Using ethnography and assemblage theory in political geography

Correspondence Pooya Ghoddousi, University College London, London, UK. Email: p.ghoddousi@ucl.ac.uk Abstract While the focus on the ‘everyday’ in qualitative human geography has greatly increased the need for, and relevance of, ethnographic methods, Megoran argued that this is particularly true for political geography as it has the potential to challenge its focus on elite discourse, allowing researchers to bring forward multiple voices to investigate the becoming of political events. More recently, assemblage theory has gained traction in political geography, not only because of its capability to include the role of the material and the affective, but also revealing the links between microand macro-politics by showing how agency emerges out of complex relations. In the first part of this paper, we present an overview of the recent uses of ethnography in political geography that have not embraced assemblage. Second, we explore the theoretical conceptualisations of, and opportunities provided by, an assemblage approach. Third, we go through the use of assemblage ethnographies in political geography, with a particular focus on Pooya's experience of research with Iranians in London. In this, he embraced a variety of ethnographic approaches, including ‘auto-ethnography’, ‘netnographies’, ‘participant sensation’, in combination with observations, participatory workshops and activism. Showing the role of ethnography as a qualitative tool for political geographers to interrogate discursive social constructions, we argue that it holds even more Received: 19 August 2019 Revised: 22 April 2020 Accepted: 30 April 2020

ticularly true for political geography as it has the potential to challenge its focus on elite discourse, allowing researchers to bring forward multiple voices to investigate the becoming of political events. More recently, assemblage theory has gained traction in political geography, not only because of its capability to include the role of the material and the affective, but also revealing the links between micro-and macro-politics by showing how agency emerges out of complex relations. In the first part of this paper, we present an overview of the recent uses of ethnography in political geography that have not embraced assemblage.
Second, we explore the theoretical conceptualisations of, and opportunities provided by, an assemblage approach.
Third, we go through the use of assemblage ethnographies in political geography, with a particular focus on Pooya's experience of research with Iranians in London. In this, he embraced a variety of ethnographic approaches, including 'auto-ethnography', 'netnographies', 'participant sensation', in combination with observations, participatory workshops and activism. Showing the role of ethnography as a qualitative tool for political geographers to interrogate discursive social constructions, we argue that it holds even more promise for analysing and intervening in the emergent politics of socio-material-affective assemblages.

K E Y W O R D S
affect, assemblage, ethnography, materials, political geography 1 | INTRODUCTION: BLURRING THE BOUNDARIES An ethnography is a form of embedded qualitative research in which the scholar spends time in the field in order to collect in-depth experiential data. Imported from Anthropology, over the last thirty years, the ethnographic method has become a significant tool for qualitative political geographers (Crang & Cook, 1995). It is more an epistemological orientation towards the field than a specific method per se. This orientation calls for the utilisation of a range of methodological tools, including (but not limited to): participant observation; interviews; field notes and diaries; surveys; focus group discussions; material collection (any mediafor instance, photographs, videos, sounds, objects, publications); auto-ethnographies; netnographies (i.e. social networks in physical or virtual spaces); mapping techniques (topographic and topologic); and Participatory Action Research.
Although practiced in the 19th century by geographers such as anarchist-geographer brothers Élie and Élisee Reclus (Ferretti, 2017;Kropotkin, 1899), ethnographic methods were not a major consideration for human geographers for most of the 20th century. Moreover, problematically, the classical anthropological conception of ethnography usually involved choosing a field in which the anthropologist immersed themselves for an extended period of time in order to gaze at, collect data about and gain insights into the lives and cultures of 'others'usually colonial subjects. The traditional bounded conception of a field site (e.g. a village)that considers it coextensive with 'a people' and 'a whole culture' (Gupta & Ferguson, 1997) was designed to emphasise its cultural difference in order to challenge presumptions of biological, psychological or social universality (Burrell, 2009: 182). At the same time, geographers were preoccupied with other methodological approaches, such as the 'quantitative revolution' in the 1960s (see Johnston et al., 2018), or the use of 'qualitative interviews, film and media analysis, as well as historical and visual accounts' by qualitative geographers since the 1970s (Nayak & Jeffrey, 2011: 117).
When, in the 1990s cultural geographers began taking to the ethnographic field (e.g. Crang & Cook, 1995), the anthropological field had already begun to change. This was due to a 'crisis of representation' that led some postmodern anthropologists (Abu-Lughod, 1991;Gupta & Ferguson, 1997) to suggest avoiding reifying representations of processes (e.g. cultural) by employing a specific method of 'writing against culture', where the accounts would be stripped of theoretical representations and general conclusions and focus solely on the particular stories told by respondents (Abu-Lughod, 1991: 262). Other non-traditional approaches to the field opened up the bounded/territorial spatiality of traditional ethnography towards more networked, relational, mobile and multiple configurations like 'multi-sited ethnography' (Marcus, 1998) 'networked field site' (Burrell, 2009) and 'mobile methods' (Elliot & Urry, 2010;Urry, 2007).
Political geographers did not immediately follow the call to ethnography, instead maintaining a focus on 'elite discourse' (Megoran, 2006, 623). , Flint (2003 made the case that researchers who 'do not identify themselves with political geography' (618) were doing some of the sub-discipline's best work. Woon (2013: 32) has attributed this to its 'apparent over-reliance on secondary accounts'. Megoran (2006) argued that the sub-discipline neglected several crucial turns happening elsewhere in human geographynot least the attention to the 'everyday', which had become a significant concern for much of human geography.
In the early 2000s, political geography began to change. In a forum discussion in Political Geography, Marston (2003: 625) asked What would it mean if more political geographers paid attention to agency, the everyday and the micro level of social life? It would mean that we would have to train our students more broadly, for one thing. It would mean that our students (and we ourselves) would need to be as adept at ethnography as they are at geopolitical analysis. Surely this couldn't be a bad thing.
Heeding this call, it was in the early 2000s that calls for ethnographic fieldwork began to proliferate in political geography (Hyndman, 2004). Such an approach, Megoran (2006: 624) argued, 'could enrich and vivify the growing, and somewhat repetitious body of scholarship on both critical geopolitics and international boundaries'. At this time, many human geographers, pioneered by feminist geopolitical scholars (Dowler & Sharp, 2001;Sharp, 2005), were emphasising the importance of the everyday power relations. This emphasis resonated with Foucauldian concepts such as 'governmentality' (for instance, McConnell, 2012;Ferguson & Gupta, 2002;Desforges, Jones, & Woods, 2005) and biopolitics (for example, Anderson, 2012;Dalby, 2013;Davis & Hayes-Conroy, 2018). Furthermore, as we shall discuss, the concepts of assemblage and affect have recently become crucial moments in the embrace of ethnography by political geographers.
We define an assemblage here as the coming-together of human and non-human 'things' that take on an emergent agency (Page, 2020). By agency, we mean the ability to do something, to affect and be affected. An assemblage approach identifies the locus of agency as emerging from changing patterns of relations among partial objects, rather than pre-existing bounded objects or entities. The multiple interpretations and styles of assemblage thinking in social scienceincluding assemblage theory (Buchanan, 2017;DeLanda, 2006), actor-network theory (Latour, 2005), and non-representational theory (Thrift, 2004)have given rise to a body of literature that is now 'a repository of methods and ontological stances towards the social' (Acuto & Curtis, 2014: 3). What groups these approaches under the umbrella of assemblage theory is a shared worldview based on relationality, complexity and openness that allows for identifying transformative potentials of human/non-human entanglements in their diverse becomings.
In the first section, this article presents an overview of selected political geographic ethnographies that have not used assemblage theory. This section will highlight the rise of feminism in human geography that has changed the meaning of 'politics' by showing that 'the personal is political'. In the second section, we discuss the theoretical underpinnings and potential benefits of an assemblage approach, both to analyse and bring about new political becomings. Third, we discuss the benefits of the marriage between assemblages and ethnographies, going on to present examples from Pooya's own research, alongside other assemblage ethnographies. In concluding, we argue that assemblage ethnographies produce potentials beyond what either approach offers. By focusing on human/nonhuman relations and affects, assemblage ethnographies open the ethnographic field, create new political subjectivities, and blur the boundaries between researcher and researched, allowing both to participate in the emergent agencies of assemblages. This can lead to a mode of Participatory Action Research in a post-structural vein (Cameron & Gibson, 2005) that aspires to transform the collective subjectivities that it studies both through participation in the material assemblages of 'fieldwork' and through feeding back the generative concepts (Robinson, 2015) produced in the research to form new material-discursive assemblages.

| ETHNOGRAPHIES
The marrying of assemblage to ethnography in the field can be understood as part of a widening of the remit of political geography since the calls by feminist geopolitical scholars in the early 2000s to embrace ethnography. This has led not just to methodological but also theoretical changes. Woon (2013: 32) argues that this was a call for a new way of 'doing' geopolitics, that ethnographies had the power to bring to light '(grounded, first-hand) storylines that are arguably obfuscated by secondary narratives, thereby attesting to the multiple geographies and politics of knowledge production'. These insights can illuminate state-personnel experiences (for instance, Weisser, 2014;Bachmann, 2016;Gruby & Campbell, 2013;Bornschlegl, 2018), economic processes (Carlisle, 2016;Connolly, 2017;Cook et al., 2004;Cook et al., 2017;Cook & Harrison, 2007;Oliveira, 2018;Watkins, 2018), or the politics of water management (Goh, 2019). However, these insights have frequently been conjoined to new conceptual approaches that have challenged what 'political' means in political geography, changing the underlying concept of the subject.
Sometimes the 'political' can be obscured as it is not always focusing, say, on the discourse of state-political actors, or the quantitative spread of elections. Instead, researchers turned to, for instance, the worker, the migrant, or the activist, blurring the line between political and social geography. Recently, this is evidenced by the political geography found in the study of consumerism in a mall, or everyday workplace experiences (Billo, 2015;Billo & Mountz, 2016;Ruwanpura, 2016), entrepreneurialism in North Korea (Wainwright, Kibler, Heikkilä, & Down, 2018), or quasianarchist practices of seed-trading in the United Kingdom to circumvent the gaze of state (Pottinger, 2018). Whether in single or multi-sited studies, these researchers (and many others mentioned in our paper) present a political geography that sparked an opening to a different spatial politics, away from the territorialised and centralised assertions of state, instead looking at the interaction of people, and the politics of relations between bodies.
Ethnography has helped bring forward an emphasis on the role of emotions in politics. This helped destabilise the normative assumptions of 'formal/informal politics, public/private spaces and the characterisation of these spaces as rational/emotional and masculine/feminine' (Schurr, 2013: 115). Bringing in considerations of emotions as part of the focus of an ethnography changes not only what politics is (the examples above), but also the understanding of how it functions. In her study of electoral campaigns in Bolivia, Schurr (2013) observed that there was also a lack of 'attention to the emotional dimension of institutionalised politics' (115). To readjust this, she presents an account of populist politics that speaks not of the rationality of state politics, nor of the irrationality of it, but of the emotion of it, of how people are swept into voting one way or another. Such a concern with emotion in electoral politics resonates beyond Latin America, into many spaces in America, Europe and beyond (cf. Page & Dittmer, 2016;Steinberg et al., 2018). This can be seen as part of what Dittmer (2017: 6) calls the 'New Statecraft', a focus 'principally concerned with the everyday crafting of the state'.
Thus, doing ethnography has also brought to light the importance of understanding the 'everyday' in the experience of the political. It is hard to understate how influential in political geography this focus has been, as it 'valourizes the mundane encounters with statist discourse that constitute our everyday lives' (Dittmer, 2017: 6). For instance, in his ethnography of the everyday of a soup kitchen in Brixton in London, Johnson-Schlee (2019: 174) found that conspiracy theories serve 'to tie the everyday to the geopolitical'. These conspiracy theories are part of an everyday survival mechanism against punitive 'British state attempts to compensate for the global economic crisis' that reveals how 'the logics of social policy terrorises the smallest daily geographies of precarious lives'. This is as true for homelessness in Athens (Bourlessas, 2018), as it is for those stateless in the borders of India and Bangladesh (Shewly, 2013(Shewly, , 2017, for migrant experiences (Ryburn, 2016) or contentious performance of borders by various political actors (Murton, 2019;Wilson & McConnell, 2015). Such work helps to undermine normative understandings of the state as something 'out there', present, and topographical, but insteadas Jeffrey (2013) asserts in his ethnography of Bosnian state-makingthat the state is 'improvised'.
This focus on the everyday has also been brought into play with some philosophical excursions (that assemblage theorists later build on). Particularly, the Foucauldian concept of 'governmentality'or the 'conduct of conduct' (Swyngedouw, 2011: 372) has shed light on the self-validating practices of some states (Lentz, 2014;McConnell, 2013). These scholars have opened politics to entwine the formal to the informal, the rational to the emotional, and the state to the everyday, without using assemblages. In other words, even before marrying ethnography to assemblages, many innovations assemblagists claim as their own had already been made by political geography ethnographers (cf. McConnell & Dittmer, 2018). And yet, as shown over the next two sections, assemblage has something to add to all these approaches, as well as challenging the understanding of politics as necessarily always antagonistic.

| ASSEMBLAGES
Thus far, we have shown the instrumental role of ethnography in reinvigorating political geography in the last two decades. It has not only refreshed the understanding of official state actors and institutions, but also brought into its purview political subjectivities hitherto ignoredthe practices of workers, migrants, and other 'ordinary' people. We also showed how some ethnographers have pushed to open the bounded spatiality of classic field sites. However, assemblage theory challenges the very concept of subjectivity: rather than an already-existing subject of study (e.g. individual humans or political entities), assemblage theory foregrounds open multiplicities of relations that constitute the subject(s). Deleuze and Guattari see these relations as open and exterior, forged by affects, always becoming, and transformed by events that are brimming with virtual possibilities; nonetheless, we also acknowledge the contribution of Actor-Network Theory (Latour, 2005;Law, 2009) that provides a methodological toolbox for the empirical analysis of actual networks of relations among humans and non-humans in case-studies that are more 'stabilising' than in flux (Müller & Schurr, 2016; see next section for more). But what do assemblages offer political geography that was not found in the above ethnographic approaches?
An assemblage approach goes beyond merely challenging the bounded representations of space and subjectivity or 'places and people' and 'culture'. The post-humanist and relational politics of assemblages disrupt the dichotomies of inside and outside, self and other, or subject and object. This is why they go beyond dialectical politics of 'us' versus 'them'that is, the aforementioned workers or migrants versus capitalists or the statethereby promising an affirmative politics of difference that has more progressive, affirmative, nuanced and inclusive potentials (Farías, 2011). On a political level, assemblages have at least two potentials: (a) producing a 'generative critique' that is 'constantly creating new associations, knowledges and alternatives' (McFarlane, 2011: 212) or emancipatory research (Rogers, 2012: 8); (b) creating new (non-state) forms of collective organisation and becoming called 'commons', 'publics', 'groupuscules' or 'left assemblages' by various authors (Amin & Thrift, 2002;Deleuze & Parnet, 1987;Farías, 2011;Hardt & Negri, 2009;McFarlane, 2011;Tampio, 2009).
Part of the political promise of assemblage theory lies in its recognition of the role of affects in forging assemblages. This allows activist scholars to analyse these flows and mobilise them to convert the emergent political agency of the assemblage into a more progressive one among their virtual possibilities. Fox and Alldred (2015: 402, emphasis in original) have argued that assemblage theory's uniqueness is that its 'concern is no longer with what bodies or things or social interactions are, but with the capacities for action, interaction, feeling and desire produced in bodies or groups of bodies by affective flows'. The capacities of these assemblages can be explained through an analysis of the patterns of their 'constitutive relations' (Deleuze, 1988: 32). The contribution of assemblage analysis is not the descriptive explanations of phenomena that offer generalised claims to truth (i.e. reified representations), but concepts that explain processes and patterns emerging in an ever-changing world (Buchanan, 2015;Deleuze & Guattari, 1994;Robinson, 2015). The conceptual toolbox thus created allows us to mobilise the affects, and thereby the constitutive relations, in an affirmative politics that holds new promises. These approaches can provide ways to understand, participate and change the world in a more-than-dialectical wayexpanding our conception of politics from antagonistic to agonistic (Amin, 2002;Mouffe, 2000), from negating to affirmative politics. Rather than defining politics based on the clash between starkly opposing sidesthat is only resolved through fighting what is wrong with the worldaffirmative politics seeks to build and strengthen what is right, alternatives that exist like 'seeds beneath the snow' (Ward, 2008: 23). To do this, affirmative politics depends on building convivial ties through exchange of positive affects (e.g. care, humour) and creating new solidarities and collective subjectivities (Cameron & Gibson, 2005), rather than negative and dividing affects (e.g. anger, hatred) that so much politics depends on.
The conceptions of assemblages we have referred to are necessarily geographic, exploring things through a topological lens, where '[t]he space between "here" and "there" is measured less by miles and kilometres (i.e. the topographic lens) and more by social relationships, exchanges and interactions involved' (Allen, 2016: 3; also see Cockayne, Ruez, & Secor, 2019). This is an intensive conception of space where the intensity of experience of these interactions is created by the flows of affect traversing assemblages. Affects are the glue that hold assemblages together, as well as the vital forces that drive the processes of becoming. In an ethnography of the Model UN, Dittmer (2013: 499) gives examples of humour as an affect that creates a 'community' through the 'energy' that binds the performer with the spectator, leading to 'the emergence of new collective subjects through the enmeshing of biology and culture'. Affects can also bind humans with non-humans to produce bodies politic with emergent agencies (Protevi, 2009). It is therefore essential to study the political role of affects in order to create new solidarities and convert the collective agencies of assemblages to actualise their more progressive potentials.
There are differences between this 'new' Deleuzian concept of affect and 'older' understandings of emotions as pioneered by feminist scholars: affects are considered to be pre-cognitive and therefore not dependent on individual human subjects or their interactions as emotions are. This conception gives affects an autonomy with new liberating potentials not afforded by human emotions that are bounded within signifying structures of language and culture (Massumi, 1995: 87). Like ANT formulations, Massumi's conception acknowledges the role of non-humans in emergent political agencies, but this focus goes beyond actual networks of things and taps into these affective intensities to create and convert the virtual potentials of assemblages (Müller & Schurr, 2016). However, this must not lead us to positing a dichotomy of emotions and affects that ignores the liberating potentials of emotions, because these non-personal affects often blur into emotions once they are sensed by human subjects (Boler & Davis, 2018;Thien, 2005) (see example of this blurring in Pooya's research below).
Admittedly, the progressive potentials of assemblage geographies mentioned above do not always materialise.
Particularly when detached from ethnography, assemblage geographies sometimes remain purely speculative, descriptive, abstract or conceptual, failing to deliver any political potentials or act as an ethos for engaged research Kinkaid, 2019;Tonkiss, 2011;. Halvorsen (2017: 446) has also challenged assemblage as 'downplaying the antagonistic nature of spatial mobilisations', arguing that it is 'ill equipped for explaining contradiction and abstracting to broader historical and geographical moments' in his study of the Occupy London social movement. However, we argue that an ethnographic approach in combination with assemblages increases the potentials for a politically engaged research that is situated in empirical realities, while opening up possibilities for generative abstraction/conceptualisation, as well as mobilising progressive political potentials.

| ASSEMBLAGE ETHNOGRAPHIES
We argue that the combination of assemblage and ethnography provides certain opportunities that were not present in either assemblage theory or ethnography. These opportunities can be summarised as: an empirically grounded focus on the everyday and mundane practices (i.e. not present in non-ethnographic assemblage research), as well as a move away from closed totalities towards open-ended and always-becoming multiplicities (i.e. present only in some recent non-assemblage ethnographic field research). Assemblage ethnography avoids reifying this complexity into static representations. It instead analyses the assemblages through generative concepts that explain actual and virtual patterns emerging in constellations of relations, thereby offering glimpses into their agencies (Bos, 2018), political potentials (Sharp, 2018) and processes of becoming (Blanco, Arce, & Fisher, 2015). Assemblagists recognise the agency of non-humans, and use affects to analyse and intervene in the emergent agencies of their human/nonhuman constellations (i.e. not present in non-assemblage ethnographic works merely focusing on humans and their emotions).
One of the key potentials of an assemblage approachespecially in their ethnographic applicationis to participate in creating ties in the real world between the researcher and participants through exchanging affects that create trust and blur the boundaries between researcher and researched. Through her non-assemblage ethnography of emotions of violence in the Philippines, Woon (2013) observed how the research can be reconstituted through emotional engagement whilst doing the research. However, assemblages go further to highlight the researcher's role in the political agencies that emerge from the human/non-human research assemblages, as the researcher is imbricated in them, peripherally or centrally, where both the researcher and the researched take part in a double process of becoming (Deleuze & Guattari, 1987). Utilising some standard ethnographic methods, assemblage ethnographies also require an experimental engagement and a temperament of openness from researchers to follow the relations that matter. As such, the ties of solidarity that situate the researcher as part of the assemblage allow the research to move from participant observation to participant sensation (cf. Miltz & Schurr, 2016)a method of sensing the waves of affect permeating the assemblage when they are sensed on the ethnographer's body (see below for examples). More fundamentally, these solidarities can provide the necessary grounds for genuine Participatory Action Research (Kindon, Pain, & Kesby, 2007) which can directly engage in converting the collective agency of an assemblage. Another way to convert the agency of assemblages is through the research outputthe aforementioned concepts that intervene in these human-non-human assemblages. These generative concepts could be created by the individual researcher or, more effectively, through collective agency of participatory research (Corson, Campbell, Wilshusen, & Gray, 2019).
In order to show some of the benefits of assemblage ethnographies in practice, we now turn to some examples from Pooya's research, focusing on the specific methods and novel approaches mentioned above. Pooya conducted ethnographies through participating in a number of assemblages of social gathering, collective action, and political campaigning among (mostly) Iranian migrants in London, UK and beyond. He used methods that are frequently utilised by non-assemblage ethnographers: participant observations, interviews, focus groups, participatory workshops, and discourse analysis. However, he also experimented with methods like auto-ethnography -'a form of selfnarrative that places the self within a social context' (Reed-Danahay, 1997: 9)to gauge the affective intensities of the assemblages he immersed himself in. This method combined with netnography (Kozinets, 2010;Matoes & Durand, 2012;cf. Conradson & Latham, 2005a, 2005b)the ethnographic tracing of social networks in physical and digital spacesallowed him to go beyond participant observation and engage in what can be called participant sensation. This was a move away from 'fly-on-the-wall' ethnography, not in the form of 'going native' (Fuller, 1999), but akin to 'deep hanging out' (Wogan, 2004) or 'insider ethnography' (Hayano, 1979), which allowed Pooya to sense the flows of affects traversing the collective body of the assemblage, while his own body was plugged into it (DeLanda, 2006: 11). For example, when Pooya took part (as a co-organiser) in a demonstration in support of the Iran nuclear deal in Trafalgar Square in London, he experienced the intensive states of elation (i.e. goosebumps) that democratised the leadership of the demonstration and fused the bodies into a collective body at specific moments during the event. This effect was strongest when this collective body spontaneously broke into singing, painting together on a large canvas, moving the canvas around the square, and finally walking to dine together in a nearby restaurant. The double becoming that blurred the boundary of researcher and researched not only allowed participant sensation as a method, it also facilitated the participatory action research element of this research. As a result, the researcher and respondents became participants at various stages of the research: from taking part in participatory workshops, co-organising conferences and colloquia, political activism, to engaging in iterative rounds of analysis and conceptualisation (e.g. in workshops, follow-up discussions or feedback of participants on draft ethnographies). In the later stages of this research Pooya and his participants mobilised these concepts in the campaigns that emerged from these assemblages. The point of this research is that they found tools (i.e. the concepts) that could be applied in other settings and open up new possibilities, rather than just explaining what a specific setting is (i.e. inductive generalisation).
Influenced by the more-than-human and material turns in human geography, assemblage ethnographies recognise the role of non-humans in the production of agency (Whatmore, 2002). In his ANT-inspired ethnographic study of Baku-Tbilisi-Ceyhan oil pipeline, Barry (2013: 1-2) asserts that the force of this argument means that 'no longer can we think of material artefacts and physical systems … as passive and stable foundations on which politics of states takes place … the unpredictable and lively behaviour of such objects and environments should be understood as integral to the conduct of politics'. While the built environment has become an important factor in assemblage ethnographies (Abourahme, 2015;Dittmer & Waterton, 2019;Miller, 2014), other ethnographers using assemblageinspired methods have focused on the agency of animals (Barau, 2014;Greenhough & Roe, 2019), mail (Davies, 2012) or campaign leaflets and posters (Page, 2019a;Page & Dittmer, 2015). For Pooya, these materials included food, passports, mobile phones, laptops, furniture, and so on. This list does not include everything found lying around, but those that bound the assemblages together and played a role in their emergent agencies. Showing how the boundaries between humans and non-humans in an assemblage blur through affective flows, this research contributes to the argument that it is not solely human cognition and intentionalityas captured by discursive methodsthat is worth studying, but also the embodied and material aspectsas captured by non-discursive methods sensitive to practices and affects (Müller & Schurr, 2016). Embracing non-discursive methods does not mean rejecting discursive methods. Instead, it advocates a more-than-discursive approach, taking into account, as Hitchings (2012) argues, that 'people can talk about their practices', particularly when these practices take the form of everyday routines. This means Pooya followed-up his participant observations with discussions where research participants reflected on their observed practices.
All the assemblages that Pooya studied involved translocal networks that he traced through netnographic approaches (e.g. use of social media) and mobile methodsthereby blurring the boundaries of the physical field sites (Page, 2019b;cf. Burrell, 2009). There was also a combination of co-present and non-co-present affects that created the intensities and consistencies of assemblages. A stark example was the intense flow of non-co-present positive affects brought about by sharing music in an email group of 40 friends scattered around the world. This led to the creation of different space-times, transporting people out of their immediate physical co-present space-times. Many recipients expressed a sense of elation felt through connecting to their loved ones that transported them away from their daily problems or sense of alienation. The elation sensed by one participant was expressed by this email to the group: 'I can't believe how synchronised we are from afar! It made me cry twice, now I'm taking my earphones to school to cry twice more during the day!'. Another recipient added that 'The more distant they keep us, the closer we get', thereby implying a political act of resistance against dominant powers emerging out of such practices of everyday life (de Certeau, 1984). This example shows how the exchange of affects blurred the boundaries of individual subjects. The collective subjectivity thereby forged exhibited glimpses of an emergent political agency in the virtual form that became visible at later stages of research when it fed into new actual assemblages of political demonstrations.
Inspired by Deleuze and Guattari (1987), Pooya and his participants conceptualised the virtual and actual political agencies of these assemblages through a continuum shifting from territorial, hierarchical and centralised modes of engagement (i.e. state apparatus) to more horizontal, democratic and participatory modes (i.e. nomadic) at will, or make their membership criteria more inclusive or exclusive depending on expedient situations. These generative concepts showed the dual tendencies of these diasporic assemblages to territorialise or deterritorialise their identities, solidarities and therefore their potentials for progressive or regressive politics. These concepts therefore allowed for engaging with these political potentials and converting their agencies in response to the situations at hand. For example, using hierarchical and territorial modes of engagement at initial stages of solidarity building, while challenging the crystallisation of these hierarchies through mobilising the nomadic and deterritorialising tendencies of the assemblages. The concepts are therefore not just scientific abstractions, but tools for converting the political agencies of assemblages.

| CONCLUSION
In this paper, we have reviewed some of the recent boom in the use of ethnographic methods in political geography.
These novel methodological and philosophical approaches to political geographical research have proven a rich source for producing new understandings of political subjectivities by emphasising the everyday and the emotional.
We then examined the contributions of assemblage theory, without ethnography, to the sub-discipline through the introduction of non-human agencies and affects. Our argument is centred around the opportunities provided by combining ethnographies with assemblage theory. While ethnographies bring a focus on the 'everyday' into assemblages, many anthropologists still struggled with opening up the bounded field site to avoid reified representations.
An assemblage approach provides the philosophical and methodological tools for blurring the boundaries of the ethnographic field: it becomes less defined by a demarcated field site and pre-defined political subjects, and more by the relations and intensities found in an open-ended field. An assemblage approach also blurs the boundary between the researcher and the researchedas the research is co-produced through said relationsopening potentials for participant sensation, empathy and solidarity thereby enabling a Participatory Action Research. Finally, it can help the research move away from producing generalisations of phenomena and focus on generating concepts that explain patterns of relations and processes of becoming which can be used by participants to convert the agency of assemblages. This is done through mobilising positive affects that forge solidarities to actualise more progressive virtual potentials of assemblages. What those potentials are is not always straightforward or clear, and we risk claiming that using this methodological toolbox or joining of activism-to-academia can bring forward a new worldthis is somewhat overblown. Nevertheless, we have shown how assemblage research and activism can help produce temporary alternatives that suggest another way is possible, however ephemeral they are.

AUTHOR BIOGRAPHIES
Pooya Ghoddousi teaches human geography at University College London (UCL). His PhD research at UCL focused on (mostly) Iranian diasporic networks of collective action in London (and beyond). This research used the Deleuzo-Guattarian concepts of assemblage, affect and nomadism in combination with ethnographic methods to explore the progressive political potentials emerging from the everyday interactions among his fellow Iranians. He is a member of the Urban Lab (UCL) and has published on nomad politics in urban settings.
Sam Page is a political geographer. His PhD research at University College London on the UK Labour Party's 2015 General Election campaign and the Deleuzo-Guattarian concepts of assemblage and affect has contributed a novel approach to political parties and electoral geography. He has also written on the topics of Donald Trump's presidency of the United States, and on the recent resurgence of breweries in London. He is currently researching the English electoral commission.