasises the importance of persistence and repetition in these practices ('But I didn't go down' , 'It was a learning curve for us'). Ultimately, both examples illustrate the importance placed by diaspora organisations on deepening and expanding the reach of their (professional) networks. Based on what we know about how organisations like OCIC and Mosaic make decisions about who to include in development events and consultations, the prioritisation by diasporic actors of social connections and networks seems warranted.
Evidently, we need to take networks seriously in the study of diaspora inclusion in development governance. That networks are of central importance for diaspora is a popular policy perspective (Kuznetsov, 2006). It suggests that the creation of networks and virtual communicational channels is the key to eradicating barriers to inclusion. Governance actors, like the OCIC or the Canadian government, have identified access to networks as a barrier to effective diaspora contributions to development. Thus, opportunities for networking are increasingly provided in the form of consultations, events, networking groups, lobbying. The proliferation of shared workspaces, networking groups and events (such as the DENG) and virtual communication tools that foster networking are considered the key to achieving inclusion in development governance. But also in academia has the power of diaspora in IR often been associated with their unique position within transnational social (Wayland, 2004) and financial networks (Elo & Minto-Coy, 2018).
Diaspora have been conceptualized as brokers or norm entrepreneurs with an 'in-between advantage' (Brinkerhoff, 2016) allowing them to act as norm entrepreneurs. But we cannot reduce diaspora power to social capital, or network relations. Connectivity to networks and channels of communication on its own does not guarantee diaspora's power and inclusion in global politics. More needs to be known about the context of these social connections, their cultural as well as material embeddedness.

Becoming legitimate partners in development governance
The above paragraphs suggest that a condition for diaspora inclusion in government-led consultations or civil society capacity building events in the development field is a diaspora's position in a social network. But whilst it might seem that the most well-networked diaspora group will automatically be the most powerful or most included, these network connections do not exist in a vacuum. In fact, I will argue that the majority of Tamil-run diaspora groups are excluded from development governance networks based on a lack of legitimacy. How this lack manifests will be explored below.
When asked about Tamil mobilization for development, one of my interlocutors laments the mobilization practices of his fellow Tamil-Canadians: one time I remember going to a meeting, where (. . . ) a lot of money was raised and it was given to an organisation in the North (of Sri Lanka) where they brought a hundred sewing machines and gave those to women. And I pretty much yelled at that meeting, saying: "Why couldn't we invest that money in sending these women to like computer classes or programming classes? Why are we reinforcing gender-stereotypes?' And the thing is, the women who have sewing machines already are producing, but they don't know how to market that and sell it. So, it is a different skill that they actually (need)not the sewing machine. They actually need somebody to tell them how to do business planning and marketing. So, that is the kind of nuanced approach to development that is really missing".
He continues by saying that: in the development space what you find increasingly now, is (Tamil) organisations that have historically done political work, getting into that space as well. So, development is in many ways being coopted for a political agenda, I would argue. So, these organisations then bring those substantial organising and resource mobilization capabilities to development work. Except they are not doing development as we understand it, they are doing development as they understand. Not based on any Sustainable Development principles. Not based on understanding the existing social inequities in SL. They are just doing it like "let's hold a party, raise money and give back".
He suggests that Tamil diaspora mobilization for development has historically taken the form 'giving money and not asking questions' . In a long interview, he very eloquently points out a range of problems that plague the 'remittance economy' , as he calls it, which has been set up between the Tamil diaspora and its homeland. In the interview, he also suggests that diaspora mobilizing for development needs to be committed to the UN's Agenda for Sustainable Development, specifically their commitment to gender equality. This is a discourse that resonates with the conversations I observe during DENG meetings. Here members often made a point of highlighting their commitment to the UN's Agenda 2030, sometimes criticizing their fellow diaspora members for lack of accountability and ineffectiveness.
Their concerns echo the above interview, and what has recently become the orthodox understanding in the development field: that diaspora remittances need to be managed in order to be effective. In mobilizing for development, the expressed willingness to move 'beyond remittances' (Newland & Patrick, 2004) towards capacity building and knowledge and skills transfer is considered central for diaspora legitimacy.
Beyond discourses that directly pertain to the 'how' of diaspora mobilization, at the time of my research the field of development governance is also structured around broader norms surrounding the concept of good governance and sustainability. In order for diasporic actors to be included in development governance or decision-making, they must signal their commitment to these ideas. This is what makes them legitimate 'partners' . Importantly, as Horst and Sinatti (2015) have suggested, in this way diaspora engagement strategies also function as a way of propagating dominant Western development norms. Further, the possession of local knowledge (of culture and language), and a sense of purpose and belonging are considered desirable traits in diasporic development actors. Horst and Sinatti (2015) have insightfully explored how diaspora are considered 'good' development actors as they purport to have a link to the homeland and local knowledge and can therefore 'represent' local populations in places where development programmes are being implemented. Meanwhile, diaspora organisations, in order to be considered legitimate, must also have technical (objective) knowledge on development practice so that they can work effectively. These discourses are often at odds with each other, making legitimacy relations hard to navigate. The 'networking group' is then evidently not just a place for making social connections. It is also a space that is hierarchically structured by dominant ideas and norms about what correct development is and how diaspora should engage in it.
Another argument that I advance in this section is that diaspora engagement in development is not only embedded in 'global' cultural fields like that of international development. Here the concept of the assemblage becomes useful, as it allows us to think about how agents are able to simultaneously embed in multiple fields. Situations where agents seem to move between different overlapping fields, suggests a flattening of scales. Naturally, a study of inclusion of diaspora in (global) development governance must be preoccupied with exploring the internal working of the development field, in order to understand diaspora access to and agency within this field. However, the reality for most diasporic actors is that they enter these 'global' fields only very infrequently. In a way, the 'global' cultural field only exists in the encounters and events that include global governance representatives, e.g. from the UN or World Bank. Tamil diaspora actors are more often concerned with their relations to their new country of residence, as illustrated in the section on networks. Diaspora organisations want to appear legitimate to the broader Canadian population and state actors. Hence, they must position themselves in a national cultural field. We know from the literature on integration that migrants are constantly negotiating their new home environment. Here, the idea of the 'model migrant' becomes central, especially for diaspora residing in liberal Western countries like Canada. Success is measured as educational attainment against structural odds, integration into mainstream culture, or the upholding of values of respectability, as explored in the literature on migrant cultural capital (Erel, 2010). Indeed, I found that the position of a diaspora group or individual within the Canadian national cultural field is shaped by their performance of Canadian citizenship, i.e. upholding Canadian values of multiculturalism, feminism and diversity, practicing aboriginal solidarity, performing 'hyphenated Canadian' identity.
But for the Tamil diaspora in Toronto, performing as a model migrant also means positioning themselves in relation to a number of fairly contentious political debates and practices. Here the issue of representation is of particular importance. Positioning in the Tamil nationalist struggle largely determines whether Tamils view an organisation as representative or not. This is actually where a lot of the development-oriented organisations seem to falter. A diaspora organisation may be so plugged in to the SDG debate and Canadian national politics that they are no longer considered legitimate 'representatives' of the Tamil diaspora or the local Tamils in Sri Lanka. Their legitimacy within the development field is traded in for legitimacy in the eyes of their Tamil-Canadian constituents, and vice versa.
Finally, the nascent Tamil diaspora organisation I encountered at the DENG also wanted to appear legitimate in the eyes of young (second or third) generation Tamils. These make up a large faction of the Toronto-based diaspora and form the core pool of applicants for their development volunteering programmes. Here, affirmations of technical capacity are less important. Rather, in order to appear legitimate and representative to these young peoplemany of whom are highly educated and hard-working young professionals -the organisation has adopted discourses and practices that display what Savage et al. (2018) describe as emerging cultural capital. Emerging (urban) cultural fields are structured around educational attainment, emphasis on youth engagement, usage of youth jargon and 'millenialisms' , highlighting innovation and 'newness' , referencing popular culture, emphasizing reflexivity and spirituality (Savage et al., 2018). In Toronto this entails, for example, quoting Drake lyrics in social media communications. These practices become particularly important when diaspora organisations seek to enlist youth volunteers for their development programmes or as they compete for municipal funds to implement projects that foster Toronto's involvement in strengthening the SDGs.
Evidently, to be considered legitimate diaspora must often position themselves within multiple overlapping cultural fields simultaneously. Those who fail to do this, struggle to gain legitimacy in the eyes of powerful development governance actors. In fact, some of the longest existing Tamil diaspora organisations, with the most extensive networks and links to all major political party leaderships, as well as the most widespread membership amongst the Toronto Tamil diaspora community, struggle to have their voice heard in global development governance fora. This is because they may have political power in the more traditional sense, as they influence federal MPs through electoral politics. However, they fare less strongly in terms of both development-related and emerging cultural capital. Finally, for diaspora individuals and organisations to exist in multiple fields at the same time, networks are important. Social connections allow diaspora to travel between different fields, proving again the importance of entanglements between different power relations.
In sum, we have now discussed two conditions that shape the ways in which diaspora are included in development governance/decision-making: The social relations between actors and the relations of these actors within overlapping fields. These two are themselves interconnected in the sense that no single actor would find themselves occupying cultural fields at different scales (global, national and local) if it was not for their networking capacity. What is missing is an understanding of a third power dimension that shapes diaspora inclusion in development governance; that of spatiality. 20

Reterritorializing diaspora inclusion in development governance
Ultimately, even if a Tamil diaspora organisation positions itself as a legitimate development actor and enters into social connections with powerful development actors in charge of diaspora engagement strategies, barriers to inclusion remain. This is because connectivity and legitimacy are also always spatially anchored. Consider the Robertson Building at 215 Spadina Avenue in downtown Toronto. On its website we can read that: 'the building is home to a cluster of community businesses, social entrepreneurs, and non-profit organizations, as well as a bio-wall (a completely living-breathing wall made of plants), and extensive green roof, greenhouse and cedar viewing deck' . 21 Approximately 30 percent of my fieldwork in Toronto took place in or in close proximity to this building located in the middle of Toronto's Chinatown. In fact, the site of the first DENG meeting described in the opening vignette, forms part of the same estate and sits right across the street. The restored warehouse with a history of manufacturing of plumbing fixtures and fittings, today contains the offices of the OCIC, Cuso International and the CSI, and thus forms a crucial component of the context in which diaspora engagement strategies are embedded.
The Robertson Building has agency as a symbolic and a social space. Is it a 'hub' where people can mingle and where organisations share office space and create and build networks. As mentioned above, OCIC and Cuso International have built much of their collaboration around sharing a large open plan office. The building also holds symbolic power.
Henri Lefebvre (1991) has famously written about how spaces can be appropriated, reappropriated and disappropriated, shaping their role in urban politics. We can think about the Robertson Building as having been reappropriated from its original function as an industrial factory, buzzing with migrant labourers, to a hub for social enterprises employing urban middle-class workers competing for capital in the knowledge economy. Importantly, and in contradiction to its original purpose, it now symbolically excludes members of the migrant working class in Toronto, who often lack the symbolic or cultural capital to legitimize their existence in these spaces. These two spatial-dimensions (symbolic and social) are captured by the socio-spatial positionality concept explored by Koinova (2019). They both affect who is included in development governance.
But the Robertson Building also has power as a material space. Who gets to enter the building is not just about who has the cultural capital to do so. The building occupies a particular territory in downtown Toronto and is embedded in the city's public infrastructure. Adler-Nissen (2012) has suggested that we must thus study the interplay between symbolic and material resources in International Politics. So how do material resources affect a diaspora inclusion in development governance? Whenever I attended an event that was hosted by Tamil diaspora, I had to plan for about 1.5 h on public transit, driving through what seemed like endless parking lots (or what Keil (2020) refers to as 'postfordist ruins'). Meanwhile, the sites of diaspora engagement for development, like the Robertson Building and the CSI are located in downtown Toronto, and not in Scarborough or North York -where the largest proportion of Tamils reside and most diaspora organisations have their offices. The representative from the Tamil organisation that was part of the DENG mentioned to me on occasion that they would like to be more closely linked to the CSI through obtaining actual office space there. Diaspora inclusion in development governance is evidently dependent on geography and economic capital needed to gain geographical proximity to sites of power. Material resources determine where diaspora organisations can afford to rent office space. In combination with geographical distance, it also determines whether they have time and money to travel to regularly attend networking events. During my time in Toronto, each time a DENG meeting was scheduled, there were some members who could not attend based on their geographical position within Toronto. Members sought to overcome this spatial barrier through measures like teleconferencing and occasionally meeting in other locations across the GTA. But this was considered second best. Members of Cuso International and OCIC could always meet up in person more spontaneously than others, consolidating already existing hierarchies within the network.
Evidently then, diaspora engagement in development governance does not float freely. Who is included is as much a question of who is connected, and who is seen as legitimate, as it is who has physical access to the sites of engagement. In sum, if a Tamil diaspora member says and does all the right things and shows up on the radar of the development governance network, barriers to inclusion -and especially sustained engagement -remain. This is because social relations, and legitimacy claims are also always spatially anchored. What the above paragraphs show is the close entanglement of conditions for inclusion of diaspora in development governance.

CONCLUSION AND AVENUES FOR FUTURE RESEARCH
The preceding paragraphs have offered original ethnographic data on diaspora engagement strategies from which were disentangled a range of power relations, namely social connectivity, legitimacy claims, and spatial embeddedness that determine the conditions for inclusion of diaspora in development governance. The paper began by arguing that the founding of the DENG networking group makes fairly obvious the importance of 'networks' to the inclusion of diaspora in development governance. Meanwhile, evidence revealed that the 'networking group' is not just a place for making social connections. It is also a place that is hierarchically structured by dominant ideas and norms about what correct development is and how diaspora should engage in it. It also offers insights into dominant ideas about Canadian identity. Empirically, the paper showed that within the international development field, there has emerged a dominant discourse of how diaspora should engage in development. Diaspora inclusion in development governance is conditioned by a diaspora actors' ability to establish themselves as a legitimate partner for development governance.
Finally, the data also revealed the importance of space (symbolic, social and material) in structuring access to and inclusion in development governance. In conclusion, whilst social connectivity is important, networks of diaspora engagement do not float freely. They are embedded in global and local cultural fields, which in turn are embedded in a material environment. A relational ontology has also allowed me to investigate and account for contingency and scale jumping.
The spaces in which development governance actors come together to 'engage diaspora' may allow for chance interactions between agents that would usually occupy different fields at overlapping scales . In this way, they can transport ideas from the local to the global scale and back again. Thus, a fairly small and unknown local diaspora organisation with good connections might gain access to people in positions of global power, thus bolstering their legitimacy. What this paper was not able to address in detail is the role that professionalization and bureaucratic power play in the politics of diaspora engagement. After all, an important means by which diaspora organisations gain legitimacy and are thus more likely to gain access to spaces of governance/decision-making, is through proving themselves to be effective professional actors in the development field. For diaspora organisations, this means spending a large proportion of their resources jumping through bureaucratic hoops. But legitimacy gained through professionalization can then sometimes also stand in the way of effectiveness. Evidence from my interviews with members of the DENG suggests that diaspora organisations are often preoccupied with reaching charity status, becoming formal members of the OCIC umbrella group, obtaining operational capacity. An interesting avenue for future research would be to ask: How does bureaucracy shape diaspora inclusion in development governance? This echoes work done by Pouliot and Thérien (2018) on accreditation practices, and Eagleton-Pierce (2018, 2020) on professionalization of NGOs at the WTO. Future research on the embeddedness of diaspora engagement strategies will also need to address in more detail questions such as: How do actors navigate the constraints that they face? How might they change their position? What forms of power allow them to do so? This requires delving deeper into the interplay of agents and the different types of structuring relations within specific contexts beyond the origin-state, but also beyond the field of development. This paper then offers a jumping off point to the study of phenomena -like diaspora engagement strategies in development governance -that do not fit into traditionally modernist and state-centric analytical frameworks, and require sensitivity towards complexity and contextual embeddedness.