(Studies in Educational Ethnography 12) Geoffrey Walford - Methodological Developments in Ethnography-Emerald Group (2007)

PREFACE
Ethnography has become one of the major methods of researching educa-

tional settings. Its key strength is its emphasis on understanding the per-
ceptions and cultures of the people and organisations studied. Through
prolonged involvement with those who are being studied, the ethnographic
researcher is able gradually to enter their world and gain an understanding
of their lives.
Each volume of Studies in Educational Ethnography focuses on a partic-
ular theme relating to the ethnographic investigation of education. Most of
the volumes have been closely linked to an annual 2-day residential con-
ference which explores various elements of ethnography and its application
to education and schooling. The series of Ethnography and Education con-
ferences began in the late 1970s, and was originally held at St. Hilda’s
College, Oxford University. The series later moved to Warwick University,
back to the Department of Educational Studies, University of Oxford in
1996, and finally returning to St. Hilda’s College in 2004.
Unusually, this present volume collects together papers that were specially
commissioned rather than being presented at this annual conference. They
are selected on the basis of their high quality, their coherence as a group and
their contribution they make to debates about ethnographic methodology.
They do not present one coherent view of how ethnography should be
conducted but, rather, present a range of sometimes competing arguments
and developments.
The series recognises that the nature of ethnography is contested, and this
is taken to be a sign of its strength and vitality. While the idea that the term
can be taken to be almost synonymous with qualitative research is rejected,
chapters are included that draw upon a broad range of methodologies that
are embedded within a long and detailed engagement with those people and
organisations studied.
Geoffrey Walford
vii
INTRODUCTION
What counts as ethnography and what counts as good ethnographic meth-

odology are both highly contested. This volume brings together chapters pre-
senting a diversity of views on some of the current debates and developments
in ethnographic methodology. It does not try to present a single coherent view
but, through its heterogeneity, illustrates the strength and impact of debate.
In the first chapter, Amar Dhand addresses the problem of access from a
learning theory perspective. It suggests that Lave and Wenger’s ‘legitimate
peripheral participation’ is a suitable lens to understand the interpersonal
dynamics when a researcher enters a site and attempts to gain an inside
perspective. Basing his discussion on his own research of a group of heroin
addicts in India, he proposes that on first arrival the ethnographer is as-
signed to designated roles that are salient to participants. Subsequently,
through a negotiation of identity meanings, the ethnographer and partic-
ipants create ‘new’ roles that are legitimatized to the group. Finally, with
key informants, the ethnographer and informant engage in mutual educative
processes where each can ‘read’ and teach the other. While he acknowledges
the limitations of applying the concept of legitimate peripheral participation
to questions of access, he argues that the theory offers a unique insight to
processes that are at the core of the ethnographic experience.
Harry Wolcott’s short chapter delves deeper into one of the commonly
accepted attributes of ethnography – the criterion of ‘intimate, long-term
acquaintance’. Basing his discussion on his well-known experience and ac-
count of ‘the sneaky kid’, he asks: ‘In ethnographic fieldwork, just how
intimate is intimate?’
In the next chapter, Susan James argues that participant observation as
a data collection technique provides for a rich encounter and that the po-
tential to meet people and learn about their lives and experiences knows no
bounds. This chapter focuses on the experiences of a researcher working as
an apprentice in three different kitchens for a doctoral study. The decision
to be a participant observer rather than a nonparticipant observer when
researching learning in the workplace is discussed and the dichotomy of
researcher and worker is highlighted. Anecdotes are provided to explain the
experiences of participant observation in terms of identity, language,
ix
x INTRODUCTION
uniform, acceptance, and initiation procedures while conducting research on

learning in the workplace.
Nick Hopwood’s chapter explores the different roles researchers can
adopt in their fieldwork, making specific reference to a school-based eth-
nographic study. The concept of territories is introduced, demonstrating
that research sites may comprise a number spatially, temporally, and/or
socially defined settings. He argues that different roles may be appropriate
in these various settings; ethnographic research will be strengthened by an
understanding of the complex combinations of and flux between roles that
may occur within short periods of time in a single research site.
Martin Forsey’s chapter explores the tendency for critical educational
ethnographers to produce accounts of schooling in which teachers either
disappear or are represented as mere cardboard cutout figures. There are
three main aims for this chapter: (1) explaining the relative lack of attention
to teachers in criticalist accounts of schooling; (2) exploring the implications
of these oversights for qualitative research in schools; and (3) suggesting
ways in which the researcher’s mindset might shift to address the issues
arising from our currently ‘averted gaze’. Advocating a mindset attuned to
the critical appreciation of social life, Forsey suggest a need to heighten the
importance of clear and precise documentation of the power-filled social
relations characterizing educational settings as an important element of the
criticalist project. Further to this, he argues for the need to position our-
selves in ways that allow us to better see, feel, and evaluate some of the
multitude of interactions that take place in schools and their surrounds.
The chapter by Sue Walters sets out the differences that she sees between
‘ethnography’ and ‘case study’ and the reasons for choosing to do case studies
using ‘ethnographic methods’ for a recent research study. After considering
the bewildering number of ways in which ‘ethnography’ and ‘case study’ have
been defined and differentiated and how these differing definitions were re-
solved for the purposes of describing and reporting her own research, she
explores what ethnographic methods can bring to a case study and why doing
case studies of particular pupils rather than of particular classrooms is an
important research decision. She then discuss how a single ethnographic case
study (or a set of single ethnographic case studies) can be of use outside of the
specific micro settings of their production, thus addressing concerns about the
generalizability and usefulness of ethnographic case study research.
The next chapter by Eric Tucker puts forward a new role for ethnography
in social and educational research. He argues that abstract measurement
is at the center of much educational research, from tests, to observa-
tional scales, to questionnaires. Contemporary psychometrics and related
Introduction xi
statistical techniques are the preeminent methodological approaches with

regard to the development of measurement instruments. Yet, in recent years,
the variety and intensity of the demands social scientists and policy makers
place on indicators have grown dramatically. Furthermore, researchers are
beginning to describe instances where there seems to exist a substantial gap
between established measurement instruments and the social phenomena
they intend to measure. This article contends that the most prominent ex-
isting procedures to secure valid indicators tend to overemphasize verifica-
tion and correspondingly de-emphasize empirical research that aims toward
generation and modification. It proceeds to make the case that the features
and strengths of ethnography specifically, and qualitative research more
generally, make it uniquely suited to contribute to the development of new
indicators and the improvement of existing indicators. To provide some
warrants for these arguments, this paper first offers a brief reading of the
state of the practice of social science empirical research on measurement. It
then turns to a specific example of this author’s ethnographic research
project that aims to improve indicators of social capital formation, in the
hopes of providing an exemplary instance of how ethnographic methodol-
ogies might contribute to the generation and modification of new, quality
measurement instruments.
Kristen H. Perry’s chapter raises many important questions about the
ways in which ethical Institutional Review Boards control what and how it
is possible to research. She focuses on the particular problem of anonymity,
which is seen as the default option by most Institutional Review Boards, yet
is highly problematic in practice. Not only are there many difficulties in
maintaining anonymity in an increasingly electronic information age, but
also some participants actually want to have their names included in the
research reports. She argues that the model of research enshrined within
Institutional Review Boards’ beliefs and procedures is outdated and inap-
propriate for much qualitative research. She draws on her own research
among Southern Sudanese refugees to illustrate her argument.
Geoffrey Walford’s chapter returns to the old problem of generalization and
ethnographic studies. He argues that the overgeneralization of ethnographic
case studies has been to disadvantage of the development of ethnography as a
methodology and field of study within education, and that ethnographers
should avoid attempts to generalize. He describes and discusses some of his
own studies and argues that their utility and merit (to the extent that they
might have both) depends on their uniqueness rather than their wider appli-
cability. He further argues that authors should name their research sites
wherever possible to try to ensure that readers do not overgeneralize.
xii INTRODUCTION
The chapter by Dalia Aralas focuses on the need for the ethnographic eye to
broaden its field of vision in grappling with particular issues in some kinds of
qualitative research. Specifically, Aralas argues for the considerable value of an
approach that differs from but complements conventional video ethnographic
approaches. To illustrate, she discusses a study which adopted a different video
ethnographic approach to pursue its objective of investigating the nature of
mathematical imagination. This made possible the generation of different
kinds of data that, in turn, facilitated an intensive interrogation of the initial
framework of the study. It is argued that the resultant data analysis opened up
opportunities for a continuous reconstruction of a more layered alternative
framework that led to a reconfiguration of the phenomenon under study. She
argues that while extending video approaches might be seen to partially un-
dermine prevailing views, the consequent production of alternative accounts
and frameworks make worthwhile contributions to our understanding.
Gretar Marinósson’s chapter is unusual in that it is based upon an eth-
nographic study of one school conducted by an educational psychologist.
Marinósson reflects on these different ways of doing educational research
and describes his own procedures. He follows a grounded theory approach
and argues that it is possible to unearth the ground rules of education
through ethnography using methods of data generation based on a recip-
rocal relationship between the whole and its parts. The behavior of indi-
viduals thus bears witness to these ground rules and the institution’s
structure and classification activity. In this way it might be said that the
drop merges into the ocean and vice versa.
Heewon Chang’s chapter focuses on how to conduct rigorous autoeth-
nographies. Autoethnography is an ethnographic inquiry that utilizes the
autobiographic materials of the researcher as the primary data. Chang
is keen to emphasize that autoethnography is different from other self-
narrative writings such as autobiography and memoir, in that it is based on
cultural analysis and interpretation of the researcher’s behaviors, thoughts,
and experiences in relation to others in society. She argues that autoeth-
nography should be ethnographical in its methodological orientation, cul-
tural in its interpretive orientation, and autobiographical in its content
orientation. She discusses the definition of this inquiry method, methodol-
ogy, and benefits of autoethnography as well as some of the pitfalls to avoid
when doing autoethnography.
Geoffrey Walford
Editor
USING LEARNING THEORY TO
UNDERSTAND ACCESS IN
ETHNOGRAPHIC RESEARCH
Amar Dhand
INTRODUCTION
This paper addresses the problem of access in ethnographic research from

a learning theory perspective. It extends a recent symbolic interactionist
approach to the problem (Harrington, 2003) by conceptualizing access as a
process of ‘legitimate peripheral participation’, broadly understood as the
processes that enable ‘newcomers’ to become part of the sociocultural
practices of a community (Lave & Wenger, 1991). I present evidence from
my journey of gaining access to three social structures of a group of heroin
addicts in India: a non-governmental organization (NGO), a small group
of ‘brothers’, and a friendship with a key informant. Using this evidence, I
argue that the ethnographer negotiates identity roles, acquires an under-
standing of the ‘rules’ of interaction, and engages in educative processes that
make him or her a legitimate peripheral participant.
Harrington (2003) defines access as ‘the process by which researchers
gather data via interpersonal relationships with participants’ (p. 593). Her
review of the methodology literature reveals a fragmented understanding
of the processes, and she attempts to integrate the formulations with a sym-
bolic interactionist framework. With particular emphasis on social identity
Methodological Developments in Ethnography

Studies in Educational Ethnography, Volume 12, 1–25
Copyright r 2007 by Elsevier Ltd.
All rights of reproduction in any form reserved
ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12001-X
1
2 AMAR DHAND
and self-presentation theories, she reveals four relationship patterns in the

access process: (1) Ethnographers are initially defined according to categories
salient to participants; (2) Ethnographers attempt to share and enhance
similarities; (3) Participants must validate the ethnographers’ identity claims;
(4) The extent of difference between the identities of the ethnographers and
participants determines the need for inside informants or covert strategies.
Lave and Wenger’s (1991) thinking is useful because it conceives changes
in Harrington’s ‘interpersonal relationships’ as a social learning process.
They suggest that social relations change during direct involvement in
activities (p. 94). Moreover, embedded within these activities and its con-
stituent relationships is the ‘place of knowledge’ or the ‘learning curriculum’
that is only mastered through participation in the activities and engagement
in the relationships. This leads them to the assertion that ‘learning occurs
through centripetal participation in the learning curriculum of the ambient
community’ (p. 100). Such a notion of knowledge acquisition through par-
ticipation is helpful because it provides insight to the acquisition process
(‘access’) and the knowledge acquired (‘data’) by ethnographers as they
become participant observers. The ethnographer’s journey and knowledge
has a relationship to the journey and knowledge of a ‘newcomer’ or ‘ap-
prentice’ who is learning to participate:
From a broadly peripheral perspective, apprentices gradually assemble a general idea of
what constitutes the practice of the community. This uneven sketch of the enterprise
(available if there is legitimate access) might include who is involved; what they do; what
everyday life is like; how masters talk, walk, work, and generally conduct their lives; how
people who are not part of the community of practice interact with it; what other
learners are doing; and what learners need to learn to become full practitioners. It
includes an increasing understanding of how, when, and about what old-timers collabo-
rate, collude, and collide, and what they enjoy, dislike, respect, and admire (p. 95).
As is indicated by the qualification ‘(available if there is legitimate access)’,

Lave and Wenger are also interested in access, but from a different per-
spective. As one of their central issues (pp. 100–105), they explore access in
terms of transparency of the community’s activities and sequestration of the
‘newcomer’. In their analysis, the focus is on the community enabling or
limiting access; the ‘newcomer’ negotiating or creating opportunities for
access is relatively unexplored. Such inherent ‘barrier’ mechanisms are rele-
vant to ethnographic access, but this paper focuses on the ethnographer’s
negotiation of those ‘barriers’.
Other methodology literature has not been unmindful of this direction of
thinking. ‘Apprenticeship as a field method’ (Coy, 1989; Downey, 2005) and
Using Learning Theory to Understand Access in Ethnographic Research 3
‘learning the ropes’ (Shaffir & Stebbins, 1991) have received some attention.
Murray Wax (1967) is particularly articulate about such ideas:
y the student begins ‘outside’ the interaction, confronting behaviors he finds bewilder-
ing and inexplicable: the actors are oriented to a world of meanings that the observer
does not grasp. Thus, with a strange language, at first all is mumbo-jumbo, then patterns
begin to emerge, and the student proceeds from halting discourse (stumbling over new
phonemes and translating inwardly from his mother tongue) to the fluency of one in-
creasingly at home in the language and situation y gradually he comes to be able to
categorize peoples (or relationships) and events y (p. 325).
However, the apprenticeship literature tends to focus on the ethnogra-

pher taking the role of an ‘apprentice’ usually in an organized situation
(weaving, shamanism, dance classes, etc.). It does not consider a more
diverse set of roles in informal, non-organized contexts. Moreover, the
‘learning the ropes’ accounts, including Wax’s comments, usually do not
utilize learning theories and remain more anecdotal, as Harrington has also
identified.
Although primarily focused on access, this paper also offers a unique
commentary on ‘legitimate peripheral participation’. Ethnographers are
distinctive ‘newcomers’ because they are more cognizant and reflexive of
their journey. Therefore, they chronicle the experience of ‘centripetal par-
ticipation’ more directly. In this way, my journey has illuminated some
points for consideration in Lave and Wenger’s theory. First, the theory may
be strengthened by exploring the robust reciprocal effects of the researcher
or ‘newcomer’ on the current ‘old-timers’, a suggestion recently supported in
the literature (Davies, 2005; Fuller, Hodkinson, Hodkinson, & Unwin,
2005). Second, the theory should be considerate of the diversity of journeys
that ‘newcomers’ may experience. For instance, the journey of the researcher
to become a ‘full participant’ (and not a ‘complete participant’ as under-
stood in the methodology literature) is quite different from the journey of
other ‘newcomers’ in this social context.
This paper first provides some contextual details of the general field
methods, setting, and participants. It then describes my journey of gaining
access through exploring relationships with NGO staff members, ‘brothers’
from a small group, and a key informant. Finally, I discuss some of the
insight gleaned from these data in relation to Harrington’s suggested rela-
tionship patterns, and the suitability of implicating Lave and Wenger’s
learning theories to those patterns.
4 AMAR DHAND
GENERAL FIELD METHODS
This paper derives from a seven-month ethnography involving three fieldwork

periods over a span of 12 months. The ethnography explored organized
(Dhand, 2006b) and naturalistic peer learning patterns (Dhand, 2006a) among
a group of heroin addicts in Yamuna Bazaar, New Delhi. I choose to name
the site to prevent results from being falsely generalized (Walford, 2002). The
principal methods were participant observation and semi-structured inter-
views. All fieldwork was conducted in Hindi and English without an inter-
preter. A native Hindi speaker transliterated and transcribed all Hindi tape
recordings into Roman Hindi. A different native Hindi speaker was consulted
in translating quotations for this paper.
Analysis for this paper was aided by the use of a qualitative software
program called TamsAnalyzer (Weinstein, 2004). Sections from fieldnotes
and transcripts were first selected and coded with the code methodology.
Then, segments relating to inter-relationships important for data generation
were recoded as methodology > access. These data were then organized into
three groups representing relationships with different social structures in the
group and recoded as methodology > access > NGO, methodology > access >
brothers, and methodology > access > Baba. These categories were then
analyzed manually for more detailed discrimination and finer analysis of
emergent themes. Items constituting similar themes were then placed to-
gether and compared using the cut and paste function of a word processor.
This led to a consolidation of the final thematic structure in each of the three
major categories.
SETTING
Yamuna Bazaar was a lively market area in Old Delhi, centred upon mostly
Hindu religious activities. The area was home to one of India’s most sacred
temples (Hanuman Mandir) and its largest crematorium (Nigam Bodh Ghat).
Believed to be referenced in the 300 BC epic poem Mahabharata, both places
were located on the sacred Yamuna River and had deep cultural meanings
of piety, charity, and the end of life. This religious milieu defined many
of the social activities including parading dead bodies down the main road
for cremation, giving food and money to the poor after prayer routines in
the temples, and conducting auspicious rituals on the Yamuna River.
The ancient area, however, was also undergoing modern development.
The construction of the Kashmiri Gate metro station and bridge intersecting
main commuter lines had caused widespread urban advancements. The

Yamuna Pusta, a large illegal colony known for its high criminal and drug
activity, had recently been destroyed with all residents forcefully relocated.
Roads were being remade, parks were being cleaned up, and local police
were more strictly displacing homeless people from the footpaths.
For heroin addicts, the clash of ancient religious elements with modern
development dynamics offered both support and brutality. They were aided
by free food, shelter options, and health services from a multitude of
religious and non-profit organizations. Conversely, they were often the vic-
tims of high criminal activity, police abuse, and stigmatization. Survival,
therefore, required learning how to take advantage of the supportive servi-
ces without falling prey to constant threats and hardships.
PARTICIPANTS
The participants were a group of active, rehabilitating, and rehabilitated

heroin addicts associated to the SHARAN NGO in Yamuna Bazaar.
Within this group, there was an internal structure of NGO staff members,
small ‘brotherhoods’ of active heroin addicts, and dyad relationships be-
tween addicts. This paper describes gaining access to each of these internal
structures by exploring the interpersonal relationships with approximately
50 NGO staff members, a 6-member ‘brotherhood’, and 1 key informant.
The NGO staff members constituted most of the rehabilitating and re-
habilitated addicts who were formerly street users in the area. The demo-
graphics of the NGO staff were, therefore, similar to the active addicts in
terms of the following characteristics: mostly male, mean age of 30-years old
(ranging from 15 to 60 years old), 3:1 Hindu to Muslim ratio, mostly mi-
grated from neighbouring villages but have lived in Delhi for approximately
5 years, live temporarily in local shelters or shared quarters, and mostly
unmarried and estranged from family. The active addicts were different
from the staff in terms of being regular users, usually homeless, working as
trash/paper collectors, rickshaw drivers, or odd job helpers, involved in both
organized and random stealing, and commonly in and out of jail.
Most of the participants were experienced poly-drug addicts who had a
history of using substances for at least five years. The two major heroin-
based drugs were (1) brown sugar/smack, an unrefined powdered form
smoked on tin-foil and available from street dealers, and (2) injection, an
intravenously administered cocktail of drugs labelled ‘norphine’, ‘avil’, and
‘diazepam’ (opioid alkaloid, anti-histamine, and CNS depressant), and
6 AMAR DHAND
available at local pharmacies. A using addict would not utilize both con-
currently, but would readily switch based on cost. Most using addicts
‘chased’ or ‘fixed’ at least two times per day, usually at the expense of food,
shelter, and clothing. The more ‘recreational’ substances used concurrently
or during intervals in between major drugs were solotion (solvents), ganja
(marijuana), charas (chewing marijuana), nitrazepam tablets (CNS depres-
sant), and daaru (alcohol), although this was not a complete list.
PLAYING A PART IN THE NGO
The process of accessing this group of heroin addicts began by establishing a

role in the NGO that provided services for the participants. In this section, I
describe the evolution of identities that I performed from ‘international
guest’ to ‘NGO worker’ to ‘interviewer’ and ‘observer’. Each transition in
this process was marked by disagreement with my performance and the
role’s associated expectations. This dissonance led to a re-negotiation of
my identity in which individuals legitimized my new position. Interestingly,
in the course of co-constructing and legitimizing the ‘interviewer’ and
‘observer’ roles, participants began to enact their renditions of these roles
and discussed their ‘findings’ with me. Although a key stage in gaining
access, the NGO-associated roles also limited access to relationships that
were important to my research questions. Therefore, over time, I began with
transition into roles with other group structures.
During my pilot study in March 2004, I navigated the network of NGOs
in Delhi to learn about target populations and the offered services. I
searched the Internet and directories, contacted leaders through e-mail, and
visited various centres and events. At one of the events sponsored by POND
(Partnership of NGOs in Delhi), I met Derek, one of the leaders of
SHARAN, who invited me to observe the harm reduction services for
heroin addicts in Yamuna Bazaar. When I visited the SHARAN Drop-
in Centre, I was an ‘international guest’, a role to which the participants
were adapted. Because the organization received international funding, and
conducted research in collaboration with Johns Hopkins University, the
members were used to visitors inspecting the services. My host, Derek, spent
half-a-day introducing me to their programs and staff members. He gave
me a tour of the facility, explained the services in English, enabled me to
observe an education session, and led me through the surrounding parks
where I watched addicts actively injecting. He also enabled me to take
pictures of the drug sets and equipment by explaining to the addicts that
I would not photograph them, just their drugs (FN 03.29.04). At this point,
my presence was not unusual for the participants. I was another Western
voyeur who was being escorted by the leader, perhaps in an attempt to
secure funding or programmatic assistance. There was no need to further
investigate this short-term visitor.
When I returned to the centre in December 2004, Derek and the staff were
surprised to see me. Derek stated, ‘Although you had said you would come
back, I didn’t think you actually would’. I was no longer following the
expectations of a guest: I returned to the site; I was hanging around for full
days without the leader, and I was speaking Hindi with the staff and clients.
This discordance with role expectations led to a flurry of questions that aimed
to re-position me into a meaningful role. Although my responses to these
questions aimed to establish a researcher’s identity, my efforts were thwarted
by more pervasive ideas about identities and motives. For example, when I
attempted to explain that I was doing research on how friends learn from
each other to a group of staff, one of the members announced to the group:
‘He’s thinking about setting up a new Centre in his country and came here to
get ideas’. The other members nodded in agreement with the statement (FN
12.21.04). Similarly, when I attempted to explain that I was born in Canada,
but my parents were from India, this was interpreted as: ‘He’s an Indian who
left India and is living in Canada’ (FN 12.21.04). As this statement reveals,
there were some intrinsic characteristics that also contributed to my role
designations. Specifically, my Indian origin and appearance (mid-twenties
age, male sex, foreign Hindi accent, single marital status) were a unique set of
characteristics that needed interpretation. One such justification was that I
was ‘an Indian living abroad who had returned to look for an Indian wife’.
My intrinsic qualities for the most part, however, could not explain my
presence in Yamuna Bazaar socializing with heroin addicts. Although one
participant mentioned that he initially thought that I was an addict (FN
02.03.05), most participants explained my membership as an ‘NGO worker’,
especially the non-addict, middle-class type who worked with the computers
in the back room. This assignment as a ‘worker’ was initially troubling for
me because it had interventionist expectations – that I would be providing
services for clients on a daily basis. I believed, however, that to force a
researcher’s role designation at this stage was neither tactical nor natural,
although I would continually remind participants of my research even though
they may re-interpret it. Therefore, I accepted the status of an ‘NGO worker’
as one of my roles, and contributed by providing feedback to the staff.
As an ‘NGO worker’ without any official responsibilities, my activities
included accompanying peer educators on outreach in the surrounding
8 AMAR DHAND
areas, sitting with staff members as they provided services at the centre, and
participating in idle ‘time pass’ routines such as drinking tea, conversing
about daily happenings with addicts, and simply watching and listening to
participants. My only informal responsibility was to provide feedback, or
commentary, on services. For instance, NGO leaders such as Derek would
request my input on educational materials such as new booklets, posters,
and videos. They would look for advice and insight from my experience,
which I tried to emphasize was very little in the NGO arena. On one oc-
casion, however, I intervened after observing NGO barbers not cleaning
shaving equipment or changing blades during their services with the clients
(FN 12.27.04). My actions included informing on-duty NGO leaders, re-
searching, and explaining the literature on contracting blood-borne viruses
to Derek, and conducting a 10-min presentation on the topic for the staff
upon his request (FN 12.28.04). Following this instance, which I felt I had
both a role and an ethical obligation to act upon, I did not make any more
presentations. Rather, my feedback came in more discreet conversations
with individual members. For example, when a worker publicly disclosed
the HIV status of a client, I took him aside and explained why that may not
be a good idea (FN 02.17.05).
Over time, it became clear that there was discordance with my role as an
‘NGO worker’ and its associated expectations: I spent little time in the back
room with the computers; I would not keep a regular schedule, and I was
more interested in listening and talking to active users than most NGO
workers. My anthropological tools, including my notepad and tape re-
corder, also became identifying symbols. Participants became interested in
what I was writing down or why their recordings would be useful. With new
inspiration, I began to explain that I was writing down stories so that I could
understand what was happening and then describe it to people in my coun-
try. Similarly, I would explain how I would listen to the recordings, think
about them, and then ‘play’ certain parts for the outside world. Some par-
ticipants became excited about the possibility of being recorded for an out-
side audience. My topic of ‘learning’, however, remained a mystery to most
participants who would again re-interpret my statements into constructs
such as ‘he is watching how drug users live’. Nonetheless, my positions
as ‘observer’ and ‘interviewer’ were becoming consolidated through these
interactions. Some of the peer educators who were keen observers of my
activities began to validate and legitimize my presence and actions. In one
instance, a peer educator explained to an addict: ‘He has come from Canada
to listen to us, and because he is from so far away, your stories are safe with
him and will not get to the police’ (FN 01.21.05). Similarly, prior to my first
recorded interview, two peer educators gained the participant’s permission

by endorsing me as a person from ‘outside’ who has come to talk to clients
and learn about SHARAN’s services (FN 12.29.04). As these examples re-
veal, specific supporters began to legitimize my questions as benign and
confidential, and my identity as deserving openness and, perhaps, respect.
Additionally, in the process of facilitating my research, some participants
were beginning to play the role of ‘co-investigators’.
Participants who were co-constructing the roles of ‘interviewer’ and
‘observer’ with me also began enacting these researcher roles. For example,
one staff member devised strategies about how to collect data:
1) Observe before 8am to see addicts before and after their first fix; 2) Listen to con-
versations at the night shelter where there is a more relaxed atmosphere; 3) Follow some
of the ‘smarter’ clients to see how they talk to their friends (FN 12.22.04).
Similarly, in a number of instances including my first recorded interview,

participants would ask questions and, occasionally, conduct full ‘interviews’
for my benefit:
Rohit and a self-appointed ‘translator’ began asking questions like when did you start,
what did you use, and where do you inject as if these were the questions that I would be
interested in. There is no doubt that I am but it’s funny that they assume that this is my
project and they are asking on my behalf. It’s also interesting that these interactions
automatically become interview interrogations of one or two users (FN 12.30.04).
One participant also became an ‘observer’ who reported his ‘findings’ in

connection with my current research aims:
After I described my current project of observing the switch [from smoking to injecting]
in the park, Derek said he observed a peer interaction in which one user was really
looking for an injection cocktail but couldn’t find one for the set price. Another drug
user then said, ‘I’ll take you there’. The man said, ‘No, you’ll give me today and then
what will happen tomorrow’. Derek thinks he said this because he and others were
present and listening (FN 03.02.05).
It is important to note that none of these instances were coordinated or

encouraged by me prior to the event; I did not recruit ‘interviewers’ or
‘observers’ for my research. Rather, in the process of co-constructing my
roles with the participants, the members themselves began naturally to enact
those roles. These instances, therefore, suggest that the legitimacy process
causes shifts in not only the researcher’s role, but also the participants’ roles.
Although the role of a liberal ‘NGO worker’ with a co-constructed ‘team
of research partners’ was advantageous on multiple levels, there were draw-
backs as well. First, as I have reported elsewhere (Dhand, 2006b), peer
educators needed to sustain a professional distance from active users to
10 AMAR DHAND
maintain a reputation of a ‘clean’ ex-user. I found that this distance was

being transferred into my relationship with active users as well. Second,
because of expectations that I should only be present with my other ‘col-
leagues’, I found that I did not ‘fit in’ so well to the context during non-
working hours (FN 08.14.05). Most importantly, I found that my research
questions, which were interested in naturalistic learning among active ad-
dicts, required a different set of roles than were available with my current
identity. Therefore, I underwent a transition period that involved changes in
spaces where I spent time, activities that I participated in, and relationships
that I fostered. I began to spend less time at the centre, limited my in-
volvement in services such as outreach, and became less connected with staff
members. Instead, I became more of a regular in the ‘backstage’ park,
played card games, and built friendships with a small group of addicts who
began to legitimize my presence as a ‘brother’.
BECOMING A ‘BROTHER’
Small groups of five to seven addicts would spend idle time together, sleep in
the same area, generate income, gather food, and usually, but not always,
use drugs together. They would call each other ‘brothers’, occasionally
with adjectives such as ‘big brother’ or ‘little brother’. Through a series of
legitimizing activities, I began an enculturation process of becoming a
‘brother’ in one small group. This section describes this process highlighting
my activities with the participants, knowledge acquired, boundaries of my
particular role, justification and validation of those boundaries, and legiti-
macy that participants gained from me.
To use Goffman’s (1959) term, the ‘backstage’ area was a public park
adjacent to the ‘frontstage’ SHARAN Centre. In this space, addicts
congregated, relaxed, socialized, gambled, and used various substances in
concealed corners on the periphery. Near the beginning of my fieldwork, my
access to the area was limited. The NGO staff typically stayed in the ‘front-
stage’ space because it was important for maintaining professional stand-
ards including a reputation of being ‘clean’. Therefore, through association,
my initial attempts to enter the ‘backstage’ was met with disapproving
comments by the staff and curious stares by the addicts. Any attempt
to strike up conversations with the addicts was awkward and uncomfort-
able. A breakthrough occurred when a key informant, with whom I built a
friendship in the NGO space, was playing cards in the park one day. I got
his attention from the outside railing and asked whether he could teach me
the game. He agreed and invited me to join the group. After introducing the
basic rules to the four-player game called seep, the key informant made me
hold his dealt hand of cards. Initially, he would gesture towards a card and
instruct me to place it on a particular pile or open space in the middle
playing area. Eventually, I would begin to gesture to a card and he would
nod affirmatively or negatively, with occasional explanations without
tipping off the other players (FN 02.17.05). Although I did not become
extremely competent at the game, the apprenticeship activity began to
legitimize my presence in the park. My actions of sitting with a small group,
watching the card game, and occasionally playing became accepted as part
of the social life of the environment. Furthermore, as the participants’
curiosity of me lessened, the activity became an ideal vantage point to
observe interactions and activities in the area.
As I learned about the range of activities that occurred in the park, I
began to participate in as many non-drug related ones as possible: I played
cricket with the addicts and street children; I listened to disagreements
and verbal fights that absorbed everyone’s attention, and I participated in
‘gossip’ about daily happenings. Of course, during the ‘gossiping’ sessions, I
could not help but ask a few more questions than the others, but I attempted
to refrain from disrupting the natural path of the conversation. The topics
of these ‘chats’ ranged from complaints about NGO services to health
problems of a particular addict to even jokes or stories for entertainment.
Topics that were relatively hidden from me at this time, probably because of
my lack of participation in drug-related activities, were conversations about
illegal income generation and drug-procurement procedures. These topics
would cause addicts to use a lower voice or to adjust their body position.
Importantly, my presence was not unnoticed by the group: many conver-
sations would be about me. Groups of addicts would ask numerous ques-
tions and engage in lengthy dialogue about my comments. In many ways, I
embraced this process and answered questions enthusiastically because the
sorts of queries and their responses to my answers provided me with insight
into the types of issues that were important to the group. Similar to the
NGO staff members, questions about my origin, family, and marital status
were some of the initial questions. However, unlike the staff, these parti-
cipants were also interested in street-related aspects of my country such as
whether fights occur between Muslims and Hindus, the degree of cleanliness
compared to India, and the variety and effects of the drugs (FN 02.09.05).
For many of these questions such as the drugs in my country, I could not
provide experience-based answers that satisfied the crowd. This would lead
to theorizing about the life and culture on the streets of Canada. One
12 AMAR DHAND
member explained the use of a drug called ‘crack’ that is placed on the top of
the head causing the body to freeze like a statue, which he demonstrated
(FN 01.20.05).
Over time, I learned about the daily routine that extended past NGO
work hours. Once the sun went down, mosquitoes would make the park
uninhabitable, so participants would co-ordinate to meet in other places
where they would have tea or continue their game. After receiving a
few invitations and making safety arrangements, such as having my rick-
shaw driver parked in the area, I began to join members for tea. During
my first night visit, I met Imran, a 38-year-old addict who had been in
Yamuna Bazaar for 20 years. Immediately, he struck me as a respected
senior in the group who was insightful of events and issues. During our
conversation, he described criminal activity under the bridge while addicts
sleep, the methods of two men in front of us who were planning to steal
supplies from the back of rickshaws, and the way that he and his friends
removed a dead body from the river one hour ago at the request of
the police (FN 08.31.05). My presence with him also stimulated other
addicts to approach us and talk about events openly and honestly. One
participant communicated that the reason he came to Delhi was to pick-
pocket to make money to survive (FN 08.31.05). The openness and richness
in stories were unparalleled in the study thus far. Participants were dis-
cussing topics that were previously off-limits with me. Additionally, on
repeated encounters with Imran at the tea stand and the park, I began
to notice a consistency to the addicts that gravitated around him, and they
began to welcome my company.
Through the initial connection with Imran, I began to spend leisurely time
with his small group of ‘brothers’ on a regular basis. I would seek out the
group in both the park and the tea stand and sit with them, drink tea, watch
card games, listen to the radio, and engage in daily conversations. I learned
that the ‘brotherhood’ consisted of five core street addicts and one or two
fluctuating ones. All the ‘brothers’ were Muslim, although two, including
Imran, had spent many years as Hindu Babas, or holy men, travelling to
pilgrimages across India. Although he was not much older than the rest,
Imran was referred to as the ‘big brother’ in the group and usually made
decisions about activities, timings, and drug use. In addition to the activities
that I mentioned above, the group would sleep in the same area and smoke
ganja (marijuana) together. The group, however, never used smack as a
group, which a member said was because Imran advised against it, although
members would do it ‘quietly’ outside of the group (FN 10.10.05). The
‘brotherhood’, therefore, served as a drug deterrent, but was probably
tolerant of private use by individuals. Throughout my time with the ‘brother-

hood’, I retained a privileged status: my tea was always paid for by Imran
or other members; my arrival was acknowledged with respectful greetings;
individuals became quiet when I talked and never argued with me as they did
with each other, and members would admonish any addicts who would ask
me for favours or money.
The group’s special treatment was complemented with a legitimization of
my participation boundaries and eccentric qualities. Most of the ‘brothers’
celebrated my lack of drug use including not taking cigarettes by openly
announcing it to curious bystanders. One of the ‘brothers’, named Veer, was
particularly vocal in my defence. He explained to another addict in my
presence that I do not play cards, smoke ganja, or drink with the ‘brothers’
because I probably do that with ‘brothers’ in my country, but cannot do it
here (FN 10.10.05). Similarly, in an admonishment of an addict asking me
to help him with NGO services, Veer stated:
He is here to meet brothers because of love. He comes here on his own accord to hang
out and meet with brothers. He will then tell his brothers [in his country] about the
brothers here and how there are good people in Delhi (FN 10.24.05).
These examples illustrate that key limitations of my involvement, namely

my non-drug use and inability to provide services, were justified in local
explanations to legitimize my unique role in the group. Similarly, my con-
stant notetaking, which many participants outside of the group associated
with police recordkeeping, was interpreted as beneficial for the ‘brothers’:
‘I think this, that our brother writes for our benefit. He does not write to
harm us’ (Veer 10.10.05). Moreover, Veer appeared to even understand my
research procedures:
Veer: You do this: you write down all of our conversations and go home at night. When
you write down from here, then at home you go at night, alone, you do it, you write.
That my brother did this today. Today, he did this and this. So you write all of it when
you go there. And that makes us happy.
AD: How did you figure this out?
Veer: You sit with us, get up with us, and you are always writing. So from this I
deciphered it. Yes, that our brother writes for us (Veer 10.10.05).
The legitimization process also had reciprocal qualities: the ‘brothers’

seemed to gain legitimacy through my presence with them.
Sentiments of love, happiness, and even pride were expressed when the
‘brothers’ described my involvement with them. More than one ‘brother’
14 AMAR DHAND
celebrated my connection with their group. During an interview, Veer de-

scribed how he felt about my presence:
That our brother has come close to us from a foreign land. He will do something or other
for us, our brother. For this reason, we get happy, by seeing you. It feels good to us.
That our brother has come to us from a foreign land. He sits here with us. Because we
have seen those types of people, who don’t stand, let alone sit. But you sit and stand with
us. It feels good to us. He is our brother. He is also one (Veer 10.10.05).
Acknowledgement of my participation in their lives was also accompanied

by an appreciation of my studies and pride for my future potential. Veer
continued:
When you go home, then our big brother Imran, he remembers you. So following him,
we also begin to remember you. This brother tells us that our brother is doing his studies
there. He is doing great studies. One day, he will become a good man, a very big man he
will become. So we also get happy, that our brother had come close to us at one time. He
is doing some good work. We are addicts. But our brother, he is one good man, sure a
great man. So, we remember you. So it makes us happy, when our big brother tells us
(Veer 10.10.05).
Although it was important for me to gain legitimacy from the group to

secure access, my scholarly accomplishments and potential for success
seemed to provide legitimacy or even hope for Veer and, perhaps, other
members of the group. It was probably the first time that they could
associate as a ‘brother’ with a person who was not tragically fixed in a world
of poverty and addiction. Perhaps through documenting and presenting
their lives to ‘brothers’ in my country, I was a vehicle for a type of tran-
scendence for them.
Becoming a brother was a gradual process of gaining access to the ‘back-
stage’ regions, building relationships with key informants such as Imran,
and spending copious amounts of leisurely time with one small group of
‘brothers’. In so doing, the ‘brothers’ became advocates of my unique quali-
ties and actions, and, in turn, I became a legitimizing presence for their
lives as well.
A FRIENDSHIP WITH BABA
Baba was a 33-year-old street addict who became a key informant in the
study. My relationship with him was an important means of experiencing
and understanding events and issues. Our relationship, therefore, not only
provided physical access, but also interpretive access to underlying meanings
that were often implicit and subsumed in everyday life. This section will
describe these aspects of access by delineating the development of our re-

lationship and how it gained interpretive functionality. It highlights initial
role-plays of a ‘guide’–‘follower’ where Baba facilitated my viewing and
experiencing of events. It follows with descriptions of how we coordinated
our daily lives and created new communication strategies to teach and learn
from each other. Lastly, it recounts the formation of ‘big brother’–’little
brother’ dynamics including protection and planning instances that both
facilitated and impeded access to data.
Baba was a holy man in appearance. He had a beard and wore a cream-
coloured turban, saffron kurta (traditional loose fitting shirt), white dhoti
(wrap-around cloth around his waist and legs), beads around his neck, and
colourful rings on his deformed fingers. I was, therefore, interested in his
reaction to an exorcism that was taking place on the banks of the Yumuna
River during the first days of my fieldwork. However, our initial attempts at
communicating were frustrating: he would speak in a rough street Hindi
dialect that I could not understand, and I would reply in formal Hindi
that would humour him. Therefore, I chose to stand next to NGO staff
members and listen to their commentary instead. However, when the ex-
orcism group returned from the other side of the river on a boat and banked
slightly downstream from our ghat (steps leading into the water), Baba
gestured to me:
Baba follows, looks back at me and nods at me, motioning me to come forward. I don’t
know if this is a good idea but I move behind him. We navigate some tight alleys. I catch
up to him and he says, ‘look at what they did to her’. We walk up the stairs and down to
the main road where we see two auto rickshaws waiting y Baba leads me to the street
and tells me to write down the number of the rickshaw (FN 12.27.04).
Baba facilitated an extended observation of the event. He had navigational

knowledge about the alleys and cultural understanding about how close to
follow and watch the events, all of which he used for my benefit. Moreover,
his facilitation of accessing these data was not through verbal explanation,
but through silent direction. He became a ‘guide’ and I became a ‘follower’.
In subsequent months, Baba guided me through a number of experiences
that were difficult and, occasionally, dangerous without his help. In the
one instance that I assumed an ambiguous identity, he led me through an
observation of how drug cocktails were acquired by injecting drug users.
Drug cocktails were ampoules of injection drugs sold by local pharmacies
for inflated prices to street addicts. I had heard the name and location of one
such pharmacy and I wanted to observe the interactions. When I asked
Baba for directions, he recognized the naivety of my plans and decided to
16 AMAR DHAND
take me. On our way, he reprimanded me for arranging to complete the

observation by myself. He explained that I could not go to the shop by
myself and just watch. The addicts or the storeowners may suspect me, and
then I would be in danger. Subsequently, Baba organized the operation:
he recruited a middleman who was a ‘regular’ and would be trusted by the
pharmacist; he instructed the middleman on the drugs to purchase and the
procedures to use, and he assembled our relative positions so that the mid-
dleman would not run away and I would have a good view of the trans-
action. He then lectured to me about the ‘rules’ that I must follow to observe
the drug procurement. His directions were specific and repeated multiple
times with illustrative analogies, examples, gestures, and signals. In total, he
gave me the following ‘rules’:
1. Don’t go to the shop by yourself.
2. Don’t let anyone know that we are doing this just for knowledge.
3. Keep quiet during the negotiation with the middleman.
4. Watch your pockets (Baba slapped my hand when I took something out of my pocket
and a 10 rupee bill fell on the ground).
5. Follow the middleman and keep walking (to make sure he doesn’t run away).
6. Don’t show your notebook.
7. Don’t let them think you are an undercover cop.
8. Don’t show the ampoules too much.
9. (Through gestures like stepping on my foot) Walk away from the middleman and
don’t talk to him too much.
10. Pretend you are a user who uses at home (FN 03.02.05).
These ‘rules’ were the implicit protocols for assuming an indigenous identity
in this situation. Baba was providing me with cultural knowledge about
interactions, motivations, and dangers that he probably acquired through
experience (in dealing drugs as I later learned). Moreover, he had tailored his
‘rules’ to accommodate my unique qualities and motives: he would probably
not tell other ‘novices’ to hide their notebooks. Through ‘insider’ knowledge
and an understanding of my traits, Baba enabled me to be legitimate and
peripheral in the social context, thereby securing access not only physically,
but also in terms of interpretation of events and personalities.
Baba used to describe our relationship as ‘two guys walking together’ (FN
03.02.05). There was coordination in our daily activities. We created our
own traditions such as buying a half litre of whole milk and sharing a kettle
of dood ka chai (tea with extra milk), exploring hidden markets of Old Delhi,
and walking together to my rickshaw at the end of the day. Baba would also
emphasize our shared Kshatriyas caste heritage and suggested that we follow
the traditional schedule of our ancestors including waking up at 4:00 am,
working from 6:00 am to 12:30 pm, napping from 1:30 to 3:00 pm, working
again from 3:00 to 9:00 pm, eating dinner and sleeping at 11:30 pm
(FN 02.17.05). During this and other suggestions such as playing the role of
‘street doctor’ which he did often and I was not comfortable doing, I had to
set a limit to our joint participation. Such boundaries were usually accepted,
although occasionally they would lead to curt disagreements. Our negoti-
ated co-ordination led to the development of new communication mecha-
nisms that bridged the comprehension gap that was initially present. During
our conversations, he would frequently interrupt himself and ask: ‘Do you
understand what I mean?’ Whenever I would say yes, occasionally he would
test me by saying: ‘Okay what? Explain’ (FN 09.15.05). Additionally,
he would reiterate some of my incorrectly pronounced words in better forms
to both understand what I meant, and to help me learn the ‘lingo’. Inter-
estingly, he also began to detect when I would use my language difficulty to
act naive with him or other participants; he would joke that I played dumb
so that others would give me more information (FN 08.20.05). Outside
of language scaffolding, we also drew diagrams and used demonstrations
to convey our meanings. Diagrams such as maps of places for exploration
were drawn in my notebook or in the dirt with both of us contributing
scribbles to make points as we talked. Demonstrations such as common
pick-pocketing strategies in crowded areas (FN 08.30.05) would involve
both of us using bodily movements to understand each other.
The co-constructed process of designing specialized communication mech-
anisms produced educative interactions that were reciprocal. As we began
developing these unique conveyance devices, much of our time was spent
teaching and learning from each other. Perhaps, we both became excited
about the ‘foreign’ information that each of us could gather from the other
person. Baba taught me a great diversity of topics. Nearly every area that I
was interested in received some commentary and interpretation from him. He
was particularly important in understanding core religious and cultural val-
ues, health beliefs, and different relationship structures. He would illustrate
his points with experiences from his roles as a ‘holy man’, ‘street doctor’,
‘drug dealer’, and ‘son’ in a ‘father’–‘son’ relationship with another street
dweller. For example, he would explain the causes of ‘cancer’ in the lungs by
drawing a diagram in my notebook (Fig. 1) depicting alcohol going into the
lungs and damaging them (Fig. 2). On another occasion, he described the
18 AMAR DHAND
Fig. 1. Baba (Right-Hand Side) Drawing a Diagram for the Ethnographer (Left-
Hand Side) (FN 09.15.05).
players involved in drug dealing by adding arrows to the English labels that I
wrote in my notebook (FN 09.14.05). Importantly, his comments would be
couched in elements of our new ‘language’ that began to ring with clarity.
Our collaborative tools were able to expose and unravel difficult concepts
providing an unparalleled access to significant meanings and processes. This
educative utility was also reciprocal. Baba would ask me to read English
instructions on drug labels and explain them to him (FN 02.04.05). I would
also conduct basic English speech and writing lessons as well as participate in
mental math games that he would usually win. On occasion, he would play-
fully call me his ‘master’ in reference to a principal in a school.
As much as we were collaborators in gaining understanding from each
other, Baba and I also constructed a ‘big brother’–‘little brother’ dynamic.
Fig. 2. Baba’s Illustration in the Ethnographer’s Notebook Depicting How Alcohol

Goes into the Lungs and Causes ‘Cancer’ (FN 09.15.05).
On multiple occasions, Baba would nurture and protect me. For example,
during my interviews with other addicts, he would occasionally leave a cup
of tea on a ledge behind me. Moreover, the specificity of his ‘rules’ during
our observations and explorations were largely to ensure my safety.
20 AMAR DHAND
Following his pick-pocketing demonstration, he emphasized that I should

learn this so that I do not become a victim. During the final months of my
fieldwork, he also became very interested in giving advice for my future. He
recommended that I should open a hospital with a pharmacy on the side to
treat patients and make money from drugs as well. He also suggested getting
married quickly, having one boy and one girl, and taking care of my mother
and father (FN 09.18.05). Such notions of ‘settling down’ seemed to taint his
perspective of my research. As I was watching a card game in the park with
Baba, Imran, and the brothers, the following dialogue occurred:
Baba then says, ‘Amar Singh, settle down.’ Imran returns the comment and says, ‘Worry
about yourself.’ Baba continues, ‘I look at him and I y don’t like it y sitting with drug
addicts. Open a store. Sitting with these men, and you will one day y [use drugs?]’ (FN
10.08.05)
In some ways, Baba wanted to save me from becoming like him. In his ‘big
brother’ role, he was not a ‘role model’ but rather a person who encouraged
me to live a traditional healthy life, and perhaps even unlearn the ‘tragic
knowledge’ that he had taught me. Although such tactics of nurturing, pro-
tecting, and planning were important for access, they also impeded access
and data generation in significant ways. Baba would get frustrated with
seeing me spend idle time with the ‘brothers’. Moreover, he would admonish
me for walking alone in some areas, coming at night, or talking to particular
people. Balancing these concerns with my objectives as a researcher took
negotiation and, at times, distancing from my key informant.
Baba provided access to core meanings and implicit knowledge of a street
addict’s life. Our relationship involved guidance through everyday activities,
sharing daily rituals, creating new communication tools, and using these
tools to teach and learn from each other. Through our friendship, he en-
abled me to be a researcher that was physically present in, and cognizant of,
some key moments in an addict’s everyday routine.
DISCUSSION
Access in ethnographic research is the participation in negotiated interper-

sonal relationships or ‘role-plays’ that enables the ethnographer to become a
legitimate peripheral participant in the socio-cultural context. The element
of negotiated interpersonal relationships is handled by Harrington (2003),
and is supported by these data. This analysis, however, reveals the impor-
tance of acknowledging the social learning processes involved in access, and
suggests the utility of Lave and Wenger’s (1991) ‘legitimate peripheral par-
ticipation’ in understanding these processes. In applying Lave and Wenger’s
thinking to the situation of access, some implications for theory develop-
ment are also revealed.
Following a literature review and analysis of the access problem,
Harrington (2003) arrives at four relationship regularities that occur
during the process of access in ethnographic research:
When ethnographers approach a research site, they will be defined in terms of social
identity categories salient to participants (p. 607).
Ethnographers gain access to information to the extent that they are categorized as
sharing a valued social identity with participants or as enhancing that identity through
their research (p. 609).
Ethnographers’ identity claims must be validated by participants in order for researchers

to gain access to information (p. 611).
The more that ethnographers’ social identities differ from those of participants, the more
likely that access will involve the use of insider informants or deception as self-pres-
entation strategies (p. 612).
All four patterns have some resonance with the data presented in this paper.
In support of the first point, the NGO staff initially categorized me as an
‘international guest’, ‘a person who was thinking about setting up a new
centre in his country’, and an ‘NGO worker’. These identities were mean-
ingful to the participants, and they were assigned to me to make sense of my
presence. The second pattern was applicable in the processes involving
co-ordination of daily activities with participants. For instance, my posi-
tions as a ‘brother’ in the ‘brotherhood’ and ‘guy who walks with me’ to
Baba became viable roles when the participants and I shared time together,
and we began to recognize and enhance our joint similarities. The third
regularity was relevant during instances of the co-construction of roles (e.g.
‘observer’ and ‘interviewer’) and the legitimization and defence of my in-
trinsic characteristics (e.g. Veer’s comments). The final assertion reflects the
necessity of Baba to enable experiences and unlock everyday meanings for
my comprehension.
Although Harrington appropriately focuses on the interpersonal rela-
tionships, she fails to illuminate the enabling mechanisms underpinning the
patterns. Especially for points two, three, and four, the following questions
arise: How do ethnographers become aware of which identity elements to
share and enhance, and how that sharing should take place? How do re-
searchers create opportunities in which their identity claims will be vali-
dated? How do investigators align perspectives and goals with insider
22 AMAR DHAND
informants so the informants become meaningful access-points and inter-

preters? My argument is that the answers to these questions involve learning
processes that are theoretically grounded in ‘legitimate peripheral partici-
pation’ (Lave & Wenger, 1991). Through participating in daily interactions,
ethnographers begin in designated roles on the periphery. From here, they
examine other roles, recognize meaningful identity elements, and become
reflexive about which of their qualities best ‘mesh’ with established iden-
tities. They then spend time in joint activities (e.g. card games), appreciating
the ‘rules’ of conduct, and testing different ‘self-presentations’ with the
participants. Slowly, they enable participants to relate to their work in ways
that are understood locally, and concurrently involve them in co-creating
legitimate positions that make ‘sense’, and may be articulated to others.
Lastly, with those members who become particularly helpful, ethnographers
co-create communication devices and teaching tools with the informants
that enable both parties to ‘read’ and teach each other in meaningful ways.
Through this process of participating in activities and engaging in
dynamic relationships, ethnographers are penetrating the ‘learning curri-
culum’ of ambient community life. They become recipients and contributors
to ‘knowledge’ that is constantly being created, modified, and stored in the
course of socio-cultural evolution. The acquisition artifacts such as skills
and information that serve as indicators of having ‘learned’ something are
absorbed along the way in a ‘subsumed’ manner. In many cases, researchers
may not even notice when or how they acquired the ability to participate
appropriately or make the correct judgements. Unbeknown to them, their
instincts become tuned to what to do in certain circumstances, who to talk
to about specific topics, and how to ask the right questions. During these
instinctual moments, reflexive ethnographers may inquire about why or
how they ‘guessed right’ about a person or situation. Only then will they
appreciate the social learning processes that have enabled them to gain
access to this level of insight.
There are some qualifications to the described process. First, the legiti-
mate roles that are constructed do not need to be identities that have a
history with the group (e.g. ‘addict’ or ‘apprentice’). Some of the con-
structed roles may be quite unique or ‘new’ to the context, and still
be legitimate (e.g. ‘observer’ and ‘brother who does not smoke with us’).
Second, as an elaboration of the first point, ‘full participation’, which Lave
and Wenger argue is where peripheral participation leads, need not be
‘complete participation’ (Gold, 1958). An ethnographer can become a ‘full
participant’ without resorting to covert research: ‘full participation’ in a
diversity of roles including one akin to ‘researcher’ can be legitimized.
Lastly, because of its negotiated nature, the learning process has reciprocal
elements. Access is successful only when the participants engage in learning
about the researcher as well. Only then will research goals be understood
and incorporated into legitimate roles and activities.
This final point reveals the unique commentary that the journey of the
ethnographer may offer to theoretical development of ‘legitimate peripheral
participation’. Although Wenger (1998) acknowledges the concept of
‘brokering’ or the ‘use of membership to transfer some element of one
practice into another’ (p. 109), there is limited consideration for the extent
and complexity of reciprocal effects (Davies, 2005). These data reveal that as
participants (‘old-timers’) learn about researchers (‘newcomers’) and co-
construct meanings with them, they may participate differently and shift the
types of roles that they perform in the community (e.g. ‘observer’ and ‘in-
terviewer’). Fuller et al. (2005) have also reported that novices may play
educational roles in relation to experienced workers in contemporary work-
place settings:
For example, our research has demonstrated that experienced workers are also learning
through their engagement with novices, and that part of the process of legitimate peripheral
participation for many novices is to help other workers to learn. This insight is of sig-
nificance as it helps undermine the view of communities of practice as unchanging (p. 64).
These findings, therefore, suggest that understanding how ‘old-timers’ learn

in relation to ‘newcomers’, and how that learning affects the legitimization
process may be valuable avenues for inquiry. A second point for consider-
ation is the diversity of journeys that different ‘newcomers’ may experience.
The original theory recognizes that there is a ‘diversity of relations involved
in varying forms of community membership’ (Lave & Wenger, 1991, p. 37).
However, it may be useful to consider the extent of difference between my
journey and the journey of an experienced addict who has just moved into
the area. To account for the variety of experiences of such dissimilar ‘new-
comers’ may be beyond the applicable boundaries of the theory. However,
exploration of the potential range of differences in journeys, and how that
range relates to the initial characteristics of ‘newcomer’ or other variables
such as time or significant events, seems to be an intriguing line of research.
CONCLUSION
This paper has attempted to provide a theoretical addition to conceptuali-

zations of access in ethnographic research. It argues that access is not only
24 AMAR DHAND
the researcher’s interpersonal relationships with the participants, but also

the process of becoming a legitimate peripheral participant in the socio-
cultural context. The analysis has revealed that the process of engaging with
the ambient community’s ‘learning curriculum’ may be strategic and nego-
tiated, but also quite subtle and implicit in nature. Therefore, careful and
reflexive accounts of access journeys with consideration of social learning
processes are needed to provide further insight into the phenomenon. Be-
cause of the unique position of the ethnographer as a cognizant ‘newcomer’,
such an approach may also be of value to the theoretical development of
Lave and Wenger’s prominent ideas.
ACKNOWLEDGMENTS
Mr. Francis Joseph transcribed and transliterated the tape recordings. I am

grateful to my advisor Professor Geoffrey Walford for his probing ques-
tions, and the staff of SHARAN and SAHARA for their co-operation and
friendship. This research was supported by the Rhodes Trust.
REFERENCES
Coy, M. W. (Ed.) (1989). Apprenticeship: From theory to method and back again. Albany: State
University of New York Press.
Davies, B. (2005). Communities of practice: Legitimacy not choice. Journal of Sociolinguistics,
9(4), 557–581.
Dhand, A. (2006a). The practice of poetry among a group of heroin addicts in India:
Naturalistic peer learning. Ethnography and Education, 1(1), 125–141.
Dhand, A. (2006b). The roles performed by peer educators during outreach among heroin
addicts in India: Ethnographic insights. Social Science and Medicine, 63(10), 2674–2685.
Downey, G. (2005). Learning Capoeira: Lessons in cunning from an Afro-Brazilian art. Oxford:
Oxford University Press.
Fuller, A., Hodkinson, H., Hodkinson, P., & Unwin, L. (2005). Learning as peripheral par-
ticipation in communities of practice: A reassessment of key concepts in workplace
learning. British Educational Research Journal, 31(1), 49–68.
Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday.
Gold, R. L. (1958). Roles in sociological field observations. Social Forces, 36(3), 217–223.
Harrington, B. (2003). The social psychology of access in ethnographic research. Journal of
Contemporary Ethnography, 32(5), 592–625.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge:
Cambridge University Press.
Shaffir, W. B., & Stebbins, R. A. (Eds) (1991). Experiencing fieldwork: An inside view of quali-
tative research. Newbury Park, CA: Sage.
Walford, G. (2002). Why don’t researchers name their research sites? In: G. Walford (Ed.),
Debates and developments in ethnographic methodology (pp. 95–107). Amsterdam,
London: JAI.
Wax, M. L. (1967). On misunderstanding Verstehen: A reply to Abel. Sociology and Social
Research, 51(3), 323–333.
Weinstein, M. (2004). TamsAnalyzer: A qualitative research tool (Version 2.49b7). Boston, MA:
Free Foundation.
Wenger, E. (1998). Communities of practice: Learning, meaning, and identity. Cambridge:
THE QUESTION OF INTIMACY IN
ETHNOGRAPHY$
Harry F. Wolcott
I have many friends who are ethnographers, but most of my friends and
colleagues are not ethnographers. So I look for ways to explain ethnography
to them without inadvertently giving a lecture or seeming too pedantic. Yet I
do not want to shortchange my response or forego an opportunity to inform
someone of what ethnography means to me and the kind of person-centered
approach that I take.
So throughout my career I have watched for definitions or descriptions or
even catch-all phrases that one can use to convey the essence of ethnog-
raphy. ‘Participant observation’ is of some help, but it hardly explains why
someone might spend a year or two among a group of total strangers,
another year writing about them, and then expecting to find an audience
eager to read the report. Nor could I find a way to explain why no two
ethnographies ever seemed to look exactly the same.
Most recently, I have turned my search to see if I can identify for myself a
set of conditions that all ethnographies seem to fulfill. It is easy to construct
such a list, but I always find that as soon as I begin to elaborate, I run into
exceptions, conditions that are NOT met by each and every ethnography.
$
An abridged version of this paper was presented at the meetings of the eleventh annual
Interamerican Symposium on Ethnographic Educational Research held in Buenos Aires,
Argentina, on 22 March 2006.

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12002-1
27
28 HARRY F. WOLCOTT
Most of the conditions that I found are met in most ethnographies most of
the time, but no one set of conditions seems to hold in every case.
Here was my beginning list. I concluded that ethnography is:
1. holistic
2. cross-cultural
3. comparative
4. based on first-hand experience
5. conducted in natural settings
6. the result of intimate, long-term acquaintance
7. non-evaluative
8. basically descriptive
9. specific
10. flexible and adaptive
11. corroborative
12. idiosyncratic and individualistic
Developing this list was fun and a challenge. Of course there could be any
number of items on it – 12 just seemed about right for my purposes. But I
wondered if every facet that I identified really had to be present every time.
Can the mix vary from one ethnographer, or one ethnography, to another? I
came up with the idea that writing an ethnography is somewhat like baking
a loaf of bread. No two recipes or loaves need ever be quite the same, and
there is no absolute set of ingredients that goes into the mix. Yeast, sugar,
salt, and fat are usually added to flours ground from wheat and other
starchy seeds, but buckwheat and quinoa, for example – both of which can
be used in bread making – are not members of the grass family.
Nor does what passes for a loaf of bread necessarily look the same the
world over. I note only that loaves of bread usually bear strong resemblance
to the variety of breads made by others in the same region. If you are baking
a loaf, or writing an ethnography, you want yours to look like the ones
made by those around you who claim to be doing something similar.
It is not only the ingredients but what is done with them – how they are
selected, and how they are combined – that makes the differences we ob-
serve. The basic materials are essentially the same.
Like the world’s breadmakers, qualitative researchers everywhere draw
from the same source – in this case, human social behavior – to obtain their
raw material. It is their choice of ingredients and how they choose to combine
and work them that make the difference, I cannot help wondering if distinc-
tions that we pride ourselves on do not seem nearly as important to those
The Question of Intimacy in Ethnography 29
outside our own circles. For example, North American ethnographers usually
combine their materials with a healthy dose of culture. ‘Culture’ is not some-
thing one sees on the ground as an observer; it is something the ethnographer
puts there to make data workable from an ethnographic point of view. But if
it is not culture, some other organizer – for instance, community, institution,
social structure – helps render the analysis, giving some characteristic way of
‘marking’ one’s work, so it looks like what it is supposed to look like.
A TEST OF THE ATTRIBUTES LIST
I cannot subject listeners to such a long list. At my age I cannot always

remember all 12 items myself. But I keep them in mind as a guide for
discussions about ethnography, especially for assessing works that seem to
be questionably ethnographic, that is, works that claim to be ethnographic
but are suspect in that regard. I do not think that whether one meets these
conditions is necessarily that important, except that recognizing what ought
to be included to warrant the label ‘ethnography’ serves as a convenient
shorthand for describing a study and judging its contribution.
My last field-based study presents a case in point. It was based on original
fieldwork. It started out as what anthropologist Stanley Brandes has
described as an ‘ethnographic autobiography’ (see a discussion in Wolcott,
2004). It is a life story of a young man whom I called ‘Sneaky Kid,’ an
unfortunate choice of names since it made him seem far younger than he
was. I ‘found’ Sneaky Kid living on my place – I live in the country on the
side of a hill with some steep forested terrain. He had constructed a cabin
out of sapling trees when I discovered him. He was a 19-year-old youth at
the time, had spent some part of his life in the local community, had long
been out of school, had never been regularly employed, and admitted to
having recently spent time in ‘reform school.’ Since he was on private
property, he asked if he could stay.
As long as he did not burn down the forest, he was not going to hurt
anything, so I let him stay. He stayed far longer than I ever imagined – two
years! When I needed to do a major construction project, digging a new
water line to my house from a spring on the hill, he asked if he could do the
work rather than my hiring it out. We began working together. I found
myself growing increasingly attracted to him, at first psychologically, then
physically. I was fascinated with his ability to eke out a living for himself
and survive the winter months in cold, damp weather. Given what seemed to
be our mutual inclination, we eventually became physically involved as well.
30 HARRY F. WOLCOTT
Then the thought occurred that here was the very kind of person I had
hoped to find for studying cultural acquisition. With our newfound inti-
macy, I assumed that he could and would talk about any aspect of his life.
He agreed to some interviewing, and I began to tape sessions with him to
recreate his life story. Eventually, I prepared a draft that he read and ap-
proved while still living on my property.
Toward the end of his second year, however, things started to go awry. He
began acting strangely, and one day he suddenly announced his decision to
leave. We both knew he had no particular place to go, but he had become
dissatisfied that his life was not leading anywhere. We parted as friends.
Frankly, I thought he might be back in a matter of weeks, if not days. I
knew I would miss him.
I tried contacting him. He had promised to keep in touch, but I
heard from him only once after he left. His travels of several months ter-
minated at the home of his mother in Southern California, a thousand miles
away. On the telephone he seemed quite animated. He told me that his
mother had committed him briefly to a mental health facility. Then I totally
lost track of him y until he returned two and a half years later to burn
my house to the ground and seriously attempt to harm me. There followed
a trial, after which he spent some time in prison. I have never seen
him since.
It is a tragic story, but that is not my point. My publisher encouraged me
to write more, beginning with the original life story but extending it now to
describe his surprising return and all that had transpired, both before and
after the fire. I accepted the publisher’s challenge. That is the account that I
have provided in Sneaky Kid and Its Aftermath: Ethics and Intimacy in
Fieldwork (Wolcott, 2002). But there were two of us in the story now, and
the new part read more like a memoir. So the question for me became:
Where does ethnography stop and the account become something else? And
does it matter?
I assume there are certain cautions concerning any group of people with
whom we intend to conduct ethnography. I am not sure what particular
precautions apply to dealing with a 20-year-old streetwise youth who had
undergone police interrogation but was not used to ethnographic inter-
viewing. We began with questions about where he had lived and how he had
gone about selecting a site and constructing both his cabin and his life. He
never did talk as freely as I assumed he would, in spite of our perceived
intimacy. I was never sure if some of the stories he told about others were
really about himself – there were parts that I simply could not corroborate.
Nor did I have to. My account starts with his story, in his words – until it
became our story, mostly in my words. The account has coherence, and his
part had his approval. But, is the account ethnography? Let me evaluate it
against my list of criteria.
The account is holistic, based on the experience of the full two years he
spent in my place. I am sure I was the only person with whom he regularly
conversed. Our formal interviews could not have been conducted in more
natural settings – I took the tape recorder up to his cabin and we talked in
the sunshine of Oregon mornings. But we also talked all the time outside the
formal interviews – at least as much as a reticent youth will talk. Our
conversations were always specific: about him and his life, both as he had
experienced it and as he sometimes liked to imagine it. Of necessity, my
stance was non-evaluative – I had to turn a blind eye to some of the ex-
periences he recounted (such as his attempt to rob a convenience store) as
well as to how he had procured some of the things that now graced his cabin
and his life – stolen bicycles topping the list.
Oddly enough, he was not into drugs except for his sporadic use of mar-
ijuana. His plan to grow a few plants of his own suffered from his failure to
water them consistently. This absence of heavy drugs is the most dramatic
contrast with anthropologist Juan Gamella’s strikingly parallel account of a
similar-aged youth in Madrid conducted during the same years and pub-
lished there, La Historia de Julián, memorias de heroina y delincuencia,
currently in its fourth edition (See Gamella, 1990).
My study is as flexible and as adaptive as I could allow myself to be in
doing it, for I had literally stumbled upon my subject in my own backyard.
I was slow to see the possibilities for preparing an account contrasting
schooling and education, but here was a golden opportunity. What may seem
missing are qualities of being cross-cultural and comparative, but in this case
– learning about a person from the same community – I brought those
qualities to the study myself from previous fieldwork as an anthropologist.
I have long insisted that we need to make better use of anthropology
in everything we do under the banner ‘anthropology and education.’ In this
case, I introduced a cross-cultural perspective by contrasting what I called
Sneaky Kid’s cultural alternatives with my own patently middle-class career
alternatives. When he decided to leave, we talked about the possibilities he
was considering as an emerging adult. They were remarkably different from
the ones I would have considered under similar circumstances. He tried
enlisting in the armed services, just as Julián in Gamella’s account had done
– until he discovered that he had not completed enough years of schooling
to qualify for enlistment. The other alternatives he mentioned included
the unlikely possibility of committing suicide, going permanently on welfare
32 HARRY F. WOLCOTT
(off and on he had been the recipient of food stamps), getting locked up in
jail or prison, being institutionalized as a crazy person – an attractive al-
ternative to him – or going ‘home.’ Home, his least favorite possibility, was
the option he finally had to settle for, although he had not been a welcome
guest in either of his divorced parents’ homes for years.
To return to the ‘tricky part,’ let me say more about ethnography as the
result of intimate, long-term acquaintance. What is ‘intimate enough?’ And
how intimate is too intimate? I remind you that Sneaky Kid and I were into
a sexual relationship before he became my informant. It was the very nature
of our relationship that prompted my asking if he would share his life story.
(It might be more accurate to say that he agreed to let me draw his life story
out of him. He originally estimated that his story would take a total of about
12 h to tell. In fact, when we started the interviews, he spoke non-stop for
about 12 min, concluding with ‘and so I decided to come up here and give it
a try.’ After that, it was entirely up to me.)
Not much has ever been said about being involved with an informant,
particularly a younger one of the same gender. Although I have tried to open
things up a bit by addressing this issue in the book, my experience is that not
much is ever going to be said. Nor do I recommend that anyone attempt to
advance the frontiers of ‘anthropology and education’ by working on this
angle. But we are human beings, and these things happen. When they do, I
think we are better off to be forthright about them than to pretend they never
occur. Your readers expect to learn about the people you have been learning
about, but they also want to know something about YOU, their storyteller.
Working with what we recognize as vulnerable age groups – children and
youth – our personal feelings about them can inspire or impede us. I am
always interested in learning all I can about those who want to inform me
about children and youth.
To this point, I have presented the problem of intimacy in fieldwork as
being mine. Now I would like to make it yours, as well. Of all the terms on
my list of conditions for ethnography, when I say that ethnography is ‘the
result of intimate, long-term acquaintance,’ I think I have pointed to eth-
nography’s single-most telling aspect, the one that sets it apart from all other
qualitative approaches. And so I ask again, ‘How intimate is intimate?’ And,
when you work with children and young people, where do you draw the line?
Years ago, an American anthropologist posed a rhetorical question: What
kind of Hopi or Kwakiutl Indian will tell his story to a White man? I now
pose a parallel question: What kind of adolescent will tell his story to an
ethnographer? Or we might turn the question around to ask, ‘What kind of
story is an adolescent willing to tell a researcher?’
I am all for using ethnographic techniques in learning about young peo-

ple. I know I learned a lot from and about Sneaky Kid, and I doubt if I will
ever understand all that I did NOT learn about him. But, as Clifford Geertz
reminded us years ago, ‘it is not necessary to know everything in order to
understand something’ (Geertz, 1973, p. 20). So, what exactly is it that we as
ethnographers wish to know? And how likely are we to find out?
REFERENCES
Gamella, J. (1990). La Historia de Julián: memorias de heriona y delincuencia. (4th ed. 2003).
Madrid: España.
Geertz, C. (1973). Thick description: Toward an interpretive theory of culture. In: C. Geertz
(Ed.), The interpretation of cultures. New York: Basic Books.
Wolcott, H. F. (2002). Sneaky kid and its aftermath: Ethics and intimacy in fieldwork. Walnut
Creek, CA: AltaMira Press.
Wolcott, H. F. (2004). The ethnographic autobiography. Auto/Biography, 12(2), 93–106.
TO COOK OR NOT TO COOK:
PARTICIPANT OBSERVATION AS A
DATA COLLECTION TECHNIQUE
Susan James
INTRODUCTION
As I sit here typing, I take a long, hard, registering look at my hands. I see them as
someone else would: dirt embedded in cracked, dry, stained skin, with small cuts, some
covered in bright blue waterproof plasters, others healing nicely. Hands that, as a child, I
imagined belonged to the wicked witch of the west or hands a gardener or other crafts-
person may be proud of y
(Diary notes, 20 May 2002)
The reasons my hands looked this way were the 50 kg of potatoes I had
peeled in the 2 weeks prior to the diary entry, the 40 kg of onions I removed
the paper shells from and the 10 kg of Jerusalem artichokes I separated their
skins in the kitchen of a local restaurant. These tasks were undertaken in the
endeavour to understand how an apprentice constructs knowledge and skill
in the workplace.
In recent years there has been a renewed focus on learning in the work-
place (Billett, 2001; Hawke, 1998) and the literature is extensive and rapidly
increasing (for example, Billett, 1995, 1996, 2001, 2002; Boud & Garrick,
1999; Eraut, 2000a, 2000b, 2001; Eraut, Alderton, Cole, & Senker, 1998;
Fuller & Unwin, 2003a, 2003b; Fuller, Hodkinson, Hodkinson, & Unwin.,

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12003-3
35
36 SUSAN JAMES
2005; Hodkinson, 2004 among others), but regardless of your epistemolog-

ical belief, ‘one broadly established truth y is that learning at work cannot
be separated out from the everyday working practices of the workplace’
(Hodkinson, 2004, p. 12):
Learning is no longer a separate activity that occurs either before one enters the work-
place or in remote classroom settings. y The behaviours that define learning and be-
haviours that define being productive are one and the same. Learning is not something
that requires time out from being engaged in productive activity; learning is the heart of
productive activity. (Zuboff, 1988, p. 395)
Consequently, to understand how elements of the workplace provide affor-

dances (Billett, 2001) for the construction of knowledge and skill it was
necessary to understand the activity of the workplace. The diary entry at the
beginning of this chapter indicates I took a very hands-on approach: par-
ticipant observation. The list of tools and methods available to a researcher
is extensive, so why did I consciously choose a method, which, considering
my field research sites, was bound to end up with some form of physical
punishment? It could be said that participant observation chose me.
The purpose of this chapter attempts to explain the experiences of a
researcher beginning her field research and putting into action the results of
the decision-making processes from the first 18 months of a doctoral study,
paying particular attention to the use of participant observation techniques
– a method that has shaped the project and the researcher irrevocably – as a
data collection tool rather than non-participant observation, which seems to
be the norm in researching and learning in the workplace.
THE CONTEXT
The aim of the research was to understand how elements of the workplace
provided affordances (Billett, 2001) for the processes of knowledge and skill
construction. To do this, young people training to be chefs were the focus to
illustrate how knowledge and skill is constructed during apprenticeship on
the trajectory towards becoming a chef. Apprentice chefs were chosen for
five reasons:
1. Labour is a key type of goal-directed activity, and it is through labour
that a kitchen functions to produce objects (meals), on time and within
budget, achieving the desired outcomes of satisfied customers and profit.
Observing apprentice chefs while they completed various tasks in the
production environment of a kitchen provided the potential to
To Cook or not to Cook 37
understand how learning is experienced and to identify the processes

involved in constructing knowledge and skill in the workplace;
2. Apprenticeship provided the opportunity to study an apprentice’s on-the-
job learning within an activity system (Engestrom, 1987) through direct
and indirect guidance (Billett, 2001);
3. The apprentice chefs in this study were all completing a Foundation
Modern Apprenticeship, which requires the completion of a National
Vocational Qualification (NVQ) at Level 2. NVQs can be completed
through a Private Training Provider (PTP) or through a day release
programme at a College of Further Education (F.E. College) conse-
quently the opportunity existed to examine off-the-job learning in an-
other activity system: a F.E. College.
4. Traditional apprenticeship was the only route to becoming a chef until
the Modern Apprenticeship programme was introduced in 1994. In re-
searching how apprentice chefs learn, the history of apprenticeship is of
significance with regard to the opportunities afforded the apprentice in
constructing his or her knowledge and skill; and
5. My background provided phenomenological validity to the study. I have
worked in the hospitality industry for a number of years and have taught
this subject in secondary schools in Australia; therefore this background
knowledge provided me with invaluable insight. My experience also pro-
vided me with a significant resource in negotiating access to sites: in times
of economic hardship, what chef is willing to turn down the prospect of a
month of free labour, while I work alongside the apprentice?
THE DECISION-MAKING PROCESS

Some of the most well-known research into learning in the workplace is
Lave and Wenger’s (1991) study of West African tailors whereby they argue
the tailors learn their trade by gradually becoming participants able to en-
gage fully as members of an existing community of practise. Similar work on
learning to move from a novice to an expert has been conducted by Billett
(1995a, 2001) focusing on hairdressers knowledge and skill construction,
Scribner’s (1999) work on dairy worker’s social construction of knowledge,
and Hutchin’s (1993) research into navigator’s learning their profession. All
these studies were conducted using non-participant observation techniques
combined with interviews or, in the case of Felstead et al. (2005) augmenting
case study work with a survey which highlights smaller, perhaps less
38 SUSAN JAMES
significant but no less important, changes in capability or understanding in

learning at work.
Consequently, to understand how learning is experienced in the work-
place, an initial methodology was constructed that was influenced by,
aligned with, and built upon previous research conducted by Billett (1995)
and Moore (1981). Billett’s research on how hairdressers structure their
knowledge through authentic activities in hairdressing salons using ethno-
graphic techniques in four case study sites greatly influenced my initial
thinking and planning. He used non-participant observation, semi-
structured interviews, a client survey and an activity-sheet to collect data
to investigate the social and situational effects of knowledge construction in
the workplace.
Moore and his team worked in The School for Experiential Learning
(SEL) to ‘develop a systematic approach to locating the pedagogy and cur-
riculum of experiential education’ (Moore, 1981, p. ii). The purpose of the
study was ‘to examine these various experiences and try to figure out a way
to analyse them in terms of education’ (Moore, 1981, p. 13) and involved 3
years of participant observation in over 35 sites where the students of the
SEL were placed as ‘apprentices’ and ‘interns’ over a set period of time,
while also spending a small portion of time attending ‘seminars’ and ‘in-
house’ courses at the school.
From reading Moore’s report, I decided that in order to understand the
knowledge and skills an apprentice chef constructs, I would need to analyse
the specific tasks assigned and completed by him/her. For ultimately, the
apprentice constructs his/her knowledge and skills by performing and ex-
periencing the various duties given by the other chefs in the kitchen, and the
conversations he/she is involved in. Furthermore, to understand the proc-
esses involved in an apprentice’s construction of knowledge and skill, it is
necessary to understand the workplace as a whole, and the apprentice’s
place within this establishment. Therefore, in order to see the apprentice
fulfilling his/her responsibilities, I would need to spend a lot of time with the
apprentice in the workplace.
THE PILOT STUDY

Access was negotiated with Rob, the head chef at The Hutch. The research
was explained to Rob and later to Nat, the apprentice, after which it was
agreed that I could spend a week observing the action in the kitchen. The
other chefs were informed of my presence in the kitchen, why I was
observing their work, and gave informed consent to be involved in the

research. The first few days of the week were spent honing my observation
skills, making decisions on what should be observed and trying to gain a feel
for the kitchen. I was able to make copious notes of the activity in front of
me as I stood in my corner of the kitchen.
However, as the week wore on I began to seriously question the authen-
ticity of the working day I observed. It is very unusual to have someone
standing or sitting around in a kitchen (Ruhlman, 1997, 2001). From having
worked in kitchens before – my phenomenological validity – I was aware
that kitchens are hectic, hot places fraught with tension that was more often
than not relieved through swearing, practical jokes and a lot of yelling (see
also Bourdain, 2000). In the first 3 days at The Hutch, I had not once seen
any of this tension relief. I needed to test my hypothesis as it was becoming
increasingly evident how painfully aware everybody in the kitchen was of
my presence, particularly my note-taking. Generally, the only people who
enter a kitchen to take notes are health and hygiene inspectors or assessors
of some kind. I needed the chefs to see me as part of the kitchen as opposed
to an outsider.
Jeffrey and Troman (2004) believe ethnographic research depends entirely
on the focus and site of the research and adhere to three ethnographic
principles:
1. research taking place over time to allow a fuller range of empirical sit-
uations to be observed and analysed and to allow for the emergence of
contradictory behaviours and perspectives. Time in the field, alongside
time for analysis and interpretation, allows continuous reflections con-
cerning the complexity of human contexts;
2. considering relations between the appropriate cultural, political and so-
cial levels of the research site and the individual’s and group’s/commu-
nity’s agency at the research site; and
3. including theoretical perspectives in order to: ‘sensitise’ field research and
analysis: provide an opportunity to use empirical ethnographic research
for the interrogation of macro and middle range theories; and to develop
new theory.
Using non-participant observation techniques at The Hutch was not
allowing me to adhere to any of the three principles. Consequently, I made
the decision to spend the second week of the pilot study as a participant
observer. Of course, I could have stayed for longer than a month. Feasibly,
how long could they have kept up this facade? Jeffrey and Troman (2004)
call this ‘compressed time mode’ and believe that, while this type of shorter
40 SUSAN JAMES
time-span ethnography provides a snapshot of what is happening, it does

not uncover much of what is really occurring. I do not agree and this will be
discussed further below but for now, more pertinently, my research time-
table did not allow more time to be spent in the field research sites to find
out whether it was a facade.
In the second week there was a noticeable difference in the atmosphere of
the kitchen such as yelling, swearing and practical jokes, which had not
occurred the week before. Although the data was not as detailed as in the
previous week (because I was unable to record exact conversations), it was a
more accurate reflection of the daily experiences of the people working in
the kitchen. What is lost in detail is gained in validity because the data
collected using participant observation techniques provides a more accurate
account of the kitchens and the workers in it.
Furthermore, working collaboratively on tasks proved more conducive to
conversations with the apprentice and was a way of gleaning data that might
otherwise have stayed under the surface if using non-participant techniques.
As Robson (1993, p. 197, emphasis in original) states,
The fact that an observer is an observer is made clear from the start. The observer then
tries to establish close relationships with members of the group. This stance means that
as well as observing through participating in activities, the observer can ask members to
explain various aspects of what is going on. It is important to get the trust of key
members of the group (key either because of their position or because of personal
qualities such as openness or interest in the ways of the group).
During the time spent working as a participant observer with the appren-
tices, I asked questions regarding the tasks rather than ‘just getting on with
it’. The questions asked were used to ascertain how the tasks influenced and
promoted learning on the progression towards becoming a chef. These
questions helped to make the process of knowledge and skill acquisition
more explicit (Eraut, 2001). Having the apprentices primarily as key in-
formants rather than subjects to be observed – although on occasions the
apprentices were this also – is an important aspect to the validity of this
research (Hammersley & Atkinson, 1995, pp. 140–141) as the data is en-
riched by the apprentices’ input as key informants (Robson, 1993, p. 197;
Taylor, 1998, pp. 53–66) and not just subjects to be interviewed. The key
informants in each of the kitchens are presented in Table 1.
Moreover, not only did participant observation allow me to collect data
on the tasks the apprentice completed and his/her wider role and experiences
in the activity of the kitchen, it also allowed my own experiences in the
kitchen to enhance the data collected because, for all intents and purposes, I
was treated as an apprentice. For example, I had to peel 50 kg of potatoes
Table 1. Key Informants.

Kitchen Key Informants
The Hutch Nat – 1st year apprentice

Chives Jack – 2nd year apprentice
Daniel – 1st year apprentice
Gastronomique Lawrence – 1st year apprentice
Sebs Clint – 1st year apprentice
during my first week at Gastronomique. The other chefs laughed at the

brown colour of my hands from all the dirt. Barry, the head chef, walked
past and mumbled, ‘‘The same task I did in my first week as an apprentice
and the same task Lawrence did when he first started here.’’ My acceptance
into the kitchens was also exemplified by the hat I wore while working at
Chives, which was required by all the chefs for health and hygiene purposes.
The hats were used as a semiotic tool; the colour of the cap signifies one’s
status in the kitchen: the head chef had a black cap, sous chefs wore blue
caps, a chef de partie would wear a purple hat, and second year apprentices
wore red and white checked caps. To me, the fact I was given a yellow cap –
the same colour a first year apprentice wore – signified authenticity, espe-
cially as work experience students were given paper hats!
The decision to continue as a participant observer was an easy one; as I
said earlier, it chose me. The kitchen came alive in the second week of the
pilot study and as I was going to be in my next three sites for a month each,
working 6 days a week on the same shifts as the apprentice, I needed the
kitchens and the workers in them to be as natural as possible. Jeffrey and
Troman (2004) may think a month is not long enough, but 4 weeks in a
stilted environment, with people not being as they normally would be,
would feel like an eternity. I also felt that one month would be enough time
to begin to understand the routines and patterns of the workplace, to feel
the tempo and workflow of a shift while also preventing me from becoming
too familiar and just another worker. As a participant observer, the di-
chotomy between my dual roles of worker and researcher (Spadley, 1980,
pp. 54–58) became apparent during the time I spent in the field research
sites. On a number of occasions the role I played was decided for me, but
there were many times when I was making the choice to be one or the other:
a researcher or a worker. Often this was a difficult choice: do I let the
conversation continue as merely a co-worker and see what is unearthed; or
do I intervene as a researcher with a question to elicit a response? There were
times when both situations occurred – when I intervened and when I did not
42 SUSAN JAMES
– but I can never be sure as to whether the most appropriate choice was
made, as the data I collected is only going to reflect the decision made in that
instance. As Hammersley and Atkinson (1995, p. 227) argue,
One must be aware of the possible effects of social location on all kinds of data, including
ethnographers’ own observational reports: we too occupy particular locations and what
we observe, what we record, and how we interpret it will be influenced by these.
Being aware of the distinct yet problematic line between worker and re-
searcher made prioritising my roles easier in the second and third field
research sites: as much as I was a worker in the kitchens, I was first and
foremost a researcher who frequently had to interrupt conversations and ask
questions. However, while the decision to be a participant observer had been
an easy one the actual experience was not.
PARTICIPANT OBSERVATION: THE EXPERIENCE
Although I had worked in kitchens before, it had been several years, and
certainly not with research motives and goals in mind. No amount of reading
of the participant observation literature could have prepared me fully for
what I was about to experience. Participant observation means so many
different things in ‘real life’. There is much literature on being a participant
observer (e.g., Bogdewic, 1999; Burgess, 1984; Delamont, 2005; Friedrichs &
Ludtke, 1975; Hodkinson, 2004; May, 1993; Parker, 2002; Taylor & Bogdan,
1998), particularly in relation to schools and educational institutions (e.g.,
Delamont, 1992; Eisner & Peshkin, 1990; Holmes, 1998; Massey & Walford,
1998, 1999; Woods, 1986), and while this literature provided some valuable
information, often the practicalities and logistics of being a participant ob-
server are not discussed. And, as mentioned, participant observation does not
seem to be a method favoured by researchers of workplace learning.
Identity
I spent 1 month in each site. By my calculations, this amount of time would

allow the routines and rhythm of the workplace to surface. After the first
few days, a feeling of ‘never having been anywhere else’ began to settle:
routines and layout were learned, and I did settle into the rhythm of the
workplace. This is deceptive. It allows one a familiarity that it should not as
it allows one to forget the purpose of being in the field and to think of
oneself as a chef, and nothing could be further from the truth. For example,
having been at Chives for a week, I was asked to do a job with one of the
apprentices. Neither of us fully understood what was required but rather
than ask for clarification, something I would do as a researcher, the ap-
prentice and I went ahead and completed what we assumed was required
because we did not want to be yelled at. Although the food produced was
fine, it was a variation on what we had been told to do, and consequently a
tirade of abuse followed. The majority of this diatribe was aimed at the
apprentice but the full force of what was being said was felt by both of us!
My identity as an apprentice had overtaken my identity as a researcher and
manifested itself by not having the gumption to ask for clarification.
Although it was generally forgotten by those in the kitchen that I was first
and foremost a researcher (or perhaps, more accurately, it was kept in the
backs of their minds), my role as a researcher did come to the fore on
various occasions. During the course of the shift, I would ask the apprentice
a number of questions regarding the tasks. Sometimes an immediate answer,
with what seemed to be very little thought, was given. Other times, the
apprentice stopped and looked, pondering the question, making him/her
step away from the task at hand and aware of the situation from my per-
spective. The act of the apprentice thinking of an answer, rather than giving
an automatic reaction, brought my role to the fore.
Occasionally comments would be made such as ‘will you write that in
your book?’ and ‘no one else asks questions like that’ or the question would
need to be re-worded for comprehension, immediately re-establishing my
role as a researcher. For example, the head chef had spent quite some time
with Lawrence, the apprentice at Gastronomique, and me showing us how
to make a risotto. It was a particularly interesting example of instruction for
my research. Afterwards I asked Lawrence about the ‘fusion’ of two par-
ticular ingredients in a risotto he was shown to make. I then needed to
explain the word fusion. On many of these occasions, the incident brought
back to my mind the role I was there to play, as I forgot also. When I was
immersed in the middle of a busy service in a hot environment churning out
meals it was easy to forget the main purpose of my presence. During such
times there is no room for error and even less for reflection (Delamont,
2005), a disadvantage of participant observation. Reflection needs to take
place after the shift. As desirable as a notebook would be, and I did carry a
small one in my pocket, the risk of burnt food compared to time out for
note-taking was too high.
The feeling of involvement that familiarity breeds resulted in another
mistake. It is one thing for a chef to criticise an apprentice, but an entirely
different story for a pseudo-chef to ‘go native’. After a week of working with
44 SUSAN JAMES
Daniel at Chives (there are seven apprentices at this site) I was becoming more
and more frustrated with his lax attitude and frequent wandering, leaving me
to complete the task at hand. Service was complete, the cleaning had been
done and a few of us were standing around joking and laughing, analysing the
day, waiting for that big hand to hit 12. The apprentice, Daniel, was trying to
defend his actions from an incident earlier in the day and comments were
flying from every which way. This apparently was de rigueur until I made a
comment about the apprentices finely honed ability to walk off at crucial times
leaving me to complete his tasks. A few laughs went around but also some
raised eyebrows and looks. In those 10 seconds, I was reminded that I was not
fully an apprentice and quickly learned a very important lesson if I was going
to be able to continue to listen to conversations such as these: in certain
instances keep my mouth shut!
Language
This topic has two aspects. I will refer to the first as ‘professional language’
and the second as ‘colourful language’.1 Professional language is two-fold:
the professional language of a chef, and my professional language as a
researcher. The language of a chef is riddled with French words and also
industry specific words, which contain absolutely no meaning outside of a
kitchen. For example, ‘the Hobart’ is the mixer – Hobart is the actual brand
of manufacturer. Lucky for me, I was au fait with the majority of the words
from my previous experience in the industry. I say lucky because I do believe
that this could have been a point of continuous ridicule throughout my stay
in these kitchens had I not been up-to-date. Of course, as an ‘apprentice’ the
odd (very few and far between!) query is allowed and I do have to admit to
listening carefully to the apprentices when they asked for clarification.
The second aspect of professional language is my own professional lan-
guage: the language of a researcher. As technical as the language of a chef
may be, chefs do not tend to use words that are considered essential to an
academic (or an apprentice academic as the case may be). The work in the
kitchen may be going along fine until I asked a question and the person I
asked would look at me blankly and say ‘‘What does that mean? Speak
English woman’’ or ‘‘You can tell who has been to university’’. Naturally
this would be said loudly so all others around could hear and I was im-
mediately differentiated. Time would need to be taken before I was ‘one of
the workers’ again. I soon learned to curb my professional language and it
was amazing how much more credible I became when I also sprinkled my
conversation with a few choice ‘colourful’ words.
In a kitchen, swear words seem to be used as nouns, verbs, adjectives and

even prepositions on occasion. It was expected to pepper one’s vocabulary
with as much colourful language as possible. By eventually joining in, the
doorway into the kitchen and its activity seemed to open just that little bit
wider. In fact, at Chives, I was informed that a bet had been taking place as
to when I would crack and swear for the first time. This was very remi-
niscent of school playground games, but nonetheless important to my being
a participant observer and accepted in the workplace.
Uniform
In the pilot site I was lent a pair of chef’s trousers to wear during my time in the
kitchen. Apart from being very comfortable, the trousers had a big, black check
pattern on them. The trousers were the same as the ones worn by everyone else
in the kitchen. Due to this similarity I did not think anything of it.
At Chives (the first field research site after the pilot study) I was informed
by the administrator (not the head chef) when negotiating access that I
would be required to provide my own uniform. Not a problem I thought. I
looked in the yellow pages for a supplier, made a phone call and ordered a
chef’s jacket and trousers, feeling very excited at having my own uniform.
The lady on the other end of the telephone inquired as to what type of
trousers I would like. I asked her to explain and she rattled off a number of
different styles. I vaguely remembered seeing the head chef wearing blue and
white but could not be sure. I decided on blue and white large check trou-
sers. Turning up on my first day I was shown the female change room. As I
emerged all the other chefs stared. I froze: what was wrong, why were they
looking at me like that? One of the chefs said ‘‘Look, it’s Jamie Oliver!’’ I
then realized they were all wearing SMALL blue and white checked trou-
sers. For the next 2 days I fielded questions on my choice of apparel.
Thankfully, after a couple of days the jokes died down BUT it was still one
thing that I did not need to deal with and could have potentially averted if I
had been more aware. I have since read Linda Measor’s (1985) article,
‘Interviewing: a strategy in Qualitative Research’ whereby she does discuss
the different clothes she wore when interviewing teachers and students.
Why, oh why, had I not read it earlier?
Acceptance
As a participant observer one wants, as much as possible, to be a part of

what is happening, to try to understand all that is occurring and the only
46 SUSAN JAMES
way to be a part of all that is happening is to be accepted. If I was to try to

gain a true understanding of what an apprentice learns, how he/she becomes
a chef through constructing knowledge and skill and what it means to be a
chef, I needed to be accepted to the fullest extent possible. Jack, one of the
apprentices at Chives, accepted me immediately. He explained the goings on
and seemed to revel in his teacher role. This allowed me to be accepted by
the other chefs much more readily as they could see how well we were
working together.
However, there was one chef at Chives who did not begin to speak to me
until the fifth day. This was difficult considering on three of these days I was
working with Daniel, another apprentice in this kitchen, and we were under
that particular chef’s direction. He communicated to me via Daniel. Even-
tually this changed and by the end of the second week full communication
was occurring. I did ask him why he had taken so long to speak to me and
his gruff reply was ‘I needed to be sure of what you are about’. Indeed, he
went on to become one of the most helpful chefs in that kitchen.
Generally though, within the first few days of being in the kitchens, once
the other chefs were sure of why I was there and were comfortable with the
fact that I was not judging them or their work in any way, tension dissolved. I
believe this was aided by the fact that I was clear from the start that the chefs
were able to ask me questions. Also, on the second day in each kitchen, I took
my field notes in from the first day to show the apprentice, to gauge his
impression of the day and whether our conclusions of the day were the same.
This was a particularly interesting form of collaboration with my key in-
formants and, I believe, helped to aid the reliability and validity of my study
as I was trying to ensure the data was as accurate as possible. A few of the
other chefs had a look and asked questions. I had my notes on subsequent
days but no one except the apprentice was interested after the first day.
Participant observation allows the researcher an authenticity to the phe-
nomena that is incalculable. As previously mentioned, I felt the quality of
the data I collected was greatly improved in the pilot site once my position
moved from non-participant observer to participant observer. Evidence for
this is the fact that in Chives and Gastronomique I was yelled at, joked
about, ridiculed, given and shown tasks, just as the apprentice was. One
example illustrates this point perfectly. My knife skills were rather rusty.
Consequently, by the end of the first week at Chives I had blue water proof
plasters on three of my fingers. For the next 2 days I had the ‘blue plaster
song’, a song made up by the chefs in that kitchen, sung (taunted) to me
when I entered the kitchen. I was, for all intents and purposes, treated as an
apprentice.
Initiation procedures
Undoubtedly, part of being accepted was the initiation procedures I had to

pass. One of these, I believe, was the colourful language aspect already
mentioned. Another was being the butt of many jokes in the first few days
and not rising to take the bait. Although more females are appearing in
kitchens, these workplaces are still very male dominated areas. Testosterone
was abundant, as one would expect with the chefs all vying for a place on the
pecking order and letting their place be known. Tolerating sexist jokes was
an aspect that had to be endured for ultimately my dual role of researcher/
chef was tenuous. One false move and I could very easily have been asked to
leave. Unfortunately, this meant smiling at tasteless comments and putting
my usually forthright demeanour on the backburner.
A far more physically painful test was the slicing of lemons, limes and
oranges into garnishes when it was common knowledge that I had a lot of
nicks and cuts on my hands. The chefs in these kitchens needed to test me. I
was encroaching on their territory and although they knew to a certain
extent what I was about, I was learning, or was about to learn, a lot more
regarding them and their work lives than these chefs were going to learn
about me. I needed to earn their trust and the most suitable way for the
chefs was through these tests. The 50 kg of potatoes mentioned earlier was a
test: did she have enough stamina and patience to peel her way through
them all? Could she be trusted to perform other tasks in the kitchen and
ultimately enter our/their domain?
TO COOK OR NOT TO COOK?

These areas I have highlighted are those that I was least prepared for in
fulfilling my role as a participant observer: knowing where boundaries for
my participation in conversations were, the way my identity as a chef and
researcher would develop, being accepted and belonging (or not), language,
initiation tests, and my trousers. It could be argued that these issues are
situation specific: a researcher conducting field research in a school would
not have to face these matters. Yes, to a certain extent the researcher in an
educational institution may not be faced with these specific challenges but
there would be issues of language, working out in one’s own mind their
position as participant observer, issues of acceptance, and uniform among
others. The words of Sarah Delamont (2005, p. 310) ring in my mind:
48 SUSAN JAMES
Good ethnographic fieldwork is characterised by ruthless, relentless, continuous self-

critical reflexivity (Coffey, 1999). Good ethnographic writing is also reflexive, and plays
with paradox (Atkinson, 1990, 1992). And, of course, the whole point of fieldwork is that
the ethnographer learns new things, develops new skills, masters novel vocabularies,
proxemics and rhythms, as well as doing the research (Delamont, 2004).
Delamont (2005), along with Becker (1971) and Wolcott (1981), have long
argued that educational research needs challenges to its familiarity and I
agree with them. Being a participant observer provided many challenges, as
outlined above, and participant observation, by its very nature of course,
will not be suitable in all research situations as Jeffrey and Troman (2004)
suggest but it should not be readily dismissed as time-consuming, being
possibly most suitable for doctoral research (Walford, 2002). So much of the
learning that occurs in workplaces, and the knowledge and skill that is
constructed through the everyday practices of work, becomes implicit and
tacit (Eraut, 2000a, 2000b, 2001). Using participant observation techniques
can help to make our understanding more explicit and need not take more
time than non-participant observation but can provide much more reward-
ing results for the research and the researcher.
NOTE
1. Colourful language is the term I use to describe the ubiquitous swearing.
REFERENCES
Becker, H. (1971). Sociological work: method and substance. London: Allen Lane.
Billett, S. (1995). Structuring knowledge through authentic activities. Brisbane: Griffith Univer-
sity.
Billett, S. (1996). Constructing vocational knowledge: History, communities and ontogeny.
Journal of Vocational Education and Training, 48(2), 141–154.
Billett, S. (2001). Learning in the workplace: Strategies for effective practice. Crows Nest, NSW:
Allen & Unwin.
Billett, S. (2002). Workplace pedagogic practices: Co-participation and learning. British Journal
of Educational Studies, 50(4), 457–481.
Bogdewic, S. (1999). Participant observation. In: B. Crabtree & W. Miller (Eds), Doing qual-
itative research. Newbury Park, CA: Sage.
Boud, D., & Garrick, J. (Eds). (1999). Understanding learning at work. London: Routledge.
Bourdain, A. (2000). Kitchen confidential. London: Bloomsbury Publishing.
Burgess, R. (1984). Methods of field Research 1: Participant observation. In: R. Burgess (Ed.),
In the field. London: Allen & Unwin.
Delamont, S. (1992). Fieldwork in educational settings: Methods, pitfalls and perspectives.

Lewes: Falmer Press.
Delamont, S. (2005). No place for women among them? Reflections on the axé of fieldwork.
Sport, Education and Society, 10(3), 305–320.
Eisner, E., & Peshkin, I. (Eds). (1990). Qualitative enquiry in education. The continuing debate.
New York: Teachers College Press.
Engeström, Y. (1987). Learning by expanding. An activity-theoretical approach to developmental
research. Helsinki: Orienta-Konsultit.
Eraut, M. (2000a). Non-formal learning and tacit knowledge in professional work. British
Journal of Educational Psychology, 70, 113–136.
Eraut, M. (2000b). The dangers of managing with an inadequate view of knowledge. turning
knowledge management downside up. Sao Paulo: 3rd International Conference for So-
ciocultural Research.
Eraut, M. (2001). Transfer of learning between education and workplace settings. Northampton:
TLRP International Workshop, University College.
Eraut, M., Alderton, J., Cole, G., & Senker, P. (1998). Development of knowledge and skills in
employment. Brighton: Institute of Education, University of Sussex.
Felstead, A., Fuller, A., Unwin, L., Aston, D., Butler, P., & Lee, T. (2005). Survey the scene:
Learning metaphors, survey design and the workplace context. Journal of Education and
Work, 18(4), 359–383.
Friedrichs, J., & Ludtke, H. (1975). Participant observation. Farnborough: Saxon House.
Fuller, A., & Unwin, L. (2003a). Learning as apprentices in the contemporary UK workplace:
Creating and managing expansive and restrictive participation. Journal of Education and
Work, 16(4), 407–426.
Fuller, A., & Unwin, L. (2003b). Creating a ‘Modern Apprenticeship’: A critique of the UK’s
multi-sector, social inclusion approach. Journal of Education and Work, 16(1), 5–25.
Fuller, A., Hodkinson, H., Hodkinson, P., & Unwin., L. (2005). Learning as peripheral par-
ticipation in communities of practice: A reassessment of key concepts in workplace
learning. British Educational Research Journal, 31(1), 49–68.
Hammersley, M., & Atkinson, P. (1995). Ethnography: Principles in practice. London:
Routledge.
Hawke, G. (1998). Learning, workplaces and public policy. Sydney: 1st Annual Conference of
AVETRA.
Hodkinson, P. (2004). Research as a form of work: Expertise, community and methodological
objectivity. British Educational Research Journal, 30(1), 9–26.
Holmes, R. (1998). Fieldwork with children. London: Sage.
Hutchins, L. (1993). Learning to navigate. In: S. Chaiklin & J. Lave (Eds), Understanding
practice: Perspectives on activity and context (pp. 35–63). Toronto: Cambridge Univer-
sity Press.
Jeffrey, B., & Troman, V. (2004). Time for ethnography. British Educational Research Journal,
30(4), 535–548.
Lave, J., & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation. Cam-
bridge: Cambridge University Press.
Massey, A., & Walford, G., (Eds). (1998). Children learning in context, studies in educational
ethnography vol. 1. Stamford, CT: JAI Press.
Massey, A., Walford, G. (Eds). (1999). Explorations in methodology, Studies in educational
ethnography vol. 2. Stamford, CT: JAI Press.
50 SUSAN JAMES
May, T. (1993). Social research. Issues, methods and process. Buckingham: Open University
Press.
Measor, L. (1985). Interviewing: A strategy in qualitative research. In: R. Burgess (Ed.),
Strategies of educational research. London: Falmer Press.
Moore, D. (1981). The social organisation of educational encounters in nonschool settings.
Chicago: Centre for New Schools.
Parker, A. (2002). Pressures, problems and the PhD process. In: G. Walford (Ed.), Doing a
doctorate in educational ethnography. London: JAI.
Robson, C. (1993). Real world research. Oxford: Blackwell.
Ruhlman, M. (1997). The making of a chef: Mastering heat at the Culinary Institute of America.
New York: Henry Holt and Company.
Ruhlman, M. (2001). The soul of a chef. New York: Penguin.
Scribner, S. (1999). Knowledge at work. In: R. McCormick & C. Paechter (Eds), Learning and
knowledge. London: Paul Chapman Publishing.
Spadley, J. (1980). Participant observation. London: Harcourt Brace Jovanovich.
Taylor, S. (1998). Introduction to qualitative research methods. A guidebook and resource. New
York: Wiley.
Walford, G. (Ed.) (2002), Doing a doctorate in educational ethnography, Studies in Educational
Ethnography (Vol. 7). Oxford: JAI Press.
Woods, P. (1986). Inside schools. Ethnography in educational research. London: Routledge and
Kegan Paul.
Zuboff, S. (1988). In the age of the smart machine: the future of work and power. Oxford:
Heinemann Professional.
RESEARCHER ROLES IN A
SCHOOL-BASED ETHNOGRAPHY
Nick Hopwood
INTRODUCTION
The roles that ethnographers adopt in their fieldwork are ‘‘perhaps the
single most important determinant of what he [or she] will be able to learn’’
(McCall & Simmons, 1969, p. 29). My purpose in this paper is to demon-
strate that these roles can be in a state of rapid flux, depending not only on
who the researcher is interacting with, but also on a complex system of
constantly changing settings for those interactions.
To this end I explore different ‘territories’ associated with my own school-
based ethnographic research, each constituting a particular setting in which
I adopted specific roles. Territories are not simply spatial areas, although
location is often an important factor in determining which roles are appro-
priate. Particular places are often used in different ways by different people
at different times in schools, and this has implications for the sorts of roles
researchers can or should adopt. The concept of territories as temporally,
socially and spatially defined settings can be used as a lens to explore de-
tailed aspects of the nature of research sites, providing a fresh understanding
of the roles ethnographers adopt in their fieldwork.
I provide contextual information about my research before considering
existing frameworks for understanding roles in ethnographic research. The
main body of this paper is then devoted to exploring in detail the roles I

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12004-5
51
52 NICK HOPWOOD
adopted in each of five different territories in school. Finally, I suggest that

the process of examining the territories of research sites has considerable
potential for enabling ethnographers to explore and exploit researcher roles
as means to access information.
CONTEXT
This paper draws on a study of how English year nine pupils (aged 13–14)
experience and conceive school geography. Fieldwork began with three
months of piloting, during which I visited the same school two or three days
per week, observing geography lessons, spending time in the playground,
staffroom and geography office, and talking to pupils about geography. I
observed over 70 lessons, taught by several different teachers, and spent over
240 hours in school just in this period of piloting.
Formal fieldwork was conducted (between April 2004 and June 2005) in
three different secondary (11–18) co-educational comprehensive schools
within a 25-mile radius of Oxford. Each consecutive period of fieldwork
lasted over three months, with two or three visits per week, most of which
took a full day. My activities were more structured than during piloting, and
in each school over 30 lessons attended by the same class and taught by the
same teacher were formally observed. Selected pupils were interviewed
about these lessons, and participated in four additional techniques: two
photo-elicitation interviews (for one of which pupils took photographs
themselves), a concept mapping exercise, and a structured interaction in
which pupils talked about a series of questions and how they relate to school
geography.
Surrounding these targeted methods of data generation were a series of
less tightly focused activities, including varying degrees of participation in
and observation of geography offices (departmental staff rooms) and areas
open to pupils during recreation (including classrooms). These wider eth-
nographic features offered opportunities for both data generation and for
developing rapport with participants, which then supported my more fo-
cused activities.
On each school visit I would normally participate in at least two lessons in
addition to the one that was the focus of my research for the day, adopting a
role that can be generally described as a classroom assistant (as was ap-
propriate given that I am not a qualified teacher). In between lessons I
would generally sit in the geography office, but also spent some time in
classrooms, dining halls and playgrounds.
Researcher Roles in a School-Based Ethnography 53
During this formal fieldwork I spent a further 700 hours in school. In-
teractions with people from the schools extended beyond school hours either
through my getting a lift to and from school with a teacher, or through
walking to the bus stop and taking a bus with pupils. The boundaries of my
research were further blurred because I have lived with one or more teachers
throughout the three years of this study and learned a lot about schools,
teachers, pupils, teaching and learning by listening to (or sometimes trying
to ignore) my housemates’ frequent complaints, moans and general verbal
purging of aspects of their working lives.
RESEARCHER ROLES IN ETHNOGRAPHY
Gold (1958) describes the roles adopted in the field as, simultaneously, de-
vices for securing information and sets of behaviours in which the researcher
him- or herself is involved. Such an understanding of research roles as both
crucial to data generation and highly personal forms of engagement in re-
search sites underpins the argument of this paper.
While roles are undoubtedly significant in many forms of social research,
they are particularly important in ethnography, wherein the researcher is
typically involved in fieldwork settings for a greater period of time, and
often more deeply than in other methodologies. The endurance of ethnog-
raphy offers opportunities for ethnographers to adopt a broad range of
roles, many of which may be dependent on levels of trust and familiarity,
which often require and result from sustained presence in the field. Notions
of endurance are also relevant to the ethnographer’s experience: ethnogra-
phy is highly intensive and demanding because roles are often hard to un-
derstand, difficult to enact, and at the same time crucial to the research
endeavour.
For several decades ethnographer’s roles have been considered and the-
orised in the context of perceived shortcomings in our understanding of
different roles. The aim of this paper is to offer a new way of thinking about
field roles in ethnographic research that avoids some of the limitations as-
sociated with existing frameworks.
Gold (1958) and Junker (1960) discuss the different sorts of roles that
ethnographers might adopt in the field in terms of a fourfold typology:
complete observer, participant as observer, observer as participant, and
complete participant. Although these ideas are useful in the way they sign-
post a range of roles and encourage researchers to think carefully about
54 NICK HOPWOOD
which roles are most appropriate given different settings and purposes, they
have been subject to criticism.
Janes (1961) suggests that ethnographers’ roles are not static but can
change over time, and within the general descriptor of ‘participant ob-
server’, he charted a sequence with increasing degrees of acceptance from
newcomer to imminent immigrant. Notions of gradual shifts in researcher
roles over time proved popular, and Oleson and Whittaker (1967; cited in
Walford, 1987) use this principle to describe evolving forms of exchange
between researchers and participants. While introducing ideas that roles can
change over time, these frameworks focus on long-term linear changes, and
imply that at any one time the ethnographer’s role can be characterised by
one particular form of interaction in a field site.
Despite these limitations, Walford (1987) notes that such simplistic un-
derstandings of roles have had a strong legacy. Gold’s (1958) typology has
been used by Punch (1998) and Robson (1993) to describe the range of in-
depth qualitative field roles, while Burgess (1984) describes his roles in
school-based research in terms of Janes’ (1961) framework.
Snow, Benford, and Anderson (1986) offer a different typology of field
roles (controlled skeptic, ardent activist, buddy-researcher, credential ex-
pert), each of which they associate with particular strengths and weaknesses
with respect to data acquisition or ‘informational yield’. They, like Gold and
Junker, suggest that researchers adopt one role in a particular project (their
typology is derived from their own experiences of doing ethnographic re-
search), and do not accommodate either the possibility that researchers’
roles can change (develop) over time, or that researchers can adopt a variety
of roles and switch between them rapidly.
Goffman (1959) provides one of the earliest examples of a more subtle
analysis of researcher roles. Rather than suggesting that researchers adopt
monolithic roles depending on the research site, or that roles follow a linear
evolution over time, he examines the nature of fieldwork sites. From this he
suggests that researchers might have to adopt different roles in order to
access the different social activities and settings that co-exist within the field
site. Of particular note is the distinction he makes between ‘frontstage’ and
‘backstage’ settings.
This paper builds on Goffman’s (1959) ideas that social activities are
often separated temporally or spatially, and I will show that research sites
may incorporate multiple frontstage and backstage arenas within a complex
collection of territories.
Walford (1986, 1987) describes a range of roles that he adopted within his
school-based ethnographic research. He notes for example how the role of
‘observer’ functioned well in lessons, but was ‘‘only acceptable within tightly
defined circumstances’’ (p. 60), notably not around the playground or
quadrangle. He also discusses different roles he negotiated with pupils, and
a further set of roles with teachers, and it is clear that his roles changed
depending on location, time and social setting. The Junior Common Room
presented challenges by virtue of its pupil-dominated quality, while informal
activities with teachers such as dinner parties or pub visits offered oppor-
tunities for different forms of interaction. Different roles permitted occa-
sional forays into more subversive aspects of school life, for example joining
pupils in using a stolen key to ‘break in’ to out-of-bounds areas in the
middle of the night (Walford, 1986).
However, the thrust of Walford’s (1987) argument was to demonstrate the
weakness of existing thinking about field roles in terms of their tendency to
portray them as something that researchers are the sole determinants of. He
concludes:
The roles open to me were severely restricted by the expectations of those being re-
searched. Moreover throughout the time at the schools I felt obliged to present an
‘overarching role’ which was somewhat in conflict with my own feelings. (p. 62)
Unlike Walford I did not find that my role-taking was severely restricted,
and I felt no obligation to present an overarching role, indeed both teachers
and pupils treated me in different ways at different times, offering me ways
to diversify the roles I undertook. This paper considers the association be-
tween roles and territories in my own experience of doing ethnographic
research in schools. The aim is to describe the concept of research territories
and to illustrate its utility as a framework for understanding fieldwork roles
such that other ethnographers might be encouraged to think through the
nature of their research sites and the implications this might have for the
roles they adopt, and ultimately, the information they acquire.
THE TERRITORIES OF RESEARCH IN SCHOOLS

In the following sections I discuss five different territories I have identified
with respect to my research, each of which can be defined in terms of spatial,
temporal and/or social characteristics. I consider my interactions with
teachers first, pupils second, although my interactions with these two groups
were complicated because:
(i) Teachers and pupils were not always separated, and I was often in
situations where I was interacting simultaneously with both.
56 NICK HOPWOOD
(ii) The ways one group perceived my interactions with the other group may
have influenced the way they engaged with me as a participant in
different aspects of their school lives.
The five territories considered below are as follows: (1) interactions with
teachers in the geography office; (2) interactions with teachers in the class-
room during formal lesson time; (3) interactions with pupils in the class-
room during formal lesson time; (4) interactions with pupils in the
classroom during fringe time; and (5) formal researcher-initiated structured
interactions with pupils. This is not an exhaustive account of all the ter-
ritories I encountered (I might add playgrounds, buses, car journeys), but it
is sufficient to illustrate both the limitations of existing writing about eth-
nographic roles and the utility of territories as a conceptual lens through
which to explore these issues.
Territory 1 – Interactions with Teachers in the Geography Office
In each school the geography teachers shared an office in which they spent
considerable amounts of time. These teacher-only spaces were characterised
by rapidly changing activities and forms social interaction, switching almost
instantly from a backstage feel (informal, private chatter) to behaviours
more suggestive of a formal frontstage setting.
During break times, pupils would often knock on the door with problems
(lost detention slips, books, coursework, shoes), requests (rescheduling of
detentions, extensions to deadlines) or documents (parental reply slips, re-
port cards). The opening of the door to a pupil would often trigger a no-
ticeable change in the atmosphere of the office and the behaviour of
teachers. They would refer to each other as Mr or Miss X, and would curtail
any conversations or activities that might undermine a formal (frontstage)
appearance to pupils. Once the door was closed and pupils shut out, ac-
tivities quickly regained their backstage character, signified by the contin-
uation of gossip or moaning about pupils from the point at which they had
been interrupted.
Even this small, physically bounded space, with its clear social boundaries
(the exclusion of pupils) was thus host to a variety of social practices which
required me to adopt a range of roles as and when appropriate.
Experience of piloting taught me that doing nothing in these contexts was
out of character and while not the cause of friction, served to distance me
from teachers. Walford (1987) notes:

I realised that it was important to ensure that staff and pupils knew that I was actually
keeping myself busy. Most ethnographic fieldwork techniques which involve simply
talking, observing and listening to people y were simply not seen as work by others. (p.
56)
While doing nothing, or appearing to do nothing was not an option, sitting

notebook in hand, avidly scribbling notes, seemed to make teachers un-
comfortable and create a stilted atmosphere. I learned that the most ap-
propriate and effective form of engagement in these contexts involved my
doing some activity (perhaps reading and making notes from a journal
paper and occasionally scribbling short field notes) that kept up norms of
busy-ness but which also maintained my position as available to join in
conversations and/or ready to help a teacher out.
In this way my roles fitted ongoing practices – I was busy working or busy
gossiping, whichever matched the general activity of the office at a particular
moment. Shifting from one role to another involved partial detachment
from those roles and being aware of changes in interactions between teach-
ers that warranted a change in my behaviour. These were sometimes obvious
switches between formal and informal (front/backstage) behaviours, but
were also more subtle, involving an unannounced consensus that, for ex-
ample, gossip had come to an end and it was time to get some serious
marking done.
Gossiping with teachers had both benefits and drawbacks. It was a key
facilitator of interactions which led to the development of rapport between
me and staff, and particularly between me and the teacher whose classes
formed the focus of my research. Informal chatter had a positive feedback
mechanism in that the more I came to know about the schools (the differ-
ence between a blue and green slip) and members of staff (their good days,
bad days, good sets, bad sets, general gripes and so on), the more easily I
could engage in office discourse.
In these moments my role was close to that described by Gold (1958) as
‘participant as observer’, in that I sought to gain the trust of teachers, and
although never actually adopting the role of teacher, my role increasingly
appeared in many respects to be more that of colleague rather than outsider.
I reminded teachers of my status as researcher, by taking out my notebook
from time to time, or by explicitly raising issues relating to my research
(such as negotiating a space for me to conduct interviews at lunchtime).
In order to become more colleague and less outsider I seized opportunities
to ‘side’ with teachers. I might agree with a teacher that pupils had been very
58 NICK HOPWOOD
well behaved in one lesson, or that a particular pupil had demonstrated

impressive thinking skills. I might also share his or her frustrations with the
prospect of a year 7 class on a windy Friday afternoon. This enabled me to
develop a shared and valued social identity within the geography office
context.
It was here that I drew strongly on what I had learned from living with
teachers for several years, and made up the deficit in my understanding of
schools that results from my status as a non-teacher. Having long endured
the blow-by-blow accounts of particular lessons freely offered on a daily
basis by my housemates, I was armed with some knowledge of the issues
teachers think are important, an understanding of the pressures and frus-
trations of school life, and a sense of how to say the right things. Everyone
(i.e. all teachers) knows that year 7s can’t be expected to concentrate for
more than a couple of minutes in the afternoons, everyone knows that Fri-
days are the worst in this respect, and everyone understands the hyperac-
tivity-inducing effect of windy weather on younger pupils.
The longer I remained in each school, the more involved I became in
various different activities and practices. This evolution of practice was
documented in my field notes: as I became a familiar face and teachers got
to know me I’d be asked to run off some photocopies before a lesson, carry
exercise books from room to room, find a particular video, make the tea, or
go to the shop to buy biscuits. In some cases I added up pupils’ coursework
marks, entered data, and even helped develop resources for teaching and
learning. Thus, I do not wish to challenge the arguments (discussed above)
that researcher roles can evolve over time. What I hope to demonstrate is
that a picture of a single role following a linear trajectory may be an over-
simplification.
While changes in roles are typically considered as attrition-like shifts that
reflect the gradual building up of trust and rapport, I experienced particular
moments after which my roles dramatically changed. This may be illustrated
by considering a day when the geography staff in one school learned that their
office was to become a classroom and they had to move everything that
afternoon to what had previously been the storeroom. Not only did I spend
several hours carting boxes, but as the only person who wasn’t teaching, I
became the only person who knew exactly where everything was. From then
on when things couldn’t be found the default option was ‘‘ask Nick’’, a re-
peated practice that solidified my rapport with teachers and significantly al-
tered the roles I adopted thereafter (essentially even more free-flowing gossip).
I was constantly conscious of the ethical implications of my role in ge-
ography office life, and of the boundaries that this imposed on forms of
practice I could adopt. Office gossip had a tricky characteristic of changing

rapidly from relatively benign subjects to others, which were more personal,
confidential and even insulting. While I jumped at some chances to side with
teachers, at other times this was less appropriate: I could not agree that Ben
was ‘‘the thickest person I’ve ever met’’ that Dom’s parents ‘‘should never
have been allowed to have children’’, or that the head is a ‘‘pathetic’’ man,
and the safest practice was to avoid entry into such discussions, either by
changing the subject of conversation, or by remaining busy-in-work rather
than becoming busy-in-gossip.
My actions and comments in the geography offices were carefully
bounded in other ways. As well as setting myself rules as to what I would
and wouldn’t discuss, I was clear with teachers and myself about what I
would or wouldn’t do. I explained that I did not want to become involved in
formal administration (dealing with detention slips, for example), or to be
set up as a user on the computer system (computers were rare enough in
these rooms and I felt they should be left for teachers). I did, however, agree
to answer the phone, and on occasion, meet teachers’ requests to say they
weren’t present so that they wouldn’t be disturbed.
The geography office constituted a challenging territory because interac-
tions with teachers within this small space were complicated by the con-
stantly changing nature of the forms of behaviour and social interaction the
teachers themselves were engaging in. My associated adoption of different
roles, and my rapid shifting between them strike at the heart of my argu-
ment that ethnographic field roles should be considered with greater sen-
sitivity than is typical in the literature. The next section shows how I
adopted other roles in interacting with the same group of people in a formal
classroom setting.
Territory 2 – Interactions with Teachers in the Classroom During Formal

Lesson Time
While the geography office was a space in which teachers could (sometimes)
let their guard down, in the classroom teachers were always ‘in role’, and
formal lesson time was always a frontstage setting. It was also frontstage for
me, as I was a visible ‘other’ in the classroom, subject to the gaze and
expectations (whether I met them or not) of teachers and pupils.
Maintaining existing codes of practice, I referred to teachers as Mr or
Miss X, rather than on first name terms as I would in the geography office.
There were obvious limitations to both the extent and format of interactions
60 NICK HOPWOOD
between myself and teachers during formal lesson time, as the teachers’
priority was, of course, to structure and guide pupils’ learning and behav-
iour rather than to talk to me. Interactions between us were almost always
short and of a formal nature. For example, I might ask quick points of
clarification about a task, or a teacher might ask me to close a blind.
Teachers also held quiet conversations with me about particular pupils or
their work, but very few of these interactions could be deemed closed in the
sense that there were always pupils nearby, and I remembered from my own
experiences as a teenager that pupils can choose to acquire acute hearing skills
when they judge that teachers are trying to be discrete. This illustrates points
(i) and (ii) made above about the complexity attendant with the simultaneous
presence of two social groups and the perceptions that one has of the other.
Many aspects of my roles in formal lesson time involved interaction with
pupils (discussed below), but my participation in classroom life was often
initiated by teachers and conducted in a way that, within the existing social
hierarchy of the school, distinguished me from pupils, and rendered the roles
I adopted closer in feel to those of a classroom assistant.
I explained to teachers that my aim was to be as helpful as possible in
lessons. One took this on board immediately and on my first day in school
asked me to act as ‘dragon’ in a lesson based on the TV programme
Dragon’s Den. This was an ideal opportunity to show shared values with the
teacher and demonstrate the sort of practices I might undertake as a re-
searcher in their school.
This illustrates the way teachers frequently engineered my role in lessons
in ways that suited them. I often undertook menial tasks that enabled the
teacher to concentrate on teaching and pupils on learning. I distributed and
collected worksheets and atlases, helped pupils fold and unfold OS maps,
and gave out the glue. More infrequently I became involved in the sub-
stantive activities of the lesson; for example, being ‘banker’ in a trading
game or ‘Chris Tarrant’ in a Who Wants to be a Millionaire? plenary task.
These instances of my involvement in formal classroom activities involve
elements of what Snow et al. (1986) call an ‘ardent activist’ role. While
neither my research sites nor questions were highly politicised, this idea is
relevant insofar as I embraced the ideology, goals and rhetoric of existing
classroom practice (it seemed reasonable to share the teachers’ concerns for
pupils’ learning). The voluntary assumption of responsibilities (although
minor in my case) and intensive participation are also among the activities
listed by Snow et al. in describing this role.
As in the geography office, my roles in formal lesson time were carefully
bounded. I never led lessons or aspects of them, and I firmly requested with
each teacher that I not be directly involved in behaviour management or

maintenance of discipline in lessons. I did not report off-task behaviour to
teachers and did not overtly engage in corrective or positive behaviour
management. On occasions when teachers addressed the whole class or
particular pupils with regards to discipline I avoided eye contact with pupils
and stood at the back of the classroom. This non-disciplinarian boundary
was less about my relationship with the teacher (although I believe it was
important to set out clear areas of territory in which the teacher’s authority
and roles were maintained), and more about my relationship with pupils.
Other researchers have reported feeling restricted in the roles they can
adopt, particularly in formal settings, and I did expect some resistance from
teachers with respect to my non-disciplinary role, especially as it evolved
into active participation in subversive classroom life (see below). Fortu-
nately my experience was less problematic, and I was able to adopt the roles
I felt were most appropriate and likely to help generate useful data. In
negotiations with teachers I emphasised my lack of teaching qualifications,
and in practice I exercised caution by avoiding certain forms of interaction
with pupils that teachers might embrace less readily until I had been around
for some time and was better able to judge where the boundaries around my
roles with pupils should lie.
Territory 3 – Interactions with Pupils in the Classroom During Formal

Lesson Time
My roles in formal lesson time were not confined to interactions with

teachers, indeed a significant aspect of being in geography lessons related to
adopting appropriate roles in interaction with pupils.
When teachers were addressing the whole class I typically sat at the back
of the room (usually on a window ledge or table) making field notes in a
school exercise book. This role of detached observer (Anderson & Burns,
1989) was appropriate at these times as it allowed the teachers to maintain
primary authority in the classroom, and gave clear signals to pupils that I
was not a teacher.
When pupils were on task I engaged in numerous spontaneous interac-
tions with pupils. These could involve approaching a pupil who had their
hand up, clarifying instructions or giving them resources they needed to
complete a task (such as paper if they had forgotten or lost their books).
Some pupils made it clear that they didn’t want assistance from me (either
by telling me, or by turning away when I approached them and calling for
62 NICK HOPWOOD
the teacher), but most actively sought my attention, some seeming simply to
enjoy showing me what they’d done. I initiated some interactions by asking
pupils to tell me what they were doing, or explain something about their
work to me.
In addition to these ‘on-task’ interactions, I also adopted more subversive
roles, although these were subject to carefully constructed boundaries. Often
rapport could be developed with pupils simply by making it obvious that I’d
seen them misbehaving but making it equally obvious that I wasn’t going to
do anything about it. This seemed to render me complicit in pupils’ sub-
versive actions in their perceptions, and was often returned with a subtle
gesture that signalled their appreciation. Misdemeanours could involve an-
ything from chewing gum, flicking rubbers, passing messages on bits of
paper, trading football cards, leaning back on a chair, or temporarily
‘stealing’ other pupils’ equipment.
Although on initial visits some students would report other pupils’ mis-
behaviour to me, expecting some sort of reprimand or sanction, they soon
stopped doing this, realising that I wasn’t going to do anything about it, and
even started actively trying to involve me in off-task or disruptive behav-
iours. This I did (within limits) judging that it would not have unwarranted
detrimental affects, but might help solidify my relationships with pupils, and
thus lead to enhanced data acquisition later on. I passed notes from one
pupil to another, and engaged in off-task conversations, answering pupils’
questions relating to my marital status, preferred football teams and tel-
evision viewing habits.
I often found that brief off-task interactions with pupils tended to be fol-
lowed by a period of more focused activity on their part. It was sometimes
quicker and less disruptive for me to pass a love note than for other pupils to
throw it or leave their seats. Similarly, I found that pupils could be quite
persistent in asking me personal questions, and my failure to respond to them
seemed to make the prospect of discovery more alluring and occupy pupils’
attention even further. By reacting in a friendly and brief manner I satisfied
pupils’ curiosity, and often reduced rather than augmented off-task behaviour.
However, this maintained the appearance of a subversive role, and it was this
appearance that was so important in developing rapport with some pupils.
On some occasions teachers commented that by engaging in short infor-
mal conversations with pupils I helped reduce the likelihood of off-task
behaviour escalating into something more serious or disruptive to other
pupils. These roles were subject to clearly defined, self-imposed boundaries,
but I was fortunate that no serious incidents arose whereby my non-dis-
ciplinarian form of practice had to be abandoned.
This engagement in more subversive elements of formal lesson time fed

directly into the roles I adopted with pupils in classroom ‘fringe’ time.
Territory 4 – Interactions with Pupils in the Classroom Curing Fringe Time
I refer to break and lunch times, and the beginning and end of lessons when
no formal teaching is taking place as ‘fringe’ time. Fringe time also incor-
porated interactions with pupils on playgrounds and out of school (on buses
etc.) – my focus on classrooms here illustrates how the same space can be
intersected with temporal distinctions and associated shifts in the forms of
social interaction which occur.
As pupils were entering and leaving a class for a lesson I chatted infor-
mally with pupils about school and home life. At break and lunch times I
sometimes hung around in classrooms that were open to pupils although I
limited this because I felt it was important to maintain some spaces as adult-
free.
I would describe this setting as both frontstage and backstage. It is
frontstage in that pupils remained very much on show to their peers and it
seemed, actively aware of their peer-group hierarchies and behaviour norms;
it is backstage in that it was characterised by informal interactions that were
not supervised or structured by teachers, and not typically accessed by
adults.
Pupils often used these opportunities to ask me personal questions (Do I
fancy Miss X? Am I a virgin?) and other more general questions (Did I
watch Little Britain?), but were also keen to tell me about themselves. Pupils
seemed to learn quite quickly that I wouldn’t comment about teachers or
other pupils, but that I wouldn’t reprimand them for expressing their views.
Their narratives about themselves or their peers were colourful, sometimes
implausible, occasionally inappropriate, and I had to make on-the-spot
judgements about appropriate boundaries to my roles in this setting, some-
times pretending not to hear or simply by finding an excuse to leave the
room.
I was sometimes called upon to be an arbitrator in disputes between
pupils, and often treated as a venting post for pupils to moan about teach-
ers, homework, detentions and the like. I heard juicy details about a fight
between pupils, although I had to be careful not to reciprocate when pupils
pried about a stabbing event – I knew more than they did (i.e. who had
stabbed who) but wasn’t in a position to reveal this information.
64 NICK HOPWOOD
Many of my lunch times, however, were spent preparing for and carrying
out structured data generation activities with particular pupils.
Territory 5 – Formal researcher-initiated structured interactions with pupils
This territory was less spatially specific than the others, incorporating a
variety of classroom, corridor, staff-room and storeroom locations. I am
referring here to contexts in which I invited particular pupils to meet me at
lunch times and participate in audio-recorded, structured activities (either
interviews about lessons, or the four techniques mentioned in the context
section above). While all my interactions with pupils were part of my broad
research agenda and atypical in the sense that pupils would not normally
interact with a doctoral student, these formal interactions were different,
and involved different fieldwork roles.
The quality of data generated through these interactions depended largely
on the rapport developed with pupils through my roles in formal lesson time
and fringe time. I asked a lot of questions to some pupils (up to 15 separate
recorded interviews with an individual), and their willingness to give up their
time may reflect to an extent the shared values that I had displayed through
talking to them about things they were interested in and/or my non-disci-
plinarian (partially subversive) classroom roles.
An important feature of all recorded interviews was that they began and
ended with an unfocused question about how pupils were feeling or how
their day was going. I participated not just as researcher needing data, but as
adult interested in children (Mayall, 2000). Cooper and McIntyre (1996)
argue that young people are unqualified experts in their own fields (i.e. in
matters relating to their ideas, opinions and experiences), and in these in-
stances I called participating pupils ‘experts’ and gave them a sheet ex-
plaining why I considered them so.
Many aspects of my roles in these interactions were aimed at enabling
pupils to participate actively in the study rather than as passive objects of
research. Such practices included asking pupils to choose names to be used
in writing about them (one pair chose the names Bart and Lisa which in-
evitably meant I used the pseudonym ‘Springfield school’). I also asked
pupils to switch the mini-disc recorder and microphone off and on and do
things like check the write protect button wasn’t pressed.
One technique in particular involved me ‘letting go’ of some control in my
research and passing responsibility for creating stimulus materials to pupils.
I gave pupils disposable cameras and asked them to take pictures showing
what they thought geography was about.1 I asked them to take between six
and ten for me, and to use the rest of the film as they wished. The cameras
came with free processing coupons, and so pupils got the pictures developed
themselves and brought in copies of those that were relevant to my study.
Pupils seemed to enjoy this task – photographs were taken from aero-
planes, ferries, and cars (signalling the fact that pupils carried the cameras
away with them on holidays). They also seemed to enjoy talking about their
pictures and seemed quite proud of them, if perhaps surprised that someone
like me was so interested in them.
During these formal recorded interactions with pupils, I also adopted
roles that were more rehearsed and more specifically attuned to my own
research agenda than the other roles I have described above. All interviews
were semi-structured in nature, and I had remits to follow my interview
guide, steer pupils’ comments so they remained relevant, and actively listen
to their responses in order to pose follow-up questions or probe for further
clarification.
The conduct of interviews in social research is well documented, and my
purpose here is not to suggest that my roles in these contexts were unique.
Rather the discussion of this fifth territory serves to demonstrate more fully
the range of roles that researchers can adopt in interaction with particular
groups of participants, and to illustrate that the different territories in which
research takes place are not purely properties of field sites, but can be
actively constructed by researchers themselves.
CONCLUSIONS
Walford (1987) notes that it is inappropriate to consider ethnographic re-
search in terms of single roles because ‘‘a variety of roles must be adopted
which will vary with the different individuals with whom the researcher
interacts’’ (p. 45). In this paper, I have explored the range of roles I adopted
in a school-based ethnographic study, and have shown that different roles
cannot be straightforwardly associated with particular people, times, or
places, or people.
I have introduced the concept of territories as settings for interaction
between researchers and participants which are defined in terms of spatial,
temporal and social dimensions. This concept is helpful in understanding the
complex nature of research sites, and the resulting insight can help re-
searchers consider the roles they adopt in the field.
66 NICK HOPWOOD
To consider a research site territorially is to think through the different

contexts in which researchers find themselves, the forms of activity and
social interaction which characterise those contexts, and the ways in which
these settings can be demarcated. Rather than simply distinguishing between
the frontstage and backstage, or different groups of people, researchers can
begin to explore ways in which the same place is host to a range of social
practices, and how these practices can change in predictable, structured
ways (as with a classroom and a school timetable) or as part of a less
predictable, dynamic setting (as with the geography offices).
As well as elucidating the way the social nature of particular places can
change over time, the process of considering the different territories within a
research site can also highlight the variety of roles that researchers adopt in
interaction with particular participants or groups of participants. I have
illustrated this by describing my participation in the gossip of the geography
office, my role as classroom assistant in formal lesson times, my informal
chats with pupils in fringe time, and my role as instigator and manager of
formal interviews with pupils. Such a description of research roles is strik-
ingly different from the typologies offered by Gold (1958) and Junker
(1960).
It seems that the very nature of ethnography (extended periods of field-
work involving a range of techniques and interaction with a variety of
participants) lends itself to the adoption of multiple different roles. Fur-
thermore Snow et al. (1986) suggest that as different fieldwork roles yield
different types and degrees of information, a variety of roles should be
utilised in any one project. However, they conclude that with few exceptions
this means that research should be conducted by teams, with each individual
adopting a role best fitting his or her strengths, rather than by a ‘lone
ranger’.
In this paper I have shown, as did Walford (1987), that researchers need
not be confined to one role. However, I hope to have additionally provided
researchers with a framework for making sense of the extremely complex
contexts in which ethnographers conduct fieldwork and of the variety of
roles that they adopt. The process of considering different territories offers a
conceptual tool, which may be applied to a variety of settings, and may
prove useful in providing more nuanced understandings of the nature of
research sites and hence field roles.
I have not described the roles I adopted in terms of a linear evolution over
time, although I have indicated both instances where ideas of gradual shifts
in relationships and roles are appropriate, and moments of more abrupt
change. My purpose is not to refute the utility of a longitudinal trajectory in
thinking about ethnographic field roles, but rather to situate it within a

different, more multi-faceted, form of understanding that accommodates
notions of longer term progression and rapid flux.
Similarly, I hope to show that while the concept of frontstage and back-
stage settings can be helpful, it may be more useful to incorporate this
distinction into a more subtle analysis of field sites, such that more than one
frontstage and backstage may be identified.
The examples I have provided all relate to three English secondary
schools, and it seems reasonable to suggest that similar territories might be
encountered in other English secondary school sites, although of course
researchers may adopt quite different roles within them depending on their
purposes and personality. What I hope to offer readers more generally is not
an understanding of researcher roles or field sites that can be simply lifted or
transplanted from one context to another, but rather the idea of thinking in
terms of territories and being open to the multiple roles that researchers may
adopt in particular spatially, temporally and socially defined territories
within a research site.
NOTE
1. The practice of giving participants cameras in research is now referred to as
‘self-directed photography’, and the earliest use of this technique was by Collier
(1957, 1967). The use of photographs and discussions about them in ethnographic
and social research has been explored by Prosser (1992) and Prosser and Schwartz
(1998).
REFERENCES
Anderson, L. W., & Burn, R. B. (1989). Research in Classrooms. Oxford: Pergamon.
Burgess, R. (1984). In the field: An introduction to field research. London: Allen & Unwin.
Collier, J. (1957). Photography in anthropology: A report on two experiments. American An-
thropologist, 59, 843–859.
Collier, J. (1967). Visual anthropology: Photography as a research method. London: Holt, Ri-
nehart & Winston.
Cooper, P., & McIntyre, D. (1996). Effective teaching and learning: Teachers’ and students’
perspectives. Buckingham: Open University Press.
Goffman, E. (1959). The presentation of self in everyday life. Garden City, NY: Doubleday.
Gold, R. L. (1958). Roles in sociological field observations. Social Forces, 36, 217–223. (Re-
printed) In: A. Bryman (Ed.) (2001) Ethnography: Volume II (pp. 141–149). London:
Sage.
68 NICK HOPWOOD
Janes, R. W. (1961). A note on the phases of the community role of the participant observer.
American Sociological Review, 26(3), 446–450.
Junker, B. H. (1960). Field research: An introduction to the social sciences. Chicago: Chicago
University Press.
Mayall, B. (2000). Conversations with children: Working with generational issues. In: P.
Christensen & A. James (Eds), Research with children. London: Falmer.
McCall, G. J., & Simmons, J. L. (1969). Issues in participant observation: A text and reader.
Reading, MA: Addison-Wesley.
Oleson, V., & Whittaker, E. (1967). Role-making in participant observation: Processes in the
researcher-actor relationship. Human Organisations, 26(4), 273–281.
Prosser, J. (1992). Personal reflections on the use of photography in an ethnographic case study.
British Educational Research Journal, 18(4), 397–411.
Prosser, J., & Schwartz, D. (1998). Photographs within the sociological research process. In: J.
Prosser & D. Schwartz (Eds), Image-based research: A sourcebook for qualitative re-
searchers. London: Falmer.
Punch, K. (1998). Introduction to social research: Quantitative and qualitative approaches. Lon-
don: Sage.
Robson, C. (1993). Real world research. Oxford: Blackwell.
Snow, D., Benford, R., & Anderson, L. (1986). Fieldwork roles and informational yield: A
comparison of alternative settings and roles. Urban life 14(4), 377–408. (Reprinted) In:
A. Bryman (Ed.) (2001). Ethnography: Volume II. London: Sage, pp. 150–169.
Walford, G. (1986). Life in public schools. London: Methuen.
Walford, G. (1987). Research role conflicts and compromises in public schools. In: G. Walford
(Ed.), Doing sociology of education. London: Falmer.
THE STRANGE CASE OF THE
DISAPPEARING TEACHERS:
CRITICAL ETHNOGRAPHY AND
THE IMPORTANCE OF STUDYING
IN-BETWEEN
Martin Forsey
In this all too unequal world the orientation of many social scientists
towards redressing the imbalances is not at all surprising. Indeed, I began my
study of schooling in Western Australia in 1997 very deliberately orientating
myself in this direction by reading Carspecken’s (1996) guide to critical
ethnography in educational research. Carspecken refers to what he calls the
‘criticalist’ approach as one that is underpinned by a shared concern about
social inequalities, and a desire to ‘work toward positive social change’
(p. 3). The emphasis on social justice, and the desire to not only ameliorate
but also fight the sometimes brutal conditions imposed on particular groups
of people by the capitalist system, articulates a research stance congruent
with my preferred standpoint. But as I will show in this paper, the criticalist
project is not without its problems. For a stance that is overtly committed to
addressing unequal distributions of power it seems remarkably unprepared
for dealing with this issue as a research objective. As Ortner (1995, p. 176)
suggests of criticalist researchers, especially those focused on resistance, they

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12005-7
69
70 MARTIN FORSEY
tend to glide over the complex and contradictory political realities of the
social groups they portray.
The major problem explored in this paper arises out of the tendency for
critical educational ethnographers to produce accounts of schooling in
which teachers either disappear or are represented as mere cardboard cut-
out figures (Forsey, 2000). This particular form of ethnographic ‘thinness’
(Ortner, 1995) is attributable to the simplistic positioning of teachers
as members of the dominant group in the resistance scripts critical ethnog-
raphers appear to carry with them upon entering schools. When ethnog-
raphers of schooling disassociate themselves from a notable group in their
field as a means of gaining the trust of their preferred subjects, which for
critical researchers are almost inevitably marginalized students (Anderson,
1989; Reed-Danahay, 1996), all too often the result is a simplistic portrayal
of teachers ‘acting on behalf of a class power not their own’ (Connell, 1995,
p. 91). Such portraits undermine the broader research enterprise in which
those of us who identify as critical educational anthropologists/ethnographers
are engaged.
There are three main aims for this chapter: (1) explaining the relative lack
of attention to teachers in criticalist accounts of schooling; (2) exploring
the implications of these oversights for qualitative research in schools;
and (3) suggesting ways in which the researcher’s mindset might shift to
address the issues arising from our currently ‘averted gaze’ (Wisniewski,
2000). Advocating a mindset attuned to the critical appreciation of social
life, I suggest a need to heighten the importance of clear and precise doc-
umentation of the power-filled social relations characterizing educational
settings as an important element of the criticalist project.
Further to this, I argue for the need to position ourselves in ways that
allow us to better see, feel and evaluate some of the multitude of interactions
that take place in schools and their surroundings. This differs somewhat with
recent pronouncements that have appeared in Anthropology and Education
Quarterly (AEQ), the flagship journal of the American Anthropological
Association’s Council of Anthropology and Education, declaring the need to
heed Nader’s (1974) relatively ancient call to ‘study up’ (Priyadharshini, 2003;
Hamann, 2003). While I think it is important that educational ethnographers
shift what sometimes seems like an exclusive focus on subaltern groups, it is
not enough to simply change the direction of our gaze. We need to pay
attention to standing and moving in the ‘in-between’ spaces occupied by
social groups. Such positioning can allow us to reflect upon the complex of
horizontal, vertical, and perhaps even diagonal, relationships that constitute
the social formations we call schools. Such positioning offers far more
The Strange Case of the Disappearing Teachers 71
productive angles from which to explore the ways in which power is con-
figured and impacts on people.
Like Ortner (1995, p. 173), I view ethnographic positioning as not just a
corporal act; it is also intellectual and moral in its intent and practice.
Ortner’s thoughts on the ‘ethnographic refusals’ she associates with critical
ethnography are of great importance to the arguments I am pursuing here.
She describes such refusals as an avoidance of the anthropological com-
mitment to ‘thickness’, by which she is referring to the ways in which Geertz
has stressed the need for ‘richness, texture and detail’ in ethnographic por-
trayal (Ortner, 1995, p. 174). These ideas are pursued later in the paper;
I first want to provide a brief overview of some of the developments of a
critical ethnography as applied to education and to uncover some of the
major strands of thought and practice in the social sciences contributing to
the development of this particular research approach.
CRITICAL ETHNOGRAPHY AND EDUCATIONAL

ANTHROPOLOGY
Critical ethnography first emerged as a distinctive research approach in
education studies in the late 1960s (Anderson, 1989, p. 250). It has now
achieved a degree of respectability and has taken its place as part of the
qualitative tradition in universities (Jordan & Yeomans, 1995, p. 399). Crit-
ical ethnography reflects what Geertz (1983) identified as a ‘blurring of
genres’. As the name suggests, it is marked by a confluence of interpretivist
field studies and critical streams of thought (Goodman, 1998, p. 51). These
converging streams, arising from a variety of sources and pushed along by
the currents of Marxist, neo-Marxist and feminist social theory, swirl to-
gether into a dynamically enriched mixture of the methods and theories of
anthropology, sociology and education. Not surprisingly the streams
formed in different parts of the globe, while composed of all of the ele-
ments just named, are configured slightly differently. As Priyadharshini
(2003, p. 421) recently noted in comparing British and American strands of
educational ethnography, the Western side of the Atlantic is marked by a
much stronger tradition of educational anthropology than in the UK. And
these differences make a difference.
I write from Australia where educational anthropology barely forms a
trickle. Nevertheless, I write as one who was trained primarily as an an-
thropologist and has applied this training to the study of schools and
72 MARTIN FORSEY
schooling. True to the swirling realities just described, I readily acknowledge

the influences of sociologists and educationalists, among others, on my re-
search and writing. I am certainly not arguing that there is a pure, or entirely
distinctive, anthropology. However, when addressing educational ethnog-
raphers, which I imagine I am mainly doing here, I think it is useful to call
attention to the centrality for anthropologists of ethnographic studies based
on participant observation. Priyadharshini (2003, p. 426) in the same paper
in which she alerts us to the differences between North American and British
educational ethnography, also makes this point, or at least she does for an
earlier era. Drawing on the writing of Nader (1974) and Marcus and Fischer
(1986), she asserts that the self-image of anthropologists was severely dis-
turbed if they could not ‘participant-observe’. According to Priyadharshini,
things have moved on from what was apparently the case in the 1970s.
In the current era, she argues, ethnographers draw upon a wide variety of
techniques; thus, their research identity is no longer defined by a specific
method. I beg to differ; at least as far as anthropology is concerned. Eth-
nography remains the signature method of anthropology, and the self-image
of the majority of anthropologists remains firmly tied to participant obser-
vation in ways that I am sure is not the case for my colleagues in sociology
and education departments. One can easily imagine being a sociologist or
educationalist without being an ethnographer. This is simply not possible
for most social and cultural anthropologists.
Having indicated the impossibility of conceptualizing an educational an-
thropology that is distinct from other disciplinary traditions, I take it as
given that anthropology has had an impact on the broader field. Therefore
while this chapter focuses on critical educational ethnography in general,
I pay particular attention to the ways in which anthropological ideals and
ideas have influenced contemporary formations of critical thought and eth-
nographic practice; they help explain why teachers tend to disappear in
critical ethnographic accounts of schools and schooling.
THE SUBALTERN GAZE AND DISAPPEARING

TEACHERS
Predictably enough, Willis’s (1977) enormously influential report of the
ways in which ‘the lads’ he followed at ‘Hammertown Boys’ learned how to
get working-class jobs was one of the first criticalist ethnographies I read.
While I was intrigued by his account of the ways in which the rebellious
‘lads’ of ‘Hammertown Boys’ penetrated the meritocratic veneer of school-

ing and ‘the role of labour in the modern structure of capitalist production’
(p. 133), I was disappointed by Willis’s treatment of the teachers in the
school. This feeling was to accompany me in my reading of other critical
ethnographies of schooling (Forsey, 2000). Willis reports conducting inter-
views with many of Hammertown Boys’ teachers, readily acknowledging
them as ‘dedicated, honest and forthright y doing an exacting job with
patience and humanity’ (p. 67); yet their presence in the text is slight. An-
ticipating this sort of criticism, Willis wisely asserts that he could not concern
himself with every variety of student culture and teaching relationship.
Rather, he suggests it is the role of other practitioners ‘to test reality with the
concepts outlined; to contextualize; to see what role different fundamental
processes play at different strengths in different situations at different times’
(p. 85).
Having one of the most influential educational ethnographers suggest that
the responsibility for holism lies not with the individual researcher but
rather with all who are involved in the research enterprise is a productive
and potentially liberating idea. Yet, the implied widening of the agenda,
aimed at filling in the gaps apparent in Learning to Labour, has happened
only to a limited extent. Attention was eventually given to young women
and girls in the ethnographic corpus. This helped shift attention away from
what one of the pioneers of this movement calls ‘boys’ own ethnography’
(Griffin, 1987, p. 87; see also Quantz, 1992, p. 456; Griffin, 1985). Ethnic
minorities also became a key locus of study, a fact reflected in my, somewhat
cursory, survey of recent editions of AEQ where 71% of the published
articles over the 6-year period survey (2000–2006) focused primarily on
ethnic minority groups.1 While some of the articles included in the count of
‘minority studies’ were written about students from White working-class
backgrounds, most converged on students of similar socio-economic status
from Latino, Black, Asian and Hispanic origins. Such data help support
recent assertions regarding the pervasive nature of the downward focus of
educational anthropology (Emihovich, 1999; Wisniewski, 2000; Priyadhar-
shini, 2003; Hamann, 2003). This ‘subaltern gaze’ (Marcus, 1998) has a
double movement associated with it, as not only do researchers give most of
their attention to marginalized groups in schools (and the vast majority of
the articles surveyed reported on research conducted in and around
schools), they also tend to concentrate mainly on students. Just over half
of the articles were written about students or youth. When those featuring
relationships between students and/or teachers and parents are considered,
this figure rises to 76%. Only 20% of the research featured teachers as the
74 MARTIN FORSEY
major interlocutors. Some of the reasons for these patterns are located in the
various theoretical, methodological and epistemological developments in
the anthropological field that are discussed below.
ETHNOGRAPHIC REFUSALS IN SCHOOLS
Ortner’s notion of ‘ethnographic refusal’ offers some important insights on

problems associated with subaltern studies. She brings together the anthro-
pological ideals of thickness and holism under the rubric of contextualizat-
ion, all of which she suggests lie at the heart of the ‘ethnographic stance’
(p. 174). The refusal of which she writes is marked by an unwillingness to
explore a field site in all of its richness. There is a baulking at thickness
and holism, resulting in the sanitizing of politics, a thinning of culture and
the dissolving of actors (p. 176). At risk of oversimplifying the argument,
ethnographic refusal can be summarized as a failure to take seriously the
complex realities that shape, and are in turn shaped by, the collective
actuality of individual lives. Ortner correlates ethnographic refusal to the
study of resistance among subaltern groups, and as she describes the sit-
uation (p. 190) such studies are often unnecessarily ‘thin’. They skate over
the politics of dominated groups as well as the cultural richness of those
groups and the intentions, desires, fears and projects that comprise the
subjectivity of social actors.
Ortner’s critique can certainly be used very productively in refining crit-
icalist studies of schooling, and help show the ambivalent, ambiguous
nature of resistance in educational settings (p. 190). However, I want to call
into question the standard anthropological line that Ortner appears to follow
in acknowledging the downward gaze to be the most appropriate intention of
ethnographic study (see p. 188). While I have no argument with her prop-
osition that we should be moving beyond the politics of representation and
the associated fears that arise about questioning and discussing the internal
politics of dominated groups (p. 190), I also think that we have to widen our
perspectives. Unless we pay attention to documenting and analysing other
parts of the whole social field, we are doomed to another level of ethno-
graphic refusal, one that prevents us from understanding the complexities of
power as well as the ambivalences and contradictions of subordinate positions
identified so cogently by Ortner (1995, p. 175).
I argue that ethnographic refusal is evident in many school ethnographies
and is most readily visible in the failure of critical ethnographers to pay close
attention to the worldview of the teachers in the school. Thickness and
holism are not taken seriously enough by many, probably because of par-
ticular biases against the so-called middle class, bourgeois lifestyles. Part of
the refusal possibly lies in not wanting to be seen to legitimize such a way of
life, or perhaps because the impetus for change for a more just, socialist
society are not thought to lie with that particular group (Davies, 1995).
Two relatively recent and influential examples of ethnographic accounts
of schooling help illustrate the points I am making here: Fordham’s (1996)
account of ‘‘Dilemmas of race, identity, and success at Capital High’’ in
Washington, DC; and Reed-Danahay’s (1996) portrayal of assertions of
local identity and resistance of state power by parents and students involved
with an elementary school in rural France (see Forsey, 2000). Both offer
powerful examples of how to do critical ethnography in schools and are
exemplary in many ways. However, insofar as they fail to take seriously the
worldview of teachers, they also offer examples of some of the problems
currently evident in the practice of critical ethnography in schools.
Reed-Danahay describes various arenas of struggle between the villagers
and Henri and Lilianne Juillard, the teachers who ran the two-room primary
school in her research site of Lavialle. She succeeds in portraying the teachers
‘as young suburbanites’, who drew varying levels of disquiet and resistance
from the parents associated with the school. One of the arenas of ‘dispute
between parents and teachers’ to which she pays particular attention is the
school lunchroom (p. 175). Reed-Danahay describes how the teachers
organize the eating of lunch according to the protocols of an idealized upper-
middle class ‘bourgeois’ household. Exemplifying the tendency for criticalist
accounts of schooling to present teachers as ‘disvoiced bodies’ (Forsey,
2000), Henri Juillard is portrayed as a stern and distant patriarch and Liliane
as the household manager who upholds patriarchal authority (Reed-Danahay,
1996, pp. 199–200). While this is probably a valid and useful interpretation
of ‘what the ethnographer saw’, there is no evidence in Reed-Danahay’s
ethnography of her ever asking the teachers why they acted in this way. At
no point do we ‘hear’ from them. Any one who has conducted research in
schools knows the importance of maintaining relationships with teachers
and administrators. Reed-Danahay is no exception and she informs her
readers that she had established cordial relations with the Juillards. Why
then do we not read of an interview, or informal questions probing the
worldview of the teachers and their reasons for behaving as they did in the
lunchroom? Not only would it be more ethically satisfying to read about
these viewpoints, exploring the motivations of the teachers would greatly
enhance our understanding of how the school works and deepen our appre-
ciation of the ways in which power is actually exercised in social settings.
76 MARTIN FORSEY
Fordham’s ethnography is set in a high school where the vast majority of

students are Black. The author aims at ‘a culturally sensitive representation
of the factors y implicated in the success of Black adolescents’ (1996, p. 23).
Opening with a passionately eloquent discussion of the importance of
reflexive engagements by ethnographers on the power-filled representations
they make, Fordham alerts her reader to the dependence of the consumers
of anthropological knowledge upon the ethnographer’s interpretations and
subsequent representations.
Unlike the Juillards in Lavialle, the Washington teachers are given a
voice, but, in an act of ethnographic ventriloquy, they emphasize the points
that Fordham wishes to make. There is no evidence I can see of the teachers
being invited into dialogue with Fordham about the representations she
makes of them, representations that are consistent with the criticalist tra-
dition of painting teachers and officials as agents of the dominant, usually
ruling class, bourgeois ‘other’, who are removed from the students and lack
understanding of their struggle (Connell, 1995). Despite most of the teachers
at Capital High being Black women, they also said to represent an assim-
ilating ‘whiteness’ (p. 191). The teachers are portrayed as resisting the
dehumanizing effects of White culture by conforming to the rules and norms
of this dominant culture. Their apparent aim is to become ‘acceptable pro-
fessionals’ (p. 202), or ‘foreign sages’, as Fordham insists on calling the
teachers. While Fordham reports that the principal went out of her way to
make her feel welcome at the school, and a number of the teachers did the
same, at no point does Fordham report inviting teachers into discussion
with her about the classificatory tools she uses to describe them. Instead,
she seems content to follow a trend evident in many critical accounts of
schooling in maintaining throughout her text a certain ironic and critical
distance from these key social actors.
ETHNOGRAPHIC DISTANCING
In their important introductory review of critical approaches to the ethno-

graphic study of schooling, Levinson and Holland (1996, p. 19) offer some
explanations for this ethnographic distancing. They point to the significant
ethical and methodological difficulties presented by school-based research
and proffer three main problems for those conducting research in schools.
Firstly, ‘schools have gatekeepers from whom permission must be obtained
and of course maintained’. Secondly, the research requires sustained obser-
vations and interviewing. Thirdly, ‘the finished product of the research may
very likely involve a critique of the very people who made the research
possible’. While claiming such difficulties to be unique to school-based re-
search is debatable, they do alert educational researchers to some important,
and oft ignored, issues. Yet, they immediately smooth down the important
threads they unpicked by asserting that, ‘a critical school ethnography can-
not help but scrutinize the actions of school officials, even if they are not
our primary research subjects’. This is true, and as I hope is abundantly
clear, not for one moment am I suggesting that the actions of teachers
and administrators should not be subject to critical scrutiny. What I am
suggesting is that critique is simply not enough and that if the critical stance
does create the sorts of ethical, methodological and epistemological dilem-
mas Levinson and Holland allude to, then we need to think about the
implications of this far more than we currently appear to be doing.
In a later essay, exploring the reasons why schooling is ignored so con-
sistently by ‘mainstream’ anthropology, Levinson (1999, p. 599) expands on
the point he made in tandem with Holland by suggesting that the problems
associated with studying schools lie in the often ‘difficult negotiation of
structurally opposed interests and alliances’ encountered by the researcher.
While he does not inform the reader about who exactly comprises the
oppositional alliances and what the various parties are fighting over, it is
reasonable to surmise that the conflict centres on defence and critique of a
hierarchical status quo.
I am arguing that teachers are rarely the main subjects of critical ac-
counts of schooling, because they tend to be viewed as protectors of the
status quo. They are simplistically portrayed as perpetrators of a system
that oppresses and undermines students, particularly those from working
class and minority backgrounds. While this is the case in many instances,
and critical ethnography has done a great service in showing this, it is time
to expand our understandings of how this happens. Very few teachers that
I know deliberately set out to become oppressors of minority groups;
indeed as Connell (1995) suggests, teachers as a group have an enormous
commitment to an ethic of care and concern for injustice. In order to
understand how educational practices are produced and reproduced real
engagement with teachers and administrators should be a vital part of the
critical ethnographic project over the coming years. I hasten to add that
this should not be done as a means of ‘getting the goods’ on teachers
(Marcus, 1998), but rather as a genuine attempt to understand the practice
of power from the purview of key players in the social dramas with which
we engage as participant observers. This theme is explored further in the
next section.
78 MARTIN FORSEY
THE SUBALTERN GAZE IN CRITICAL

ETHNOGRAPHY
In their seminal account of the ways in which social and cultural anthro-
pology has developed into a form of ‘cultural critique’ Marcus and Fischer
(1986, p. 1) argue that anthropologists promised their largely Western
readership enlightenment on two fronts:
1. The salvaging of distinct cultural forms of life from a process of rampant
Westernisation; and
2. A form of cultural critique of and for this Western readership.
Later, they turn their attention to the ways in which anthropologists

shifted their attention away from small-scale societies in former European
colonies, towards studies ‘at home’ in the so-called Western societies.
Referring to this process as ‘repatriation’, Marcus and Fischer make a telling
observation:
y as they have written detailed descriptions and analyses of other cultures, ethnog-
raphers have simultaneously had a marginal or hidden agenda of critique of their own
culture, namely the bourgeois, middle-class life of mass liberal societies, which industrial
capitalism has produced (p. 111).
Nader (1974) asserts that anthropologists tend to favour the underdog, a

positioning she describes using the ocular metaphor of ‘studying down’, and
which Marcus (1998) names as ‘the subaltern gaze’. Such preoccupations are
certainly not exclusive to anthropology, as Bell (1978) noted nearly 30 years
ago, sociology is also replete with top-down perspectives. However, it is
quite probably the case that anthropologists by personal inclination and
training are more likely to be dismissive of Western cultural formations than
are sociologists and educationalists. Helping to support this assertion,
Priyadharshini (2003, p. 420) notes that while social scientists in general
have increasingly shifted their attention upwards, the subaltern gaze is still
very evident among educational anthropologists (see also Hamann, 2003).
The cultural imperatives that push and prod anthropologists towards
critique of bourgeois lifestyles help explain the lack of attention to teachers
in anthropological accounts of schooling. But, as Priyadharshini (2003,
pp. 423–425) also shows, a lack of sympathy with middle-class existence is
not the exclusive domain of anthropologists. Her PhD research, which
involved a critical ethnographic study of MBA programs at Indian univer-
sities, caused a degree of consternation among her research colleagues.
Priyadharshini reports that fellow social scientists often interrogate her
about her values and theoretical stance. Linking this curiosity to her interests
in a pedagogical practice (the MBA) that in the view of many ‘supplied
managers for exploitative, undemocratic and corrupt multinational corpo-
rations’ (p. 424), she argues that her colleagues puzzlement arises from
assumptions about qualitative researchers needing to develop a close rapport
with those they study.
Priyadharshini’s experiences echo concerns Marcus has been raising for
more than two decades. In conjunction with Fischer (1986), he suggests that
commitments to the cultural critique of bourgeois existence push social
researchers towards finding and emphasizing the negative or the absurd
when investigating the worldviews of middle-class people. In an earlier
work, drawing our attention to the problems raised by ethnographic studies
among elites, Marcus (1983) points to the methodological difficulties posed
by ‘ideological distancing from one’s subjects to the point of disapproval’
(p. 23). Importantly, he links the failure to engage with elite subjects to a
fear on the part of the researcher to charges of elitism, a situation which he
persuasively argues arises from confusion between working empathy and
ideological empathy with one’s research subjects.
More recently, Marcus (1998, p. 27) has issued a timely warning against
research approaches committed to ‘getting the goods’ about elite groups,
and ‘probing the interior dynamics of how power shapes [the subjects] lives
and is produced by them so that they can be opposed’. Social researchers
are obliged to interrogate as thoroughly as possible their motivations for
carrying out their research. This is not only an ethical issue; it is method-
ological and epistemological as well. Understanding and comprehending
how people come to grips with their particular social location and realities,
no matter where they are positioned in social hierarchies, is surely a key
component of our jobs as social researchers. It is helpful to recognize that
criticism is usually the easy part of our project, one that is often taken up all
too quickly by criticalist researchers (Goodman, 1998). This ‘race to theory’
(Christian, 1987) tends to undermine what should be the primary challenge
of social research – understanding social formations in all of their imperfect
glory.
While many of the teachers I know would find laughable any suggestions
that they are part of society’s elite, especially when they compare their
salaries with those of other university graduates, obviously one’s social
standing is always a relative concept. This relativism helps explain why a
number of critical ethnographic accounts of schooling treat teachers as part
of a local elite, with all of the ideological distancing of which Marcus writes
so cogently. At a time when the calls for deeper, richer, ‘thicker’, more
80 MARTIN FORSEY
nuanced understandings of power and its social effects are being issued
(Abu-Lughod, 1990; Ortner, 1995), those drawn to a critique of the power
relations produced by industrial capitalism often miss important opportu-
nities to deepen their understanding of the complexities of power. Indeed,
in so far as those of us committed to a politically engaged and critical
anthropology of schooling ignore teachers and their formative relationships
with students’ parents and administrators, we condemn ourselves to ‘eth-
nographic thinness’ (Ortner, 1995), and a concomitant theoretical and
epistemological impoverishment.
THE PROBLEMS WITH REPRODUCTION AND

RESISTANCE
Critiques of the status quo mark some of the most important contributions
of critical research. This is not the place to replay the development of critical
thought as applied to the evaluation of schooling, but the key trope of
reproduction is worth exploring because of the way in which it directs social
researchers towards investigations of the beliefs and values of all social
actors and their interactions with the social systems they are subject to.
Initial uncovering of the ways in which schooling induces worker power-
lessness that emerged in reproductionist arguments in the 1970s, and the
idea that schools reproduced rather than ameliorated iniquitous social
structures, offered an important corrective to liberal fantasies of education
as an unproblematic ‘springboard for upward mobility’ (Levinson & Holland,
1996, p. 5; see also Althusser, 1971; Bowles & Gintis, 1976; Bourdieu &
Passeron, 1977). It helps shift blame away from the victims of such structures
to the system that produced them (Connell, Ashenden, Kessler, & Dowsett,
1982, p. 28). However, the trap of overdeterminism that plagues reproduc-
tionist accounts of schooling is now well recognized by most social researchers
(Anderson, 1989; Davies, 1995; Levinson & Holland, 1996).
Willis (1977) is often acclaimed as the one who shattered ‘the image of the
passive, malleable student implicit in reproduction theory’ (Levinson &
Holland, 1996, p. 9). It was the rebellious ‘lads’ of ‘Hammertown Boys’, and
their penetration of the meritocratic veneer of schooling and ‘the role of
labour in the modern structure of capitalist production’ (Willis, 1977,
p. 133) that helped critical analysts to escape their own deterministic trap
(Anderson, 1989, p. 251). Resistance emerged as the key theoretical trope of
such approaches, with Willis’s ‘grounded version of resistance theory’
becoming the touchstone work for critical ethnographers seeking to marry

phenomenological concerns about human agency to Marxist preoccupations
with social structure (Anderson, 1989, pp. 255–256). Resistance became
synonymous with agency in many ethnographic accounts of schooling writ-
ten in the criticalist paradigm, apparently providing the ethnographer with
the means to move beyond static, deterministic explanations of human be-
haviour insinuated in cultural reproduction theories (Hoffman, 1999, p. 675).
While Willis’s Learning to Labour was, and remains, an important text,
especially for the way in which it allowed subsequent researchers to think
beyond the traps set by an overly neat and impenetrable set of structures
(p. 175), its privileging of resistance helped create a new set of conceptual
problems. Resistance approaches have become hegemonic in many parts of
the educational research corpus, partly because of their promise of locating
agency among the subjects of the ethnographer’s downward gaze. Willis is
clearly not to ‘blame’ for this, but his influence is profound and undoubtedly
helped set the agenda for many a critical researcher. The problem I have
with Learning to Labour lies not only in Willis’ representation of the teachers
at Hammertown Boys, but also with those dismissed by the resistant
‘lads’ as the ‘ear’oles’, the complicit ‘prats’ with ‘five As and one B’ on their
report (Willis, 1977, p. 14). These ‘conformists’, who display a relatively
high acceptance of the authority of the school (pp. 98–99), receive scant,
almost dismissive, treatment from Willis. His ethnography is thin in various
directions.
Responding to later criticisms about this, Willis (1981) implores his read-
ers to recognize that his inattention to the ‘ear’oles’ does not mean that he
denies their involvement in cultural production: ‘it seems hard to assume
that I would forget in one instance what I’d been at pains to emphasize in
another’ (p. 61). That said he then goes on to admit that the ‘ear’oles’ were
presented as a foil for ‘the lads’, as a stylistic device rather than a theoretical
necessity (p. 62). As such, the conformity of the ‘ear’oles’ has a pejorative
ring to it, probably because their lack of resistance qua agency means that
they have no role to play in the critique of capitalism (Shweder, 2001). There
is apparently no potential for ‘a working-class movement with leftist
undertones’ in this group (Davies, 1995).
If it is true that for many critical researchers agency is usually, to be found
in resistant behaviours, then, intentional or not, this equation is problem-
atic. Such readings of the world render conformity as a form of submission
to dominant structures. The agency of the conformist is denied; they become
cogs in the wheel, mere ghosts in the machine. Yet, as Walford (1986, p. 23)
astutely observes of the ‘ear ‘oles’, by virtue of their being able to transcend
82 MARTIN FORSEY
their own working class backgrounds, they may well have been more active
and creative producers of original cultural forms than ‘the lads’ were. This
dichotomous typology of ‘resistant’ and ‘conformist’ individuals, no matter
how ideal these types might be, is too simplistic. It misses all of the subtle
nuances in the ‘tender interval’ between structure and agency (Deeds-Ermath,
2001, p. 44). For it is a delicate space, difficult, if not impossible, to discern.
As Deeds-Ermath (2001, p. 44) argues, we need to move away from dualistic
representations as ‘such either/or formulae hamper language’s capacity for
both/and’. As I discuss in the next section, this movement away from simple
binaries into more complex dimensions that pay attention to what exists
on the continuum between domination and cooperation, allows social
researchers to position themselves not only in more interesting places, they
are probably more productive as well.
STUDYING UP, STUDYING THROUGH
In the vibrant American anthropology of education scene over recent years

there have been a number of contributors to the Anthropology and Education
Quarterly advocating the need to heed more closely Nader’s (1972) exhor-
tation for anthropologists to ‘study up’. Emihovich (1999, p. 483) argued the
need to ‘confront issues of power and hierarchy in all forms of educational
setting’. Wisniewski (2000, p. 8) identified the need to conduct thick eth-
nographic research of university academics, while Priyadharshini (2003) and
Hamann (2003) explore Nader’s ideas in some depth. Hamann asserts that
Nader’s research agenda offers an opportunity to ‘revitalise and guide the
anthropology of education’ (p. 438) and Priyadharshini uses Nader’s ideas
more critically as a means of illuminating some of the obstacles faced by
those choosing to study those ensconced in powerfully lofty positions.
‘Studying through’ is how Shore and Wright (1997, pp. 14–15) in their
overview of anthropological research of policy name what they see as being
the most useful way of approaching the study of organisations. They em-
phasize the importance of recognizing ‘policy communities’ as ‘contested
political spaces’, of producing multisited ethnographies that ‘trace policy
connections between different organizational and everyday worlds’, and of
addressing questions about how particular discourses become prominent
and whose voices prevail in the policy setting. Taking a lead from Shore and
Wright, as well as Deeds-Ermath (2001, p. 44), I argue here the need to
move away from the either/or formulae of dualistic representations of social
life implied in Nader’s up and down dichotomy towards a ‘both/and’
commitment to research conducted in the in-between spaces that help

structure our social world. Priyadharshini (2003) reaches some similar con-
clusions, and in so doing advocates a move away from adversarial research.
However, her final comment, that an inquisitorial approach to power offers
‘an opportunity for critique from within that may not be available other-
wise’ (p. 434), helps illustrate how difficult it can be for social researchers
to escape the oppositional imperatives that seem to guide so much of the
anthropological project in educational settings. Critique is important, but,
as Goodman (1998, p. 55) argues, emancipatory ideology regularly replaces
portrayal as the central purpose of ethnographic research:
All of us are simply trying as best we can to make sense of the experience of living y [so]
clear, authentic ethnographic portrayal will likely be more educative than the ideological
insights of the researcher who made the observations in the first place (pp. 56–57).
CULTURAL RELATIVISM, CRITICAL

APPRECIATION AND THE IMPORTANCE OF
STUDYING ‘IN BETWEEN’
British sociologist of education, Maı́rtı́n Mac an Ghaill (1996) is one crit-
icalist researcher who has engaged in extensive ethnographic research
project among teachers. The teachers in his research site were experiencing
tumultuous change in their working lives and conditions and, based on their
responses to the neo-liberalist organisational changes taking place, Mac
an Ghaill identified three broad groups of teachers: Professionals; Old Col-
lectivists; and New Entrepreneurs. His sympathies clearly lay with the two
former groups in their struggle against the ascendant New Entrepreneurs.
Expressing his awareness that educational researchers tend to ‘caricature
social actors in terms of goodies and baddies’ (p. 184), he attributes his
apparent favouring of two groups over the other to the unintended effects of
using ideal types. Interestingly, he does not entertain the possibility that his
dualistic portrayal is reflected in, rather than explained by, the conceptu-
alisation of ideal types. However, it is in the next sentence, in which Mac
an Ghaill distances his thinking from ‘the adoption of a relativist position’
(p. 184, my emphasis), where the tendency to caricature social actors in
terms of ‘goodies’ and ‘baddies’ really lies.
Mac an Ghaill’s rejection of relativism was very evident in a lecture he
delivered to the Western Australian Institute for Educational Research in
August 1998. Responding to a question about power and the researcher, he
84 MARTIN FORSEY
made the disarmingly candid observation that: ‘When I am with students,

I hate teachers. When I am with teachers, I hate administrators. When
I’m with principals y well who can you hate?’ Along with most of my fellow
audience members, I laughed at this comment. It resonated strongly with
me. But the laughter in part reflects a failure on the part of many critical
ethnographers to take seriously the views of those we oppose. Mac an
Ghaill’s comment reflects a privileging of the subaltern that characterizes
much of the educational research enterprise. Coupled with rejection of rel-
ativism, the almost permanently downward focus of the ethnographic gaze
has the unfortunate epistemological consequence of limiting the researcher’s
abilities to comprehend what many criticalist researchers are surely all too
aware of, that power is marked by complex relationships between similar
and different groups located in various sections of the social hierarchy. As
Ortner (1995, p. 175) argues, subaltern groups are never singular or unitary
and their relationships with dominant groups are always characterized by
some degree of ambivalence. After all, ‘in a relationship of power, the
dominant often has something to offer, and sometimes a great deal’. Unless
we find ways of positioning ourselves, physically, emotionally and theoret-
ically, in the spaces between these sometime oppositional groups, we may
never come to recognize these sorts of realities. We condemn ourselves to the
sorts of ethnographic thinness and the concomitant failure to really grapple
with, and reach, deeper understandings of the practice of power.
Social researchers, especially those studying so-called Western societies,
often conflate cultural relativism with moral relativism (Fernandez, 1990).
For anthropologists at least, cultural relativism in the main seems reserved
for study of ‘the other’; yet it is arguably more important in so-called home
settings. I argue that ethnographers researching middle-class people need to
suspend their political critique for at least a little while, learning instead to
understand the people they work with and their particular life path. In other
words, in the search for a rigorous cultural relativism, researchers of social
spheres that incorporate people from elite and middle class walks of life
need to develop a mindset attuned to critical appreciation of their research
sites and the people in them.
While criticism is vastly more fashionable than appreciation in analyses of
industrial capitalist societies, with the former term carrying an air of gritty
realism in opposition to appreciation’s wide-eyed, functionalist innocence,
they have the potential to interlock. Considering the words side by side, they
interweave as different threads of the same tapestry. Committing myself to
appreciating peoples lived experience, no matter what their social situation,
is not to naively accept all that they believe and do. Rather, it helps create
a mindset attuned to giving social settings and the people living in them their
due value. Critique clearly has the potential to perform this analytical role,
but not in the way it currently shapes social thought. Social researchers
entering so-called middle class settings are encouraged to find and empha-
size the negative or the absurd (Marcus & Fischer, 1986; Goodman, 1998),
paying little heed to the fact that, as with villagers on remote islands, we
are observing ways in which people come to grips with their lives in the
particular historical and social moments in which they find themselves.
For those researchers committed to social justice and evoking subaltern
voices, such a stance can feel very awkward, perhaps even unnatural and
unethical. However, if we are to heed the earlier mentioned calls for deeper,
richer, ‘thicker’, more nuanced understandings of power and its social
effects, a commitment to critical appreciation of all aspects of social life is
vital. Sympathetic downward gazes, punctuated by highly critical upward
glances, may be emotionally and ideologically satisfying, but this does not
necessarily make for a satisfactory research process.
Haraway’s (1991) influential ideas on ‘situated knowledges’ alert us to the
need for social researchers to position themselves in methodologically and
epistemologically useful spaces. Problematizing the premium so often placed
on ‘establishing the capacity to see from the peripheries and the depths’
(p. 191), she argues for ‘passionate detachment’ and ‘mobile positioning’ as
the most useful response to ‘the impossibility of innocent identity politics and
epistemologies as strategies for seeing from the standpoints of the subjugated
in order to see well’ (p. 192). Drawing on scientific metaphors to make her
case, Haraway argues that, ‘One cannot ‘‘be’’ either a cell or molecule – or
a woman, colonized person, labourer, and so on – if one intends to see and
see from these positions critically. ‘‘Being’’ is much more problematic and
contingent’ (p. 192).
Studying ‘in-between’, in the interstices of social life is what I am advo-
cating (see Mulcock, 2001; Forsey, 2004). It is easy to say or write of course,
but not so easy to operationalize. What I am calling for in the shift in
mindset from critique to critical appreciation is a commitment to rich, thick
portrayal of the complexities of the social sites we find ourselves in and a
preparedness to report the views of a variety of social actors with tough
minded fairness (see Herskovits, 1977). In other words, we need to try to see
the social world from a variety of perspectives. More especially we need to
be prepared to stand in spaces that allow us to view our research sites from
directions towards which we are perhaps not naturally inclined.
There is nothing radical in these proposals of course; most would rec-
ognize what I am suggesting as solid methodological practice for any
86 MARTIN FORSEY
ethnographer. For a variety of reasons, however, many criticalists seem

to have forgotten this when it comes to portraying schools. One obvious
consequence of this is that the teachers tend to disappear from critical eth-
nographies of schooling, and our collective understanding of these important
sites of social action is the poorer for it. As Nader argued all those years ago,
it is time to shift our points of focus.
NOTE
1. Included in the count is issue 1 of the 2006 volume. Number 3 of the 2002 volume
was excluded because it was a special issue focused on an exceptional topic – the
aftermath of the September 11 attack on the World Trade Centre.
ACKNOWLEDGMENTS
My thanks to Bradley Levinson, Jane Mulcock and Geoffrey Walford for
their perceptive comments on earlier versions of this paper.
REFERENCES
Abu-Lughod, L. (1990). The romance of resistance: Tracing transformations of power through
Bedouin women. American Ethnologist, 17(1), 41–55.
Althusser, L. (1971). In: B. Brewster (Trans.), Lenin and philosophy, and other essays.
New York: Monthly Review Press.
Anderson, G. L. (1989). Critical ethnography in education: Origins, current status, and new
directions. Review of Educational Research, 59(3), 249–270.
Bell, C. (1978). Studying the locally powerful. In: C. Bell & S. Encel (Eds), Inside the whale.
Sydney: Pergamon.
Bourdieu, P., & Passeron, J.-C. (1977). Reproduction: In education, society, and culture. Beverly
Hills, CA: Sage.
Bowles, S., & Gintis, H. (1976). Schooling in capitalist America: Educational reform and the
contradictions of economic life. New York: Basic Books.
Carspecken, P. F. (1996). Critical ethnography in educational research: A theoretical and
practical guide. New York: Routledge.
Christian, B. (1987). The race for theory. Cultural Critique, 6(1), 51–63.
Connell, R. W. (1995). Transformative labour: Theorizing the politics of teachers work.
In: M. Ginsberg (Ed.), The politics of educators’ work and lives. New York: Garland
Publishing.
Connell, R. W., Ashenden, D. J., Kessler, S., & Dowsett, G. W. (1982). Making the difference:
Schools, families and social division. Sydney: George Allan & Unwin.
Davies, S. (1995). Leaps of faith: Shifting currents in critical sociology of education. American
Journal of Sociology, 100(6), 1448–1478.
Deeds-Ermath, E. (2001). Agency in the discursive condition. History and Theory Theme Issue,
40(December), 34–58.
Emihovich, C. (1999). Studying schools, studying ourselves: Ethnographic perspectives on
educational reform. Anthropology and Education Quarterly, 30(4), 477–483.
Fernandez, J. W. (1990). Tolerance in a repugnant world and other dilemmas in the cultural
relativism of Melville J Herskovits. Ethos, 18, 140–164.
Fischer, A. (1969). The personality and subculture of anthropologists and their study of U.S.
Negroes. In: S. Tyler (Ed.), Concepts and assumptions in contemporary anthropology.
Proceedings of the Southern Anthropological Society (3), 12–17.
Fordham, S. (1996). Blacked out: Dilemmas of race, identity, and success at capital high.
Chicago: University of Chicago Press.
Forsey, M. (2004). ‘‘He’s not a spy, He’s one of us’’: Ethnographic positioning in a middle-class
setting. In: L. Hulme, & J. Mulcock (Eds), Ethics, engagement and ethnography:
Anthropologists confront the field, New York: Columbia University Press.
Forsey, M. (2000). The anthropology of education: Cultural critique or ethnographic refusal?
Anthropological Forum, 10(2), 201–221.
Geertz, C. (1983). Local knowledge. New York: Basic Books.
Goodman, J. (1998). Ideology and critical research. In: J. Smyth, & G. Shacklock (Eds), Being
reflexive in critical educational and social research (pp. 50–66). London: Falmer.
Griffin, C. (1987). Review symposium – ‘schooling as a ritual performance’. British Journal of
Sociology of Education, 8(1), 87–90.
Griffin, C. (1985). Typical girls: Young women from school to the job market. London:
Routledge.
Hamann, E. T. (2003). Imagining the future of anthropology of education if we take Laura
Nader seriously. Anthropology and Education Quarterly, 34(4), 438–449.
Haraway, D. J. (1991). Simians, Cyborgs and women: The reinvention of nature. New York:
Routledge.
Herskovits, M. J. (1977). Cultural relativism: Perspectives in cultural pluralism. New York:
Random House.
Hoffman, D. (1999). Turning power inside out: Reflections on resistance from the (anthro-
pological) field. Qualitative Studies in Education, 12(6), 671–687.
Jordan, S., & Yeomans, D. (1995). Critical ethnography: Problems in contemporary theory and
practice. British Journal of Sociology of Education, 16(3), 389–408.
Levinson, B. A. (1999). Resituating the place of educational discourse in anthropology.
American Anthropologist, 101(3), 594–604.
Levinson, B., Holland, D. (1996). The cultural production of the educated person: An intro-
duction. In: B. Levinson, D. Foley, & D. Holland, (Eds), The cultural production of
the educated person: Critical ethnographies of schooling. New York: State University of
New York Press.
Mac an Ghaill, M. (1996). Sociology of education, state schooling and social class: Beyond cri-
tiques of the new right hegemony. British Journal of Sociology of Education, 17(2), 163–176.
Marcus, G. (1998). Ethnography through thick and thin. Princeton, NJ: Princeton University
Press.
Marcus, G. E. (1983). Elite as a concept, theory and research tradition. In: G. Marcus (Ed.),
Elites: Ethnographic issues. University of New Mexico Press: Alberquerque.
88 MARTIN FORSEY
Marcus, G. E., & Fischer, M. M. (1986). Anthropology as cultural critique: An experimental

moment in the human sciences. Chicago: University of Chicago Press.
Mulcock, J. (2001). Ethnography in awkward spaces: An anthropology of cultural borrowing.
Practicing Anthropology, 23(1), 38–42.
Nader, L. (1974). Up the anthropologist – perspectives from studying up. In: D. Hymes (Ed.),
Reinventing anthropology. New York: Vintage Books.
Ortner, S. B. (1995). Resistance and the problem of ethnographic refusal. Comparative Studies
in Society and History, 37, 173–193.
Priyadharshini, E. (2003). Coming unstuck: Thinking otherwise about ‘‘studying up’’. Anthro-
pology and Education Quarterly, 34(4), 420–437.
Quantz, R.A., (1992). On critical ethnography (with some postmodern considerations).
In: M. LeCompte, W. Millroy, & J. Preissle, (Eds), The handbook of qualitative research
in education (pp. 447–505). San Diego: Academic Press.
Reed-Danahay, D. (1996). Education and identity in rural France: The politics of schooling.
Cambridge: Cambridge University Press.
Shore, C., Wright, S. (1997). Policy: A new field of anthropology. In: C. Shore, & S. Wright
(Eds), Anthropology of policy: Critical perspectives on governance and power (pp. 3–39).
London: Routledge.
Shweder, R. A. (2001). Rethinking the object of anthropology and ending up where Kroeber
and Kluckhohn began. American Anthropologist, 103(2), 437–440.
Walford, G. (1986). Life in public schools. London: Methuen.
Willis, P. (1981). Cultural production is different from cultural reproduction is different from
social reproduction is different from reproduction. Interchange, 12(2–3), 48–67.
Willis, P. E. (1977). Learning to labour: How working class kids get working class jobs. Aldershot,
Hampshire: Gower.
Wisniewski, R. (2000). The averted gaze. Anthropology and Education Quarterly, 31(1), 253–274.
‘CASE STUDY’ OR
‘ETHNOGRAPHY’? DEFINING
TERMS, MAKING CHOICES AND
DEFENDING THE WORTH OF A
CASE
Sue Walters
There is often a lack of clarity about the terms ‘ethnography’ and ‘case
study’ and what differentiates them as research terms. This is acknowledged
by many (e.g., Burns, 2000, p. 459; Hammersley, 1992, p. 183; 1998, p. 1;
Punch, 1998, pp. 152, 162; Yin, 2003, pp. 12–13; Platt, 1992, p. 46 cited in
Yin, 2003). The lack of clarity arises partly because the terms ‘ethnography’
and ‘case study’ can be used to describe a research strategy, a research focus
(the choice of phenomena to be studied), the research methods (or proce-
dures) used and/or the results of a research study.
For a new researcher, particularly those preparing, conducting and writ-
ing up their first qualitative study, the research literature available, and I am
thinking here of those books that introduce or discuss research methodology
(and are aimed at research methods course reading lists), can be bewildering.
In my own research into the educational experiences of a small group of
early years Bangladeshi pupils at school in the east of England, searching for
the correct way to describe my study, I found that different authors defined

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12006-9
89
90 SUE WALTERS
he terms I was encouraged to use, ‘ethnography’ and ‘case study’, in pro-

foundly different ways. There was a lack of clarity across and within texts
about what differentiated these terms and even whether the terms referred to
methods of data collection, research strategies, objects of study or the final
product of research. While the different uses of the terms may reflect a
healthy dynamic engagement with the emerging and fluid forms of
qualitative research work, or point to the contested nature of research
terms, they can also create great confusion to new researchers starting to
work in the qualitative field. Mindful of Daniels and Akehurst’s comment
that ‘learning to be a researcher is like learning to speak a language’
(in Birbili, 2001) I start by detailing the thinking that went into sorting out
the way to describe and document my research study. I briefly consider how
the terms ‘ethnography’ and ‘case study’ were used in the available literature
and how I went on to develop clear and workable definitions in order to
present my work. I am not here describing the process of creating an ap-
propriate research design; I knew what I wanted to do, how to do it and
why. I am discussing the process I went through to find a way of talking
about and representing my research, in order to decide which words to use
when the words on offer could potentially mean many different things.
What is outlined here is my process of working out how to name and
describe what I was doing. I then turn to a consideration of, first, the
importance in my research of conducting ethnographic case studies of
individual children and not of classrooms and second, of how, what I came
to term, ‘ethnographic case study’ can be of use outside of the specific micro
setting of its production, thus addressing concerns about the generalizability
and usefulness of this kind of research.
RESEARCH LITERATURE: DEFINITIONS AND

DISCUSSIONS1
Writers present very different descriptions of what constitutes ‘ethnography’

and ‘case study’. While not attempting to review the literature, nor to com-
ment on or endorse the overall quality of argument and presentation of
research methodology made by each of the authors considered here, the
following demonstrates how different the explanations and definitions
offered can be.
Cresswell defines ethnography as ‘a description and interpretation of a
cultural or social group or system’ in which the researcher studies the
‘Case Study’ or ‘Ethnography’? 91
Ethnography Case Study

An Object of Research A social group or entire cultural or A bounded system (a programme, an
Study social system event an activity, or individuals but not a
system of people)
A Method of Research Prolonged observation of the group, Can use different methods that focus on
Study typically through participant the collection of in-depth data from
(Approach/Process) observation and one-to-one interviews multiple sources rich in context -
(using anthropological concepts) documents, archieval records, interviews,
– always qualitative. direct and/or participant observation,
physical artefacts
- can include quantitative methods and
analysis (may or may not use
anthropological concepts)
A Product of Research A holistic portrait of a group or system Can result in very different products - an
Study in-depth analysis of a programme, an
(Outcome) event, an activity or individuals
With the Aim Of Understanding the way a group or Understanding the uniqueness of a case
social system works, the meanings it or to providing an illustration of an issue.
gives to actions, artefacts and so on.
Ethnography investigates people in
interaction in ordinary settings, it looks
for patterns of daily living (culture),
what people do, say and use to find out
what a stranger would have to know in
order to be able to take part in the
group or society in a meaningful way.
Fig. 1. A Summary of Cresswell’s Delineation of ‘Ethnography’ and ‘Case Study’.
meanings of behaviour, language and interactions of the culture sharing

group (Cresswell, 1998, p. 58) case study as an ‘exploration of a bounded
system y over time’ (Cresswell, 1998, p. 61) and understands them as trying
to accomplish different things: see Fig. 1. While Cresswell’s notion of the five
traditions, within which ‘ethnography’ and ‘case study’ are placed, seems
flawed, his discussion of the two terms helpfully highlights the fact that, in his
view, while ethnography is both an object of study (that object being
a cultural or social group or a system); a method (a research approach or
process) and a product of research (a holistic portrait) and cannot be quan-
titative, case study can utilize ‘multiple sources of information’, it can result
in very different products (outcomes) but it has a specific object of study
(a bounded system). Such a differentiation between the two terms seemed a
good working definition (and Cresswell himself notes the potential overlap
between ‘ethnography’ and ‘case study’ when researchers study a system). He
further differentiates the terms by stating that in an ‘ethnography’ the entire
92 SUE WALTERS
cultural or social system is the focus of attention, whereas in ‘case study’ a

system of people ‘is typically not the case’ (Cresswell, 1998, p. 66).
However, Hammersley (1992, p. 183; 1998, p. 1), while acknowledging
the overlap and fluidity in the meanings of the two terms uses ‘ethnog-
raphy’ to mean the much broader ‘qualitative method’ and ‘case study’
becomes ‘a selection strategy’ alongside ‘experiment’ and ‘survey’. A case
being ‘the phenomenon (located in space/time) about which data are col-
lected and/or analysed and that corresponds to the type of phenomena to
which the main claims of a study relate’ (Hammersley, 1992, p. 184). Sim-
ilarly, for Silverman (2005, p. 126), the term ‘case study’ refers simply to
the particular topic or phenomena the research sets out to study. In this
respect every research study is a study of a case. Along similar lines,
Brewer, like both Hammersley and Silverman in writing about qualitative
research, states that ‘While not all case studies are qualitative, all
ethnographic research involves case study’ (Brewer, 2000, p. 77). Stake
(1995, p. 86) prefers to take the line that case study is a choice of what is to
be studied rather than being a methodological choice and, like Cresswell,
that what distinguishes case study is its focus on a bounded system.
Merriam follows this by defining ‘case study’ as ‘an examination of a
specific phenomenon, such as a programme, an event, a person, a process,
an institution or a social group using both qualitative and quantitative
methods’ (Merriam, 1988, p. 9). She defines ‘ethnography’ as a set of
methods used to collect data, and as a written record, but not as an object
of study (Merriam, 1988, p. 23). This makes her definition different to
Cresswell’s. Travers entitles his book Qualitative Research Through Case
Studies but ‘case study’ does not appear in the index nor in the contents
page and is used in the Preface to refer to research studies that are used as
an ‘example’ to illuminate a particular research approach (Travers, 2001,
p. viii). Yin, while also acknowledging that ‘case study’ is often confused
with ‘ethnography’ (Yin, 2003, p. 12), defines ‘case study’ as ‘an empirical
inquiry that investigates a contemporary phenomenon within its real life
context, especially when the boundaries between phenomenon and context
are not clearly evident’ (Yin, 2003, p. 13 my emphasis).
Presented with so many differing or overlapping definitions and descrip-
tions I was unable to simply utilize the terms to describe my research design
and methodology without thinking deeply about the delineations I was
offered and how I intended to use the terms in my work. What follows is an
outline of my thinking and how I developed workable definitions as I ‘learnt
to speak the language’ of contemporary research methods.
DEFINING TERMS
For the sake of clarity I chose to follow Stake who, as we have noted, presents
‘case study’ as not a methodological choice but a choice of object to be
studied (Stake, 1995, p. 86). I took as the choice of object to be studied in
an ‘ethnography’ a social or cultural group or system and the aims of
‘ethnography’ as understanding the way a group or social system works, the
meanings it gives to actions, artefacts and so on. ‘Ethnography’, thus, inves-
tigates people in interaction in ordinary settings, it looks for patterns of daily
living (culture), what people do, say and use in order to find out what a
stranger would have to know in order to be able to take part in the group or
society in a meaningful way. This results in a holistic cultural portrait of the
social group (taken from Cresswell, 1998, pp. 58–60). ‘Case study’ on the
other hand, following Stake, takes as the focus of its study a bounded system
of some kind. This can be an individual, an event, an organization or a
cultural group (or a number of these). Thus I began to conceive of the
difference between the terms as this: ‘case study’ does not ignore context but it
focuses on the case situated within a context. The emphasis of the investi-
gation is on the case. Thus, in this view, ‘ethnography’ provides a holistic view
of a social group or culture, ‘case study’ an in-depth study of a bounded
system or case or set of cases. ‘Case study’ is able to explore a range of topics,
while (traditionally) ‘ethnography’ focuses on cultural behaviour, practices
and artefacts. In this way ‘case study’ differs from ‘ethnography’ in that it is
not seeking to understand a social group or system (unless a group is taken as
a case) it is seeking to understand a case within an acknowledged social
system. This can be represented in the following manner.
social group social group the case
or culture or culture
or social or social
system system (context)
Ethnography Case Study
MAKING A CHOICE
Either choice of study would have been appropriate for my study of the
experiences of Bangladeshi pupils in school in the context of their perceived
94 SUE WALTERS
underachievement. I was interested in exploring the experiences of a small

group of Bangladeshi pupils in school in order to understand what was
helping and hindering them in their learning. Previous statistical research
had highlighted the fact that Bangladeshi pupils as a group were
underachieving in the school system (e.g., Kysel, 1988; Nuttall & Varlaam,
1990; MacIntyre, Bhatti, & Fuller, 1993; Demack, Drew, & Grimsley, 2000;
Pathak, 2000) but was unable to properly account for why this might be so.
Various qualitative studies in the 1980s and 1990s had considered the
learning experiences of minority ethnic pupils and concluded that such pu-
pil’s poor achievement could be accounted for in terms of the different
amounts or kinds of attention that different (ethnically defined) groups of
pupils received from their teachers or that setting and banding arrangements
in schools were key in creating disadvantage for certain minority ethnic
groups (e.g., Wright, 1992; Gillborn, 1990; Troyna, 1991; Connolly, 1998a,
1998b). However, none of these studies had focused on Bangladeshi pupils
and their findings did little to illuminate what might be happening for
Bangladeshi pupils in terms of their learning and achievement. I could
therefore have chosen to conduct what I have defined above as an
‘ethnography’ of a school or classroom, looking, like previous qualitative
researchers, at the complex interactions, behaviours and meaning making of
teachers, pupils and so on but focusing on Bangladeshi pupils in order to
understand something of the ways in which pupils were supported or hin-
dered in their learning and achievement. Or I could have chosen to conduct
a number of what I have defined above as ‘case studies’ of individual
Bangladeshi pupils with the same ends in sight. A choice of study that
focused specifically on a child, or a number of children, rather than on the
social group or the social system of a classroom seemed a more appropriate
focus for understanding what the experiences of being a pupil in a classroom
and of learning and taking part in the classroom were like. I was not in-
tending to get a holistic view of a classroom, or a number of classrooms or
school cultures but I did want to get at the experiences of specific children in
the context of the classroom. The previous qualitative research noted above,
which had considered minority ethnic children in the classroom, had started
with the social system, the culture of the classroom and through that had
shown how identities were produced and resources subsequently allocated to
pupils. I wanted to look ‘right up close’ at specific children’s experiences of
being in the classroom by starting with a specific child or number of iden-
tified children and seeing what was happening to them in specific groups and
locations rather than exploring the dynamics of particular classrooms. The
starting point for the study would be a specific child or a group of children
not classrooms. In this respect I decided that my research would be called

‘case study’.
Case study as a research strategy allows the researcher to undertake a single
case study or a number of case studies. I decided to undertake a number of case
studies rather than a single case study so that I would have a range of ‘portraits’
to illustrate the experiences of Bangladeshi pupils.2 An important aspect of the
research was to consider the differing experiences Bangladeshi pupils might
have in relation to each other and to get away from the notion that children,
once ascribed to an ethnic group, must have much the same experience and
needs in the classroom. In this respect, choosing to use ‘case study’ rather than
‘ethnography’ as my research strategy was an important research decision. In
addition to this, a number of case studies of individual children would allow me
to carry out ‘within-case analysis’ (a detailed description of each case and
themes within the case) followed by a cross-case analysis (a thematic analysis
across the cases) (Cresswell, 1998, p. 63) and a discussion of ‘lessons learnt’
(Lincoln & Guba, 1985). By being able to do ‘within-case analysis’ I would be
able to avoid the danger of only considering Bangladeshi pupils in relation to
each other and be able to consider them as individual pupils in a classroom.
Combining ‘within case analysis’ with ‘cross-case analysis’ would allow for a
level of depth as there would be the opportunity, later in the research, to
explore any emerging themes or patterns or dissimilarities. In this respect,
conducting the research as a number of case studies rather than as an eth-
nography was an important research decision.
However, before turning to the issue of the use of a small number of
portraits of individual pupils as a means of contributing to knowledge about
Bangladeshi pupils’ schooling experience and the issue of underachievement it
is necessary to say why I chose to describe my research strategy as
ethnographic case study rather than simply as ‘case study’ or ‘qualitative
case study’.
ETHNOGRAPHIC CASE STUDY AND MOVING

BEYOND NOTIONS OF ‘BOUNDEDNESS’
At the heart of the definition of case study I chose to adopt (Stake, 1998,
p. 86) there is an emphasis on the notion of ‘boundedness’ with which I was
uncomfortable (Stake, 1995, p. 9; Cresswell, 1998, p. 61). Stake speaks of the
aim of case study being ‘to thoroughly understand the bounded case’
(Stake, 1995, p. 9) and goes on to say that a child can constitute a case
because s/he is ‘a working combination of physiological, psychological,
96 SUE WALTERS
cultural, aesthetic and other forces’ (Stake, 1995, p. 436). Although indeed a
child could be considered to be a bounded unit, I am in favour of seeing
children in context and in interaction, as created in and by the social proc-
esses in which they take part and not as separated, bounded identities (Yon,
2000). Case study, as defined by Stake, would seem to have an unspoken
affinity with modernist conceptions of subjectivity which do not sit easily
with my approach to understanding identity and subjectivity as multiple and
fluid. In the light of this I chose to think of my research strategy as case
study because the phenomena that the research was investigating was the
child or the children rather than the classroom or the school and as eth-
nographic because I did not consider the child or children to be bracketed off
from the culture, the interactions and the context around them. Case study,
as presented by Stake, does recognize the importance of context, but for my
study I wanted to be clear that context was not some kind of background
scenery or ‘noise’ but was an integral part in constituting who and what the
child, or children, were. In this respect using the term ‘ethnographic’ with
‘case study’ signals that the social system or cultural group that the child
studied was part of was an important part of constituting who and what that
child was and could be. This understanding of ethnographic case study
could therefore be represented in the following way (where I have taken the
previous elements of ethnography and case study and combined them).
social group
or culture the case
or social
system
Ethnographic Case Study
I also chose to term my research strategy ‘ethnographic case study’ be-

cause my research focus required me to use many of the same research
methods as an ethnographer, these being:
the need to enter into a close and relatively prolonged interaction with
people in their everyday lives (Tedlock, 2000) and actively participating as
a member of the social group (Emerson, Fretz, & Shaw, 1995);
suspending premature judgements about what should be selected as data,
creating analytical frameworks from interactions with informants and
allowing for new questions and avenues to be explored to emerge from the
fieldwork (Walford & Massey, 1998, pp. 5–9; Stake, 1995, pp. 3, 9, 22, 33;
Spradley, 1980, p. 29);
looking at the research situation from many perspectives by gathering data
from many sources and in a variety of forms (Walford & Massey, 1998,
pp. 5–9; Merriam, 1988, p. 69) thus capturing multiple realities
(Stake, 1995; Adelman, Kemmis, & Jenkins, 1980); and
paying attention to subtlety, complexity and embeddedness (Adelman et
al., 1980).
and because the final outcome of the study would share a lot with the outcomes
of an ethnographic study, namely, an account of processes, practices and ex-
periences and what happens over a period of time through which phenomena
and processes can be understood within their particular social, historical and
spatial contexts. Such an outcome would include a richness of detail rather
than the ‘forgetting’ of statistical research (Keith, 1993) and provide an ac-
count of ‘the values, practices, relationships and identifications of people’ and
explain ‘What is going on here? How does this work? How do people do this?
(Walford & Massey, 1998, p. 5) although presented as a series of case studies.
A SINGLE CASE, MULTIPLE CASES AND THEIR

WORTH
A criticism of ‘case study’, one that has an important bearing on my own

study, is that findings cannot be generalized or applied to other settings and
because of this they are of little use to policy and practice and to wider
understandings of social phenomenon (e.g., educational achievement). Case
study and qualitative research in general, including ‘ethnography’, have
been defended from this criticism in a number of ways. These can be
grouped as follows, running from defences which claim that such research
need not be generalizable in and of itself, through those that say it can be
generalizable in ways that are different to survey and quantitative work to
those that claim that ‘case study’ research is generalizable in the same way as
these forms of research.
1. Exploratory
This defence of the importance of ‘case study’ and other forms of qualitative
research sees the findings of such work important in providing insights,
98 SUE WALTERS
language and categories that can be utilized in the design of other, usually,
quantitative studies. For example, such a study can reveal important aspects
of a person or group of people’s experiences and the language that they use
in order to talk about these experiences. These experiences and the language
used can then be incorporated into the design of a survey, which is then
administered to a large, statistically sampled group. In this respect, such a
case has an important exploratory role to play in advance of another study.
2. Particularity
Some researchers have defended ethnography and qualitative case study by

speaking of the importance and necessity of having things depicted in their
particularity (Stake, 1995; Keith, 1993). Connolly (1998a, 1998b) claims that
it is a mistake to expect to be able to generalize from ethnographic research
and, like Stake and Keith, he defends the remit of such work as being
concerned with identifying and understanding particular social processes
and practices. Sharrock and Anderson (1982, pp. 172–173) defend their
research on these grounds too.
3. ‘Naturalistic Generalization’
Related to this focus on particularity, Stake and Donmoyer have both de-
fended ‘case study’ by claiming that such work can have a ‘general relevance’
(Stake, 1995) through the manner in which it is consumed by its readers
and how such a study becomes part of a reader’s knowledge of the world.
Stake refers to this as a process of ‘naturalistic generalization’. According to
Stake, such studies provide ‘vicarious experience’ in the form of a ‘full and
thorough knowledge of the particular’ and in this way facilitate ‘naturalistic
generalization’ (Gomm, Hammersley, & Foster, 2000, p. 7). Through reading
a case study, people are able to build up a body of tacit knowledge on the
basis of which they can act. This depends on the case study being described
properly ‘in a way that captures its unique features’ (Gomm et al., 2000, p. 7).
Donmoyer bases his approach to generalization, like Stake, on ‘experiential
knowledge’ (Donmoyer, 2000, p. 55). He sees case study as able to facilitate
learning in the reader by substituting for first hand experience (Gomm et al.,
2000, p. 9). The reader consumes a case study through assimilating and ac-
commodating what is read and integrating and/or differentiating this with
what is already known (Donmoyer, 2000, pp. 59–61).
4. Transferability
Lincoln and Guba (2000) claim that there are ways of stating conclusions
arising from the study of one case or context that might hold in another
case or context. For them, case study research produces ‘working
hypotheses’ a term they take from Cronbach (Lincoln & Guba, 2000,
p. 39). The transferability of conclusions from one case to another is a
function of the ‘fit’ between the two contexts. Therefore, case study
researchers have to provide ‘thick descriptions’ of their research case,
particularly of the context of the case, for this to be possible (Gomm et al.,
2000, p. 8). This differs from Donmoyer and Stake in that ‘transferability’
as defined here can only be made if the settings and contexts of the studies
are similar. Punch also chooses to talk of ‘transferability’ rather than
‘generalization’, researchers needing to be mindful of sampling, the
concepts used in data analysis and of describing the context in enough
detail for the reader to be able to judge whether the findings are trans-
ferable to other settings (Punch, 1995, pp. 255–256).
5. Fuzzy Generalization
Bassey defends qualitative case study against the criticism that one cannot
generalize from it by claiming that case study can make ‘fuzzy generaliza-
tions’, by this he means ‘the kind of statement which makes no claim to
knowledge, but which hedges its claim with uncertainties y in some cases it
may be found that y .’ (Bassey, 1999, p. 12).
6. Analytical Generalization
This defence argues that while a case study cannot be used to generalize
to a wider populations (through statistical generalization) it can be used
to generalize to theoretical propositions i.e., case studies can be used to
contribute to or expand on current explanations in relation to social issues
(Yin, 2003, p. 10). ‘Whilst survey research relies on statistical generaliza-
tion, case study research and experiments rely on analytical generalization
where the researcher is striving to generalize a particular set of results to
some broader theory’ (Yin, 2003, p. 37). However, a case study, just
like an experiment, has to be replicated in order for generalizations to be
made.
100 SUE WALTERS
7. Empirical Generalization
Some defend case study by claiming that it is possible to make generalizations

from case study in the same way that it is possible to make generalizations from
survey research (Gomm et al., 2000, pp. 5, 9). Starting with an acceptance that a
great deal of everyday knowledge building consists of ‘naturalistic generaliza-
tion’, Gomm, Hammersley, and Foster note that such generalization is subject
to a great deal of error when considered alongside ‘statistical generalization’
(Gomm et al., 2000, p. 104). They suggest that case study researchers can
improve on ‘naturalistic generalization’, and therefore on the accuracy of their
informal empirical generalizations, by taking account of the relevant hetero-
geneity within the population with which they are concerned (Gomm et al.,
2000, p. 105) and consider the typicality or non-typicality of the case that is to
be studied in relation to the relevant population. ‘In short, it is necessary to
compare the characteristics of the case(s) being studied with available infor-
mation about the population to which generalization is intended’ (Gomm et al.,
2000, p. 105). They also suggest that case study researchers, instead of focusing
on a case, or cases, that are typical, can choose to study a small number of cases
that are atypical and at the extreme ends of the range of heterogeneity identified
in the relevant population (Gomm et al., 2000, p. 107). They argue that there
are two strategies, used by experimental and survey researchers, for drawing
conclusions from the case, or cases, to the larger population these being ‘the-
oretical inference’ and ‘empirical generalization’ (Gomm et al., 2000, p. 102).
Key to this form of ‘empirical generalization’ is the need to know as much as
possible about the relevant wider population that the case or cases represent and
to select and present the case or cases carefully in the light of this knowledge.
8. ‘Holographic Generalization’
Using the metaphor of holographic film, Lincoln and Guba suggest that, as
with the film, which if cut into pieces each piece will contain within itself all
of the information and image stored on the whole film, ‘the full information
about a whole is stored in its parts’ (Lincoln & Guba, 2000, p. 43). Thus, in
case study research ‘samples need not be representative in the usual statis-
tical sense to render generalizations warrantable; any part or component is a
‘perfect’ sample in the sense that it contains all of the information about the
whole that one might ever hope to obtain’ (Lincoln & Guba, 2000, p. 43).
It is worth noting that some of the defences above refer to the importance
of the writing up of the case so that the reader can gain ‘vicarious
experience’ (Stake) or ‘transfer’ knowledge (Lincoln and Guba; Punch) or

‘learn’ (Donmoyer) yet these comments tend not to develop issues concern-
ing the writing and reading of research. First, there is an assumption that
research writing acts as a straightforward conduit carrying what happened
and what was said ‘in the field’ directly (and untainted) into the reader’s
mind. Yet writing, by its very nature, is restricted by its generic forms and
styles; what one writes on the page is an act of framing, editing and pres-
entation just as the way a photograph brings its own styling, framing, ed-
iting, focusing and ‘leaving out’ to the image that is consumed. Neither a
written case study, nor a photograph, is neutral; both have contributed to
and construct the vision ‘seen’ by the consumer, the reader. In a similar
manner, while Stake and Donmoyer have considered how a reader ‘deals’
with reading a piece of case study research (through integrating what they
read with what they already know or have experienced), other writers do not
address how case study research is read and consumed, i.e., what readers do
with case study research descriptions.
The defences of case study as outlined above also need to be considered in
relation to understandings of the social world. The writers who argue that
one can generalize from case study research as from survey research or by
using the strategy of ‘theoretical inference’ (Gomm et al., 2000, p. 102) are
working within a view of the social world that can confidently expect case
study researchers to be able to work with notions of universality, known and
fixed populations and so on, while many writers, including this one, do not
share this epistemological stance. For post-positivist researchers (those that
reject universality and the notion of a reality beyond a socially produced
one) these defences are less acceptable ones. Rather they will defend case
study and its usefulness on other grounds (including some of those noted
above) as I do here.
DEFENDING THE WORTH OF CASE STUDY: AN

EXAMPLE
In order to consider how a case study can be of use outside of its setting, I
present a brief overview here of one of the case studies from my research
study. There were six case studies in total all conducted at the same time
during the course of one academic year. The six pupils were the total pop-
ulation of Bangladeshi pupils in Year 3 in the county, attending three
different schools in the city. The children lived in various parts of the city
102 SUE WALTERS
with their immediate families. Fathers were the main breadwinners and were
all working in the restaurant trade. The children’s mothers were all house-
wives. Rahul3 the subject of this case study, was one of the two boys in the
study.
Rahul’s case study was structured to present Rahul in relation to the
following headings: as a learner, as a reader, as a pupil taking part in
classroom interactions, as a multi-lingual language user, as a social member
of his class, at home and at school and teachers’ response to Rahul as a
pupil. The headings reflected the manner in which the data was analysed and
arose out of the initial readings of the data. (The same headings were used
for all of the case studies.) Through these headings Rahul’s case study de-
picted the spaces and the opportunities provided by the curriculum and by
the teacher for pupils to ‘take part’ in classroom interactions, how Rahul
‘took part’, his needs as a learner and as a bilingual pupil, the manner in
which his teachers’ assessed his needs and the support that he was subse-
quently provided with and how Rahul subsequently became positioned as a
particular kind of pupil, one unlikely to be a successful achieving pupil. A
brief summary of what is described in Rahul’s case study is provided below.
Rahul was a seven-year-old Bangladeshi boy who, because of a number of long absences
from school in the first two years of his schooling was beginning Year 3 of primary
school with a little knowledge of spoken English and a very limited experience of the
curriculum and of the English that he encountered in his classroom.
The three Year 3 classrooms I conducted my research in all followed a very similar use
of space in the classroom and in Rahul’s classroom, like the others, introductions to the
different subjects such as Literacy, Numeracy, History and so on were conducted by the
teacher with the children sitting before her on a carpet in a corner of the classroom.
Rahul’s teacher would introduce a topic and build into her talk questions to the children
that the children had to raise their hands to answer. After this ‘introduction’, which in
Rahul’s classroom usually lasted between twenty minutes and half an hour, the children
would be sent to work in their groups at desks in order to undertake a piece of work set
by the teacher and explained during the introductory talk. Rahul’s teacher used a lot of
talk and very little visual demonstration or example of what she was talking about at
these times and the amount and level of English language in use in the classroom was
well beyond Rahul’s experience and understanding. Rahul presented himself during
these times as a very disengaged pupil. After about five minutes of teacher talk he would
‘switch off’ and slide to the back of the group of children. He did not raise his hand to
answer questions and he ceased to look at the teacher. Whilst aware of Rahul’s beginner
bilingual status, Rahul’s teacher came to understand and assess Rahul as a lazy pupil
who couldn’t be bothered to make an effort to follow what was happening in class. Her
early efforts to support him by holding spaces for him in class discussions and encour-
aging him to answer in any way he could, changed to simply supporting him by asking
him at the end of every introductory session if he knew what he had to do when he left
the carpet and went to his desk. The teacher usually found that Rahul did not know
what he was expected to do and so she went through the instructions for the task with
him before sending him off to his seat.
Rahul’s teacher also attempted to support him as a learner by placing him in the lower
‘ability’ Literacy and Numeracy groups in the classroom as she worked with these
groups more often than with others and she felt that she could offer Rahul more support
in this way. However, the teacher was frequently distracted and ‘taken away’ from her
work with these groups by the demands of other children or the demands of particular
children in these groups and Rahul did not gain much personalised support at these
times. Later in the school year when Rahul’s reading had improved and he would, by the
teacher’s own admission, have been better served by moving up into a higher Literacy
group, he was forced to stay in the low level group he had been placed in because of a
school policy – the next group up was funded for a particular category of pupil to which
Rahul did not belong. Thus Rahul became stuck in a low group for Literacy to which he
no longer really belonged and which was not able to stretch him sufficiently.
Rahul’s teacher and other school adults lacked specific training in working with bi-
lingual pupils and so made judgements about bilingual children’s level and competence
in English based on the children’s social fluency in English. As Rahul began to speak and
use English in social situations more frequently his teacher thus judged that his English
was developing enough for him to deal with the English that he was hearing and ex-
pected to work with in the classroom. As a result of this Rahul’s teacher did not think
that he required very much additional support from the specialist peripatetic English as
an Additional Language teacher that visited the school once a week, except for some
additional reading. She and the other school adults also thought that Rahul’s sister in
the class below was more in need of support from this teacher because ‘she doesn’t get as
much as Rahul at home’. Thus Rahul received very little support for his English and
learning from this service and when he did the manner in which such support was
delivered meant that Rahul was taken out of the classroom by the visiting teacher half
way through an introductory session and returned when the children were at their desks
half way through completing the task they had been set. This also happened when the
classroom assistant was asked to provide some additional support by hearing Rahul
read. At these times Rahul who had missed the full introduction to what was being done,
the instructions about what was to be done and a good deal of the time allotted to
completing the task, became very frustrated at not being able to complete his work
alongside the other children. As part of the research I was able to document in great
detail how Rahul’s school day became an extremely fragmented day as he moved in and
out of the classroom as part of the support that was provided to him.
Over the course of the year the case study depicts how Rahul’s teachers increasingly come
to describe him as a ‘lazy’ pupil, reading his disengagement as a lack of motivation rather than
as a difficulty with working with the English of the classroom and with the fragmented nature
of his school day. It also documents how Rahul came to receive little appropriate support.
The case study was also able to show how Rahul’s isolation from other children, he
had few friends in the class, and his lack of opportunities to use and practice his English,
affected his development of English. It was also able to show his isolation outside school,
he did not see very much of his few relatives that lived on the other side of the city and
did not often play with other children, except his sister. It was able to show the differ-
ences between the kind of support that Rahul had from his parents in relation to learning
in his English school and that of his class peers whose parents were usually more richly
104 SUE WALTERS
resourced in the ways of the English education system and time available. Through the
case study, which followed Rahul through a year of his schooling, it became possible to
see the subtle and complex ways in which Rahul became positioned as a potentially
underachieving pupil.
Using Rahul’s case study as an example I want to argue that this case study
has worth and can be useful outside of its setting for the following reasons:
It can provide concepts, language to other studies. It allows other re-
searchers to see issues that need to be included in other studies (as issues
or variables) for example, teachers’ decision-making about their pupils
and their needs, teachers’ tacit ways of assessing pupils. It can raise new
questions and areas for study. (This can be linked back to the defence of
case study and qualitative work discussed under the heading of Explor-
atory above. For each of the reasons that follow I provide, in brackets at
the end, this link back to the previous discussion).
It gives and account of the processes of everyday life for a pupil in a
classroom in all their subtlety. It presents the reader with a description of
practices of a classroom, how the teachers, pupils and Rahul understand,
make meaning and accomplish tasks. It provides a thorough account of
Rahul’s experiences during his Year 3 (and those of his teachers, peers,
parents and siblings that pertain to Rahul as a pupil) in this way illu-
minating the key processes and practices that Rahul’s experience is em-
bedded in. In this way, policy and practitioners can deepen their
understanding of how underachievement can come to be played out in
a classroom, how national and school policy can play a part in this (for
example, setting arrangements, the curriculum, teachers’ training, EAL
support, policy and assessment). It also, through its placement alongside
five other case studies of Bangladeshi pupils of the same age and attending
school in the same area, challenges the assumption that all Bangladeshi
pupils have the same learning needs. (Particularity)
Readers of the case study are asked to what extent, if any, it fits with what
they already know or think/believe. Does it challenge what they already
know? Reading the study and how it depicts Rahul will raise issues and
questions in relation to what the reader already knows or has read (and
which the reader will go on to read or experience for themselves) even if it is
only to disagree. The study raises sensitivities or questions that a person
might take with them into another setting after reading the case study and its
presentation of Rahul. (Naturalistic Generalization and Transferability)
The case study, and its presentation of Rahul, makes a case for certain
theoretical issues to be included or accepted in other research. For
example, it asks for the manner in which the positioning of pupils, the
spaces they are offered by teachers, the curriculum and by the socio-
spatial organization of the school in which to show who they are and what
they can do, to be considered in other research and in policy and practice.
It makes a case for looking at things in a particular way and demonstrates
how this can be illuminating and useful. (Analytical Generalization)
The case study’s presentation of Rahul provides a description of a sit-
uation and the study can thus be of use in thinking about this situation
and what action can be taken by policy makers and practitioners in re-
lation to Rahul’s opportunities as a learner and achieving pupil. In this
way, the study can be used as a learning tool in professional training and
in policy discussions (much as case studies are used in the management,
medical and legal worlds). It is a way of presenting a scenario and asking
trainees, professionals and policy makers how they could work with the
situation presented – and how they understand it. It can at the same time
be used to encourage ‘critical’ readings of research accounts through
asking what is missing from the account and how the account (re)disposes
the reader to certain understandings and leads them away from others.
This last reason is not covered in the earlier discussion of the ways in
which case study and qualitative research can be defended from the criticism
that it is of little use to policy and practice because of its lack of gener-
alizability. The defences of ‘particularity’, ‘naturalistic generalization’ and
‘transferability’ are certainly applicable but I prefer to refer to this under-
standing of the worth of case study as ‘pragmatic insight’. I take the use of
the term ‘pragmatic’ in relation to case study work from Kenny and
Grotelueschen (1980, p. 11, cited in Merriam, 1988, p. 20). They use the term
to emphasize the applied nature of case study research and I choose to use it
because it highlights the practical use of case study research to policy and
practice. In addition to this, I see as integral to the usefulness of case study
work the opportunities it offers for engagement with how the case is pre-
sented and for critical readings of what is and is not included in the account.
CONCLUSION
In this chapter, I have outlined the process I went through in order to find a
way of talking about and representing my research. I have discussed the
process and importance of ‘defining terms’ so that it was possible to describe
my research design and process despite the range of competing and differing
106 SUE WALTERS
definitions and delineations of ‘case study’ and ‘ethnography’ on offer. At the

same time, following from this initial emphasis on defining terms, I have
emphasized the importance of research design choices and how these have an
impact not only on how something is to be studied but also on the kinds of
findings that are possible. Finally, I have considered how case study work can
be defended from criticisms that it is of very little use outside of the micro
settings of its own production. I have discussed a variety of ways in which
case study has been and can be defended, related these to one of the case
studies that emerged from my own study, and have argued for an additional
defence of case study research which I have termed ‘pragmatic insight’.
NOTES
1. This is by no means an exhaustive review of the literature. The books discussed
here are the ones I was directed to at the time and the books that I commonly find
shelved under ‘Research Methods’ in bookshops and libraries. They cover a range of
publication dates and a number are second editions, which suggests something of
their popularity and use.
2. In introducing the study a great deal of care was taken with the terms that were
used to describe pupils. Using the term ‘Bangladeshi’ covers up the fact that the
children themselves may not name themselves in this way and may refer to them-
selves as British or Muslim. This was explored in the study. For the purposes of
clarity in writing up the research, but with the above proviso in mind, I used the term
‘Bangladeshi’ to describe those pupils at the centre of the study who were born in
England but whose parents had been born in Bangladesh and had moved to England
when they were teenagers or adults (see Walters, 2003).
3. The pupil’s name has been changed.
ACKNOWLEDGMENT
The research study referred to above was funded by a much-appreciated

studentship from the ESRC between 1999 and 2003.
REFERENCES
Adelman, C., Kemmis, S., & Jenkins, D. (1980). Rethinking case study: Notes from the second
Cambridge conference. In: H. Simons (Ed.), Towards a science of the singular: Essays
about case study in educational research and evaluation (pp. 45–61). UEA: Centre for
Applied Research in Education.
Bassey, M. (1999). Case Study Research in Educational Settings. Buckingham: Open University
Press.
Birbili, M. (2001). Teaching educational research methods. Available at http://www.esca-
late.ac.uk/exchange/EdResMethods/.
Brewer, J. D. (2000). Ethnography. Buckingham: Open University Press.
Burns, R. B. (2000). Introduction to research methods (4th ed.). London: Sage.
Connolly, P. (1998a). ‘Dancing to the wrong tune’: Ethnography, generalization and research
on racism in schools. In: P. Connolly & B. Troyna (Eds), Researching racism in edu-
cation: Politics, theory and practice. Buckingham: Open University Press.
Connolly, P. (1998b). Racism, gender identities and young children: Social relations in a multi-
ethnic, Inner City Primary School. London: Routledge.
Cresswell, J. W. (1998). Qualitative inquiry and research design: Choosing among five traditions.
Thousand Oaks, CA: Sage.
Donmoyer, R. (2000). Generalizability and the single-case study. In: R. Gomm, M. Hammers-
ley & P. Foster (Eds), Case study method: Key issues, key texts. London: Sage.
Demack, S., Drew, D., & Grimsley, M. (2000). Minding the gap: Ethnic, gender and social class
differences in attainment at sixteen, 1988–95. Race, Ethnicity and Education, 3, 117–143.
Emerson, R., Fretz, R., & Shaw, L. (1995). Writing ethnographic fieldnotes. Chicago: University
of Chicago Press.
Gillborn, D. (1990). ‘Race’, ethnicity and education: Teaching and learning in multi-ethnic
schools. London: Unwin Hyman.
Gomm, R., Hammersley, M., & Foster, P. (Eds). (2000). Case study and generalization. In:
Case study method: Key issues, key texts. London: Sage.
Hammersley, M. (1992). What’s wrong with ethnography? Methodological explorations. London:
Routledge.
Hammersley, M. (1998). Reading ethnographic research (2nd ed.). Harlow: Longman.
Keith, M. (1993). Race, riots and policing: Lore and disorder in a multi-racist society. Occasion
London: UCL Press.
Kenny, W. R., & Grotelueschen, A. D. (1980). Making the case for case study. Occasion Paper,
Office for the Study of Continuing Professional Education. College of Education, Uni-
versity of Illinois Urbana-Champaign.
Kysel, E. (1988). Ethnic background and examination results. Educational Research, 30, 83–89.
Lincoln, Y. S., & Guba, E. G. (1985). But is it rigorous. Trustworthiness and authenticity in
naturalistic evaluation. In: D. D. Williams (Ed.) Naturalistic Evaluation. San Francisco:
Jossey-Bass.
Lincoln, Y. S., & Guba, E. G. (2000). The only generalization is: There is no generalization. In:
R. Gomm, M. Hammersley & P. Foster (Eds), Case study method: Key issues, key texts
(pp. 27–44). London: Sage.
MacIntyre, D., Bhatti, G., & Fuller, M. (1993). Educational experiences of ethnic minority
students in the City of Oxford. Oxford: Oxford University Department of Educational
Studies.
Merriam, S. (1988). Case study research in education: A qualitative approach. San Francisco:
Jossey-Bass.
Nuttall, D., & Varlaam, A. (1990). Differences in examination performance. London: Inner
London Education Authority Research and Statistics Branch.
Pathak, S. (2000). Research topic paper: Race research for the future, ethnicity in education,
training and the labour market. London: DEE.
108 SUE WALTERS
Platt, J. (1992). Cases of casesyof cases. In C. C. Ragin & H. S. Becker (Eds.) What is a case?
New York: Cambridge University Press.
Punch, K. (1998). Introduction to social research: Quantitative and qualitative approaches.
London: Sage.
Silverman, D. (2005). Doing qualitative research: A practical handbook (2nd ed.). London: Sage.
Sharrock, W., & Anderson, B. (1982). Talking and teaching: Reflective comments on in-class-
room activities. In: G. C. F. Payne & E. C. Cuff (Eds), Doing teaching: The practical
management of classrooms. Worcester: Billing and Son.
Spradley, J. P. (1980). Participant observation. New York: Holt Rinehart & Winston.
Stake, R. E. (1995). The art of case study research. London: Sage.
Tedlock, B. (2000). Ethnography and ethnographic representation. In: N. Denzin & Y. Lincoln
(Eds), Handbook of qualitative research. London: Sage.
Travers, M. (2001). Qualitative research through case study. London: Sage.
Troyna, B. (1991). Underachievers or underrated? The experience of pupils of South Asian
origin in secondary schools. British Educational Research Journal, 17, 361–376.
Walford, G., & Massey, A. (1998). Studies in educational ethnography: Children learning in
context. London: JAI Press.
Walters, S. (2003). Bangladeshi pupils: Experiences, identity and achievement. Unpublished
DPhil thesis, University of Oxford.
Wright, C. (1992). Race relations in primary school. London: David Fulton.
Yin, R. K. (2003). Case study research: Design and methods (3rd ed.). London: Sage.
Yon, D. A. (2000). Elusive culture: Schooling, race and identity in global times. Albany, NY:
State University of New York Press.
MEASUREMENT AS A RIGOROUS
SCIENCE: HOW ETHNOGRAPHIC
RESEARCH METHODS CAN
CONTRIBUTE TO THE
GENERATION AND
MODIFICATION OF INDICATORS
Eric Tucker
INTRODUCTION
This article begins with a brief reading of the state of the practice of em-
pirical social science research on measurement before proceeding to the
discussion of an exemplary instance of this researcher’s ethnographic effort
to improve indicators of social capital formation. Given the central role
measurement plays in social science research, it is appropriate, that a volume
on methodological innovations in ethnography would contain a chapter
about the relationship of ethnography to measure development. However, it
is worth acknowledging that the line of argumentation advanced in this
chapter is unconventional. The central tenant of this chapter – that
ethnography has much to offer to the field of measurement and that
ethnographers ought to take the contribution that they have the potential to

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12007-0
109
110 ERIC TUCKER
make to the field of measurement seriously – at present might be thought to

have little agreement either among those researchers whose primary focus is
measurement or among ethnographers. This chapter contends that the fea-
tures and strengths of ethnography specifically, and qualitative research
more generally, makes it uniquely suited to contribute to the development
of new indicators and the improvement of existing indicators. This chapter
modestly hopes to encourage discussion of this contention and illustrate
how this author sees his own ethnographic research into indicators of
social capital formation as an attempt to address a pressing metho-
dological dilemma within the field, more general of social scientific measure
development.
Measurement is at the centre of much educational research, from tests to
observational scales, to questionnaires. A concern to develop and improve
measurement exists in each social scientific discipline. Indeed, measure
development is a continual focus of academic debate. The first part of this
chapter locates a number of relevant streams within the field of measure-
ment. The second half attempts to situate my own research enquiry into
indicators of social capital formation within these streams.
In the case of social capital, the need to develop new, valid, and rigorous
indicators emerged both when the concept was born and on an ongoing
basis as the demands placed on the concept’s performance changed. Indeed,
these factors influence demand for quality indicators more generally. Un-
fortunately, the method for such systematic, purposeful generation of in-
dicators and measures is insufficiently understood.
THE CHALLENGES ARISING FROM THE CONTEXTS

OF MEASUREMENT
This article rests on the assumption that, in recent years, the variety and
intensity of the demands social scientists and policymakers place on indi-
cators have grown dramatically. The main argument that supports this
claim is that measurement instruments increasingly must respond to the
newly global reach of social science and the diversity of cultures, contexts,
users, and participants this brings. Globalisation demands a social science
that is fluent in comparative and international measurement, capable of
exploring the challenging nature of the problems and issues faced by re-
searchers across cultures and nations, and eager to test the robustness of
indicators in different contexts within a broad comparative framework.
Measurement as a Rigorous Science 111
Researchers increasingly ask themselves: What steps can be taken to incor-

porate cultural perspective and difference into the process of measure
development, validation, and use? How can researchers develop rigorous
measures with minimal error when used in diverse cultural and language
communities? When working across space and time, what must researchers
do to enhance measure quality? An increased focus on measurement is nec-
essary, given the explosion of research needs, agendas, and sites.
Measurement tools should be applicable for a range of times, spaces,
contexts, and users. The economist Angus Maddison (1995a, 1995b, 2001,
2003) addresses the question of how to measure across vast expanses of time
and space – a topic with which the social sciences must increasingly grapple.
Meanwhile, researchers such as Bill Hillier (1996) have taken on the the-
orisation and operationalisation of particular notions of the social use of
urban space – causing the development of measures that capture aspects of
peoples’ perceptions and uses of space. The context within which a
measurement instrument is used also influences what gets measured. Some-
times the context is inhibiting or reactive. For instance, Strange, Forest, and
Oakley (2003) look at the use of surveys with youth in schools, teasing out
the influence of the social context on the self-completion of questionnaires.
The context of collection may also differ from the context of development
in ways that skew the validity and appropriateness of the instrument
(Allalouf & Ben-Shakhar, 1998).
Context changes when participants have distinct characteristics and traits
(for instance, see Maller & French, 2004). A range of papers have explored
the importance of measurement capable of navigating gender-related
variation (e.g. Ben-Shakhar & Sinai 1991; Bolger & Kellaghan, 1990; Lowe
& Reynolds, 2005; and Fletcher & Hattie, 2005). Gender differentials with
regard to item-response tendency and concept dimensionality remind us that
who the participants’ subjected to measurement matters. The significance of
gender reminds one about the issue of measuring attributes, attitudes, and
behaviour within sub-populations more generally. For instance, sometimes
items have differential functioning for sub-populations (Sehmitt & Dorans,
1990; Stevens & Tallent-Runnels, 2004; Davis, Capobianco, & Kraus, 2004).
In addition, translations of instruments to new language contexts and
assessments of the validity of measurement across languages deserve men-
tion. The development and validation of translated scales is an issue ad-
dressed throughout the literature (for instance, Collazo, 2005; Boles, Yoo,
Ebacher, Lee, & Ashton, 2004; Hong & Wong, 2005; Moneta & Yip, 2004;
Grégoire, 2004; Sireci, 1997). Differences of gender, race, ethnicity,
language, and culture of the research participants complicate measurement.
112 ERIC TUCKER
The variety of contexts of measurement has increased the expectations

placed on measurement instruments. These raising expectations have
emerged in tandem with a concern among some researchers that in certain
instances there seems to exist a gap between established measurement in-
struments and the social phenomena they intend to measure.
THE DEMAND FOR HIGH QUALITY INDICATORS:

UNDERSTANDING AND MINIMISING
MEASUREMENT ERROR BY SECURING VALIDITY
AND RELIABILITY
I have established that the demands placed on measures have been increasing
in recent years. Before discussing the widening gap between the social world
and the instruments we use to capture them, I will discuss social scientists’
demand for high-quality indicators. The consensus view of social scientists
today might be thought to define an indicator as high quality when its va-
lidity and reliability have been adequately established. The prominence of
psychometric approaches to securing validity and reliability can be seen
through even a casual perusal of the measurement literature. A first group of
literature is methodological in bent, and uses specific issues to raise general
concerns about measurement. For instance, Clarke (2002) writes about
falsely precise scores. McGuckin and Peck (1993) demonstrate how construct
integrity is influenced by social change. Hogan and Agnello (2004) take
empirical stock of the degree to which established psychological instruments
are validated. Howrey (1996) discusses how measurement error plagues cer-
tain types of data. A second group seeks to identify and reduce measurement
error for particular instruments. For instance, Fortunato (2004), Clapham
(2004), DeCarlo (2005), and Fleming, Jordan, and Lang (1996) Fleming et al.
(1996), are a few articles out of the hundreds that use psychometrics to
minimise the measurement error of instruments. Researchers discuss this
issue generally and offer methodological, definitional, and practical solutions
to measurement error.
Viswanathan (2005), advancing a cornerstone contribution to the field,
proposes that measurement error should be at the centre of discussions
of measure design. He argues that the ‘measure development process con-
sists of a series of steps to develop reliable and valid measures. Rather
than proceed directly from an abstract construct to its concrete measure-
ment, the distance between the conceptual and the operational has to be
traversed carefully and iteratively. Measure development ideally should

combine empirical assessment with conceptual examination’ (p. 383). This
means that ‘measurement error should be understood in each stage of
developing measures and their use.’ He continues, ‘errors in developing
and using measures are closely interrelated’ (Viswanathan, 2005, p. 386).
This insight is crucial to methodological writing on measurement.
Researchers and methodologists alike, Viswanathan suggests, must reflect
on the specific threats to validity and reliability that they encounter, and
offer a nuts and bolts discussion of how to develop indicators that are
correct for measurement.
The gap between established measurement instruments and the social
world raises acute concerns, precisely because social scientists and policy-
makers demand high-quality indicators and it is unclear whether existing
development procedures have allowed the research community to keep pace
with demand. Next, I discuss the potential objectives of generating valid
indicators to the broader challenges faced by the social and applied sciences.
Rather than offering partisan advocacy for these positions, I merely offer a
synthesis of academic and policy perspectives on the need and utility of
high-quality indicators.
Although I do not defend each of these positions, it is worth noting that a
variety of reasons are offered to support the contention that quality
measures matter. First, indicators facilitate learning and information ex-
change, allowing researchers to collect, organise, and present information on
a jurisdiction’s assets and vulnerabilities. Second, indicators raise the profile
and comprehensibility of certain social phenomena (Mills, 1980). For in-
stance, I would argue that the decision to measure domestic violence as a
‘crime’ put the issue on the radar screen and facilitated social investment.
Third, indicators suggest social trends and allow researchers to monitor the
direction a jurisdiction is moving (Kiuranov, 1982). Fourth, indicators help
identify gaps in understanding about social issues. For instance, Sen (1999,
2002) demonstrated that Gross Domestic Product was not a sufficient in-
dicator for social well-being and economic development – and thus increased
the understanding of the role of health care, education, and compensation in
measuring progress towards development. Fifth, indicators improve policy
decision making by facilitating planning and enhancing evidence-based
consensus (Archer, Kelly, & Bisch, 1984). They are useful for monitoring
and reporting, but also for social engineering, program evaluation, social
accounting, and goal setting (Smith, 1981). Sixth, quality indicators allow
analysis and facilitate the development of models. Finally, quality indicators
enhance intervention design, evaluation, and performance.
114 ERIC TUCKER
THE GAP BETWEEN ESTABLISHED MEASUREMENT

INSTRUMENTS AND THE SOCIAL WORLD
A number of small-scale studies have begun to suggest that in certain areas,
substantial gaps exist between social scientists and decision makers’ desire to
understand the world and the capacity of existing indicators to approximate
the quantity of the phenomena consistently and accurately. If these studies
are taken seriously, then it appears as though many arenas of the social
sciences have a real need for better measurement tools as many extant in-
dicators are of rather low quality. This gap between our desire to under-
stand the social world and the quality of established measures provides
perhaps the primary rationale for the use of ethnographic research methods.
Several factors inform this conclusion. First, many researchers have an
expressed preference towards using existing measures. Given that a signifi-
cant number of researchers narrowly focus on what is currently measured in
administrative data sets, I suggest that ethnographic research into how a
given concept should be operationalised will beneficially broaden the con-
versation. Second, the use of indicators that do not fit, work, or produce
quality data raises the risk of misinterpretation, misinformation, and un-
intended errors. Ethnographic enquiry into how fitting and workable
existing indicators are can result in steps to modify and improve those
instruments. Third, I contend that the development of robust measures will
result from the exploration of a broad number of specifically selected case
studies to generate indicators that better capture particular concepts.
Measure development is more innovative and open when it identifies,
acknowledges, and grapples with the gap between the social world and
established indicators.
The notion of a gap between established instruments and the social world
has been posited within disciplines across the social sciences. The field of
measurement regularly identifies areas where the quality of established
measures is low. Some examples may illustrate the nature of gaps between
indicators and the social world. For instance, Whiting and Harper (2003, p.
14) use qualitative research to argue that existing indicators of social capital
are not relevant for young people – that voting and sitting on the executive
committee of a charity are bad measures of youth engagement in political
and social life. Shriver (1995) discusses how the measurement of current cost
data in financial accounting has a systematic overstatement bias. Powers,
Fowles, Farnum, and Ramsey (1994) study how typed replicas of hand-
written essays receive lower average scores, demonstrating the challenge of
using written essays as an indicator of writing skill. Frisbie and Cantor
(1995) look at the validity of alternative methods for assessing spelling

ability finding differential outcomes for different instruments. Potenza and
Stocking (1997) explore the challenge of dealing with flawed questions on
tests, which intend to capture a particular skill but fail. Galvez and McLarty
(1996) work around the limitations of demographic indicators that omit
temporary residents. Whether we talk about social capital, current costs,
handwriting, spelling, flawed questions, or temporary residents – the gap
between the reality of a social phenomena and what the indicator captures is
a recurring challenge in measurement.
RESEARCH FOCUSED ON THE GENERATION AND

MODIFICATION OF INDICATORS
I have demonstrated both the increasing demands placed on indicators and

existence of a gap between certain established measurement instruments and
the social phenomena they intend to measure. I will now make the argument
that the field of measurement tends to over-emphasise verification and cor-
respondingly de-emphasise empirical research that aims towards generation
and modification. Although I suggest the real need for measurement liter-
ature that speaks to the process of generating and modifying new indicators,
I also argue that this gap in the literature does not reflect contem-
porary innovations in measure design. Researchers involved in cutting-edge
research in fields ranging from psychology and economics to environ-
mental policy and city planning are developing novel approaches to meas-
urement within their particular domain of social science. Published
measurement methodologies thus lag behind measurement practice, and
there is a need for research highlighting the researchers’ perspectives on
important methodological considerations, theoretical insights, and perva-
sive challenges in measurement design. The goal is a body of methodological
literature that takes the purposeful systematic generation and modification
of indicators seriously. When researchers get it right, they do not begin with
an indicator and then establish its validity; they conduct empirical research
to generate and modify indicators that are faithful to and illuminate the
concepts and phenomena under study.
Developing new measures for constructs should be an ongoing priority
for the social sciences. Examples of researchers who aim to generate and
modify new indicators abound. Putnam (1993, 2000), Florida (2002, 2005),
Laumann (1994), and Sen (1973, 1987, 1999, 2002) represent figures at the
116 ERIC TUCKER
top of their field who take this challenge seriously. But across the social
sciences, operationalising concepts stand as a priority. For example, the
design of the Human Development Index of well-being (which grew from
Sen’s writings) has spurred a robust research community focused on mod-
ifying the computation process (see Gormely, 1995; Acharya & Wall, 1994).
Noble et al. (2004) describes the development and refinement of a scale to
assess attitudes towards working single parents. Le, Casillas, Robbins, and
Langley (2005) construct a student readiness inventory that measures the
psychosocial and academic-related skills that predict university’s academic
performance and retention. These examples represent larger trends regard-
ing the generation of new indicators and scales across the social sciences.
CONDUCTING RESEARCH TO DESIGN AND

IMPROVE INDICATORS
The field of measurement places high priority on the statistical validation of

indicators. However, indicators themselves are often based on logical de-
duction, assertion, and speculation, rather than upon rigorous scientific re-
search. Empirical research that aims to improve the design and enhance the
quality of an indicator is a small fraction of published work. Journals of
measurement print few articles reporting on research geared towards em-
pirical generation and modification of indicators. Instead, scales are con-
ceived of as ‘already always’ existent and in need of subjection to statistical
verification. Social scientists, it seems, spend little time tackling the question
of how to operationalise constructs through concept definition, domain de-
lineation, item generation, scale development, and modification. They focus
on running statistical tests to secure validity and reliability.
Examples of articles that focus on how constructs are conceptualised and
then developed represent one strand of current research to design and im-
prove indicators. Stevenson and Evans (1994) write about the conceptuali-
sation and measurement of Cognitive Holding Power. In the article, theory
and research inform the conceptualisation, two dimensions of holding
power are distinguished, an instrument is developed, and then the reliability,
validity, and factor structures are examined. Giancarlo, Blohm, and Urdan
(2004) report on the development of a new instrument to measure mental
motivation. Dowson and McInerney (2004) write about the development
and validation of the Goal Oriented and Learning Strategies survey, which
measures student motivational goal orientation and metacognitive
strategies. Asher, Defina, and Robert (1995) examine the shortcomings of

measures of income inequality that relate to age–income profiles and pro-
pose improvements. Wolfe, Ray, and Harris (2004) compare the quality of
three scales of measures of teacher perception of instructional environments.
Articles describing the design and improvement of indicators abound.
A distinct body of qualitative research aims to improve the design of items
and scales. Pearce (2002) discusses how the salience of findings can be en-
hanced by combining survey research with ethnographic methods. She uses
ethnographic methods iteratively to improve measurement of the influence
of religion on childbearing preference in Nepal. Farrall (2004) calls into
question a range of studies showing a majority of citizens in European and
North American countries report experiencing ‘fear’ or feeling ‘angry’ about
the possibility of criminal victimisation ‘all the time.’ He uses ethnographic
and qualitative data to suggest that fear and anger are less common and that
the inappropriate use of survey indicators has systematically exaggerated the
emotional response. These articles represent the conduct of qualitative
research to improve measurement.
MY RESEARCH AS AN EXEMPLARY INSTANCE OF

ETHNOGRAPHIC RESEARCH THAT AIMS TO
DEVELOP AND IMPROVE INDICATORS
In this second half of the paper, I make the case that the features and
strengths of ethnography specifically, and qualitative research more gener-
ally, make it uniquely suited to contribute to the development of new in-
dicators and the improvement of existing indicators. I accomplish this by
turning to one specific example of this author’s ethnographic research project
that aims to improve indicators of social capital formation, in the hopes of
providing an exemplary instance of how ethnographic methodologies might
contribute to the generation and modification of new, quality measurement
instruments. Thus, rather than making a global appeal, I embed my
argument that ethnographic methodologies have unique features and strengths
that make it well suited to contribute to the development of new indicators
within a discussion of the particular methodology I employed in a particular
research project. This section thus outlines the research methodological
perspective and approach employed in this education research project.
The data generation, analysis, and discussion in this account were derived
from two compressed ethnographic case studies exploiting two substantially
118 ERIC TUCKER
different programmes analytically linked by their ability to illuminate

aspects of the types of social practices that contribute to the formation of
social capital. The ethnographic research strategy and design, the theoretical
approach to sampling and case selection, the two-phased data collection,
and the three-tiered data analysis allowed me to move from the generation
of comprehensive, synthetic structural and textural descriptions of the
programme to the subsequent solicitation of practitioner and expert review
and culminating in the generation and modification of categories. The rest
of this chapter describes and justifies the methodological conduct of my
investigation.
In the empirical domain, my research project asked: What are some of the
major categories, generated from the analysis of each set of case study data,
which describe the social practices that potentially contribute to the formation
or deformation of social capital? Such categories, it was hoped, would illu-
minate potentially operable dimensions of the concept and be suggestive of
features that merit further consideration. Methodologically, I departed from
the contemporary emphasis on verification as the primary mechanism to
secure the validity and reliability of measures. I contended that minimising
measurement error might best be achieved by a greater emphasis on the
generation and modification phases of measure development as comple-
mentary to verification efforts. In other words, I called for a greater em-
phasis on empirical social research that focuses upon the generation and
modification, in addition to the verification, of indicators. I thus used the
empirical work to justify the case that ethnographic research methods, and
other flexible and emergent approaches, might be used to generate more
thorough understandings of the particular social phenomena, which might
in turn be used to develop higher quality indicators.
MY RESEARCH STRATEGY AND DESIGN
Broadly speaking, the research strategy I embraced with regard to the

selection of questions, settings, cases, sampling procedures, and data-gener-
ation methods was an ethnographic case study. My data generation, analysis,
and discussion thus emerged from two compressed ethnographic cases
(Walford, 2001) exploiting two substantially different programmes analytically
linked by their potential to illuminate aspects of the types of social practices
that contribute to the formation of social capital. This ethnographic approach
contrasts with experimental or survey research in that I collected in-depth,
unstructured data from the existing social settings to describe some of the
unique, sensitising aspects of each case that lay bare potentially essential fea-
tures of the relevant social phenomena.
This project employed a case study approach, in that there are two cases
examined and a significant amount of detailed information from a number
of sources has been collected for a range of dimensions for each case
(Gomm, Hammersley, & Foster, 2000). In this sense, the case study
approach was aligned with the research questions in that it facilitates
description, understanding, and explanation in a thorough, open-ended
manner. Although I do not see the case study approach more generally as
embodying a particular research paradigm (Gomm et al., 2000, p. 5), the
belief that in-depth qualitative enquiry into a case can be used to generate
more thorough understandings of the particular social phenomena, which
might in turn be used to develop higher quality indicators, seems supportable.
What was the relationship between my ‘research strategy’ and the actual
research I conducted? The research strategy was, I suggest, the ‘scaffold’
that organised and supported the investigation, and kept me from straying
too far off the path of the question and the compressed ethnographic
methodology driving the project. Here, I echo McKenzie’s (2000) use of the
term ‘scaffolding’ within the context of education. He defines scaffolding,
saying, it conventionally refers to structures put in place to aid construction.
‘Scaffolding,’ as I use it here, might thus be said to provide a clear structure
and precisely stated expectations, without a methodological straight jacket
that destroys initiative and resourcefulness. By analogy, scaffolding (i.e. the
research strategy) is not the research project itself, but simply the supporting
structure. McKenzie argues that ‘scaffolding,’ ‘provides clear directions,’
‘clarifies purpose,’ keeps one on task, ‘points towards worthy sources’
(in this case of methodological texts), reduces disappointment by maxim-
ising learning and efficiency, and ‘creates momentum.’ These features of
‘scaffolding’ mirror role that the compressed ethnographic case study strat-
egy played in this research project.
THE NATURE OF ETHNOGRAPHIC ENQUIRY

To win the argument that ethnographic research methods have the potential
to contribute to the development of measurement instruments, I will offer a
definition of ethnography. In doing so, I hope to outline the foundation for
the argument that runs through the second half of the chapter: ethnographic
120 ERIC TUCKER
methods can be used to generate more thorough understandings of the

particular social phenomena, which might in turn be used to develop higher
quality indicators.
Walford writes that, ‘(t)hrough prolonged involved with those who are
being studied, the ethnographic researcher is able gradually to enter their
world and gain an understanding of their lives’ (Walford, 2002, p. VII).
I reject the notion that ethnography is synonymous with qualitative research
more broadly. Still, ethnographic researchers have drawn upon a broad
range of theoretical lenses, data-collection tactics, and data-analysis meth-
odologies. As Hammersley and Atkinson point out, some authors offer
a liberal interpretation of the definition of ethnography. My research fits
within such a broad interpretation. For Hammersley and Atkinson, this
refers to participating ‘in people’s daily lives for an extend period of time,
watching what happens, listening to what is said, asking questions – in fact,
collecting whatever data are available to through light on the issues that
are a focus of the research’ (Hammersley & Atkinson, 1995, p. 1). My
research was designed and conducted in a manner that prioritised
participant subjective experience – how they perceived and gave meaning
to their participation.
What constitutes ethnographic research? Walford (2001) offers a com-
pelling broad outline, suggesting ethnographers to ask: What is going on
here? How does this work? How do people do this? How are things done
around here? He suggests elements of ethnographic research worth refer-
ence. First, ethnography studies culture, stresses that humans move within
social worlds, and ‘balances attention to the sometimes minute everyday
detail of individual lives with wider social structures.’ A culture, for
Walford, is ‘made up of certain values, practices, relationships and
identifications’ (p. 7). Second, ethnography is constituted of multiple meth-
ods and diverse forms of data. He writes that to ‘gain a multidimensional
appreciation of the setting, the ethnographer must be prepared to consider
many different types of data’ (p. 8). Third, ethnography entails engagement
and observation in the situation where the activity actually occurs. This
requires, ‘human connection with participants, and an investment of time
(p. 8). Fourth, in ethnography the researcher is the instrument and the
primary source of data. The researcher’s subjectivity is seen as an inevitable
aspect of research. When rigorous, the researcher reflects upon why par-
ticular decisions were made, ‘why certain questions were asked or not asked,
why data were generated a particular way and so on.’ Walford cites Dey
(1993) who argues that ‘The danger lies not in having assumptions but in
not being aware of them’ (p. 8). Fifth, in ethnography, participant accounts
have a high status. Interactions with informants are used to generate and
create the analytical frameworks used to understand and portray the
phenomena under study. Sixth, ethnography proposes a cycle of hypothesis
and theory building. Walford describes this as the ‘ethnographer’s constant
commitment to modify hypotheses and theories in light of further data.’ This
process of formulation, testing, reformulation, retesting, means that ‘what
needs to be looked at and reported on may change and explanations of what
is going on may be supplanted by ones which seem to fit better’ (p. 8). In this
sense, ethnography is emergent and flexible. Seventh, the ethnographer aims
‘to discover how people in the study area classify or label each other, how
they find meaning in activities they care about in life, and how they engage in
processes in which they individually and collectively define (antecedents
and consequences of) their situations’ (p. 8). It enquires into how parti-
cipants perceive and give meaning to the situation. Walford does not offer
an exhaustive description, but these seven elements distinguish ethnography
from qualitative work more generally.
THE LENS OF EMPIRICAL PHENOMENOLOGICAL

RESEARCH
Epistemic orientations are connected to methodologies and types of

evidence that stem naturally from the type of knowing postulated. One
lens I brought to this investigation is that of empirical phenomenological
data analysis. It is both a particular approach to ethnography and a research
tradition that did much to inspire the emergence of ethnographic method-
ologies more generally. Phenomenology is a direction of inquiry, what I call
an epistemic orientation (Husserl, 1960). Through this approach, empirical
phenomenological analysis attempts to strike at some of the essential sim-
ilarities between participants’ descriptions of experiences without diminish-
ing the significance of variance. It aims to give credence to the direct
experience of participants, experience that is necessarily relative to their
position in the midst of the world. In Merleau-Ponty’s (1962, p. xix) words,
‘to say that there exists rationality is to say that perspectives blend,
perceptions confirm each other, a meaning emerges.’
Inspired by some of the pioneering insights of Husserl, empirical
phenomenological approaches hope to discover insights ‘by reference to
the things and facts themselves, as these are given in actual experience and
intuition’ (Moustakas, 1994, p. 47). Moustakas (1994, p. 47) assures us that
122 ERIC TUCKER
empirical phenomenology attempts to provide ‘a logical, systematic, and

coherent resource for carrying out the analysis needed to arrive at essential
descriptions of experience.’ Farber describes phenomenology, with regard to
social science research, as seeking to ‘‘identify presuppositions and ‘put
them out of play;’’’ concerning itself not with matters of fact but with
determining meanings; and dealing with and offering direct insight into real
essences and ‘possible’ essences (cited in Moustakas, 1994, p. 49). The social
scientist inspired by empirical phenomenology might thus set out to expli-
cate the constituent components and meanings of phenomena, discerning
features of consciousness, and types of experience.
Empirical phenomenology as an orientation and a lens, is geared towards
the development of ideal types. In this case, that would mean categories that
refer to social practices that may tend to contribute to the formation or
deformation of social capital. Types are ideal in the sense of a non-real model
that is ‘limited but useful for the purpose of inquiry.’ Types approximate
‘reality sufficiently for the purpose of interpreting and understanding reality,
but never as the sum-total of reality’ (Gordon, 1995, p. 55). Take the example
of a trigonometry class, where triangles are used as visual aids to illustrate
the range of potential trilaterals. The construction paper figures are ‘tokens
for the types that constitute triangles.’ Cut-outs, as things (tokens) can be
counted. However, it is ‘incoherent to ask how many types of right triangle
are there.’ Types are unique. They are eidetic essences. The search for types
faces the problem of ‘identifying social phenomena without reducing them
solely to their singular identifications’ (Gordon, 1995, p. 56). Typification
(the search for essence) constitutes a fundamental feature of human science.
Phenomenologically inspired enquiry, I suggest, is well suited to the task
of generating and modifying indicators of social phenomena for several
reasons:
1. Empirical phenomenological analysis attempts to focus on units of

meaning. Developing valid indicators of social phenomena often requires
interpreting the meaning of people’s experiences. The same action may
mean and ‘indicate different things for different people.’ Empirical phen-
omenological analysis has the potential to help access phenomena
through the eyes of respondents, learning from their concerns and ways
of understanding. Empirical phenomenological analysis asks data (indeed
people) directly ‘why they behave in a particular way or why they express
a particular attitude’ (De Vaus, 1995, pp. 57–58). Although informants
are not always aware of the reason, answers provide valuable insights
regarding the structure of the phenomenon.
2. Empirical phenomenological analysis seeks to return to the phenomena

themselves. Too often, the development of indicators simply refers to
existing measurement indices. Empirical phenomenological analysis
returns to the things themselves and focuses on the social world as it is
given. It brackets everyday routines, assumptions, and biases. This is
crucial for the task of generation.
3. Empirical phenomenological analysis seeks to identify essential fea-
tures of phenomena. The research undertakes reflective analysis and
interpretation of descriptions with the aim of providing a comprehensive
description of the general meanings derived – in other words, the essences
or structures of the experience. The phenomenological approach employed
here relates back to the effort to identify the types, to provide a portrayal
of the phenomenon that is vital, rich, and layered (Moustakas, 1994).
By describing the potential benefits of empirical phenomenology as a

theoretical lens, I am not assuming the burden of defending the entire his-
tory and trajectory of phenomenology as a tradition. By sharing the lenses
and orientations that influenced how I approached this particular ethno-
graphic research project, the hope is to support the relevance of that small
corner of the tradition that directly applies to my research project as one
exemplary instance of how ethnographic methodologies might contribute to
the generation and modification of new, quality measurement instruments.
MY APPROACH TO SAMPLING AND CASE

SELECTION
My approach to sampling and case selection was aligned with my effort to
generate evidence and understanding that would facilitate the development
of new measurement instruments. In my research project, I employed the-
oretical sampling to inform case selection and data collection oriented as it is
towards category development. As such, selection of cases and identification
of discrete data sources was purposive and ongoing (Glaser & Strauss, 1967,
p. 45). Glaser and Strauss (p. 49) argue that with theoretical sampling, the
‘researcher chooses any groups that will help generate, to the fullest extent,
as many properties of the categories as possible, and that will help relate
categories to each other and to their properties.’ Initial enquiry inspired
subsequent selection of research sites, cases, and informants geared towards
the generation of data that facilitates comparisons that build categories.
124 ERIC TUCKER
In selecting cases, sites, and data sources, I did not aim to capture a
representative slice of the phenomena. Instead, samples were selected in an
effort to arrive at a deeper understanding of categories and dimensions of
constructs. I tapped into veins of data in an iterative, progressive manner.
This approach to data collection is deliberatively emergent and avoids the
embrace of a pre-conceived framework. I used theoretical sampling to
facilitate category development, modification, refinement, and solidification.
In other words, the clarity, parsimoniousness, workability, fit, and level of
abstraction and relevance of categories developed over time – and the mod-
ification and refinement of these categories guided the selection of additional
sites and data sources.
During initial data collection for each case significant categories began to
emerge. Subsequent data were collected to develop these categories and
better understand their properties and function. On the basis of the data
analysis of that case, the second, complementary case was selected to test
and extend the proposed categories (Strauss & Corbin, 1990, p. 192). Yin
specifies the purposes of case selection within theoretical sampling, identi-
fying the appropriateness of choosing a case to: (a) fill theoretical categories,
(b) test the emerging categories, (c) select a case that is the opposite to raise
contrary perspectives. (Yin, 1994, pp. 53–54).
Glaser and Strauss (1967) suggest that theoretical selection of comparison
groups provides ‘simultaneous maximization or minimization of both the
differences and the similarities of data that bear on the categories being
studied’ (p. 55). Cases reveal common factors and relevant differences.
Maximising differences between cases, accordingly ‘increases the probability
that the researcher will collect different and varied data bearing on a cat-
egory, while yet finding strategic similarities among the groups’ (p. 56). I
maximised differences between cases to cover a range of types, variations,
conditions, relationships, processes, mechanisms, etc. By studying new or-
ganisations, regions, cities, or nations, I saw differences and similarities
which lent insight into the categories originally developed (p. 57). Categories
were thus crafted by handling a multiplicity of groups and situations.
My research’s sampling approach entailed collecting a variety of ‘slices of
data,’ echoing the approach advocated by Glaser and Strauss (1967, p. 65)
who state:
In theoretical sampling, no one kind of data on a category nor technique for data
collection is necessarily appropriate. Different kinds of data give the analyst different
views or vantage points from which to understand a category and to develop its prop-
erties; these different views we have called slices of data. While the [researcher] may use
one technique of data collection primarily, theoretical sampling for saturation of a
category allows a multifaceted investigation, in which there are no limits to the tech-
niques of data collection, the way they are used, or the types of data acquired.
Rigorous data collection occurred simultaneously to data analysis. Collec-

tion and analysis were iterative and mutually informing.
DATA COLLECTION PHASE ONE: INNOVATIVE AND

ESSENTIAL PRACTICE EVALUATION
In each case, my phase of data collection was to conduct an innovative and

essential practice evaluation (IEPE). IEPE is a phrase I coined to capture
what amounts to a specific approach to process evaluation. As such the term
IEPE does not represent a methodological innovation, but denotes my
approach to dealing with questions such as what activities are essential,
when and where they occur, and who delivers them. It examines whether or
not the programme is delivered as intended to the target population.
According to Rossi, Lipsey, and Freeman (2004), ‘program process mon-
itoring is the systematic and continual documentation of program perform-
ance that assesses whether the program is operating as intended or
according to some appropriate standard y’ (p. 171). It asks how the pro-
gram is intended to be delivered, and if it is delivered as intended.
I define IEPE as a systematic investigation into programmatic imple-
mentation, as an effort to identify and map the core features and approaches
of programming. IEPE is an exploratory and systematic analysis of the
characteristics of the organisation under study and an account of the factors
that variously contribute to organisational health and sickness. Esman and
Uphoff (1984) offer questions that IEPE takes on board, including the types
and level of performance, the environments within which an organisation
operates, the structure of the organisation, the type and extent of member
participation, and its overall performance. Rossi et al. (2004, 7th edition, p.
171) also inspired my approach, calling for description, identification, and
specification of the need for a programme, the programmatic theory, the
issue to be addressed, the targets of the programme, and the target pop-
ulation. IEPE grew from my training in so called ‘innovative practice’ ed-
ucation policy research (Bardach, 2000; Stokey & Zeckhauser, 1978;
Weimer & Vining, 1999; MacRae & Whittington, 1997). ‘Innovative prac-
tice’ research aims to distinguish between elements that are essential and
those are merely supportive and optional, to distinguish general vulnerabil-
ities, and allow for variation and complexity (Bardach, 2000).
126 ERIC TUCKER
In Phase 1, I conducted ongoing participant observations amounting to a

compressed ethnography. This included time in the field documenting ob-
servations, chance conversations, overheard remarks, and informal inter-
views. The review of tens of thousands of pages of primary documents
provided a complementary, unobtrusive method of gaining insight into
the values, actions, and beliefs of community participants (Marshall &
Rossman, 1995, p. 85). Robson (2002) suggests that archived documents
have the advantage of being subject to reanalysis, providing documentation
over time and from a range of contexts.
IEPE might be thought of as attempting to improve the validity of in-
dicators teased out of interview data. Walford (2001, p. 89) indicts interview
data, urging researchers to see it as only one of several ways to generate data
rather than as automatically reliable, ‘hard,’ or sufficient. As precursor to
interviews, process evaluation has advantages similar to two-phase, mixed-
method research. First, the consistencies between interviews and observa-
tional data can be triangulated. Second, process evaluation findings can
clarify and illustrate interview findings complementarily. Third, interview
questions and focus are developed, and new lines of questioning initiated.
Fourth, evaluation data provides richness and detail of understanding, ex-
panding the quality of interviews (Green, Caracelli, & Graham, 1989; Green
& Caracelli, 1997; Kidder & Fine, 1987).
DATA-COLLECTION PHASE TWO: IN-DEPTH

INTERVIEWS
The accounts offered by informants of how they perceive and give meaning
to their actions are the most important source of data. Actions are acts of
human consciousness – enacted in the midst of structures, memories, goals,
values, and oppression. I encouraged participants to consider the questions
as my equal and mutual investigator, and treated them as an expert with rich
stores of experience relevant to the research question. Many interviews were
repeated/multi-event encounters.
I conducted interviews as an informal, interactive process. Open-ended
questions and comments were used to solicit feedback from respondents. I
prepared an interview guide containing a series of questions ‘aimed at
evoking a comprehensive account of the person’s experience’ (Moustakas,
1994, p. 114). These interview questions formed the research protocol and
focused interviews on ‘the experience of a situation’ (Von Eckartsberg,
1998). Efforts were made to build relationships with informants through

extensive informal contact before commencing or even suggesting an inter-
view. Interviews often ‘began’ with social conversation over several weeks,
building trust and creating a relaxed atmosphere to establish a rapport with
the research participant, making him or her comfortable enough to answer
questions honestly and in detail. Questions were altered (or not asked at all)
when appropriate. Sometimes, interviews focused on a particular question
or set of questions to the exclusion of others.
The individuals were selected for interviews purposively, in accordance
with the following priorities (which are similar to those upon which the two
cases were selected). The selection process for interviewees was influenced by
the following priorities. Interviews
Helped ‘flesh out’ existing hypothesis by seeking individuals who can offer
a ‘full account’ of a particular viewpoint.
Sought insights that helped tease out why emerging data and themes
disagree with established understanding.
Sought interviews that helped to develop categories, map dimensions, and
explore similarities and differences, coherence and incoherence, and the
relative importance of a given category.
Sought negative instances and dis-confirmatory examples to mitigate the
risk of bias.
Sought confirmatory evidence by checking and re-checking categories and
exploring possible rival hypothesis.
Sought guidance from ‘experts’ in the field by asking critical informants
about categories and concepts under development.
In addition, for both cases, I used Internet-based text communications as
an object of analysis and used the Internet to collect data from individuals
(Bryman, 2004, p. 457). This was uniquely appropriate because within both
case studies, Internet-based communication was central to network activity.
Web-data generation involving website, webpage, blog post, electronic ar-
ticle, bulletin board, and listserv post-related sources of data as the object of
analysis. Materials were identified and archived. In addition to my broader
orientation to data analysis, with e-data I prioritised analysis through the
following lenses:
Examination of material as embodying a dimension of the networking
and connection between group participants.
Examination of content as an expression of the types of social practices
that occur in the network.
128 ERIC TUCKER
Examination of e-material as participant engagement in a conversation

relevant to the themes, categories, and research questions of this study.
THREE TIERS OF DATA ANALYSIS

The first tier of data analysis was my effort to generate comprehensive,
synthetic, structural, and textural descriptions of case activities. I conducted
data analysis with an empirical phenomenological orientation. I will not
describe the procedures in detail. With the transcripts of interviews, obser-
vation notes, primary documents, innovative and essential practice evalu-
ation related documents, blog data, and case-relevant literature, I conducted
a recursive comparative analysis: I took a sample of interviews, read them,
and developed concepts and categories; and then checked these categories
against another pile of transcripts, making necessary adjustments. I sorted
through physical stacks of interview data, echoing Dey’s (1993) proposals to
split and splice piles, dividing and combining evidence until piles and cat-
egories genuinely cohere. The specific data-analysis procedure I conducted
was an expanded and modified version of the empirical phenomenologically
informed analysis process proposed by Moustakas (1994). The result of Tier
1 was several comprehensive, synthetic IEPE Reports. As outlined above,
these reports:
Charted the dimensions of the programme, including essential features
and relevant innovations.
Sought to capture programmatic and experiential variance and complex-
ity while omitting the purely idiosyncratic or particular.
Sought to describe how participants understood and perceived what was
attempted, what happened, and how it happened.
These comprehensive, synthetic reports are an end product in and of

themselves. In some minds’ research projects, these reports would be the end
product of an ethnographic research process. However, because the aim is
not to describe case-related phenomena, but to develop categories that aim
to operationalise concepts with minimal measurement error – these reports
were only the first step.
In the second tier, I presented practitioners in the field with comprehen-
sive, synthetic descriptions of the innovative and essential features of the
case activities. I solicited practitioner commentary and analysis with regard
to whether the synthesis offered ‘fit,’ was accurate, and was comprehensive.
Glaser and Strauss emphasise the importance of joint coding and joint
analysis as a mechanism to arrive at accurate approximations of the context
being studied. It was in this spirit that I solicited feedback from relevant
practitioners.
Glaser (1992) might argue that reviewers add social sensitivity and the
ability to offer feedback that maintains analytic distance while drawing
upon experience to suggest modifications and validate asserting. Glaser and
Strauss (1967, p. 27) might contend that this approach allowed me to ‘check
out’ the ‘emergent set of propositions’ represented by the report. They might
also contend that the process is confirmatory, adding a ‘purposeful’ and
‘systematic’ character to the analysis (p. 28). Glaser and Strauss might fur-
ther make the case for practitioner response as allowing descriptions that are
‘quite rich, complex, and dense’ to develop further and become more fitting
and relevant (p. 32). On the basis of the ongoing feedback and review from
these practitioners, I finalised the descriptive, synthetic reports. These re-
ports served as the foundational data set for Tier 3, Category Generation
and Modification. It is in this sense that I have a genuinely multiple-tiered
process of data analysis.
The third tier involved the generation of categories of social practices that
may contribute to the formation of social capital. A category is a specifically
defined dimension of concept, a portion of a system of classification of the
divisions articulated in a schematic framework. Social capital can be defined
as the relationships and networks that facilitate collective action and access
to resource. My project proposed the use of social scientific research to
distinguish this concept’s categories, in other words, an empirical approach
to the delineation of this concept’s dimensions. Concepts are descriptive and
explanatory devices designed to facilitate communication. Indicators ope-
rationalise concepts. A category stands by itself as a conceptual element of
a concept, and as a necessary stage for indicator development. The best
approach to category development, I argue, is the systematic discovery of
categories that emerge from social scientific evidence.
Categories are simply a classificatory structural unit of a concept that
social scientists distinguish to aid the development of indicators. I sought
categories for indicators that will fit the concept and the situation being
researched and work when put into use through measurement. Following
Glaser and Strauss, by ‘‘‘fit’ I mean that the categories must be readily (not
forcibly) applicable to and indicated by the data under study; by ‘work’
I mean that they must be meaningfully relevant to and be able to explain the
behavior under study’’ (1967, p. 3). Categories, I argue, must be clear
enough to be readily operationalised. They should also be coherent and
130 ERIC TUCKER
clear to other social scientists and to significant laypersons. One caveat

should be made. Despite the fact that categories should fit both the concept
and the situation being researched, they should aim for wide applicability.
Categories are generated from the data but entail a degree of conceptual
abstraction.
CONCLUSION: TOWARDS A GROUNDED-

INDICATOR APPROACH
This essay proposes the development of a ‘grounded-indicator approach’ to

public policy research. In coining the term, this essay defines a ‘grounded-
indicator approach’ as an approach that entails grounding indicators in
social research (i.e. the purposeful systematic generation of indicators from
the data collected in the field). A grounded-indicator approach derives in-
dicators from the study of phenomena the concept represents. Grounded
indicators are discovered, developed, and provisionally verified through
systematic collection and analysis of data that pertain to those phenomena.
The developer of grounded indicators does not begin with an indicator and
then establish its validity. Rather, he or she begins with fieldwork and allows
relevant insights to emerge. The purpose of a grounded-indicator approach
is to generate indicators that are faithful to and illuminate the concepts and
phenomena under study.
Researchers measuring social capital formation with standard indicators
often feel as though they are fitting an indicator from an unrelated theory to
the phenomenon being studied; or, ‘‘forcing ‘round data’ into ‘square cat-
egories’’’ (Glaser & Strauss, 1967, p. 37). To ground indicators researchers
tease out why emerging data disagrees with established understanding,
allowing measures to develop gradually as data and interpretations
accumulate. This means continually redesigning and reintegrating indica-
tor-related notions and constantly checking out tentative indicators as data
pour in. Because current understanding of many social phenomena is not at a
point of theoretical saturation, researchers should aim to collect data to
develop categories, map dimensions, and explore similarities and differences,
coherence and incoherence, and the relative importance of a given measure.
Grounded indicators, when put into practice, provide us with relevant
predictions, explanations, interpretations, and applications. In addition to
the extant standards for verification of indicator validity, common grounds
for assessing indicator quality exist: consistency, clarity, fit, integration,
parsimony, scope, and workability. Grounded indicators additionally aim to

be usable in practical applications, describe accurately and thoroughly, en-
able prediction and explanation, provide a novel perspective on action and
behaviour, and facilitate theoretical advances. Put another way, grounded
indicators seek to be both analytic (designating characteristics of entities)
and sensitising (illuminating the phenomena and yielding a ‘meaningful’
picture).
As noted at the beginning, measurement is at the centre of much edu-
cational research. Yet researchers are beginning to describe instances where
there seems to exist a substantial gap between established measurement
instruments and the social phenomena they intend to measure. This chapter
argued that the most prominent existing procedures to secure valid indica-
tors tend to over-emphasise verification and correspondingly de-emphasise
empirical research that aims towards generation and modification. It pro-
ceeded to make the case that the features and strengths of ethnography
specifically, and qualitative research more generally, make it uniquely suited
to contribute to the development of new indicators and the improvement of
existing indicators. To provide some warrants for these arguments, this pa-
per first offered a brief reading of the state of the practice of social science
empirical research on measurement. It then turned to a specific example of
this author’s ethnographic research project that aims to improve indicators
of social capital formation, in the hopes of providing an exemplary instance
of how ethnographic methodologies might contribute to the generation
and modification of new, quality measurement instruments. In concluding
with a proposal for a ‘grounded-indicator approach’ to social science,
I acknowledge that my thinking on this topic is still tentative. If the idea
of a ‘grounded-indicator approach’ proves feasible, however, I hope to
strengthen the mandate for generating grounded indicators, broadening the
picture of what ethnographers, among others, do with their efforts and
contribute to public policy research.
REFERENCES
Acharya, A., & Wall, H. J. (1994). An evaluation of the United Nations’ human development
index. Journal of Economic and Social Measurement, 20(1), 51.
Allalouf, A., & Ben-Shakhar, G. (1998). The effect of coaching on the predictive validity of
scholastic aptitude tests. Journal of Educational Measurement, 35(1), 31–47.
Archer, S. E., Kelly, C. D., & Bisch, S. A. (1984). Social indicators. In: S. E. Archer (Ed.),
Implementing change in communities: A collaborative process (pp. 118–140). St. Louis:
The C.V. Mosby Company.
132 ERIC TUCKER
Asher, M. A, & Defina, R. H. (1995). Age-adjustment of income inequality trends: A meth-

odology critique. Journal of Economic and Social Measurement, 21(1), 33.
Bardach, E. (2000). A practical guide for policy analysis. The eightfold path to more effective
problem solving. New York, NY: Chatham House Publishers, Seven Bridges Press.
Ben-Shakhar, G., & Sinai, Y. (1991). Gender differences in multiple-choice tests: The role of
differential guessing tendencies. Journal of Educational Measurement, 28(1), 23–35.
Boles, K., Yoo, T.-Y., Ebacher, A., Lee, K., & Ashton, M. C. (2004). Validity studies:
Psychometric properties of scores on the French and Korean versions of the Hexaco
Personality Inventory. Educational and Psychological Measurement, 64(6), 992–1006.
Bolger, N., & Kellaghan, T. (1990). Method of measurement and gender differences in
scholastic achievement. Journal of Educational Measurement, 27(2), 165–174.
Bryman, A. (2004). Social Research Methods (3rd ed.). Oxford: Oxford University Press.
Clapham, M. M. (2004). The convergent validity of the Torrance tests of creative thinking and
creativity interest inventories. Educational and Psychological Measurement, 64(5), 828–841.
Clarke, M. (2000). News or noise? An analysis of U.S. news and world report’s ranking scores.
Educational Measurement: Issues and Practice, 21(4), 39–48.
Collazo, A. A. (2005). Translation of the Marlowe–Crowne social desirability scale into an
equivalent Spanish version. Educational and Psychological Measurement, 65(5), 874–900.
Davis, M. H., Capobianco, S., & Kraus, L. A. (2004). Measuring conflict-related behavior:
Reliability and validity evidence regarding the conflict dynamics profile. Educational and
Psychological Measurement, 64(4), 707–731.
DeCarlo, L. T. (2005). A model of rater behavior in essay grading based on signal detection
theory. Journal of Educational Measurement, 42(1), 53–76.
De Vaus, D. A. (1995). Surveys in social research. St. Leonards, NSW: Allen and Unwin.
Dey, I. (1993). Qualitative Data Analysis. London: Routledge.
Dowson, M., & McInerney, D. M. (2004). The development and validation of the goal ori-
entation and learning strategies survey (GOALS-S). Educational and Psychological
Measurement, 64(2), 290–310.
Esman, M. J., & Uphoff, N. T. (1984). Local organizations: Intermediaries in rural development.
Ithaca: Cornell University Press.
Farrall, S. (2004). Revisiting crime surveys: Emotional responses without emotions? Or, look
back at the anger. International Journal of Social Research Methodology, 7(2), 157–171.
Fleming, M., Jordan, J. S., & Lang, K. M. (1996). The impact of measurement error in the U.S.
National Income and Product Accounts on forecasts of GNP and its components.
Journal of Economic and Social Measurement, 22(2), 89.
Fletcher, R., & Hattie, J. (2005). Gender differences in physical self-concept: A multi-
dimensional differential item functioning analysis. Educational and Psychological Meas-
urement, 65(4), 657–667.
Florida, R. (2002). The rise of the creative class: And how it’s transforming work, leisure, com-
munity and everyday life. New York: Basic Books.
Florida, R. (2005). The flight of the creative class: The new global competition for talent (1st ed.).
New York: HarperBusiness.
Fortunato, V. J. (2004). A comparison of the construct validity of three measures of negative
affectivity. Educational and Psychological Measurement, 64(2), 271–289.
Frisbie, D. A., & Cantor, N. K. (1995). The validity of scores from alternative methods of
assessing spelling achievement. Journal of Educational Measurement, 32(1), 55–78.
Galvez, J., & McLarty, C. (1996). Measurement of Florida temporary residents using a
telephone survey. Journal of Economic and Social Measurement, 22(1), 25.
Giancarlo, C. A., Blohm, S. W., & Urdan, T. (2004). Assessing secondary students’ disposition
toward critical thinking: Development of the California measure of mental motivation.
Educational and Psychological Measurement, 64(2), 347–364.
Glaser, B. G. (1992). Emergence vs. forcing: Basics of grounded theory analysis. Mill Valley,
California: Sociology Press.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qual-
itative research. New York: Aldine de Gruyter.
Gomm, R., Hammersley, M., & Foster, P. (2000). Case study method: Key issues, key texts.
Thousand Oaks: Sage.
Gordon, L. R. (1995). Fanon and the crisis of European man: An essay on philosophy and the
human sciences. New York: Routledge.
Gormely, P. J. (1995). The human development index in 1994: Impact of income on country
rank. Journal of Economic and Social Measurement, 21(4), 253.
Greene, J. C., Caracelli, V. J., & Graham, W. D. (1989). Toward a conceptual framework for
mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11(3),
255–274.
Greene, J. C., & Caracelli, V. J. (1997). Defining and describing the paradigm issue in mixed-
method evaluation. In: J. C. Greene & V. J. Caracelli (Eds), Advances in mixed-method
evaluation: The challenges and benefits of integrating diverse paradigms (New Directions
for Evaluation, No. 74) (pp. 5–17). San Francisco: Jossey-Bass.
Grégoire, J. (2004). Factor structure of the French version of the Wechsler Adult Intelligence
Scale-III. Educational and Psychological Measurement, 64(3), 463–474.
Hammersley, M., & Atkinson, P. (1995). Ethnography: Principles in practice (2nd ed.). London:
Routledge.
Hillier, B. (1996). Space is the machine: A configurational theory of architecture. Cambridge:
Hogan, T. P., & Agnello, J. (2004). An empirical study of reporting practices concerning
measurement validity. Educational and Psychological Measurement, 64(5), 802–812.
Hong, S., & Wong, E. C. (2005). RASCH rating scale modelling of the Korean version of the
Beck depression inventory. Educational and Psychological Measurement, 65(1), 124–139.
Howrey, E. P. (1996). Forecasting GNP with noisy data: A case study. Journal of Economic and
Social Measurement, 22(3), 181.
Husserl, E. (1960). Cartesian meditations: An introduction to phenomenology (D. Cairns, Trans.).
The Hague: Martinus Nijhoff.
Kidder, L. H., & Fine, M. (1987). Qualitative and quantitative methods: When stories converge.
In: M. M. Mark & R. L. Shotland (Eds), Multiple methods in program evaluation: New
directions for program evaluation (pp. 57–75). San Francisco: Jossey-Bass.
Kiuranov, C. (1982). Social indicators as an instrument of social change. Madrid, Spain: In-
ternational Sociological Association.
Laumann, E. (1994). The social organization of sexuality: Sexual practices in the United States.
Chicago. Chicago: University of Chicago Press.
Le, H., Casillas, A., Robbins, S. B., & Langley, R. (2005). Motivational and skills, social, and
self-management predictors of college outcomes: Constructing the student readiness in-
ventory. Educational and Psychological Measurement, 65(3), 482–508.
134 ERIC TUCKER
Lowe, P. A., & Reynolds, C. R. (2005). Factor structure of AMAS-C scores across gender
among students in collegiate settings. Educational and Psychological Measurement, 65(4),
687–708.
MacRae, D., & Whittington, D. (1997). Expert advice for policy choice: Analysis and discourse.
Washington: Georgetown University Press.
Maddison, A. (1995). Explaining the economic performance of nations: Essays in time and space.
Aldershot, Hants, England: Edward Elgar.
Maddison, A. (2001). The World Economy: A Millennial Perspective. Paris, France: Develop-
ment Centre of the Organisation for Economic Co-operation and Development.
Maddison, A. (2003). The World Economy: Historical Statistics. Paris: Development Centre of
the Organisation for Economic Co-operation and Development.
Maller, S. J., & French, B. F. (2004). Universal nonverbal intelligence test factor invariance
across deaf and standardization samples. Educational and Psychological Measurement,
64(6), 1030–1052.
Marshall, C., & Rossman, G. B. (1995). Designing qualitative research. Thousand Oaks, CA:
Sage.
McGuckin, R. H., & Peck, S. (1993). Manufacturing establishments reclassified into new in-
dustries: The effect of survey design rules. Journal of Economic and Social Measurement,
19(2), 121.
McKenzie, J. (2000). Scaffolding for success. [Electronic version] Excerpted from: Beyond
Technology, Questioning, Research and the Information Literate School Community.
Retrieved October 31, 2006, from http://fno.org/dec99/scaffold.html.
Merleau-Ponty, M. (1962). Phenomenology of perception. New York: Humanities Press.
Mills, C. A. (1980). On social indicators and development. Tokyo: United Nations University.
Moneta, G. B., & Yip, P. P. Y. (2004). Construct validity of the scores of the Chinese version of
the need for closure scale. Educational and Psychological Measurement, 64(3), 531–548.
Moustakas, C. E. (1994). Phenomenological research methods. Thousand Oaks, CA: Sage.
Noble, C. L., Eby, L. T., Lockwood, A., & Allen, T. D. (2004). Attitudes toward working single
parents: Initial development of a measure. Educational and Psychological Measurement,
64(6), 1030–1052.
Pearce, L. D. (2002). Integrating survey and ethnographic methods for systematic anomalous
case analysis. Sociological Methodology, 32(1), 103–132.
Potenza, M. T., & Stocking, M. L. (1997). Flawed items in computerized adaptive testing.
Journal of Educational Measurement, 34(1), 79–96.
Powers, D. E., Fowles, M. E., Farnum, M., & Ramsey, P. (1994). Will they think less of my
handwritten essay if others word process theirs? Journal of Educational Measurement,
31(3), 220–233.
Putnam, R. (1993). The prosperous community: Social capital and public life. The American
Prospect, 4(13), 35–42.
Putnam, R. (2000). Bowling alone: The collapse and revival of American community. New York:
Simon and Schuster.
Robson, C. (2002). Real world research: A resource for social scientists and practitioner-
researchers (2nd ed.). Oxford: Blackwell.
Rossi, P. H., Lipsey, M. W., & Freeman, H. E. (2004). Evaluation: A systematic approach
(7th ed.). Thousand Oaks: Sage.
Sehmitt, A. P., & Dorans, N. J. (1990). Differential item functioning for minority examinees on
the SAT. Journal of Educational Measurement, 27(1), 67–81.
Sen, A. K. (1973). On economic inequality. Oxford: Clarendon Press.

Sen, A. K. (1987). On ethics and economics. Oxford: Basil Blackwell.
Sen, A. K. (1999). Development as freedom (1st ed.). New York: Alfred Knopf.
Sen, A. K. (2002). Rationality and freedom. Cambridge: Belknap Press.
Shriver, K. A. (1995). The measurement of current cost data: Implications of economic anal-
yses. Journal of Economic and Social Measurement, 21(1), 17.
Sireci, S. G. (1997). Problems and issues in linking assessments across languages. Educational
Measurement: Issues and Practice, 16(1), 12–19.
Smith, T. W. (1981). Social indicators: A review essay. Journal of Social History, 14(14),
739–747.
Stevens, T., & Tallent-Runnels, M. K. (2004). The learning and study strategies inventory high
school version: Issues of factorial invariance across gender and ethnicity. Educational
and Psychological Measurement, 64(2), 332–346.
Stevenson, J. C., & Evans, G. T. (1994). Conceptualization and measurement of cognitive
holding power. Journal of Educational Measurement, 31(2), 161–181.
Stokey, E., & Zeckhauser, R. (1978). A primer for policy analysis (1st ed.). New York:
W.W. Norton.
Strange, V., Forest, S., & Oakley, A. (2003). Using research questionnaires with young people
in schools: The influence of the social context. International Journal of Social Research
Methodology, 6(4), 337–346.
Strauss, A. L., & Corbin, J. M. (1990). Basics of qualitative research: Grounded theory pro-
cedures and techniques. Newbury Park: Sage.
Viswanathan, M. (2005). Measurement error and research design. Thousand Oaks: Sage.
Von Eckartsberg, R. (1998). Introducing existential–phenomenological psychology. In: R. Valle
(Ed.), Phenomenological inquiry in psychology (pp. 1–20). New York: Plenum Press.
Walford, G. (2001). Doing qualitative educational research: A personal guide to the research
process. London: Continuum.
Walford, G. (2002). (ed.). Doing a doctorate in educational ethnography. Oxford: JAI Press.
Studies in Educational Ethnography, 7.
Weimer, D. L., & Vining, A. R. (1999). Policy analysis: Concepts and practice. Upper Saddle
River, N.J.: Prentice Hall.
Whiting, E., & Harper, R. (2003). Young people and social capital. London: Office for National
Statistics.
Wolfe, E. W., Ray, L. M., & Harris, D. C. (2004). A RASCH analysis of three measures
of teacher perception generated from the school and staffing survey. Educational and
Psychological Measurement, 64(5), 842–860.
Yin, R. K. (1994). Case study research: Design and methods (2nd ed.). Thousand Oaks: Sage.
‘I WANT THE WORLD TO KNOW’:
THE ETHICS OF ANONYMITY IN
ETHNOGRAPHIC LITERACY
RESEARCH
Kristen H. Perry
When I first began exploring literacy practices among Sudanese refugees

three years ago, I approached several orphaned refugee youth – commonly
referred to as the ‘Lost Boys’ – in the state of Michigan in the US. Many of
these young men were quite happy to share their stories with me, to open
their lives to examination. I explained to these youth that I would be
changing their names to protect their identities, in line with the research
requirements agreed upon by my university’s Institutional Review Board
(IRB). One young man, however, questioned why I needed to change his
name. I explained that this was the standard procedure, and that it was done
to protect research participants from discrimination, persecution, or other
harm that might befall them if someone recognized their identities as a result
of my research. I explained that I really had no choice, given the nature of
the IRB approval our project had received. This young man then told me
that he was not willing to participate in my study if I was going to change his
name. He said, ‘I have something important to say, and I want the world to
know that I am the one who said it.’ We parted ways, and I found other

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12008-2
137
138 KRISTEN H. PERRY
refugee youth who were willing to participate anonymously (see Perry,

2005a, 2005b, 2006).
This Sudanese young man’s refusal to participate in my study bothered
me greatly, for two reasons. First, this particular individual was a commu-
nity leader among the refugee youth in the area; he did indeed have inter-
esting and important things to say, and I was saddened that I would not be
able to include his voice in my study, even if anonymously. More impor-
tantly, however, the young man’s words caused me to question the common
practice of changing participants’ names in research – particularly in eth-
nographic studies and other deeply personal qualitative methodologies.
Why is anonymity the ‘default’ assumption in research? What is privacy,
and who really is being protected by current guidelines for anonymity? What
does it mean to ‘protect’ participants? When is anonymity truly necessary,
and when does it actually silence participants’ voices? When, and why,
should participants’ real names be used? Who should have control over
issues of naming and anonymity in research – researchers, IRBs, or the
participants themselves? Do participants fully understand the implications
of this decision? For that matter, do researchers or IRBs?1
THE TRADITION OF ANONYMITY IN RESEARCH
Current practice, reified through typical IRB guidelines, has its roots in
traditional positivist frameworks about research. The positivist research
paradigm is traceable back to Enlightenment epistemologies, which em-
phasized the fact-based, value-free nature of knowledge (Christians, 2000;
Cunningham & Fitzgerald, 1996; Howe & Moses, 1999; Muchmore, 2000).
Christians (2000) suggests that researchers who were grounded in positivist
approaches used utilitarian perspectives on research ethics. The utilitarian
approach suggested that a single set of moral considerations could guide all
inquiry; these considerations were outlined in a generally accepted code of
ethics that emphasized informed consent, privacy/confidentiality, and ac-
curacy, and that opposed deception in research. Many of these conventions
were codified in national legislation in the USA beginning in 1974, in re-
sponse to several experiments that had mistreated research participants
(Hecht, 1995; UCRIHS, 2004). According to Christians (2000), ‘Three
principles, published in what became known as the Belmont Report, were
said to constitute the moral standards for research involving human sub-
jects: respect for persons, beneficence, and justice’ (p. 140). These principles
were intended to ensure that people participated in research voluntarily and
‘I Want the World to Know’ 139
anonymously; that researchers protected the well-being of their participants;

and that both the benefits and the burdens of research be distributed eq-
uitably (Christians, 2000). This legislation established requirements for IRBs
that would review and monitor federally funded research conducted by
universities and other institutions. Most universities and other research-
conducting institutions have since expanded the purview of their IRBs to
monitor all institutional research – not just that which is federally funded
(Hecht, 1995).
Like the larger code of ethics, the specific belief that researchers must
protect participants’ privacy and confidentiality at all costs is deeply rooted
in Enlightenment ideals. Christians (2000) suggests that these ideals reflected
beliefs in the ‘nonnegotiable status’ of privacy and ‘a sacred innermost self ’
(p. 139). The positivist tradition, encoded in university IRBs, relies on
safeguards to protect people’s identities and those of the research locations. Confiden-
tiality must be assured as the primary safeguard against unwanted exposure. All personal
data ought to be secured or concealed and made public only behind a shield of an-
onymity. Professional etiquette uniformly concurs that no one deserves harm or em-
barrassment as a result of insensitive research practices (Christians, 2000, p. 139).
Although rooted in the Enlightenment, these ideals are held no less deeply
today; current debates in the USA surrounding the tension between national
security and individual rights to privacy reflect just how dearly Americans
hold beliefs about privacy. However, certain beliefs about the nature and
conduct of research are becoming less and less appropriate as ethnographers
and other qualitative methodologists challenge the dominant positivist par-
adigm that shapes research in the human sciences. The issue of anonymity is
one such research convention that must be re-examined in our postmodern,
poststructuralist world.
CHALLENGING THE TRADITION
Questions and challenges to the tradition of ‘protecting’ research partici-

pants by altering their identities have arisen from a variety of fields including
anthropology, feminist studies, sociolinguistics, and sociology. While some
American researchers have questioned the practice, much of the scholarly
debate challenging anonymity comes from British researchers. Some chal-
lenges to the standard of anonymity raise practical questions: Is true an-
onymity even possible, particularly in this age of instant digital information?
Other challenges stem from deeper epistemological, ethical, and critical
methodological beliefs, particularly those held by feminist and other critical

scholars.
Practical Considerations: The Digital Revolution
Many scholars have wondered whether true anonymity is even possible in

research (Christians, 2000; Fine, 1990; Muchmore, 2000; Walford, 2005),
particularly in the digital information age (Hecht, 1995). Even when re-
searchers take great precautions to protect identities, pseudonyms and lo-
cations may be recognized (Christians, 2000; Walford, 2005). This is
particularly true when researchers study highly specific communities or or-
ganizations. Fine (1990), for example, argues that ‘the group of apocalyptic
flying saucer cultists that Festinger, Riecken, and Schachter (1956) describe
in such robust detail in When Prophesy Fails, can be linked to their mun-
dane, real, and earthly existence by a search of Chicago-area newspapers of
the period’ (p. 77).
What was true for a study in 1956 is even more so for current and future
research, as new digital technologies make information and identities highly
accessible to anyone with an Internet connection and a Google search bar.
Walford (2005) cites the example of Peshkin’s (2001) study of Edgewood
Academy, the pseudonym of an elite private school in New Mexico. Using
the Internet, Walford not only quickly identified the school as Albuquerque
Academy, he even discovered that the school’s library web entry for
Peshkin’s book announced that Peshkin had spent a year studying the
academy for his book. So much for preserving anonymity! Not only do
digital technologies make identifying sites and participants easier, but re-
searchers themselves are increasingly using research technologies that are
more likely to identify participants. A decade ago, Hecht (1995) argued that
researchers had turned to new technologies – he named microcassette re-
corders, palm-sized camcorders, and tiny radio-transmitting microphones –
in increasing numbers. As Hecht commented:
The presentation of data is also changing, as researchers turn back to these new tech-
nologies for conveying the important messages gathered through their studies. It is not
unusual to attend a regional or national meeting where a researcher no longer just reads
the findings from his or her study but will display photographs, audio or video record-
ings, or electronic interactions as part of their presentations (p. 5).
A decade later, this is even more true, as data are increasingly recorded in
digital formats that can be sent around the globe with the click of a mouse –
or even via cell phone.
Hecht (1995) rightly questions, ‘Can a subject remain truly anonymous,

and what does it mean to insure a subject of his or her confidentiality, given
the expanding role of research technology?’ (p. 5). He also suggests that
current legislation concerning research activities, and the IRB regulations
that ensue, may not adequately account for this technological revolution in
data gathering. While these standards may be somewhat behind the times in
terms of technology, they are certainly outdated in relation to recent schol-
arship concerning power dynamics and the nature of the researcher–par-
ticipant relationship, particularly in ethnographic and other qualitative
methodologies.
Power and Research
For the past 15 years or so, scholars – particularly those working within
critical traditions – have considered issues of power within research. Many
of the issues raised relate to concerns about respect, beneficence, and justice,
the same principles that Christians (2000) suggests guided the establishment
of positivist-based research guidelines. However, when viewed through crit-
ical lenses, these principles take on different meanings and raise different
questions. What does it mean to respect those who participate in research?
How is respect embodied in the researcher–participant relationship? Who is
benefiting from this research – and how can participants, both individuals
and communities, share more equitably in the benefits? How can research be
designed so that it not only increases knowledge, but also furthers the cause
of justice in the world?
Researchers working within feminist perspectives have challenged as-
sumptions that are based in a positivist framework for research (Christians,
2000; Powell & Takayoshi, 2003). Traditional research practices typically
produce unequal power distributions between researchers and participants;
in most studies, researchers hold most of the authority and control in
the research relationship, and they also gain the most from the process
(Cameron, Frazer, Harvey, Rampton, & Richardson, 1993). Christians
(2000) describes the feminist communitarian model, where ‘participants
have a say in how the research should be conducted and a hand in actually
conducting it’ (p. 145). In contrast to positivist expectations of distance,
objectivity, and anonymity, feminist researchers advocate empirical research
that fosters reciprocity, collaboration, and the formation of meaningful re-
lationships between researchers and participants (Powell & Takayoshi,
2003).
Cameron et al. (1993) present three models of research – ethical, advo-

cacy, and empowerment – and they label these ‘research on’, ‘research for’,
and ‘research with’ participants. The authors suggest that the ethical
model produces the most asymmetrical power distributions between re-
searcher and researched; the advocacy model increases the likelihood that
research participants may benefit from the research study, but still positions
them as objects of study. Only the empowerment model, according to the
authors, actively seeks to more equitably redistribute power among partic-
ipants and researchers. The authors describe several aspects of power
that need to be considered in the empowerment model of research. One
important power consideration involves determining how people are rep-
resented in a study. The authors argue that by sharing these decisions with
participants, ‘the researcher maximizes their opportunities for defining
themselves in advance of being represented’ (p. 90). Likewise, when more
reciprocal relationships are achieved in qualitative research, Christians
(2000) suggests that ‘invasion of privacy, informed consent, and deception
are nonissues’ (p. 149).
Scholars also have questioned who benefits most from research. Under
typical frameworks – what Cameron et al. (1993) refer to as ethical models –
researchers benefit the most from the conduct and publication of research in
the social sciences (Cameron et al., 1993; Fine, 1990). Fine argues, ‘Authors
have a power that is rarely matched by the poor soul who is being depicted’
(p. 76). Of course, Fine’s choice of ‘poor soul’ to refer to research partic-
ipants is interesting in itself, although he uses it to make a point about how
participants are frequently depicted by researchers. Fine argues that because
researchers hold all (or most) of the power in research relationships in
typical models, they also therefore hold great responsibility to protect par-
ticipants. However, Fine also suggests, that this process often denies par-
ticipants proper credit for their contributions to the study and the
production of knowledge – which are, of course, substantial. Walford
(2005) even suggests that anonymity is more protection for the researcher
than the researched: ‘Perhaps the idea of anonymity allows researchers to
write their books and articles with less concern for absolute accuracy and to
base their arguments on evidence which may not be as strong as desirable’
(p. 89).
Critical scholars suggest that current practices and regulations are biased
toward patriarchal structures and are paternalistic in nature (Christians,
2000; Grinyer, 2002). These practices and regulations assume that research-
ers and IRBs know – better than participants themselves – what the best
interests of research participants are and how to protect them.
Real Participation: Empowering Participants
Paternalistic beliefs about research assume that all participants not only
deserve protection, but that they desire to remain anonymous (Grinyer,
2002). This is clearly not always the case, as the young man who refused to
participate in my study shows. Other researchers have encountered similar
desires among participants to be named. Grinyer (2002), for example, con-
ducted research among families of young adults who had been diagnosed
with cancer; she found that three-quarters of her participants wished to use
their own names in the study. In fact, Grinyer reports that one participant
who elected to use a pseudonym later regretted it. This participant,
Gabrielle, wrote a letter to Grinyer that explained her regret:
Looking back I was very disappointed to not see Stephen’s and my name in print. Even
though my words were there, I felt as though I had somehow lost ownership of them and
had betrayed Stephen’s memory. That was entirely my own fault. I was also upset
because my family and friends found it odd as well. They expected and wanted y our
names too (Grinyer, 2002, p. 3).
Grinyer, like many critical scholars, concludes that it is problematic for

researchers to ‘make judgements [sic] on behalf of others, however, well
intentioned’ (p. 4).
Others have also recognized the need to increase participants’ agency
within the participant–researcher relationship. Powell and Takayoshi (2003)
argue that qualitative researchers, particularly those who have developed or
are developing reciprocal relationships with participants, must follow their
instincts in designing and conducting research – even when these instincts
challenge accepted practice. Powell and Takayoshi acknowledge that IRBs
are important influences in the processes of designing and making decisions
about research. However, they also argue that ‘what is ethical in the midst of
the study might not necessarily fall within IRB guidelines governing the
researcher’ (p. 418). In cases such as these, Powell and Takayoshi advocate
what they term ‘informed disobedience’ (p. 418) on the part of the re-
searcher, who likely knows better than the anonymous IRB reviewers what
is ethical in the context of the research.
CASES FROM ETHNOGRAPHIC STUDIES
Powell and Takayoshi (2003) suggest that ‘not enough of our scholarly
reflection considers the potential complications of real people and real
situations’ (p. 412). In order to look at real people and the situations that
face them, I now present examples from ethnographic literacy research
contexts, particularly from my own studies, that help to illustrate why an-
onymity is not always appropriate, and why participants might choose to
reveal their identities.
Southern Sudanese Refugees
My work with Southern Sudanese refugees began shortly after I returned

from two years as a US Peace Corps Volunteer in Lesotho, Southern Africa.
I learned about a local community of the so-called ‘Lost Boys’ who were
seeking academic tutors, and I began working closely with three orphaned
youth, one just entering college and the other two still in high school. After a
year and a half of this work, one of my students asked me to serve on the
board of the Southern Sudan Rescue and Relief Association, a local or-
ganization comprised of both Sudanese refugees and Americans. My close
relationship with these young men also led to a study of how they practiced
literacy in their everyday lives, particularly comparing their literacy lives in
Africa with their literacy practices in the US and how these literacy practices
compared to those of formal schooling. This work, and my deepening re-
lationship with the local Sudanese community, led to my current research
project, which involves examining the role of culture in the development of
literacy practices of young children in intact Sudanese refugee families.
Historical and Cultural Context

The history of the Sudan is a troubling one, and it is also one that is coming
to be increasingly prominent in both international news media and global
politics. The Southern Sudanese are members of various tribes located in
Southern Sudan; these southerners, typically black African Christians, have
been engaged in a civil war against the northern-dominated Arab Muslim
government for over 20 years.2 The war that displaced the Southern
Sudanese was the result of centuries of deep ethnic and religious divisions.
The country’s government is controlled by the North, and it has system-
atically worked to subjugate the African South by imposing Muslim sharia
law, making Arabic the official national language, and turning a blind eye to
the traditional practice of enslaving Southern Sudanese (Bok, 2003; Deng,
1995). In 1983, the Southern Sudanese rebelled against the atrocities
of the northern-dominated government by creating the Sudanese People’s
Liberation Movement and Army (SPLM/A), which engaged the government
in a civil war that lasted over two decades. This war completely devastated
Southern Sudan. At least two and a half million people have been killed since
the beginning of the conflict, and five million people have been displaced as
refugees. Militias bombed, pillaged and destroyed villages and crops, slaugh-
tered families, raped women, and captured both women and children to be
taken to the North, where they were kept as slaves and forced to convert to
Islam (Bok, 2003; Yang, 2002). Like the current conflict in Darfur, the con-
flict in Southern Sudan caused a mass exodus of southerners, many of whom
ended up as refugees in Egypt or other Middle Eastern nations or in the
Kakuma Refugee Camp near Lake Turkana in Kenya. Southern Sudanese
refugees typically spend many years in these nations of refuge. Many of these
refugees are lucky enough to be granted asylum in countries such as the
United States, Canada, Great Britain, and Australia, and these countries
have made a particular effort to resettle the thousands of orphaned youth
commonly known as the ‘Lost Boys’ (U.S. Department of State, 2001).
These ‘Lost Boys’ are a special case among the Southern Sudanese
refugees. Tens of thousands of Sudanese children, mainly boys, began a
mass exodus from the south in 1987. The group was comprised mainly of
boys for two primary reasons: First, boys fled their villages in reaction to
news that the armies on both sides were abducting boys and forcing them
to fight. Second, many young boys were away from home, tending to herds
of animals in remote cattle camps, when militias descended upon their
villages, destroying the villages and slaughtering their families (Yang, 2002).
The orphaned youth began an arduous and dangerous journey of over
1,000 miles – entirely on foot. After a treacherous crossing of a crocodile-
infested river, where thousands of boys drowned, approximately 33,000
Lost Boys reached refugee camps in Ethiopia, where they remained for
nearly four years. Following a coup of the Ethiopian government, the
refugees were forced back into the Sudan. Only 7,000 of the original
group survived to reach the Kakuma Refugee Camp in Kenya in 1992
(Yang, 2002).
In January of 2005 the Sudanese government and the SPLM signed a new
peace accord that outlines an agreement for power-sharing between the
North and the South and that makes provision for an eventual election
where the South may determine whether or not to remain part of the Sudan
(Embassy of the Republic of the Sudan, 2005). At that time, many Southern
Sudanese expressed the hope that this accord will finally bring peace to the
region. The sudden death of Dr. John Garang, the longtime leader of the
SPLM and the newly instated vice-president of Sudan, in an airplane ac-
cident on July 30, 2005, along with the subsequent uprising of ethnic
violence and the ongoing conflict in Darfur have significantly dampened the
optimism that accompanied the signing of the peace accord.
The Research Projects

Although my work with the orphaned youth and with the refugee families
comprised two distinct studies, both projects were informed by a sociocul-
tural perspective on literacy. Within this framework, literacy is a social
practice shaped by social, cultural, economic, political, and ideological fac-
tors (Barton & Hamilton, 1998; Barton, Hamilton, & Ivanič, 2000; Street,
2001). Researchers who study literacy practices shed light into the myriad
ways in which individuals navigate and negotiate the various literacy land-
scapes and contexts that surround them, and they offer implications for the
ways in which literacy instruction may be made more relevant to learners’
real-world lives. In the ‘Lost Boys’ study (Perry, in press, 2005a, 2005b), I
sought to understand (1) the meaning of literacy to the refugees; (2) the life
domains that contextualized their literacy practices; (3) how they used
different languages across and within those domains; and (4) the extent to
which school literacy practices aligned with the everyday literacy practices
of these refugees. With my current and ongoing study of refugee families
with young children (see Perry, 2005c), I am pursuing similar issues. In
addition to the questions explored with the orphaned youth, I am exploring
the ways that the refugee experience has shaped the literacy practices of the
parents, as well as the ways in which the young children are appropriating,
transforming, and otherwise making sense of the literacy worlds across the
contexts of home, community, and formal schooling.
For both studies, I used a fairly typical ethnographic case study meth-
odology. Both studies relied heavily on participant-observation, various
types of interviews, and the collection of different types of literacy artifacts.
Participant-observation in both studies involved spending a great deal of
time in participants’ homes; one way I gained access and fostered reciprocity
with the participants was by offering academic tutoring and community
mentoring or cultural broker services to the participants. Although I cen-
tered my study around focal participants and focal families in both studies,
the research also involved a great deal of ‘hanging out’ at community events
such as church services and traditional celebrations. In the study of refugee
families, I also observed the focal children in their school classrooms. In-
terviews in both studies focused on ascertaining the types of texts and lit-
eracy practices that participants engaged with, both in their previous lives in
Africa and in the USA, as well as collecting life histories and other impor-
tant cultural information from focal participants. I collected a variety of
literacy artifacts, including texts that were read and written by both focal
participants and other community members, audio recordings of events such
as church services, and photographs of literacy events and the literacy en-
vironments.
Like most ethnographic studies, these methods had the potential to be
highly invasive to participants’ lives. I not only spent a great deal of time
with the participants, but I also asked them deeply personal questions that
often led them to recount awful experiences or to reflect on the difficulties of
their lives in the USA. Using photographs and audio recordings also had the
potential to be invasive. In both studies, however, I found that instead of
balking at a potentially invasive situation, participants actually welcomed
the opportunity to talk about their experiences and to educate me (and,
by extension, the larger world) about their people, their culture, and their
history.
Anonymity Decisions
In many cases, such as the study among families of young people diagnosed
with cancer (Grinyer, 2002), revealing identities by using real names allows
participants to retain ownership of their voices. As Fine (1990) suggests, it
also helps to give credit where credit is due. While working with these
refugees, I came to believe that using participants’ real identities not only
would allow them to retain ownership and credit, but it might serve to
empower them at the same time. Because of this belief – largely sparked by
the young Sudanese man I described at the beginning of this chapter – I
chose to challenge the practice of automatic anonymity in my study of the
refugee families. Instead of telling participants that I would be automatically
changing their names, I helped the families decide whether or not to use
their real names or to remain as anonymous as possible. Although it is
impossible for researchers to know the full extent of potential benefits and
risks in a study (Muchmore, 2000), I explained what I foresaw as possible
benefits, as well as the possible drawbacks of participation in my study.
Much to my surprise, all three families clearly wanted to use their real
identities in the project. I was surprised largely because I knew the histories
of these families – their lives had not been easy, and they had sought refuge
in the United States as a result of great oppression and violence in the
Sudan. Although I believed it was important for participants to be given the
choice about how they were represented, I expected that most would choose
to protect their identities, given what they had endured. My earlier
experience with the Sudanese youth should have taught me otherwise! Viola,
one of the three mothers in my study, explained her decision:
We don’t have any political problems or anything to hide our names. Our problems – we
need people to know what is going on in our country and how we are suffering here. It’s
not easy for new refugees here to be like American people. They struggle in many
ways y We need them to know our culture, our tradition, because of our kids.
Viola’s words startled me – how could she claim that she did not have any
political problems? Viola’s very existence in the USA was the result of
political problems. Viola herself had told me about one of her relatives who
still works for the Sudanese government, but who does underground work
for Christian organizations in the Sudan; Viola explained that if the gov-
ernment found out about this, her relative likely would be assassinated. To
me, this alone would be reason enough to wish to remain anonymous!3 Yet,
this is precisely why researchers – let alone IRBs – should not make de-
cisions about naming and anonymity on behalf of participants; my as-
sumption would have been totally wrong. Viola’s words indicate that the
naming issue directly relates to issues of identity and empowerment for her;
she believed that using her real name would directly benefit both her own
children and the wider Sudanese community, who are resettling in this
country in ever greater numbers.
Viola’s desire to be identified may also be an attempt to retain her sense of
who she is. Refugees, unlike other types of immigrants, typically settle in
new countries not so much because they wish to, but because they have to –
their very lives are often in danger in their native lands or in surrounding
countries. Resettlement and the adjustment to a new place can be difficult
experiences, as Viola notes. Maintaining a voice and an identity may be a
powerful way to cope with the challenges faced by refugees. Indeed, Francis
Bok describes in his autobiography, Escape from Slavery (Bok, 2003), how
he was captured by Arab northerners in the Sudan and forced into slavery as
a young boy. After he left the Sudan, Bok was surprised that some refugees
did not want to talk about their experiences. Bok felt differently: ‘My story,
however, was all I had with me, the only remnant of my past’ (p. 105).
In some cases, name-changing can be its own form of oppression. Forcing
individuals to change their names has, in fact, been a common technique of
oppression used in the Sudan. Bok’s captor forced him to convert to Islam
and changed his name to Abdul Rahman; Bok reclaimed his own name after
he escaped from slavery. In my research with the ‘Lost Boys’ – indeed, this
term, while used by many of the youth to refer to themselves, was bestowed
upon them by a Western journalist who compared them to the orphans in
Peter Pan – Chol4 told me about a name-changing experience his uncle

encountered while studying at a university in Khartoum, Sudan’s capital.
Chol’s uncle continually found himself failing at the bottom of the class,
although he studied hard and knew the material. When he complained
about this, an acquaintance suggested he change his name to something
more Arab-sounding; after Chol’s uncle changed his name, he suddenly
found himself at the head of the class. Given these contexts, it is under-
standable why many Sudanese refugees might take a dim view toward hav-
ing their names changed.
Salvadoran Campesinos
In other contexts, revealing participants’ true identities may actually offer

them more protection than anonymity would. In their book, Now We Read,
We See, We Speak (Purcell-Gates & Waterman, 2000), the authors con-
ducted an ethnographic study among campesinos in El Salvador. The book
focuses on eight poor rural women who were becoming literate after the
brutal civil war in that country. In their book, the authors chose to reveal
the women’s true identities – not only using their real names, but also
including photographs of the women in the book. Purcell-Gates and
Waterman believed that revealing the women’s true identities would actually
offer them more protection than would pseudonyms; government reprisals
against the women could actually have been easier to perpetrate if the
women’s identities had been hidden from the world’s view. In addition,
Purcell-Gates explained: ‘We also asked the women and people in the com-
munity which they would prefer and they all said they wanted their names
attached to their stories. They had fought long and hard (in a very bloody
civil war) for this and they wanted it out there!’ (personal communication,
February 15, 2006).
CONCLUSION: NEED TO CHALLENGE STANDARD

PRACTICE
Qualitative researchers, particularly those who work within ethnographic

and other deeply personal traditions, need to seriously rethink our research
practices that accept anonymity as the norm. Allowing participants – not
IRBs – to make informed decisions about how they are represented has
many potential benefits. The discussion and decision about naming versus
anonymity can provide one step toward equalizing power relationships be-
tween researchers and participants, and it also helps researchers to under-
stand that they cannot necessarily assume what their participants’
perspectives or best interests may be. Offering this choice to participants
allows participants to retain ownership of their words, stories, and iden-
tities; it respects them as individuals rather than as ‘subjects of study’. In
some extreme cases, naming participants may actually do more to protect
them than anonymity would. Revealing identities also forces researchers to
think even more carefully about how they represent their participants, what
they choose to describe, and how they describe it (Walford, 2005).
Although I believe that revealing participants’ real names may be im-
portant (and that offering participants that choice is imperative), I am in no
way suggesting that researchers should always reveal participants’ identities.
There are cases and contexts where it would be inappropriate – indeed,
irresponsible – to do so, for example, while I know the name of Viola’s
relative who does underground work in the Sudan, I cannot in good con-
science put that person’s name in print. Many individuals and communities
are vulnerable, and the threat of discrimination, persecution, or other harm
is very real. However, researchers and IRBs cannot necessarily assume that
they know ahead of time which individuals or communities may be threat-
ened, as these cases illustrate. All of the cases described in this paper are
potentially vulnerable populations – refugees, young children, the terminally
ill, the rural poor who are becoming literate – yet, each group contained
individuals who were eager to share their stories and their names.
The issue of how we represent young children and who makes the decision
about that representation is, I think, particularly complex. David, Edwards,
and Alldred (2001) advocate for allowing children to have a greater role in
the decision-making processes that accompany participating in research.
However, the children who participated in their research were 10 years old
and above, and it was relatively easy for these researchers to argue that the
children had a good understanding of what was occurring in the research
process. Researchers who, like me, work with children who are younger
than seven may not be able to claim that these young participants fully
understand the research process or the potential implications that may arise
from their participation in the project. (Of course, it could also easily be
argued that adults may not fully understand these implications, either!)
Although I agree with David, Edwards, and Alldred that children should
have a voice in the research process, I also must acknowledge that this belief
is clearly situated in my White, middle-class, Enlightenment-based ‘pro-
gressive’ philosophies about individual rights and the nature of childhood.
Most IRBs, in contrast, believe that children have the right to assent to
participate, but that they cannot consent to participate (UCRIHS, 2004);
thus, consent is left to parents. From an IRB’s perspective, therefore, chil-
dren should not have the sole responsibility to make decisions about how
they are represented in research. Likewise, many cultures, particularly non-
Western ones, believe that parents should have full control over any de-
cisions made about their children’s lives – not only when the children are
young, but as they reach adulthood, as well. For example, among the
Sudanese refugees that I work with, not only is it common for parents to
choose their children’s spouses, but it is also common for parents to choose
their children’s careers. Viola, for example, explained to me that her father
played a very traditional role in choosing her career – he decided that she
would become a lawyer – but that he had been highly unusual in allowing
Viola to choose her own spouse. In communities such as these, where par-
ents exert a great deal of power in decision-making about children, allowing
children to have the sole voice in deciding how they are represented likely
would be a culturally foreign concept. For my study, I dealt with this issue
by asking the family to discuss the issue and to come to a common
understanding. In all three families, however, the parents made the final
decision.
This cultural issue of children’s right to make decisions leads to a larger
point, which is that different ethnic and cultural communities do not nec-
essarily share the same European ideological history that shaped research
ethics in Europe and the US. beliefs about privacy and individual rights are
quite literally foreign to many communities. Anthropologists and cultural
psychologists have demonstrated that the nature of the self, of individual-
ism, and of community vary widely across cultures (see Markus &
Kitayama, 1991; Nisbett, 2003; Oyserman, Coon, & Kemmelmeier, 2002).
For example, Nisbett (2003) argues that the insistence on freedom of in-
dividual action and a strong belief in universal rules of proper behavior (e.g.,
that all research participants should be anonymous) are Western ideals that
are not shared by cultures in other regions of the world. Many cultures, such
as those found in Africa, Asia, and Latin America, tend toward more col-
lective, interdependent belief systems, rather than individualistic ones
(Markus & Kitayama, 1991; Oyserman et al., 2002). This may partly ex-
plain why the refugees that I work with or the campesinos described by
Purcell-Gates and Waterman (2000) chose to use their real identities; for
them, privacy or an individual’s ability to make an independent decision
may not have been a prominent concern. And, as Viola’s rationale for using
her real name illustrates, participants from communities such as these may
feel that the ultimate benefit to the larger community greatly outweighs any
individual personal risk that they may face by participating in the research
study. Each of the cases that I have described in this chapter therefore
presents a strong argument for the consideration of cultural context when
making decisions about how participants are represented in research.
Questions to consider
When should researchers use participants’ real names? There are no

straightforward answers to this question; naming and anonymity must be
negotiated on a case-by-case basis. Some situations, however, should lead
researchers to strongly consider using participants’ own names. First and
foremost, researchers clearly should seriously consider using participants’
own names whenever participants wish to be named. Researchers should
also consider this when identifying participants either (1) empowers their
voices, as in the cases of many of the Sudanese refugees described; or (2)
affords them more protection than anonymity might, as in the example of
Salvadoran campesinos.
In the ideal future, IRBs will come up with a new set of standards and
questions about participant identities in qualitative research, rather than
requiring that ‘researchers must propose to protect human subjects’ rights
with regard to privacy by using research designs that safeguard subjects’
privacy during the gathering and storage of data’ (UCRIHS, 2004). Until
that time, researchers may find it useful to use some questions, in conver-
sation with participants, when determining whether or not to use partic-
ipants’ names or to have them remain anonymous: What are the possible
benefits and risks of using participants’ real names? What are the possible
benefits and risks of using pseudonyms? How would using real names or
anonymity empower participants (and their families/institutions/communi-
ties)? How would using real names or anonymity threaten them? With either
decision, how can we ensure that participants are respected and represented
fairly?
NOTES
1. Some researchers and IRBs distinguish between ‘confidentiality’ and ‘anonym-
ity’ (Walford, 2005). The Michigan State University IRB, for example, explains:
‘Anonymity means that no one, including the principal investigator, is able to as-
sociate responses or other data with individual subjects. Investigators may promise
anonymity only under this condition. Data collection involving face-to-face or taped
recording is not anonymous. Confidentiality means that although subjects’ identities
may be known to the investigator(s) and the research staff, subjects’ identities will be
kept confidential and reports of research findings will not permit associating subjects
with specific responses or findings’ (UCRIHS, 2004). This definition is clearly
grounded in a positivist, quantitative framework. For the purposes of this paper,
therefore, I use the term ‘anonymity’ to refer to the practice of using pseudonyms to
mask participants’ identities.
2. This conflict has existed far longer than the current conflict in Darfur, although
the issues are similar in both conflicts.
3. This is also an excellent reason to keep this particular relative’s identity hidden.
4. ‘Chol’ is a pseudonym, as anonymity was required by the IRB approval agree-
ment for that study.
REFERENCES
Barton, D., & Hamilton, M. (1998). Local literacies: Reading and writing in one community.
London: Routledge.
Barton, D., Hamilton, M., & Ivanič, R. (Eds) (2000). Situated literacies: Reading and writing in
context. London: Routledge.
Bok, F. (2003). Escape from slavery. New York: St. Martin’s Press.
Cameron, D., Frazer, E., Harvey, P., Rampton, B., & Richardson, K. (1993). Ethics, advocacy
and empowerment: Issues of method in researching language. Language and Commu-
nication, 13(2), 81–94.
Christians, C. G. (2000). Ethics and politics in qualitative research. In: N. K. Denxin & Y. S.
Lincoln (Eds), Handbook of qualitative research, (2nd ed). Thousand Oaks, CA: Sage
Publications.
Cunningham, J. W., & Fitzgerald, J. (1996). Epistemology and reading. Reading Research
Quarterly, 31(1), 36–60.
David, M., Edwards, R., & Alldred, P. (2001). Children and school-based research: ‘Informed
consent’ or ‘educated consent’? British Educational Research Journal, 27(3), 47–365.
Deng, F. M. (1995). War of visions: Conflict of identities in the Sudan. Washington, DC: The
Brookings Institution.
Embassy of the Republic of the Sudan. (2005). Sudan government and SPLM/A sign a com-
prehensive peace agreement ending Africa’s longest conflict [Electronic version].
SudaNews, February–March.
Festinger, L., Riecken, H., & Schachter, S. (1956). When prophesy fails. New York: Harper &
Row.
Fine, G. A. (1990). Credit and blame in ethnographic publishing. The American Sociologist,
21(1), 76–79.
Grinyer, A. (2002). The anonymity of research participants: Assumptions, ethics, and prac-
ticalities [Electronic version]. University of Surrey Social Research Update, 36. Retrieved
July 8, 2004, from http://www.soc.surrey.ac.uk/sru/SRU36.html
Hecht, J. B. (1995). The Institutional Review Board in social science research. Paper presented
at the annual meeting of the Mid-Western Educational Research Association, Chicago,
IL.
Howe, K. R., & Moses, M. S. (1999). Ethics in educational research. Review of Research in
Education, 24, 21–59.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emo-
tion, and motivation. Psychological Review, 98, 224–253.
Muchmore, J. (2000). Methodological and ethical considerations in a life history study of
teacher thinking. Paper presented at the annual meeting of the American Educational
Research Association, New Orleans, LA.
Nisbett, R. E. (2003). The geography of thought: How Asians and Westerners think differ-
ently y and why. New York: The Free Press.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and col-
lectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological
Bulletin, 128, 3–72.
Perry, K. (2005a). ‘Helping our people’: The role of literacies in mediating community among
Sudanese refugees. Paper presented at the meeting of the American Educational
Research Association, Montreal, Canada.
Perry, K. (2005b). From storytelling to writing: Transforming literacy practices among Suda-
nese refugees. Paper presented at the 14th world congress of applied linguistics/inter-
national association of applied linguistics, Madison, WI.
Perry, K. (2005c). Making sense of literacy landscapes: A case study of a Sudanese refugee
child’s literacy practices in America. Paper presented at the meeting of the national
reading conference, Miami, Florida.
Perry, K. (2006). Sharing stories, linking lives: Literacy practices among Sudanese refugees. In:
V. Purcell-Gates (Ed.), Cultural practices of literacy: Case studies of language, literacy,
social practice, and power. Mahwah, NJ: Lawrence Erlbaum.
Peshkin, A. (2001). Permissible advantage? The moral consequences of elite schooling. Mahwah,
NJ: Lawrence Erlbaum.
Powell, K. M., & Takayoshi, P. (2003). Accepting roles created for us: The ethics of reciprocity.
College Composition & Communication, 54(3), 394–422.
Purcell-Gates, V., & Waterman, R. (2000). Now we read, we see, we speak: Portrait of literacy
development in an adult Freirean-based class. Mahwah, NJ: Lawrence Erlbaum.
Street, B. (2001). The new literacy studies. In: E. Cushman, G. R. Kintgen, B. M. Kroll & M.
Rose (Eds), Literacy: A critical sourcebook (pp. 430–442). Boston: St. Martin’s Press.
UCRIHS, Michigan State University. (2004). Instructions for completing the application for
initial review of a project involving human subjects. Retrieved February 26, 2006, from
http://www.humanresearch.msu.edu/applications/Initial_Application_Instructions.doc
U.S. Dept of State. (2001). Fact sheet: Sudanese (Kakuma) youth, June 11. Retrieved February
28, 2004, from http://www.state.gov/g/prm/rls/fs/2001/3398.htm
Walford, G. (2005). Research ethical guidelines and anonymity. International Journal of
Research & Method in Education, 28(1), 83–93.
Yang, D. C. (2002). Kakuma Turkana. St. Paul, MN: Pangaea.
EVERYONE GENERALIZES, BUT
ETHNOGRAPHERS NEED TO
RESIST DOING SO$
Geoffrey Walford
During the early part of 2006 I attended two international conferences

on ethnography and education. Both were large conferences with several
parallel sessions, so the papers I attended may have been idiosyncratic and
non-representative, but one issue that struck me was the frequency of ci-
tations to Paul Willis’ book Learning to Labour, first published in 1977. I felt
it was remarkable that this particular ethnography should still be so widely
quoted, and it re-emphasized my view that, while it probably has been the
most influential ethnography of education of the last half century, this par-
ticular ethnographic study’s success has much to do with over-generaliza-
tion–both by the author and by readers. Further, that over-generalization of
the experiences of some non-conformist lads has led to a lack of under-
standing (or even problematization) of why, at that time, the bulk of work-
ing-class boys actually did get working-class jobs.
Of course, I am far from being alone in critiquing certain aspects of
Learning to Labour. Soon after its original publication various authors crit-
icized it for what was seen as its glorification of laddish culture, its lack
of consideration of ethnicity, its sexism and so on. Indeed, the amount of
$
This chapter draws in part on extracts from Walford (2001, 2002, 2005).

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12009-4
155
156 GEOFFREY WALFORD
attention to it given by critics was probably one of the reasons it became

recognized as a classic study. Few would agree with Durkheim’s conclusions
in his classic study of Suicide, but the continued attention given to it, in part,
acts to solidify its reputation. Similarly, the collection of papers linked to
Willis’ work that was originally presented at a special session at the American
Education Research Association’s Annual Conference in 2003, to mark the
25th anniversary of the book, and later published in book form served to
reinforce the book’s classic reputation (Dolby, Dimitriadis, & Willis, 2004).
My concern with the book is embodied within the subtitle of the book:
‘How working class kids get working class jobs’. For the book has been
taken, and is still being taken by many ethnographers, as showing exactly
this – how working-class young people (at best, restricted to boys and at that
particular time in history) creatively constructed a culture that echoed in
many ways the culture of the shop floor jobs that they eventually occupied.
The cultural form that the lads created was seen to have within it a ‘partial
penetration’ of the conditions of existence of its members and of their
position within a class-based society. Their behaviour and attitudes towards
school were seen as having a form of logic and rationality, yet, there is
a moment ‘in working class culture when the manual giving of labour power
represents both a freedom, election and transcendence, and a precise inser-
tion into a system of exploitation and oppression of working class people’
(Willis, 1977 p. 120).
It is worth briefly examining another book that was published by Paul
Willis (1978). While it was published a year after Learning to Labour (and
actually by a far more prestigious publisher), his far less well-known Profane
Culture had its genesis several years prior to it.
Profane Culture was based on Willis’ doctoral thesis that was awarded in
1972. It presents the results of two separate ethnographic studies of two very
different groups of young people who he calls ‘the motor-bike boys’ and ‘the
hippies’. The first group were located through contact with an unnamed
motor-bike club linked to a church in an unnamed large English city. He
conducted field work and interviews over a nine-month period with a loose
and variably numbered group which centred round one person and had a
core of five people with another two who were often present. The second
group were contacted through attendance at a public house in a large un-
named industrial city. Over a period of five months Willis spent time with
three different small groups of three young people and interviewed them in
their homes. Both of these studies were thus small scale – which is exactly
what would be expected for a doctorate – and the book documents elements
of the cultures of these two small groups of young people.
Everyone Generalizes, but Ethnographers Need to Resist Doing So 157
But the book claims far more than is warranted. The introduction to the
book states:
This book presents two important cultures generated during the 1960s and still widely
influential today - the motor-bike boys, sometimes known as rockers, and the hippies,
sometimes known as heads or freaks. The form of the book is of two ethnographic
accounts of the inner meanings, style and movement of these cultures, but the essential
theme of the book is that oppressed, subordinate or minority groups can have a hand in
the construction of their own vibrant cultures and are not merely dupes: the fall guys in a
social system stacked overwhelmingly against them and dominated by capitalist media
and commercial provision.
From two very limited case studies, the claims have moved to being about
entire sub-cultures. From the small-scale and particular, claims about whole
cultural phenomena are made.
Of course, once a book is published the author has little control over how
it is interpreted. Many of the ways in which Learning to Labour has been
used to bolster arguments, find little support in the actual text, and the book
is far more nuanced than most readings suggest. However, the pattern
started in Profane Culture is repeated in Learning to Labour. The larger
study of which Learning to Labour was a part involved six separate groups
of boys (mostly non-conformist boys) in five different schools, but the heart
of the ethnography is a group of 12 non-academic (sometimes used syn-
onymously with non-conformist) working-class young men at an uniden-
tified secondary modern school in an unidentified town. The main school
they attended was at the heart of an inter-war housing estate and had ‘sub-
stantial West Indian and Asian minorities’ (p. 4) which were not studied;
the town itself had a population of about 60,000 but was part of a large
industrial conurbation. On page 4 Willis states ‘I wanted to be as certain as
possible that the group selected was typical of the working class in an in-
dustrial area’, but he also states that the 12 boys themselves (‘the lads’) were
selected on the basis of friendship links and membership of some sort of
oppositional culture within the school. These ‘lads’ were, in fact, deliberately
selected to be not representative of the wider body of working-class boys in
the school or of working-class boys in general; they were selected because
they were perceived to be part of a particular oppositional school culture.
Although we are not told how regular or extensive the contact was, these 12
were followed from the second term of their penultimate year in school, and
through their final year. The ‘lads’ and three other young men were then
followed into work where they were observed for a short period and in-
terviewed about their work. Teachers and parents were also interviewed.
There is little reason to doubt the quality and quantity of data generated
in this study or the depth of relationship between Willis and the lads; the
problem is in the analysis of these data and the conclusions that are drawn.
The problem of inappropriate generalization from the experiences of a small
non-randomly selected sample of boys in one particular school is a central
part of the text. The move from the particular culture of these 12 lads to the
much more general ‘working-class culture’ is made continually, when all
that can be realistically claimed is that this particular cultural form is one of
the many possible working-class cultural forms that could be generated -
indeed, only one of the many possible non-conformist forms that could
be generated. While the book does include some discussion of alternative
cultures that some other working-class students develop in school, the
reader is led to understand that it is only this particular form ‘created’ by the
lads that has continuities with working-class jobs involving manual labour.
The experiences and cultural forms created by the other boys are presented
as the other end of a bipolar ‘lads’ and ‘ear’oles’ dichotomy.
Of course, we all generalize. Everyday life requires it if we are to make our
way through each day. Take the first paragraph of this chapter for example.
I state that I attended two large ethnography and education conferences and
that I was surprised by the large number of times that Learning to Labour
was cited. I give no indication of where these conferences were held, how
many sessions I actually attended, or even what I mean by ‘large’. I also
state that those papers that I saw might be idiosyncratic and unrepresenta-
tive, but my guess is that most readers were prepared to accept my claim
that my experience supported the ‘classic’ nature of the book. We use an-
ecdotes to try to generalize in everyday life, but ethnographers should be
more careful. Indeed, I feel that we should make every attempt not to
generalize but to bring out the specifics of each case that is studied.
THE DILEMMA OF GENERALIZATION
Willis presents just one example of a fundamental and longstanding di-

lemma within qualitative and ethnographic research (Walford, 2001). The
method requires a focus on a very small number of sites, yet there is often a
desire to draw conclusions which have a wider applicability than just those
single cases. Within the ethnographic literature about education there is a
plethora of examples where schools, or particular groups of children or
teachers within schools, are researched because they are seen as ‘typical’, or
because they can offer ‘insights’ into what may be occurring in other
schools. Thus, for example, we have many studies of single schools where
racism and sexism has been shown to occur. Strictly, such studies can only
show that events that have been interpreted as racist (usually by the re-
searcher alone, but sometimes by those involved as well) occurred in that
one particular school, with those particular teachers and students, at a time
that is often many years before the publication of the academic article or
book. Such studies cannot provide information about what might be hap-
pening now even in that same school and, most importantly, cannot provide
any evidence about what is happening or what was happening in any other
schools.
There are several standard attempts to deal with this dilemma. The first is
to argue that, while strict generalizability is not possible in the statistical
sense for one (or a small number of cases) cannot possibly be an adequate
sample drawn from a wider population of schools or classrooms, ethno-
graphic and qualitative studies can achieve ‘transferability’ through ‘thick
description’. If the authors give full and detailed descriptions of the partic-
ular context studied, it is claimed that readers can make informed decisions
about the applicability of the findings to their own or other situations.
Broadly, although in different forms, this view is presented by authors such
as Stake (1995), Becker (1990) and Lincoln and Guba (1985). However, in
order to be able to judge whether a particular finding from an ethnographic
study in one school is applicable in another, it is actually necessary to know
as much about that second school as about the first. While Agar (1996, p. 44)
and Gomm, Hammersley, and Foster (2000) argue that it might be possible
to use survey data to gain such information, this can only give the most
general of statistical information and cannot give information on school
cultures or other detailed school-specific features. Lincoln and Guba (1985,
p. 316) follow their logic through and argue that this means that one cannot
generalize from a case study to a wider population unless one makes un-
warranted assumptions about the wider population.
The second main way in which commentators have tried to deal with the
problem of generalizability is to argue for theoretical generalization. Nu-
merous authors (e.g. Bryman, 1988; Schofield, 1990; Silverman, 1993; Yin,
1994) have sought to move away from statistical or empirical generalization
from case studies, and have proposed that the wider significance of findings
from a particular ethnographic study can be derived through the strength of
logical argument for each case. A case is significant only in the context of a
particular theory, and logical inference replaces statistical inference. An ex-
trapolation can be made between a particular case and a wider population
only if there is a strong theoretical or logic connection between them. The
strength of the theoretical reasoning is seen to be crucial. This means that

the selection of the case to be studied is crucial. Rather than attempt to find
a school or classroom that is ‘typical’ or for which ‘there is no reason to
believe it is untypical’, theoretical generalization requires a clear theoretical
basics for the choice. This may be seen as a version of Glaser and Strauss’
(1967) theoretic sampling, applied here to the initial choice of research site
rather than the individuals and situations to be sampled once the site has
been chosen. However, the attempt to by-pass empirical generalization
through the idea of ‘theoretic gereralization’ is, in the end, a failure. For
there to be a convincing theoretical argument for extrapolation, there needs
to be empirical evidence about the wider population. This requires a great
deal of empirical work with the wider population as well as with the case
study school. Indeed, one problem is usually defining what any ‘wider pop-
ulation’ actually is in relation to the cases studied.
SOME PERSONAL EXAMPLES
Much of my own ethnographic work has been dogged by considerations of

generalizability. When I published my study of boys’ private boarding
schools I still tried to argue that, while it was based on ethnographic re-
search in just two schools, there were aspects that applied to a wider range
of schools. I claimed that a limited generalizability was possible. Although
the book was given the title of Life in Public Schools (Walford, 1984, 1986a,
1986b) I made it clear that any generalization could only be made to the
limited wider population of the 29 schools then members of the Eton and
Rugby groups of schools which I believed, shared many similarities. I
explicitly excluded Eton, bringing the population down to 28, and argued
that there was sufficient information available on all of these 28 schools for
me to be able to assess generalizability. I spent four weeks in one uniden-
tified school and a term in another, made visits to several of the others in the
two groups, and was able to interview members of staff who had previously
been teachers in practically all of them. The frequent meetings between
teachers within the two groups meant that all schools were well known to
teachers in the others. I now believe that it would have been better to have
restricted my account to the two schools and have made no further attempts
at generalization.
In contrast, my next major study was of a named institution – my own
university at that time – Aston University in Birmingham (Walford, 1987).
Aston was one of the universities most badly affected by the cuts in
government funding implemented in 1981, and I recognized that there was

an important story to tell about the ways in which the institution was being
restructured. There was no question about generalization, as this was a
study of the reactions within one particular university. Its wider interest was
simply that the severity of the reductions in government funding were such
that extreme changes were being made. It was an indication of what could
happen, but not a prophesy of what would happen within other universities.
My following ethnographic study was of a named school – the City
Technology College, Kingshurst (Walford & Miller, 1991; Walford, 1991a,
1991b, 1991c). This ethnography was undertaken as part of a linked project
to a wider study of the City Technology College (CTC) initiative (Whitty,
Edwards, & Gewirtz, 1993) but, again, there was no intention of making
generalizations about other CTCs from the study of this one. The school
was important in itself, simply because it was the first of this new type of
school to be established and because it was at the centre of the initial con-
troversy over funding and selection of children. I expected that there would
be a wide audience for an account of selected aspects of ‘what it was like’.
I had no desire to make any generalizations, but pressures were put on me
by the publisher to make statements about the whole scheme and to draw
general conclusions. In particular, they wished the book to be called City
Technology Colleges and to include a detailed account of the wider initiative.
I resisted the pressure, even though I might have led to higher sales, and
stuck to the singular title, City Technology College, to emphasize the re-
stricted nature of the study.
More recent work, which has been qualitative but not ethnographic, has
focussed on a group of small private evangelical Christian schools and their
involvement as a pressure group in influencing government legislation
(Walford, 1995a, 1995b). This was then followed by a somewhat similar
linked study that examined the implementation of the particular aspects of
the 1993 Education Act that made it possible to obtain state funding for
some private schools (Walford, 1997, 2000). In both cases, the work in-
volved visits to schools where interviews with headteachers, teachers and
others were conducted and documentary material collected. Only 15 schools
eventually benefited from the legislation and achieved state funding, and the
success of each school was widely publicized in both the local and national
press. Each story was, in itself, worth telling. There was no temptation to
make any generalizations beyond each school itself.
Thus, my qualitative and ethnographic work is unusual in that most of it
has dealt with named schools, for it was impossible and undesirable to
conceal their identities. In most cases, the choice of site was made because
these particular schools were of interest in themselves, and not because I had
any expectation of being able to generalize from them.
THE IMPORTANCE OF NAMING
I have argued elsewhere (Walford, 2005) that ethical guidelines produced by

various organizations and university ethical review boards are misplaced in
assuming that the default position with ethnographic research should be that
sites are not named. There are various ethical reasons why such an assump-
tion is open to question, but here I wish to consider the more practical issue
of what the effects of not naming sites might be on authors and readers.
There is little doubt that a major reason for not naming sites or people
involved in ethnographic research is part of an access strategy. Most re-
searchers probably believe that it is easier to obtain access if anonymity is
promised, for it suggests that it is possible to reduce any possible harm by
making it impossible for the site to be identified. It is open to severe question
as to how reasonable any promise of anonymity actually is in practice,
but reduction of possible harm to participants may not be the main reason
for such pledges. Several authors have proposed further reasons for not
naming sites or individuals. In a recent article, Scheper-Hughes reflects on
her research in Ireland (1979) and her revisit after 20 years. She writes
(Scheper-Hughes, 2000, p. 128):
Still, were I to be writing the book for the first time and with hindsight, of course there
are things I would do differently. I would be inclined to avoid the ‘cute’ and ‘conven-
tional’ use of pseudonyms. Nor would I attempt to scramble certain identifying features
of the individuals portrayed on the naive assumption that these masks and disguises
could not be rather easily decoded by the villagers themselves. I have come to see that the
time-honoured practice of bestowing anonymity on ‘our’ communities and informants
fools few and protects no one – save, perhaps, the anthropologist’s own skin. And I fear
that the practice makes rogues of us all – too free with our pens, with the government of
our tongues, and with our loose traditions and interpretations of village life.
Anonymity makes us unmindful that we owe our anthropological subjects the same
degree of courtesy, empathy, and friendship in writing as we generally extended to them
face to face where they are not our ‘subjects’ but our book companions without whom
we quite literally could not survive. Sacrificing anonymity means we may have to write
less poignant, more circumspect ethnographies, a high price for any writer to pay. But
our version of the Hippocratic Oath – to do no harm, in so far as possible, to our
informants – would seem to demand this.
In this article Scheper Hughes begins to suggest less altruistic reasons for
why anonymity might be offered to participants and organisations that are
involved in ethnographic research. It may benefit the researcher more than

the researched.
Perhaps the idea of anonymity allows researchers to write their books and
articles with less concern for absolute accuracy and to base their arguments
on evidence which may not be as strong as desirable. If named schools and
people are being discussed, the need for very strong evidence before claims
are made becomes obvious. At the extreme, writers could be sued for libel in
a way that is difficult to do where names are not used. Researchers are able
to hide poor evidence behind the pseudonyms without those researched
being able to make a challenge.
But there are further worrying possibilities. Jan Nespor (2000) sees
anonymization as a representational strategy with interesting ontological
and political implications, the most striking of which, he believes, have to do
with the way anonymization naturalizes the decoupling of events from his-
torically and geographically specific locations. In other words, the fact that
we do not name a site gives the findings of the research a spurious gen-
eralizability. If we attempt to conceal details about a school, it becomes a
more ‘general’ place – a school that ‘could be any school’, a school which is
just one example of many. The pseudonyms selected by some authors often
exaggerate this tendency (e.g. Anytown, Hightown, Beachside). Ethnogra-
phers thus implicitly invite readers to see their findings as being applicable to
other situations. Yet, to be able to understand any school, readers really
need to know the school’s history and geographical location, its physical
facilities and appearance and the nature of the students it serves and the
staff who teach there. Each school is unique in structure, culture and or-
ganization. The way it responds to change can only be understood in the
context of its history and socio-cultural and political location.
While, as readers, we intellectually accept the lack of generalizability of
ethnographic work, we are seduced by the lack of specific details about the
site and situation such that the significance of particular pieces of research
expands to fill our general understanding of the issues. Thus Learning to
Labour (Willis, 1977) has been widely taken to explain why ‘working class
kids get working class jobs’, yet it is based mainly upon a study of only 12
young men in a single school in a particular social, political and economic
context. For example, it is usually forgotten that these lads were actually in
the first year who were required to stay at school until 16 rather than 15. To
be required to stay at school for a year more than any young people had
been forced to stay before might well have had important consequences
for the culture created by these particular working-class boys. Similarly,
Beachside Comprehensive (Ball, 1981) is seen as giving information on the
effects of banding and streaming in secondary schools, and Rebels without a

Cause (Aggleton, 1987) is taken to explain some of the middle-class expe-
riences of the transition from school to work. More recently, Jo Boaler’s
(1997) work in two somewhat contrasting schools has been widely accepted
as indicating the relationship between teaching styles, setting and gender
and success in mathematics teaching, and Gillborn and Youdell’s (2000)
ethnographic study of two schools is taken to have shown the nature of the
‘A-C economy’ in British secondary schools as a result of recent policy
changes. As it is about mathematics, Boaler’s study is sometimes referenced
as if it had relevance way beyond even its British context.
The fact that none of the research schools is identified, implicitly gives the
writer and reader the chance to broaden the findings of each study beyond
the situation investigated. It gives a spurious generalizability of time and
space to the results of specific studies. While giving the actual names of sites
(and maybe of people involved) does not guarantee that readers will not
generalize, it might act to reduce the more outlandish examples and it would
certainly make writers more circumspect.
CONCLUSION
Ethnography of education draws from two rather different origins –
anthropology and sociology. In general, those who look more towards an
anthropological justification have few problems with the idea that single
ethnographic studies cannot be generalizable, for example, the work by
Stambach (2000) on a village school on Kilimanjaro or Foley’s (2004) ac-
count of schooling of Mesquakis from a Native American reservation.
What is more, anthropologists usually name their sites and are more likely
to see what they are doing as a straightforward ‘contribution to knowledge’.
Their work may have both long- and short-term benefits to the groups
studied, but that is not seen as a primary purpose.
Lack of generalizabilty should not be seen as a weakness. Named case
studies have a major contribution to make. They can inform us about, and
vicariously let us experience, settings to which we would normally have
no access (Gomm et al., 2000, p. 61). They can take us to a school on
Kilimanjaro, or on an Indian reservation, or to the first of a new type of
British school. They can allow us to appreciate the cultures created in each
setting through someone else’s eyes. They can give us information about a
wide diversity of cultures that it would be impossible for us to understand if
we had to experience them by ourselves.
My view is that ethnographic sites should be chosen because they or the

activities in them are intrinsically interesting or important in themselves, for
themselves. Generalization, whether it be empirical or theoretical, need not
be the goal at all. It might be that particular cases are selected because they
are important to understand in the light of their unique significance. This
might be, for example, their role within policy formulation and develop-
ment. Or it might be that the site is selected because particular activities
or cultural forms are known to be present.
Learning to Labour has been damaging to the development of educational
ethnography because Willis, and the readers of the book, made claims that
went beyond those that could justifiably be made from the study. It gen-
eralized from close involvement with 12 lads in a particular school, in
a particular social and physical location, at a particular time, such that it
appeared to show how working-class young people created cultural forms
where there were continuities between school and work. Both Willis and
readers largely ignored the fact that many working-class boys failed to cre-
ate cultures with these particular continuities. What was some of the di-
versity of cultural forms actually created by working-class young people? To
what degree were there continuities and discontinuities between the cultures
they created and the work they eventually took up? To what extent were the
specifics of the school, economic situation, time and place related to the
cultural forms? What are the limits to wider population that might have
created similar cultural forms?
It is very unlikely, of course, that these questions could all be answered,
but it is of great importance that they are asked. To not do so is to simply
accept the results of a small study as completely generalizable to an un-
known wider population. Willis’ book set a precedent for very many books,
theses and articles that followed to take a small-scale study and make ex-
aggerated general claims. I would argue that it is better for ethnographers
not to seek generalizabilty at all, but to study sites, groups and cultures for
what they. A major advantage of naming any site is that it reduces the
chance that either the author or the reader will be tempted to make spurious
generalizations.
REFERENCES
Agar, M. H. (1996). The professional stranger (2nd ed.). San Diego: Academic Press.
Aggleton, P. (1987). Rebels without a cause. London: Falmer.
Ball, S. J. (1981). Beachside comprehensive. Cambridge: Cambridge University Press.
Becker, H. S. (1990). Generalising from case studies. In: E. Eisner & A. Peshkin (Eds),
Qualitative inquiry in education. New York: Teachers College Press.
Boaler, J. (1997). Experiencing school mathematics. London: Falmer.
Bryman, A. (1988). Quantity and quality in social research. London: Unwin Hyman.
Dolby, N., Dimitriadis, G., & Willis, P. (Eds) (2004). Learning to labor in new times. London:
Routledge Falmer.
Foley, D. E. (2004). The heartland chronicles (with 2nd epilogue). Philadelphia: University of
Pennsylvania Press.
Gillborn, D., & Youdell, D. (2000). Rationing education. Buckingham: Open University Press.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory. Chicago: Aldine.
Gomm, R., Hammersley, M., & Foster, P. (2000). Case study and generalization. In: R. Gomm,
M. Hammersley & P. Foster (Eds), Case study method. London: Sage.
Lincoln, Y. S., & Guba, E. (1985). Naturalistic enquiry. Beverley Hills: Sage.
Nespor, J. (2000). Anonymity and place. Qualitative Inquiry, 6(4), 564–569.
Scheper-Hughes, N. (1979). Saints, scholars, and schizophrenics: Mental illness in rural Ireland.
Berkeley: University of California Press.
Scheper-Hughes, N. (2000). Ire in Ireland. Ethnography, 1(1), 117–140.
Schofield, J. (1990). Increasing the generalizability of case study research. In: E. Eisner
& A. Peshkin (Eds), Qualitative inquiry in education. New York: Teachers College
Press.
Silverman, D. (1993). Interpreting qualitative data. London: Sage.
Stake, R. (1995). The art of case study. London: Sage.
Stambach, A. (2000). Lessons from Mount Kilimanjaro. New York: Routledge.
Walford, G. (1984). The changing professionalism of public school teachers. In: G. Walford
(Ed.), British public schools: Policy and practice. Lewes: Falmer.
Walford, G. (1986a). Life in public schools. London: Methuen.
Walford, G. (1986b). Ruling-class classification and framing. British Educational Research
Journal, 12(2), 183–195.
Walford, G. (1987). Restructuring universities: Politics and power in the management of change.
Beckenham: Croom Helm.
Walford, G. (1991a). Researching the City Technology College, Kingshurst. In: G. Walford
(Ed.), Doing educational research. London: Routledge.
Walford, G. (1991b). City Technology Colleges: A private magnetism? In: G. Walford (Ed.),
Private schooling: Tradition, change and diversity. London: Paul Chapman.
Walford, G. (1991c). Choice of school at the first City Technology College. Educational Studies,
17(1), 65–75.
Walford, G. (1995a). Faith-based grant-maintained schools: Selective international policy
borrowing from The Netherlands. Journal of Education Policy, 10(2), 245–257.
Walford, G. (1995b). The Christian Schools Campaign – a successful educational pressure
group? British Educational Research Journal, 21(4), 451–464.
Walford, G. (1997). Sponsored grant-maintained schools: Extending the franchise? Oxford
Review of Education, 23(1), 31–44.
Walford, G. (2000). Policy, politics and education – sponsored grant-maintained schools and
religious diversity. Aldershot: Ashgate.
Walford, G. (2001). Doing qualitative educational research. London: Continuum.
Walford, G. (2002). Why don’t we name our research sites? In: G. Walford (Ed.), Educational
ethnography and methodology. Amsterdam, New York, Oxford: Elsevier.
Walford, G. (2005). Research ethical guidelines and anonymity. International Journal of

Research and Method in Education, 28(1), 83–93.
Walford, G., & Miller, H. (1991). City Technology College. Buckingham: Open University
Press.
Whitty, G., Edwards, T., & Gewirtz, S. (1993). Specialisation and choice in urban education. The
City Technology College experiment. London: Routledge.
Willis, P. E. (1977). Learning to labour. How working class kids get working class jobs.
Farnborough: Saxon House.
Willis, P. E. (1978). Profane culture. London: Routledge and Kegan Paul.
Yin, R. K. (1994). Case study research (2nd ed.). London: Sage.
EXTENDING VIDEO
ETHNOGRAPHIC APPROACHES
Dalia Aralas
The generation of insight for the advancement of ethnographic knowledge

arises not only from the capacity of researchers to avail themselves of in-
novative mediational means in the pursuit of new research objectives in new
research sites, but also it arises from the capacity of ethnographers to rethink
and reconceptualize contemporary approaches, whether the objectives be
new or old, the research sites familiar or exotic. This chapter focuses on the
need for the ethnographic eye to broaden its field of vision in grappling with
particular issues in qualitative research. Specifically, it will argue for the
considerable value of an approach that differs from but complements pre-
vailing approaches in current usage; this approach argues for a re-viewing of
directions within video-mediated research of an ethnographic orientation.
The discussion will make reference to a study undertaken to investigate the
nature of mathematical imagination within a classroom context at the upper
secondary level in Sabah, Malaysia. This chapter discusses specific method-
ological questions related in particular to the use of video recordings; the
substantive contribution of the research study is drawn upon to a limited
extent for the purposes of the present discussion. The chapter comprises five
parts. Part I argues for extending conventional approaches to the use of
video recording, discussing considerations for the use of a different, com-
plementary approach. Part II illustrates the approach of employing video
recording to generate different data. Part III discusses some of the ways that

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12010-0
169
170 DALIA ARALAS
the different data generated made possible an intensive interrogation of the

initial framework, leading to a different reframing of the phenomenon. Part
IV emphasizes the drastic difference in result from the employment of a
different approach, that is, a reconstruction of a more layered, alternative
framework that leads to a reconfiguration of views regarding the phenom-
enon under study, namely, a reconfiguration of mathematical imagination.
Part V concludes the chapter.
PART I
Current ethnographic research is marked by an expanding variety of ap-

proaches that indicates not only the infusion of paradigmatic proliferation
into the field but also the expansion of technologies mediating ethnographic
exploration as well as the growth in research agendas of educational re-
search. Current researchers therefore have much to draw upon but the di-
versity of approaches indicates and presents major challenges (Walford,
2002). In particular, the double crisis of representation and legitimation that
has become apparent in qualitative research has problematized the very
possibility of valid/trustworthy/authentic/useful research. Thus, before an
ethnographer even steps one foot, gingerly, in the research site, she would
have had to grapple with weighty issues, such as the transparency of lan-
guage, that have considerable bearing on the formulation of research ques-
tions and on the construction of an initial frame for her study. More
importantly, the agency of the researcher, as well as the agency of her
research subjects – vis-à-vis the structural constraints of epistemic access to
the nature of reality – has been cast into doubt, as has the capacity of the
researcher and the research subjects to produce a meaningful, dare we say
truthful, account of their ways of knowing.
In connection with attempts to gain access to the nature of the world or of
social realms, prevailing methodological wisdom emphasizes the necessity of
unobtrusive methods of data collection, whether as alternatives to or in
combination with direct elicitation methods. For example, Lee (2000) argues
that the major problems that beset the act of direct eliciting of data by
means such as interviews and questionnaires have been attributed to the
researcher’s presence. Lee (2000) cites works by many researchers such as
Webb for whom concerns about ‘researcher reactivity’ have prompted the
search for ‘unobtrusive measures’ that are deemed ‘non-reactive’. It is in the
current climate of advocacy of unobtrusive measures that video-mediated
research has been conducted. The works of Lee (2000), Kelly and Lesh
Extending Video Ethnographic Approaches 171
(2000), and Powell, Francisco, and Maher (2003) provide illustrations of

works that emphasize the importance of unobtrusive methods in pursuit of
the objectives of video-mediated research. To this end, the researchers
adopted the following approaches: the study of naturally occurring phe-
nomena, or the study of phenomena in naturalistic settings, or the use of
methods that were deemed least intrusive or disruptive of the natural phe-
nomena under observation. Unobtrusiveness is deemed desirable for the
research seeks to uncover the nature of educational or social processes as
they are, that is to say, in conditions that are either natural or in naturalistic
settings (classrooms without the presence of the researcher) or in controlled
conditions where the effects of significant variables are amenable to meas-
urement or be subject to conventional analysis in the production of research
accounts of natural phenomena.
For the purposes of this chapter, it is worth examining examples of such
works to illustrate the high significance conferred on unobtrusive video-
mediated approaches, without rejecting the worthwhile concerns raised
about research quality. The review of Powell et al. (2003) on literature of
‘‘videotape methodology’’ indicated that there were deliberate strivings,
made either explicitly or implicitly, in video-mediated research, in education
and other fields, to maximize unobtrusiveness and minimize researcher re-
activity. Examples included works from the 1950s that ‘‘involved the study of
detailed transcripts of ‘cinema film of naturally occurring interactions’’’
(Powell et al., 2003, p. 406). The authors also cited the TIMSS Videotape
Classroom Study that had a ‘‘goal to understand how teachers construct
and implement eighth-grade mathematics lessons in Germany, Japan and
the United States’’. The review also cited studies by many researchers that
raised concerns about and strove ‘‘to ameliorate the tendency of video im-
ages to falsely portray typicality’’ (Powell et al., 2003, p. 412). To explain the
popularity of videotape as a ‘‘medium for capturing and archiving data’’ in
many areas of research, Powell et al. (2003, p. 407) quoted Pirie (1996, p. 2)
for whom videotaping, as a method for researching learning in classroom
contexts, is ‘‘the least intrusive, yet most inclusive, way of studying the
phenomenon’’. It is instructive to consider Pirie’s study which explicitly
linked ‘‘all choices as to methods of data gathering y to one’s research
questions’’ which were about ‘‘the phenomenon of ‘folding back’ and
‘collecting’ within the Pirie–Kieren theory of mathematical understanding’’
(Pirie, 1996, p. 2). Her research reflected concerns about obtrusiveness
and researcher reactivity that are widespread in video-mediated research
that purports to examine teaching experiments or relationships between
teaching methods, pupil learning styles and pupil achievement. With regard
172 DALIA ARALAS
to classroom observations, a researcher is instructed to guard against

obtrusiveness by sticking strictly to the role of a fly-on-the-wall, as it were;
video recordings are to be made from distances deemed safely unobtrusive.
In practice, this involves attaching video cameras to stationary stands at
the back or sides of the classroom so as to not disrupt ‘normal’ classroom
proceedings. In the less usual case where the attention shifts from classroom
contexts or processes to the activities of focus groups, video cameras are
still attached to stationary stands and the researcher strives to minimize
reactivity. For example, Pirie (1996, p. 3) stated that she ‘‘chose to focus a
fixed camera’’ on ‘‘pre-selected groups of students’’ at a distance ‘‘close
enough to carry the outline of what the students were writing, but not
so close that I was not able to see all their faces and body language’’. The
video camera is fixed at a distance that enables visibility of facial expres-
sions and bodily gestures and audibility (technically enhanced) of pupil
utterances, but no closer. It is noteworthy too that Pirie discussed the
merits of the use of the methods of ‘‘video-stimulated recall’’ and ‘‘think-
aloud’’ in terms of obtrusiveness; the value of such methods was deemed
lessened to the degree they promoted changes in the pupils’ thinking proc-
esses (Pirie, 1996, pp. 6–7). These changes were deemed obstructive of the
goal of studying natural phenomena. To summarize, on the whole, the lit-
erature indicates an explicit and implicit advocacy of the importance of un-
obtrusiveness. If there were any works that advocated the deliberate
employment of obtrusive, disruptive methods, they do not appear on the
research radar.
The emphasis on unobtrusiveness is linked to concerns with validity or
other measures of research quality. A higher degree of unobtrusiveness is
deemed an indication of a higher degree of quality. The value of approaches
that purposefully use obtrusiveness or disruption to generate data has been
underplayed in mainstream research. Notable exceptions are works in the
traditions of ethnomethodology by Garfinkel (1967) and to a lesser extent,
ethnography. There is significance in this, for both traditions do not place
limitations on the study of naturally occurring phenomena by non-disrup-
tive means in naturalistic milieu; both traditions go so far as to problematize
the very nature of naturally occurring phenomena. In brief, there are differ-
ences in the value conferred upon unobtrusiveness and disruption in re-
search approaches because there are not only differences in the purposes of
particular studies. More importantly, the differences in valuing disruptive
approaches are based upon different methodological considerations regard-
ing the quality of research within fundamentally different perspectives about
the nature of social reality itself.
An ethnographer undertaking research that seeks to generate insight into

the ways learners make sense of the world cannot ignore the possibilities
that disruptive methods may afford. Contrary to prevailing wisdom in
mainstream video-mediated research, an ethnographer must consider and
make decisions in the field to tolerate obtrusiveness and deliberately act to
use disruption as one of the significant means of generating different sorts of
data if that were ethically justifiable and methodologically warrantable.
Ethical considerations regarding the right to privacy, safety, psychical and
emotional well-being of research participants were of central importance in
my study and these were given due and proper attention. Further, ethical
considerations in visual research are context specific (Banks, 2001) and are
beyond the scope of this chapter. What is noteworthy here is that there are
methodological considerations for a complementary approach, that is, a
video-ethnographic approach that deliberately uses obtrusive means and
capitalizes on disruption of the ‘natural’ flow of phenomena for the gen-
eration of different data in pursuit of worthwhile research purposes.
An important methodological consideration in support of such an ap-
proach pertains to the need to use methods that suit research purposes even
in the face of contrary research wisdom. Hodkinson (2001) has taken issue
with the apparent claim of the ‘‘new research orthodoxy in the United
Kingdom’’ that ‘‘method can ensure objectivity in research’’. Walford (2001,
p. 9) argued that the issue is not about ‘‘researcher subjectivity’’ which is ‘‘an
inevitable feature of the research act’’. The question should rather be di-
rected to warrants for methodological decisions. These considerations sup-
port the stance that flexibility regarding research approaches, including the
use of disruption, is not inherently unjustifiable. It may even be an imper-
ative that flexibility be warrantable in attempts to capture a social reality
that is problematic in every which way, and which would be unwise to
assume composed of structures possessed of an objectivity awaiting the
researcher’s cold impassioned gaze. Flexibility and the use of more than one
approach in a study, even if these approaches be contrary and lead not to
triangulation of data but offer more complication, are worthwhile in that
they generate all sorts of data that are important for the ethnographic ob-
jective of obtaining insight into the ways actors make sense of their social
worlds in the very process of building such worlds. Furthermore, taking on
different approaches generates alternative views of complex realities which
may afford researchers opportunities for the construction of different, if
contrary, models (Walford, 2001). This can only be positive for the growth
of understanding regarding phenomena that given their complex nature are
irreducible to any model within prevailing or existing perspectives. The
174 DALIA ARALAS
necessity of interrogating theories and revising, building new frameworks

give weight to the adoption of a complementary or alternative approach.
The specificities of these considerations for the research study conducted
into mathematical imagination are discussed below.
PART II
My research study needed to seek different ways of seeing social worlds, not
to replace conventional ways of seeing but to make additional layers to
enrich such seeing. That is, it is imperative that a researcher seeks to extend
the capacity of the study to address research questions in ways more pow-
erful than might be, if prevalent restrictions were adopted. Thus, the re-
search study used a combination of complementary approaches, but for the
purposes of this chapter, the disruptive approach will be highlighted. In-
troductory remarks about the study are given to provide a context for the
discussion.
The research had been undertaken primarily to investigate the nature of
mathematical imagination, with a view to explore and build an initial
framework for the analysis of mathematical imagination; the analysis was to
be informed by an interrogation of the interplay between theoretical and
empirical works, where the latter was derived from a qualitative study pri-
marily ethnographic in orientation. The data were generated mainly from
the classroom observations of mathematics classes at upper secondary level
(Year 10 and 11); the pupils were aged 16 on average and three-quarters of
them lived in relatively disadvantaged areas. Four teachers, two of whom
had been my former students as undergraduates, were involved in the study.
The data generated included video- and audio recordings of classroom ob-
servations, informal conversations with pupils during and after such obser-
vations, pupil written work, informal teacher conversations and interviews
before and after classroom observations and pupil and teacher question-
naires. The fieldwork – conducted in four schools in three towns in different
districts of Sabah – was undertaken in two phases in one year; the total
time in the field was six months. The length of time was deemed adequate
in further consideration of my background as a mathematics educator,
who has had extensive experience of observing similar classrooms in the
course of my professional work. There was a large amount of data gen-
erated; this was by design. There has not been any previous research that has
set out to study ‘mathematical imagination’ or ‘x imagination’, either of a
community of learners or experts, that I could turn to for reference. It made
sense then, for a study such as this was, to prepare as wide a ‘net’ as possible
to ‘capture’ the dynamics of real-life mathematizing by pupils in classroom
contexts.
At the beginning of the fieldwork, the following sub-questions (derived
from the main research question pertaining to the nature of mathematical
imagination) were asked: What pupil actions are potentially describable as
acts of mathematical imagination? What does mathematical imagination in-
volve? I had made a decision at the start of the study to focus on what I then
called dialogical instances that might appear to indicate shifts in mathemat-
ical engagement. The study had begun with an initial framework that con-
ferred central importance on dialogicality; this was continuously
interrogated. The study adopted a methodological stance different from
other studies in important ways. Notably, the study placed great importance
on the assumption that the very concept of imagination implies a shift in
engagement that is based on a rupture in the flow of ways pupils made sense
of situations. This implied an examination of discontinuity of action, of
processes as well as shifts in engagement. Such notions were interrogated
during analysis. Furthermore, with regards to the generation of data: whereas
other studies that examine pupil learning in classroom situations take as given
that data ought to be generated in a ‘naturalistic’ setting, the study argued for
the addition of a different approach to generate different data in the follow-
ing ways: first, the employment of video recording by the researcher in close
proximity to pupils and their work, second, the video recording (from both
far and very close distances) of unusual set-ups involving unfamiliar math-
ematical situations that pupils initially found bewildering.
The obtrusiveness of the approach adopted did not present difficulties and
did not give rise to less useful data. On the contrary, the approach made
available different data that served the purpose of studying the nature of
mathematical imagination in greater depth. The richness of the data ob-
tained thus allowed for a deeper analysis and afforded insights into a phe-
nomenon in ways that were not anticipated and made possible the
reconstruction of a richer model. It is noteworthy too that the approach
did not cause pupils discomfort, largely due to the great rapport that had
developed between the teachers, pupils and me. This aspect is discussed
more fully in a forthcoming report. It suffices here to say that the partic-
ipants were excited that they were taking part in the first ever video-me-
diated research of classrooms in Sabah. The pupils were sympathetic to my
plight as a researcher that their teachers reminded them of frequently. The
pupils gave their cooperation in ways that eased the generation of data and
they showed eagerness to talk about, and to display their work.
176 DALIA ARALAS
The researcher, with video camera held in her hands in free movements
(with the video camera secured by means of strong straps around one of her
arms) moved around the classroom freely, filming the whole class or groups
or individual pupils or their work. Depending on the decisions made about
the potential significance of events that involved mathematical imagination
in play, the researcher hovered above the pupils as they worked. The video
camera was at times held very close to the pupils, just above their heads or
sometimes even lower by their sides, as they grappled with the mathematical
situations individually, in silence or as they conversed with their neighbours.
The researcher recorded the pupils’ every move, as they wrote, gestured,
spoke to themselves or their peers, as they used rulers, compasses, models;
that is, as pupils employed all sorts of mediational means: material, ima-
gistic and linguistic during their work. The researcher used the zoom lens to
record every stroke of the pen or pencil as pupils made marks in their own
books or the common worksheets sometimes given to groups to work on;
the recordings included the pupils’ movements, utterances, hesitations and
pauses. At times the pupils made modifications to their writings and their
utterances, after some moments (perhaps on reflection) or after bursts of
conversations with their neighbours or after seeing a neighbour’s work. It is
noteworthy that the researcher recorded the pupils’ utterances not just with
the video but with an audio tape recorder. When the pupils had seemingly
completed a phase in the ways they grappled with situations, the researcher
moved to make video recordings of other students. The researcher moved
between the rows of desks of the pupils after video recordings of significant
phases. The researcher also used the zoom lens to capture, as closely as
possible, pupils’ movements and writings, when they were asked to show
their ‘working’ or to attempt a mathematical problem that they were re-
quired to solve at the blackboard in front of the classroom.
It is important to note that by breaching the safe distance of unobtru-
siveness deemed desirable in mainstream video research, the researcher was
able to gain access to the inside of the very ‘space’ of origination and un-
folding of pupil action, of the development of imaginative ways of pupil
grappling with situations. Standing very close to the pupils and using the
zoom lens, the researcher was able to access the precise forms of work
produced by pupils in the very unfolding of the ways pupils made sense of
the situations they confronted and invented or produced their own takes
and solutions and remade the frames they used for grappling. As the pupils
worked, the researcher recorded, in precise detail, the ways pupils began to
form letters, numerals even as they began to formulate their utterances to
themselves, to their peers or to their teachers, individually as well as
collaboratively. The recordings were made moment-by-moment from the

beginning of the formulation of the writings and sayings to the development
and conclusion of these works, with the accompanying pupil gestures, ut-
terances and hesitations about their work. These data are therefore seen in
their very form, in the very ways they are formulated in the very moments of
invention or production by the pupils. They are also available for a ‘genuine’
recall afterwards, to trace the development of the pupils’ movements, ut-
terances, playful trials and submitted written work, for a variety of
purposes. The latter include availability as a resource for the researcher,
for further conversations with pupils (for further data generation) and
equally importantly, availability as a resource for repeated re-viewing
during data analysis.
PART III
The use of complementary approaches, that is, the disruptive, intrusive ap-
proach detailed above, in conjunction with a conventional, unobtrusive ap-
proach where the researcher took the position of a fly-on-the-wall, as it
were, generated a collection of various sets of data. The collection of data,
including sets of data different from those that would have been obtained
using mainstream approaches, afforded an extension of the capacity of the
study to pursue the chosen research purposes that were different from ob-
jectives of previous research. The variety of data availed the researcher of a
different resource for analysis and interrogation of the initial framework
and opened up opportunities for reinterpreting data, reconceptualizing an-
alytical lenses for the examination of the phenomenon under study.
The availability of a different collection of data has implications for the
evaluation of the validity/authenticity/trustworthiness of the research, for it
impacted on writing descriptions of classroom observations, on the initial
stages of data analysis (initial coding, memoing, constant comparison) as
well as on the constant interrogation, construction, revision and reconfig-
uration of an analytical framework of mathematical imagination. This was
made possible by undertaking a hyper-spiral research path. This comprises
multiple reconstructions of cycles of looking at the research in a holistic
way, at every cyclical stage rebuilding the initial framework after every run
of data analysis. Such multiple cyclical reconstructions were therefore built
layer by layer, wherein at each layer a holistic reframing is undertaken.
There are many video-mediated research studies that make clear the ways
they conducted analyses of video data; Powell et al. (2003) and Pirie (1996)
178 DALIA ARALAS
are examples. The analysis conducted in the study clearly shared common-
alities as well as differences with the analyses in the literature. This chapter
focuses on analysis of the different sets of data generated in the study.
At each cycle of the data analysis, the researcher made repeated re-vie-
wings of video recordings with combinations of other data and previous
analysis of previous findings at previous cycles. Here, the re-viewings of
video recordings are discussed, for they are not repetitive (in a mechanistic
way) but are substantially different. First, during re-viewings and re-analysis
of data, further attempts were made to capture or revise particular relations
of significance between pupil action and spaces of action that did not ‘leap’
out immediately to me while in the classroom and were thus not recorded in
my field notes, or in previous visits during data analysis. The significance of
deferred judgements of relations is important. The significance is heightened
in the context of the hyper-spiral path undertaken in this study. The sig-
nificant data, relations among data, were not lost at a point during the
fieldwork path but had been made available as a resource at a further point
along the research path, particularly during different cycles of analysis at
different levels. Second, during repeated re-viewings and re-analysis of data
the researcher traced the moment-by-moment twists and turns in the de-
velopment of pupils’ convoluted actions. It was possible to form chains of
connections between instances, episodes and events so that they were in-
terpretable. The relations between action and spaces that shape or constrain
would not have been describable (much less interpretable in meaningful
ways) without recourse to a history of the development of the action, in-
cluding the difficulties related to the shifts, ruptures, discontinuities and the
development of spaces of action. During analysis, the researcher looked for
chains of connections between movements, utterance; chains of connections
between actions and spaces of origination of actions (difficulties or devel-
opment of shifts, ruptures, discontinuities); chains of divergence, conver-
gence, transformations of pupils’ works made possible descriptions as well
as interpretations of mathematical imagination in play. Third, during re-
analysis at different points of the hyper-spiral of cyclical revisions of anal-
ysis, the researcher added more details regarding different aspects of math-
ematical imagination made manifest, adding layer upon layer of insights
from previous analyses based on previous viewings of the video and other
data. Action was taken to revise and rename codes; reclassify, revise memos;
revise chains of linkages; revise the themes generated. These actions made
manifest the salience of relations that were not previously attended to.
Fourth, the analysis was deepened by revisiting the review of literature. The
findings generated, which were contrary to expectations, necessitated
rethinking aspects of mathematical imagination. In this regard, the literature

review that became increasingly more relevant as data analysis deepened
were primarily the works of Ibn al-Arabi (Chittick, 1989; Coates, 2002),
Wittgenstein (2001), Ernest (1998) and Kearney (1998). These works made
available important resources on imagination and mathematics and signifi-
cant ways of seeing that resonated with the data, the analysis, the inter-
pretations and the findings produced during the study. The creative use and
extensions of the resources offered by the works of Ibn al-Arabi (Chittick,
1989; Coates, 2002), Wittgenstein (2001), Ernest (1998) and Kearney (1998)
to the hyper-spiral path taken enabled the researcher to try on different
perspectival lenses that were more appropriate to the data, analysis and
interpretation. The different data generated, brought about findings that
were contrary to expectations and made necessary alternative framings. In
time, such multiple reframing enabled a rethinking of mathematical imag-
ination in terms of an alternative view comprising different modalities and
dimensions.
The hyper-spiral path resulted in greater depth and breadth of seeing. It
brought to the fore the significance of many themes; the most important of
these were the final themes generated at the higher levels of analysis, themes
that I called ‘‘creative agency’’, ‘‘creative intentionality’’ as well as themes
related to ‘space of the origination of action (leading to different levels of
transformation)’, namely ‘‘generative space’’. The pupils’ work evoked a
notion of ‘‘mathematical poesies’’. In brief, the different data generated by
the different approach made imperative the drastic reconceptualizing of the
phenomenon of mathematical imagination. The data analysis in the study
generated three main forms of mathematical imagination.
PART IV
To begin exploration, I had prepared a skeletal framework which posited a

notion of mathematical imagination as a form of mediated action, drawing
on a notion of mediated action as theorized by Wertsch (1998). There was a
variety of possible foci in classroom observations. At the start of the study, I
made a decision to focus on what I then called dialogical instances that
might appear to indicate shifts in mathematical engagement. An important
assumption that I had built into my initial working framework had been a
notion of dialogicality that focussed on utterances, deemed to be of the
utmost importance in the development of mental functioning (Bakhtin,
1981). The primary stress on dialogicality was largely derived from my
180 DALIA ARALAS
reading of the literature, mainly of socio-cultural orientation, related to

views of learning. Another assumption that I had built into my working
framework was an emphasis on the conceptual dimension of mathematical
imagination. This too was largely derived from my reading of the literature,
in the vein of analytical philosophy, related to views of imagination.
Specifically, I had drawn on a formulation of imagination provided by
White (1990). White, unfazed by centuries of complex, if convoluted, lit-
erature on imagination, had defined imagination in terms of a singular di-
mension. His analysis severed imagination from imagery and visualization;
imagination referred primarily to the capacity of persons to imagine. For
White (1990), ‘‘to imagine something is to think of it as possibly being so’’
(p. 184). White’s analysis became the point of departure for my study.
Drawing upon these considerations, the initial framework postulated a no-
tion of mathematical imagination as a form of mediated action, the gen-
eration of possibilities, mainly in terms of a singular, conceptual dimension;
the development of mathematical imagination was assumed to be mainly a
processual function of dialogicality.
These assumptions were challenged during the course of the study. The
data generated, in particular the sets of data generated through the different,
disruptive approach, provided a collection of data that threw up findings
that undermined the central assumptions of the skeletal framework. It was
necessary not only to revise the initial framework in the sense of fortifying it,
but drastic changes were required to take account of the unexpected findings
that were building up, indicating the necessity of rethinking the basis of the
framework. The hyper-spiral path – of continuous cycles of rebuilding of the
framework, on layer upon layer of cycles of analysis, interpretation, re-
framing after wave upon wave of data that were contrary to my initial
expectations – brought forth a framework that bears scant resemblance to
that with which the study began.
However, perhaps because the data generated from the different video
ethnographic approach adopted in this study was so surprising, to me as to
others, it has not been easily accepted. Different approaches generate
different data. In fairness to these data, frameworks must be reconceptu-
alized. I had anticipated examining in greater depth the processes by
which dialogicality shaped the genesis and/or the development of mathe-
matical imagination. But the data did not support the centrality of dialog-
icality for the study. I had anticipated examining the nature of the
conceptual basis of mathematical imagination and the links of this basis
with some particular mediational means. Yet because the approach
breached the safe distance of prevailing video research, it gained access to
the very spaces of origination of pupil action and generated sets of data that,
upon analysis, could not offer interpretations and findings in support of the
central assumptions of my study. The approach offered a window into the
constraints of particular spaces wherein actors struggle to make sense of the
world in ways that their customary communities have not accustomed them
to. It made possible to trace the variety of ways that pupils sought to
mathematize situations that were initially meaningless or bewildering, the
nature of the divergent ways pupils take as well as the ways such divergence
is brought about. The resultant analysis made clear that for pupils to
make sense of situations in innovative ways, in their struggle to become
fledgling mathematicians, they had to co-construct an alternative space of
mediated action. That involved resistance and disruptions of ‘‘customary
space’’. It called forth the ‘‘creative agency’’ of the pupils in collaborative
acts of co-constructing a generative space that affords them the possibility
of mathematical imagination. The pupils produced inventive mathe-
matical subsystems of their own. The data that was generated using
the different video ethnographic approach made manifest mathematical
imagination in play.
However, data from unfamiliar approaches are themselves unfamiliar and
also require immersion and multiple cycles of viewings and analysis and
interrogation and revisions and reconceptualizations of the ways we make
sense of phenomena. Such data are not easily acceptable, much less inter-
pretations of such data and lesser still frameworks based on interpretations
of such data. It is unheard of, at least at present, to confer on ordinary
pupils, struggling to make sense of mathematics, with fancy notions such as
creative agency and creative intentionality. The framework that will be in
my forthcoming report will discuss four dimensions. It is worth mentioning
that pupils’ works were the focus of early conference presentations of the
study (Aralas, 2003, 2004).
That is, a research study that adopts a different video ethnographic ap-
proach in pursuit of particular research objectives will generate different
kinds of data that lead to the production of different accounts and models.
Such differences make for results that may be so different or contrary so as
to be unacceptable. It is instructive to discuss a presentation of some of the
data generated in the study at a BSRLM (British Society for Research into
Learning Mathematics) conference in 2003. Some samples of pupils’ work
which, on analysis, were inventive and formed coherent (if unusual) sub-
systems, were displayed for discussion. The mathematics educators there
opined that the pupils’ solutions were incorrect and that this indicated a lack
of mathematical ability on the part of the pupils who produced such work.
182 DALIA ARALAS
The educators made references to literature on mathematical ability. These

responses are not unrepresentative of prevailing views on pupil learning.
Pupils have been evaluated for the longest time in terms of ‘mathematical
ability’; their work has been judged in terms of correct or incorrect math-
ematical reasoning. This state of affairs is explicable in view of some con-
siderations. One significant consideration is that the prevailing perspectives
of learning are dominated by two metaphors: first, ‘‘acquisition’’ and sec-
ond, ‘‘participation’’ (Sfard, 1998). That is, the first prevalent view stresses
the acquisition of correct knowledge in correct ways by pupils, whether
these ways be algorithmic or dialectical reasoning. The second prevalent
view emphasizes the importance of ‘legitimate peripheral participation’ of
pupils in communities of practice. It may be that research that offers data or
analysis that are unfamiliar or differ markedly from expectations based on
prevalent views are not acceptable. The notion that pupils may themselves
act with creative agency to generate subsystems that may be genuine forms
of knowledge, mathematical poesies, albeit idiosyncratic or quirky ones, is
an unfamiliar notion.
In fairness, the significance of creative agency on the part of pupils is
not remarkably easy to infer. Conceivably a charge could be levelled against
the notion of creative agency in that it may be as much a construct as it is
a real attribute of pupils. Certainly, creative agency is not a ‘solid’ attribute
of pupils as the colour of their hair, nor is it as easily inferable as fami-
liar concepts such as dispositions or habits or correct/incorrect ways of
‘doing’ mathematics or other sorts of learning. However, the authenticity or
usefulness of concepts such as creative agency lies in their explanatory
effectiveness (as a part of a framework that makes explicable pupil action
and furthers understanding of pupil ways of grappling with the world), not
in the ‘solidity’ or ‘visibility’ of such concepts. Different perspectives, the
result of different sorts of research using different approaches, make ‘visible’
(and if prevalent, make familiar) different notions, for they engage with
the world in different ways. The important point is that creative agency is
a notion that is abstract. It is ephemeral in nature. Its ephemerality is a
source of difficulty for the purposes of research. It was however, abstrac-
ted in the course of the research undertaken, significantly at the stage of
data analysis from the data generated. That it was possible to abstract the
notion despite its ephemerality derives from the advantageous resources
(a collection of different sets of data) that arose from the use of a different
video ethnographic approach that broke away from conventional research
practice.
PART V
The research study would not have been possible to conduct in its com-
plementary wholeness if the researcher had not availed herself of the differ-
ent video ethnographic approach described above. It would not have been
possible to undertake an effective investigation of the nature of mathemat-
ical imagination with the restrictions imposed by conventional video-me-
diated research to conduct observations of naturally occurring phenomena
in normal contexts in unobtrusive, non-disruptive ways.
The chapter has discussed the way that a different video ethnographic
approach, as described, made possible a different sort of research that re-
quired answers to specific research questions that had not been the subject of
research before. The research made possible the generation of different data
with consequences for the generation of deeper insights. These were that
there is an interplay of agency and a diversity of ways and means of grap-
pling with, enacting and establishing the world (comprising multiple realms),
within a complex of cultural spaces that are infused with socio-historical
contingencies, with consequences for analytical framing and reconfiguration.
For these insights, which led to the development of theoretical framework
that is, in turn, deemed useful in the ways we may understand the phenom-
enon of mathematical imagination in play, the research was worthwhile. It
was made possible by extending video ethnographic approaches.
To conclude, extending video approaches may undermine prevailing
views but the consequent production of alternative accounts and frame-
works of phenomena under study do make worthwhile contribution to the
advancement of ethnographic knowledge.
ACKNOWLEDGMENT
I am deeply grateful to Professor Geoffrey Walford for his critical com-
ments on drafts of this chapter. I owe him a tremendous debt of gratitude
for his supervision of the research study on which this chapter is based.
REFERENCES
Aralas, D. (2003). Mathematics and rules: pupil constructed rules. Unpublished, Conference of
British Society for Research into Learning Mathematics (BSRLM), 6–7 June, University
of Oxford.
184 DALIA ARALAS
Aralas, D. (2004). Investigating mathematical imagination. Poster 45 and Round Table Dis-
cussion at the Tenth International Congress on Mathematical Education (ICME-10),
4–11 July, DTU Technical University of Denmark, Copenhagen.
Bakhtin, M. M. (1981). The dialogic imagination: Four essays. In: M. Holquist (Ed.), C.
Emerson and M. Holquist (Trans.). Austin: University of Texas Press.
Banks, M. (2001). Visual methods in social research. London: Sage.
Chittick, W. C. (1989). The Sufi path of knowledge: Ibn al-Arabi’s metaphysics of imagination.
Albany, NY: State University of New York Press.
Coates, P. (2002). Ibn ‘Arabi and modern thought. Oxford: Anqa.
Ernest, P. (1998). Social constructivism as a philosophy of mathematics. Albany, NY: State
University of New York Press.
Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Hodkinson, P. (2001). The contested field of educational research: The new research orthodoxy
and the limits of objectivity. Paper presented to the British Educational Research As-
sociation Annual Conference, University of Leeds, 13–15 September 2001. http://
www.leeds.ac.uk/educol/documents/00001875.doc.
Kearney, R. (1998). Poetics of imagining: Modern to postmodern. Edinburgh: Edinburgh Uni-
versity Press.
Kelly, A. E., & Lesh, R. A. (2000). Handbook of research design in mathematics and science
education. Mahwah, NJ: Lawrence Erlbaum.
Lee, R. M. (2000). Unobtrusive methods in social research. Buckingham: Open University Press.
Pirie, S. E. B. (1996). Classroom video-recording: When, why and how does it offer a valuable
data source for qualitative research? paper presented at the Annual Meeting of the North
American Chapter of the International Group for the Psychology of Mathematics Ed-
ucation. Panama City, Florida. 14 October. ERIC Reports: ED401128. Washington DC:
U.S. Department of Education, Office of Educational Research and Improvement.
Powell, A. B., Francisco, J. M., & Maher, C. A. (2003). An analytical model for studying the
development of learners mathematical ideas and reasoning using videotape data. Journal
of Mathematical Behaviour, 22, 405–435.
Sfard, A. (1998). On two metaphors for learning and on the dangers of choosing just one.
Educational Researcher, 27(2), 4–13.
Walford, G. (2001). Doing qualitative educational research: A personal guide to the research
process. London: Continuum.
Walford, G. (2002). Introduction: Debates and Developments. In: G. Walford (Ed.), Debates
and developments in ethnographic methodology. Studies in Educational Ethnography
(Vol. 6.). Oxford: Elsevier.
Wertsch, J. V. (1998). Mind as action. New York: Oxford University Press.
White, A. R. (1990). The language of imagination. Oxford: Blackwell.
Wittgenstein, L. W. (2001). G. E. M. Anscombe (Trans.), Philosophical investigations [German
text, with a revised English translation.] (3rd ed.). Oxford: Blackwell.
THE OCEAN MERGES INTO THE
DROP: UNEARTHING THE
GROUND RULES FOR THE SOCIAL
CONSTRUCTION OF PUPIL
DIVERSITY
Gretar L. Marinósson
All know that the drop merges into the ocean but few know that the ocean merges into
the drop. (Kabir 1450–1518 cited by (Baldock, 1995)
ETHNOGRAPHY OF ONE SCHOOL
Working as a teacher and an educational psychologist in England and in

Iceland for a number of years, I slowly realised that I had to abandon much
of what I had learned during my training in psychology and research meth-
odology. The social world of school children was simply far more complex
and uncharted for the known theories and methods to suffice. I had to
function at a different level to that of a traditional scientist or a clinical
psychologist if only to be accepted by the children I worked with let alone
gain answers to the type of questions I had been asking for some time
(Marinósson, 1998): Why are so many children and youths disaffected by
school since the purpose of education is ostensibly to serve their needs? How
do children who find their life in school difficult perceive the school and

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12011-2
185
186 GRETAR L. MARINÓSSON
other people there? How do the adults at school understand these pupils and
their work? What shapes their reactions in the school context?
It became gradually clear to me that through listening, and above all
through caring, I obtained a totally different kind of information and at-
tained a radically different kind of understanding of the meanings and in-
terplay of complex factors to that obtained through experimental or other
methods based on logical positivistic epistemology. Factors such as self-
image, social role, status, power, acceptance, social relationships and hier-
archy emerged as crucial elements in the lives of the pupils, teachers and
families that I talked to. The role of meaning and values in child develop-
ment and upbringing became clearer. And oddly enough, in that confusing,
context-bound social world, discernable patterns emerged, more subtle and
complex than in the experimental world, but equally valid, it seemed, and
even transferable to other contexts. I was looking for a method that would
help me seek explanations of how these phenomena inter-linked; ethnog-
raphy seemed promising.
In the study on which this paper is based my purpose was to analyse the
work of a mainstream compulsory school as it bore upon its response to
pupil diversity (Marinósson, 2002). I wanted to see how things were and ask
why rather than look at how things could be and ask why not; I therefore
intended the study to be descriptive and explanatory rather than normative.
I attempted to create a comprehensive picture of a school and its relation-
ship to its community, focusing on its handling of pupils’ diverse learning
needs. This involved examining the dynamic process of definition and pro-
vision for individual needs as this was dealt with by the teachers, the school
management, specialised services, pupils and their parents.
Common knowledge of schools tends to be at the level of anecdotes; their
organisation being complex and little understood, as are the processes that
define the construction of pupil diversity. Moreover, many practices become
quickly routinised. Examples are the categorisation of pupils, the use of
space for teaching, the division of labour among the staff and methods of
instruction and examination. I wanted, however, to go beyond description
of such routines and analyse the underlying assumptions. Mercer and Ed-
wards (1981) called these kinds of assumptions ‘‘educational ground-rules’’.
I realised that as it was the pupils most difficult to teach who threw into
relief the structure and morality of the school some of its ground rules could
be studied by looking at the way it attempted to meet the needs of its
peripheral members. They test the function of individual parts, the link
between them and the organisation’s relations with wider bureaucratic and
social systems. Thus, a diversity of needs is the moment of truth for any
The Ocean Merges into the Drop 187
school and its efforts to meet them an optimal way to study its structure and
function. I was thus
looking for evidence of the ‘taken-for-granted’ assumptions which, although rarely or
never explicitly invoked or discussed by participants, nevertheless define the process of
‘doing education’ (Mercer, 1991).
I wanted to attempt to unveil the educational ground rules, formulate and

present them for discussion. This meant looking for basic (often unspoken)
assumptions that are the central elements of education as a social construc-
tion or influence this construction fundamentally. Examples are that chil-
dren’s learning capacity is seen to be largely inherited and to that extent
immutable; that parents ability in handling their children’s upbringing is at
best questionable; that the school has the role of a moral agent in this
upbringing and that there is a need for tight control of pupil groups par-
ticularly to prevent disturbance as learning is considered to take place in
quiet surroundings.
As the task of ethnography is to determine how cultural factors and
human agency interact with each other to co-determine social life, it is
singularly well suited to such a study. Ethnography is literally a description
of a (foreign) culture; thus, this is a study of the (foreign) culture of a school.
There is only one way to understand another culture: to live with it. To take up residence
in it, ask to be tolerated as a guest, to learn the language. Then there is hope that
understanding will come sooner or later. It will, however, always be a mute kind of
understanding. For, as soon as one understands the unknown, the need for an expla-
nation disappears. To explain a phenomenon is to distance oneself from it. (Hoeg, 1992)
This is an example of a traditional kind of ethnography where a researcher

spends some considerable time in the field gathering data about real people
in order to understand and explain the social processes in the school as they
relate to the phenomenon under study (Wolcott, 1999). To paraphrase
Woods (1996), the research is contextualised within settings in the school
and the researcher’s self is inextricably bound up with the research (p. 51).
Thus the researcher is both the major research instrument and the one who
monitors how this instrument is being used. He must therefore be careful to
evaluate the influence of his own values and theoretical organisation on the
interpretation of the data.
Accounts can be read backwards to uncover and explain the consciousness, culture and
theoretical organisation of the observer. (Willis, 1984, p. 90)
Thus, a classroom observation says as much about the observer as it does

about the pupils and the teacher observed, and what they are doing in that
classroom. As Delamont (1992) has pointed out, data collection is never

atheoretical.
The study was based on the critical realist view shared by Bashkar, Harré
and others (Danermark, Ekström, Jakobsen, & Karlsson, 2002) that ‘‘there
is a world of events out there that is observable and independent of human
consciousness [and that] knowledge of that world is socially constructed’’
(Denzin & Lincoln, 2005, p. 13). Bhaskar thus holds that reality is onto-
logically stable but epistemologically unstable; although reality is in itself an
intransitive phenomenon, our knowledge of it is subject to social and po-
litical influences and therefore transitive (Bhaskar, 1989). Critical realists
believe that scientific work must go beyond statements of regularity to
analysis of the mechanisms, processes and structures that account for the
patterns that are observed (Denzin & Lincoln, 2005).
In what follows the methods of producing data and explaining the un-
derlying social processes through a grounded theory procedure are described
and discussed. The argument is that it is possible to unearth the ground rules
of education through an ethnographic lens using methods of data collection
based on an assumption of a reciprocal relationship between the whole and
its parts. These ground rules govern the behaviour of individuals in the same
way as the institution moulds people’s thinking (Douglas, 1986). Con-
versely, the behaviour of individuals bears witness to these ground rules and
the institution’s structure and classification activity. Thus the interaction of
the whole and its parts is a key to unearthing the basic assumptions that
determine to a large extent the response of the school to the diversity of its
pupil group. Accordingly one may through close scrutiny of, say, observa-
tion data related to micro-events or interviews that revolve around daily
experiences in the classroom pick up evidence of fundamental social con-
structions that are based on value-laden assumptions that may be termed
the ground rules of responding to pupil diversity.
THE CASE OF MOSSY MOUNT SCHOOL
Choosing a case is critical to understanding the phenomenon under study.

Hamel (1991) maintains that case studies must ‘‘locate the global in the
local’’ thus making the careful selection of the research site the most critical
decision in the analytic process. Stake suggests that we select a case from
those instances, from which we think we can learn the most. This may or
may not be a case typical of such cases in general. Attributes, such as ease of
entry and likelihood of occurrence of the phenomenon, make the case the
one most likely to be learned from (Stake, 1994, p. 243). This suggests that
selecting a case is essentially a matter of sampling.
Case study is not a methodological choice but a choice of object to be studied. [y] As a
form of research, case study is defined by interest in individual cases, not by the methods
of inquiry used. [y] We could study it in many ways. [y] The name case study is
emphasized by some of us because it draws attention to the question of what specifically
can be learned from the single case. (Stake, 1994, p. 236)
Choosing one school for the study has a corollary in the archetypal village
chosen as a unit for case studies in anthropology. It is of a convenient size,
thus giving a comprehensive view, and it is assumed that it illustrates the
particular traits of the culture that the research is attempting to describe or
explain (cf. Clark, Dyson, & Millward, 1995). As Spindler and Spindler
argue in their foreword to Wolcott’s book on the Man in the Principal’s
Office
y it is only through a case study in depth of this kind that the dynamics of the system
and the interactions of people within it can be seen in their functional totality. (Wolcott,
1973)
Thus, I use this single school to present important aspects of mainstream

compulsory schools in Iceland, saying to my readers:
Here is the underlying structure, the skeleton of the case, unencrusted by the idiosyn-
crasies of specific instances, and here is a model of the dynamic that sets its parts in
motion. (Weiss, 1994, p. 173)
There are, however, certain problems associated with the metaphor of a

village, for although the villages studied may have been reasonably homog-
enous examples of the cultures they operated within, this is not the case with
schools today (or indeed villages). Societies are complex and changing and
individual schools have their own cultural attributes, which are different
from other schools.
I approached Mossy Mount School from a wide perspective, first looking
at the local community and the school catchment area; then at the school as
a whole, and finally activities and events inside the school, only to return to
the larger backdrop later. The district where the school is located is divided
into six school catchment areas, some of the schools still being built. The
area that Mossy Mount serves is a housing estate that has been built up in
the last 25 years. The houses are mixed, terraced houses and blocks of flats
as well as some bungalows. Most of the houses are privately owned but
there are also some flats bought with social support for those on limited
income. The district is overwhelmingly populated by people of lower income

and lower social status compared to the municipality as a whole. Most of
the parents work long hours outside the home, leaving the children to child
minders or to fend for themselves outside school hours.
The school was chosen because it was a fully comprehensive compulsory
school (having 600–700 pupils of 6–16 year age groupand 70 teaching staff);
had experience of working with pupils with disabilities or special needs and
an organised response to these. Although most Icelandic schools are much
smaller (300–400) and not situated in an urban area, a school like Mossy
Mount sheds a better light on a school’s response to diversity than many
others, as it is a complex organisation, containing different levels of inter-
action. The diversity of pupils’ needs is large and so are the pressures and
attempts, from teachers, pupils, parents and specialists, to respond in some
way. The main disadvantage in selecting such a large school is its complexity
and the difficulties in description and analysis that this entails. In a case of
this type the sampling is moreover bound to be nested (Miles & Huberman,
1994, p. 27), in that the individual pupils or teachers sampled are a part of
the pedagogic events or activities, which again are a part of the school. This
again reflected the assumption of a reciprocal relationship between the
whole and its parts.
STUDYING THE SOCIAL CONSTRUCTION OF PUPIL

DIVERSITY
This intensive, qualitative case study of one school was framed within an
interpretive paradigm (Bogdan & Biklen, 1992; Lincoln & Guba, 1985) in-
formed by social constructionism (Berger & Luckmann, 1966; Burr, 1995;
Gergen, 1985, 1994, 1999; Nightingale & Cromby, 1999) and symbolic in-
teractionism (Blumer, 1969; Hargreaves, 1978; Rock, 1979; Woods, 1996).
One of my assumptions was that the construction and reconstruction of
pupils’ diverse learning needs and the school’s response to these occur
through a process of negotiation where symbols play a central role. Ac-
cording to this perspective our criteria for identifying behaviour or events
are highly circumscribed by culture, history or the social context, or are
altogether non-existent. Thus, we are invited to consider critically the social
origins of our taken for granted assumptions about the ‘real’ world and
about knowledge. I assume, as mentioned above, that there may be ‘‘reality
out there’’ but that my knowledge of it is partial, fleeting, socially produced
and temporally situated.
I assumed that participants constructed and reconstructed meaningful con-

cepts about children’s behaviour, abilities, culture, gender, etc. and possible
responses by the school to these individual qualities. Pupils with disabilities
are identified within a process of negotiation from the initial hunch or a feeling
developed by a teacher in the course of her interaction with her pupils to the
formal diagnosis by an outside expert. This is an emergent process, which the
pupils and the parents enter into in response to expectations, which they
perceive to come from the school and the education system.
Symbolic interactionism deals with small-scale events and psychological
processes. Fundamental to it is the view that people construct their own and
others’ identities through social encounters. Social constructionists take a
wider view by claiming (Berger & Luckmann, 1966) that all social phenom-
ena, including knowledge of ‘reality’, are constructed through social inter-
change. This occurs, they claim, through externalisation, ‘objectivation’ and
internalisation. An example of this process is provided by the way categories
of disability have developed. Sleeter, in discussing the emergence of the
category of ‘learning disabilities’ in the US, notes:
During the late 1960s and early 1970s, minority groups pressured schools to discard the
notion of cultural deprivation and stop classifying disproportionate numbers of minority
children as mentally retarded. In 1973, the category of mental retardation was redefined,
lowering the maximum IQ score from one standard deviation below the mean to two
(Grossman, 1973), which dissolved the category of slow learner. [y] Instead, many
students who previously were or would have been classified as retarded, slow or cul-
turally deprived were now classified as learning disabled (Sleeter, 1986, p. 52).
Thus, new ideas about the nature of learning difficulties were externalised
and ‘objectified’ through changes in criteria for the categorisation of pupils.
Then, the new criteria were internalised or accepted as ‘real’ by later gen-
erations of parents and teachers. A similar story can be told of the social
construction of other, more recent categories, such as attention deficit, hy-
peractivity disorder (ADHD). Thus, special educational needs are not dis-
covered but defined. Diagnostic practices produce their clients rather than
uncover their disorders. Once pupils belong to a client group, special ed-
ucational provision can be established. There would not, for example, have
been seen to be much justification for building institutions for ‘idiots’ had
they been viewed as happy individuals (Borjeson, 1997, p. 17, his emphasis).
Similarly, our understanding of the nature of difficult behaviour in children,
which has changed in the last 100 years from ‘immoral’ to ‘disordered’,
influences our reactions to individuals (Reid, Maag, & Vasa, 1993). As
sufferers of a biological aberration they are now seen to be deserving of
expert attention.
Learning disabilities exist, of course, but they are specific to individuals

more than to medical or psychological conditions, and they shade into the
general population (Goodey, 1999). Pupils’ behaviour is influenced by such
factors as the school as an organisation, the meanings individual actors read
into the relevant social context as well as personal genetic makeup. Rec-
ognising that social constructionism and symbolic interactionism cannot be
expected to account for all behaviour or all situations observed, they nev-
ertheless deal with how behaviour is constructed at the level of a ‘‘theory of
mind’’, taking into account what the participants know about the thinking
of those in question, about the social circumstances, about biological de-
terminants, etc.
DOING ETHNOGRAPHY
Data collection and analysis started the moment I gained the first shred of
information about the school. This was a classical type of ethnography
where one investigator stations himself in the middle of an action zone and,
with himself as the main instrument of data collection, gathers information
through interaction with the locals that may help him understand what is
going on and why. After gaining access I adopted the role of a researcher
with a low profile. I tried to present myself openly and honestly but at the
same time showing an interest in the people. I hoped that they would feel
that my presence as a friendly human being had a benign effect and thus
counterbalanced their natural apprehension, and that my genuine affection
for children and respect for teachers and their job would be felt sufficiently
so that my potential informants would trust me to treat their lives in con-
fidence. Naturalistic studies that rely on observation of people and personal
informants must rely on trust for gaining access to the field. Johnson found
the existing theories of trust unsatisfactory: the exchange theory, which
maintains that trust results from a quid pro quo relationship; the individual
morality theory, which asserts that trust comes from building up a ‘good
guy’ image; the adoption of a membership theory, which claims that trust is
invested in persons willing to commit themselves to a group; and the psy-
chological need theory, which asserts that trust is based on a need fulfilment
by the respondent. Instead, Johnson sees the relationship of trust as a de-
velopmental process, to some extent biographically specific, which means
that the enquirer has to build and attend to the development of trust with
each respondent continuously to prevent it from being destroyed (Johnson,
1975) referred to by Lincoln and Guba (1985, p. 256). The approach I
adopted recognises the importance of the developmental and personal na-

ture of trust and its maintenance. It also underlines the crucial role of the
personal image portrayed by the researcher both before and after personal
trust is established.
The study spanned over three years in the life of the school, the fieldwork
lasting for over two years of that time, on average 1–2 days a week. I
started by making myself acquainted with the school as a whole, its col-
laborators and district services. This involved a great deal of observation,
both general on corridors, in staff room and in classrooms as well as fo-
cused in one class. It also meant interviewing key informants and observing
crucial meetings and a variety of lessons. Gradually, I focused more on
contexts including pupils considered to have disabilities or special needs.
Finally, I concentrated on interviewing pupils, their parents, experts and
administrators outside the school as well as informants within it. I collected
documentary material throughout the study. Thus data collection can be
seen to have shifted several times and oscillated between concentrating on
the whole to focusing on details and back again. Similarly the use of
methods alternated from a mixture of methods to an emphasis on obser-
vation and then to interviews. Thus, despite the assumption of a relation-
ship between the whole case and its individual parts it was deemed
necessary to check and recheck data from both observations and interviews
through triangulation. This study was characterised by a process of explo-
ration where theory development was postponed until it could be grounded
in the data.
THE AMBIVALENCE OF THE OBSERVER

Observation involves a complex array of processes. Sanger points out, for
instance, that there is a world of difference between observation and seeing.
He describes the latter as creating order out of the chaos of signals reaching
the brain, gleaning information from the environment by placing signifi-
cance on some things and not others, making meaning out of a situation,
even noticing the absence of things, being aware of one’s in-built biases and
taking steps to counteract them (Sanger, 1996, pp. 1–8). An observation, the
result of looking, seeing and interpretation is in turn based on the observer’s
assumptions, his cultural knowledge (Spradley, 1980). In comparison with
interviews and document analysis, observations have distinct characteristics
as a data collection method: they are one way, and therefore void of the
negotiation with another about the choice of information recorded; they are
fleeting, the observer being unable to catch everything that is happening and
thus must choose without a moment’s hesitation one aspect in place of
another; and the data they produce are the observer’s own interpretations of
a scene. Interpretation thus enters into the observation earlier than it does in
the interview or document analysis.
Some researchers equate ethnography with participant observation as
this method is a necessary (although not sufficient) ingredient of the eth-
nographic approach. Others argue that there is no such thing as a complete
participation or a complete observer: in research complete participation
would prevent the research act and no one can avoid taking part to some
extent. Although the prospect of complete participation may be alluring to
would-be researchers, considering the anonymity and protection it affords
as well as minimising reactivity on behalf of the participants, the advan-
tages are, as I have discovered, heavily outweighed by the restrictions it
places on mobility, access to various parts of the settings under study and
relationships with at least some of the actors. This, incidentally also applies
to the role of a complete observer. It can easily work against the purposes
of the research to be an obvious outsider (Hill, 1995). On the other hand, it
can be a considerable advantage over being seen to be one of the staff
(Hargreaves, 1967). The trick is to be seen to be part of the scene while
maintaining a certain distance from it. This can, however, prove to be
immensely challenging (Hammersley & Atkinson, 1983, p. 93). When ob-
serving in classrooms I usually sat at a table like the pupils’, either on my
own or beside a pupil, and made notes in my little black book. I usually
greeted the class upon entering, told them my name, and that I was con-
ducting research in the school and asked whether I could stay with them for
a while. Being in doubt as to whether this was at all within their authority
to decide they usually said nothing, sometimes a vague yes or a single no
could be heard. Often, some pupils came up to me at the end of the lesson
and asked what I was doing or what I was writing. I told them honestly that
I wrote down everything that I saw and heard. They asked no more but I
understood that some of them thought this was yet another assessment,
perhaps because they had not behaved well enough. I tried to be aware of
the effect I had on my surroundings but did not make a habit of asking the
teachers or pupils about the influence of my presence and participation in
class. As Sanger (1996) points out the mechanics of research of this type
remain firmly in the ethnographer’s possession despite his attempts at re-
maining detached (p. 16).
I am not at all certain that open observation of this kind in classrooms is
ethically acceptable. It raises, among other things, the question of whether a
classroom is a private or a public place. In a traditional school, like Mossy

Mount, with one teacher per class in one classroom where the door is usu-
ally closed, one gets the feeling that one is entering a private place of a kind,
like somebody’s office. One needs permission or a good reason to enter. In
the words of Eisner (1998): Classroom is like a ladies’ boudoir. This feeling
is strengthened where the classroom door is locked from the inside (to
prevent disturbance by pupils from other classes), as was often the case in
this school. Perhaps the classroom fits the third form of territory in the
public realm suggested by Lofland (1989), the parochial form, characterised
by a sense of commonality among the participants who are involved in
interpersonal networks that are located within ‘‘communities’’. This is not
so in classes which take place in open-plan schools. I have frequently walked
through such classes without so much as being noticed. The closed-circle
feeling of the classroom is thus defined by the physical structure and how it
is perceived, first and foremost by the teacher. The acceptability of the
observation depends to a large extent upon how (un)obtrusive it is to the
members of that community. Signs like the observer’s glance and his stance;
his reactions to what is happening and his method of recording are signals to
the pupils about whether there is reason to react to the observer’s presence.
With my demeanour and explanation, I tried to indicate to the pupils that I
was doing the most natural thing imaginable, making notes of what was
happening in their classroom. As this was not at all natural to them, they
probably ascribed it to the vagaries of adults.
WALKING A TIGHTROPE BY QUALITATIVE

INTERVIEWING
The type of interviewing employed in this study has most frequently been
named in-depth interviewing. I am not content with that term as research
interviews of this kind can be of any depth. For lack of a better term I prefer
qualitative interviewing (Weiss, 1994, appendix). The aim of the interview is,
in Sanger’s (1996) words
creating the conditions in which the interviewee says what he/she means, means what he/
she says, says what he/she thinks, and thinks about what he/she says. p. 66)
The challenges presented by unstructured interviews are three, according to

Sanger. First, how to achieve a penetrative conversation with a relative
stranger in a short time; second, how to strike a balance between one’s desire
to know and the respondent’s right to some measure of protection; and

third, there is the problem of ‘‘what claims to truth are associated with the
results’’ (Sanger, 1996, p. 65). The first problem is the one of overcoming
obstacles to being accepted in a strange culture by presenting a sympathetic
mien, the second relates to confidentiality and the third to validity. Qual-
itative interviews do raise particular issues compared to other methods of
data collection. They are a testing ground for the dilemma experienced by
the ethnographic investigator as s/he tries on the one hand to see a situation
through the eyes of a stranger and on the other attempts to keep in abeyance
‘‘precisely those assumptions that must be taken for granted’’ in the inter-
view (Hammersley & Atkinson, 1983, p. 93). Keeping this balance is
important as the way the interviewee sees the researcher, as a stranger or a
friend, determines to a large extent the quality of the material he imparts to
him. I think, I managed to assure my teacher informants that I understood
their professional assumptions and their daily preoccupations, but that I still
wanted to know in greater detail of their reflections on their job, the school
and the pupils. The same can be said about the people outside the school
whom I interviewed, the parents, the professionals, the administrators and
the politicians; this contact was generally a pleasant and meaningful expe-
rience. The pupils were a little more difficult: despite my attempts at
explaining my purpose (I am doing a research study on the school, trying to
understand how the people here think about pupils who have disability
or special needs), they did not understand my point well enough. They
probably saw me as an outsider, asking odd questions, which did not con-
cern them. This is where the greatest distance lay between my informants
and myself. Nevertheless they generally did as well as could be expected and
enjoyed the attention and the chance of missing a lesson.
During interviews I attempted to get to the heart of my informant’s
meaning, ‘‘capturing the ‘essence’ of the account’’ (Miles & Huberman,
1994, p. 8). There was a constant search going on for an understanding of
why things were as they were; a quest for some meaningful relationships. I
used whatever threw light on the research questions, be it something I heard,
saw, read or simply perceived through insight as a result of being immersed
in the school for so long, assuming that the drop would merge into the
ocean. The focus was on the way my informants understood and accounted
for the phenomena under study and the situation as it was how it had come
to be and their attitude to it. Despite my preoccupation with using each
interview to the full for data collection, I tried also to make it into a con-
versation where I joined the other person in her or his thoughts and stated
my own position on the matter under discussion.
I did not manage very well to conceal my own agenda when talking to my
informants and in writing about my observations. This agenda consisted of
urging mainstream schools to try harder in educating all pupils irrespective
of learning needs, disability or other differences. I assumed a pluralistic
community outside the school and that this mainstream school should host
such a community. I wanted the school to take an active stand on social
justice and equality among its pupil group. However, the immediate purpose
of the study was to describe and understand the way the school reacted to
the diversity of pupils’ learning needs at the time of the study. The crucial
question was what facilitated and what restricted the school in taking up the
challenges of inclusion?
USING WHATEVER TEXT HAS ALREADY BEEN

CREATED
I collected a number of documents on paper and websites about individual

pupils, the school, the school district as well as laws, regulations and
National Curriculum Guides. I used this material in three ways: to provide
clues about dominant themes in the operation of the school; as starting
points for discussion of issues with participants and as instruments for tri-
angulation in analysing data from observations and interviews. Most of the
documentary data was thus dealt with in interviews during fieldwork.
MAKING SENSE OF THE DATA
I utilised grounded theory, or the constant comparison method, as a me-

thodical, systematic process of data analysis and viewed it as such rather
than as a methodological framework. Glaser and Strauss (1967) describe
this as a method for generating theory from data, and argue that a theory
thus produced would be more suited to its purposes than theory generated
by logical deduction from a priori assumptions. It would ‘fit’ the data in the
sense that the categories produced would be readily applicable to and in-
dicated by the data and it would ‘work’ in the sense that it would explain the
behaviour under study. The process of grounded theory study is well de-
scribed by Bartlett and Payne (cited by Scott, 1996). It involves basically 10
steps, which may be viewed as a heuristic device rather than a cook-book
prescription. The theory develops gradually, from the first attempts at cat-
egorisation to the step of theoretical integration; where it should be com-
plete, only to be tested as a whole against the data.
The observation data, the documentary evidence and some of the inter-
views formed a base for a detailed description of the school. In addition a
great deal of the interviews was used for the purpose of attempting to glean
answers to some of the major research questions. This was done through
interpretation of the meanings of participants in combination with inter-
pretation of observations and analysis of documentary evidence. Transcripts
of interviews were analysed in more than one way. First, all elements of each
interview were coded and sections coded in a similar way were interpreted as
a whole. Second, each salient part of an interview (paragraph or longer) was
paraphrased in a few words in the margin and labelled with a broad cat-
egory of an emergent theme. These themes were then listed on the front
cover of the transcript together with the major thread running through the
interview, if there was such a thread. The themes formed emerging hypoth-
eses, which I then tested in further interviews or observations. The same
applied to field notes. Thus the process is a bottom-up procedure, where
larger, more complex, categories are based on smaller and simpler ones,
again reflecting the desert in the gain of sand.
Gradually, as my overview of the school and its form of operation
developed, some themes started emerging. These were at first too numerous
to keep control of but I wrote them down in sketchy format and gradually
grouped them together. The themes formed emerging hypotheses, which
I then tested in further interviews or observations. The same applied to field
notes. The approach can therefore be described as issue-focused analysis
(Stake, 1994, p. 239). Some concepts emerged early and disappeared
early, others remained longer, still others did not crystallise until later.
Thus local integration gradually led to inclusive integration (Weiss, 1994,
p. 160).
It is the researcher who decides at every point what is meaningful in the
data and how it can be used for constructing theory. He alone is responsible
and cannot blame the method if the results are not as he might wish. This is
indicated by Strauss and Corbin (1990) as they speak of the need for the
researcher to ‘‘see with analytic depth what is there’’ as he or she performs
the crucial task of coding (p. 76). Strauss and Corbin therefore seem to
assume that insight and/or a priori theoretical knowledge is an advantage to
the creation of a grounded theory. The theory is thus not entirely grounded
in the data but at least partly based on the researcher’s preconceptions. This
may mean, for instance, that the categories will not necessarily turn out to
be as discrete as Glaser and Strauss prescribe, but instead be loosely over-
lapping themes. This is how the method is used in this study: as a guideline
for the development of emergent themes.
UNEARTHING THE GROUND RULES: AN

EMERGENT THEORY
My emergent theory concerned itself with explaining why the school re-
sponded to its pupils as it did. This involved uncovering assumptions, the
ground rules, about the functioning of the school but it also sought an
understanding of their provenance. However, in trying to trace the origin of
perspectives or policies, attitudes or actions, it is important to recognise the
inherent obstacles. For one thing proving causation in the social realm is
complicated and often tenuous. For another, participants who act on their
ideas or express particular viewpoints may not be aware of the link between
their action and an existing theory; all those who accept a particular theory
are, for example, not necessarily in agreement about its content, purpose or
application. A theory explaining how a composite of social constructions,
such as a school’s response to pupil diversity, has developed must therefore
at best be tentative.
HOW DID THE SCHOOL RESPOND TO DIVERSITY?

The first crucial finding to emerge was a confirmation of the ‘foreshadowed
problem’ that pupils were categorised into two groups, the majority, and a
minority who were considered in need of additional resources for a variety
of reasons. Diversity was treated as a nuisance rather than strength in
management as well as in the composition of classes. The second finding was
something that I was not looking for and therefore came at first as a sur-
prise: the school laid particular emphasis on the control of behaviour. Thus
the school instituted mechanisms for the prevention and containment of
disruption at all levels from classrooms to management. Thus, although the
general ethos was one of the incorporation and inclusion (Reynolds &
Sullivan, 1979, p. 50), there were some situations where coercive exclusions
were considered necessary, at times for the common good. The third im-
portant finding was that the school operated an efficient system of man-
agement that offered those pupils, who were not able to adjust to the
mainstream curriculum a continuum of special needs provision within the
school although some pupils were not catered for, particularly those whose
behaviour caused concern. The fourth important finding was the promi-
nence the head teacher gave to raising standards in all aspects of school-
work. The final major finding was the surprising isolation of the school from
its local community, particularly the parents, despite its considerable
activity that aimed to establish and maintain links with local social agencies
and expert services. While the school maintained a collaborative network
with these outside agencies for the ostensible purpose of preventing sub-
stance abuse among its youth, it neglected its relationship with their parents,
who took a passive consumer role towards the school, treating it as a service
institution for their children. This appeared to be the result of a long-
standing unspoken policy that parents keep a certain distance from school-
work. Their primary role was to make sure that their children were properly
prepared for school and keep well informed about the schools’ decisions.
They should certainly not meddle too much in school business.
WHY DID THE SCHOOL RESPOND TO DIVERSITY

AS IT DID?
During the study the educational ground rules governing the function of
Mossy Mount School as an institution gradually emerged. With the help of
the literature I then attempted to explain the socio-historical context in
which these ground rules were formulated. These I considered the major
factors in the social process of constructing pupil diversity:
1. The school is a moral institution based on Christian values. The dominant
value system of the community is thus the guiding light for the aim of the
school to improve its pupils, morally, mentally and physically. The
school’s primary role of upholding these values in society explains its pre-
occupation with behaviour over and above learning. This means that pu-
pils are assessed on their behaviour as much as on their school attainments.
2. The school is a bureaucratic institution structured around principles of
efficacy and run according to set rules and routines. This explains the
school’s preoccupation with teaching hours and pupil/teacher ratio in
place of curricular matters particularly as regards special needs. Here
economic means take precedence over educational ends. The bureaucratic
structure of the school is strengthened by the adoption of the positivist
paradigm in its conception and dissemination of knowledge.
3. The school is a pedagogic institution, caring for and disciplining its pu-
pils. This is closely associated with the child-centred and tolerant attitude
of teachers on the one hand and the emphasis on the act of learning on
the other. The caring attitude is more prominent, reflecting the impor-
tance placed on well-being. It also means that teachers operate both as
proxy parents and as professionals.
4. The school is a professional institution run on principles of professional

knowledge. The role of the school as a socialising agent has in recent
years been played down. Instead, teaching the subjects, as defined by the
National Curriculum Guide, has been seen as the school’s major task,
based upon positivistic conceptions of truth, objectivity, reason, knowl-
edge and learning (Ajalnámskrá grunnskóla, 1999). Traditional methods
of grouping pupils, whole class teaching to the mean and the emphasis on
school subjects are indications of this perspective. This explains the tech-
nical approach adopted, the psycho-medical model governing the
school’s management of special educational needs. It explains why teach-
ers call in experts when pupils do not fit a pre-determined mould.
5. The school is under the strong influence of professional interests of the
teachers and the experts, who challenge the management’s endeavour to
maintain control of pupils. The teachers’ claim for control over peda-
gogic activity in the classroom is founded on their autonomous profes-
sional role and the power vested in their Trade Union as reflected in their
collective agreements on work conditions. The claim by outside services
for control over decisions about difficult pupils is defended on the basis of
their expert knowledge and their gatekeeper function in assessing indi-
vidual pupils for special provisions.
The difficulties inherent in changing schools in the direction of inclusion

while continuing to premise their work on the psycho-medical paradigm
(Dyson & Millward, 1997, p. 55) may be seen as an example of a clash of
values and power placing the school in a set of dilemmatic situations. If one
makes someone too special he may lose general rights and if one tries to play
down his or her disability, (s)he may lose the benefits of being recognised as
being disabled (the individuality–commonality dilemma). Similarly, placing a
pupil in a special unit may benefit him/her educationally but scar him/her
socially by stigmatisation (the placement dilemma). These dilemmas, which are
temporarily resolved but not necessarily solved, arise from trying to match the
full range of individual needs into a common system. They concern conflicting
perspectives and basic values underpinning the diverse purposes of the school,
which people are uncertain about how to reconcile or choose between.
In response to these dilemmas, if and when some limit of attainment or
behaviour is reached, individuals are excluded through the means of referral
to support services. This is usually done, in line with the dominant notions
of scientific knowledge, under the guise of acquiring or acting according to
expert advice, although in most cases separation and demarcation have
already taken place. Thus medical–psychological diagnosis and placement,
supported by bureaucratic arrangements, habitually become the final stages

of labelling and aggregation in a process of resolution of dilemmas that are
fundamentally of an ethical nature.
DOES THE OCEAN MERGE INTO THE DROP?
The basic assumption of the ethnographic approach to the study of a single

school used in this research is concisely stated in the epitaph from the Little
Book of Sufi Wisdom. The drop of rain becomes part of the ocean but
similarly the ocean becomes part of the drop. Thus the interplay between the
smaller elements of the social world of the school and the school as an
institution together with its related agencies is reciprocal. The concept of the
hermeneutic circle (Gadamer, 1979) is useful in this context. It involves the
continuous process of alternating between interpreting the meaning of each
part of the text (anything containing a message that can be read) and the
text as a whole. The assumption is that interpretation of the text as a whole
helps interpretation of the parts and vice versa. Shifting from foreground to
background and back again in this way certainly applies both to data col-
lection and analysis. The minutiae of social interaction observed, the finer
details of personal constructions heard, the particulars of discourse read in
written documents dealing with the identities of individuals and the school’s
response to these are all clues to the prevailing social constructions of pupil
diversity. In this sense keys to the understanding of the universe may often
be found in its microcosms (Hamel, 1991) the global in the particular, the
world in the grain of sand, provided that the case is carefully selected.
Institutions, conversely, guide perception and legitimate knowledge
through the use of symbols for classification (Douglas, 1986). The catego-
ries produced involve power relations and social consensus within the in-
stitution. Mary Douglas’ work is particularly relevant to the study of a
school as an institution and its production of special needs through its
classification systems.
A classification system is made for use. How classes are made depends entirely on the
classifiers’ interests. Straight away this approach links knowledge to action. (Douglas, 1999)
For Douglas, categorisation is closely related to reasoning. The first step

towards opening common reasoning up for analysis is to identify the
method of categorisation, language. Language, or the symbolic universe, is
therefore treated as data in the search for understanding of how reality is
constructed and reconstructed. Although the institution is originally
fashioned by agreement on labile loosely organised activities (Selznick,

1996), its patterns of action soon become stable and inert, in turn affecting
the future behaviour of all its members who perceive it as an objective reality
(Berger & Luckmann, 1966, p. 71). Thus, the power of the school as an
institution and its related agencies impregnates, and in that sense merges
into, the school’s daily activities and constitutes the observed incidences, the
statements heard and discourses read in a study such as this.
The argument proposed here is the general one of the reciprocal relation-
ship between the whole and its parts, which makes it possible to infer some-
thing about the backdrop, the ground rules of education, from data gleaned
in specific events studied and vice versa. This relates to the discussion about
the value of case studies (Ragin & Becker, 1992). While I claim that there is
an internal reciprocity in the case studied (the theoretical concept of school
response to pupil diversity), I do not maintain that it is representative of
other similar cases elsewhere. Nevertheless, I believe that there is much to
learn here about other similar cases particularly at an analytic level. One of
my basic conclusions is that Mossy Mount School’s response to pupil di-
versity is a reflection of basic values and assumptions, which have their origin
in the complex development of Western societies in the past several hundred
years. These include governmental approbation of school systems for the
purposes of maintaining moral order in the community. One can therefore
appreciate that it may take considerable time and effort to fundamentally
alter the school’s response to the problem of pupil diversity.
Ultimately, of course, this is a matter of the kind of knowledge produced,
of truth and the validity of the study. If a strategy such as was employed in
the case of Mossy Mount School produces a trustworthy holistic picture of a
well-selected case and manages to unearth some of the ground rules of
education through careful data collection and interpretation one may con-
clude that the whole and its parts are linked reciprocally and that the drop
merges into the ocean and vice versa. If not, a more laborious approach
must be adopted to make an authentic case. It is left to the reader to judge
whether the above argument is borne out by the methods and processes
described and by the conclusions and inferences produced.
ACKNOWLEDGMENTS
I am indebted to Anne Nevøy for her valuable suggestions regarding an
institutional approach to analysing the data. I also wish to recognise a
considerable debt to my supervisors, Ingrid Lunt and David Scott.
REFERENCES
Ajalnámskrá grunnskóla. (1999). Almennur hluti [National Curriculum Guide. General Part].
Reykjavı́k, Menntamálarájuneytij [Ministry of Culture and Education].
Baldock, J. (1995). The little book of Sufi wisdom. Dorset: Elements Books.
Berger, P., & Luckmann, T. (1966). The social construction of reality: A treatise in the sociology
of knowledge. Harmondsworth: Penguin.
Bhaskar, R. (1989). Reclaiming reality. London: Verso.
Blumer, H. (1969). Symbolic interactionism: Perspective and method. Englewoods Cliffs, NJ:
Prentice Hall.
Bogdan, R. C., & Biklen, S. K. (1992). Qualitative research for education. An introduction to
theory and methods. New York: Allyn and Bacon.
Borjeson, M. (1997). Om Skolbarns olikheter. Diskurser kring ‘‘sarskilda behov’’ i skolan-med
historiska jamforelsespunkter. Stockholm: Skolverket.
Burr, V. (1995). An introduction to social constructionism. London: Routledge.
Clark, C., Dyson, A., & Millward, A. (Eds) (1995). Towards inclusive schools? London: Fulton.
Danermark, B., Ekström, M., Jakobsen, L., & Karlsson, J. C. (2002). Explaining society:
Critical realism in the social sciences. London: Routledge.
Delamont, S. (1992). Fieldwork in educational settings. Methods, pitfalls, perspectives. London:
Falmer Press.
Denzin, N. K., & Lincoln, Y. S. (2005). The Sage handbook of qualitative research. Thousand
Oaks, CA: Sage.
Douglas, M. (1986). How institutions think. New York: Syracuse University Press.
Douglas, M. (1999). Implicit meanings. Selected essays in anthropology. London: Routledge.
Dyson, A., & Millward, A. (1997). The reform of special education or the transformation of
mainstream schools? In: S. J. Pijl & C. J. W. Meijer (Eds), Inclusive education. A global
agenda. London: Routledge.
Eisner, E. (1998). Educational uses of assessment and evaluation Conference on Educational
Assessment. University of Iceland, Háskólabió, Reykjavik.
Gadamer, H. (1979). Truth and method. London: Sheed and Ward.
Gergen, K. J. (1985). The social constructionist movement in modern psychology. American
Psychologist, 40, 266–275.
Gergen, K. J. (1994). Realities and relationalships. Surroundings in social construction. New
York: Harvard University Press.
Gergen, K. J. (1999). An invitation to social construction. London: Sage.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory. Strategies for qualitative
research. New York: Aldine De Gruyer.
Goodey, C. (1999). Learning disabilities: The researcher’s voyage to planet Earth. In: S. Hood,
B. Mayhall & S. Oliver (Eds), Critical issues in social research. Power and prejudice.
Buckingham: Open University Press.
Grossman, H. (Ed.) (1973). Manual on terminology and classification in mental retardation
(revised edition). Washington, DC: American Association on Mental Deficiency.
Hamel, J. (1991). Case study methods. Newbury Park: Sage.
Hammersley, M., & Atkinson, P. (1983). Ethnography principles in practice. London:
Tavistock.
Hargreaves, D. (1967). Social relations in a secondary school. London: Routledge.
Hargreaves, D. (1978). Whatever happened to symbolic interactionism? In: L. Barton &

R. Meighan (Eds), Sociological interpretations of schooling and classrooms. Diffield:
Nafferton.
Hill, J. (1995). Entering the unknown: Case study analysis in special schools. In: P. Clough &
L. Barton (Eds), Making difficulties. Research and the construction of SEN. London: Paul
Chapman.
Hoeg, P. (1992). Miss Smilla’s feeling for snow. London: Penguin.
Johnson, J. M. (1975). Doing field research. New York: Free Press.
Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry. Beverley Hills: Sage.
Lofland, L. (1989). Social life in the public realm: A review. Journal of Contemporary Ethnog-
raphy, 17, 453–482.
Marinósson, G. L. (1998). The ethnographic approach. Educational and Child Psychology, 15, 9.
Marinósson, G. L. (2002). The response to pupil diversity by a mainstream compulsory school in
Iceland. Unpublished Ph.D. thesis, Insititute of Education, University of London.
Mercer, N. (1991). Reserching common knowledge. Studying the content and context of educa-
tional discourse. In: G. Walford (Ed.), Doing educational research. London: Routledge.
Mercer, N., & Edwards, D. (1981). Ground-rules for mutual understanding: A social psycho-
logical approach to classroom knowledge. In: N. Mercer (Ed.), Language in school and
community. London: Edward Arnold.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook.
London: Sage.
Nightingale, D. J., & Cromby, J. (1999). Social constructionist psychology: A critical analysis of
theory and practice. Buckingham: Open University Press.
Ragin, C. C., & Becker, H. S. (Eds) (1992). What is a case? Exploring the foundations of social
inquiry. Cambridge: Cambridge University Press.
Reid, R., Maag, J. W., & Vasa, S. F. (1993). Attention deficit hyperactivity disorder as a
disability category: A critique. Exceptional Children, 60, 198–214.
Reynolds, D., & Sullivan, M. (1979). Bringing schools back in. In: L. Barton & R. Meighan
(Eds), Schools, pupils and deviance. Driffield, UK: Nafferton Books.
Rock, R. (1979). The making of symbolic interactionism. London: Macmillan.
Sanger, J. (1996). The Compleat observer. London: Falmer.
Scott, D. (1996). Methodology in educational research. A distance learning package. Southamp-
ton University, Unpublished.
Selznick, P. (1996). Institutionalism ‘old’ and ‘new’. Administrative Science Quarterly, 41, 270–277.
Sleeter, C. E. (1986). Learning disabilities: The social construction of a special education cat-
egory. Exceptional Children, 53, 46–54.
Spradley, J. P. (1980). Participant observation. New York: Harcourt Brace Javanovich.
Stake, R. E. (1994). Case study. In: N. Denzin & Y. S. Lincoln (Eds), Handbook of qualitative
research. Thousand Oaks: Sage.
Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and
techniques. London: Sage.
Weiss, R. S. (1994). Learning from strangers. The art and method of qualitative interview studies.
New York: The Free Press.
Willis, P. (1984). Notes on method. In: S. Hall, D. Hobson, A. Locker & P. Willis (Eds),
Culture, media and language. London: Hutchinson/Centre for contemporary cultural
studies, University of Birmingham.
Wolcott, H. (1973). The man in the principal’s office: An ethnography. New York: Holt, Rinehart
and Winston.
Wolcott, H. F. (1999). Ethnography a way of seeing. Walnut Creek, CA: Alta Mira Press.
Woods, P. (1996). Researching the art of teaching. Ethnography for educational use. London:
Routledge.
AUTOETHNOGRAPHY: RAISING
CULTURAL CONSCIOUSNESS OF
SELF AND OTHERS
Heewon Chang
WHAT IS AUTOETHNOGRAPHY?
Autoethnography is ethnographical and autobiographical at the same time.

Here I intentionally place ‘ethnographical’ before ‘autobiographical’ to
highlight the ethnographical character of this inquiry method. This character
connotes that autoethnography utilizes the ethnographic research methods
and is concerned about the cultural connection between self and others
representing the society. This ethnographic aspect distinguishes autoeth-
nography from other narrative-oriented writings such as autobiography,
memoir, or journal.
Ellis and Bochner (2000) define autoethnography as ‘autobiographies that
self-consciously explore the interplay of the introspective, personally en-
gaged self with cultural descriptions mediated through language, history,
and ethnographic explanation’ (p. 742). Although their definition leans too
far toward the autobiographical than the ethnographic end, their observa-
tion of ‘connecting the personal to the cultural’ accurately points to the
important mission of autoethnography (p. 739). This important linkage
between ‘the self and the social’ is also emphasized in Reed-Danahay’s
(1997) oft-quoted book, Auto/Ethngoraphy: Rewriting the Self and the

ISSN: 1529-210X/doi:10.1016/S1529-210X(06)12012-4
207
208 HEEWON CHANG
Social. In these works of Ellis, Bochner, and Reed-Danahay, the ‘self ’ refers
to an ethnographer self.
However, when the term ‘auto-ethnography’ was first introduced by an-
thropologist Heider (1975), ‘self’ did not mean the ethnographer self, but
rather the informant self. In his study of Dani people, he called their cultural
accounts of themselves, the Dani’s autoethnography. The term was used in a
similar way when Butz and Besio (2004) discussed the colonized people’s
self-understanding. Hayano (1979) employed the term ‘autoethnography’
differently when he studied his ‘own people.’ Wolcott (2004) informs us that
his ‘own people’ were card players who spent ‘leisure hours playing cards in
Southern California’s legitimate card rooms’ (p. 98).
Since then, an extensive list of labels has been used to refer to autobi-
ographical applications in social science research according to Ellis and
Bochner (2000, pp. 739–740). For Reed-Danahay (1997), the label of auto-
ethnography includes at least three varieties: (1) ‘native anthropology’ pro-
duced by native anthropologists from the people group who were formerly
studied by outsiders; (2) ‘ethnic autobiography’ written by members of ethnic
minority groups; and (3) ‘autobiographical ethnography’ in which anthro-
pologists interject personal experience into ethnographic writing (p. 2).
These varied autoethnographic studies do not place an equal emphasis on
autobiography (content) and ethnography (inquiry process). Ellis and
Bochner offer an insightful triadic model to illustrate the complexity of the
autoethnography nomenclature. They state that ‘[a]utoethnoraphers vary in
their emphasis on the research process (graphy), on culture (ethno), and on
self (auto)’ and that ‘[d]ifferent exemplars of autoethnography fall at differ-
ent places along the continuum of each of these three axes’ (p. 740). Some of
them place more value on the ethnographic process; others on cultural in-
terpretation and analysis; and yet a third kind on self-narratives. Keeping in
mind the triadic balance, I argue that autoethnography should be ethnog-
raphical in its methodological orientation, cultural in its interpretive ori-
entation, and autobiographical in its content orientation. This implies that
self-reflective writings deficient in any one of these ingredients would fall
short of ‘auto-ethno-graphy.’
METHODOLOGY OF AUTOETHNOGRAPHY
Various methodological strategies of autoethnography have been developed
in a variety of qualitative research traditions and listed under different
names (Ellis & Bochner, 2000, p. 740). The list of the names is also extensive
Autoethnography: Raising Cultural Consciousness of Self and Others 209
according to these authors. Regardless of different origins and representa-

tions, all the methodological strategies share the commonality of being the
qualitative, ‘narrative inquiry’ (Clandinin & Connelly, 2000). Some are more
ethnographic than others in terms of its ethnographic intent and research
process. The autobiographic inquires with the ethnographic orientation are
the ones I focus on in this chapter.
Like ethnography, autoethnography pursues the ultimate goal of cultural
understanding underlying autobiographical experiences. To achieve this
ethnographic intent, autoethnographers undergo the usual ethnographic
research process of data collection, data analysis/interpretation, and report
writing. They collect field data by means of participation, self-observation,
interview, and document review; verify data by triangulating sources and
contents; analyze and interpret data to decipher the cultural meanings of
events, behaviors, and thoughts; and write autoethnography. Like ethnog-
raphers, autoethnographers are expected to treat their autobiographical
data with critical, analytical, and interpretive eyes to detect cultural under-
tones of what is recalled, observed, and told of them. At the end of a
thorough self-examination within its cultural context, autoethnographers
hope to gain a cultural understanding of self and others. Autobiographical
narratives will add live details to this principled understanding, but narra-
tion should not dominate autoethnography. In the following subsections,
I will break down the research process into two interconnected, not always
sequential, steps: (1) composing autobiographical field texts and (2) turning
autobiographical field texts into autoethnography.
Composing Autobiographical Field Texts
The initial step of research involves collecting data, which continues

throughout the research process with different intensity at different points.
Here, I cautiously introduce a new term ‘field texts’ by Clandinin and
Connelly (2000) to refer to ‘data.’ I will sometimes use ‘field texts’ in lieu of
‘data’ when ‘composing field texts’ describes more accurately what auto-
ethnographers do. Since the term ‘data’ has been traditionally associated
with quantitative research inquiries and autoethnographers accumulate vo-
luminous texts as multiple data-collection activities progress, the term ‘field
texts’ is justifiably adopted as an alternative to ‘data.’ At the same time, I am
cautious of replacing ‘data’ with ‘field texts’ completely because autoeth-
nographical ‘fieldwork’ is different from other qualitative inquiries. Whereas
qualitative/ethnographic fieldwork is likely to take place in an environment
210 HEEWON CHANG
where the researcher comes in direct contact with others, autoethnographic

fieldwork often involves others in the researchers’ recollection and reflection.
Memory is both a friend and foe of autoethnographers. Whereas it allows
researchers to tap into the wealth of data to which no one else has access,
memory selects, shapes, limits, and distorts. Memory fades as time goes,
blurring the vitality of details. Dillard (1987) (cited Clandinin & Connelly,
2000) recognizes ‘blurring’ as ‘smooth[ing] out details, leaving a kind of
schematic landscape outline’ (p. 83). Memory also triggers aversion when it
attempts to dig deeper into unpleasant past experiences. Foster testifies to the
excruciating pain she experienced in recalling her childhood tainted with her
mother’s schizophrenia and the obliterating effect of the pain on her memory
until her autoethnographic study helped her heal (Foster, McAllister, &
O’Brien, 2005). Memory can also select and embellish pleasant moments.
Omission and addition are natural occurrences in our recalling. In the same
way as subjectivity, they are detrimental to our autobiographic research
endeavor unless they are properly recognized and disciplined.
Composing field texts helps researchers become aware of the limiting
nature of memory and bring details to the ‘schematic landscape outline.’
Clandinin and Connelly (2000) concur that field texts ‘help fill in the rich-
ness, nuance, and complexity of the landscape, returning the reflecting
researcher to a richer, more complex, and puzzling landscape than memory
alone is likely to construct’ (p. 83).
Autoethnographers can use various techniques to facilitate their recalling,
organize memories, and compose field texts as data. The techniques of data
collection include, but are not limited to, (1) using visual tools such as free
drawings of significant places, ‘kinsgrams,’1 and ‘culturegrams’2; (2) inven-
torying people, artifacts, familial and societal values and proverbs, mentors,
cross-cultural experiences, and favorite/disliked activities; (3) chronicling
the autoethnographer’s educational history, typical day and week, and an-
nual life cycle; (4) reading and responding to other autoethnographies and
self-narratives; and (5) collecting other field texts such as stories of others,
‘storied poems,’ personal journals, field notes, letters, conversation, inter-
views with significant others, family stories, documents, photographs, mem-
ory boxes, personal–family–social artifacts, and life experiences (Clandinin
& Connelly, 2000, p. 101). These techniques are elaborated in my book,
Autoethnography, to be published by Left Coast Press in 2007. Autoeth-
nographers are commended to develop their own techniques of data col-
lection to meet their research goals.
One of the commonly used data-collection techniques for ethnography is
participant observation, in which researchers participate in the lives of their
informants while observing the latter’s behaviors. In a similar fashion to

this, autoethnographers can observe their own behaviors and document
their thoughts while living them. Rodriguez and Ryave (2002) argue that
self-observation as a data-collection technique is useful because it gives ac-
cess to ‘covert, elusive, and/or personal experiences like cognitive processes,
emotions, motives, concealed actions, omitted actions, and socially re-
stricted activities’ (p. 3) and brings to the surface what is ‘taken-for-granted,
habituated, and/or unconscious manner that [they] y are unavailable for
recall’ (p. 4). Self-observation may be used in the form of self-introspec-
tion when autoethnographers are alone or in the form of ‘interactive in-
trospection’ while the researchers interact with others. In the interactive
introspection, the researchers and the others can interview each other ‘as
equals who try to help one another relive and describe their recollection of
emotional experiences’ (Ellis, 1991 cited Rodriguez & Ryave, 2002, p. 7).
Although Rodriguez and Ryave’s technique of ‘systematic self-observation’
is originally suggested for studies that utilize multiple informants who are
instructed to conduct their own self-observation, this technique can be ap-
plied to autoethnography that focuses on one informant, none other than
self. Field journals or a self-developed recording form may be used to doc-
ument unstructured or structured self-observation.
Interviewing is another vital data-collection technique employed in eth-
nographic fieldwork (Ellis, 2004; Fontana & Frey, 2000; Schensul, Schensul,
& LeCompte, 1999). Through interviewing with myriad informants, eth-
nographers gather information unavailable from participant observation.
When applied to autoethnography, interviews with others fulfill a different
goal. The interviews provide not only outsider perspectives, but also exter-
nal data to confirm, complement, or dispute internal data generated from
recollection and reflection. One caveat, however, is that face-to-face inter-
view can hamper honest exchanges between interviewers – autoethnogra-
phers themselves – and interviewees. To obtain more candid perspectives on
autoethnographers from interviewees, external interviewers or other creative
alternatives such as e-mail survey or questionnaire compiled by a third party
may be adopted.
Turning Autobiographical Field Texts into Autoethnography
In qualitative research, the step of data collection is not always sequential

to or separate from that of data analysis/interpretation. Rather, the data-
collection process is often intertwined and interactive with data analysis and
212 HEEWON CHANG
interpretation. In other words, these activities often take place concurrently

or inform each other in a web-like fashion. For example, when autoeth-
nographers recall past experiences, they do not randomly harvest bits of
fragmented memories. Rather, they select some according to their research
focus and data-collection criteria. Evaluating certain experiences against the
criteria is an analytical and interpretive activity that is already at work
during data collection. During this data-collection process, the researchers
are also able to refine their criteria, which will in turn shape the analysis and
interpretation process.
When analyzing and interpreting autoethnographic field texts, autoeth-
nographers need to keep in mind that what makes autoethnography eth-
nographical is its ethnographic intent of gaining a cultural understanding of
self that is intimately connected to others in the society. The cultural mean-
ings of self’s thoughts and behaviors – verbal and non-verbal – need to be
interpreted in their cultural context. Interpretation begs a question of ‘why’
to be answered: ‘Why does a self perceive, think, behave, and evaluate the
way it does and how does the self relate to others in thoughts and actions?’
Autoethnographic data analysis and interpretation involves moving back
and forth between self and others, zooming in and out of the personal and
social realm, and submerging in and emerging out of data. Like other eth-
nographic inquiries, this step of research process is methodologically neb-
ulous to describe and instruct because analysis and interpretation require
ethnographers’ holistic insight, creative mixing of multiple approaches, and
patience with uncertainty. Yet some simple strategies – searching for recur-
ring patterns, applying existing theoretical frameworks, and compare–con-
trasting with other autoethnographies – can be adopted as a starter in the
process of analysis and interpretation.
The interweaving of data collection, analysis, and interpretation ulti-
mately leads to the production of autoethnography. This means that au-
tobiographical writing cannot come without a methodical process of
ethnography and its focus on cultural understanding. However, it does
not mean that writing can begin only when analysis/interpretation is com-
pleted. Wolcott (2001) suggests that ethnographers begin writing earlier in
the ethnographic process, even during the early stage of fieldwork, because
writing stimulates, helps organize, and facilitates the subsequent data-col-
lection/analysis/interpretation process. This suggestion is useful for auto-
ethnographers.
The writing style of autoethnography can vary, falling somewhere in the
continuums between ‘realist’ description and ‘impressionist’ caricature and
analytical description and ‘confessional’ self-exposure. Van Maanen’s (1988)
classification of ethnographic writings may help autoethnographers exper-

iment with different styles such as ‘realistic tales,’ ‘confessional tales,’ and
‘impressionist tales.’ Realistic tales refer to matter-of-fact accounts and
representations that ethnographers give about people whom they have
studied first hand. Realist tales are characterized by ‘minute, sometimes
precious, but thoroughly mundane details of everyday life among the people
studied’ (p. 48) and include ‘accounts and explanations by members of the
culture of the events in their lives’ (p. 49). Ethnographers who employ realist
tales tend to speak of the people they have studied with the authority of an
expert. In reaction to realist ethnographers’ unabashed claim of authority
over other people’s culture, those who practice confessional ethnography
expose ‘how particular works [really] came into being’ (p. 74) in confessional
tales. ‘Personal biases, character flaws, or bad habits,’ which Van Maanen
dubs as ‘embarrassing,’ are candidly displayed to demystify the ethno-
graphic process and to augment a ‘reasonably uncontaminated and pure
[ethnography] despite all the bothersome problems exposed in the confes-
sion’ (p. 78). Impressionist tales highlight ‘rare’ and ‘memorable’ fieldwork
experiences (p. 102). If realist tales focus on ‘the done’ and confessional tales
on ‘the doer,’ ‘impressionist tales present the doing of fieldwork’ (p. 102). If
autoethnographers keep in mind that these tales are originally identified
with ethnography and thus need to be modified when applied to autoeth-
nography, the different tales may provide alternatives in autoethnographic
writing. Whichever style autoethnographers decide to employ, autoethnog-
raphers are advised not to lose the sight of the quintessential identity of
autoethnography as a cultural study of self and others.
BENEFITS OF AUTOETHNOGRAPHY
Autoethnography is becoming a useful and powerful tool for researchers
and practitioners who deal with human relations in multicultural settings:
e.g., educators, social workers, medical professionals, clergies, and counsel-
ors. Benefits of autoethnography lie in three areas: (1) it offers a research
method friendly to researchers and readers; (2) it enhances cultural under-
standing of self and others; and (3) it has a potential to transform self and
others toward the cross-cultural coalition building.
Methodologically speaking, autoethnography is researcher-friendly. This
inquiry method allows researchers to access easily the primary data source
from the beginning because the source is the researchers themselves. In
addition, autoethnographers are privileged with a holistic and intimate
214 HEEWON CHANG
perspective on their ‘familiar data.’ This initial familiarity gives researchers

an advantageous edge in data collection and in-depth data analysis/inter-
pretation.
Autoethnography is also reader-friendly in that the personally engaging
writing style tends to appeal to readers more than the conventional scholarly
writing. According to Nash (2004), ‘scholarly personal narratives’ liberate
researchers from abstract, impersonal writings and ‘touch readers’ lives
by informing their experiences’ (p. 28). Gergen and Gergen (2002) also
eloquently state, ‘In using oneself as an ethnographic exemplar, the re-
searcher is freed from the traditional conventions of writing. One’s unique
voicing – complete with colloquialisms, reverberations from multiple rela-
tionships, and emotional expressiveness – is honored’ (p. 14). This unique
voice of the autoethnographer is what readers respond to.
Secondly, autoethnography is an excellent vehicle through which re-
searchers come to understand themselves and others. I found this benefit
particularly applicable to my teaching of multicultural education. As a
teacher educator, I feel compelled to prepare my students to become cross-
culturally sensitive and effective teachers for students of diverse cultural
backgrounds. Self-reflection and self-examination are the keys to self-
understanding (Florio-Ruane, 2001; Nieto, 2003). Kennet (1999) concurs
with other advocates of self-reflection, saying that ‘[writing cultural] auto-
biography allows students to reflect on the forces that have shaped their
character and informed their sense of self’ (p. 231). The ‘forces’ that shape
people’s sense of self include nationality, religion, gender, education, eth-
nicity, socio-economic class, and geography. Understanding ‘the forces’ also
helps them examine their preconceptions and feelings about others, whether
they are ‘others of similarity,’ ‘others of difference,’ or even ‘others of op-
position’ (Chang, 2005). Others of similarity refer to members of cultural
groups that one belongs to, feels comfortable with, and share common values
with. Others of difference are those who belong to groups that have different
cultural standards than the self. Others of opposition are those who are
considered as ‘enemies’ to the self due to seemingly irreconcilable differences.
Not only writing one’s own autoethnography but also reading others’
autoethnography can evoke self-reflection and self-examination (Florio-Ruane
2001; Nash, 2002). Connelly shares a poignant story of how reading the self-
narrative of his doctoral student of Chinese heritage stirred up his child-
hood memory of a Chinese store owner from his rural hometown in Canada
(Clandinin & Connelly, 2000). Through self-reflection, he discovered shared
humanity between this stranger of his childhood and himself. This discovery of
self and others is a definite benefit of doing and sharing autoethnographies.
Thirdly, doing, sharing, and reading autoethnography also help trans-

form researchers and readers (listeners) in the process. The transformation
of self and others is not necessarily a primary goal of autoethnography but a
frequently occurring, powerful by-product of this research inquiry. Coia and
Taylor’s (2006) experimentation with ‘co/autoethnography’ illustrates this
benefit. In this participatory process, the researchers involved their educa-
tion students in writing their personal narratives, meeting in small groups
weekly to share the narratives aloud and conduct a cultural analysis col-
laboratively, exchanging newly acquired self-awareness on ‘their past,
present, and future selves,’ and ultimately ‘strengthen[ing] perspective on
teaching’ (p. 21). In the end, the authors witnessed that students’ self-
awareness and cultural understanding were broadened and their teaching
philosophies and practices became more inclusive and sensitive to others’
needs.
Self-transformation may be manifested in different ways in the edu-
cation field. Some may become more self-reflective in their daily praxis
(Florio-Ruane, 2001; Nieto, 2004; Obidah & Teel, 2001). Others may adopt
‘culturally relevant pedagogy’ when selecting curriculum content and ped-
agogical strategies, and interacting students, peer teachers, and the com-
munity (Ladson-Billings, 1994). Self-transformation may also take place as
they seek to reach out to unfamiliar others and pursue a new learning of
unfamiliar cultures. As their understanding of others increases, unfamiliar-
ity diminishes and perspectives on others change. As a result, others of
difference and of opposition may be reframed to be included in their notion
of community, ‘extended community’ in Greene’s (2000) term.
Another type of self-transformation may accompany healings from the
emotional scars of the past, which Foster illuminated in her writing (Foster
et al., 2005). By sharing with others her painful experience of growing up
with a mother with schizophrenia and understanding the cultural root of her
‘wounds,’ Foster experienced liberation and relief from the burden of iso-
lation, loneliness, and shame. The liberating force of autoethnography was
the foundation of self-empowerment for Foster.
When manifested in increased self-reflection, adoption of the culturally
relevant pedagogy, desire to learn about ‘others of difference,’ development
of an inclusive community, or self-healing, the self-transformative potential
of autoethnography is universally beneficial to those who work with people
from diverse cultural backgrounds. Through the increased awareness of self
and others, they will be able to help themselves and each other correct
cultural misunderstandings, develop cross-cultural sensitivity, and respond
to the needs of cultural others effectively.
216 HEEWON CHANG
PITFALLS TO AVOID IN DOING

AUTOETHNOGRAPHY
In the shadow of the growing interest and support of autoethnographic
research methods, critiques are lurking. The criticism of autoethnography
does not necessarily imply that this inquiry is inherently faulty. Rather, it
reminds researchers to look out vigilantly for appropriate application of this
research inquiry and to avoid potential pitfalls. Here are five pitfalls that
autoethnographers need to watch out: (1) excessive focus on self in isolation
of others; (2) overemphasis on narration rather than analysis and cultural
interpretation; (3) exclusive reliance on personal memory and recalling as a
data source; (4) negligence of ethical standards regarding others in self-
narratives; and (5) inappropriate application of the label ‘autoethnography.’
The first pitfall relates to the very notion of culture. In the minds of
anthropologists, culture is inherently a group-oriented concept. Culture and
people have a symbiotic relationship according to de Munck (2000) who
says: ‘Culture would cease to exist without the individuals who make it
up y. Culture requires our presence as individuals. With this symbiosis, self
and culture together make each other up and, in that process, make mean-
ing’ (pp. 1–2). Therefore, the notion of culture predisposes the co-presence
of others even in a discussion of ‘individual culture’–‘propriospect’ in
Goodenough’s (1981) and Wolcott’s (1991) term and ‘idioverse’ in
Schwartz’ term (1978) (cited de Munck, 2000). By these authors, an indi-
vidual culture is an individual version of their group cultures, which they
construct in relationship with others. Autoethnography, therefore, should
reflect the interconnectivity of self and others. Unfortunately, the method-
ological focus on self is sometimes misconstrued as a license to dig deeper in
personal experiences without digging wider into the cultural context of the
individual stories commingled with others. Autoethnographers should be
warned that self-indulgent introspection is likely to produce a self-exposing
story but not autoethnography.
Second, autoethnographers swept by the power of story telling can easily
neglect the very important mission of autoethnography – cultural interpre-
tation and analysis of autobiographic texts. Self-narration is very engaging
to writers as well as readers and listeners (Foster et al., 2005; Nash, 2004;
Tompkins, 1996). Yet, as Coia and Taylor (2006) say, ‘It is not enough
simply to tell the story or write a journal entry’ (p. 19) for the cultural
understanding of self to take place. Unless autoethnographers stay focused
on their research purpose, they can be tempted to settle for elaborate nar-
ratives with underdeveloped cultural analysis and interpretation.
Third, autoethnographers can fall into the pitfall of over-relying on their

personal memory as the source of data. Personal memory is a marvelous and
unique source of information for autoethnographers. It taps into the res-
ervoir of data to which other ethnographers have no access. Yet, Muncey
(2005) reminds us, ‘Memory is selective and shaped, and is retold in the
continuum of one’s experience, [although] this does not necessarily consti-
tute lying’ (p. 2). Memory can censor past experiences. When data is col-
lected from a single tool without other measures for checks and balances,
the validity of data can be questioned. When the single tool is the researcher
self, the unbridled subjectivity of autoethnographers can be more severely
challenged. Although the obsession with objectivity is not necessary for
qualitative research, autoethnographers need to support their arguments
with broad-based data like in any good research practice. For this reason,
they can easily complement ‘internal’ data generated from researchers’
memory with ‘external’ data from outside sources, such as interviews, doc-
uments, and artifacts. Multiple sources of data can provide bases for tri-
angulation that will help enhance content accuracy and validity of the
autoethnographic writing.
The fourth pitfall stems from a false notion that confidentiality does not
apply to self-narrative studies because researchers use their autobiographical
stories. Playing the multi-faceted role of a researcher, informant, and
author, autoethnographers may be tempted to claim full authorship and
responsibility for their stories without hesitation. Clandinin and Connelly’s
(2000) poignant question to narrative inquirers, ‘Do they own a story
because they tell it?’ should equally challenge autoethnographers. Since
autoethnographers’ personal stories are often linked to stories of others,
however explicit the linkage is, the principle of protecting confidentiality of
people in the story is just as relevant to autoethnography. Since main char-
acters reveal their identities in autoethnography, it is extremely difficult to
protect fully others intimately connected to these known characters. Yet,
autoethnographers, like other researchers of human subjects, are charged to
adhere to the ethical principle of confidentiality. This inquiry method
demands more creativity in practicing it.
The last pitfall concerns the confusion in using the term ‘autoethnography.’
As I discussed earlier in this chapter, the term has been used to refer to a
variety of narrative inquiries sprung up in different academic disciplines. The
mixed bag labeled with ‘autoethnography’ has confused researchers as well as
readers. Since no one can claim an exclusive license to use this label, it is the
researcher’s responsibility to become informed of the multiple usage of the
term and to define its use clearly to avoid confusion. That is precisely what
218 HEEWON CHANG
Wolcott (2004) does in his article, ‘Ethnographic Autobiography.’ Although

my use of autoethnography differs from what he proposes – leaving the term
to the original meaning by Hayano who refers to autoethnography as a study
of the researcher’s own people – his conscious clarification of the term clearly
orients readers. With the rigorous effort to distinguish autoethnography from
other self-narrative inquiries, readers will be able to understand this research
method for what it stands for, distinguishing it from highly descriptive self-
narratives such as autobiography and memoir.
CONCLUSION
As the outgoing president of the Council on Anthropology and Education
(a subdivision of American Anthropological Association), LeCompte (1987)
once asserted in her speech, later published as article, that all research end-
eavors are autobiographic. I understood her remark to imply that research
topics, methods, and processes that we select reflect our personal interest,
biases, and circumstances. When I reflect on my ethnographic and quali-
tative works, I find her comment insightful and accurate.
My preference of case-specific narratives, field-based research methods,
and adolescents has led me through a series of ethnographic/qualitative
studies. Beginning with American high school students on the West Coast of
the U.S.A. (Chang, 1992a), I moved on to a study of Korean female stu-
dents in a vocational high school (Cho & Chang, 1989; Chang, 2000), to
Korean–American high school students (Chang, 1992b), and to Christian
high school students in Pennsylvania (Chang, 1998). Venues have changed,
but my dogged commitment to ethnography and adolescents has persisted.
What is the cultural root of my persistence?
Searching for answers to questions such as this is what autoethnographers
do and what I have done in my autoethnography. Although I reserve an
elaborate cultural analysis of myself for another place, here I can safely note
that the research process has been empowering and transformative. My
teaching and doing autoethnography has helped my students and me (1)
connect our individual past with our individual and collective present, (2)
understand culturally rooted reasons for our comfort with others of similarity,
discomfort with others of difference, and aversion with others of opposition,
and (3) expand this understanding into culturally unfamiliar territories. Open-
ing-up to the new understanding and new possibility is the definite benefit of
autoethnography, which gives a foundation to cross-cultural coalition build-
ing to embrace others, even others of opposition, in the multicultural society.
NOTES
1. The ‘kinsgram’ refers to a kinship diagram that visually shows one’s relation to
other members in his/her kinship structure, created or dissolved by birth/death,
marriage/divorce, and other forms of attachment/separation. Kinship diagram is
often adopted in anthropological fieldwork (see Bates & Fratkin, 2003, p. 282; and
Puples & Bailey, 2003, p. 190).
2. The ‘culturegram,’ a self-coined term, refers to a web-like chart on which au-
toethnographers display their multi-faceted self-identity in terms of multicultural
categories such as nationality, ethnicity, race, language, gender, religion, socio-eco-
nomic class, and other interests (Chang, 2002).
REFERENCES
Bates, D. G., & Fratkin, E. M. (2003). Cultural anthropology (3rd ed.). Boston, MA: Pearson.
Butz, D., & Besio, K. (2004). The value of autoethnography for field research in transcultural
settings. Professional Geographer, 56(3), 350–360.
Chang, H. (1992a). Adolescent life and ethos: An ethnography of a US high school. London:
Falmer Press.
Chang, H. (1992b). Persist or perish in the qualitative studies of adolescents: Strategies to win a
silent negotiation. 1992 QUIG conference proceedings. Athens, GA: University of Georgia.
Chang, H. (1998). Is cultural diversity a friend or a foe? An ethnography of Christian schools.
Paper presented at the 1998 academic festival, Eastern College, St. Davids, PA.
Chang, H. (2000). Triple jeopardy in identity formation: Korean urban girls in an evening
commercial high school. Paper presented at the urban girls conference, SUNY, Buffalo,
NY.
Chang, H. (2002). Cultural autobiography: Self-reflective practices for multicultural educa-
tors. Paper presented at the CCTE Education 2002 Conference, Calvin College, Grand
Rapids, MI.
Chang, H. (2005). Self-narratives for Christian multicultural educators: A pathway to under-
standing self and others. A Journal of the international community of Christians in
teacher education [online], 1 (1). Available at http://www.icctejournal.org/ICCTEJour-
nal/past_issues/vol1issue1/v1i1chang [Accessed 26 February 2006].
Cho, E., & Chang, H. (1989). Life and value system of evening girls’ commercial high school
students in Korea (written in Korean). Annual Journal of Soong Sil Arts and Sciences
Research Institute, 19, 11–38.
Clandinin, D. J., & Connelly, F. M. (2000). Narrative inquiry: Experience and story in qualitative
research. San Francisco, CA: Jossey-Bass.
Coia, L., & Taylor, M. (2006). From the inside out and the outside in: Co/autoethnography as a
means of professional renewal. In: C. Kosnik, C. Beck, A. R. Rease & A. P. Samaras
(Eds), Making a difference in teacher education through self-study (pp. 19–33). Amsterdam,
Netherlands: Springer.
de Munck, V. (2000). Culture, self, and meaning. Prospect Heights, IL: Waveland.
Dillard, A. (1987). An American childhood. Boston, MA: G.K. Hall.
Ellis, C. (1991). Sociological introspection and emotional experience. Symbolic Interaction, 14,
23–50.
220 HEEWON CHANG
Ellis, C. (2004). An ethnographic I: A methodological novel about autoethnography. Walnut

Creek, CA: Altamira.
Ellis, C., & Bochner, A. P. (2000). Autoethnography, personal narrative, and personal reflex-
ivity. In: N. Denzin & Y. Lincoln (Eds), Handbook of qualitative research, (2nd ed.).
Thousand Oaks, CA: Sage.
Florio-Ruane, S. (2001). Teacher education and the cultural imagination. Mahwah, NJ:
Lawrence Erlbaum.
Fontana, A., & Frey, J. H. (2000). The interview form structured questions to negotiated
text. In: N. Denzin & Y. Lincoln (Eds), Handbook of qualitative research, (2nd ed.)
(pp. 645–672). Thousand Oaks, CA: Sage.
Foster, K., McAllister, M., & O’Brien, L. (2005). Coming to autoethnography: A mental health
nurse’s experience. International journal of qualitative methods [online], 4 (4), Available
at http://www.ualberta.ca/ijqm/backissues/4_4/html/foster.htm [Accessed 9 February
2006].
Gergen, M., & Gergen, K. (2002). Ethnographic representation as relationship. In: Bochner,
A. P. & C. Ellis (Eds), Ethnographically speaking: Autoethnography, literature, and
aesthetics. Walnut Creek, CA: Altamira.
Goodenough, W. (1981). Culture, language, and society. Menlo Park, CA: Benjamin/
Cummings.
Greene, M. (2000). The passions of pluralism: Multiculturalism and the expanding community.
In: J. Noel (Ed.), Sources: Notable selection in multicultural education. Guilford, CT:
Dushkin/McGraw-Hill.
Hayano, D. M. (1979). Auto-ethnography: Paradigms, problems, and prospects. Human
Organization, 38(1), 99–104.
Heider, K. G. (1975). What do people do? Dani auto-ethnography. Journal of Anthropological
Research, 31, 3–17.
Kennet, C. L. (1999). Saying more than hello: Creating insightful cross-cultural conversations.
In: D. C. Elliott & S. D. Holtrop (Eds), Nurturing and reflective teachers: A Christian
approach for the 21st century. Claremont, CA: Learning Light Educational Consulting
and Publishing.
Ladson-Billings, G. (1994). The dreamkeepers: Successful teachers of African American children.
San Francisco, CA: Jossey-Bass.
LeCompte, M. (1987). Bias in the biography: Bias and subjectivity in ethnographic research.
Anthropology and Education Quarterly, 18(1), 43–52.
Muncey, T. (2005). Doing autoethnography. International Journal of Qualitative Methods
[online], 4 (1), 1-12. Available at http://www.ualberta.ca/iiqm/backissues/4_1/pdf/
Muncey.pdf [Accessed 9 Februrary 2006].
Nash, R. J. (2002). Spirituality, ethics, religion, and teaching: A professor’s journey. New York:
Peter Lang.
Nash, R. J. (2004). Liberating scholarly writing: The power of personal narrative. New York:
Teachers College Press.
Nieto, S. (2003). What keeps teachers going? New York: Teachers College.
Nieto, S. (2004). Affirming diversity: The sociopolitical context of multicultural education
(4th ed.). Boston, MA: Allyn and Bacon.
Obidah, J. E., & Teel, K. M. (2001). Because of the kids: Facing racial and cultural differences in
schools. New York: Teachers College.
Puples, J., & Bailey, G. (2003). Humanity: Introduction to cultural anthropology. Baltimore,
MD: Wadsworth.
Reed-Danahay, D. E. (Ed.) (1997). Auto/ethnography: Rewriting the self and the social.
New York: Oxford.
Rodriguez, N., & Ryave, A. (2002). Systematic self-observation. Thousand Oaks, CA: Sage.
Schensul, S. L., Schensul, J. J., & Lecompte, M. D. (1999). Essential ethnographic methods:
Observations, interviews, and questionnaires. Walnut Creek, CA: Altamira.
Schwartz, T. (1978). Where is the culture? Personality as the locus of culture. In: G. De Vos
(Ed.), Making of psychological anthropology. Berkeley, CA: University of California.
Tompkins, J. P. (1996). A life in school: What the teacher learned. Reading, MA: Perseus Books.
Van Maanen, J. (1988). Tales of the field: On writing ethnography. Chicago, IL: University of
Chicago Press.
Wolcott, H. F. (1991). Propriospect and the acquisition of culture. Anthropology and Education
Quarterly, 22(3), 251–273.
Wolcott, H. F. (2001). Writing up qualitative research (2nd ed.). Thousand Oaks, CA: Sage.
Wolcott, H. F. (2004). The ethnographic autobiography. Auto/Biography, 12, 93–106.
ABOUT THE AUTHORS
Dalia Aralas is a doctoral research student at the University of Oxford

Department of Educational Studies. Her doctoral research study, ‘Investi-
gating Mathematical Imagination’, is supported by a four-year JPA
Malaysia scholarship. She is a pure mathematician by training. She is a
lecturer of Universiti Malaysia Sabah (UMS) where she was the Coordi-
nator of mathematics and science education before she left on study leave.
She had academic and administrative responsibilities for coordinating ac-
ademic programmes run jointly by the two Schools (faculties) of Education
and Science. She has taught many undergraduate and postgraduate courses
at UMS, including mathematics education, educational foundations and
issues (philosophical perspectives). She was the head of the teacher training
programme in mathematics and science. She directed the first of statewide
UMS mathematics camps which have been held annually. She has also
tutored and examined an undergraduate class in a course in mathematics
education at the University of Oxford. Besides additional training in teach-
ing mathematics, she also has a teaching certificate in dance. Her research
interests include mathematics education, particularly the philosophical
strand; imagination and agency; and research methodology.
Heewon Chang (PhD, University of Oregon) is associate professor of Ed-

ucation and chair of Graduate Education Programs at Eastern University,
Pennsylvania, USA, where she teaches courses on multicultural educa-
tion, research design, gender equity education, and global education.
She founded an open-access e-journal, Electronic Magazine of Multicultural
Education (http://www.eastern.edu/publications/emme), in 1999 and has
served as Editor-in-Chief. Her book, Adolescent Life and Ethos: An
Ethnography of a US High School (1999, Falmer Press), and subsequent
research projects reflect her interest in adolescent culture. Trained as an
educational anthropologist, she keeps her research focus on multicultural
education, anthropology and education, and ethnographic and autoethno-
graphic methodology. Her lived experience with the Korean, US, and
German cultures inform her teaching and research agenda.
223
224 ABOUT THE AUTHORS
Amar Dhand holds a doctorate from the Department of Educational Studies

at Oxford University where he was a Rhodes Scholar from Canada, and on
leave from medical studies at Harvard Medical School. His current research
focuses on naturalistic peer learning processes among heroin addicts in
India. He has written and continues to write about a diversity of practices
among a particular group of addicts including poetry performances, street
medicine, and brotherhood structures. More generally, he is interested in the
interface between education and medicine in theoretical, empirical, and
pragmatic dimensions. Following his medical training, he hopes to indulge
his curiosity for everyday life with sociological and educational research
projects that intersect with his medical life.
Martin Forsey lectures in Anthropology at the University of Western

Australia. He teaches introductory anthropology as well as Australian
culture and society and the anthropology of organisational enterprises.
Martin’s research interests include the organisational culture of schools,
school choice and educational reform and he has written about ethno-
graphic method, neo-liberal reform of schooling and organisational change.
A monograph titled Challenging the System? A Dramatic Tale of Neo-liberal
Reform in an Australian Government High School will be published in 2006
by Information Age Publishing.
Nick Hopwood’s research interests include geography education, environ-

mental education, and professional development in academia. As well as
studying pupils’ experiences of school geography (the focus of his current
work), he has been involved in projects investigating geography teachers’
conceptions and practice, and student teachers’ ideas about teaching con-
troversial issues in geography and science. Nick’s developing interests in
research methodology focus on the implications of conducting research with
children, on issues of researcher roles and processes of data generation, and
on the analysis of qualitative data. He is now working on a project aimed at
understanding and enhancing processes of professional development for
post graduate research students aiming for academic careers (part of Oxford
University’s Centre for Excellence in Preparing for Academic Practice).
Susan James is a postdoctoral researcher at the ESRC funded SKOPE

(Skills, Knowledge and Organizational Performance) Research Centre,
based jointly at Oxford and Warwick Universities. Her research interests
include vocational education and training, VET systems, work-based learn-
ing, on-the-job and off-the-job training, school-to-work transitions, and
About the authors 225
theories of learning. She is currently working on a European-wide project

researching opportunities for progression in low wage, low-skill work.
Gretar L. Marinósson is a professor of Education at the Iceland University of

Education in Reykjavı́k. He is also dean of research and director of IUE
Research Centre. He was educated as an educational psychologist (in the UK)
and a teacher (in Iceland) and worked as a teacher and school psychologist
in England and in Iceland for 15 years before becoming a university lecturer
in 1987. He obtained his masters and doctoral degrees in Manchester and
London, respectively. Major research interests are special needs and inclusive
education, pupils’ social interaction and behaviour, and parent–school
collaboration. At present he leads a research team studying the education
of pupils with learning disability in Iceland. Recent works include Pathways
to Inclusion, a Guide to Staff Development (edited with Rósa Eggertsdóttir).
Kristen H. Perry is a doctoral candidate at Michigan State University. Her

work focuses primarily on literacy development and culture, particularly in
African communities both in the US and abroad. Her specific areas of
interest include the ways in which culture and literacy development transact
in diverse communities. Her research also investigates the various ways in
which home and community practices of literacy align with school practices
of literacy, particularly for immigrant and refugee children. She is the
recipient of Michigan State’s University Distinguished Fellowship and a
Spencer Research Training Grant Fellowship. She also has taught in mul-
tigrade classrooms in the US state of Colorado; worked as a resource
teacher and HIV/AIDS educator in Lesotho, Africa through the US Peace
Corps; and tutored African refugees in the US.
Eric Tucker completed his doctorate in Educational Studies at the

University of Oxford with the support of a Marshall Scholarship. He gradu-
ated with distinction from Oxford with a Masters of Science in Education
Research Methodology, writing a dissertation entitled Towards Grounded
Indicators of Social Capital Formation. He holds honours degrees in Public
Policy and Africana Studies from Brown University where he received the
Truman Scholarship, Royce Fellowship, Noah Krieger Prize for the best
Public Policy thesis, and graduated Magna Cum Laude and Phi Beta Kappa.
As a director of publications for the National Association for Urban Debate
Leagues he served as editor for the Urban Debate Chronicle, co-authored
Teaching Argumentation and Debate: An Educator’s Activities Manual (2004)
and authored Building an Urban Debate League: A Practitioner’s Handbook
226 ABOUT THE AUTHORS
(2006). He also wrote How to Build a Debate Program: An Organizer’s

Manual.
Geoffrey Walford is professor of Education Policy and a Fellow of Green

College at the University of Oxford. His research foci are the relationships
between central government policy and local processes of implementation,
private schools, choice of schools, religiously based schools, and qualitative
research methodology. His books include Life in Public Schools (Methuen,
1986), Restructuring Universities: Politics and Power in the Management
of Change (Croom Helm, 1987), City Technology College (Open University
Press, 1991, with Henry Miller), Doing Educational Research (Routledge,
ed., 1991), Choice and Equity in Education (Cassell, 1994), Educational
Politics: Pressure Groups and Faith-Based Schools (Avebury, 1995),
Policy, Politics and Education-Sponsored Grant-Maintained Schools and
Religious Diversity (Ashgate, 2000), Doing Qualitative Educational Research
(Continuum, 2001), Private Schooling: Tradition and Diversity (Continuum,
2005) and Markets and Equity in Education (Continuum, 2006). He was joint
editor of the British Journal of Educational Studies from 1999 to 2002, editor
of this annual volume, Studies in Educational Ethnography, and has been
editor of the Oxford Review of Education since January 2004. He is also one
of the deputy editors of the new journal Ethnography and Education.
Sue Walters is lecturer in Educational Studies at Edinburgh University. She

was previously employed as a Research Associate at Lancaster University
Literacy Research Centre working with Dr. Uta Papen on a study of literacy,
learning, and health. Prior to that, and following her completion of a doctorate
on the educational experiences of Bangladeshi pupils, she worked at the Open
University with Dr. Sarah Neal looking at issues of ethnicity and rural identity.
Harry F. Wolcott is professor emeritus of Anthropology and Education at the

University of Oregon in Eugene, Oregon, where he has spent his entire career.
He had the good fortune to meet George Spindler when he went to Stanford
to study for his doctorate. He realized that ‘anthropology and education’ was
the right field for him, although there really was no ‘field’ as such at the time.
Under Spindler’s direction he did his dissertation study among the Kwakiutl
Indians of British Columbia and he has been doing ethnography ever since,
ranging from a study of urban African beer drinking in Bulawayo, Rhodesia
(now Zimbabwe) to a study of a school principal across town. He is a past
president of the Council on Anthropology and Education and was the fourth
editor of its journal, the Anthropology and Education Quarterly.

(Studies in Educational Ethnography 12) Geoffrey Walford - Methodological Developments in Ethnography-Emerald Group (2007)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Studies in Educational Ethnography 12) Geoffrey Walford - Methodological Developments in Ethnography-Emerald Group (2007)

Uploaded by

Copyright:

Available Formats

PREFACE

Ethnography has become one of the major methods of researching educa-

What counts as ethnography and what counts as good ethnographic meth-

uniform, acceptance, and initiation procedures while conducting research on

statistical techniques are the preeminent methodological approaches with

This paper addresses the problem of access in ethnographic research from

Methodological Developments in Ethnography

and self-presentation theories, she reveals four relationship patterns in the

As is indicated by the qualiﬁcation ‘(available if there is legitimate access)’,

However, the apprenticeship literature tends to focus on the ethnogra-

GENERAL FIELD METHODS

This paper derives from a seven-month ethnography involving three ﬁeldwork

main commuter lines had caused widespread urban advancements. The

The participants were a group of active, rehabilitating, and rehabilitated

PLAYING A PART IN THE NGO

The process of accessing this group of heroin addicts began by establishing a

recorded interview, two peer educators gained the participant’s permission

Similarly, in a number of instances including my ﬁrst recorded interview,

One participant also became an ‘observer’ who reported his ‘ﬁndings’ in

It is important to note that none of these instances were coordinated or

maintain a reputation of a ‘clean’ ex-user. I found that this distance was

tolerant of private use by individuals. Throughout my time with the ‘brother-

These examples illustrate that key limitations of my involvement, namely

AD: How did you ﬁgure this out?

The legitimization process also had reciprocal qualities: the ‘brothers’

celebrated my connection with their group. During an interview, Veer de-

Acknowledgement of my participation in their lives was also accompanied

Although it was important for me to gain legitimacy from the group to

A FRIENDSHIP WITH BABA

describe these aspects of access by delineating the development of our re-

Baba facilitated an extended observation of the event. He had navigational

take me. On our way, he reprimanded me for arranging to complete the

3. Keep quiet during the negotiation with the middleman.

6. Don’t show your notebook.

7. Don’t let them think you are an undercover cop.

8. Don’t show the ampoules too much.

Fig. 2. Baba’s Illustration in the Ethnographer’s Notebook Depicting How Alcohol

Following his pick-pocketing demonstration, he emphasized that I should

Access in ethnographic research is the participation in negotiated interper-

Ethnographers’ identity claims must be validated by participants in order for researchers

informants so the informants become meaningful access-points and inter-

These ﬁndings, therefore, suggest that understanding how ‘old-timers’ learn

This paper has attempted to provide a theoretical addition to conceptuali-

the researcher’s interpersonal relationships with the participants, but also

Mr. Francis Joseph transcribed and transliterated the tape recordings. I am

Methodological Developments in Ethnography

A TEST OF THE ATTRIBUTES LIST

I cannot subject listeners to such a long list. At my age I cannot always

I am all for using ethnographic techniques in learning about young peo-

Methodological Developments in Ethnography

2005; Hodkinson, 2004 among others), but regardless of your epistemolog-

Consequently, to understand how elements of the workplace provide affor-

understand how learning is experienced and to identify the processes

THE DECISION-MAKING PROCESS

signiﬁcant but no less important, changes in capability or understanding in

THE PILOT STUDY

observing their work, and gave informed consent to be involved in the

time-span ethnography provides a snapshot of what is happening, it does

Table 1. Key Informants.

The Hutch Nat – 1st year apprentice

during my ﬁrst week at Gastronomique. The other chefs laughed at the

PARTICIPANT OBSERVATION: THE EXPERIENCE

I spent 1 month in each site. By my calculations, this amount of time would

In a kitchen, swear words seem to be used as nouns, verbs, adjectives and