Ethnography in the plural: Experimenting with collective ethnography in research on politics and politicians in Brazil

Abstract

Developments in qualitative research have increasingly recognized the value of modes of ethnographic inquiry that go beyond the conventional model of the lone ethnographer. Many examples involve parallel, collaborative enquiry in which ethnographers come together around frameworks, share data and collaborate in analysis – rather than working as a collective throughout the research process, including fieldwork, what we refer to here as ‘collective ethnography’. Less has been written on doing collective ethnography as a practice. Through a series of vignettes, this article tells the story of a research experiment in collective ethnography, focusing on politicians and their social worlds in São Paulo, Brazil. We suggest that an engaged process of collective enquiry offers sources of insight that usually lie beyond the ethnographic gaze: including into our own conduct and the choices we make in our ethnographic encounters and interpretive avenues.

‘What would ethnographic fieldwork look like if it was shaped around the epistemic practice of experimentation?’ ask the #Colleex collective. ¹ This captures the possibilities of a more innovative, exploratory approach to ethnographic research. As #Colleex suggest, ‘to invoke ethnographic experimentation is not necessarily to signal a methodological rupture with conventional forms of ethnography’, to ‘inject an experimental sensitivity in fieldwork’ and, with it, to ‘renovate the discipline’ of anthropology. ‘Experiments’, they note, ‘are singular events producing the unexpected: throwing us questions we didn’t have, changing our notions of values and creating the conditions of their own appreciation’. ² This article is an account of just such an experiment, bringing together two Brazilian political scientists and two British social anthropologists in an intensive encounter with collective ethnography, in the context of the methodological explorations of a multi-country research project, Ethnographies of Parliaments, Politicians and People. This experience threw us questions we didn’t have and brought us insights that we might not otherwise had.

Brazilian anthropology has a more pluralistic, innovative and collaborative character than the versions of the discipline found in Britain, France, the United States and elsewhere in the global north, as Mariza Peirano’s (1992) Uma Antropologia no Plural – echoed in our title – makes clear. In contrast to the largely individualistic approach or more rarely team-based comparative approach to ethnography found in the global north, there is a rich history of collective and collaborative research in Brazilian anthropology dating back to the 1970s, in which the Museu Nacional in Rio de Janeiro played a catalytic role. Some examples include the Projeto Emprego e Mudança Social no Nordeste (1975–1977), coordinated by Moacir Palmeira with a team from the Museu Nacional (Palmeira, 1979); Afrânio Garcia and colleagues’ work on the Hábitos Alimentares em Camadas de Baixa Renda (food habits in low-income groups) project in a collaboration between the Museu Nacional and the University of Brasilia; another collaborative project in the 1990s between the Museu Nacional and the French École Normale Superieure (ENS, Paris) on land occupations and agrarian reform (Sigaud and L’Estoile, 2006), and a large collective project from 1997 to 2005 that instigated the Núcleo de Antropologia da Política, a collaboration between the Museu Nacional, the Federal University of Rio de Janeiro, the University of Brasilia and the Federal University of Ceará (UFC), which went on in the 2000s to work on agrarian reform and agribusiness, and in the mid 2010s on social movements and the public sphere (Leite Lopes and Heredia, 2014). ³

From the early work of Moacir Palmeira and his analyses of electoral dynamics and political mediation, to more recent studies of participatory governance, activism, and policy implementation, Brazilian anthropology has also long engaged with the study of politics. The field has contributed a nuanced understanding of how political relationships are formed and sustained. Scholars such as Lygia Sigaud, João Pacheco de Oliveira, and Licia Valladares have brought attention to the everyday spaces in which politics unfolds—favelas, rural settlements, administrative offices—emphasising the relational, embodied, and affective dimensions of political life. This work often attends to how political legitimacy is performed and contested, and how categories such as ‘citizenship’ and ‘representation’ are produced in specific encounters. Studies by Abers and Von Bülow (2011) and Leite Lopes and Heredia (2014) have shown how policy is co-produced through negotiations between bureaucrats, politicians, social movement actors, and citizens.

This body of work reflects a broader commitment in Brazilian anthropology to investigating how power is exercised, negotiated, and resisted across multiple scales. It also aligns with collective research practices where teams of researchers investigate different sites or aspects of a shared political process. By situating our work within this tradition, we hope to contribute to ongoing discussions about the methodological and epistemological possibilities of collective ethnographic approaches to studying politics. In bringing our discussion of collective ethnography into dialogue with longstanding Brazilian traditions of political ethnographic research, we aim to expand Anglophone methodological conversations on collective ethnography beyond the global north and highlight the contributions of multi-voiced, team-based inquiry.

Within scholarship on politicians and in parliamentary studies, ethnographic research remains relatively rare, in contrast to scholarship on other political domains, such as civil society and social movements (Cornwall, 2008; Crewe, 2021). Two exceptional examples of teams of political science making serious use of ethnographic methods in parliamentary research are Rai (Rai and Johnson, 2014) and Geddes. ⁴ But collective ethnography is unheard of within scholarship on politicians. By foregrounding the methodological textures and tensions of working as a group, and by reflecting on the uneven roles and shifting dynamics within our team, we seek to contribute to broader debates about the uses of ethnography in research on politics.

Through a series of vignettes drawn from intensive methodological experimentation in multiple sites in the Brazilian metropolis of São Paulo, we tell the story of our experiment in such a way as to enable the reader to see elements, routes and connections that take dimensions of the collective ethnographic experience as a starting point for reflexive analysis. We suggest that collective ethnography offers a way not only of diversifying who produces ethnographic knowledge, but also of recognising how that knowledge is relationally and processually made.

Collaborative, ‘many-handed’ and collective ethnography

The craft of ethnography is generally practiced by lone researchers, but the process invariably involves collaborations and moments of collective engagement and enquiry (Wasser and Bresler, 1996). At the very least, the researcher is often influenced by encounters with others with whom they have worked, studied, discussed, reasoned, argued, quarrelled, celebrated or shared a table. And intensive engagement in meaning-making with interlocutors lies at the very heart of ethnography in practice. Since the late 1990s, there has been growing experimentation with collaborative and collective ethnography, especially in the field of education and health research. Out of this has come a literature that explores questions of negotiation, meaning-making and the relational dynamics of the collective process (see Austin, 2003; Bresler et al., 1996; Buford May and Pattilo-McCoy, 2000; Clerke and Hopwood, 2014; Creese et al., 2008; Erickson and Stull, 1998; Gerstl-Pepin and Gunzenhauser, 2002; Gordon et al., 2006; Jirón and Imilan, 2016; Martínez et al., 2021; Mauthner and Doucet, 2008; Pardee et al., 2018).

Yet, as Jarzabkowski et al. (2015) note, although there is growing interest in collective and collaborative ethnography there has been less written on how it is actually done. Piecing strands from the literature together, these experiments appear to have taken different forms. Many have involved ethnographers working in different field sites and collaborating by sharing insights, communiques and, sometimes, their own fieldnotes. Some have been team-based projects led by a senior colleague, sometimes as a PI or PhD supervisor, where individual projects have been brought together through the interactions the PI’s grant has made possible. Others have involved the co-presence of ethnographers in a single fieldsite, each pursuing individual enquiries and writing individual fieldnotes to later compare what each of them saw. Less commonly, it seems, have ethnographers carried out participant observation together simultaneously in the same fieldsite, working in plural rather than in the customary ethnographic singular; in a rare example, ethnomusicologists Justice and Hadley (2015) describe what they call ‘collaborative fieldwork… scholars going into the field together and working through the ethnographic process as a team’ (2015: 64), going on to comment on how unusual such experiences remain.

The terminology that has developed to describe these experiences lacks sharp distinctions between practices. One strand, ‘collaborative ethnography’ has its own literature. ‘Collaboration’ in ‘collaborative ethnography’ is generally between an ethnographer and their interlocutors (Arribas Lozano, 2022; Lassiter, 2005); the process of enquiry bears strong resemblance to that of participatory research, shifting power relations in the design, implementation and ownership of the research process (Cornwall and Jewkes 1995). Indeed, Lassiter (2005) positions collaborative ethnography as part of a cluster of related approaches that include feminist methodologies and participatory action research. Elaine Lawless (1993) coins the term ‘reciprocal ethnography’ to describe a similar process of involving participants in shaping the research process, and in analysis and theorisation. Joanne Rappaport (2008) narrates a particularly powerful example of co-theorisation and collaboration with interlocutors in the production of ethnographic knowledge, describing it as ‘collaborative ethnography’.

The term ‘collective ethnography’ is used less commonly in the methodological literature. As we use it here, it speaks to something different: collaboration among ethnographers in collective ethnographic fieldwork. The instances we came across of collaboration among ethnographers generally did not often involve co-presence in the field. Instead, they often referred to the simultaneous process of individual fieldwork in different contexts, where the collective aspect involves sharing an initial collective goal and then exchanging ideas, findings, insights, feedback on drafts and analyses as the work develops. This is a relatively familiar model for large-scale international research projects, where a common framework guides individual studies that are then brought into dialogue and emerging insights synthesised; two of us have worked in this way in numerous projects over the years, including in the project within which our experiment took place. Harkness (2017) and Martínez et al. (2021), for example, write of how as five anthropologists from different parts of the world, they came together to frame a shared research initiative that took shape through communications with each other by email, calling it ‘many-handed ethnography’.

In a similar vein, Pardee et al. (2018) describe how 12 scholars working on independent projects studying a common phenomenon came together around shared goals, met once or twice a year for 6 years, reviewed each other’s work and had twice-yearly teleconferences, in

… an integrated, reflexive process of research design and implementation in which a diverse group of scholars studying a common phenomenon yet working on independent projects engage in repeated theoretical and methodological discussions to improve (1) transparency and accountability in the research and (2) the rigor and efficacy of each member’s unique project. (2018: 672)

Pardee et al. situate their experience in established feminist principles, including a commitment to reflexivity, especially in relation to power and positionality, as well as a willingness to engage with a plurality of voices and perspectives. These are principles that have also guided this and our other collective research projects, and are core to feminist participatory research (Maguire, 1987). Underemphasised in this literature are accounts of the relational dynamics of collective engagement in ethnographic research amongst teams of people who may have very different investments in the outcome, statuses, perspectives and experiences. This includes tensions that may need to be navigated in terms of collective ownership of material, resolving differences of opinion in analysis or reluctance to challenge each other’s assumptions as well as concerns about sharing data, especially in a raw form (Gordon et al., 2006).

In Brazilian anthropological research there are precedents for collective ethnographic practice dating back to the late 1970s, as noted earlier. These involve coordinated fieldwork with subsequent joint analysis (Palmeira, 1979; Sigaud and De L’Estoile 2006). Unlike collaborative ethnography, which explicitly redistributes epistemic authority to interlocutors (Lassiter, 2005; Rappaport, 2008), these projects maintained an academic-driven research structure while facilitating collective modes of interpretation. These studies have advanced the discipline, while strengthening rigour and relevance of the individual pieces of research as a collective endeavour. Furthermore, Brazilian scholars such as Leite Lopes and Heredia (2014) have examined how collective methodologies enhance the ethnographic study of political institutions and actors, highlighting the analytical advantages of this approach.

Building on this earlier work, our argument in this article is that collective ethnographic enquiry that involves the co-presence of ethnographers and a deliberative process that runs through each stage of ethnographic research, animating participant observation, reflexive analysis and meaning-making, can enrich the way ethnographic knowledge is produced. Having several eyes, ears and pens engaged in capturing what is going on can improve the quality of the data and inspire new forms of imagination for meaning-making. Depending on how the team is composed, the co-presence of people who are differently positioned offers the possibility of asking different kinds of questions, in different ways (cf. Justice and Hadley, 2015).

Collective ethnography also creates the opportunity for a dynamic form of iterative peer review in which emerging insights can be aired, challenged, picked apart, added to, and revised. Plural engagements in the field of action create multi-faceted relational dynamics and these bring a deliberative quality to ethnographic research. The mere fact that we can see other ways of doing and thinking at close hand in the actions and reflections of fellow collective ethnographers can create space and permission to pursue avenues of thought about which we might otherwise be insecure or hesitant. As such, collective ethnography becomes in effect a complementary means of producing knowledge to that of lone or comparative ethnography: one in which the rigour and trustworthiness (Lincoln and Guba, 1985) of ethnographic research develops through the iterative process of enquiring together.

Collective ethnography: An experiment

Our experiment in collective ethnography began when the four of us convened in São Paulo to spend a week together thinking about methodologies for researching politics and politicians, in the context of the multi-sited international research project Ethnographies of Parliaments, Politicians and People. It was envisaged as a kind of hands-on methods workshop, an opportunity for focused reflection on methodology at a crucial point in the life of the project. The idea was to spend some of our time together actually doing political ethnography and using that as an entry point for conversations about epistemology and methodology. What we hadn’t anticipated at that stage was quite where the experiential approach we took to having these conversations would lead us.

While none of us would contemplate regarding a mere week of doing collective ethnography as a means to produce viable or reliable ethnographic findings (and it is important to emphasise this here, as it is not the intention of this article to suggest that this could replace the longer, harder, slog of fieldwork focused on producing new knowledge rather than experimenting with method), we were surprised at just how much we learnt, how much working together brought in terms of new perspectives and insights, and how enjoyable the process was. As it turned out, what started with a modest ambition of a few shared ethnographic encounters and a lot of conversation became an intensive, immersive fieldwork experience, taking us into legislatures, the offices of elected representatives, municipal government buildings, political rallies, street-based electoral canvassing and place-based constituencies. The insights we gained felt rich enough to work up into an article to share with other ethnographers.

For a group of researchers interested in politics and politicians, we could not have chosen a more opportune time. It was election time, what Heredia and Palmeira (2012) term ‘the time of politics’. ⁵ Campaigns for the presidential election and to elect deputies to the federal and state chambers were in full swing. Our interlocutors were deeply immersed in the social life and affect of electoral politics; the intensity of our method matched that of a closely fought election, one with stakes that could not have been higher. Our plan was to spend as much time as we could over the course of the week observing and participating in political spaces, carrying out informal interviews, analysing our experiences and using them to explore the broader methodological and epistemological questions that were the reason for us working together in this setting.

We had little sense at the outset of what collective ethnography might look or feel like, nor what we might get out of it; it was an open-ended experiment. We’d all worked collaboratively. For the anthropologists amongst us, these experiences had usually been in the mode of the solitary anthropologist working as an ethnographer alongside people from other disciplines who were using their own methods and approaches. While both the anthropologists had done collaborative ethnographic research, including joint fieldwork, and one of us had experience training and facilitating co-enquiry by interdisciplinary teams using Participatory and Rapid Rural Appraisal, being part of a collective ethnographic endeavour involving collective fieldnotes and co-presence in ethnographic encounters with more than one other researcher took these experiences to another level.

We were at a mix of career stages: Telma had just finished a PhD in Political Science rooted in an ethnographic approach to researching the municipal chamber and political territories of São Paulo. Her knowledge was deep and local; our encounters and interviews built on her pre-existing relationships and rapport with interlocutors. Andrea was a newcomer to the project, but not to Brazil, having spent more than 20 years engaging as a political anthropologist with the unfolding and unravelling of the democratic promise of Brazil’s 1988 “Citizens’ Constitution”. Along with being the project’s Principal Investigator (PI), Emma brought deep international expertise on the study of parliaments and people. And Cristiane, coordinator of the Brazilian team, knew federal-level politics from the inside, and came equipped with experience and understanding of the dynamics of Brazilian politics at the national level.

The relations of power that might otherwise have operated in such a situation, given the hierarchical nature of academia, were turned on their head by the fact that the most junior member of the team was the one with the greatest proximity to and knowledge of the field, having recently emerged from an intensive study of the places and people who were to become our field sites and interlocutors, and the PI was completely new to Brazil and did not speak Portuguese. This inversion of traditional academic hierarchies has been observed in Brazilian collective projects, such as the studies coordinated by Afrânio Garcia (2003) on agrarian reform and agribusiness in Brazil. Garcia noted that younger scholars and those with local expertise often became methodological anchors in ways that disrupted the presumed authority of senior researchers. Similarly, in our project, while academic rank remained a structural factor, proximity to the field often outweighed institutional seniority in shaping ethnographic insight and methodological direction.

Equally, being ranged on a continuum from a first-time foreign visitor to a foreigner familiar with Brazil, to a Brazilian visiting the city from the capital, to a local with deep contextual knowledge served not only to invert academic hierarchy. It also offered us interpretive as well as ethnographic opportunities. As we note earlier, foreigners may ask naïve or apparently stupid questions. This can be extremely useful. Foreigners can also ask about things that a local might be curious to know that would be really odd for them to ask. Deep contextual knowledge can guide inquiry into places a foreigner might take months to even realise exist. For anthropologists trained in social or cultural anthropology in the global north, ethnographic research often begins with ignorance and with rudimentary language skills that are honed over time as the ethnographer becomes more familiar with what is initially strange and difficult to grasp. This is considered core to the process of anthropological learning. Collaborative ethnography involving people who are familiar with and newcomers to a research context offers both proximity and distance at the same time, and – most importantly – the opportunity for dialogue and reflection that engages these differences of perspective.

With the exception of the first vignette, our interactions with interlocutors were in Portuguese; our fieldnotes and conversations were in English; we moved between the two languages in translation, which brought interpretive opportunities none of us had anticipated. Telma, Andrea and Cristiane took turns to translate for Emma; we found ourselves pausing to pick up on a particularly resonant term or one especially difficult to translate directly. This pushed us to reflect on fields of meaning, including on the interpretation of terms we might otherwise have simply assumed shared meaning across our languages. So it was with the terms mandato and gabinete, each of which had their own distinctive fields of meaning in our languages that required not only an account that contrasted our respective political systems but also took us into reflection on the principles that underpinned them and what the practices associated with them in this context told us about how democracy was constructed and enacted. So too with concepts of representation and political office, which seemed at first to be simple enough to map across contexts but ended up taking us into complex and interesting discussions on political philosophy and the politics of political practice.

It was also interesting to pause to reflect on everyday turns of speech in Portuguese that captured something about political relationships. Conveying these concepts, terms and idioms to Emma and explaining the contexts of their use had the effect of raising these dimensions to consciousness, making them the subject of reflection in a way that working entirely in Portuguese or English would not have permitted. This brought those of us who spoke Portuguese and were culturally familiar with Brazil and Brazilian politics unexpected insights, provoked by Emma’s questions about things we might otherwise have simply taken for granted, as well as by our attempts to explain.

We’d all struggled with the norms and frames of political science and the methodological limitations of the positivist approach to the study of political representation. One of the challenges we’d also all faced in different ways was having our methodological process as ethnographers questioned and at times undermined, and our ethnographic methods derided as “anecdotal”, “impressionistic”, “subjective” and (even) “soft and fluffy”. This tension between ethnographic method and the expectations of political science has been a recurring theme in Brazilian research on the anthropology of politics. Leite Lopes and Heredia (2014) argue that ethnographic methods offer unique access to the micro-politics of governance, brokerage, and representation - an approach that remains underutilized in mainstream political science.

Our process sought to make as much of what we did a collective endeavour as possible. Emma suggested we write collective fieldnotes in a shared Google Doc. This served as a collective resource that we could all access and add to simultaneously and sequentially. Each of us adopted a colour for our entries. Over the days that followed, we worked in the early morning and evening to write up our fieldnotes from our interviews and encounters. Prose was interspersed with comments, questions, thoughts, echoes and reminders. We signed up at the outset to a collective code that would govern our work together, also at Emma’s suggestion. We’d said we would aim for pithy insights, rather than lengthy monologues. That we’d distil in the fieldnotes the main lines of insight and inquiry, the main points from our interviews and the main headlines from our discussions. And that what emerged would be our collective property, something we’d made together and that none of us could take off and do something of their own with unless they’d got the permission of the others. Each day we talked through our ethnographic encounters, using the convivial spaces of shared meals, bus rides and walks to and from bus stops as opportunities for collective reflection. In one of these moments, Emma reflected that adding three more people to a piece of research doesn’t just make it a sum of four: there is the interaction between each one in the group and between each one and interlocutors in the field, making it a much more complex equation.

We found ourselves constantly ruminating on what we were learning and on our positionalities: our process of interrogation of our observations and insights was seamless, intensive, iterative and interactive. And in the early mornings and evenings, we continued this process in our collective fieldnotes. Together, we wrote more than 12,000 words, a rich record of commentaries, notes and reflections on the sequence of events that took us all over the city into encounters with a diversity of people – bureaucrats in the city’s public administration, would-be and experienced elected representatives, their brokers, agents, teams and voters. Some entries paint pictures, with one person leading and others sketching in further details, and at times correcting, revising or contesting the initial interpretation. Others were more about us as ethnographers, what we had come to know, what we had come to recognise as the right or wrong way to do ethnography, and what we were learning from working together like this. And others still are more methodological, from notes on process to sharing what we learnt from experiences of doing ethnographic research in other contexts.

This iterative model of fieldwork reflection resonates with the approach described by Andrea Barbosa (2019) in her ethnographic research on political activism in Brazil, where small research teams used nightly discussions and group analysis sessions to reassess their methodological choices in real time. The Brazilian tradition of reflexive ethnography, particularly in studies of urban politics (Leite Lopes and Heredia, 2014), similarly emphasises the importance of these shared analytical moments, not just as a means of data validation but as generative of new theoretical insights. Additionally, Barbosa (2019) highlights how collective ethnography enables researchers to adapt their inquiry dynamically, an approach we also found valuable as we adjusted our questioning strategies and site selections based on ongoing collective discussions. Our collective ethnographic approach builds on this tradition, demonstrating that embedding multiple researchers in fieldwork focused on politicians and politics allows for an enriched, layered understanding of institutional and informal political practices.

In what follows, we take a series of what Erickson (1990) calls ‘analytic vignettes’ and use them to draw out some of what our collective ethnography taught us. Although the vignettes are independent of each other, there is a thread that connects them that gets thicker as it is woven. The sequence of encounters we describe here happened chronologically; with each, we progressively added more layers of interpretation as we got to know each other better and deepened our understanding through collective discussion through which we iteratively interpreted each encounter. As the vignettes unfold, they add further layers to the story about the process of our experiment in collective ethnography. In the first, we explore the aspect of the speed of connection an insider can provide vis-à-vis the possibility of posing questions that an insider simply could not or would not think to ask. The second vignette takes this as a point of departure and delves into discussions between us about languages and meanings that arose because through the collective process, we began to interrogate our frames of reference and taken-for-granted understandings of key terms. And the third and last vignette discusses how much contextual knowledge can be called for to understand politics in a setting. It speaks again to the speed of connection and the benefit of having more eyes, ears and imaginations.

The bureaucrat

We found him alone in an eyrie far above the dense urban landscape, landmark buildings in the foreground flanked by a distant mountain range of white tower blocks. Signs of life in the streets below were hard to spot from 18th floors up. Plate glass on all four sides of the building ensured the dominion could be viewed from all angles. Rows of desks sat unattended; desktop computers disconnected. We had the space to ourselves, he said. The pandemic imperative to work from home had lingered. Most days, he said, it’s only him and the office assistant, who was due in an hour. At first, we chatted, propped up a little awkwardly by the windows, pulling up blinds to gasp at new vistas and snap photographs like giggling tourists, marvelling at the view. Then he led us into the small box that was his office, and we settled down to work, taking our seats on brown polythene-clad chairs around an imitation-wood round table.

The bureaucrat’s mask was so tightly fitted we could not see the expression on his face. Only the dance of his eyes, lighting up with mirth from time to time. He had the dry humour of someone who has been inside the bureaucratic life and found those wry comic moments that can be a lifeline. Emma led the questioning in English. Telma chipped in from time to time; her friendship with the bureaucrat and the warm familiarity between them expanded the scope for our enquiry. It invited a different dynamic to that of the conventional ethnographic interview, with Telma shifting from researcher to informant and back to researcher almost in the same breath, and the bureaucrat telling us stories, then stepping back to join us in their analysis, prompting more questions and reflections.

A foreign researcher might find it hard to get into that particular Secretariat. Not only did we need to come prepared with Covid vaccination certificates and passports in hand. But our bureaucrat also had to come down to the lobby to meet us in person to take us to his office and accompany us out at the end. Nor would foreigners and strangers likely have been so bold as to rove around the empty office – something that broke the ice as brilliantly as a shared joke, enabling us to hover, chatting, before relocating around the formality of an office table. Telma’s familiarity with the scene and friendship with the bureaucrat, both part of a cohort of people who’d studied together, enabled her to not only pitch her questions just right, but also to switch between being interviewer and informant in explaining to the foreigners how things worked. She was also able, with relational capital accumulated over years, to prod our interviewee to open up and tell us some of the things a formal interview might never have yielded.

What we learnt juxtaposed the formal structure of decision-making with the doing of public finance – the capacity challenges, the frustrations of allocations that could not be vied across financial years and had to be re-budgeted, the politics of managing complex demands and contrary political pressures. We were able to probe, using the bridge Telma had built for us, into some of the more challenging political dynamics of public administration and gain insights enriched by our plural positionalities. One aspect of this was an in-depth discussion of participatory budgeting to which we could bring the diversity of the knowledges of the four of us, asking questions framed by a deep familiarity with the context of São Paulo, a specialist’s intimacy with participatory processes, an insider knowledge of federal policy processes, and those that only a complete stranger - but one with considerable knowledge of political institutions elsewhere - would be able to ask.

It was that continuum of familiarities that we came to recognise as one of the aspects of collective ethnography that was especially powerful. It included being able to mobilise questions that spanned from those that only someone who knew the situation from the inside out could pose through to those that an insider simply could not or would not think to ask. As we found, what this offered us was an ability both to rapidly enter into the world of the interviewee and to ‘make strange’ practices, concepts and ways of working that we might otherwise have taken for granted. Most of all, the richness of insight we took from this ethnographic encounter came from the plurality of our subject positions and how the interactions between them enabled us to ask, listen and probe.

It was in this very first interview that Telma began to problematise a concept – mandato - that had dominated her previous research, because it became evident that it was so difficult to translate into English for Emma to understand. Continuities across bureaucratic worlds enabled Andrea to frame questions that drew on intimacy with budgets and planning from her recent life as a university senior manager, as well as her research into institutionalised participation. Cristiane brought insights from immersion in the life of the Brazilian National Congress and Emma from her studies of the UK Parliament.

Our own collaborative process was iterative enough to open a space for the bureaucrat to join us, stepping back from the social facts he had created for our inspection and into the process of theorising and analysing. As the conversation evolved, he editorialised his observations with reflections that made us comment later that it was almost as if he’d joined the research team and become part of the process of puzzling over and theorising what was going on. It helped that brought an analytical gaze towards his everyday life inside bureaucracy. But there was no doubt in our minds that it was the collective process – the fact that there were four people around a table instead of just one - that helped to create an environment of reflective conversation instead of question-response.

The provocateur

Dressed in scuffed white Converse, space invader socks, dark blue chinos and an immaculate white shirt, he perched on the edge of his chair, animated. It was all getting a bit boring, he confided. There wasn’t any conflict. There wasn’t even any controversy. Time, he thought, to spice things up a bit. Make sparks fly. Take to social media and stoke a debate. Get people agitated. It went a bit far that time. As far as the president. And that’s not at all what he wanted; there’s something icky about being amplified by Bolsonaro; he pulled a face as if to say, “yuk!”. But it certainly gave him visibility. And visibility gives him votes. He reeled off the number of followers he’d acquired since he started running his campaigns on social media.

The first time he got elected to the municipal chamber it was almost an exact match of followers and votes. The next time, he’d got more followers than votes; his popularity was starting to pick up in other parts of the country. Although 80% of his followers are Paulistanos, he’s beginning to get a national profile. And that’s what he’s after. He’s got his ambitions set on being part of setting the national agenda. That’s why, he said to us, he doesn’t bother to credit-claim when he sorts things out for people who get directly in touch with him. Unlike the traditional-style politicians who make a meal out of what they do for their voters. That’s because it’s not part of his strategy to turn local-level actions into votes. He’s playing a different game. Three of his team sat at desks with desktop computers, tracking social media rather than compiling spreadsheets with lists of voters as the teams of the traditional politicians around the building might do.

In furnishing his office, he’d gone for a long table rather than a desk, dark grey paint on the walls, plants, a hipster cafe look. On the main table was a plastic foetus, representing his anti-abortion views. On the table to his right, prominently placed and visible to anyone peering into the room from the door, a collection of plaster saints, attestation of his Catholic faith. His bookshelf on the other side of the room bore the full set of Harry Potter books, neatly arranged. Behind the door, in the main office, was a blackboard with bold white chalk entries. In the middle of the board was a slogan that proclaimed, “whatever I’m for, they’ll be against”.

Our post-interview debrief focused first on how the interview went and only then on what was said, and the interest he’d piqued in all of us. Smoothly, was the verdict; also artfully, a clever wending through the questions by Telma, looping back to pick up on what he said. There was something in his way of being that resonated with her, she reflected. We considered this and how it had enabled the interview, but also how it might have restrained the others from asking the unasked questions in our minds. Because the interview flowed so well, Andrea, Emma and Cristiane became, instead, observer-participants rather than co-protagonists. This is a dynamic in collective ethnography that’s worth unpacking. Does one researcher lead, and the others assume roles as observers? Does a division of labour emerge organically, or is it better to plan a potential sequence of questions and question-askers prior to informal discussions or interviews?

Our process was emergent: we found ourselves shifting gear in interviews, moving between intense observation to posing questions and leading and observing, in turn. Between the four of us, this gave us a capacity to experience the nuance and also the tensions in the spaces between the roles of participant and observer that go unmarked in the idea of participant observation as research practice. As Telma wrote later in our collective fieldnotes:

On the issue of reflexivity about how we interview that we have been discussing, I was thinking that my way of operating, besides incorporating the humour we mentioned, is by seeking the spark of what I have in common with my interlocutors, so that I can be true to myself and respectful to the other person and we can create something together in the moment. In this case, we have in common the fact that he likes to create conflict, and I love to unpack/reveal it, so we are both obsessed by conflicts each in their own way.

The idea that a successful interview works when the interviewer latches on to what they have in common with the interviewee got us thinking. Andrea had been observing how Telma navigated the questioning so as to create an ambience in which the interviewee could come into their own. She fed back to Telma how successfully she had been able to do this, how quickly this worked to build an intimacy even with six of us – the four ethnographers, the politician and one of his minders – around the table. There’s much here, we reflected, that’s similar to the way in which we build rapport in other situations, finding commonalities and working with them. These reflections led to more about preparing for interviews and what exactly we all did to get ourselves ready to interview particular people. Telma went on to reflect, in our fieldnotes:

This goes back to the topic about researching beforehand or going “naive” to an interview. If I only have one or few shots with someone, I will definitely go to an interview “informed”, but that doesn’t mean much actually… What is built in our conversation might not allow/encourage me to “use” previous information I have and confront people with it. My “finding a spot of contact obsession” can lead the conversation to some other subject. This is a problem if we don’t have much time to do research because I constantly need to go back to people and ask for things I forgot about or didn’t have the opportunity to ask before.

What Telma identified as a problem can be reframed as a positive attribute; responsiveness that builds on what’s being said, rather than being stuck in effect reading out a list of questions that can end up coming over as non-sequiturs. Even the outcome of only remembering afterwards what else there was on the mental checklist is not necessarily a problem for the ethnographer. If an interview takes an unexpected path, it does not mean that the original questions can’t be asked another time. We compared notes. All of us had had experiences of repeated contact with people where in successive interviews or informal interactions arising from the initial, more formal, interview, we gained a deeper understanding simply because we’d come back. The point, we came to reflect, was that interviews are relationship-building activities that go on to bear fruit as relationships develop and deepen. Working collectively, the rhythm of building connection seemed to be accelerated.

In our interview with the provocateur, we were struck by what else was revealed when we paused to consider the words we were choosing in translation. Having to translate for an English speaker meant having to reach into the meaning of a term in a way that made it curious and strange. Doing this served to frame questions which became the basis of further conversations with interviewees to explore whether those meanings were shared, a process also described by Chiumento et al. (2017). And this in turn led to further reflection on how collective ethnography can work to make strange that which is otherwise taken for granted, through the co-presence of four different minds, working together. The difficulty of conveying these terms in English opened up an enquiry within our enquiry into the rich field of meanings that a question posed by a foreigner could open up. We began posing a question to all of our interviewees: ‘what does mandato mean to you?’. It was Telma’s question. But it was a question she no longer felt able to ask: to the interviewee it would seem puzzling and strange for a fellow insider, one who they knew to have familiarity with politicians and their mandates, to ask such a question. The asking and the answering were facilitated by becoming part of our collective enquiry.

The broker

‘I’ve got a mission,’ the broker told us. She saw herself as a connector. A conduit of people, information, needs, and resources. And of candidacies and votes. From an upstairs office in a busy shopping street in one of the poorer neighbourhoods on the periphery of the city, she spun her web through WhatsApp, phone calls and a schedule of regular meetings of the social movement in which she was a leader. We’d arrived bearing coffee and cake, as had become Telma’s habit when visiting her. Telma had acted as our broker. Cristiane led the questioning. Andrea translated for Emma for a while, but was then forced to concede defeat as the broker’s stories became ever more multistranded and complicated. “This is too complex for me to translate properly,” she said, appealing to Telma to take over. It was a reminder of how much contextual knowledge can be called for to understand politics in this setting. As Telma translated for Emma, the broker jumped onto the phone to carry on sorting things out for people, the work she’d described to us that she spends her life doing, getting stuff done.

Just before we said our thanks and goodbyes, Emma asked a question about what the broker’s advice would be to someone taking over her role from her. And then the energy shifted. She spoke about integrity. About values. About being consistent, following through. About difficult choices. About between remaining true to principles. About inducements, the dirty games played by the powerful to buy themselves votes and cut out the opposition. About being offered easy money that would lift her off the breadline and solve all of her financial problems. About refusing it. We got so vivid a picture of the pressures she was living with that the rest of the interview faded into the background. Suddenly, from a conversation that had hovered on the surface, we were plunged into the depths of the brokerage dilemmas with which she has to contend.

When collective ethnography involves an invitation into an ethnographic encounter from a researcher who is deeply embedded in ongoing research relationships, it presents a lot of risks. There’s the risk that those who are invited in say or do the wrong thing, complicating or even jeopardizing the trust that has been won. There’s also a risk that by bringing in outsiders, relationships may be changed by events beyond the anticipation and control of any party. There’s a risk, too, that these encounters reveal something about the researcher that causes irrevocable damage to their credibility. It’s perhaps especially risky when foreigners are involved. By inviting three comparative strangers into carefully nurtured relationships, Telma was opening herself up to these risks. But there were also benefits. Foreigners could also ask infelicitous questions that no-one would dare to ask, as well as out-of-the-box ones that open up a whole new seam of insight that might otherwise have remained buried.

From our ethnographic encounter with the broker, we came to reflect more deeply on the power of relationships, and the political value of those who were able to mobilise that connectivity to generate votes. We also learnt about the political economy of vote mobilisation. The sales of lists of WhatsApp numbers. The payments made by candidates to the assistants who hand out leaflets. The neighbourhood job lots. The practicalities of the work of connecting candidates to maximise visibility and coverage. In our fieldnotes, we explored the ways the broker was able to use her connectivity to expand that of others, extending in the process her own networks, and the significance of the work of brokerage.

Doing ethnography, as anthropologists know so well, goes beyond verbal interchange. It’s also about scanning the room, noting the little mattress under the desk where the broker’s grandchildren take naps when she looks after them there, the boxes of medicines in a stack at the back of the room that are political currency as well as palliative and cure, the framed photographs and certificates of attendance placing her at political and religious occasions and positioning her within networks of influence. Participant observation enables the researcher to put together what’s said with visual clues or prompts for questions that might otherwise never occur to a researcher or enumerator pursuing a fixed programme of questions. Here too, our collective ethnography came into its own: four pairs of eyes could take in so much more, bringing details to our collective attention and analysis, and positing and challenging the interpretations that came with what we saw.

Working alone, would we have noticed that the medicines were almost all marked up as prescription-only, that the certificates and photos were a mix of presence at government-run trainings and evangelical Church events, attesting to how interwoven the two spheres had become? Perhaps. But it was in the running commentary we gave each other after the interview that we were able to chew over the significance of those details, weaving them together with Telma’s embedded knowledge to build with them a picture that it might have taken a lone ethnographer many months to bring into view.

Conclusion

‘Collective’, ‘collaborative’ and ‘team’ ethnography all imply working together. But as Mauthner and Doucet (2008: 977) note, much of the time ‘many researchers work in individualistic ways such that “the project may in effect become a federation rather than a unified whole” (Platt, 1976: 90)’. As they go on to observe (2008: 977), ‘it is rare for team researchers to come together systematically as an interpretive community where the multiple, situated and distinctive subjectivities and perspectives of the researchers are exchanged in an “interpretive zone” (Wasser and Bresler, 1996: 6)’. Along with the epistemological advantages of triangulation (Tiainen and Koivunen, 2006), there is in the emotional and intellectual process of sharing thoughts and feelings in this ‘interpretive zone’ something significant that distinguishes the collective from the lone ethnographic encounter.

The conviviality built between us through the intensiveness of the process brought us together as research collaborators. It gave us the opportunity to develop practices that could help us sustain this mutuality across the distance that would come after this intensive research moment. Jarzabkowski et al. write of how emotional sharing helped them to feel less isolated, ‘building an ethos of openness within our team that was foundational to other forms of sharing’ (2015: 19). ‘In this way,’ they note, ‘sharing all types of emotional experience, tiredness, humour, and so forth became the norm… such sharing provided a foundation for team-members to engage in constructive and robust debate, which became habitual’ (2015: 19–20). Researching politics and politicians together in this ‘time of politics’, in the context of the intense pre-election build-up gave our work an added intensity; a dimension that is worth highlighting, as collective ethnography might arguably be used to good effect at such political moments, including in the build-up to or the wake of major legislative or policy changes.

This intensive, reflexive engagement helped nudge us out of our conceptual comfort zones, testing assumptions and enabling us to build an enlarged collective understanding. Together, we saw differently. Our interlocutors responded differently than they would have done to an encounter with a sole ethnographer; we did not feel at any time that engaging with a group of ethnographers rather than an individual jeopardised rapport. Indeed, we experienced the opposite: the conviviality of our encounters producing more deliberation and dialogue than question-and-answer. Studies by Palmeira (1979) and Garcia (2018) highlight how ethnographic engagement with political actors in Brazil requires continuous negotiation, particularly when research teams include both insiders and outsiders to the political sphere. This process of negotiation and the interplay of proximity and distance was central to our study, as it allowed us to capture nuances of political discourse and practice that might otherwise be overlooked in conventional political ethnography.

Creese et al. (2008: 200) draw attention to the interactive and negotiated dynamics of what they call ‘interpretive knowledge building’ through the ‘interaction of different identities/values/histories’ of different members of the team. One aspect of our collective ethnography that emerged as an unexpected benefit was just how valuable each of us found the conceptual and contextual insights that came from working in and across two different languages – which could map across to working in different conceptual languages, where collective ethnographers come from different disciplines, and be as rich in terms of insight. Whatever the ways we have of addressing language differences in ethnographic fieldwork, working solo does not provide much of an opportunity to ruminate on the way a term or a figure of speech can be translated, or to pluck one of these out of a stream of speech to show it to someone else and inspect it for its meaning. Being able to do this as part of our collective analysis offered us a richness of insight that might otherwise have eluded us.

Our experiences suggest that making ethnographic research an engaged, collective, process of enquiry may not only improve the quality of the data and analysis. It also, we argue, offers ethnographers a source of insight that usually lies beyond our own reflexive ethnographic gaze: into our own conduct and the choices we make in our interpretive avenues. Of course, not everything can be shared; we remained in that respect four discrete participant-observers, each of us privy to our own unique ethnographic experience (cf. Hannerz, 2003). But conducting joint fieldwork in an intensively interactive way, and sharing thoughts, feelings and reflections through shared fieldnotes as well as face-to-face conversations, permitted us a proximity to each other’s experience of the same ethnographic encounter that would be difficult to replicate in multi-sited or serial ethnographic form. By communicating not only in person but also via a collectively generated set of fieldnotes, we added a layer of depth to our engagement that built a foundation for the analysis that was to follow. We can only speculate on what effect consciously composing a collective ethnography team from a more heterogenous group of co-ethnographers – including ethnographers from other disciplines, as well as more diversity of genders and ethnicities - would have had on what we took away from this experience. Maybe what we perceived as an acceleration in the rhythm of interpretation would not have happened in the same way, but when this more diverse collective of ethnographers finally arrived at a shared view, it would be one that potentially conveyed even more perspectives and even more insight.