Research Methods Literature Review PLEASE SEE DOCUMENTS NEEDED ATTACHED!!!

Everleigh

ethnographydesign-qualitativemethod.pdf

Home >Psychology homework help >Research Methods Literature Review PLEASE SEE DOCUMENTS NEEDED ATTACHED!!!

Doing Ethnography: Ways and Reasons

In: The SAGE Handbook of Qualitative Data Collection

By: Marie Buscatto

Edited by: Uwe Flick

Pub. Date: 2018

Access Date: August 16, 2021

Publishing Company: SAGE Publications Ltd

City: London

Print ISBN: 9781473952133

Online ISBN: 9781526416070

DOI: https://dx.doi.org/10.4135/9781526416070

Print pages: 327-343

This PDF has been generated from SAGE Research Methods. Please note that the pagination of the

online version will vary from the pagination of the print book.

https://dx.doi.org/10.4135/9781526416070

Doing Ethnography: Ways and Reasons

Marie Buscatto

Introduction

More than a century ago, anthropologists adopted ethnography as their main method (Taylor, 2002). Early

on, they defined general scientific principles guiding the study of ‘foreign’ societies (Gobo, 2008). But

ethnography has, also very early on, been adopted by other disciplines, such as sociology, opening new ways

to study contemporary societies (Madge, 1963). For instance, at the end of the nineteenth century, Frederick

W. Taylor founded his ‘scientific method’ based on his personal observation of workers (see Wästerfors,

Chapter 20, this volume). In the 1920s Elton Mayo and his colleagues founded the ‘Human Resources School’

following the ‘Hawthorne experiment', which was mainly based on observations led among workers in a

plant.1 After World War II, prominent researchers in Chicago, such as Erving Goffman, Howard S. Becker or

Donald Roy based some of their most renowned works – respectively dealing with mental illness, marijuana

smokers, or plant workers – on long-term systematic observations.

This method has now become quite legitimate in sociology. What, then, is defined as ethnography? How

may the use of ethnography enrich one's scientific results? How is an ethnographic study to be conducted

over time? What are the issues to be dealt with to avoid ethnography's potential limits? What are the new

developments? These are some of the questions to be dealt with in this chapter, based on current and past

literature and my reflexive experience as an ethnographer.

What is ethnography?

Ethnography is mainly defined as long-term observation (see Wästerfors, Chapter 20, this volume) personally

conducted in situ by the researcher and aimed at producing specific data:

In its most characteristic form it involves the ethnographer participating, overtly or covertly, in

people's daily lives for an extended period of time, watching what happens, listening to what is said,

asking questions – in fact, collecting whatever data are available to throw light on the issues that are

the focus of the research. (Hammersley and Atkinson, 1995 [1983], p. 1)

Ethnography may take several forms – participant, semi-participant, or external. It may be conducted in

various ways – for example, following the same people through different places (see Kusenbach, Chapter 22,

this volume) or being focused on one or several specific settings. It may be implemented in foreign countries

over long periods of time (as in most anthropological works) or in the researcher's own country over shorter

periods of time (as in most sociological works). It may be done overtly – those being observed know about it

SAGE

SAGE Research Methods

Page 2 of 20 Doing Ethnography: Ways and Reasons

– or covertly. Ethnography can also be used to study society, adopting quite differing approaches. While Gary

Alan Fine promotes a ‘peopled ethnography', to be distinguished from ‘postulated’ or ‘personal ethnographies’

(2003), Loïc Wacquant defends the need for ‘carnal sociology’ (2005). Whereas James Clifford and George

Marcus argue for the development of ‘postmodern ethnographies’ (1986), Howard Garfinkel is in favour of

an ‘ethnomethodological approach’ (1967). If Michael Burawoy develops the ‘extended case method’ (1998),

Barney Glaser and Anselm Strauss earlier expose the virtues of ‘grounded theory’ (1968).

In other words, even if one tends to associate ethnographic sociology with the interactionist paradigm, due to

the powerful influence of the interactionist ‘School of Chicago’ researchers – Hughes, Goffman, Becker most

notably – ethnography can also be used to develop any theoretical framework, be it Marxism, postmodern

theory, ethnomethodology, phenomenology, conversation analysis (see Jackson, Chapter 18, this volume),

pragmatism, or social fields. Those different uses are not necessarily compatible with each other since

ways to collect, to interpret, and to present data may vary strongly among those sociologists, as powerfully

discussed in Katz's (2004) controversial article. But whatever researchers’ ‘ethnographic styles’ (Wacquant,

2003) and main theoretical framework, the two central features of ethnography remain the effective presence

of the ethnographer in the world she studies and its long-term systematic observation:

‘We characterize ethnography as sustained, explicit, methodical observation and paraphrasing of

social situations in relation to their naturally occurring contexts.’ (Weick, 1985, p. 568) (…) It

excludes when the fieldworker does not remain in the field (for months or for years) to become

saturated with first-hand knowledge of the setting. (Glaser and Strauss, 1967) (in Morrill and Fine

1997, p. 425)

Ethnographic data are mainly constituted from the researcher's personal in situ systematic observations

carried out over time. But they also include any other data collected in the field during the course of

observation, such as informal conversations, formal interviews (see Roulston and Choi, Chapter 15, this

volume), documents (see Rapley and Rees, Chapter 24, this volume) or objects which are part of those

observations (Gobo, 2008).

What are the main contributions of ethnography to the study of social

phenomena?

One may summarize ethnographic main contributions to the study of social phenomena around three main

ideas.

First of all, ethnography provides a privileged access to ‘invisible’ or difficult to access social phenomena.

Observation, as opposed to questionnaires or interviews, gives access to people's practices, and not to their

oral justifications or representations (LaPiere, 1934), that is, to all those ‘natural', hidden, taboos or difficult to

express practices which people have difficulty in describing (or would not like to describe even if they were

aware of them). While interviews or questionnaires enable people to express opinions or to describe practices

SAGE

SAGE Research Methods

Page 3 of 20 Doing Ethnography: Ways and Reasons

as they remember them, direct observation gives access to practices while they are performed. This has been

key in unveiling dance musicians’ or marijuana smokers’ practices (Becker, 1963), the ways mental illness

is daily fabricated through ritual practices in mental institutions (Goffman, 1961) or plant workers’ resisting

practices (Roy, 1952; see next section).

Moreover, since ethnography is led over time, the researcher is also able to discover the ways that practices

are socially produced through action and interaction, how they are legitimated, how they get transformed

through negotiation, conflict, and confrontation between actors. One may then better grasp key social logics,

even if sometimes contradictory, which underline people's practices through time. As stated by Michael Rosen

about organizational ethnography, ‘by definition, ethnography is a longitudinal method, geared towards a

process-based understanding of organizational life’ (Rosen, 1991, p. 12). Following this argument, specific

organizations can then be thoroughly defined as collective worlds which get produced, legitimated and

transformed through organizational members’ actions and interactions such as the specific neighbourhood

investigated by Whyte (1943). Thanks to long-term personal observations, the researcher is able to produce

a complete understanding of how a specific observed world (such as a hospital, a school, a neighbourhood

or a restaurant) or a specific group of people (such as artists, sick people, drug dealers or homeless people)

collectively produce their lives over time.

One last, but not the least of the advantages of ethnography, is to get access to multiple perspectives

and practices as developed by observed people over time. In all social activities, logics of action may vary

depending on the time of the day, the week, or the year. Actors’ rationalities may also vary depending on

several factors, which may change And since the ethnographer is to observe people's activities in the long

run, she will be in a good position to distinguish exceptional facts from recurring activities, general rules and

norms from disrupting behaviours… and see how those recurring and exceptional events work together to

produce, legitimate and transform the observed world, group of people, or actions.

Machine workers’ resisting strategies: a historical illustration

Let's take a famous historical example presented by the Chicagoan interactionist Donald Roy (1911–1980),

which has highly influenced the development of the sociology of work. Based on the analysis of covert

participant observations made during eleven months of work as a radial-drill operator in the machine

shop of a steel-processing plant in 1944 and 1945, he showed how workers were fighting a rational and

collective war against management through what he called ‘restriction’ practices. In his 1952 article, he

describes the active strategies collectively shaped by machine workers in order to defend their economic

and personal interests against management rules and orders: his ethnographic work unveiled two specific

workers’ resisting strategies, ‘quota restriction’ and ‘goldbricking', as will be shown now.

Based on his own productive output, he first found that his behaviours were quite simple to describe: either

he was getting a $1.25 to $1.34 an hour pay, or he was getting a salary far below the 0.85 cents official basic

salary per hour. And what he called ‘this bimodal pattern of hourly earnings for the ten-month period’ (Roy,

SAGE

SAGE Research Methods

Page 4 of 20 Doing Ethnography: Ways and Reasons

1952, p. 428) appeared to be very close to those of his main fellow workers. It ‘was the rule of the shop’ and,

based on his daily observations, this rule was clearly produced by two specific resisting strategies: ‘quota

restriction’ and ‘goldbricking'.

On the one hand, the concentration of the $1.25 to $1.34 level was due to ‘quota restriction’ behaviours: the

fixed quota being considered as feasible to earn at least 1 dollar, workers did produce more than expected,

but avoided producing too much above a 1.25 dollars quota, so that the set quota would remain as such (and

would not be downgraded by the Methods Department). On the other, the concentration of hourly payments

far below the normal pay of 0.85 cents was due to ‘goldbricking’ behaviours: when the quota was considered

as too demanding to earn at least 1 dollar an hour, workers preferred to ‘hold back’ and to earn much less, so

that the Methods Department would have to improve the quota. In both cases, workers then had ‘free time',

sometimes hours to spend, in order to respect both strategies and ensure the quotas remain feasible and

acceptable to them. As observed by Roy, ‘I reached the peak in quota restriction on June 27, with but three

and a half hours of productive work out of the eight’ (Roy, 1952, p. 433).

If those strategies were economically sound for those machine workers, they were also collectively

developed. As regularly detailed by Roy, workers were spending a lot of time assessing jobs collectively,

making sure that new hires did not work too fast and did not threaten their resisting strategies, and explaining

to them all the gains to be drawn from such strategies (protect feasible quotas, avoid working too hard, having

‘free time’ on the job). Workers led a collective, even if hidden, war against management and the Methods

Department.

Ethnography has been key in unveiling those practices since workers tend to hide them. Long-term

observations also enabled Roy to discover the bimodal pattern of restriction and the ways it was collectively

built-up over time. He also showed that despite the multiple jobs to be dealt with in this plant, this pattern did

not change over time and was driven by machine workers’ economic and personal interest. Roy's results have

been key, right from the 1950s, in developing a better understanding of how first-level workers are able to

collectively fight to defend their working conditions – including their level of earning – using quite rational and

well-organized strategies as is now regularly demonstrated in all organizations (for example, see Buscatto,

2008).

Better grasp ‘GENDER AT WORK': a contemporary illustration

Following this first historical example, I will now show how contemporary ethnography has been quite

instrumental in better grasping gendered differences in work settings or situations. Overall, ethnography

does enrich the identification and analysis of gender in work settings or situations precisely by inducing the

researcher to focus on ‘doing gender’ (West and Zimmerman, 1987).2 It is in daily interaction that ‘sex-class

makes itself felt, here in the organization of face-to-face interaction, for here understanding about sex-based

dominance can be employed as a means of deciding who decides, who leads, and who follows’ (Goffman,

1997, p. 208). Ethnography enables the researcher to unveil how gender segregation is produced by actors

SAGE

SAGE Research Methods

Page 5 of 20 Doing Ethnography: Ways and Reasons

in the work context. Gender segregation, a complex, dynamic phenomenon operative at different points in the

social order, is actually enacted by actors in the work situation. To deepen this initial conclusion, I will now

describe the three specific contributions that ethnography can make to the study of gender in work situations

in enabling researchers to spot disparaging gendered stereotypes, occupational networks, and norms (for

detailed explanations, see Buscatto, 2010).

First of all, direct observation proves extremely effective not only in identifying existing gender stereotypes but

also to grasp how they are constructed, how they operate in work situations and help produce professional

legitimacy. For example, in my study of French jazz singers (Buscatto, 2007a, 2007b), through observing

singers at work – rehearsing, performing on stage, jamming, – and in leisure situations – drinking, partying,

going out, joking in informal situations, I progressively came to not only unveil two key feminine pejorative

stereotypes associated with seduction and lack of technicality, but also the ways those pejorative stereotypes

favoured a strong devaluation of those female singers, and their marginalization in the French jazz world. Due

to those two stereotypes, female singers tended to be considered as potential sexual objects while their voice

was not viewed as a ‘real instrument', which negatively affected their reputation as professional musicians. By

focusing my attention on daily interactions between musicians, ethnography helped identify this tie between

specific negative gender stereotypes and their effects on men's and women's professional reputations.

Second of all, ethnography is useful not only in identifying the difficulty women have in getting into and

remaining a part of ‘masculine’ social networks, but also and above all, in spotting the social ‘reasons’ for this,

as shown in my study of women jazz instrumentalists (Buscatto, 2007a). Social networks are crucial when

it comes to keeping one's place in fluid art worlds and constructing a relatively good professional reputation

in them (Becker, 1982). I was able to show that French women jazz instrumentalists over the age of 30,

whatever their musical reputation, could not make a living primarily from jazz, partly because they had such a

hard time getting into stable social networks; in this regard they were extremely dependent on their male jazz

musician (programmer, agent, or producer) husbands or partners (when they had one). It was by comparing

my observations with statements from these men and women, collected in interviews and informal situations,

that I realized that, except for friendly moments spent with their husbands’ or partners’ musician friends, these

women instrumentalists were not likely to have any prolonged friendships with male jazz musicians. One

reason appeared to be that, when playing music and sharing moments outside of musical work, jazzmen felt

more comfortable with other jazzmen. Masculine socializations had prepared them to share more topics of

interest, to behave in more similar ways on a daily basis. A second reason explaining female marginalization

was their high seductive power – as for female singers – which made them difficult to envision as regular

colleagues. Interviews did enable us to unveil this phenomenon: when musicians were asked to describe their

professional connections (groups and invitations), female musicians were not often mentioned by their male

counterparts as regular colleagues and female musicians had fewer opportunities to work on a regular basis

as compared to their male colleagues. But observations were key in analysing why female musicians tended

not to be treated as regular colleagues. Male-only interactions appeared much more relaxed, ‘natural’ and

easy to handle for male musicians than mixed interactions, both in work situations – including dealing with

conflicts and fights over music, – and in leisure situations – when partying, drinking, travelling. Male norms

SAGE

SAGE Research Methods

Page 6 of 20 Doing Ethnography: Ways and Reasons

made it easier for men to work with other men in this jazz world, partly explaining female instrumentalists’

marginalization over the age of 30.

Third, women may become ‘fed up’ with working in male milieus that they consider unlikely to include them

professionally or where they find it difficult to gain recognition for their professional abilities. This difficulty

is explained in part by the maleness of work norms that make it harder for women to participate than men.

Here again, ethnography helps in identifying these gender norms and their possible effect on the construction

of occupational work positions, as shown in my study about unionists (Buscatto, 2009). The very way that

union activity operated was strongly marked by norms that are socially constructed as ‘masculine’ – topics of

conversations, valued public behaviours, flexibility outside of work. All this made it difficult for women to gain

a foothold in union worlds. As this example clearly showed, it was by comparing data collected in interviews

– concerning areas of interest, modes of acceding to union positions, ideas about union activity – with data

collected through observation – union member gatherings, social behaviour, the mutual inviting that union

members do – that I could spot which social norms were valued and/or activated by sex and what their

gender-related effects were on unionists’ trajectories. Unionism was a ‘male’ activity, and as such it was less

likely to attract women than men. When it did attract them ‘anyway', it afforded them a position that they were

then hard-pressed to construct in a ‘masculine’ way.

Reflexivity as a scientific model

But then, how does one conduct an ethnographic study scientifically? The anthropologists who invented field

studies in the late nineteenth century, most notably Bronislaw Malinowski and Edward E. Evans Pritchard,

defined their way of proceeding to be scientific (Gobo, 2008). The stated aim was to collect ethnographic data

in an objective way. But anthropology was destabilizing the positivist scientific model and ethnography soon

appeared as incomplete, lacking and imperfect (Buscatto, 2016).3 It thus could not be practised with a blind

respect for the positivist method based on the presumed superiority of the hypothetico-deductive model and

the understanding that it would ensure objectivity for the collected ‘data'.

In the positivist method, researchers construct hypotheses, indicators and questions in order to ensure that

they will not be caught up in, or tripped up by, preconceptions, which may conflict with the objectivity of the

experimental method. But very soon, ethnographic researchers could not be considered as being protected

from their theoretical or personal biases. They were potentially considered as systematically falling victim to

them. And being integrated into the community they were studying offered no external guarantee that they

would be able to break free of their ideologies, assumptions, or preconceptions. It was considered, very early

on, that researchers’ past experience, personality, and cultural orientations necessarily influence what they

see and how they interpret it. Realizing this, anthropologists were quick to construct a reflexive scientific

approach.

Reflexivity implies reconciling the opposites of involvement and detachment (Elias, 1956; Hughes, 1971).

The intention of constant and rigorous adherence to this principle is to provide a solid basis for research

SAGE

SAGE Research Methods

Page 7 of 20 Doing Ethnography: Ways and Reasons

without eroding its primary quality: the researcher's ability to adapt to the experience and realities of the

people she is observing. Though involvement does imply constant flexibility and adaptability on the part

of ethnographic researchers, they must still be vigilant about maintaining distance, remaining detached.

Involvement is a construction that the researcher needs to analyse and negotiate with actors in connection

with the rules structuring the social system in which they operate. Detachment is a continuous activity and

guides the researcher's choices in all phases of his involvement: determining how to behave, what strategies

to implement, the limits to impose on her action and analysing collected observations in relation to the

roles played. But in her involvement, the researcher may go through phases of intense identification. The

processes underlying that kind of identification then have to be analysed, deciphered, and contextualized

so as to apprehend, if only after the fact, the workings of the action structuring the negotiated relationship

between the researcher and the observed actors and to assess their implications for the study. These last

activities are what is meant by detachment. Detachment thus requires the researcher to describe and critically

question her way of viewing the world and the social conditions in which ‘data’ are produced in the research

act (for developments, see Gobo, Chapter 5, this volume).

How does one start an ethnographic research?

Following reflexivity as a scientific model, ethnography is supposedly an inductive method, which implies

that one is to start observations as openly as possible in order to let analyses emerge from observations

(as opposed to a deductive method which entices the researcher to start by developing strong theoretical-

driven hypotheses to be tested empirically (for a detailed discussion of induction, deduction, abduction, see

Kennedy and Thornberg, Chapter 4, this volume). As nicely stated by Atkinson ‘we create the fields we

study, in collaboration with hosts and informants, through the research acts we enact’ (Atkinson, 2015, p. 26).

However, as also asserted by Atkinson, even if ethnography does suggest that the observer is to let the field

‘speak’ through time, she is also to guide those observations whenever necessary. Indeed, sheer induction

never happens since researchers, as educated human beings, do hold preconceived ideas about the world.

So, one is to follow two main principles when entering his or her ‘field', in order not to get trapped in such a

sheer inductive illusion.

Before starting his field, the ethnographer is to wonder what to observe and how. Since she is far from being

a neutral and fully open-minded human being, what is best here is to build up and define specific ‘theoretical

orientations’ and ‘target-question’ in order to avoid, as much as possible, being influenced by preconceptions

without being aware of them. Whether he is aware of them or not, ‘theoretical orientations’ shape the way

the ethnographer defines his discipline scope, society as a whole, the ways individuals are shaped and how

they participate in shaping society collectively. Ethnography being an inductive method which implies that the

ethnographer is guided by what she observes in order to progressively fabricate sociological results, she does

start conducting her observations as openly as possible, so that theoretical orientations are not to be confused

with a specific school of thought. But theoretical orientations do participate in shaping how the ethnographer

starts observing and it is important that she is aware of them so as to be able, if needed, to transform

SAGE

SAGE Research Methods

Page 8 of 20 Doing Ethnography: Ways and Reasons

them over time. They also enable the ethnographer to define a ‘target-question’ being as adequate as it can

be, right from the beginning. Indeed, the ‘target-question’ is directly linked to the ‘theoretical orientations',

while specifying the topic to be investigated over time, even in quite general terms (the ‘target-question’

is to be enriched, transformed, confirmed throughout the research). Thanks to this ‘target-question’ – for

instance social mobility, illness trajectories, organizational socialization, gender at work, or urban dynamics

– the ethnographer will be able to decide what is to be observed and how. The ethnographer can then more

easily define where to start observations and how to lead them over time. It will guide the ethnographer

in deciding where to observe, how and in which conditions, and can also be transformed over time based

on a first analysis of initial observations (for instance if the target-question appears too narrow-minded or

misconstructed or biased, based on a first analysis of data as detailed below).

If right from the beginning the observer is as open, fluid and available as possible and tries to be as close to

and familiar with the people she observes, ‘theoretical orientations’ and ‘target-questions’ do enable her to be

as efficient as she can be in developing rigorous sociological findings. Glaser and Strauss, while defending

in their 1968 seminal book, ethnography as an inductive method – the ‘grounded theory’ – did define such

guiding principles when illustrating their use of ethnography empirically:

The general problem to be investigated was whether different ideologies did exist among the various

psychiatric professionals. […] Field work was begun first at the private hospital and directed initially

only by the frameworks of ideas known as the ‘sociology of work’ and ‘symbolic interactionism'.

(Glaser and Strauss, 1968, p. 157)

In this example, they not only identify clear ‘theoretical orientations’ – ‘sociology of work’ and ‘symbolic

interactionism’ – but also a well-defined ‘target-question’ – ‘whether different ideologies did exist among the

various psychiatric professionals'. Those choices led them to define one specific field – a psychiatric hospital,

people to be observed, that is, all the people involved in treating psychiatric patients – and a long-term

observation strategy in order to identify ways psychiatric professionals were collectively defining psychiatric

illnesses and negotiating their roles in treating them through time.

As clearly stated by Atkinson:

We do not pull theories and hypotheses out of thin air. We derive them from a variety of sources:

our own prior exposure to phenomena, the work of others (published or otherwise) and our first

engagements with our research field. (Atkinson, 2015, p. 56)

Based on his initial clearly-stated theoretical orientations and target-question, the researcher will decide what

to observe – focusing on one or a few specific settings or following people (on ‘Go-alongs', see Kusenbach,

Chapter 22, this volume); how to do it – as a participant or non-participant observer, covertly or not; and

along which timings – on a regular basis, evenings only, at different moments throughout the year. Studying

freelance musicians who work in several separated places from a gendered perspective (Buscatto, 2007a,

2007b) led me to observe as many different situations as possible – concerts, rehearsals, parties, training

SAGE

SAGE Research Methods

Page 9 of 20 Doing Ethnography: Ways and Reasons

sessions, jam sessions, private moments – as a participant and non-participant observer, while studying

organizational socializations affecting first-level and middle-manager insurance workers led me to focus on

two specific departments over a full year at different moments of their organizational life (Buscatto, 2008).

So researchers need to be aware how they build up their theoretical orientations and target-questions, how

relevant they are, how they influence them in the process of entering the field and decide what to observe

and how. This will help them make the most relevant decisions throughout the full research process.

How does one build up grounded analyses?

If the researcher is to be as open as possible while leading his first observations, he also needs to analyse

them as soon as possible in order to produce the most appropriate working hypotheses or categories to orient

his observational strategies over time. As stated very early on by Glaser and Strauss:

Whether the sociologist as he jointly collects and analyses qualitative data, starts out in a confused

state of noting almost everything he sees because it all seems significant, or whether he starts out

with a more defined purpose, his work quickly leads to the generation of hypotheses. (Glaser and

Strauss, 1968, p. 39)

Those first analyses, which are of course enlightened by initial theoretical orientations and target-question

(see above), will help the ethnographer decide which situations to focus on in the next days and which

actors to include (or not) in her next observations (Charmaz and Bryant, 2016). If the topic is familiar to the

ethnographer (because similar research has been done and published in the past), she may start with what

Giampetro Gobo calls ‘working hypotheses’ (Gobo, 2008; see Gobo, Chapter 5, this volume). If the topic is

new and has not been explored yet or if the researcher wants to renew the current approach to this topic,

she will more probably develop general ‘guiding hypotheses'. Those hypotheses do help the ethnographer to

better define what is to be observed, when and how, but should never constrain the observer completely, so

that alternative hypotheses may be defined if they seem to fit better with what comes out of the analysis of

the first observations conducted in the field. If unexpected observations ask for new concepts or categories

(not to be found in the current literature), one is to develop an abductive reasoning. Abduction is key when

one observes situations which cannot be explained by current concepts, leading the researcher to create new

ones along the way so as to make sense of his field (Agar, 2006; see Kennedy and Thornberg, Chapter 4,

this volume).

But those analyses do not mean anything if they are not based on notes which are to be as detailed and

factual as possible (for detailed explanations and examples, see Wästerfors, Chapter 20, this volume). Notes

may be taken on the spot (visibly if they do not affect observed people since they are accepted behaviours,

such as in concerts, rehearsals, meetings or administrative work) or may be done in hidden places (toilets

for instance) or afterwards (at home, in a café, at school…) when they might affect people's behaviours or

are impossible to do, as in mobile work, illegal settings, or places of leisure. In all cases, one is to take as

SAGE

SAGE Research Methods

Page 10 of 20 Doing Ethnography: Ways and Reasons

detailed and precise notes as possible of observed activities, observed people's characteristics (age, clothes,

status, sex, links with other observed people if any), and observed places (size, configuration, or objects).

Who is interacting with whom? Who is doing what and how? What are the objects or documents involved?

Those are some of the questions those notes try to document. Of course, what is to be observed and reported

highly depends on the target-question and the working hypotheses, which will help the ethnographer fine-

tune them over time (and transform them according to what is observed). The researcher will be able to

compare what can be compared, to distinguish recurring facts from exceptional facts, major actors from minor

ones, to explain similarities and differences in rigorous and documented ways. If the ethnographer is eager to

mention a personal remark (emotions, feelings, working hypotheses, or intuitions, methodological questions,

categorizations…), he would rather indicate it in a way that clearly distinguishes it from factual observed

details (a different colour, a noticeable sign, a large interval).

While observations are piling up, working hypotheses are getting fine-tuned, sociological analyses are getting

specified, the ethnographer may decide to focus on specific situations or moments or people in order to

confirm, enrich, or transform her target-question, her theoretical orientations, or her final analyses. For

instance, after a few weeks observing workers at Hermes, an insurance company, I noticed that middle

managers were experiencing quite a difficult situation and were feeling both humiliated and threatened. I

then decided to observe more middle managers than expected in order to both better understand Hermes

organizational socialization at work (my main target-question), and better grasp the experience of middle

managers in big companies (the literature devoted to this specific question did not seem convincing to

me, which enticed me to add this secondary target-question to my main one). Later on, in opposition to

the literature devoted to middle management at the time, I was able to identify a ‘social psychologization’

phenomenon which was both producing and legitimating middle managers’ difficulties in such companies

(Buscatto, 2008).

The ethnographic process then appears as a circular one, oscillating between initial questions, observations,

hypotheses, and conceptual analyses. First, theoretical orientations and target-questions lead the researcher

to pick specific observational conditions and working hypotheses. Initial observations, once analysed, help the

ethnographer question those orientations, questions, and hypotheses and either confirm or transform them.

They progressively lead to the formulation of first conceptual categories and sociological results. This iterative

and systematic research process, aimed at developing sound sociological explanations, may be labelled as

either an actual version of the grounded theory (Charmaz and Bryant, 2016) or ‘analytic induction’ (for a

discussion of analytic induction as compared to grounded theory, see Hammersley, 2010).

How does one conduct a scientific ethnographic research?

But what does one exactly do when conducting a scientific ethnographic research?

‘A cookbook approach towards collecting data and writing up one's results is rejected here’ (Rosen, 1991,

p. 11). Indeed, there is no single set of formulas or practical guidelines that will suffice to attain the desired

SAGE

SAGE Research Methods

Page 11 of 20 Doing Ethnography: Ways and Reasons

scientific quality or goal. While finely detailed techniques may be developed to help researchers achieve

rigorous, controlled practice, in the reflexive approach those techniques cannot be seen as mere procedures

to be followed, standards to evaluate or rules to follow and check off:

What we need to do now is to show the ways in which quality research can be every bit as credible

as the best quantitative work. Part of that will involve recognizing that good quality research depends

upon craft skills that ultimately transcend the kind of lists of factors that we have been reviewing in

this chapter. (Silverman, 2006, p. 311)

How is one to proceed then?

There are many techniques that are useful in adopting a reflexive approach in ethnography. They include,

for instance, ‘saturation’ (see below), ‘triangulation’ (see Flick, Chapter 34, this volume), ‘counting’ and

‘measuring’ (see Buscatto, 2016; and Gobo, Chapter 5, this volume), analysis of the social conditions in which

‘data’ are produced and analysed (see below), collective research, combining comparable cases (see below,

as well as Sørensen et al., Chapter 10, this volume about comparison), note-taking technique (see above and

see Wästerfors, Chapter 20, this volume), or sharing and ‘discussing’ research findings with observed actors

(see Buscatto, 2008). To convey how reflexive approaches are effectively practised, I will here discuss two

specific major techniques in detail: the first, comparison and the second, the analysis of social conditions in

which ‘data’ are produced and analysed.

Using comparison throughout the scientific process

Comparative analysis is an ancient scientific practice, considered as key in developing reliable scientific

results. For instance, very early on, Emile Durkheim asserts that ‘la méthode comparative est la seule qui

convienne à la sociologie’ (Durkheim, 1986 [1895], p. 15).4 Indeed, comparative analysis does enable the

researcher to step from a specific case study to a higher level of generalization (conceptualizing, theorizing, or

categorizing) in controlled ways. For Everett Hughes, ‘the essence of the comparative frame is that one seeks

differences in terms of dimensions to all the cases’ (Hughes, 1971, p. 420). By connecting several specific

observed groups or settings with each other, the researcher is able to unveil more conceptual explanations

in rigorous ways. For Glaser and Strauss ‘the constant comparing of many groups draws the sociologist's

attention to their many similarities and differences. Considering this leads him to generate abstract categories

and their properties, which, since they emerge from the data, will clearly be important to a theory explaining

the kind of behaviour under observation’ (Glaser and Strauss, 1968, p. 36).

But the ways groups or settings are to be compared fully depend on the research question. When he attempts

to grasp the influence of state policies on the production of specific ‘Factory regimes', the ethnographer

Michael Burawoy compares companies with similar activities and organizations evolving in different countries

(Burawoy, 1985). When I attempted to understand how organizational socializations were shaped within

SAGE

SAGE Research Methods

Page 12 of 20 Doing Ethnography: Ways and Reasons

companies, I chose to compare two French insurance companies whose size, activities, and formal rules

were quite similar, so that specific organizational socializations built-up by workers, unions and employers

could be unveiled (Buscatto, 2008).

The researcher will have to decide how different or similar the groups or settings being compared should

be. ‘Minimizing differences among comparison groups also helps establish a definite set of conditions under

which a category exists’ (Glaser and Strauss, 1968, p. 56). It is best when one is studying an exploratory

topic and wants to focus on a limited set of interpretive possibilities. This idea guided my choice of two French

insurance companies sharing several traits (size, turnover, business objectives, types of customers) and of

two similar departments dealing with the same activity in both companies, so that my study would focus

on organizational socialization only. On the other hand, ‘when maximizing differences among comparative

groups (thereby maximizing differences in data) he possesses a more powerful means for stimulating the

generation of theoretical properties once his basic framework has emerged’ (Glaser and Strauss, 1968, p.

57). In this same study, this idea led me to observe very different types of workers (by sex, age, work, or time

spent in the company) and diverse working units within both departments, so that organizational socialization

could be defined thoroughly despite all individual differences and work units.

Comparison is not a natural reality, it is constructed by the ethnographer, based on her research question

and working hypotheses. But since ethnographic research is not a chemistry experiment, the analysis is to be

as rigorous as possible to avoid over-interpretation or mis-interpretation since compared groups, settings, or

traits, even if precisely defined, are never fully similar.

Analysing the social conditions under which a research study is

conducted

Another key technique to ensure the quality of one's research is to analyse the social conditions under which

‘data’ and data analysis are produced. As stated by Sudhir Venkatesh, reflecting on his ethnographic study of

a poor public housing development in Chicago in the 1990s:

Reconstructing the informants’ point of view – in this case their perceptions of the fieldworker as,

variously, academic hustler, ‘nigger just like us’ and ‘Arab’ – can aid the researcher in determining

patterns of structure and meaning among the individual, group, and/or community under study.

(Venkatesh, 2002, p. 91)

For reflexive sociologists, the ethnographic study and its interpretation are founded on the relation between

the researcher(s) and the observed actors. The ways in which an ethnographer is allowed to observe people

through time (or not), as well as ways she may interpret those observations, are indeed widely dependent on

reciprocal social positions or roles negotiated (or not) between the observer and observed.

In reflexive sociological practice, the social conditions for producing empirical material are thus a fundamental

component of the analysis when it comes to checking which data you produce and constructing a findings-

SAGE

SAGE Research Methods

Page 13 of 20 Doing Ethnography: Ways and Reasons

based analysis. Ethnographers have identified several stages in the process in which it is helpful to analyse

the observed–observer relationship: constructing the observed population; opening doors in the field;

maintaining oneself in the field; determining how the researcher–observed actors’ relationship is produced;

and choosing how to analyse the collected data. The process unfolds over time and, as explained, it cannot

follow predefined rules.

The ethnographer is thus to try out social positions or roles in the field in order to open closed doors, elicit

invitations, circumvent prefabricated discourse, gain access to the unsaid or taboo, and interpret rationally

what she has observed once the study is over. Those social positions are both defined by the field's

possibilities and by the researcher's specific characteristics and strategies. For instance, if an ethnographer is

to observe an organization, one may try to develop different social positions in the field – as an expert, as an

academic, as a friend, as a worker, as a trainee – depending on what the ethnographer is observing, when,

and how (Buscatto, 2008).

But social positions are also defined by the ethnographer's social characteristics as perceived by the

observed actors – age, professional status, physical appearance, sex, or prior relationships with observed

actors. People will act differently with the ethnographer depending on his perceived age, sex, social position,

or physical attributes, in relation to the way those characteristics are defined and interpreted by people in the

field. It is key for the ethnographer to become aware of perceived positions throughout his study, day after

day, and to assess how they affect what is observed and what it tells him about social relationships within the

field. Moreover, those social positions and social characteristics may also influence the way one is to define

his research question.

For instance, being a woman in a very masculine world may entice one to study gendered relations in

this social world due to the repeated gendered experiences the researcher is going through (Buscatto,

2007b), while being a dance musician (Becker, 1963) or a plant worker (Roy, 1952, see above) may lead the

researcher to study ways dance musicians or plant workers define work on a daily basis. It not only makes

one aware of those questions, but also opens up opportunities to do so by allowing one to observe several

situations that enrich the analysis. While social relationships that are developed with the actors are constantly

being constructed and reconstructed in a chaotic, experimental way, the researcher is also analysing it in

depth so as to discover the best way to observe what needs to be observed, and interpret the experienced

situations, whatever they happen to be.

How is one to decide when to stop collecting ethnographic data?

While defining one's field, based on her theoretical orientations and target-question, the researcher is able to

guide her observational strategies as appropriately as possible and may even transform them over time based

on the analysis of ongoing observations. She is progressively building up categories, or even concepts, which

enable her to present sound and rigorous scientific results. But how does one assess she is finished with

observations? How does one decide that the analysis is well grounded and the observation period is over?

SAGE

SAGE Research Methods

Page 14 of 20 Doing Ethnography: Ways and Reasons

One analysis technique to be used here is what Glaser and Strauss have called ‘saturation’ (see Charmaz

and Bryant, 2016 as well as Schreier, Chapter 6, this volume).

While the researcher, based on his observations, has progressively shaped categories, or even concepts

throughout the study, which enable him to answer the ‘target-question’ in adequate terms, he is experiencing

at some point not learning anything ‘new’ regarding those preliminary analyses when observing new

situations and people. One is then to implement a few more observations in order to repeatedly experience

this feeling of ‘saturation’ thoroughly and make sure it is not linked to boredom or over-interpretation. The

ethnographer may consider this final phase as a perfect phase to test all working hypotheses, categories, or

concepts in quite confortable ways. Most of the time, those first sociological conclusions are not yet formally

expressed in academic terms, but they are sufficiently well-defined in sociological terms to be tested in the

field.

For instance, while I was studying jazz singers from a gendered perspective, I did experience the saturation

phenomenon in the last stages of my field observation, before starting to write the first scientific article

presenting my findings in academic terms. I had not yet defined their situation as a gendered marginalization

and had also not clearly defined the three key processes that were producing and legitimating this situation

in the jazz world. But, thanks to my ongoing analysis of field observations and formal interviews, I had

already assessed the gendered construction of their professional reputation and of musical relations shaped

by female singers and male instrumentalists through interaction. The last weeks of observation made me

feel ‘saturation’ and I decided to stop observing the jazz world in order to focus on writing this first academic

article. But I was ready to renew observations if a thorough analysis had showed contradictions or conflicting

interpretations which could not be settled based on collected data, which did not happen, as is the case if the

saturation phenomenon is based on a systematic and rigorous analysis of observations aimed at answering

one's target-question. But I did start new lengthy observations of the jazz world a few years later when I

developed a new target-question – why are there so few female instrumentalists? This new target-question

could not be answered based on my former observations, which had mainly focused on jazz singers, even if

some data did already help me in doing so. I could then not only confirm what had already been demonstrated

in this first article on female jazz singers, but also describe and explain how the marginalization of female jazz

instrumentalists was daily produced and legitimated in the French jazz world (Buscatto, 2007a, 2007b).

So saturation is to be understood in relation to a specific target-question and is possible only if one is

analysing his observations on a daily basis, enlightened by his explicit theoretical orientations and working

hypotheses. Of course, saturation may be impeded either by lack of time (results are expected before this

phenomenon is experienced thoroughly), or because the field is not available anymore (at the end of a

contract or if people refuse to be observed any longer for instance). The researcher may then have to

redefine the target-question, or reduce the scope or express the limits of his analysis, in order to ensure that

sociological results which are presented in a written form do answer scientific expectations.

Ethnography and its potential limits

SAGE

SAGE Research Methods

Page 15 of 20 Doing Ethnography: Ways and Reasons

Ethnography is often denigrated for its limited access to social phenomena and its difficulty in assessing the

quality of results with rigor due both to the ethnographer's strong involvement in the field and the limited scope

of this field. But those potential limits have long been discussed by ethnographers to show that ethnography

may not only produce high-quality results, but also open the road to strong generalizations. Indeed, due to

her long-term involvement in the field, the researcher has many opportunities to test her conclusions over

time (saturation being one of them as shown above). Moreover, he may use observed people's feedback

to improve his conclusions over time (Buscatto, 2008). Regarding generalizations, several ethnographers

have demonstrated how ethnographic studies have built-up new strong concepts such as ‘total institutions’

(Goffman, 1961), ‘moral entrepreneurs’ (Becker, 1963) or ‘factory regimes’ (Burawoy, 1985) which enabled

researchers using questionnaires or interviews to better analyse social phenomena afterwards.

Ethnography is also often criticized for its potential ethical limits since observed people are often very close

to the observer and give her access to the most personal parts of their lives. But, as stated by Atkinson

(2015), ethnography appears as quite an ethical method in that it not only implies a strong commitment of

the researcher towards her observed people, but also tends to give the most accurate account of what is

going on in the observed field. However, it does not prevent the researcher from answering some tricky

ethical questions: how does one get people to consent to the observations they are enduring? How does one

ensure that observed people do not get hurt along the way? Regarding the issue of consent, the majority of

the time the situation is quite difficult to deal with. If one may easily ask consent of people observed on a

recurring basis (the chore of most observations such as nurses in a hospital, street-people or employees in an

insurance company), he will most of the time observe people who cannot be informed of his position without

disturbing the observed activity (such as customers, patients, passers-by, temporary workers, the audience,

etc.). Therefore, one has to admit that, if informed consent should be definitely obtained from people who are

regularly observed, it may not be required from occasional observed people who are just passing through.

This leads us to what we consider as the most important issue in ethnography: how to make sure people

do not get hurt when results are published? Published results may affect people in two main ways. On the

one side, if one is to be recognized by colleagues, friends, or family, his or her reputation may be denigrated,

even if it was not intended (for such a case see Whyte, 1943). The solution here is to ensure each individual's

anonymity, not only by erasing all real names, but also by making sure that the people who know those

persons can trace neither example nor portrait back to one specific individual. In observed worlds where

people know each other well – such as the jazz world, an organization or an orchestra – the ethnographer is

to give few details or to present composed portraits – mixing elements drawn from different observed people

– so that observed people cannot be recognized by their colleagues, their friends, their family. On the other

hand, the unveiling of social dynamics may appear to be giving a bad image of specific environments (as

sexist, racist, or corrupted for instance). Here, the researcher is to describe social dynamics as rigorously as

possible, and to present precise and neutral examples, which may never be connected to one or the other

person involved. This will not avoid some observed people being unhappy with the ways their social world

is described from a sociological perspective, but it will cancel the negative consequences those published

results may have on individuals (including their self-esteem). Our final advice here would be to present

SAGE

SAGE Research Methods

Page 16 of 20 Doing Ethnography: Ways and Reasons

intermediary results to observed people as often as possible (see Buscatto, 2008 and 2016). It will not only

help the researcher avoid misinterpretations and enrich his final scientific results, but it will also give observed

people the opportunity to get accustomed to sociological reasoning, present their points of view (even if

disagreeing) and get the researcher to write in the most distanced and neutral style to avoid as many negative

consequences as possible.

Conclusion

As demonstrated throughout this chapter, if ethnography is a quite powerful and insightful method, which has

historically proven its scientific efficiency, it also requires a strong personal involvement of the ethnographer

in the field to be maintained over a long period of time. Ethnography appears as a quite time-consuming

method, which one is to adopt if one is ready to involve oneself, both intellectually and personally. But

the involvement is worth the effort, even if not always academically rewarded, since results are often quite

powerful in unveiling how observed worlds function in the long run.

But ethnography is also evolving quite radically in relation to several changes affecting contemporary

societies. On the one hand, since more and more people spend time playing, exchanging ideas, developing

relationships or sourcing information through the Internet (see the chapters in Part VI, of this volume),

ethnographers are developing new ways to study virtual life. The question is, then, to assess if virtual

ethnography is about observing people interactions with each other or about reading documents produced by

people whose identities are not known (see the section dedicated to this issue in this volume). On the other

hand, new techniques such as video or sound data are more and more often used by ethnographers to enrich

their analysis (for video data see Knoblauch et al., Chapter 23, this volume, and for sound data see Bull,

Chapter 27, this volume). And how those different data are to be articulated is to be worked out thoroughly

(see Hesse-Biber, Chapter 35, this volume).

The future of ethnography thus relies on its ability to answer those new empirical and epistemological

challenges in reflexive and rigorous ways, as fully demonstrated by several chapters in this volume.

Notes

1. If Taylor or Mayo are hardly considered as ethnographers by contemporary sociologists, they did pave the

way to the use of ethnography in sociology.

2. This discussion on how ethnography helps better grasp gendered phenomena is not to be confused

with (otherwise insightful) discussions on either how women (as opposed to men) may develop specific

ethnographies (for instance due to their privileged access to specific fields or phenomena, as thoroughly

discussed below and in Buscatto, 2007b), or on the construction of a ‘feminist ethnography’ led by women

only and dedicated to the study of women (due to the fact that, on the one hand, women tended to be

SAGE

SAGE Research Methods

Page 17 of 20 Doing Ethnography: Ways and Reasons

http://methods.sagepub.com/book/the-sage-handbook-of-qualitative-data-collection##i2314

http://methods.sagepub.com/book/the-sage-handbook-of-qualitative-data-collection##i2319

underrepresented among sociologists, and still are at the highest academic levels, and, on the other, women

have been understudied in sociology, up until at least the early 2000s).

3. All ideas briefly presented here are thoroughly developed and illustrated in a chapter about reflexivity in

ethnography (Buscatto, 2016).

4. ‘The comparative method is the only one which suits sociology’ (our translation). Durkheim devotes a full

chapter to this question in the mentioned book.