Ethnographic Encounters: Using Ethnography to Study Brazil’s Participatory Governance Institutions

Abstract This article explores the contributions that ethnography and ethnographers can make to understanding governance institutions. Told through the story of a collaborative research project on participatory health governance in the municipality of Cabo de Santo Agostinho in north-east Brazil, we explore how the principles of ethnographic research intersected with other principles: of collaborative enquiry, participatory research and a methodological ethics grounded in the reversal of the conventional relationship between researcher and researched. Picking up on some of the core defining features of ethnography as a methodology, we examine how these came to inform our understanding of the dynamics of politics, power, and participation. In doing so, we reflect on how it is the pliable and improvisational nature of ethnography in practice that contrasts most substantially with the conventional tools of the political scientist and that makes ethnographic encounters so rich and valuable for the study of political institutions.


Introduction
It is only relatively recently that political scientists have begun to recognise the usefulness of ethnography to the study of politics (Auyero, 2006).Over the last decade, there has been growing interest in its possibilities.
This article seeks to contribute to the rapidly expanding literature on the ethnography of political institutions (Crewe, 2021).In it, we tell the story of using ethnography as part of a collaborative research project on one of Brazil's many thousands of participatory sectoral councils, the Conselho Municipal de Saúde (CMS) in the municipality of Cabo de Santo Agostinho in the north-eastern state of Pernambuco.The project was part of an innovative research programme called Olhar Crítico (' A Critical Look') that convened social movement activists, non-governmental organisations and scholars of anthropology, political economy and political science to take a critical look at practices of citizenship and participation in Brazil. 1ur contribution to Olhar Crítico began as an unconventional collaborative enquiry in which ethnography played a relatively minor background role.Gripped by the unfolding story, Andrea, the ethnographer amongst us, continued the process of critical enquiry through episodic fieldwork visits to Cabo.With this, the project came to revolve largely around the application of the traditional tools of the ethnographic researcher: participant and direct observation.Yet the original framing of the project, the social movement-led collaboration with which it began and the continued engagement that became part of the dynamic of the research as it evolved, were all decisively different from a conventional anthropological research project.Narrating the biography of the project as it unfolded over the years, we examine what an ethnographic approach can offer in understanding political institutions.We suggest, in conclusion, that it is the pliable and improvisational nature of ethnography and its attentiveness to what Auyero (2007) has called 'the grey zone' of politics that contrasts most substantially with the conventional tools of the political scientist and that makes ethnographic encounters so rich and valuable for the study of political institutions.

Ethnographies of Democracy
When this project began in the early 2000s, there was little literature on the use of ethnography in studies of political institutions to inspire us.Writing in 2006, Javier Auyero describes a 'double absence: of politics in ethnographic literature and of ethnography in studies of politics ' (2006: 258).Anthropologists preferred to 'study down' than to 'study up'.Political Science was dominated by a positivistic mindset that cast aspersion on ethnography as 'impressionistic' and lacking 'rigour'.And yet there was also growing recognition of the limitations of the conventional toolkit of the political scientist.As Auyero observes: … the pace of political action, the texture of political life, and the plight of political actors have all been cast into the shadows created by the unnecessary and deleterious over-reliance on quantitative methods in both political science and political sociology (2006: 258).
By the time the 2000s drew to a close, there was a veritable flowering of ethnographies of politics.The merits of an ethnographic approach to the study of political institutions -understood here in the narrow sense of the term as structures in which binding decisions are made that concern the administration of government, as well as their rituals, rules and practices -had become evident (Wedeen, 2010;Crewe, 2021).
What ethnography contributes to the study of political institutions, above all, is to bring them to life.
Ethnographic writing tells stories.Ethnographers show characters grappling with power, privilege, convention and contestation.These stories enable us to hear unheard voices and get a sense of the 'hidden transcripts' (Scott, 1990) as characters move in and out of formal political arenas.In her 2010 survey of the emerging literature, Lisa Wedeen highlights one of the most valuable aspects of an ethnographic approach, arguing that 'ethnography adds value to political analyses in part by providing insight into actors' lived experiences ' (2010: 61).It helps us understand motivations and get a sense of actor-networks (Latour, 2005) spanning formal and informal political spaces and the 'discourse coalitions' (Hajer, 1995) that form within and connect them.Fu and Simmons refer to ethnography as 'one particularly crucial methodological tool in the study of contention' (2021: 1695) and suggest that By providing a fuller appreciation of the "what"-uncovering hidden processes, exploring social meanings, and giving voice to unheard stories-ethnography and "ethnography-plus" approaches not only help us to appreciate the "what" but help us answer the "how" and the "why" (2021: 1697).
Annavarapu and Levenson describe ethnography as 'a method uniquely positioned to make sense of people's lifeworlds, worldviews, and political logics ' (2021: 343).Asking people to speak about their experience, narrate examples and explore their understandings of what is going on can offer insights that can be difficult to gain from methods such as formal, structured surveys.Interviews with political actors offer the researcher only a partial picture.They can be one-sided, selectively remembered, glimpses from the perspective of people whose positionality influences what they notice, remember and recount.Survey methods share this attribute, even as quantitative political scientists use the numbers they are able to generate in this way as fact.
Teixeira, Crunivel and Fernandes highlight the challenges of access to politicians and observe what they describe as the 'recurrent opacity of political life' (2020: 547), noting that the worlds of politics are ones 'where lies and omissions are part of the game' (ibid).The ethnography of the everyday lives of political institutions brings the dynamism of these institutions to life, in all their messy complexity.Luhtakallio and Eliasoph (2017: 749) speak of 'ethnography's capacity to open windows that traditional analysis of political institutions leaves shut'.Through these windows, to continue the metaphor, researchers can peer into the lives of institutions as lived, with all the 'details and effects of different forms of political action, networks and tactics,' as Auyero and Joseph (2007: 3) put it.' After all,' Auyero observes, ethnography is uniquely equipped to look microscopically at the foundations of political institutions and their attendant sets of practices, just as it is ideally suited to explain why political actors behave the way they do and to identify the causes, processes, and outcomes that are part and parcel of political life (2006: 258).
Ethnography is described in this literature as 'particularly crucial', 'uniquely positioned' and 'uniquely equipped' for the study of politics.This begs the question: what is ethnography?It is worth establishing that there is no single commonly understood definition of ethnography (Forsey, 2010).Some researchers set boundaries with other forms of social research much more strictly than others.For some, ethnography consists of the application of a particular method, participant observation.This involves the ethnographer becoming part of the lifeworld they are studying to such an extent that they are able to understand how it works from the perspective of a participant.This is a form of understanding referred to as 'emic' (from within) versus 'etic' (from outside), from a distinction originally drawn by linguist Kenneth Pike in the 1950s.For others, ethnography involves a rich mix of methods that go beyond participant observation, as described in Pelto and Pelto's classic 1978 anthropological methods manual, including: semi-structured, unstructured and key informant interviews; group discussions; direct and participant observation; visualisation methods borrowed from psychology; and even the use of surveys to establish how widely certain practices or beliefs are found.
It is, however, not just a matter of method.In a compelling account of what the ethnographic research process involves, the sociologist Loïc Wacquant distinguishes ethnography from other forms of social research by focusing on proximity, describing it as … social research based on the close-up, on-the-ground observation of people and institutions in real time and space, in which the investigator embeds herself near (or within) the phenomenon so as to detect how and why agents on the scene act, think and feel the way they do (Wacquant, 2003: 5).Schatz (2013) identifies another distinguishing feature: it is the opportunity ethnography presents for immersion, he argues, that makes it such a powerful method for getting under the skin of institutions and for understanding the biographies and social lives of these spaces.What makes an ethnographic approach so valuable for the study of political institutions is not only that it can get close up and personal.It is also that ethnography enables us to understand better the subterranean workings of power through the shaping of discourse and forms of exclusion in these political spaces.In doing so, it allows researchers to get something of an insider's view into the institutions they are studying.In another more recent review article that captures the rise and uses of political ethnography, Benzecry and Baiocchi (2017) note another key quality to ethnography: a richness of detail that can provide vital clues into how political institutions actually work.
Where previous studies of politics used broad strokes to paint a picture of political life, political ethnography allows the researcher to bring up the mundane details that can affect politics, providing a 'thick description' where one was missing (2017: 232).
A further set of defining features positions ethnography as an emergent, pliable, and adaptable practice through which the researcher evolves an approach that best fits the context, rather than a rigid, narrowly defined enactment of strict rules and restrictive methodological rituals.Timothy Pachirat neatly captures these attributes of the ethnographic process, emphasising the unruly elements of the ethnographer's craft.
Ethnography as a method is particularly unruly, particularly undisciplined, particularly celebratory of improvisation, bricolage, and serendipity, and particularly attuned to the possibilities of surprise, inversion, and subversion in ways that other methods simply are not (Cited in Wedeen, 2010: 256).
From these descriptions of ethnography, a set of family resemblances can be drawn.These stand in sharp contrast with the conventional methodological toolbox of the political researcher.Ethnography is a methodology that is adaptive to the researcher and the circumstances, flexible, informal, improvisational, emergent.It is also, as a number of the authors cited here allude to, a way to zoom in on what's really going on, getting a rich, detailed, picture.It works through immersion and proximity, with those who practice it seeking to see things from the point of view of those they are studying.For these reasons, ethnography relies on the ethnographer(s): they participate, they observe, they write up their fieldnotes, interpret their findings and corroborate them with others, interview, watch and hang out with people.The very embeddedness of the method of the ethnographer in their own personality and persona has been one of the reasons for those of a more positivist bent to be suspicious about the validity -or, perhaps more accurately, pace Lincoln and Guba (1985), the trustworthiness -of ethnographic research.The centrality of the ethnographer to ethnographic research raises other questions about the use of ethnography in research on political processes and institutions.Does the ethnographer need to be trained in social research to practice ethnography, and if not, how does an ethnographer make their ways of working legible to people from disciplines where the collection of data is much more narrowly bound to the creation of artefacts such as the answers to questions, whether in questionnaire surveys or interviews?Much ethnographic research is effectively kept secret; it takes place in the observations, 'headnotes' (Sanjek, 1990) and fieldnotes of the ethnographer.How might the practice of ethnography be democratised, decentering the expert ethnographer?Is this possible?
How does the ethnographer collaborate with others?How does ethnography work in co-creative, participatory research?Is 'participatory ethnography' an oxymoron or a rich field of possibility?
These are some of the questions this article explores.In what follows, we look at the contribution that ethnography and what we call 'ethnographic encounters' made to our shared understanding of the play of politics and power in the everyday life of Cabo's CMS.

Introducing Olhar Crítico
Led by the anthropologist Jorge Romano, then Director of Programmes at ActionAid Brasil and supported by imaginative DfID Social Development Adviser, Sue Fleming, also an anthropologist, the aim of Olhar Crítico was to learn from the unfolding experience of Brazil's participatory governance institutions and from social movements engaged in creating, shaping, occupying and brokering intermediary spaces between citizens and the state.It drew together social movement actors, staff from international and local civil society organisations, distinguished Brazilian researchers from the disciplines of anthropology and political science, and a couple of international political anthropologists.
In early 2003, we found ourselves in the verdant environs of the Hotel Sete Colinas in Olinda at the inception workshop of the programme.Immediately, we came to recognise that this was to be an entirely different kind of way of doing research.The workshop brought together activists and practitioners representing some of Brazil's most prominent social movements, including the movement that had come together to defend usufruct rights to the babaçu nut in the north-eastern state of Ceara, where aggressive commercialisation had violently driven the women and men who had traditionally harvested and processed this oil-rich fruit off the land, and the right to the city movement, who were mobilising neighbourhoods in Brazil's many urban slums.In Gramscian terms, the process centred these 'organic intellectuals' and their immersive, experiential knowledge.Then there were academics, 'traditional intellectuals', who brought their theories, concepts and methodologies.
Turning the conventional relationship between researcher and researched inside out, the workshop took shape around a series of dialogues that centred the activists' and practitioners' experiential expertise.The researchers were invited into the centre of the circle to sit with a leader of a social movement and interview them, asking them questions about their vision for change; the dynamics of leadership and activism in the movement; forms of engagement with the state; campaigns, struggles and victories.The biographies of the movements told in this way were gripping.With their sharply honed questions and interviewing skills, the academics facilitated the activists and practitioners to tell their stories; they then reflected on what those stories might tell us about the politics of democratisation and the realisation of the promise of participation and inclusion.
The research process came to be shaped out of these stories.Four teams were formed, matching activist/ practitioner with academics.Drawn to Silvia's story of the CMS in Cabo, Andrea and Nelson joined Silvia to form one of the teams.Silvia led a social movement organisation called Centro das Mulheres do Cabo (Cabo Women's Centre, CMC), which had been a partner of ActionAid Brasil for some years.Silvia herself was a veteran of the movimento sanitarista, the public health reform movement that led the struggle for a national health service that was free at the point of delivery and accessible to all.She'd joined the movement as a medical student, as social movements came together to mobilise for an end to military rule and a return to democracy in the 1980s.Sanitaristas -activists of the movement -had moved in the intervening years into the service of the state, practising medicine and influencing public health.Passionate about access to health services for all, sanitaristas were everywhere: inside and outside the health system, vital allies for the transformation of public health services and the realisation of the inclusive vision of the Brazilian national health service, the Sistema Único de Saúde (SUS).Some, like Silvia, went on to lead social movement and non-governmental organisations working in partnership with the state, but remaining sufficiently at a distance to play a role in holding it to account.Others entered the state, serving in the commissioning and regulation of local, regional and national health services.They were very present in these participatory governance spaces, acting for both the state and for civil society.
The 'Citizens' Constitution' of 1988 enshrined the dream of a national health system in which organised civil society and health service professionals themselves had a role in financial and policy oversight, which came to be referred to as controle social.A complex institutional architecture was put in place.Participatory conferences, conferências de saúde, were held at regular intervals at municipal, state and federal levels.Tens of thousands of representatives of civil society and the state across the country were convened to deliberate health policy.The creation of conselhos de saúde, participatory health councils, was mandated for each of Brazil's more than 5000 municipalities, 26 states and at the national level.The composition of the conselhos was specified as consisting of four segments.Half of the membership should be from civil society, elected from organised civil society groups registered with the state.A quarter should be health providers, including representatives of the private sector.And the final quarter should be health managers, including the Secretary of Health at municipal, state and federal level.Silvia's organisation CMC was one of the civil society actors in the municipality elected to the conselho.As the representative of this organisation, Silvia served on the conselho and from there, was elected as the conselho's President.She was one of the first civil society leaders to serve in this position nationally.
Nelson knew Jorge from working together at the Federal Rural University of Rio de Janeiro.An economist by training, he'd got a broad expertise in rural development and interest in social movements.He was part of a network of academics brought into the project by Jorge, which included colleagues from universities across the country, anthropologists with expertise on local social movements and political scientists studying the emergence and consolidation of new political institutions given possibility by the 1988 Constitution.
Andrea was part of Jorge's international network, a political anthropologist with an interest in the politics of institutionalised participation who had worked for ActionAid International and had done some work with ActionAid Brasil.With Alex Shankland, a colleague from the UK-based think-tank the Institute of Development Studies (IDS), Andrea joined Olhar Crítico as an international counterpart researcher; their role was to link Brazilian experiences and experiments in democratisation with other global south contexts in which democratic innovation was taking shape.
Staying with the process that had begun in the workshop, our plan was for Silvia to re-tell the story and for us to press pause at different points, so that we could be directed by her to other sources and voices, guided by her questions.In this way, we would explore how political actors of different kinds viewed, used, and engaged with the CMS.Together, we'd reflect on and write about what we learnt in these encounters, using them to deepen our collective understanding of the way institutionalised participation played out in practice.

Ethnographic Beginnings
Within a few weeks, we reconvened in Cabo.Silvia had prepared a mountain of documents, reports, papers, municipal government documents and training materials.We dipped into the material with her.In a conventional research project, this might have taken the form of us carting off the papers to work on ourselves.
In this one, it involved interviewing her to find out about elements that struck her as significant, with her leading the enquiry.This enabled Silvia to point out things that we should take note of and to reflect on the correspondence or dissonance of what was in the documents with what she remembered from the meetings of the CMS or what was going on behind the scenes at the time.Through this, she helped to shape the questions that Nelson and I would go on to ask the people with whom she suggested we speak.
From there, we went on to conduct a series of interviews.The administrator of the CMS took the list of the conselheira/os that Silvia had written down on a piece of paper.She tracked them down one by one, scheduling meetings with them that gave us a run of days packed with meetings.We worked in the room where the CMS held its meetings, with its light green walls, plain wooden tables, dim light from a naked bulb and white plastic chairs.A hand-written poster with ground-rules for interactions in meetings hung on the wall.The administrator had seen it prudent to put it together, she told us, when one meeting had got a bit too feisty and people had begun shouting at each other.We learnt from Silvia that she had tried to expand civil society engagement by having a system in which two sets of representatives were elected: titulares (titleholders) and suplentes (substitutes), representing different organisations, who would stand in if titulares were not able to attend or were stood down.A system was in place where three counts of non-attendance would result in the titulares being replaced by their suplente.Only titulares could vote.But suplentes could and did attend meetings, along with members of the general public.
On that first visit, we spent a week in Cabo.Days full of interviews.Evenings spent in intense conversation reflecting on what we were learning.In all but one of the interviews, Nelson and Andrea went without Silvia.
It was only in interviewing a high-level official in the municipal secretariat that Silvia felt her presence might help open a door and introduce a level of frankness into a conversation that could easily have become more superficial without her.Each day, we'd reflect together on where we'd got to and share emerging insights.Silvia would throw in some more questions to get us to investigate and some more people we might want to speak with.The pace was intense, the process iterative, the learning intensive.We quickly got much, much deeper into understanding the micro-politics of the CMS and the dynamics of institutionalised participation that could have been possible if either of the external researchers had landed up in Cabo to do their own research project.Our academic questions helped to frame reflections, offering theories that were useful to think with or as a foil for disagreement.But Silvia's insights took us to another level of analysis.
In a mere week of interviews, a researcher working alone would only hope to scratch the surface.By the end of the week, guided by Silvia in this way, we'd got a good handle on the story of how the CMS had come into being and how it was working.We'd heard from a wide variety of civil society and state actors.Silvia's social and political capital had brokered the possibility of extensive, illuminating, and frank conversations.
We'd got a picture of the different phases of development, the interface with other governance institutions and the interplay between conselho and conferência in deliberations over policy and in the dynamics of accountability.We'd also tapped into some of the challenges -political, logistical, epistemic, procedural -faced by the participatory governance system in Cabo seen in microcosm in this way.Over the course of the week, we'd also introduced an innovation into the way the CMS meetings were minuted that would be significant further down the line.The administrator had watched how we'd used a small recording device to record the interviews.She commented that she could do with something like this to produce more accurate minutes, so Andrea gave her the recording device as a contribution to the development of the CMS.
But we hadn't had the opportunity to do what ethnographers do: bear witness through participant observation.
We hadn't seen the CMS at work.
A month or so later, Nelson and Andrea returned to Cabo for a workshop that brought the conselheira/os together.Using a participatory methodology, Silvia facilitated an interactive timeline that re-told the story of the formation of the CMS and then put people into small groups to consider the challenges identified in our research.Our principal focus was on how the CMS put into actual practice the principles of controle social in a context where accountability of the state to citizens had been so lacking in the past.Part of the institutional design of this form of participatory governance was that civil society organisations would act as the organised interests of different publics.This, however, was also the CMS's Achilles Heel.If people needed to constitute themselves as representatives of a civil society organisation in order to participate, one registered officially with the appropriate paperwork in place, what potential exclusions might bureaucratising democracy in this way produce or exacerbate?Who didn't get to be seen or heard in these spaces because they hadn't constituted themselves as an interest group in order to seek representation?And who gained space for voice on the back of a registered association who they'd come to affiliate with for this purpose?
We'd interviewed and interacted with those who could make time for us, as well as for the work of the CMS.Was time to participate an inequitably distributed resource that would come to affect the quality of democracy in this space?Many of the people who were present in the CMS as civil society representatives were less well off; they were often working class or lower middle class, some had disabilities and didn't work, others were pensioners.What was it that people got out of being part of the CMS? Quite some number of those who worked were employed by the state, claiming their statutory right to have the time off work to attend CMS meetings.As one confided later, dull as the meetings sometimes were, they were a lot more interesting and less taxing than going to work.Another conselheira showed off her photo album, full of pictures of gatherings and trips to Recife and Brasilia as a member of the CMS.We became curious about the relationship members of the CMS had with other interested parties.These included political parties and particularly the parties of the left for whom these forms of participation formed such an important part of their electoral strategy, and for whose members the CMS could provide a stepping-stone into elected office as a local government councillor (vereador).To what extent, we wondered, was this space laced through with political patronage networks and relationships, and with what implications?These were some of the questions that bubbled up out of our initial enquiry.Olhar Crítico moved on, fast paced, bringing the case studies together and interlacing analysis of the local dynamics of different forms of citizen participation with larger-scale studies of social movements and their role in creating a broader political landscape for transforming inequalities.It was a time in which much was changing for the better in Brazil.These years were the high point of Inácio Lula da Silva's presidency.The Workers' Party (Partido dos Trabalhadores) was transforming urban and rural life with imaginative social policies.The ambiance of governance was thick with progressive buzzwords Cidadania.Participação.Controle social.It was too interesting to just stop there.

The Ethnographic Presence
Out of Olhar Crítico and work with Brazilian political scientist Vera Schattan Coelho and IDS-based political sociologist John Gaventa came a theorisation of 'invited spaces' (Cornwall, 2002;Cornwall and Coelho, 2004;Cornwall and Gaventa, 2006): intermediary spaces between state and society in which state and non-state actors participate in the process of governance.In their ideal form, 'invited spaces' are deliberative arenas in which diverse publics engage with the state and seek responsiveness, listening, accountability and action.
Collaborating with Vera brought a new set of questions, and with it, a return to more conventional style of ethnographic research.Vera had been looking at institutional design in the municipal conselho of São Paulo, generating intriguing findings on how different rules of representation changed the nature of democratic deliberation and decision-making within these spaces (Coelho, 2004).Our collaboration surfaced the interplay between institutional design and the animation of these spaces by political agents whose styles of political action and registers were shaped in domains of discourse heavily inflected with party politics.The interviews that Nelson and Andrea had done were invaluable.But hearing the reflections of positioned political actors on the conduct of others could only lend a partial perspective.To approach this empirically, we needed to get to the heart of the action.We needed to see what was going on in the CMS meetings.
The project took a new turn.Ethnography requires presence, above all.Little else can substitute for it.
It is not something that can be easily delegated or outsourced, although it can be done collaboratively and is especially rich when multiple pairs of eyes and ears observe and witness together.Recordings can capture some of the basic lines of argumentation that might be mobilised in such a space.Narrations and exegeses of documents presented at or produced at meetings can add another layer of insight.But to make any sense of what is going on, you need to sit there and watch, and use being present and witnessing as a jumping off point for ruminations and reflections with others who were also present, to dissect and chew over what was going on.Fortunately, ethnographic study of a conselho is less difficult than studying the day-to-day life of a parliament or other kind of elected chamber.It just requires showing up once a month.
Andrea was able to spend a couple of months in Brazil with her small children and their dad.This made it possible to attend two consecutive meetings of the CMS, get to know the secretariat staff, line up formal interviews and chat with as many conselheira/os as possible as often as possible.Relationships could be built that proved later on to be a vital source of gossip, analysis and leads that could be followed up with other social and political actors in the municipality.During this time, a municipal conferência was held.Convenors from Centro das Mulheres do Cabo and their activist colleagues from Recife used the space creatively, framing key deliberations over maternal health, humanised birth, abortion and contraception.By that stage, Andrea had got to know some of the participants well enough to benefit from running commentaries and back stories throughout the event, as well as to witness the impact of practices of Brazilian participatory governance on the interplay of personality, power and politics.There was, for example, frequent use of a rule that gave people only three minutes to speak before they were closed down; the technique provided chairs with a tool to manage grandstanding, political posturing and mansplaining.
In the years that followed, Andrea was involved in a number of projects with Brazilian colleagues in Rio, Salvador and São Paulo.She took to flying into Brazil via Recife, scheduling her trips to coincide with the meetings of the CMS.She'd drop by the secretariat, catch up on gossip, get a copy of the recordings of the meetings and the minutes, which were extensive and detailed on account of the recordings.This gave her vital clues as to what had been happening since her last visit.She'd bump into conselheira/os in the street, shoot the breeze, trade titbits of gossip.She'd pop into the clinic or the pharmacy or the municipal government secretariat to catch up with the key players.She became familiar enough to the main characters that they'd hail her in the street, call her over for a chat and keep her abreast of the latest twist or turn.

The Failed Survey
Ethnographers face one challenge in making their research legible to political scientists and another in relation to comparability of their findings.Both are substantial.This makes comparative work difficult.For all that epistemological differences can present an unbridgeable divide, collaboration can be facilitated by methods that meet in the middle.And so it was that Andrea decided that it would be worth conducting a survey.
The idea was to take questions from surveys applied by Vera and IDS & CEBRAP colleagues Peter Houtzager and Adrian Gurza Lavalle, who were doing some interesting work on civil society participation in participatory governance spaces.A great deal of time was spent researching and putting together a questionnaire that would be applied to the list of civil society organisations who were registered as associations and appeared on the list of eligible organisations from whom representatives could be put forward for election onto the council.
Andrea pre-tested the survey, finding herself straying time and again from the questions.Something that is under-examined in the literature is what doing ethnography does to the researcher; it becomes difficult, as a habituated ethnographer, to simply plough through a list of questions without wanting to probe and to follow up answers with unscripted questions.There is nothing more affirming of the ethnographic method than being exposed to the limitations of the conventional questionnaire survey at first hand by administering it in person.
Care was put into thinking about who might best administer the survey.The prospect of wool being pulled over the interviewer's eyes was high.The administrator of the CMS initially offered to do it.Then she found a colleague who'd been helping out in the office.The colleague had enough knowledge of the local scene to put the questions credibly.We were set to go.The first step seemed simple enough: to compile a list of all the civil society organisations in the municipality who were eligible to enter candidates to be elected to the CMS.
From this, we'd be able to compose a purposive sample.As it turned out, this was hardly a neutral undertaking.
There were many reasons for the creation of such a list to be politically undesirable, as we would soon find out.The survey process faltered at the first hurdle.What began to emerge from what would seem a relatively innocuous task became too difficult for the enumerator who had been hired for the task to persist with.The politics was such that it would likely have cost her at least her job.
Getting an accurate list of civil society organisations does not sound that tricky a task until you begin to reflect on what such a list might reveal.It was rumoured that some of those registered as associations were allegedly fronts for private medical companies or people with other kinds of political interests in who was on the CMS.Having dud organisations on the list of civil society organisations eligible for election offered two benefits.It supplied votes, as the system worked by identifying first the pool of potential representatives and then inviting everyone in that pool to vote.This could be useful for the administration if they were bent on pushing through a particular policy, as well as for getting accounts signed off without trouble.It also supplied seats for candidates who might have difficulty getting elected on the ticket of the association they were more closely associated with, either because that association had held a seat that was rotating or because the association was unpopular with a significant number of the electorate.Rules of representation requiring the presence of organisations representing particular kinds of civil society organisation -those working with children, social care, older people and so on -were a way in which candidates could find themselves a seat, as they operated effectively as quotas.All the candidates had to do was to stand on behalf of that organisation, and they were in.
The failed survey was a disappointment.There would be no numbers to compare, no comparative answers to the same questions to yield material to mull over.But if we hadn't tried to create a census in order to sample civil society organisations in the municipality it might never have come to light that 'civil society participation' was not all that it seemed.All of this was itself an 'ethnographic encounter' and grist for the ethnographer, rather than a major blow.It yielded some fascinating ethnographic material.

Ethnographic Interactions
Like an old friend, ethnography was there to cushion the fleeting dismay caused by having to abandon the survey.Andrea returned to it with gusto, sitting in CMS meetings taking as close to verbatim notes as possible, recording dialogue, interactions, everything that was going on 'front stage'.along with 'back-stage' looks, whispers and facial expressions.To aid what the anthropologist Clifford Geertz (1973) so memorably described as 'thick description', she also began experimenting with other methods for capturing what she was seeing.Emma Crewe writes: … ethnography entails far more than interviews, shadowing, and a dash of observation.Ethnography requires whatever methods and theories assist in an analysis of the cultural logic of institutions rather than a set recipe of tools and techniques (2021: 3/14) One of the methods that proved particularly interesting was an interaction map that built up a cumulative picture of the meeting.This started with a picture of the room with an x for each person.Each time people spoke, a circle would be drawn around their x.Arrows would indicate where they were aiming their remarks, whether into the centre of the room, towards the Chair or at neighbours.The map provided a visual record of the meeting, showing how key actors were shaping the debate.It was also useful to situate and trace the monopolising of the dialogic space by particular characters.It gave a far more accurate picture than what might be picked up from just observing the meeting.
It was amply clear from the interaction map that there was not only uneven but very partial participation.
There was also some 'strategic non-participation' (Cortez Ruiz 2004) going on.Some conselheiras/os remained silent throughout.Some never got a word in edgeways.Some failed to get the President's attention for their raised hand and didn't follow the most assertive of the members of the CMS by simply barging into the conversation by talking loudly over others as their voices tailed off, in that short interval in which interjections can steal into the space before the President once again notes a raised hand.But others simply sat there without ever seeking to interject.The best-known silence was from the woman who was nicknamed "the silent nun".She turned up bang on time for meetings and left as promptly, without ever exchanging any words.Rumour had it she was one of the clients of the municipal administration, who allegedly sought to stack the CMS with people who wouldn't vote against their budget or policies.Yay-sayers.There were also the nay-sayers, the ones who would object out of the pleasure of objection.One of them, an older Black man, would spin long animated tales of contestation and conquest in the run up to the meeting and object vociferously if there was the slightest variation in procedure being proposed.But he would often fall asleep during the presentation of the budget and programmatic priorities.
While he was the only one who dropped off during the main business of the meeting, he wasn't alone in being reduced to dull silence when such matters were under discussion.This in itself was intriguing.The ostensible business of the conselho was to deliberate and then to sign off on policy, the accounts and the allocated health budget.But rarely were any questions raised about the formal business being put before the meeting.Indeed, it was relatively rare for any observations at all to be made about any other regular agenda item concerning the core business of the CMS.Was it that people were ill-equipped to ask questions about the accounts?Were they bored by the detail?Or was there something else at play -especially where an alignment of political party affiliation and state-civil society patron-client networks (after all, many of the organisations present were contracted by the state to deliver services and may well have had something to lose if they started asking too many questions) might make studied silence or obfuscation a political strategy in itself ?
In contrast, there was frequently animated debate about ways of working.The conduct of conduct of the meetings appeared to be of much more concern to participants than what they were ostensibly there to do.
In fact, so much so that this effectively became the core business of the CMS, with the formal business of the meeting relegated to a swift run-through close to the end.Of course, the art of agenda-packing and strategic placement of key items at the end of an agenda is as well known in these spaces as in any others.But what the ethnographer was able to witness was something else.It was the very deliberability of the processes and procedures of participatory governance itself that was evident in these meetings.Participatory governance in this context took the form of vigorous participation in contesting, debating, and redefining the very fabric of the governance institution itself.Elsewhere, Andrea uses an extended case study of a challenge to the rules of representation to explore this theme and its implications for how we think about deliberation and democracy in settings like these (Cornwall, 2008).

Ethnographic Sensibilities
Dropping in, catching up, connecting the dots, Andrea attended as many meetings as she could and worked hard to make sense of the ones in between.Silvia stopped going so regularly to the CMS, although the CMC retained a seat as a civil society representative.The CMC were to go on to use that seat tactically to bring in a range of less prominent civil society actors to gain experience of working with government in these spaces.
They were able in this way to open otherwise closed or colonised spaces to interest groups who would otherwise barely gain any opportunity for voice or presence: the fledgling LGBT association in the town, the sex workers' association, local women's organisations from the low-income settlements at the edges of town.Arguably, they were drawing on and amplifying their own patron-client network in this way.But through the art of substitution, with mentoring and support, they were able to deepen and broaden democratic engagement across the various conselhos in the municipality.Mapping those spaces, strategizing over deployment into them, and then coordinating across them, served to enable CMC to work in a joined-up way across a range of sectoral conselhos.
Equally, the political parties were able to operate in a similar fashion: mapping the spaces and placing candidates in them under the guise of a cluster of civil society organisations that could be used interchangeably so when the tenure of one ran out, the same person could re-appear under the badge of another.Links through into the formal political arena were evident, both in people's aspirations and career trajectories and in terms of sponsorship and support.Aspirants for elected political office saw the CMS as a soft entry into a representative role that could give them visibility with the municipal government and with the citizens in their areas, a springboard into electability as vereadores (municipal councillors).Understanding the role of political affiliations and appreciating the interconnectedness of political spaces and the actors who traverse them is vital in decoding the political dynamics of the space of the CMS.The conventional instruments of the political scientist might well miss the way in which the traces of interactions in other spaces are etched into 'invited spaces' (Cornwall, 2002) such as these, simply because these are questions that are tricky to ask and difficult to frame without having witnessed these dynamics in play.
Participant observation revealed the extent to which religious organisations and political parties kept the conselhos and associated institutions like the conferências in a relatively tight grip.Andrea saw this play out on a larger national stage when she travelled to Brasilia to the Conferência Nacional de Saúde and watched the Catholic Church skilfully mobilise to block progressive policy moves in reproductive health.It was also evident in the way Lula's government worked to frame the space.At the Conferência, the slogan "Aqui é permitida sonhar" -"Here, it is permitted to dream" -hung in a large banner over the proceedings, speaking to the democratising impulse of the participatory governance architecture as a space for the imagination, the spirit and for the diversity of Brazil's citizenry to find a place and a voice.This vision of participatory governance is one that is inherently party-political: strongly associated with the Partido de Trabalhadores (PT) on whose platform Lula came to power, although also emblematic of a philosophy of governance embraced by other leftist parties.The power of that imaginary was regularly invoked through the language used to describe what was happening in these spaces and the very idea of them being deliberativo -not deliberative, in the English sense of the word, but spaces with decision-making power in which binding decisions can be made.
Participatory governance institutions are arenas where agendas are rehearsed, demands are framed and constituencies sought and built on issues that play out in formal political spaces.They are also places where dense patronage networks that span different kinds of political spaces are activated and operationalised, where political points are scored, and political narratives mobilised.These are spaces in which it is those with speaking skills that are honed through party-political gatherings will likely stand the most chance of being heard, and where the less eloquent or fervent can find themselves talked over or their points left hanging in the air as the debate moves onto ground that has been better rehearsed elsewhere.There's a gender dynamic to this, but it is one so inflected with political affiliation that it is complex to discern or disentangle.Ethnography provides powerful methodological tools to make sense of how all this works to create a particular ambiance, with particular political effects.Andrea attended the training.It was very dry.No-one asked questions.In the next meeting, the debate was all about the rules again, rather than the content about health policy and the setting and approval of budgets on which the cohort had been trained.Deliberating democracy continued to be of far more interest to members of the CMS than poring over the accounts or participating in deliberation over the finer details of health policy (Cornwall, 2008).
Over time, the project morphed into a longitudinal ethnographic study of the CMS, tracing its fortunes through different conjunctures: changing municipal administrations, with a sequence of different political parties taking charge of the municipality; changing Presidents of the CMS, with a succession of three civil society Presidents and one two-year period in which the Secretary of Health took charge, as was standard elsewhere in the country; changing national and state-level politics, with one period of alignment that ran all the way through from municipal to state to national and in which a host of democratising changes seemed possible, but in which the emerging political dynamics that were to undermine that potential were all too evident.The term 'ethnographic sensibility' has gained currency in Political Science over the last decade or so (Pader, 2006;Jourde, 2009;Simmons and Smith, 2017;Herzog and Zacka 2019).Herzog and Zacka write: Adopting an ethnographic sensibility involves being interested in what people do as well as why they do it.It is to be concerned with how they perceive, think about, and ascribe meaning to their environment and behavio[u]r… it involves observing how people respond to specific situations and trying to make sense of what these situations look like to them -interpreting their interpretations of the social world (2019: 764).
Approaching the study of political institutions with this sensibility enables a closer reading of what is going on.An ethnographic sensibility can render the participant observer alert to the dialect of political activists from the left parties that saturates these spaces, or the party-political gestures and rituals that create a continuity with other, more informal, spaces, where people come together with members of their political party or gather for rallies.This sensibility infused conversations about the CMS between us, and with other observers and analysts of the CMS with whom we engaged.These were reflective, descriptive, interpretive discussions worthy of any ethnographic research project.Participants in those conversations brought examples, adduced evidence, lent critique and drew out broader conclusions.That these ethnographic reflections were collaborative made them more pithy, nuanced, and insightful.

On Political Ethnography: Finalising Considerations
As this article has sought to argue and affirm, ethnographic research has much to offer the study of political institutions.The different phases through which this research project evolved illustrate a number of different modes of using ethnographic methods for researching political institutions.The initial, more collaborative, phase brought together an experienced social movement leader with deep knowledge of the institution, a senior Brazilian political economist equipped with comparative insights, and a British anthropologist with an interest in the micro-dynamics of power and the application of political theory.The process reversed the relationship between researcher and researched, as the social movement leader deployed the researchers as research assistants, framing questions and guiding analysis.Autoethnography of sorts came into play as the social movement leader reflected on her own experience and worked together with the researchers to tease apart the dynamics in the creation and development of the CMS.
As the project morphed into a biography of the CMS over a longer period, researched through intermittent ethnographic encounters, it became more conventionally ethnographic.In the process, ethnography brought us rich material to think with.Being able to engage with the ongoing saga of the CMS with new questions, new conceptualisations and through new ethnographic encounters provided a deeper understanding of the wider locus of these institutions in a crowded political landscape, and into some of the otherwise hidden dynamics and unintended consequences of their institutional design (see Cornwall, 2008).Those moments of close observation, the hastily scribbled notes seeking to capture every detail, every interjection and response, every scowl and smile, all evoked the texture, the dynamics and the feel of what was going on in ways that descriptions in interviews would not have been able to capture.The collaboration and the relationships built in the first phase enabled a greater depth and intimacy of understanding that would have taken a much more immersive ethnographic experience to build.Over time, our collaboration brought other insights, as we engaged with different audiences and framings.
What did we learn that the conventional toolbox of political science would not have taught us?Some of what we learned might have been hypothesised, but would have been difficult to verify using surveys: that there was a dense, capillary network that linked actors in different political spaces and made the conselho a site for a number of projects, including the pursuit of personal agendas, that played out in other arenas -what Auyero calls the 'grey zone of politics' (in Auyero and Joseph, 2007); that modes of 'doing politics' in the conselho were imbricated with cultures of politics in other spaces, notably those of political parties; that leadership played a determining role in realising or dampening the democratic potential of spaces such as these.But there were things that we found out that might never have been asked: that the 'civil society' presence in the conselho served many purposes, not least the maintenance of existing clientelistic relations with the government and that very few, if any, of the 'civil society' participants engaged with the ostensible purpose of the conselho; that participation in participatory governance spaces was less about civil society organisations exercising 'citizen voice' than a project of civic education that could equip 'uncivil society' with the knowledge and understanding to play their part in the designs for democracy that the conselhos represent; and, ultimately, that institutional design can be both a bulwark against the incivility of 'civil society' and in its undoing, the unmaking of democracy within spaces such as these (Cornwall, 2008).
More than anything, our experience affirmed the value of ethnographic research in offering the political theorist and scholar of political institutions precious glimpses inside these spaces, bringing political culture into view and into analysis with all the 'details and effects of different forms of political action, networks and tactics', as Auyero and Joseph (2007: 3), cited earlier, suggest.Ethnography animates the scene with a cast of characters, their gestures and rhetoric, their raised and dimmed voices, their disagreements, and their silences.
With this, we gain an expanded understanding of the texture of political life, and with it the political dynamics within institutions that are difficult to otherwise perceive or grasp.What ethnography permits the researcher to do is to be curious about concepts that might otherwise be taken for granted, like 'civil society', 'democracy' or 'accountability', and to explore the meanings they have for those who inhabit political institutions.This, too, is a way in which ethnography contributes to the study of politics: by 'making the familiar strange and the strange familiar'. 2The techniques of the ethnographer, the pliable and improvisational qualities of ethnographic method, an 'ethnographic sensibility' and the pursuit of 'ethnographic encounters' all, we contend, make an invaluable contribution to the methodological toolkit of researchers studying political institutions, whether as part of collaborative, participatory, multi-methods, or ethnographic research.It is time more political scientists began broadening their horizons and using them.
Received: June 28, 2022 Approved: January 20, 2023 The CMS became for Andrea like a long-running telenovela.It didn't quite feel like the right time to stop watching.Her visits continued.She'd drop in on the office, meet up with Silvia and colleagues in the Centro das Mulheres do Cabo to chew over the latest, and touch base with conselheira/os old and new to hear what they made of what was going on.A new President was elected.He was a Black man from a low-income neighbourhood who was physically disabled and used a wheelchair.He took the conversation back to equipping the newly elected conselheira/os with training, so they knew what they were doing.With each new President and each new cohort of conselheira/os, this demand would resurface.This time, though, it seemed to be happening.